Choose timezone
Your profile timezone:
28th IAEA Fusion Energy Conference (FEC 2020)
→
Europe/Vienna
Virtual Event
Virtual Event
Contact
-
-
-
The 2019-2020 scientific and technological programme exploits JET’s currently unique capabilities: Tritium handling and ITER-like wall (ILW: Be wall and W divertor). It is the culmination of years of concerted scientific and engineering work, with the ILW installation in 2010, improved diagnostic capabilities, now fully available, a major Neutral Beam Injection (NBI) upgrade providing record power in 2019 (P$_{NBI}$ up to 32MW), and the technical & procedural preparation for safe operation with tritium. Research along three complementary axes yielded a wealth of new results since last IAEA. Firstly, the JET plasma programme delivered scenarios suitable for high fusion power (P$_{FUS}$) and alpha particle ($\alpha$) physics in the coming D-T campaign (DTE2), with record sustained neutron rates, as well as plasmas for clarifying the impact of isotope mass on plasma core, edge and plasma-wall interactions, and for ITER pre-fusion power operation (e.g. L-H transition in He plasmas). The efficacy of the newly installed Shattered Pellet Injector (SPI) [ref1] for mitigating disruption forces and runaway electrons was demonstrated, informing ITER disruption management. Secondly, research on the consequences of long-term exposure to plasma in the ILW was completed, with emphasis on wall damage and fuel retention, and including analyses of wall materials and dust particles. This will help validate assumptions and codes for the design & operation of ITER and DEMO. Thirdly, the nuclear technology programme aiming to deliver the maximum technological return from operations with D, T and D-T [ref3] benefited from the high D-D neutron yield in 2019 (2.26x10$^{19}$n), securing new results for validating radiation transport and activation simulation codes, and nuclear data for ITER. Measuring systems are ready for collecting data in T and in D-T campaigns producing 14MeV neutrons.
Integrated scenarios preparation for high P$_{FUS}$ sustained for 5s (i.e. relevant to energy confinement times in JET) progressed significantly for the two routes investigated: ‘Baseline’ (q$_{95}$~3, I$_P$≥3MA, $\beta$$_N$<2) and ‘Hybrid’ (tailored q-profile, q$_{95}$~4.5, I$_P$ ≤2.7MA, $\beta$$_N$≥2.4). Peak neutron rate of 4.2x10$^{16}$n/s (2.7x10$^{16}$n/s averaged over 5s) are obtained simultaneously with tolerable divertor temperatures and controlled high/medium Z impurity for the full pulse duration in Baseline plasmas at 3.3T/3.5MA, with P$_{TOT}$=34MW (NBI and Ion Cyclotron Resonance Heating (ICRH)). Pellets help controlling the ELM frequency (f$_{ELM}$) needed for impurity flushing, with low total D$_2$ throughput for high confinement. Hybrid plasmas developed to 3.4T/2.3MA reached 4.8x10$^{16}$n/s but MHD avoidance and f$_{ELM}$ control must be optimised for improved, steady performance. The equivalent P$_{D-T}$ for these pulses is consistent with past predictions at same B$_T$, I$_P$, P$_{TOT}$ [ref4], giving confidence in the theory-based modelling. Further gains are likely with 40MW now reachable and higher I$_P$, with divertor heat loads controlled by strike-point sweeping, thus prospects for reaching the target (5x10$^{16}$n/s) are good. In these conditions P$_{D-T}$=11-16MW is predicted by theory-based physics models, with range due to uncertainties in the pedestal predictions and to whether isotope effects are included or not. Fast particle diagnostics, significantly improved since DTE1, can now detect small amount of $\alpha$’s, as shown in dedicated experiments making use of 3-ion ICRH scheme (D-(D$_{NBI}$)-3He) to create MeV range particles, with $\alpha$ (≈10$^{16}$s$^{-1}$) from D+$^3$He reactions. Simultaneous detection of He and hydrogen isotopes with an enhanced high resolution sub-divertor residual gas analyser, as planned for ITER, has been demonstrated.
Experiments and modelling preparing the T campaign. Observations that the impact of isotope on H-mode plasmas comes mainly from the pedestal [ref5] motivated recent gyrokinetic (GK) theoretical investigations of JET pedestals showing that the toroidal branch of the ETG instability can be driven at ion-scale poloidal wavelengths and may be responsible for significant inter-ELM pedestal heat transport. However, in some regimes, isotope effects on core plasma may also be important [ref4]. New experiments in D$_2$ and H$_2$ L-mode plasmas and related core GK modelling show that, in plasmas with a strong stabilizing effect of fast particles, differences in fast particle content with isotope mass may lead to strong deviations from the gyro-Bohm scaling of core transport. Recently developed ICRH-only H-mode plasmas (low input torque, dominant e-heating) show the same normalised confinement factor (H$_{98(y,2)}$) and T$_e$ profiles as their NBI-only counterpart at same PTOT, but n$_e$ profile for the NBI case is 50% higher due to NBI fuelling and possibly different particle transport. Work to clarify the impact of edge/divertor was performed. Experiments at 2MA/2.3T, low triangularity ($\delta$) demonstrated that changes in H$_{98(y,2)}$ and pedestal T$_e$ due to divertor configuration can be condensed into a single trend when mapped to the target T$_e$ as the main parameter governing recycling conditions rather than D$_2$ fuelling rate. A high performance neon seeded scenario (2.7T/2.5MA, high-δ shape) with edge conditions closer to ITER was developed. Neon seeding leads to significant increase (by ≈50%) in pedestal pressure and T$_e$ (from 0.4keV to 0.8keV) and in H$_{98(y,2)}$ (from 0.6 to 0.9) with mitigated divertor power loads. The well diagnosed discharges are used for validating physics-based SOL-edge modelling, increasing confidence in ITER divertor design basis and supporting deployment of neon over chemically reactive N$_2$ as seed gas in ITER.
JET disruption management programme is in two parts: 1) disruption avoidance based on improved termination techniques and on real-time detection of unhealthy plasmas with jump to controlled termination, causing significant reduction in disruption rate in baseline (60% to 20%) and hybrid plasmas (9%), and 2) disruption mitigation with SPI performed as part of a collaboration between ITER, US and Europe. After successful installation and commissioning, extensive experiments took place with the JET SPI demonstrating very good reliability. By varying the neon content in the SPI pellets, the disruption current quench time can be controlled efficiently in JET, scaling to the range required by ITER. High Z impurity SPI also demonstrated run-away electron suppression. Additionally, it was discovered that D$_2$ SPI applied to a high current run-away beam leads to benign impacts on the wall, suggesting a new potential solution for run-away electron control in ITER.
Plasma-facing components (PFC) long term exposure in ILW. Retrieval of PFC, wall probes (including test mirrors) and dust for ex-situ studies after three ILW campaigns provide deep insight into material erosion and deposition. Low mobilization of dust during in-vessel operations is shown. Emphasis was placed on material damage such as melting of the Be upper dump plates (UPD) and the identification of factors triggering this process. Comprehensive studies including imaging survey, morphology changes, mass loss and fuel inventory analysis on the most affected UDP tiles was performed. The undisputed reason for melting was unmitigated disruption events which tend to move the melt layers in the poloidal direction resulting in formation of upwards going waterfall-like structures of molten metal. The halo current is believed to provide the j x B force driving the melt layer motion. Global material migration results constitute a unique dataset for modelling and thus improved predictions for ITER.
[ref1] L. R. Baylor, et al., Nucl. Fusion 68 (2019) 211, [ref2] I. Jepu et al., Nucl. Fusion 59 (2019) 086009, [ref3] P. Batistoni et al, Fusion Engineering and Design Vol 109-11, 2016, [ref4] J. Garcia et al. Nucl. Fusion 59 (2019) 086047 [ref5] C. F. Maggi et al., Plasma Phys Control Fusion 60 (2018) 014045
Speaker: Dr Joelle Mailloux (UKAEA) -
DIII-D physics research addresses critical challenges for operation of ITER and the next generation of fusion energy devices through a focus on innovations to provide solutions for high performance long pulse operation, development of scenarios integrating high performance core and boundary plasmas, and fundamental plasma science and model validation. Substantial increases in off-axis current drive efficiency from an innovative top launch system for EC power (Fig. 1), and in pressure broadening for Alfven eigenmode control from a co-/counter-Ip steerable off-axis neutral beam, both improve the prospects for optimization of future long pulse/steady state high performance tokamak operation. A high beta-p optimized-core scenario with an internal transport barrier that projects nearly to Q=10 in ITER at 9 MA was coupled to a detached divertor, and a Super H-mode optimized-pedestal scenario with co-Ip beam injection (Fig. 2), was coupled to a radiative divertor. Fundamental studies into the evolution of the pedestal pressure profile, and electron vs. ion heat flux, measuring both density and magnetic field fluctuations, validate predictive models of pedestal recovery after ELMs (Fig. 3).
Link to High Resolution Figures 1, 2, 3
The achievement of more than double the off-axis ECCD efficiency using top launch geometry compared with conventional low field side (LFS) launch, as predicted by quasi-linear Fokker-Planck simulations, is due to the longer absorption path for the EC waves which also interact with higher v|| electrons that suffer fewer trapping effects than outside launch. In addition, the new unique co-/counter-Ip steerable off-axis neutral beam broadens the energetic particle (EP) pressure profile and reduces Alfven eigenmode (AE) drive in scenarios with both high toroidal rotation and those with net zero average input torque. New EP measurements show a beam current threshold for Compressional AEs, insensitivity of Beta-induced Acoustic AEs to fast beam ions, and resolution of phase-space flows caused by AEs, from first-of-a-kind Ion Cyclotron Emission (ICE) and Imaging Neutral Particle Analyzer (INPA) data.
Studies of high current runaway electron (RE) beams reveals excitation of current-driven (low safety factor) kink instabilities that promptly terminate the RE beam on an Alfvenic time-scale, offering an unexpected alternate pathway to RE beam mitigation without collisional dissipation. Newly developed real-time stability boundary proximity control and neural-net-based Vertical Displacement Event (VDE) growth-rate calculations are shown to prevent VDEs. The effectiveness of emergency shutdown and disruption prevention tools projects to at least 50% of ITER disruptions being delayed until normalized-Ip is at safe levels, and demonstration of a novel technique for healing flux surface with 3D fields shows promise for providing current quench (CQ) control. Single and multiple Shattered Pellet Injection particle assimilation rates and current quench (CQ) densities are shown to be predictable from 0-D simulations and empirical scaling laws.
Several core-edge integration scenarios demonstrate coupling of a high performance core and radiative divertor operation for target heat flux control. High density and stored energy plasmas with Super H-mode edge pedestals were made both in a lower single null shape accessible by JET and in a higher triangularity near double null shape coupled to a radiative divertor for target heat flux control using nitrogen injection in a core-edge integrated scenario. High-performance plasma with high poloidal beta, large Shafranov shift, and Te and Ti internal transport barriers coupled to a detached divertor with active feedback-controlled Nitrogen puffing also demonstrated integration of core-edge solutions. A high performance hybrid core demonstrated compatibility with radiative divertor operation using Neon or Argon gas injection. Core impurity peaking in the hybrid was substantially reduced using near-axis electron cyclotron heating.
The ability to predict the impurity seeding needed for divertor dissipation has advanced through new capability for measuring charge-state resolved densities of impurity species in the divertor. Also electric drifts in detached divertors with convection dominated heat transport lead to expanded radiative volume. Using these advances, SOLPS-ITER simulations show the synergy between SOL drifts and the SAS divertor geometry for achieving lower density detachment. Modeling of intra-ELM tungsten gross erosion with an analytic Free-Streaming plus Recycling Model is now validated in ITER-relevant mitigated-ELM regimes using pellet pacing and RMPs. SOL tungsten transport in plasmas with both BT directions is consistent with strong entrainment in SOL flows and ExB drift effects.
Advances in pedestal physics through new measurements of density and internal magnetic fluctuations suggest a possible role for micro-tearing and trapped electron modes in DIII-D pedestal transport. Main ion CER measurements indicate ion heat flux is anomalous at low collisionality and transitions to near neoclassical levels at high collisionality. Plasma rotation scans, and both new non-linear analytic theory and 2-fluid code simulations, confirm that ELM suppression by RMPs requires near zero ExB velocity at the top of the pedestal, and achieving suppression appears to be closely linked to a high field side plasma response. The wide pedestal QH-mode regime was obtained with zero input beam torque, electron heating, and LSN shape, consistent with requirements for ITER.
Recent fundamental research on L-H mode power threshold physics shows that turbulence driven shear flow through Reynolds stress and the coexistence of modes associated with various instabilities can lower the L-H power threshold across multiple parameters: eg. q95 and ion grad-B drift direction. Application of RMPs raises turbulence decorrelation rates and reduces Reynolds stress driven flow and flow shear, hence increasing the L-H power threshold. Finally, plasmas with negative triangularity show weak power degradation of H-mode level core confinement while maintaining an L-mode-like edge without ELMs.
In 2020 and beyond DIII-D will install additional tools for optimizing tokamak operation through current and heating profile control using a low field side 1 MW helicon high harmonic fast wave CD system, a unique high field side Lower Hybrid CD system, increased ECH power, and coupling to boundary advances using a new high power closed divertor and a wall insertion test station. Experiments will continue the optimized coupling of high performance core and high power density divertor solutions.
This work was supported in part by the US DOE under contracts DE-FC02-04ER54698 and DE-AC52-07NA27344Speaker: Max Fenstermacher (LLNL @ DIII-D) -
Since the last IAEA-FEC, the EAST research programme has been, in support of ITER and CFETR, focused on development of the long-pulse steady-state (fully non-inductive) high beta H-mode scenario with active control of stationary and transient divertor heat and particle fluxes $^{[1]}$. The operational domain of the steady-state H-mode plasma scenario on EAST has been significantly extended with the ITER-like configuration, plasma control and heating schemes. Several important milestones on the developments of the high beta H-mode scenario and its related key physics and technologies have been achieved.
-
A minute time scale long-pulse steady-state high beta H-mode discharge (shown in figure 1) with the major normalized plasma parameters similar to the design of the CFETR Phase-III 1GW fusion power operation scenario ($\beta_P$ ~ 2.0, $\beta_N$ ~ 1.6, $f_{bs}$ ~ 50%, $H_{98(y2)}$ > 1.3 at $q_{95}$ = 6.5~7.5) has been successfully established and sustained by pure RF heating on EAST with the ITER-like tungsten divertor $^{[2]}$, as shown in figure 1. The important feature of this high beta H-mode plasma scenario is that, due to the stabilization effect of the Shafranov shift on the plasma turbulence, a higher $\beta_P$ results in a better plasma confinement as shown in figure 2. Further simulation suggested that a high density gradient promotes the ITB formation in high $\beta_P$ plasmas, which might further benefit the development of this high beta plasma scenario towards a high density regime.
-
Active control of divertor radiation has been successfully integrated into the high beta H-mode ($\beta_P$ ~ 2.5, $\beta_N$ ~ 2.0, $f_{bs}$ ~ 50%) plasma scenario without a degradation of the plasma confinement ($H_{98(y2)}$ > 1.2) at high density ($n_e$/$n_{GW}$ ~ 0.7) and moderate edge safety factor ($q_{95}$ ~ 6.7) $^{[2]}$. The peak heat flux on the tungsten divertor was reduced by ~30% with active impurity seeding of a mixture of 50% neon and 50% deuterium. The high-Z impurity concentration in the plasma core has been well controlled in a low level by applying the on-axis ECRH and reducing the fast ion losses through beam energy optimization.
-
The grassy-ELM regime has been extended to the normalized parameter space designed for the CFETR 1GW fusion power operation scenario. This regime exhibits good compatibility with high $f_{bs}$ and fully non-inductive operation, being characterized by a low pedestal density gradient and a wide pedestal, which prevents large-ELM crashes due to an expansion of the peeling-ballooning boundary after the initial pedestal collapse, indicated by BOUT++ nonlinear simulations. High separatrix density makes this regime especially suitable for operation with divertor detachment. Several feedback control schemes have been developed to achieve sustained detachment with good core confinement $^{[3]}$. This includes control of total radiation power, target electron temperature, and particle flux measured by divertor Langmuir probes or a combination of the control of target electron temperature and AXUV radiation near the X point. Integration of these detachment feedback control schemes with the grassy-ELM regimes and the high $\beta_P$ scenario has been demonstrated with neon seeding, which provides an integrated high beta scenario applicable to long-pulse operation $^{[4]}$.
-
ELM suppression has been achieved using various different methods on EAST. Full suppression of ELMs has been demonstrated, for the first time, for ITER-like low torque injection plasmas by using n=4 resonant magnetic perturbations (RMPs) $^{[5]}$. A moderate reduction (~5%) of the energy confinement has been observed together with significant reduction of both the plasma density(20%)and the Tungsten concentration (a factor of 2) during ELM suppression. The ELM suppression window agrees well with the prediction by MARS-F modelling. Robust ELM suppression by Boron powder injection without confinement degradation or even with confinement improvement has been achieved in a wide parameter range. EHO-like edge coherent modes were excited during the ELM suppression phase by Boron powder injection. In addition, simulation results from BOUT++ confirm that both the helical current filaments (HCF) driven by lower-hybrid waves (LHWs) and RF sheath effects on the ICRF antenna contribute to ELM suppression.
-
A flowing liquid lithium (FLiLi) limiter plate has been successfully assembled and tested in EAST for the first time. This plate was designed based on the concept of liquid metal infused trenches (LiMIT), which is using thermoelectric MHD to drive liquid Li flow along the surface channels. The preliminary results show that with the increase of Li flow rate, the fuel particle recycling is gradually reduced, and the plasma performance is slightly improved. There was no obvious Li burst and limiter damage even at a high injection power of 5.5 MW. In addition, ELM mitigation was observed with FLiLi operation.
-
For the first time, EAST has been operated with helium to support the ITER needs. The H-mode power threshold in a helium plasma is found to be 1.2-2.2 times higher than the scaling law in deuterium plasma with pure RF-heating $^{[6]}$. The $C_{He}$ plays a crucial role in determining the energy confinement and pedestal characteristics in helium H-mode plasmas. Divertor detachment is more difficult to achieve in He than in D. The control of divertor heat loads and W sources is achieved by RMP along with neon impurity seeding. The inter-ELM W erosion rate in He is about 3 times that in D with similar divertor conditions, while the intra-ELM W sputtering source shows a strong positive correlation with the ELM frequency $^{[7]}$.
-
A new lower tungsten divertor with a higher closure has been designed and to be installed on EAST in 2020. Several key subsystems, including the heating & current drive, cryogenic, plasma control and diagnostics will be upgraded accordingly for achieving the next milestones i) >400s long-pulse H-mode operation with ~50% bootstrap current fraction, and ii) demonstration of power exhaust at ~10 MW power injection for >100s.
Reference:
1) B. N. Wan et al 2019 Nucl. Fusion 59 112003
2) X. Z. Gong et al 2020 this conference
3) L. Wang et al 2020 this conference
4) G. S. Xu et al 2020 this conference
5) Y. W. Sun et al 2020 this conference
6) B. Zhang et al 2020 this conference
7) R. Ding et al 2020 this conference
Speaker: Baonian Wan (Institute of Plasma Physics, Chinese Academy of Sciences) -
-
-
-
The ADITYA/ADITYA-U tokamaks are equipped with state-of-art spectroscopic diagnostics in the visible and vacuum ultraviolet (VUV) region of the spectra. These spectroscopic systems are used to study several physics problems in ADITYA tokamak as well as in ADITYA-U, which is an upgraded version of ADITYA, having capability of producing shaped plasmas. The physics studies addressed in this paper are studies of impurity behaviour, dynamics of neutrals, measurement of plasma rotation, a novel technique to estimate electron temperature in the edge region by simulating the temporal profile of H-alpha intensity observed at the time of hydrogen gas puff, estimation of particle confinement and ion temperature etc.
The ADITYA/ADITYA-U tokamak plasma is diagnosed mainly by detecting the radiation emanating from the plasma using spectrometers, filter-PMT combination and fast imaging cameras. Several spectroscopic diagnostics were developed for ADITYA, which was having a poloidal ring limiter, and its present upgraded version ADITYA-U tokamak with toroidal belt limiter. The line averaged ne and core Te of the plasmas are in the range of 1 - 6.0 x 1019 m-3 and 300-700 eV, respectively. The toroidal magnetic field was ~ 0.75 to 1.4 T in these studies.
The fast visible imaging camera was attached to tokamak using an imaging fiber bundle on a re-entrant view port to acquire images of plasma evolution at ~ 14 kHz frame rate. This system enabled to observe the thick toroidal filaments (figure 1a) during the disruptive phase of ADITYA tokamak plasma and those were explained in term of interchange modes 1. Further the boundary of the plasma was also detected by the imaging camera and was used to control plasma position. Furthermore, the fast visible imaging was used to study the plasma collapse during micron-sized particle injection through an electromagnetically driven payload for disruption mitigation studies. Photo multiplier tube (PMT) based system, which is normally used to monitor temporal evolution of spectral lines, such as H, from OII and CIII and visible continuum emission, was extensively employed to study the hydrogen recycling, impurities behaviour and Zeff [2]. It was found that the Zeff were reduced 1.5 to 3.0 for the analyzed discharges after the Li coating of plasma facing components of ADITYA tokamak compared to the Zeff values of 2.0 - 4.5 for the discharges before the Li coating [3]. The radial profiles of H measured using 8 channels PMT array based diagnostic were modelled using DEGAS2 neutral transport code. The details studies on the contribution from various atomic and molecular processes show that substantial amount of hydrogen molecules survive inside the plasma up to 4 cm from the limiter of ADITYA tokamak having typical edge Te of 10- 15 eV [4]. A PMT based system is recently developed to monitor the radial profile of visible continuum for the Zeff profile estimation of ADITYA-U tokamak plasma.
An indigenously developed impurity transport code based on semi–implicit numerical scheme was used for modelling of the radial profile of O4+ emissions at 650 nm from ADITYA plasma and the studies has revealed the higher oxygen diffusion coefficient at plasma edge than the neoclassical values as shown in figure 1b [5]. The O4+ emission profile were recorded using a high resolution visible spectroscopy system having 1 m focal length visible spectrometer and charge coupled device (CCD) camera, which capable of multi-track measurement simultaneously from eight lines of sight. The radial profiles of impurity toroidal rotation velocities from ADITYA-U plasmas were also obtained by measuring Doppler shifted passive charge exchange emission at 529 nm from C5+ ion. The lights were collected using a front-end optic placed inside the re-entrant tube from a tangential port of the tokamak. The data were converted using an Abel-like matrix inversion to get radial profile of impurity rotation and ion temperature. The maximum C5+ rotation velocity and ion temperature values were ~ 15 km/s and ~ 120 eV, respectively, for the analyzed discharges and the reversal of toroidal rotation was observed for the discharges having electron densities 3x1019 m-3 [6] as depicted in figure 1c. The spatial profile of temperature of neutral hydrogen and low ionized impurities ions of ADITYA-U plasmas were also determined from the Doppler broadened H and spectral lines from lowly ionized carbon and oxygen impurities after incorporating the influence of Zeeman Effect into the apparent broadening of spectral lines. It is found that the neutral temperature (TH) is having two components, warm and hot and their values vary in the range of 3 - 5 eV (shown in figure 1d) and 18 - 25 eV, respectively [7].
A toroidal grating based VUV survey spectroscopy system working in the 10 - 180 nm range regularly monitored the survey spectrum from low and highly ionized impurities in the plasmas enabling to do the details studies of impurity transport in the central region of the plasma. The measured VUV spectral lines from highly ionized irons, such as Fe14+ and Fe15+, were analyzed to study the ion impurity behaviour in high density ADITYA tokamak plasmas and its concentration was found to be lower than 0.1% [8]. For the first time, ADITYA-U has experimentally demonstrated the use of electromagnetic pellet injection for firing the impurity pellets into the tokamak plasmas for disruption mitigation studies. Lithium titanate (Li2TiO3) and Lithium carbonate (Li2CO3) particles of 50 – 80 micron particles are injected in to the plasma. The spectral lines of Li showed a self-absorption peak on top of the emission spectra during the injection, which has been diagnosed thoroughly to understand the disruption mechanisms.
References:
1. Banerjee S. et al. Phys. Plasmas 24, (2017) 102513.
2. Chowdhuri, M. B., et al. Plasma Sci. Technol. 15 (2013) 123.
3. Chowdhuri, M. B., et al. Plasma Phys. Control. Fusion 62 (2020) 035015.
4. Dey, R., et al. Nucl. Fusion 59 (2019) 076005.
5. A Bhattacharya, Phys. Plasma 27, ( 2020) 023303
6. Shukla, G., et al. Nucl. Fusion 59 (2019) 106049.
7. Yadava, N., et al. Atoms 7 (2019) 87.
8. Patel, S., et al. Nucl. Fusion 59 (2019) 086019.Speaker: RANJANA MANCHANDA (INSTITUTE FOR PLASMA RESEARCH) -
The EUROfusion Work Package PFC (Plasma-Facing Components) focuses on critical plasma-surface interaction studies and components qualification in view of upcoming ITER operation and in preparation for DEMO exhaust solutions. This poster gives an overview of the latest main results in WP PFC, as well as their implications for ITER and DEMO.
Helium-Tungsten Interaction
The Helium-Tungsten (He-W) interaction was studied extensively in the full W tokamak WEST (CEA), in dedicated experiments in the divertor manipulator of ASDEX Upgrade with He pre-damaged W (MPG), in dedicated high fluence experiments on ITER monoblocks in the MAGNUM-PSI device (DIFFER), and compared with other laboratory studies and first principle modelling. The goal is the predictability towards He-W interaction in the different phases of ITER (He plasmas and DT plasmas with He ash). Figure 1 shows an example of the formation of W fuzz in samples exposed to ASDEX Upgrade H-mode plasmas. The results, showing pronounced evolution of the surface morphology and the creation of He nanobubbles below the surface, will be used to benchmark models relevant for lifetime estimations of the first ITER divertor. This work related to He-W interaction is done jointly with the ITER Organisation.
Fig.1) Formation of W fuzz in AUG [Brezinsek et al.]Tungsten surface morphology and erosion
The erosion yield of the W fuzz has not only been studied under tokamak conditions, but also in laboratory arrangements with monoenergetic ion beam bombardment. These experiments provide the sputtering yield of 3D structured surfaces, as well as W fuzz, as a function of impact energy and impact angle. Complementary simulations of the surface evolution of e.g. W fuzz was performed with the TRI3DYN code. The reference binary-collision approximation code 3DTRIMSP has been benchmarked in 2018 against ion beam experiments and 3D structures and provides the input basis for larger scale modelling with ERO2.0 and WallDYN 3D. Overall, the combination of these laboratory and tokamak experiments will be used to benchmark also global migration codes with novel surface roughness modules (e.g. in ERO2.0).
Fig. 2) Preferential sputtering direction at rough surfaces [Eksaeva et al.]Fig. 2 shows an example of the so-called needle formation under perpendicular impact on low-Z surfaces (Be), modelled with the roughness surface module implemented in ERO2.0. This example is an outcome of the long-standing PISCES-B cooperation between WP PFC and UCSD with regular guest scientists from EU send to the US to perform studies related to Be.
Finally, 3D modelling utilising the ERO2.0 code without the surface roughness module has been applied to benchmark JET (Be, W) and WEST (W) experiments as well as to predict the Be migration in ITER. A typical example from the WEST application is shown in fig. 3 providing the experimental plasma information, the plasma background as well as the corresponding ERO2.0 run. In the future, dedicated runs with the surface roughness of technical surfaces and appropriate material mixing will be done for the different devices in support of predictions for ITER and DEMO.
Fig. 3) ERO2.0 modelling of W erosion in WEST [Gallo et al.]Qualification of ITER-like monoblocks
Tests of ITER-like monoblocks at elevated temperature were performed in high fluence experiments in MAGNUM-PSI, reaching record values of D+ fluence of 1031m-2, comparable to a year of ITER divertor operation without any visible damage and extremely low fuel retention. Re-crystallisation of W occurred in the high fluence experiments in D at the theoretically predicted material temperatures, which gives confidence in the lifetime predictions of the ITER divertor. Combined exposure of He and D ions in PSI-2 showed no impact on the recrystallization temperature, but on changes in the morphology. Additional emphasis was put on the study of monoblock castellation and shaping. ITER monoblocks with different shapings were installed in WEST for this purpose. The local damage of optical hot spots, which are predicted to occur in ITER at the projection of toroidal gaps, has been observed. Dedicated PIC modelling was applied and identified the particle and power flux to these locations. The same modelling was used for ITER predictions about castellation and shaping.
Parallel to these studies, ELM simulations by combined laser and plasma exposition and e-beam impact were carried out up to the record number of 106 ELM-like pulses, to study the response of W to repetitive transient heat loads. All three W PFC studies contribute to the determination of the operational window for the W divertor in ITER regarding heat loads, transients and surface temperature conditions.
Complementary, studies with plasma and high heat flux exposure on advanced materials such as PIM-W, WfW etc. - developed under WP MAT- started in WP PFC in 2019. Multiple disruption-like loads on W were carried out in the plasma gun facility QSPA in KIPT. The surface damage evolution was compared to previous studies on the ITER-grade reference W regarding synergistic effects with combined plasma and heat load.
Fuel retention
Finally, dedicated experiments were carried out to quantify the fuel retention in W for combined D and N2 exposure in laboratory and linear plasma conditions. A dramatic enhancement of the retention in the near-surface area has been identified, with local fuel content in the percentage range depending on the exposition conditions. WN formation has been observed near the surface, acting in some cases as permeation barrier. Thus, the actual enhancement in fuel retention is solely near the surface, which can interact during plasma operation with the recycling and fuelling of tokamak plasmas. The second system investigated regarding the fuel retention by implantation is the mixed He-D system in W in combination with neutron damage proxies. Self-damaging by W ions was applied to simulate defects in W and mimic the impact of fusion neutrons. Enhanced retention by the synergistic effects has been documented. Detailed modelling was applied to simulate the observed enhancement of retention in the near surface. The fuel release efficiency was documented by stepwise heating of the W material. Fuel recovery at high temperatures is challenging if He impact and neutron damage are present simultaneously.
In order to quantify in-situ or in-operando the retention of T and D in the first wall material of fusion devices, a dedicated task utilising LIBS and LIA-QMS was executed successfully in 2019 (see contribution of H. van der Meiden et al., “LIBS - monitoring of tritium and impurities in the first wall of fusion devices”). The experiments demonstrated the capability to resolve D and T as well as quantify the fuel in different types of ITER-like codeposits containing in particular Be and W as well as impurities as found in JET samples.
Bibliography
S. Brezinsek et al., PSI 2020, Jeju, KoreaA Eksaeva et al. PSI 2020, Jeju, Korea
A Gallo et al., 2020 Phys. Scr.2020 014013
Speaker: Michael Reinhart (EUROfusion) -
Achieving net energy production in magnetic confinement fusion devices is a key milestone in the quest for fusion energy. With the mission of demonstrating net fusion energy, the SPARC tokamak is being designed jointly by the MIT Plasma Science and Fusion Center and Commonwealth Fusion Systems. Its study of reactor-relevant, alpha-heating-dominated scenarios and high power density regimes will help retire risk for ITER operations and for fusion power plants. A team of over 100 engineers and scientists is on track to deliver a toroidal field model coil using high-temperature superconductor (HTS) technology by 2021, with the engineering design of the tokamak progressing in parallel. Negotiations with potential host sites in the Northeast US are underway, with start of construction planned in 2021 and operation expected in 2025.
SPARC will be a pulsed machine operating with Deuterium-Tritium (DT) fuel and with ICRF auxiliary heating. The high strength of the magnetic field ($B_T>12.0T$ on axis), will allow operation at high plasma current and high absolute density, leading to net fusion output in a device with a size comparable to current tokamaks ($R_0<2.0m$). In particular, the SPARC mission objective has been established as demonstration and study of Q>2 plasma conditions, where Q is the ratio between the total fusion power and the external power absorbed in the plasma. Figure 1 depicts the poloidal cross-section of the Version 1C (SPARC V1C) design iteration, and Figure 2 indicates main plasma parameters for the baseline DT H-mode plasma discharge.
Following a traditional design workflow (1), SPARC parameters are first selected using empirical scaling laws and plasma operation contour (POPCON) analysis. Figure 3 represents the operational space for SPARC V1C for its baseline scenario, demonstrating that $Q\approx11$ can be reached with conservative assumptions ($H_{98}=1.0$ confinement, and $\nu_{Ti}=2.5$, $\nu_{ne}=1.3$ profile peaking factors, consistent with empirical predictions). Total fusion power remains below administrative limits for the machine ($P_{fus}<140MW$), and safety factor ($q^*=3.05$), normalized density ($f_G=0.37$) and normalized pressure ($\beta_N=1.05$) are at reasonably safe levels of operation.
The development and validation of theory-based reduced models allow integrated simulations to also inform the design of SPARC. To this end, simulations with the TRANSP code (2) coupled with the TGLF model (3) for turbulence and EPED (4) for pedestal stability are performed. Figure 4 depicts simulated temperature and density profiles. $H_{98}\approx1.0$ is predicted, and fusion gain results in $Q\approx8.2$. The good agreement between the two independent workflows (empirical and theory-based simulations) provides high confidence that SPARC will accomplish its $Q>2$ mission. During nominal operation, D-T(3He) ICRF minority heating at 120 MHz will be utilized for on axis heating of both 3He and T. AORSA and CQL3D (5) simulations are in good agreement with TRANSP, which uses TORIC (6) to model ICRF. Single-pass absorption is excellent, and minimum losses of ICRF power (~1%) to alphas are predicted.
Loss of fast ions (alphas and RF-tail ions) due to toroidal field (TF) ripple can be a major issue for the design of DT tokamaks, as it can lead to excessive localized wall heating. The effect of first-orbit, classical and TF ripple in SPARC has been studied with ASCOT (7). Simulations indicate that the total losses are small (<2%), due to the low edge TF ripple (0.15%) in the SPARC design. There is no concentration of losses toroidally and only modest concentration poloidally in the current TF design.
Managing divertor heat flux will be challenging in SPARC, but unmitigated levels are comparable to ITER. To ensure divertor survivability with conservative assumptions (i.e., not relying on partial or complete divertor detachment), the poloidal field coil set and central solenoid are being designed to ensure that a fast strike point sweep (~1 Hz) can be achieved. Thermal simulations of the divertor target indicate that sweeping during the flat-top is sufficient to ensure divertor survivability with only a moderate divertor radiation fraction. SPARC will be equipped with impurity gas injection to attain detached-divertor scenarios. The feasibility of an “advanced” divertor is also being assessed.
In summary, SPARC will be an important experiment to study burning plasma physics and will be a proof-of-principle for high-field, compact fusion power plants. The SPARC design is converging towards a self-consistent model of the machine with robust engineering and physics. Conservative estimates of fusion gain show significant margin for the Q>2 mission, leaving room for extensive exploration of burning plasma physics regimes.
This work was supported by Commonwealth Fusion Systems.
(1) ITER Physics Expert Group on Confinement and Transport et al. Nucl. Fusion 39, 2175 (1999).
(2) J. Breslau et al. USDOE SC-FES (2018).
(3) G.M. Staebler et al. Phys. Plasmas 14, 055909 (2007).
(4) P.B. Snyder et al. Phys. Plasmas 16, 056118 (2009).
(5) E. F. Jaeger et al. Phys. Plasmas 13, 056101 (2006).
(6) M. Brambilla et al., Plasma Phys. Con. Fus. 41, 1 (1999).
(7) A. Snicker et al. Nucl. Fusion 52, 094011 (2012).Speaker: P. Rodriguez Fernandez (MIT Plasma Science and Fusion Center) -
Inertial confinement fusion (ICF) aims to assemble and confine a dense, high pressure fusion fuel over a relatively short timescale (≪1 μs) compared to magnetic confinement fusion (> 1 s). This is typically accomplished by imploding a spherical capsule at high implosion velocities (>350 km/s) to obtain the fuel temperatures (>4 keV) and areal densities (ρR >0.3 g/cm2) required for ignition.1 Magneto-inertial fusion (MIF) utilizes magnetic fields that relax these requirements by limiting thermal conduction losses and introducing magnetic confinement of charged fusion products. On the Z Machine at Sandia National Laboratories, we are pursuing a specific pulsed-power2 driven MIF concept called Magnetized Liner Inertial Fusion (MagLIF).3 MagLIF is the first MIF concept to demonstrate fusion-relevant temperatures, significant neutron production, and magnetic trapping of charged fusion products4,5, and has the potential to generate multi-MJ yields and significant fuel self-heating on a next-generation pulsed power machine.6
In MagLIF, a centimeter-scale cylindrical tube, or “liner,” is filled with a fusion fuel (typically deuterium gas), pre-magnetized using an axial magnetic field of 10-20 T using Helmholtz coils, pre-heated to an average temperature of 100-200 eV via a kilojoule-class laser, and finally radially imploded over ~100 ns via the Lorentz force to velocities of ~70 km/s using 15-20 megaamperes of current from the Z Machine. This process is schematically demonstrated in Figure 1. The laser preheat increases the initial adiabat of the fuel, which is then compressed in a quasi-adiabatic implosion to reach fusion-relevant conditions. The axial magnetic field, which flux-compresses to >1000 T near peak convergence, limits thermal conduction losses from the hot fusion fuel to the comparatively cold liner walls during the implosion and simultaneously increases trapping of charged fusion particles in the narrow radial direction during stagnation.
The first MagLIF experiments demonstrated that the axial magnetic field, the laser preheat, and the implosion are all required to generate thermonuclear fusion yields. Without any of these inputs, no significant yield could be generated. Utilizing pure deuterium fuel, these experiments produced multi-keV fuel temperatures, neutron energy spectra consistent with a thermonuclear plasma, and produced yields up to 2x1012 neutrons—a DT equivalent energy of 0.3 kJ.4 Analysis of the secondary deuterium-tritium (DT) neutrons demonstrated the fuel column was highly magnetized, with the fuel radius exceeding the average Larmor radius of DD-produced fast tritons, a promising and necessary requirement for trapping of alpha particles.5,7
Subsequently, our efforts have focused on improving the stability of liner implosions, increasing the fuel preheat energies, and simultaneously increasing the applied magnetic field and current delivered. Controlling the magneto Rayleigh-Taylor (MRT) instability is required to achieve uniform compression and minimize areal density variations in the liner. Simulations suggested that applying a dielectric coating to our beryllium liners would mitigate the early-time electrothermal instability,8 which is believed to seed the more deleterious MRT instability. Experiments verified this stabilizing effect9 and have produced highly uniform stagnation columns compared to uncoated targets, as shown in Figure 2.
To increase the laser preheat, we developed a new laser platform that incorporates phase smoothing10 of the laser beam in addition to a ~20 J pre-pulse that disassembles the laser-entrance window (a thin polyimide foil) on the top of the target that is required to initially contain the fusion fuel. This platform demonstrated an energy coupling efficiency of ~50% of the main ~2 kJ laser pulse to the fuel and simultaneously reduced mix from the polyimide foil into the fusion fuel.11 The axial magnetic field and current delivered to the target were simultaneously increased from 10 to 15 T and 16.1 MA to 19.4 MA, respectively, by reducing the inductance of the power-feed leading up to the target, increasing the anode-cathode gaps to reduce parasitic current losses, and employing more powerful Helmholtz coils. Simultaneously implementing the increased preheat, axial magnetic field and current delivered resulted in a record MagLIF performance of 1.1x1013 neutron yield (equivalent to 2 kJ DT energy produced), nearly an order of magnitude greater than previous experiments on this platform.
Numerical simulations show additional improvements are attainable by further increasing the applied magnetic field to 30 T, the laser-preheat energy coupled to 6 kJ, and target current to 22 MA, with >100 kJ DT fusion yields produced in 2D simulations using parameters that should be achievable on the Z Machine. Our efforts in the next five years are directed at increasing these input parameters simultaneously. Scaling this target to a next-generation facility is encouraging—fusion yields in excess of 10 MJ are possible with currents of 60 MA, laser preheat energies of 40 kJ, and applied magnetic fields of 19 T. Even larger yields approaching a gigajoule may be possible by propagating the fusion burn into a layer of frozen DT ice on the inside surface of the liner.6,12
Despite the impressive predicted yields, developing MagLIF as an energy producing source will be challenging.13 While the electrical energy delivered to the target is efficient (5-10%), it also generates a destructive post-shot explosion (even in the absence of fusion yields) that destroys the metallic power feed that delivers current to the target, limiting present-day operations to a single shot per day. Gigajoule producing targets would produce a tremendous number of neutrons that could adversely affect pulsed power components while activating materials to potentially hazardous levels. Conceptual solutions to these problems have been investigated, including recyclable transmission lines14 and neutron-absorbing blankets;15 however, our present efforts in MagLIF are directed towards demonstrating the fundamental physics and possibility of attaining significant fusion gains on larger-scale facilities. Surmounting this scientific difficulty remains the first obstacle towards harnessing a magneto-inertial fusion-energy producing system.
References
1. J. Lindl, Phys. Plasmas 2, 3933 (1995).
2. R.D. McBride et al., IEEE Trans. Plasma Sci. 46, (2018).
3. S.A. Slutz et al., Phys. Plasmas 17, 056303 (2010).
4. M.R. Gomez et al., Phys. Rev. Lett. 113, 155003 (2014).
5. P.F Schmit et al., Phys. Rev. Lett. 113, 155004 (2014).
6. S.A. Slutz and R.A. Vesey, Phys. Rev. Lett. 108, 025003 (2012).
7. P.F. Knapp et al., Phys. Plasmas 22, 056312 (2015).
8. K.J. Peterson et al., Phys. Rev. Lett. 112, 135002 (2014).
9. T.J. Awe et al., Phys. Rev. Lett. 116, 065001 (2016).
10. M. Geissel et al., Phys. Plasmas 25, 022706 (2018).
11. A.J. Harvey-Thompson et al., Phys. Plasmas 25, 112705 (2018)
12. A.B. Sefkow et al., Phys. Plasmas 21, 072711 (2014).
13. D.B. Sinars et al., J. Fusion Energy 35, 78 (2016).
14. S.A. Slutz et al., Phys. Plasmas 10, 429 (2003).
15. B.G. Logan, Fusion Engineering and Design, 22, 1953 (1993).This paper describes objective technical results and analysis. Any subjective views or opinions that might be expressed in the paper do not necessarily represent the views of the U.S. Department of Energy or the United States Government.
Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525.
Speaker: D. A. Yager-Elorriaga (Sandia National Laboratories) -
Abstract
The COMPASS tokamak, operated in the Institute of Plasma Physics of the Czech Academy of Sciences in 2009 – 2020, is one of few devices with an ITER-like plasma shape. Its flexibility, extensive set of diagnostics, and NBI heating allow to address key issues in the fusion research in support of ITER and DEMO, such as edge and SOL physics, the L-H transition, runaway electrons and disruption studies, plasma-wall interaction.
Recent results related to these fields, obtained in the last two years, till the final COMPASS tokamak shutdown in September 2020, are the subject this contribution.L-H transition
Extensive sets of experiments have been performed recently in order to study the L-H transition.
The dependence of the L-H power threshold on the X-point height above the divertor PLH(|X-div|) was analyzed in the framework of causal graphical modelling. This motivated the separation (conditioning) of transitions into those with q95~3 and those with higher or lower q95 and also the normalization (counterfactual reasoning) of PLH to a common reference density in order to block confounding effects. The results (see figure) show a clear linear trend where PLH increases by 30-40 kW (~18%) per 1 cm of the X-point height above the divertor. While the trend with |X-div| is similar for all the discharges, those with q95~3 have a base value of PLH larger by 50 kW. This 30% increase in PLH around q95~3 seems to be related to the presence of intrinsic error fields.
![Density-normalized L-H power threshold P<sub>LH</sub> as a function of the X-point height above the divertor |X-div| (left). Critical radial electric field E<sub>r</sub> in the SOL at which the L-H transition occurs (right). In both figures, several outliers are observed for q<sub>95</sub> close to 3. [1, 2]](https://www.hron.name/test/figure.jpg)
Simultaneous measurements of the radial electric field Er in the SOL and inside the separatrix shows that Er increases (in absolute value), both upstream and downstream and inside the separatrix, with decreasing X-point height. This is qualitatively consistent with transport modeling in [3] and also consistent with the idea that the consequent change in the ExB shearing rate could be responsible for the change of PLH.
Furthermore, the effect of controlled HFS error field (EF) on the L-H transition was studied in detail, utilizing the unique COMPASS HFS 3D coils. Using these coils, a displacement of the central solenoid was simulated, while different sets of coils were used to assess the error field correction (EFC) from the LFS and top/bottom of the vessel, which is of high importance for ITER, having in mind the detrimental effects of HFS EF observed on NSTX-U recently [4]. It was shown that the residual EF, after the correction of HFS EF by LFS EFC, can lead to disruptions if present during the L-H transition. While in case of NBI-assisted L-H transitions the disruption rate was around 50% as reported in earlier COMPASS results, in case of purely Ohmic L-H transitions (induced by X-point ramp) the disruptions were inevitable. The critical parameter appears to be the plasma rotation which is very low during Ohmic L-H transitions in COMPASS (and is also expected to be generally low in ITER), but injection of even a small external momentum (PNBI < 100 kW) is sufficient to prevent the disruption.
Run-away electrons (RE) physics
Both experiments and modelling effort contributed significantly to the understanding of crucial topics, such as more efficient mitigation, suppression of RE beam generation, its feedback control, beam detection, transport of RE and their interaction with MHD.
Radial stability of the relativistic RE beam and the role of RE energy was studied and based on this, a new algorithm for the RE beam feedback control was implemented [5]. Vertical electron cyclotron emission (V-ECE) diagnostics has been commissioned [6] and used to monitor RE seed population in the plasma, which is helpful in the mitigation scenarios. The evolution of the RE beam impact and its overall energy was measured using a new dedicated calorimetry head. Studies of low and high Z impurity material impact on the RE beam dynamics showed very promising results thanks to a new vacuum pellet injector system. Besides systematic surveys of the effects of the massive gas injection and impurity seeding, the effect of externally applied resonant magnetic perturbation on the RE dynamics was addressed [7].
A new full orbit particle tracking code taking into account radiation damping and 3D perturbed magnetic field was developed. RE transport model in presence of natural magnetic perturbations based on fractional diffusion theory was developed and qualitative comparison with experiments was made, evaluation of the RE diffusion coefficient with a guiding-center particle code is ongoing. The conditions under which the radiation reaction acting on the charged particle prevails over the collisions were investigated analytically.
Disruptions studies
Disruption studies are focused on electromagnetic loads on the machine. For the first time simultaneous measurements of plasma current asymmetries as well as toroidal and poloidal vessel currents including their poloidal distribution have been performed [8]. New diamagnetic technique for measurements of poloidal current in the wall has been successfully validated in experiments on COMPASS [9]. The disruption forces modelled by CarMa0NL code have been validated against the COMPASS data, and then applied for the design of COMPASS-U vacuum vessel [10].
Two special divertor tiles were installed in COMPASS to directly measure currents flowing between these tiles and the plasma during vertical displacement events (VDEs). They allow better understanding of current pattern distribution within vessel structure and divertor and aim on testing the model of asymmetric toroidal eddy currents (ATEC) [11].
The vacuum vessel motion during disruptions has been studied simultaneously by displacement sensors and accelerometers.
Liquid Metal Divertor (LMD) experiments
Liquid metals are considered as an option to overcome issues related to conventional plasma facing components (PFC), such as melting of leading edges, cracking, morphology and heat capacity deterioration subsequent to neutron irradiation or re-crystallization. One candidate technology is the capillary porous system (CPS) where the liquid metal is impregnated in a metallic mesh and confined against MHD effects by capillary forces. However, this potential solution comes with new issues such as resilience to transients, tritium retention, evaporation, etc. Most of these issues were investigated experimentally for the first time under ELMy H-mode conditions in a tokamak divertor. [12, 13].
A specially designed CPS module filled with liquid Li and with a liquid Sn alloy (LiSn) was installed in the COMPASS divertor for two separate dedicated power exhaust experiments. Good power handling capability of both liquid metals was observed for an averaged deposited, perpendicular heat flux up to 12 MW/m2. The CPS module was exposed to ELMs with relative energy ~3% and a local peak energy fluence ~15 kJ/m2 locally. No droplets were ejected from the CPS surface and no damage of the CPS mesh was observed, as well as no contamination of the core and SOL plasmas by Sn.
References
1 J Seidl, H-mode workshop, 2019
[2] O Grover, EFTSOMP workshop, 2019
[3] AV Chankin, NME, 2017
[4] CE Myers, 2016
[5] O Ficker, NF, 2019
[6] M Farnik, RSI, 2019
[7] J Mlynar, PPCF, 2019
[8] E Matveeva, 46th EPS Plasma Physics, 2019
[9] VV Yanovskiy, 46th EPS Plasma Physics, 2019
[10] VV Yanovskiy, IAEA FEC 2020
[11] R Roccella, NF, 2016
[12] R Dejarnac, PSI, 2020
[13] J Horacek, PSI, 2020Speaker: Dr Martin Hron (Institute of Plasma Physics of The Czech Academy of Sciences) -
The TCV tokamak continues to leverage its unique shaping capabilities, flexible heating systems and modern control system to address critical issues in preparation for ITER and a fusion power plant. For the 2019-20 campaign its configurational flexibility has been enhanced with the installation of divertor gas baffles and its diagnostic capabilities with an extensive set of upgrades. Experiments are performed in part by topical teams, under the auspices of the EUROfusion medium-size tokamak programme and local teams at the Swiss Plasma Center together with international collaborators, resulting in a rich and focused scientific programme.
Auxiliary heating is provided by NBI and ECRH. An improved acceleration grid for the NBI system reduced power losses in the duct allowing for a 2.5x increase in injected energy with an injected power up to 1.3MW. The legacy ECRH system of 1.4MW from two 83GHz gyrotrons (X2) and 0.9MW from two 118GHz gyrotrons (X3) was enhanced by a 1MW dual frequency (84/126GHZ) gyrotron for X2 or X3 heating. The new gyrotron performs as designed validating the numerical models used for its development.
The most conspicuous upgrade was the installation of removable gas baffles that separate the vessel into main and divertor chambers, Fig. 1. The baffles seek to increase the divertor neutral pressure and thereby facilitate the extrapolation to future devices, such as ITER that will rely upon operation with high divertor neutral pressure [1]. They allow for various divertor and limited configurations. Experimentally, Ohmically heated diverted discharges confirm SOLPS-ITER predictions with up to a 5x increase in divertor neutral pressure. Optional fuelling into the divertor or the main chamber can disentangle the effects of fuelling rate, divertor or main chamber neutral pressure, $p_\mathrm{n,div/main}$, and plasma density, $n_\mathrm{e}$.
As predicted, increasing $p_\mathrm{n,div}$ in the baffled divertor facilitates access to detachment, which commences at ~30% lower $n_\mathrm{e}$. Scanning the plasma plugging by displacing the X-point with respect to the baffles indicates that the installed divertor closure may be close to optimal. Experiments also provide further evidence that the onset of detachment is determined by $p_\mathrm{n,div}$ rather than $n_\mathrm{e}$. These experiments ultimately seek to validate edge models. Reversing the toroidal field direction reveals changes in the target currents and the formation of a potential well below the X-point in reverse field as predicted by SOLPS-ITER including drifts.
All previously obtained alternative configurations were achieved in the baffled divertor with detailed investigations first focusing on the Super-X divertor [1]. Specific configurations designed to disentangle the effects of a large target radius, $R_\mathrm{t}$, and the angle between the divertor leg and the target surface confirm an expected strong dependence of the detachment onset on the angle, whereas a predicted dependence on $R_\mathrm{t}$ remains elusive.
Two new gas-puff imaging (GPI) systems, diagnosing the X-point region and the outboard midplane, have greatly increased the ability to investigate scrape-off layer (SOL) transport. Characterisation of the SOL turbulence was also extended into H-mode, linking $p_\mathrm{n,div}$ to the formation of the density shoulder [2].
H-mode studies were facilitated by an apparent reduction in the power threshold, $P_\mathrm{LH}$, for the baffled divertor. Here, the inter-ELM target temperatures were lower and nitrogen seeding led to detachment. At the H-mode density limit, a full MARFE develops as a dense strongly radiating region at the X-point that subsequently moves up the HFS edge.
Particular attention was dedicated to the pedestal in type-I ELMy H-modes, where the baffles lead to significantly higher $T_\mathrm{e,ped}$ and, hence, higher $p_\mathrm{e,ped}$, Fig. 2. The pedestal degrades with fuelling that increases $n_\mathrm{e,sep}$, but decreases $T_\mathrm{e,ped}$, consistent with previous findings. The role of $n_\mathrm{e,sep}$ in the pedestal is further highlighted in discharges with a range of $R_\mathrm{t}$ that require different fuelling rates to obtain the same $n_\mathrm{e,sep}$, but then display the same pedestal characteristics, highlighting alternative divertors’ weak effect on pedestal and core properties.
In a continued effort to extrapolate ELMy H-mode performance to the ITER baseline (IBL) scenario, NBI and X3 heated H-modes succeeded in matching the ITER targets of $\kappa$=1.7, $\delta$=0.4, $\beta_\mathrm{N}$=1.8 and $q_{95}$=3.0 whilst retaining good confinement ($H_\mathrm{98y2}$~1) [3]. The Greenwald fraction reached 0.6, but at those densities X2 ECH absorption becomes unreliable and ELM triggered NTMs were only avoided by lowering $I_\mathrm{P}$ with a stationary demonstration at $q_\mathrm{95}$=3.6.
In the quest to ELM-free regimes, negative triangularity (NT), Fig. 1 (c), is confirmed as an attractive scenario with an L-mode confinement matching that of H-modes for positive triangularity. NBI heating extended the operating space of NT plasmas to the IBL value of $\beta_\mathrm{N}$. However, similarly to the low-$q_{95}$ IBL in TCV, these discharges are prone to NTMs, but inaccessible to X2 ECH control. Reciprocating probe plunges past the LCFS of Ohmic NT plasmas confirm a reduced turbulent E×B flux extending from the core into the plasma edge. The measurements are corroborated by measurements from the new mid-plane GPI system.
TCV continues to address critical aspects of the discharge evolution that may limit plasma performance or even pose a danger to components in ITER and future power plants. This includes NTMs, with a successful validation of a new analytical model of the classical island stability, which will facilitate NTM pre-emption in future devices. Enabled by NBI, fast-ion studies are gaining prominence with the development of robust scenarios that display rich, fast-ion driven, MHD spectra and the commissioning of a fast-ion loss detector. Further experiments aim at understanding and controlling runaway electrons (REs) [4], whether created at low density or by mitigated disruptions, as they may cause severe damage in larger devices. Filtered imaging for multiple wavelength ranges provided the first measurements of synchrotron radiation on TCV, Fig. 3, revealing information about location, size and energy distribution of REs and even allowing to detect pre-disruption seed distributions. RE scenarios were extended from Ne and Ar to He, Kr and Xe injection, and to NT and diverted configurations, increasing the space for model validation. In addition, strategies to purge the impurities after the RE beam formation with further $\mathrm{D}_2$ injection are being explored.
TCV also continues to employ its flexible digital control system to enhance available control solutions [5]. With a view to future long-pulse tokamak discharges, a generic plasma control framework has been developed, implemented and applied to avoid density limit disruptions by controlling the NBI power based upon an estimated proximity to the disruptive boundary. An ability to re-assign EC sources to $\beta_\mathrm{N}$ or NTM control was demonstrated. Plasma exhaust control, for future reactors, was explored using an estimate of the C-III radiation profile along the divertor leg, indicative of the local $T_\mathrm{e}$, with a feedback control of the distance of the radiation from the target to the X-point demonstrated using gas injection as actuator for both L- and H-mode scenarios.
Various short baffled and un-baffled campaigns are planned for 2020 highlighting yet another dimension in TCV’s signature flexibility.[1] C. Theiler, et al., this conference, [2] N. Vianello, et al., this conference, [3] O. Sauter, et al., this conference, [4] G. Papp, et al., this conference, [5] F. Felici, et al., this conference.
Speaker: Holger Reimerdes (Ecole Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC)) -
KSTAR$^{1,2}$ program has been focused on resolving the key physics and engineering issues for ITER and future fusion reactors utilizing unique capabilities of KSTAR. First of all, a new advanced scenario was developed targeting steady-state operation based on the early diverting and heating during the ramp-up phase of plasma current and significant progress has been made in shape control to address the MA level of plasma current and stationary ITER-similar shape (ISS). It is demonstrated effective use of the H&CD with instrumented plasma control and shaping parameters became a key to access to the advanced operation scenarios such as high $β_p$, high $l_i$, high $q_{min}$, hybrid, internal transport barrier (ITB) and low $q_{95}$ operation. The examples of advanced scenarios are shown in Figure 1. The stationary ITB (fig 1a) is successfully reproduced with comparable confinement as H-mode level ($H_{89}$ ~ 2) both in limited and USN configuration, a low qmin scenario (fig 1b) is developed based on early diverting and delayed core heating approach and finally stable long-pulse H-mode operation (fig 1c) was extended upto 88 sec.
Recent KSTAR 3D experiments have focused on several ITER-relevant issues, such as divertor heat flux broadening in 3-row vs 2-row resonant magnetic perturbations (RMPs) on ELM-crash suppression, RMP-driven ELM-crash-suppression on ITER-like low $q_{95}$ (~3.2-3.4) and the characterization of ELM-crash suppression window in terms of normalized electron collisionality ($\nu^*_e$) and plasma toroidal rotation ($V_{tor}$) at pedestal top. Strong up-down asymmetry in 3-row configuration was identified and effect of the kink/anti-kink configuration was also clarified for ELM suppression in LSN plasmas. We have demonstrated the ISS-compatible RMP control in KSTAR using n=2, +900 phasing RMP, although the ISS has been more vulnerable to mode-locking than typical KSTAR configuration. A detailed study of the KSTAR database (where RMP configuration of all the discharges belongs to n=1, +900 phasing) showed that the ELM-crash suppression phase in KSTAR is in the range of 0.2 < $\nu^*_e$ < 1.2 and $V_{tor}$> 40 km/s. During the ELM suppression phase, coexistence of filamentary mode and smaller scale turbulent eddies at pedestal with broad-range of wave number ($k_θ$<1.1 $cm^{-1}$ and frequency (f<100 kHz) is identified by ECE imaging (ECEI) and strong energy exchange of the filamentary and turbulent modes was measured. The bicoherence analysis of the edge harmonic oscillations (EHOs) at natural ELM-less mode shows that there is a strong nonlinear interaction between EHOs, and the nonlinear interaction of EHOs has a significant effect on the ELM structure and dynamics.
Cross-validation between the advanced diagnostics and the modeling provides new insight on the basic transport process at KSTAR. For example, in the recent MHD-quiescent KSTAR plasmas non-diffusive avalanche-like electron heat transport events are observed by the ECEI and these observations have been successfully reproduced by gyrokinetic simulations indicating the broad range of spatial scales up to the minor radius. In addition, various studies utilizing the KSTAR fluctuation diagnostics demonstrated the importance of the turbulence characteristics in plasma rotation and confinement. The extensive study of the intrinsic rotation in Ohmic plasmas found a clear link between the counter-current toroidal rotation direction and the quasi coherent mode (QCM) which is measured by the Microwave Imaging (MIR). The improved confinement in the low rotation experiment was correlated with the suppression of the broadband (~200 kHz) ECEI fluctuations, and Collective Thomson Scattering provides a detailed measurement on the high-k density turbulence which is suppressed during the typical LH transition. Finally, strong interaction between fast-ion and EP driven MHD mode was identified with Fast ion $D_α$ (FIDA) diagnostics.
KSTAR provided unique demonstration on the performance of symmetric multiple Shattered Pellet Injections (SPIs) which is the main strategy of ITER for disruption mitigation. It was shown successfully the current quench rate changes proportionally as the time difference varies from several percent to several tens of percent of the thermal quench (TQ) duration (1~2 ms) and it was demonstrated that peak density was increased twice with dual SPIs compared with a single SPI and energy can be radiated when multiple SPIs are injected simultaneously, as planned in ITER.
Lastly, the research plan in near term will be addressed with the machine upgrades. KSTAR will focus on the development of the DEMO/ITER relevant operational scenario, i.e., high-beta steady-state operation with benign MHD activities which will require robust plasma control in strong shaping, control of MHD modes and thorough analysis of the fundamental physics processes. In these regards, KSTAR upgrades will includes extensive NBI (off-axis, 6MW) & RF (Helicon CD, 4MW) heating & current drive capabilities and the installation of new tungsten divertors with active cooling.References:
$^1$G.S. Lee et al, Nucl. Fusion 40 575 (2000) 575
$^2$H. K. Park et al, Nucl. Fusion 59 (2019) 112020 (13pp)Speaker: Si-Woo Yoon (National Fusion Research Institute) -
The 2019-2020 scientific and technological programme exploits JET’s currently unique capabilities: Tritium handling and ITER-like wall (ILW: Be wall and W divertor). It is the culmination of years of concerted scientific and engineering work, with the ILW installation in 2010, improved diagnostic capabilities, now fully available, a major Neutral Beam Injection (NBI) upgrade providing record power in 2019 (P$_{NBI}$ up to 32MW), and the technical & procedural preparation for safe operation with tritium. Research along three complementary axes yielded a wealth of new results since last IAEA. Firstly, the JET plasma programme delivered scenarios suitable for high fusion power (P$_{FUS}$) and alpha particle ($\alpha$) physics in the coming D-T campaign (DTE2), with record sustained neutron rates, as well as plasmas for clarifying the impact of isotope mass on plasma core, edge and plasma-wall interactions, and for ITER pre-fusion power operation (e.g. L-H transition in He plasmas). The efficacy of the newly installed Shattered Pellet Injector (SPI) [ref1] for mitigating disruption forces and runaway electrons was demonstrated, informing ITER disruption management. Secondly, research on the consequences of long-term exposure to plasma in the ILW was completed, with emphasis on wall damage and fuel retention, and including analyses of wall materials and dust particles. This will help validate assumptions and codes for the design & operation of ITER and DEMO. Thirdly, the nuclear technology programme aiming to deliver the maximum technological return from operations with D, T and D-T [ref3] benefited from the high D-D neutron yield in 2019 (2.26x10$^{19}$n), securing new results for validating radiation transport and activation simulation codes, and nuclear data for ITER. Measuring systems are ready for collecting data in T and in D-T campaigns producing 14MeV neutrons.
Integrated scenarios preparation for high P$_{FUS}$ sustained for 5s (i.e. relevant to energy confinement times in JET) progressed significantly for the two routes investigated: ‘Baseline’ (q$_{95}$~3, I$_P$≥3MA, $\beta$$_N$<2) and ‘Hybrid’ (tailored q-profile, q$_{95}$~4.5, I$_P$ ≤2.7MA, $\beta$$_N$≥2.4). Peak neutron rate of 4.2x10$^{16}$n/s (2.7x10$^{16}$n/s averaged over 5s) are obtained simultaneously with tolerable divertor temperatures and controlled high/medium Z impurity for the full pulse duration in Baseline plasmas at 3.3T/3.5MA, with P$_{TOT}$=34MW (NBI and Ion Cyclotron Resonance Heating (ICRH)). Pellets help controlling the ELM frequency (f$_{ELM}$) needed for impurity flushing, with low total D$_2$ throughput for high confinement. Hybrid plasmas developed to 3.4T/2.3MA reached 4.8x10$^{16}$n/s but MHD avoidance and f$_{ELM}$ control must be optimised for improved, steady performance. The equivalent P$_{D-T}$ for these pulses is consistent with past predictions at same B$_T$, I$_P$, P$_{TOT}$ [ref4], giving confidence in the theory-based modelling. Further gains are likely with 40MW now reachable and higher I$_P$, with divertor heat loads controlled by strike-point sweeping, thus prospects for reaching the target (5x10$^{16}$n/s) are good. In these conditions P$_{D-T}$=11-16MW is predicted by theory-based physics models, with range due to uncertainties in the pedestal predictions and to whether isotope effects are included or not. Fast particle diagnostics, significantly improved since DTE1, can now detect small amount of $\alpha$’s, as shown in dedicated experiments making use of 3-ion ICRH scheme (D-(D$_{NBI}$)-3He) to create MeV range particles, with $\alpha$ (≈10$^{16}$s$^{-1}$) from D+$^3$He reactions. Simultaneous detection of He and hydrogen isotopes with an enhanced high resolution sub-divertor residual gas analyser, as planned for ITER, has been demonstrated.
Experiments and modelling preparing the T campaign. Observations that the impact of isotope on H-mode plasmas comes mainly from the pedestal [ref5] motivated recent gyrokinetic (GK) theoretical investigations of JET pedestals showing that the toroidal branch of the ETG instability can be driven at ion-scale poloidal wavelengths and may be responsible for significant inter-ELM pedestal heat transport. However, in some regimes, isotope effects on core plasma may also be important [ref4]. New experiments in D$_2$ and H$_2$ L-mode plasmas and related core GK modelling show that, in plasmas with a strong stabilizing effect of fast particles, differences in fast particle content with isotope mass may lead to strong deviations from the gyro-Bohm scaling of core transport. Recently developed ICRH-only H-mode plasmas (low input torque, dominant e-heating) show the same normalised confinement factor (H$_{98(y,2)}$) and T$_e$ profiles as their NBI-only counterpart at same PTOT, but n$_e$ profile for the NBI case is 50% higher due to NBI fuelling and possibly different particle transport. Work to clarify the impact of edge/divertor was performed. Experiments at 2MA/2.3T, low triangularity ($\delta$) demonstrated that changes in H$_{98(y,2)}$ and pedestal T$_e$ due to divertor configuration can be condensed into a single trend when mapped to the target T$_e$ as the main parameter governing recycling conditions rather than D$_2$ fuelling rate. A high performance neon seeded scenario (2.7T/2.5MA, high-δ shape) with edge conditions closer to ITER was developed. Neon seeding leads to significant increase (by ≈50%) in pedestal pressure and T$_e$ (from 0.4keV to 0.8keV) and in H$_{98(y,2)}$ (from 0.6 to 0.9) with mitigated divertor power loads. The well diagnosed discharges are used for validating physics-based SOL-edge modelling, increasing confidence in ITER divertor design basis and supporting deployment of neon over chemically reactive N$_2$ as seed gas in ITER.
JET disruption management programme is in two parts: 1) disruption avoidance based on improved termination techniques and on real-time detection of unhealthy plasmas with jump to controlled termination, causing significant reduction in disruption rate in baseline (60% to 20%) and hybrid plasmas (9%), and 2) disruption mitigation with SPI performed as part of a collaboration between ITER, US and Europe. After successful installation and commissioning, extensive experiments took place with the JET SPI demonstrating very good reliability. By varying the neon content in the SPI pellets, the disruption current quench time can be controlled efficiently in JET, scaling to the range required by ITER. High Z impurity SPI also demonstrated run-away electron suppression. Additionally, it was discovered that D$_2$ SPI applied to a high current run-away beam leads to benign impacts on the wall, suggesting a new potential solution for run-away electron control in ITER.
Plasma-facing components (PFC) long term exposure in ILW. Retrieval of PFC, wall probes (including test mirrors) and dust for ex-situ studies after three ILW campaigns provide deep insight into material erosion and deposition. Low mobilization of dust during in-vessel operations is shown. Emphasis was placed on material damage such as melting of the Be upper dump plates (UPD) and the identification of factors triggering this process. Comprehensive studies including imaging survey, morphology changes, mass loss and fuel inventory analysis on the most affected UDP tiles was performed. The undisputed reason for melting was unmitigated disruption events which tend to move the melt layers in the poloidal direction resulting in formation of upwards going waterfall-like structures of molten metal. The halo current is believed to provide the j x B force driving the melt layer motion. Global material migration results constitute a unique dataset for modelling and thus improved predictions for ITER.
[ref1] L. R. Baylor, et al., Nucl. Fusion 68 (2019) 211, [ref2] I. Jepu et al., Nucl. Fusion 59 (2019) 086009, [ref3] P. Batistoni et al, Fusion Engineering and Design Vol 109-11, 2016, [ref4] J. Garcia et al. Nucl. Fusion 59 (2019) 086047 [ref5] C. F. Maggi et al., Plasma Phys Control Fusion 60 (2018) 014045Speaker: Dr Joelle Mailloux (UKAEA) -
A. Bhattacharjee(a), B. Allen(g), C.-S. Chang(a), H. Chen(f), Y. Chen(f), J. Cheng(f), E. D’Azevedo(b), P. Davis(e), J. Dominski(a), M. Dorf(c), M. Dorr(c), S. Ethier(a), A. Friedman(c), K. Germaschewski(h), R. Hager(a), A. Hakim(a), G. Hammett(a), J. Hittinger(c), S. Janhunen(d), F. Jenko(d), S. Klasky(b), S. Ku(a), R. Kube(a), L. LoDestro(c), N. Mandell(a), G. Merlo(d), A. Mollen(a), M. Parashar(c), J. Parker(c), S. Parker(f), F. Poli(a), L. Ricketson(c), A. Scheinberg(a), M. Shepherd(i), A. Siegel(g), S. Sreepathi(b), B. Sturdevant(a), E. Suchyta(b), P. Trivedi(a), G. Wilkie(a), and M. Wolf(b)
(a) Princeton Plasma Physics Laboratory, Princeton, NJ, (b) Oak Ridge National Laboratory, Oak Ridge, TN, (c) Lawrence Livermore National Laboratory, Livermore, CA, (d) Institute for Fusion Studies, University of Texas at Austin, Austin, TX and Max Planck Institute for Plasma Physics, Garching, (e) Rutgers University, New Brunswick, NJ, (f) University of Colorado, Boulder, CO (g) Argonne National Laboratory, Chicago, IL
(h)University of New Hampshire, Durham, NH (i) Rensselaer Polytechnic Institute, Troy, NYWhole Device Modeling (WDM) is generally described as assembling physics models that provide an integrated simulation of the plasma. All components that describe a magnetic confinement device, from macroscopic equilibrium to micro-turbulence and control systems, are included in WDM, which describes the evolution of a plasma discharge from start-up to termination. Economical and safe operation of burning plasma devices requires predictive WDM with a confidence level established by validation and uncertainty quantification. Simulations covering the whole device, while certainly not a substitute for experiments, are much more cost-effective than building multiple billion-dollar facilities to test new ideas or concepts, similar to how aircraft manufacturers used simulations to reduce the number of physical wings they needed to build in designing superior aircraft (1).
The High-Fidelity Whole Device Model of Magnetically Confined Fusion Plasma is an application (hereafter referred to as WDMApp) (2,3) in the DOE Exascale Computing Project (ECP). The ECP is a DOE 413.3b project---the largest in the DOE Office of Science--- and is governed by the same rigorous rules of operation as major experimental facilities. The ultimate problem target of the project is the high-fidelity simulation of whole device burning plasmas applicable to an advanced tokamak regime (specifically, an ITER steady-state plasmas with ten-fold energy gain), integrating the effects of energetic particles, plasma-material interactions, heating, and current drive. The most important step in this project, and one that involves the highest risk, is the coupling of two existing, well-established, extreme-scale gyrokinetic codes – the GENE continuum code for the core plasma, and the XGC particle-in-cell (PIC) code for the boundary plasma. We have accomplished this challenging milestone for the first time in the magnetic fusion community. Fig. 1 demonstrates a coupled GENE-XGC in a WDMApp simulation for nonlinear ITG turbulence.
These developments would not be possible without the remarkable advancements in edge turbulence simulation codes for which XGC is an exemplar, along with COGENT and GKEYLL.
COGENT is a continuum gyrokinetic code for edge plasmas (4, 5). The code is distinguished by its use of fourth-order conservative discretization and mapped multiblock grid technology to handle the geometric complexity of a tokamak edge. It solves full-f gyrokinetic equations for an arbitrary number of plasma species, which can also be coupled to a set of lower-dimensionality fluid equations in cases where a reduced fluid model is adopted to describe electrons or neutrals. The code offers a number of collision models, ranging from the simple Krook operator to the fully nonlinear Fokker-Plank operator, and includes an ad-hoc anomalous transport model that can be utilized for the case of 4D axisymmetric transport calculations. Recent applications of the COGENT code to the analysis of cross-separatrix edge plasma properties include (a) 4D calculations, which demonstrate the values of radial electric field and toroidal rotation comparable to those observed on the DIII-D facility, and (b) 5D calculations of ITG turbulence, which elucidate the role of magnetic shear stabilization in the X-point region.
The GKEYLL project (6) is developing a continuum gyrokinetic capability that can evolve the electromagnetic gyrokinetic equations in the tokamak edge. The code uses a Hamiltonian form of the full-f equations, and for electromagnetic terms, uses a symplectic formulation. A novel version of a high-order discontinuous Galerkin scheme is used, ensuring that total energy (particles plus fields) is conserved by the spatial discretization. GKEYLL has performed the first fully nonlinear full-f continuum simulations of electromagnetic gyrokinetics in the scrape-off layer (SOL), including sheath boundary conditions. In Fig. 2 we show a snapshot of the turbulence profiles when statistical steady-state has been obtained. Shown are density and temperature contours near the midplane. Intermittent blob-like structures are seen ejected from the source region as they propagate outwards. Comparisons with electrostatic simulations show that the turbulence is larger amplitude and much more intermittent in the electromagnetic case. Also shown are magnetic field lines between the top and bottom divertor plate being stretched by blobs. Full details of the scheme and detailed description of the results are described in (7).
References
(1) “Case Study: Boeing Catches a lift with High Performance Computing,” Report by Council on Competitiveness, 2009
(2) G. Merlo, J. Dominski, A. Bhattacharjee, C.-S. Chang, F. Jenko, S. Ku, E. Lanti, and S. Parker, Phys. Plasmas 25, 062308 (2018)
(3) J. Dominski, S.-H. Ku, C.-S. Chang, J. Choi, E. Suchyta, S. Parker, S. Klasky, and A. Bhattacharjee, Phys. Plasmas 25, 072308 (2018)
(4) M. Dorf and M. Dorr, Contr. Plasma Phys. 58, 434 (2018)
(5) M. Dorf and M. Dorr, Contr. Plasma Phys. (2020); available online DOI: 10.1002/ctpp.201900113
(6) A. Hakim, N. Mandell, T. Barnard, M. Francisquez, G. Hammett, and E. Shi, Continuum Electromagnetic Gyrokinetic Simulations of Turbulence in the Tokamak Scrape-Off Layer and Laboratory Devices, submitted to Phys. Plasmas. (2020)
(7) N. Mandell, A. Hakim, G. Hammett, and M. Francisquez, J. Plasma Physics 86, 149 (2020)Speaker: Mr Amitava Bhattacharjee -
A new era of predictive integrated modeling has begun. The successful validation of theory-based models of transport, MHD stability, heating and current drive, with tokamak measurements over the last 20 years, has laid the foundation for a new era where these models can be routinely used in a "predict first" approach to design and predict the outcomes of experiments on tokamaks today. The capability to predict the plasma confinement and core profiles with a quantified uncertainty, based on a multi-machine, international, database of experience, will provide confidence that a proposed discharge will remain within the operational limits of the tokamak. Developing this predictive capability for the first generation of burning plasma devices, beginning with ITER, and progressing to tokamak demonstration reactors, is a critical mission of fusion energy research. Major advances have been made using this predict first methodology. Extensive predictive modeling has informed the planning for the JET D-T campaign. This includes integrated modeling of JET hybrid regimes with newly upgraded heating sources, for various concentrations of deuterium D and tritium T. The self-consistent profiles of tungsten, ion and electron temperature, toroidal rotation and densities, have been predicted using theory-based turbulence and neoclassical transport models. The EPED model predicts it is possible to access the, higher pressure, super-H pedestal regime for JET achievable shapes. This prediction has been confirmed with DIII-D experiments. Super-H experiments on JET are planned. A new high accuracy neural network fit to the QuaLiKiz transport model has been completed, opening the way to time dependent predictions, at near real time speed, of complete tokamak discharges. Neural network fits to the TGLF and Multi-Mode models are progressing. The EAST tokamak is using predictive modeling to optimize the high bootstrap fraction regime for fully non-inductive operation and to plan future upgrades of power and current drive systems. A new integrated modeling workflow called TRIASSIC is being developed and tested on the KSTAR tokamak. Predictive modeling of CFETR is informing the design activity. ITER is using predictive modeling to simulate phases of the experimental operations plan. An overview of several of these recent advances will be presented, providing the integrated modeling foundations of experimental successes, as well as progress towards the goal of integrated predictive modeling for experimental design. Two examples, selected from the many advances in the prediction of tokamak experiments, are summarized in this synopsis.
1st example: The fast response of cold pulses due to impurity injection in tokamaks, with an inversion of the inward electron temperature pulse from decrease to increase, has long been argued to be inconsistent with a local transport paradigm [1]. The first demonstration that the cold pulse temperature response could be captured by a local turbulence transport model (TGLF [2]) was performed for the C-MOD tokamak [3]. Only electron and ion temperatures were predicted in these cases, with the density profile being evolved in a prescribed way. It was found that the inversion of the electron temperature pulse from decrease to increase was caused by the stabilization of the trapped electron mode (TEM) by the flattening of the electron density profile. In discharges where the TEM mode was not dominant there was no inversion in agreement with experiment. The transport model was then used in predict first method to simulate the cold pulse response [4] in the DIII-D tokamak. The very fast, high spatial resolution, density profile data on DIII-D confirmed the speed of the prescribed density response and the electron temperature response predictions were confirmed. The final step was to prove that the TGLF model could predict the fast density response to the impurity injection. This required adding the injected impurity density to the transport modeling. This integrated modeling was performed for experiments on the ASDEX Upgrade tokamak [5]. The predicted electron temperature response is compared with data in
. It was found that the destabilization of the ion temperature gradient mode (ITG) by the transiently hollow impurity profile increased the speed of propagation of the electron density pulse into the core. Thus, the speed of the combined electron, ion, and impurity, temperature and density pulses were accurately modeled and new physics insights were discovered. This is a convincing proof that local turbulence transport can account for the paradoxical cold pulse phenomenon.
2nd example: The new upgrades to off-axis NBI current drive capability on DIII-D were preceded by state of the art integrated modeling [6] illustrated by the advanced tokamak predictions in
. The profiles in Fig. 2 are a steady state self-consistent solution of the pedestal structure (height and width), core transport, MHD equilibrium and heating and current drive using validated theory-based models. An iterative high performance workflow IPS-FASTRAN was developed to find the integrated optimum solution [6]. Well validated theory-based models for MHD equilibrium (EFIT) and stability (DCON), turbulent transport (TGLF), pedestal structure (EPED1), neutral beam heating and current drive (NUBEAM) and electron cyclotron heating and current drive (TORAY-GA) were integrated. The IPS-FASTRAN modeling predictions have been confirmed with experiments showing good agreement that will be reported at the FEC 2020 conference. Verification of the accuracy of these predict first method simulations are a valuable test of the new capabilities. The same integrated modeling workflow is being used in the design of the CFETR, CAT Fusion Pilot Plant and SPARC tokamaks and to predict ITER plasmas.
This work was supported by the US Department of Energy under DE-FG02-95ER54309, DE-FC02-04ER54698, DE-SC0019736
[1] K. W. Gentle, W. L. Rowan, R. V. Bravenec, G. Cima, T. P. Crowley, et al., Phys. Rev. Lett. 74 (1995) 3620
[2] G. M. Staebler, J. E. Kinsey, and R. E. Waltz, Phys. Plasmas 14, (2007) 055909.
[3] P. Rodriguez-Fernandez, A. E. White, N. T. Howard, B. A. Grierson, G. M. Staebler, et al., Phys. Rev. Lett. 120 (2018) 075001.
[4] P. Rodriguez-Fernandez, A. E. White, N. T. Howard, B. A. Grierson, L. Zheng, et al., Phys. Plasmas 26, (2019) 062503.
[5] C. Angioni, E. Fable, F. Ryter, P. Rodriguez-Fernandez, T. Putterich, and the ASDEX Upgrade team, Nuclear Fusion 59 (2019) 106007.
[6] J. M. Park, J. R. Ferron, C. T. Holcomb, R. J. Buttery, W. M. Solomon, et al., Phys. Plasmas 25 (2018) 012506.Disclaimer-This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
Speaker: Gary Staebler (General Atomics) -
Since the last IAEA-FEC, the EAST research programme has been, in support of ITER and CFETR, focused on development of the long-pulse steady-state (fully non-inductive) high beta H-mode scenario with active control of stationary and transient divertor heat and particle fluxes $^{[1]}$. The operational domain of the steady-state H-mode plasma scenario on EAST has been significantly extended with the ITER-like configuration, plasma control and heating schemes. Several important milestones on the developments of the high beta H-mode scenario and its related key physics and technologies have been achieved.
-
A minute time scale long-pulse steady-state high beta H-mode discharge (shown in figure 1) with the major normalized plasma parameters similar to the design of the CFETR Phase-III 1GW fusion power operation scenario ($\beta_P$ ~ 2.0, $\beta_N$ ~ 1.6, $f_{bs}$ ~ 50%, $H_{98(y2)}$ > 1.3 at $q_{95}$ = 6.5~7.5) has been successfully established and sustained by pure RF heating on EAST with the ITER-like tungsten divertor $^{[2]}$, as shown in figure 1. The important feature of this high beta H-mode plasma scenario is that, due to the stabilization effect of the Shafranov shift on the plasma turbulence, a higher $\beta_P$ results in a better plasma confinement as shown in figure 2. Further simulation suggested that a high density gradient promotes the ITB formation in high $\beta_P$ plasmas, which might further benefit the development of this high beta plasma scenario towards a high density regime.
-
Active control of divertor radiation has been successfully integrated into the high beta H-mode ($\beta_P$ ~ 2.5, $\beta_N$ ~ 2.0, $f_{bs}$ ~ 50%) plasma scenario without a degradation of the plasma confinement ($H_{98(y2)}$ > 1.2) at high density ($n_e$/$n_{GW}$ ~ 0.7) and moderate edge safety factor ($q_{95}$ ~ 6.7) $^{[2]}$. The peak heat flux on the tungsten divertor was reduced by ~30% with active impurity seeding of a mixture of 50% neon and 50% deuterium. The high-Z impurity concentration in the plasma core has been well controlled in a low level by applying the on-axis ECRH and reducing the fast ion losses through beam energy optimization.
-
The grassy-ELM regime has been extended to the normalized parameter space designed for the CFETR 1GW fusion power operation scenario. This regime exhibits good compatibility with high $f_{bs}$ and fully non-inductive operation, being characterized by a low pedestal density gradient and a wide pedestal, which prevents large-ELM crashes due to an expansion of the peeling-ballooning boundary after the initial pedestal collapse, indicated by BOUT++ nonlinear simulations. High separatrix density makes this regime especially suitable for operation with divertor detachment. Several feedback control schemes have been developed to achieve sustained detachment with good core confinement $^{[3]}$. This includes control of total radiation power, target electron temperature, and particle flux measured by divertor Langmuir probes or a combination of the control of target electron temperature and AXUV radiation near the X point. Integration of these detachment feedback control schemes with the grassy-ELM regimes and the high $\beta_P$ scenario has been demonstrated with neon seeding, which provides an integrated high beta scenario applicable to long-pulse operation $^{[4]}$.
-
ELM suppression has been achieved using various different methods on EAST. Full suppression of ELMs has been demonstrated, for the first time, for ITER-like low torque injection plasmas by using n=4 resonant magnetic perturbations (RMPs) $^{[5]}$. A moderate reduction (~5%) of the energy confinement has been observed together with significant reduction of both the plasma density(20%)and the Tungsten concentration (a factor of 2) during ELM suppression. The ELM suppression window agrees well with the prediction by MARS-F modelling. Robust ELM suppression by Boron powder injection without confinement degradation or even with confinement improvement has been achieved in a wide parameter range. EHO-like edge coherent modes were excited during the ELM suppression phase by Boron powder injection. In addition, simulation results from BOUT++ confirm that both the helical current filaments (HCF) driven by lower-hybrid waves (LHWs) and RF sheath effects on the ICRF antenna contribute to ELM suppression.
-
A flowing liquid lithium (FLiLi) limiter plate has been successfully assembled and tested in EAST for the first time. This plate was designed based on the concept of liquid metal infused trenches (LiMIT), which is using thermoelectric MHD to drive liquid Li flow along the surface channels. The preliminary results show that with the increase of Li flow rate, the fuel particle recycling is gradually reduced, and the plasma performance is slightly improved. There was no obvious Li burst and limiter damage even at a high injection power of 5.5 MW. In addition, ELM mitigation was observed with FLiLi operation.
-
For the first time, EAST has been operated with helium to support the ITER needs. The H-mode power threshold in a helium plasma is found to be 1.2-2.2 times higher than the scaling law in deuterium plasma with pure RF-heating $^{[6]}$. The $C_{He}$ plays a crucial role in determining the energy confinement and pedestal characteristics in helium H-mode plasmas. Divertor detachment is more difficult to achieve in He than in D. The control of divertor heat loads and W sources is achieved by RMP along with neon impurity seeding. The inter-ELM W erosion rate in He is about 3 times that in D with similar divertor conditions, while the intra-ELM W sputtering source shows a strong positive correlation with the ELM frequency $^{[7]}$.
-
A new lower tungsten divertor with a higher closure has been designed and to be installed on EAST in 2020. Several key subsystems, including the heating & current drive, cryogenic, plasma control and diagnostics will be upgraded accordingly for achieving the next milestones i) >400s long-pulse H-mode operation with ~50% bootstrap current fraction, and ii) demonstration of power exhaust at ~10 MW power injection for >100s.
Reference:
1) B. N. Wan et al 2019 Nucl. Fusion 59 112003
2) X. Z. Gong et al 2020 this conference
3) L. Wang et al 2020 this conference
4) G. S. Xu et al 2020 this conference
5) Y. W. Sun et al 2020 this conference
6) B. Zhang et al 2020 this conference
7) R. Ding et al 2020 this conference
Speaker: Baonian Wan (Institute of Plasma Physics, Chinese Academy of Sciences) -
-
Construction of JT-60SA is progressing on schedule towards completion of assembly in March 2020 and the first plasma in September 2020. As of January 2020, manufacture and assembly of all the main tokamak components have been successfully completed satisfying technical requirements including functional performances and dimensional accuracies. Development of plasma actuators and diagnostics is also going well such as achievement of long sustainment of high energy intense negative ion beam. Commissioning of the power supply and the cryoplant has also satisfied requirements. Development of all the control systems and evaluation procedures of tokamak operation has been completed towards the Integrated Commissioning starting in April 2020, and plasma operation scenarios in the first plasma phase have been established. Unique importance of JT-60SA for H-mode and high-beta steady-state plasma research has been confirmed using advanced integrated modellings. These experiences of assembly, integrated commissioning and plasma operation of JT-60SA contribute to ITER risk mitigation and efficient implementation of ITER operation.
Introduction
The JT-60SA (R/a =3m/1.2m, Ip-max =5.5MA, heating power = 41MW x 100s) project [ref.1] was initiated in 2007 under the framework of the Broader Approach agreement by EU and Japan for early realization of fusion energy by conducting supportive and complementary works for ITER towards DEMO. Construction of JT-60SA is progressing successfully towards completion of assembly in Mar. 2020 and the first plasma in Sep. 2020 by the very close collaboration between QST in Japan, F4E in Europe, EU Voluntary Contributors and EUROfusion. The JT-60SA Research Plan [ref.2] covering its machine lifetime of ~ 20 years coordinated with ITER and DEMO schedules has been established with variety of plasma prediction using integrated modeling codes [ref.3]. Recently in Nov. 2019, a new collaboration arrangement between ITER and JT-60SA was signed which covers assembly, integrated commissioning and operation/experiments for finalization of ITER component design, risk mitigation and efficient implementation of ITER operation.Tokamak Construction
After the last IAEA FEC [ref.1], manufacture of all remaining tokamak components has been completed successfully including, superconducting Centre Solenoid (CS), thermal shields, Cryostat Top Lid, Cryolines, etc. As of Dec. 2019, the closure of the vacuum vessel has been accomplished, and the tokamak has been covered by the Cryostat Vessel Body (Fig.1). All the tokamak components have been assembled with excellent dimensional accuracy of ±1mm thanks to careful and smooth positioning using specially designed jigs, high accuracy measurement by Laser trackers, and fine adjustment utilizing sims. The magnetic field error is now expected below 10-4 Bt as designed. Commissioning operation of all large power supply systems, the Quench Protection Cirquit, the Switching Network Units and Super Conducting Magnet Power Supplies, has also been progressed with few residual commissioning activities still ongoing. The commissioning operation of the Cryoplant (equivalent refrigeration capacity of 9 kW at 4.4K) has also been successfully completed by satisfying the required performances.Plasma Hating Systems
For the heating systems, Positive-ion source NBs (85keV, 100sec, 20MW by 12 unit), Negative-ion source NBs (500keV, 100sec, 10MW by 2 units), ECH with multiple frequency Gyrotron (110GHz & 138GHz for 100s and 82GHz for 1 sec) and movable launchers, R&D have been steadily progressing and the targets of their development have been achieved. In particular, high energy intense hydrogen negative ion beams with 500 keV, 154 A/m2 for 118 s, which exceeds the requirement for JT-60SA, has been demonstrated by using a semi- cylindrical negative ion source with a three-stage accelerator. This result was realized by integration of i) stable voltage insulation by suppression of arching, ii) precise beam control and iii) stable negative ion production by maintaining the temperature balance in the negative ion source.
Integrated Commissioning and Control Systems
From April 2020 to Feb. 2021, the integrated commissioning is planned with the first plasma in Sep. 2020 and subsequent 5 months of machine commissioning with plasmas (‘the first plasma phase’). In this phase, the goal of plasma operation is to demonstrate equilibrium controllability of MA-class (<2.5MA) diverted plasmas with the full performance superconducting coil systems, 1.5MW ECH and upper divertor. For such tokamak operations, the Supervisory Control System and Data Acquisition System (SCSDAS) has been developed having the roles of (a) plant monitoring and machine state management, (b) discharge sequence management, (c)real-time plasma control, (d) device protection and human safety, (e) data storage, archive, and database management etc. For plasma controls, we have simulated operation scenarios of the first plasma phase with a newly developed advanced codes with control logics, such as pre-magnetic optimization scheme, plasma equilibrium control with iso-flux control method, control gain optimization method, and strategies for accessing stable operational regimes. Figure 2 shows the discharge scenario at Ip=2.5MA. EC wall cleaning operations and EC-assisted breakdown are also explored with optimized EC injection and toroidal / horizontal field. These results in the JT-60SA first plasma phase will contribute to highly valued subjects in ITER first plasma/subsequent operations.
Scenario Development for ITER and DEMO and Risk Mitigation of ITER
After machine enhancements in 2021-2022, physics experiments will start in 2023 using in-vessel coils, particle fueling and pumping with lower divertor, enhanced diagnostics and high heating power of 26MW at Ip up to 5.5 MA. Toward this phase, variety of predictions of H-mode and high-beta steady-state plasmas covering divertor-SOL-Pedestal-Core have been progressing using advanced integrated modellings including newly-developed globally optimized steady-state transport solver GOTRESS coupled with turbulence models and pedestal models, Gyrokinetic theory based neural-network transport modeling DeKANIS, etc. These studies have confirmed the unique and important characteristics of JT-60SA (highly-shaped, high-beta, 500keV high energy ions, electron heating, controllable rotation etc.) for study of fusion plasma physics such as impacts of fast ions and plasma shape on microturbulence. As for operation scenario development of high-beta steady-state with controlled divertor heat load, an important result has been achieved using the integrated divertor code SONIC upgraded to treat multiple impurity species simultaneously. The result has shown that ‘mixture-seeding of Ar with small amount of Ne’ can keep the peak heat load below allowable 10MW/m2 together with smaller Ar concentration in the SOL and core plasmas than an Ar-only case. These studies have also confirmed significant roles of JT-60SA for ITER risk mitigation (disruption and ELM mitigation) including magnetic perturbation effect on both transient and stationary heat load, vertical displacement event, plasma response to massive gas injection, pedestal and ELM stability and control with Pellet and RMP.[ref.1] P. Barabaschi, Y. Kamada, H. Shirai and JT-60SA Intrgrated Project Team,
Nucl. Fusion 59 (2019) 112005.
[ref.2] JT-60SA Research Plan - Version 4.0, Sept. 2018,
http://www.jt60sa.org/pdfs/JT-60SA_Res_Plan.pdf
[ref.3] G. Giruzzi, M. Yoshida et al., Plasma Phys. Control. Fusion, 62 (2020) 014009.Speaker: Yutaka Kamada (National Institutes for Quantum and Radiological Science and Technology) -
DIII-D physics research addresses critical challenges for operation of ITER and the next generation of fusion energy devices through a focus on innovations to provide solutions for high performance long pulse operation, development of scenarios integrating high performance core and boundary plasmas, and fundamental plasma science and model validation. Substantial increases in off-axis current drive efficiency from an innovative top launch system for EC power (Fig. 1), and in pressure broadening for Alfven eigenmode control from a co-/counter-Ip steerable off-axis neutral beam, both improve the prospects for optimization of future long pulse/steady state high performance tokamak operation. A high beta-p optimized-core scenario with an internal transport barrier that projects nearly to Q=10 in ITER at 9 MA was coupled to a detached divertor, and a Super H-mode optimized-pedestal scenario with co-Ip beam injection (Fig. 2), was coupled to a radiative divertor. Fundamental studies into the evolution of the pedestal pressure profile, and electron vs. ion heat flux, measuring both density and magnetic field fluctuations, validate predictive models of pedestal recovery after ELMs (Fig. 3).
Link to High Resolution Figures 1, 2, 3
The achievement of more than double the off-axis ECCD efficiency using top launch geometry compared with conventional low field side (LFS) launch, as predicted by quasi-linear Fokker-Planck simulations, is due to the longer absorption path for the EC waves which also interact with higher v|| electrons that suffer fewer trapping effects than outside launch. In addition, the new unique co-/counter-Ip steerable off-axis neutral beam broadens the energetic particle (EP) pressure profile and reduces Alfven eigenmode (AE) drive in scenarios with both high toroidal rotation and those with net zero average input torque. New EP measurements show a beam current threshold for Compressional AEs, insensitivity of Beta-induced Acoustic AEs to fast beam ions, and resolution of phase-space flows caused by AEs, from first-of-a-kind Ion Cyclotron Emission (ICE) and Imaging Neutral Particle Analyzer (INPA) data.
Studies of high current runaway electron (RE) beams reveals excitation of current-driven (low safety factor) kink instabilities that promptly terminate the RE beam on an Alfvenic time-scale, offering an unexpected alternate pathway to RE beam mitigation without collisional dissipation. Newly developed real-time stability boundary proximity control and neural-net-based Vertical Displacement Event (VDE) growth-rate calculations are shown to prevent VDEs. The effectiveness of emergency shutdown and disruption prevention tools projects to at least 50% of ITER disruptions being delayed until normalized-Ip is at safe levels, and demonstration of a novel technique for healing flux surface with 3D fields shows promise for providing current quench (CQ) control. Single and multiple Shattered Pellet Injection particle assimilation rates and current quench (CQ) densities are shown to be predictable from 0-D simulations and empirical scaling laws.
Several core-edge integration scenarios demonstrate coupling of a high performance core and radiative divertor operation for target heat flux control. High density and stored energy plasmas with Super H-mode edge pedestals were made both in a lower single null shape accessible by JET and in a higher triangularity near double null shape coupled to a radiative divertor for target heat flux control using nitrogen injection in a core-edge integrated scenario. High-performance plasma with high poloidal beta, large Shafranov shift, and Te and Ti internal transport barriers coupled to a detached divertor with active feedback-controlled Nitrogen puffing also demonstrated integration of core-edge solutions. A high performance hybrid core demonstrated compatibility with radiative divertor operation using Neon or Argon gas injection. Core impurity peaking in the hybrid was substantially reduced using near-axis electron cyclotron heating.
The ability to predict the impurity seeding needed for divertor dissipation has advanced through new capability for measuring charge-state resolved densities of impurity species in the divertor. Also electric drifts in detached divertors with convection dominated heat transport lead to expanded radiative volume. Using these advances, SOLPS-ITER simulations show the synergy between SOL drifts and the SAS divertor geometry for achieving lower density detachment. Modeling of intra-ELM tungsten gross erosion with an analytic Free-Streaming plus Recycling Model is now validated in ITER-relevant mitigated-ELM regimes using pellet pacing and RMPs. SOL tungsten transport in plasmas with both BT directions is consistent with strong entrainment in SOL flows and ExB drift effects.
Advances in pedestal physics through new measurements of density and internal magnetic fluctuations suggest a possible role for micro-tearing and trapped electron modes in DIII-D pedestal transport. Main ion CER measurements indicate ion heat flux is anomalous at low collisionality and transitions to near neoclassical levels at high collisionality. Plasma rotation scans, and both new non-linear analytic theory and 2-fluid code simulations, confirm that ELM suppression by RMPs requires near zero ExB velocity at the top of the pedestal, and achieving suppression appears to be closely linked to a high field side plasma response. The wide pedestal QH-mode regime was obtained with zero input beam torque, electron heating, and LSN shape, consistent with requirements for ITER.
Recent fundamental research on L-H mode power threshold physics shows that turbulence driven shear flow through Reynolds stress and the coexistence of modes associated with various instabilities can lower the L-H power threshold across multiple parameters: eg. q95 and ion grad-B drift direction. Application of RMPs raises turbulence decorrelation rates and reduces Reynolds stress driven flow and flow shear, hence increasing the L-H power threshold. Finally, plasmas with negative triangularity show weak power degradation of H-mode level core confinement while maintaining an L-mode-like edge without ELMs.
In 2020 and beyond DIII-D will install additional tools for optimizing tokamak operation through current and heating profile control using a low field side 1 MW helicon high harmonic fast wave CD system, a unique high field side Lower Hybrid CD system, increased ECH power, and coupling to boundary advances using a new high power closed divertor and a wall insertion test station. Experiments will continue the optimized coupling of high performance core and high power density divertor solutions.
This work was supported in part by the US DOE under contracts DE-FC02-04ER54698 and DE-AC52-07NA27344Speaker: Max Fenstermacher (LLNL @ DIII-D) -
We present here recent highlights from Wendelstein 7-X (W7-X), the most advanced and largest stellarator in the world, in particular stable detachment with good particle exhaust, low impurity content, and energy confinement times exceeding 100 ms, maintained for tens of seconds, as well as proof that the reduction of neoclassical transport through magnetic field optimization is successful. W7-X, which has a magnetic field strength of 2.5 T and a plasma volume of 30 m$^3$, started operation in 2015 [1-5]. Following the installation of a full set of in-vessel components, in particular 10 passively cooled fine-grain graphite test divertor units, it was operated again in 2017 and 2018. Plasma pulses up to 100 s were successfully sustained [6], despite the lack of active cooling. Stable and complete detachment was achieved routinely. The pumping efficiency was initially relatively low [7-9] but it significantly improved later. Detachment with high pumping efficiency was achieved for up to 28 seconds at a heating power of 5 MW with a very low impurity content [10] (Figure 1), indicating control of divertor-heat-flux, plasma density, and impurity content, and giving confidence for reaching the foreseen high-performance, quasi-steady-state (30 minutes) discharges in the future [11]. The performance of the W7-X divertor, and the behavior and parameters of the edge- and scrape-off-layer plasma are now understood in quite some detail [eg. 12], thanks to measurements from a suite of diagnostics [eg. 13-17].
The earlier-reported stellarator triple-product record discharge [18] has now been shown to provide proof that the optimization for reduced neoclassical transport in W7-X was successful, Figure 2 [19]: The high temperature (appr. 3.5 keV for both ions and electrons in the center) and high hydrogen ion density (appr. 7x1019 m-3 in the core) were achieved with 5 MW of heating, and an energy confinement time of 0.22 s corresponding to about 1.4 times the energy confinement time expected from the ISS04 stellarator scaling [20]. For other, less optimized stellarators scaled to the W7-X size and magnetic field strength, similar plasma temperature and density profiles would have required significantly higher heating power to balance neoclassical transport, in particular in the mid-radius (strong gradient) region.
A number of discharges with similar performance to the triple-product-record discharge have since been achieved. These are generally characterized by core density peaking, and a reduction of turbulent density fluctuations. Without such turbulence reduction, the central ion temperature appears to be clamped to appr. 2 keV [21]. These findings are consistent with W7-X transport usually being dominated by ITG turbulence, but stabilized by strong density gradients in a so-called stability valley [22], as exemplified in Figure 3 for the W7-X standard configuration.
During turbulence-dominated phases, impurity confinement times are low (of order the energy confinement time) and no impurity accumulation is seen, but they can be very large if the turbulence is suppressed, and this then leads to impurity accumulation [23]. Recent findings from the Large Helical Device (LHD) show that ITG-dominated discharges readily mix hydrogen isotopes, whereas electron-scale (trapped-electron mode) turbulence does not [24]. It is tentatively concluded that a non-negligible amount of ITG turbulence is beneficial for impurity control as well as for fuel (isotope) exchange and helium exhaust in a stellarator fusion reactor, whereas too much ITG turbulence could potentially clamp the ion temperature below the burn point. These and other recent results [see eg. 25-31] will be put into the context of future goals for the W7-X, the world stellarator program, and the magnetic confinement fusion program in general.
References
1 T. Klinger et al, Plasma Phys. Controlled Fusion 59(1) 014018 (2017)
2 H.-S. Bosch et al, Nuclear Fusion 57, 116015 (2017)
3 R. C. Wolf et al, Nuclear Fusion 57 102020 (2017)
[4] T. Sunn Pedersen et al, Physics of Plasmas 24 055503 (2017)
[5] A. Dinklage et al, Nature Physics (2018) https://doi.org/10.1038/s41567-018-0141-9
[6] T. Klinger et al, Nuclear Fusion 59 112004 (2019)
[7] T. Sunn Pedersen et al, Nuclear Fusion 59 096014 (2019)
[8] D. Zhang et al, Phys. Rev. Lett. 123, 025002 (2019)
[9] D. Zhang et al, this conference (2020)
[10] M. Jakubowski et al, submitted to Phys. Rev. Lett. (2020)
[11] M. Jakubowski et al, this conference (2020)
[12] F. Reimold et al, this conference (2020)
[13] V. Perseo et al, Nuclear Fusion 59 124003 (2019)
[14] V. Perseo et al, this conference (2020)
[15] T. Barbui et al, JINST 14 C07014 (2019)
[16] C. Killer et al, Plasma Phys. Control. Fusion 61 125014 (2019)
[17] C. Killer et al, this conference (2020)
[18] T. Sunn Pedersen et al, Plasma Phys. Control. Fusion 61 014035 (2019)
[19] C. Beidler et al, in preparation (2020)
[20] H. Yamada et al Nuclear Fusion 45 1684 (2005)
[21] M. Beurskens et al, this conference (2020)
[22] J. A. Alcusón et al, Plasma Phys. Control. Fusion 62 035005 (2020)
[23] A. Langenberg et al, this conference (2020)
[24] K. Ida et al, Phys. Rev. Lett. 124, 025002 (2020)
[25] J. Geiger et al, this conference (2020)
[26] Y. Feng et al, this conference (2020)
[27] G. Fuchert et al, this conference (2020)
[28] K. Aleynikova et al, this conference (2020)
[29] H. Laqua et al, this conference (2020)
[30] S. Lazerson et al, this conference (2020)
[31] A. Dinklage et al, this conference (2020)Speaker: Prof. Thomas Sunn Pedersen (Max Planck Institute for Plasma Physics) -
Experiments on ST40 towards burning plasma conditions
M. P. Gryaznevich for TE.Ltd team
Tokamak Energy Ltd, 173 Brook Drive, Milton Park, Abingdon, OX14 4SD, UK
e-mail: mikhail.gryaznevich@tokamakenergy.co.ukSpherical Tokamak (ST) path to Fusion has been proposed in R Stambaugh et al, Fus. Tech. 33 (1998) 1, and experiments on STs have already demonstrated feasibility of this approach. Advances in High Temperature Superconductor (HTS) technology (M Gryaznevich et al, Fus. Eng. & Design 88 (2013) 1593) allows significant increase in the Toroidal Field (TF) which was found to improve confinement in STs. The combination of the high beta, which has been achieved in STs, and high TF that can be produced by HTS TF magnets, opens a path to lower-volume fusion reactors, in accordance with the fusion power scaling ~ beta^2Bt^4V. High field spherical tokamak ST40 (design parameters: R=0.4-0.6m, R/a=1.6-1.8, Ipl=2MA, Bt=3T, k=2.5, pulse~1-2sec, 2MW NBI, 2MW ECRH/EBW, DD and DT operations) is the first prototype on this path and is now operating, Fig.1.
Plasma current > 0.5MA at 2T TF, electron and ion temperatures in a several-keV range produced using merging-compression formation, solenoid-assisted ramp-up and 1MW of 25kV NBH, and densities up to 2x10^20m^-3 have been achieved in the first experimental campaigns in 2018-2020. At the flat-top, measured Ti increases with TF, in agreement with observations on other STs. However, on ST40, at TF > 1 – 1.1T we observe sharp increase in Te, Ti and W(EFIT), Fig.2.
TF Cu magnet in ST40 and some PF coils are LN2 cooled and research is on-going on development of full-HTS magnets. HTS prototype magnet with 24.4 T (at 21 deg K) has been built, Fig.3, and now we are planning to increase the field. LN2 cooling of present Cu magnets, installation of the second 1MW 50kV beam and upgrades of power supplies are on-going and will allow increase of TF to 3T and the pulse duration from ~0.3sec (at present) to 1-2sec.
Experiments are carried out to study transport properties in ST at higher TF, higher heating power (up to 4MW), low collisionality, in aim to bring plasma parameters close to burning conditions. Transport simulations with ASTRA, NUBEAM and TSC codes have been performed to model ST40 parameters and to support the physics basis of the compact high field ST path to Fusion. We show that high confinement regimes with just collisional (neoclassical) transport can be expected even when only ohmically heated. In an auxiliary heating regime, we find a hot ion mode with Ti in the 10keV range to be achievable in ST40 with as low as 1MW of absorbed power, Fig.4.
Issues connected with specific features of the high field ST are discussed, i.e. limitations of applicability of confinement scalings for prediction of performance of ST40. However, we show that if the performance achieved on other spherical tokamaks can be extended to ST40 conditions, up to 1 MW of Fusion power can be expected in DT operations. Studies of fast ions and alpha particle transport, heating and current drive, torque deposition and momentum transport have been performed using ASCOT, NUBEAM, Monte Carlo code NFREYA and the Fokker - Planck code NFIFPC. Different NBI energies and launch geometries have been studied and optimised. The confinement of thermal alphas in ST40 3T/2MA scenario is studied with full orbit following (which is necessary because of the large, compared to the plasma size, alpha particle gyro radius). The first orbit losses are seen to be above 60% even in the high-performance scenario illustrating that the alpha confinement in a small device is very difficult even at the highest available fields and plasma currents. However, DT experiments on ST40 will provide useful information for verification of such simulations.
We are intensively working on the design of our next tokamak, ST-F1, with plasma volume ~0.5 of JET. This device is aimed to demonstrate Q>3, will have HTS magnets, tritium blanket, and the goal is to build it by 2025.
Speaker: Mr Mikhail Gryaznevich -
Operating a full tungsten actively cooled tokamak:
overview of WEST first phase of operation
J. Bucalossi and the WEST Team (http://west.cea.fr/WESTteam)
CEA, IRFM, F-13108 St-Paul-Lez-Durance, France.E-mail: jerome.bucalossi@cea.fr
WEST is a MA class superconducting, actively cooled, full tungsten (W) tokamak. Equipped with two up-down symmetric divertors, it operates at 3.7T, up to 1MA, with a plasma volume of 15 m3 and an aspect ratio between 5 and 6. CW RF power is installed: up to 9 MW of ICRH power and 7 MW of LHCD.
In support of ITER operation and DEMO conceptual activities, WEST aims at power exhaust studies in long and steady-state pulses, in various divertor configurations (LSN, USN, DN), in a full W environment. The lower divertor, partially made of ITER-grade Plasma Facing Units (PFU) complemented with inertially cooled W-coated elements in phase 1 (2017-2019, see Fig. 1), will be replaced by a complete ITER-grade divertor in phase 2, starting autumn 2020. This paper reports on the main findings from WEST phase 1, in terms of operational domain, plasma performance achieved, and first tests of the ITER grade PFU.
Initial phase of operation was hindered by the production of runaway electron beams, when the ohmic current failed to rise quickly enough. Interestingly, start-up runaway electrons have been avoided by reducing the prefill pressure. Severe damages on PFC were observed [Diez] and one runaway beam impact induced radiation even quenched one superconducting coil [Reux].
Additionally to 200°C baking and glow discharge cleaning, boronizations have been performed in the second campaign leading to long lasting improved breakdown conditions and higher density operational domain.
Apart from the few post-boronization pulses, the fraction of radiated power in LHCD, ICRH or LHCD+ICRH pulses remains high, around 50%. Tungsten is, in most cases, the major radiating species [Goniche]. Remarkably, in presence of W antenna limiters, the fraction of radiated power using LHCD or ICRH is similar [Colas], as long as the ICRH coupling conditions are optimized [Hillairet].
In the ohmic phase, W radiation can lead to central cooling hence deleterious (2,1) MHD modes [Maget]. Nitrogen injection during this early phase, by increasing the edge resistivity, leads to more peaked electron temperature, hence reduced MHD [Manas], allowing for higher performance of the RF-heated phase. Up to 9.2 MW of combined ICRH and LHCD power has been achieved and up ~5MW/1s separately [Hillairet, Liang].
In L-mode, the stored energy, WMHD, increases according to the ITER96 L-mode scaling law up to 350kJ. L-H transitions are observed after fresh boronization, when combining 4MW of LHCD with 1MW of ICRH [Goniche], for a power crossing the separatrix of the order of the Martin 2008 scaling law [Martin J. Phys. Conf. Ser. 2008]. It results in a significant increase of the particle confinement time (30% increase of plasma density with gas injection turned off). The Doppler reflectometry ExB velocity well gets deeper, reaching -5km/s [Vermare]. But, in most cases, the plasma radiation increases leading to an oscillatory regime.
On the actively cooled upper W divertor, long pulses lasting up to 55 s have been routinely achieved (see Fig. 2), with a loop voltage down to 90 mV [Goniche]. In these L mode, electron heated, torque free plasmas, no W accumulation is reported despite peaked density profiles attributed to dominant TEM turbulence [Manas].
The heat flux level and pattern on the lower and upper divertors have been characterized thanks to embedded thermal measurements, IR and flush-mounted Langmuir Probes. The maximum heat flux currently reported on the W-coated graphite components is slightly above 5 MW.m-2 [Gaspar]. This was obtained with a conducted power on the divertor of ~2MW in two different scenarios: 1/ with combined LH and ICRH power after boronization at low X-point height (dX = 40 mm), 2/ with LH power only and high X-point height (dX = 120 mm).
In SOLEDGE2D simulations, target temperature profiles measured by Langmuir probe, as well as the radiated power measured by bolometry, were well reproduced with 3% of Oxygen as effective medium Z charge [Ciraolo]. The simulated asymmetry of O between the inner and outer targets is in qualitative agreement with UV measurements. The force balance analysis shows that friction dominates over thermal gradient forces at the inner target, while, at the outer target, the repelling thermal gradient forces dominate.
On the ITER-grade PFU, cracking and local melting have been observed for misaligned PFU. In addition, optical hot spots, which have been predicted to occur in ITER at the projection of the toroidal gaps on the subsequent PFU, have been observed experimentally, even for PFU aligned within specifications [Diez].
Finally, a He campaign has been run to investigate interactions between He plasmas and W PFC, in particular the formation of W fuzz [Tsitrone, Pegourie, Douai]. More than a hundred of 20-30 s pulses were repetitively performed in LSN. The conditions for W fuzz formation have been reached in the outer strike point area on the inertial W divertor (Einc > 20 eV, fluence > 1e24 He.m-2, Tsurf > 700 °C). Articulated Inspection Arm inspections before and after the He campaign have shown no macroscopic sign of surface modification. Post mortem analysis of the W components is ongoing to characterize the He induced nanostructures formed.
WEST phase 2 will start in autumn 2020, to address long pulse / high fluence operation on the newly manufactured ITER-grade actively cooled divertor, up to 10 MW/1000 s.
References:
[Diez], [Reux], [Goniche], [Colas], [Hillairet], [Maget], [Manas], [Liang], [Gaspar], [Ciraolo], [Tsitrone], [Pegourié], [Douai], this conference. [Martin 2008] J. Phys.: Conf. Ser. 123 012033Speaker: Jerome Bucalossi (CEA) -
Spherical tokamak (ST) research in Japan 1 is being conducted as a nationally coordinated program of university-scale ST devices under the ST Research Coordination Subcommittee organized by National Institute for Fusion Science (NIFS). The roles of university ST research include: (1) unique and challenging research through creativity and innovation which might be considered too risky for large ST devices, (2) establishment of the scientific basis for achieving ultra-high beta and ultra-long pulse (Fig. 1),
![Japanese ST research strategy [1].](https://workshop.nifs.ac.jp/fec2020/image/101-Takase-image-fig_1.jpg)
(3) contribution to the scientific basis for practical and economically competitive fusion power, complementing the mainline tokamak research (JT-60SA, ITER, etc.), and (4) development and training of a future generation of world-leading tokamak scientists. Specific research topics include: (a) development of start-up, current drive, and control techniques without the use of the central solenoid (CS), (b) formation and sustainment of very high beta plasmas, and (c) demonstration of steady-state operation and the study of steady-state issues such as heat and particle control, divertor physics, and plasma-wall interaction.
(a)-1. Plasma current ($I_p$) start-up by RF waves: Electron cyclotron wave (ECW) at 2.45 GHz and 5 GHz are used to excite the electron Bernstein wave (EBW) via O-X-B mode conversion on LATE (Kyoto U.). Highly overdense ST plasmas (up to 7 times the plasma cutoff density) are formed when the fundamental EC resonance layer is located in the plasma core, and EBW is excited in the 1st frequency band ($\omega_{ce} < \omega < 2\omega_{ce}$). Whereas EBW in the 1st frequency band heats the bulk electrons, EBW in the 2nd frequency band is absorbed by high-energy electrons and drives $I_p$. Intermittent plasma ejections across the plasma boundary synchronized with poloidal field decrement were observed in highly overdense plasmas. Oscillations in the Alfven frequency range and potential increase were observed, suggesting the loss of high-energy electrons. The 28 GHz RF injection system on QUEST (Kyushu U.) can regulate wave polarization and parallel index of refraction ($N_{||}$) with a beam radius focused down to 50 mm. $I_p > 100\ kA$ was achieved by injecting X-mode with $N_{||} = 0.78$, assisted by poloidal field induction. The RF power is likely absorbed by energetic electrons. Electron temperature of up to 0.5 keV was obtained by injecting X-mode with $N_{||} = 0.1$, indicating that effective bulk heating is possible. In $I_p$ start-up experiments using the lower hybrid wave (LHW) on TST-2 (U. Tokyo), top-launch was found to be more efficient for Ip ramp-up than outboard-launch. A 2-dimensional phase space model explaining X-ray emission shows that LHW driven radial transport is the dominant loss mechanism of fast electrons, and higher density is preferable in the present situation [2]. This is the first clear demonstration of RF driven electron transport.
(a)-2. CS-less $I_p$ start-up by non-RF methods: In transient CHI, magnetic reconnection plays an important role in the formation of closed flux surfaces during $I_p$ start-up. Plasmoid-driven reconnection following the tearing instability of the elongated current sheet and associated ion heating in the presence of the toroidal guide field were investigated on HIST (U. Hyogo) [3]. Several small-scale plasmoids generated during the injection phase merge with each other to form one or two large-scale closed flux surfaces during the decay phase. Transient CHI is also investigated on QUEST (US-JA collaboration). Investigation of synergistic effects of electron beam injection and EBW current drive in overdense plasmas has begun on LATE [4].
(a)-3. Optimization of inductive plasma start-up: In low voltage inductive $I_p$ start-up with ECW pre-ionization on TST-2, application of a weak vertical field with positive decay index during breakdown was found to be beneficial at low pre-fill pressure and high ECW power. Application of ECW power extended the low pressure limit for breakdown as well as the high pressure limit for burn-through. An MHD equilibrium model with fast electron orbits taken into account, and a model to simulate electron diffusion in both velocity space and real space are being developed.
(b) Access to high temperature and/or beta regime: Reconnection heating for direct access to burning plasmas, being investigated on TS-3U, TS-4U, UTST (U. Tokyo), will be reported in Ref [5].
(c) Demonstration of steady state operation by high-temperature wall: A high-temperature wall plays an essential role in reducing wall-stored hydrogen and facilitates hydrogen recycling. A clear extension of pulse duration at the wall temperature of 473 K was observed on QUEST by water cooling, indicating that recycling can be controlled by wall temperature. During long duration discharges, a high concentration of neutral particles was achieved behind the bottom divertor plate [6].
Stabilization using helical field coils: Suppression of the oscillation and the outer displacement in the radial position was observed by applying the helical field to the tokamak plasma on TOKASTAR-2 (Nagoya U.), which is an ST-helical hybrid device equipped with parallelogram-shaped partial helical field coils [7].
ST research in Japan has produced many innovative results including (i) $I_p$ start-up by LHW (TST-2), EBW (LATE), ECW/EBW (QUEST), CHI (HIST, QUEST), electron beam (LATE); (ii) optimization of ECW-assisted inductive start-up and pre-ionization by AC operation of the Ohmic coil (TST-2); (iii) extension of ion heating by plasma merging (TS-3U, TS-4U, UTST); (iv) hydrogen recycling control with high-temperature wall (QUEST); and (v) radial position stabilization by superposed helical field (TOKASTAR-2).
1 Y. Takase et al., Nucl. Fusion 57, 102005 (2017).
[2] A. Ejiri et al., this conference; N. Tsujii et al., this conference.
[3] M. Nagata et al., this conference.
[4] H. Tanaka et al., this conference.
[5] Y. Ono et al., this conference.
[6] T. Onchi et al., this conference; K. Hanada et al., this conference.
[7] K. Yasuda et al., Plasma Fusion Res. 13, 3402072 (2018).Speaker: Yuichi Takase (University of Tokyo) -
The report provides an overview of the results obtained at the upgraded Globus-M2 spherical tokamak 1 since the last IAEA conference. The tokamak was designed to reach the toroidal magnetic field as high as BT =1 T and the plasma current Ip = 0.5 MA having a small plasma minor radius a = 0.22-0.23 m. Currently 80% of highest magnetic field and plasma current value are reached, so during the reported period the experiments were performed with the toroidal magnetic field up to 0.8 T and plasma current up to 0.4 MA. The plasma breakdown conditions were improved noticeably with regard to the Globus-M ones, 30% breakdown loop voltage decreasing was achieved. The discharge duration was increased due to higher central solenoid volt-second consumption. The plasma column magnetic configuration explored was the divertor lower null with the aspect ratio A = R/a = 1.5-1.6, triangularity up to δ~0.35 and elongation up to κ~2.2.
The first neutral beam heating experiments on Globus-M2 have demonstrated an increased efficiency, comparing with the Globus-M ones, at the same NBI parameters (deuterium beam with particle energy 28 keV and the heating power 0.8 MW). The electron and ion central plasma temperatures exceeded 1 keV at the central density as high as 1×10^20 m-3. The diamagnetically measured plasma thermal energy increased up to 10 kJ, which is nearly triple as high as in Globus-M (BT =0.4, Ip = 0.2 MA). NPA spectra demonstrating improved fast particle confinement are presented. The energy confinement time increased more than two times that is significantly higher than the IPB98(y,2) scaling predicts. The effect is due to the strong dependence of the energy confinement time on the toroidal magnetic field in accordance with the Globus-M experimental scaling that is found to be valid for a wider range of BT. The regression fit of the Globus-M/Globus-M2 data yields the following scaling for energy confinement time:
τE ~ Ip^(0.58)BT^(1.23)Pabs^(-0.66)ne^(0.63)
where Pabs is the absorbed heating power and ne is the line average density. The scaling confirms weak τE dependence on Ip that emphasizes the major role of BT on heat perpendicular transport in spherical tokamaks, Enhanced plasma parameters allowed us to obtain regimes with much lower collisionality. That make possible investigation of dependence of the normalized energy confinement time (BTτE) on collsionality (ν~ne/T^2) in the wide range of plasma collisionalities 0.018<ν< 0.23. This dependence turned out to be rather strong BTτE ~ ν^(-0.8) for a fixed values of safety factor q ~ BT/Ip, normalized ion gyroradius ρ ~ T^(0.5)/BT and parameter βT ~ W/BT^2. The power balance analysis carried out using ASTRA transport code indicates the reduction of both electron and ion heat diffusivity with collisionality decrease while the ion heat diffusivity remains near the neoclassical level.
Important results are related to non-inductive current drive. About 30% of the loop voltage drop was recorded during the NB injection, which indicates a noticeable amount of non-inductively (mainly bootstrap) driven current. For the first time in spherical tokamaks a non-inductively driven current was recorded during the launch of the electromagnetic waves of the lower hybrid (LH) range (2.45 GHz) with the help of toroidally oriented grill. The fraction of noninductively driven current has exceeded 30% in the discharge with the total current of 0.2 MA. The modelling results of the experimental data by means of the ASTRA transport code and Fast Ray Tracing Code incorporated to ASTRA 2 are presented.
Plasma scrape of layer (SOL) and divertor characteristics were investigated in new experimental conditions of enhanced magnetic field and plasma current. Heat and particle fluxes together with currents and potentials in SOL and divertor plate vicinity were measured with a divertor Langmuir probe array and movable Langmuir probe. The plasma parameters in SOL were also modelled with the fluid version of the SOLPS-ITER code. Currents and drifts were included in the simulations. Comparison of experimental and simulated heat flux power density decay length (λqt) in SOL with the well-known scalings is presented.
The study of Alfvén modes (АМ) was continued during the reported period. An increase in plasma parameters led to a change in the nature of AM and the expansion of their frequency spectrum (50–300 kHz). Together with the toroidal Alfvén eigenmodes (TAE), observed earlier on Globus-M, the so-called Alfvén cascades (AC or RSAE) were identified. Observation of ACs made it possible to apply the method of MHD spectroscopy to determine the evolution of qmin in a discharge. In experiments on current drive by the LH waves, modes with a frequency of about 1 MHz, excited by fast electrons, were detected. To study the spatial structure of AM, Doppler backscattering diagnostics was used [3] with application of a multi-channel microwave scheme. Using the neutral particle analyzer and a neutron detector, we studied the dependence of fast particle losses initiated by TAEs on the magnetic field and plasma current. It was shown that losses decrease significantly with increasing field and current, demonstrating dependence favorable for compact neutron sources.
Also presented are new diagnostics designed to fill in the missing data on plasma parameters and improve the quality of the simulation, such as: diagnostics Z eff, laser interferometer, charge-exchange resonance spectroscopy (CXRS), etc.
1. V.B. Minaev et al 2017 Nucl. Fusion 57 066047
2. A.D. Piliya, A.N. Saveliev, JET Joint Undertakin Abingdon, Oxfordshire, OX14 3EA, 1998
3. V.V. Bulanin et al 2019 Tech. Phys. Lett., v.45, 11 p.p. 1107-1110Speaker: Yury Petrov (Ioffe Institute) -
The ADITYA Upgrade (ADITYA-U) is a medium sized (R0 = 75 cm, a= 25 cm) tokamak having toroidal graphite limiter, configured to attain shaped-plasma operations with an open divertor in single and double-null configurations [1]. The foremost objective of ADITYA-U is to prepare the physics and the technological base for future larger tokamaks by expanding the ADITYA-U operating space and by performing dedicated experiments for validation of physics models. Since the 2018 IAEA-FEC Conference, ADITYA-U operations have been mainly devoted on realizing the plasma parameters close to the design parameters of circular plasmas in limiter configuration and also the initiation of shaped plasma operation. Emphasis has been given to novel experiments on runaway electrons (REs) and disruption control in the ADITYA-U [2]. Furthermore, experiments on radiative-improved modes using Ne, Ar gas injection, modulation of MHD modes [3] and edge turbulence using periodic fuel gas-puffs, density dependence of plasma toroidal rotation reversal [4], fuelling using SMBI etc.
For the typical discharges in ADITYA-U, in absence of any strong pre-ionization, the gas breakdown and successful plasma start-up is normally achieved with peak loop voltages of ~ 18 – 20 V (Electric field ~ 4.5 V/m). 42 GHz ECR [5] assisted low loop voltage (~10 – 12 V, Electric field ~2.1 V/m) start-up was successfully achieved with wave launched in fundamental O-mode from low field side. The toroidal magnetic field is ~ 1.4 T. Plasma discharges having plasma current ~ 170 kA, plasma duration ~ 330 ms, chord-averaged electron density ~ 2 – 6 x 10^19 m^-3 and central electron temperature ~ 300 – 500 eV has been achieved. The time evolution of typical high current, longer duration discharge of ADITYA-U is shown in Figure 1 and the overall progress of plasma current and duration enhancement during the year 2018-2019 are shown in Figure 2(a) and 2(b) respectively. Repeated cycles of vacuum vessel baking up to 135° C, followed by extensive wall conditioning using novel techniques [6] along with lithium coating [7] resulted in substantial reduction in partial pressures of various mass species and achievement of lower base vacuum of ~ 6 x 10^-9 Torr. Successful recovery of volt–sec along with adequate control of real-time horizontal plasma position and multiple gas puffing led to the achievement of longer discharge duration discharges in ADITYA-U.
_Fig.2(b)_RLTanna.jpg)
Over the last two years, significant progresses have been achieved in ADITYA-U experimental research including, 42GHz ECR assisted low loop voltage start-up and heating experiments, electromagnetically driven pellet impurity injector for injecting micron size particles at high velocity (~220m/s) to understand disruption mitigation, wall conditioning by using different techniques of lithium coating, formation of runaway beam and its avoidance, Neon and Argon impurity injection for radiative improved modes, suppression of electrostatic fluctuations using hydrogen gas puff and its correlation with hard X-rays, effect of positive edge electrode biasing on drift tearing modes and runaway transport and effect of SMBI with edge safety factor (qedge), toroidal rotation reversal threshold studies etc. Dependence of current quench time (CQT) on qedge during disruptions and relation of CQT on the prevailing MHD activities prior to disruption has been studied in detail. Discharges with low qedge showed a high CQT compared to those observed at high qedge. Exploring the RE formation mechanisms, controlled RE generation experiments have been carried out by lowering the plasma density and adjusting the vertical magnetic field as shown in Figure 3 (#33061).
Fast visible imaging video camera, used for 2D tangential viewing, captured images which showed features of the RE beam formation with high spatial and temporal resolution as shown in Figure 4. This RE beam generation in controlled fashion is very useful in studying the mitigation techniques of REs using different techniques as the prevention of such RE beam is of a vital importance in future tokamaks, especially in the ITER, because of its potential danger to the plasma facing components.
In another significant RE experiment carried out in ADITYA-U, correlated suppression of RE loss, evident from hard X-rays intensity measurements, has been observed with suppression of edge electrostatic fluctuations in discharges where magnetic fluctuation amplitudes are not sufficient for affecting REs. Multiple periodic gas puff, which are used for plasma fuelling, suppresses the electrostatic fluctuations in the floating potential as well as the density fluctuations (measured with Langmuir probes) in the edge region. Figure 5 shows the suppression of edge turbulence due to gas puff. This is also observed clearly in the spectrogram of density fluctuations. The hard X-rays flux is also seen to be highly correlated with the edge turbulence. The decrease in edge turbulence is accompanied by a decrease in HXR flux. This observation of correlated effect of electrostatic fluctuation on RE transport throws light on new mechanism of RE loss and may be exploited to design novel RE mitigation methods.
In another interesting experiment, the REs are confined by applying a voltage to an electrode placed at the edge region of the plasma prior to the disruption. Figure 6 shows the multiple pulses of biasing applied to an electrode placed ~ 2.5 cm inside the LCFS for the shot #33336. The last pulse of biasing coincides with the plasma disruption as shown in Figure 7, describes that in presence of biasing during the disruption, the HXR intensity as well as HXR flux persists for ~10 ms even after the termination of the plasma current. A possible mechanism of confinement is the Er x BΦ motion of REs after the disruption, where the Er is generated by electrode biasing.
For the first time, ADITYA-U has experimentally demonstrated the use of electromagnetically driven payloads for particle injection into the tokamak plasmas for disruption mitigation studies. The impurity particles reached the core of the plasma within ~ 1.25 ms and causes fast termination of plasma current and radiate the whole plasma energy in ~ < 2 ms. Furthermore, the preliminary experiments related to plasma shaping by charging the divertor coils during plasma current plateau, confinement improvement with ion cyclotron resonance (ICRH) assisted auxiliary heating and deuterium injection is undergoing and the results of the same will be presented. This paper summarizes the experimental research of ADITYA-U tokamak in the key areas of thermo-nuclear fusion over the last two years.References
[1] Tanna R. et al 2018 Plasma production and preliminary results from the ADITYA Upgrade tokamak Plasma Sci. Technol. 20 074002.
[2] Tanna R. et al 2019 Overview of operation and experiments in the ADITYA-U tokamak Nucl. Fusion 59 112006.
[3] Raj Harshita et al 2020 Effect of periodic gas-puff on drift-tearing modes in ADITYA/ADITYA-U tokamak discharges Nucl. Fusion 60 036012.
[4] Shukla G. et al 2019 Observations of toroidal plasma rotation reversal in the ADITYA-U tokamak Nucl. Fusion 59 106049.
[5] Shukla B.K. et al 2019 Commissioning of Electron Cyclotron Resonance Heating (ECRH) system on tokamak ADITYA-U Fusion Eng. Des. 146 2083 – 86.
[6] Jadeja K.A. et al 2019 Novel approach of pulsed-glow discharge wall conditioning in the ADITYA Upgrade tokamak Nucl. Fusion 59 086005.
[7] Jadeja K.A. et al Lithium Wall Conditioning Techniques in ADITYA-U Tokamak for Impurity and Fuel Control (In this conference).Speaker: RAKESH TANNA (Institute for Plasma Research) -
Using its unique flexibility and advanced plasma diagnostics, the TJ-II stellarator is contributing to the understanding and solution of critical challenges in fusion plasmas. Next, we highlight some of the most relevant recent results in the framework of its research programme.
Towards validation of gyrokinetic and neoclassical simulations. Aiming at the validation of the instability properties predicted by gyrokinetic (GK) simulations and of the electrostatic potential variations on the flux surface, φ1, calculated by neoclassical (NC) codes, dedicated experiments have been carried out in TJ-II for a systematic characterization of turbulence wavenumber spectra and perpendicular rotation velocity measured by Doppler Reflectometry (DR) at poloidally separated positions on the same flux surface [1]. Poloidal asymmetries in the intensity of the wave number spectrum that depend on plasma conditions have been characterized and compared with global linear GK simulations by the code EUTERPE. Model and experiment qualitatively agree in the radial dependence of the turbulence intensity, in the turbulence dispersion relation and in showing a poloidal asymmetry that depends on the magnetic configuration. Recent experiments exploring configurations with different magnetic ripple have shown a reduction in the turbulence asymmetry at configurations with reduced ripple. Besides, the influence of base ion mass has been investigated in hydrogen and deuterium plasmas. The ion mass in TJ-II plasmas does not affect the properties of the turbulence, neither the amplitude nor the spectral shape or the poloidal asymmetry. The lack of dependence of the turbulence spectrum on the ion mass is also found in GK simulations. Model validation will also benefit from the effort of verification of GK simulations in different computational domains [2] as well as from the application of the recently developed GK code stella to multispecies turbulent transport calculations in TJ-II [3].
Poloidal asymmetries in radial electric field, Er, are found that depend on the plasma collisionality. These results have been compared with the contribution to Er arising from −φ'1 as calculated with the NC version of the code EUTERPE. These results show variations in Er comparable in size to those found in the experiments, but there is a disagreement regarding the sign of the Er correction. Recent simulations performed with the newly developed NC code KNOSOS [4] show that the effect of kinetic electrons on φ1 has to be taken into account due to the strong Te dependence of the electron contribution to φ1 when the electrons are in the 1/ν regime [Fig. 1]. KNOSOS (KiNetic Orbit-averaging SOlver for Stellarators) is a freely available, open-source code that calculates neoclassical transport in low-collisionality plasmas of three-dimensional magnetic confinement devices by solving the radially local drift-kinetic and quasineutrality equations.
Experiments in TJ-II with cryogenic and TESPEL pellets show that post-injection particle radial redistributions can be understood qualitatively from neoclassical predictions while also providing a means to benchmark the HPI2 code [5]. TJ-II measurements provide 2-D maps of plasma potential and fluctuations to address the question of how hollow density profiles, created by pellets, affect turbulent transport. Fluctuations are stronger in the negative density gradient than in the positive one in consistency with TEM linear simulations [6].
Towards the validation of fast ion induced stabilization and the identification of Alfvén Eigenmode actuators. An ambitious research programme is in progress to investigate the relation between zonal structures and Alfvén eigenmodes (AE) and its role on the nonlinear dynamics of AEs and transport as well as to develop and demonstrate AE control strategies using ECRH and ECCD in TJ-II. The unique TJ-II experimental set-up using a dual HIBP has shown that, in some conditions, long range correlations (LRC) are detected both at the AE frequencies and low frequencies (<10 kHz) [Fig. 2]. LRC are observed in plasma potential fluctuations but not in density fluctuations as expected in zonal flow structures [Fig. 2]. It is an open question whether those zonal structures are directly driven by fast particle effects or/and are the consequence of plasma scenarios with reduced damping of zonal flows. Experiments in TJ-II have demonstrated the effectiveness of ECRH and ECCD actuators to modify AE activity [7].
Towards the characterization of the interaction between neoclassical and turbulent transport mechanisms. We have investigated the impact of Er on turbulence propagation and the coupling between the plasma edge and the scrape-off layer (SOL) during electron–ion root transitions where Er is changed in a controlled manner from positive to negative values. It is shown that Er does not only affects the radial turbulence correlation length but it is also capable of reducing the propagation of turbulence from the edge into the SOL. This result was obtained using a technique based on the transfer entropy, which quantifies the propagation of information [8]. These observations are highly relevant for the understanding of the mechanisms that determine the SOL width.
The interplay between of NC radial electric fields, Reynolds stress gradients and LRC has been investigated in different plasma scenarios in TJ-II. Turbulent driven acceleration alone cannot explain the dynamics of zonal flows whose radial width is affected by the isotope mass [9]. These results are in line with the expectation that the interplay between turbulent and neoclassical mechanisms is an important ingredient of the dynamics of edge zonal flows.
Power exhaust physics: liquid metals. Solid and liquid samples of Li/LiSn/Sn, in a Capillary Porous System (CPS) arrangement, have been exposed to the edge plasma [10]. A simple 1D model was applied to the data, allowing for the evaluation of the kinetic energy (Ek) of ejected atomic species while their residence time at the edge was determined by monitoring the ratio of first ion/neutral emission light intensities. A clear evolution of Ek with sample temperature was deduced for Li atoms, this being associated to the different relative contributions of sputtered/evaporated atoms. SnI emission into the plasma has also been measured with radial and toroidal resolution. The deduced mean free paths for the ejected Sn atoms under sputtering conditions (low T) imply unrealistic high energies if the bibliographic data for the ionization rate constant of Sn are assumed. For the LiSn case, Li as well as Sn emissions were simultaneously detected and analysed. Plasmas under cut-off (collapsing) conditions were also investigated to check for the sensitivity of the recorder line intensities and ionization rates to the edge electron temperature.
[1] T. Estrada et al., Nuclear Fusion 59, 076021 (2019)
[2] E. Sánchez et al. Gyrokinetic simulations in stellarators using different computational domains, 28th IAEA Nice 2020
[3] J. M. García-Regaña et al., Turbulent transport of impurities in 3D devices, 28th IAEA Conf. Nice 2020.
[4] J. L. Velasco et al.,KNOSOS, a fast neoclassical code for three-dimensional configurations, 28th IAEA Nice 2020
[5] K. McCarthy et al., Pellet studies in TJ-II, 28th IAEA Nice 2020
[6] A. Melnikov et al., 2-D mapping of fluctuations and plasma profiles in TJ-II, 28th IAEA Nice 2020.
[7] A. Cappa et al., AE control strategies in TJ-II, 28th IAEA Conf., Nice 2020.
[8] G. Grenfell et al., Nucl. Fusion 60 014001 (2020)
[9] R. Gerrú et al., Nucl. Fusion 59 106054 (2019)
[10] F. Tabarés et al., Liquid metal studies in TJ-II, 28th IAEA Conf., Nice 2020Speaker: Prof. Carlos Hidalgo (CIEMAT) -
Achieving ignition and high fusion yield in the laboratory is a central goal of the U.S. Inertial Confinement Fusion (ICF) Program. Three major and credible approaches are currently being pursued: laser indirect-drive (LID), laser direct-drive (LDD), and magnetic direct-drive (MDD). While the three approaches use very different means for driving a spherical or cylindrical implosion that can compress and heat a mass of deuterium-tritium (DT) fuel (by laser-generated radiation drive, direct laser irradiation, or direct magnetic acceleration, respectively), they share many common challenges, including how to efficiently couple the kinetic energy of the implosion to internal energy of the fuel at stagnation, and how to assemble the fuel at the necessary conditions, in pressure, temperature, and areal density to initiate self-heating and ignition.
Significant progress has been made in each of these approaches in recent ICF experiments on the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory, the OMEGA laser facility at the Laboratory for Laser Energetics, and the Z pulsed power facility at Sandia National Laboratories. This includes progress in both advancing the absolute fusion output and, as importantly, improving our knowledge and understanding of the implosion behavior and causes of deviations from ideal or theoretical performance.
In this overview, we will review some of the major results from each facility, with a particular focus on recent advances in diagnostic measurement techniques and analysis that have improved our understanding of the states of the assembled fuel and confining shell at stagnation. Particular emphasis is placed on recent advances in determining the 3D spatial morphology and thermodynamic properties of the fusion fuel, including pressure, temperature, areal density, and mix, that determine the degree of alpha-particle production, confinement, and self-heating, and proximity to ignition.
As an example, Fig. 1 shows the hot spot temperatures and areal densities inferred from some of the principal LID experimental campaigns performed on the NIF since 2011. Improved performance can be observed as the implosion design evolved from the original low-adiabat† (a~1.6) CH ablator design (2011-2013), to a mid-adiabat (a~2.3) CH design (2013-2015), and then to a mid-adiabat (a~2.7) high-density-carbon (HDC) ablator design (2016-2019) [1-3]. The principal gains resulted from improved hydro-instability at higher adiabat, and improved energy coupling and implosion symmetry with HDC designs. As of today, the highest performing HDC implosions have achieved hot spot temperatures of ~4.5 keV, hot spot areal densities of ~0.3 g/cm2, hot spot pressures of ~350 Gbar, and fusion energy output of >50 kJ. At these conditions, the self-heating of the hot spot by alpha-particle deposition is estimated to be amplifying the total fusion output by a factor of ~2.5-3x. As can be seen in Fig. 1, these conditions are now quite close to the static self-heating boundary, dT/dt>0, where for a static hot spot at the time of peak compression, the time derivative of temperature is positive due to alpha-particle heating exceeding energy losses from radiation and thermal conduction. Whilst reaching this boundary will be a very noteworthy physics achievement, it is not a sufficient condition for ignition. For ignition, the alpha-heating rate must exceed all losses including mechanical work from expansion of the hot spot after peak compression. This condition, equivalent to the requirement d2T/dt2>0, depends on the shell confinement and necessitates on the order of a ~1 keV higher hot spot temperature [4,5].
In the LDD approach, remarkable progress has been made using a data-driven statistical mapping model to optimize input laser and target parameters that has led to a 3x increase in neutron yield over the past two years on OMEGA [6]. The best-performing spherical direct drive (SDD) implosion has a hydrodynamically-scaled predicted fusion yield in the 0.5 MJ range at NIF laser energy, approaching the burning plasma regime. Polar direct drive (PDD) implosions and laser-plasma coupling research are performed on OMEGA and the NIF for the LDD approach [7,8]. In the MDD approach, recent magnetized liner inertial fusion (MagLIF) experiments on Z have produced record yields and stagnation parameters through steady enhancements in the target initial magnetization, laser preheat energy delivered to the fuel, and electrical current delivered to the imploding shells [9]. In addition to record fusion performance, MagLIF implosions have also demonstrated significant potential for DT alpha-particle confinement and stopping via the strong magnetic fields entrained in the fuel [10,11], a necessary requirement for self-heating and ignition.
This work was performed under the auspices of Lawrence Livermore National Security, LLC (LLNS) under Contract DE-AC52-07NA27344, the Sandia National Laboratories, a multi-mission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. DOE NNSA under contract DE-NA0003525, and upon work supported by the U.S. DOE NNSA under Award Number DE-NA0003856, the University of Rochester, and the New York State Energy Research and Development Authority.
† The adiabat is the ratio of the pressure of the DT fuel to the Fermi degenerate pressure.
References
1. J. Lindl et al., Phys. Plasmas 21, 020501 (2014)
2. O. A. Hurricane et al., Nature 506, 343 (2014)
3. S. Le Pape et al., Phys. Rev. Lett. 120, 245003 (2018)
4. P. T. Springer et al., Nucl. Fusion 59, 032009 (2019)
5. P. K. Patel et al., Phys. Plasmas (submitted, 2020)
6. V. Gopalaswamy et al., Nature 565, 581 (2019)
7. F. J. Marshall et al., Phys. Plasmas 23, 012711 (2016)
8. P. B. Radha et al., Phys. Plasmas 23, 056305 (2016)
9. M. R. Gomez et al., manuscript in preparation (2020)
10. P. F. Schmit et al., Phys. Rev. Lett. 113, 155004 (2014)
11. P. F. Knapp et al., Phys. Plasmas 22, 056312 (2015)Speaker: Pravesh Patel (Lawrence Livermore National Laboratory) -
The mission of the spherical tokamak NSTX-U is to advance the physics basis and technical solutions required for optimizing the configuration of next-step tokamak fusion devices, and to advance the development of the ST concept towards a compact, low-cost Pilot Plant 1. NSTX-U will operate at up to 2 MA and 1 T with up to 10 MW of Neutral Beam Injection (NBI) power for 5 seconds and 4 MW of High Harmonic Fast Wave (HHFW) power. NSTX-U has three main objectives: to explore confinement and stability at low aspect ratio and high beta at low collisionality, to develop the physics understanding and control tools to ramp-up and sustain high performance plasmas in a fully-non-inductive regime, and to develop and evaluate conventional and innovative power and particle handling techniques to optimize plasma exhaust in high performance scenarios. Following the initial 2016 NSTX-U run campaign, analysis has continued during NSTX-U Recovery to address physics issues to develop understanding and capabilities that, once operation commences, will aid in achieving these three objectives.
Stability and 3D physics: The resistive DCON model for calculating tearing mode stability ($\Delta'$) has been developed, benchmarked against extended-MHD (M3D-C1) simulations, and used to identify regions in $q_{95}-\beta_N$ that are simultaneously stable to both 2/1 tearing modes (Fig.1) and n=1 ideal kink.
M3D-C1 has been extended to predict Te and Ti independently in the presence of impurities during a thermal quench 2. The simulations show a contraction of the current channel that is sensitive to temperature, and a fast stochastization of the B-field, highlighting the difficulty in cooling the plasma while avoiding a thermal quench using impurity injection. Continued analysis on 3D error fields have shown that misalignment of the toroidal field (TF) was the largest source of the error field on NSTX-U during initial operation in 2016, and these calculations, guided by constraints on the field line pitch at the divertor plates in order to mitigate potentially high heat fluxes, were used to drive the engineering tolerances for TF shift and tilt for NSTX-U Recovery [3]. Additional calculations show that misalignments in other PF coils lead to extended divertor footprints but which are contained within the divertor region designed to handle high heat fluxes [4]. Energetic particle (EP) physics: The phase-space-resolved reduced EP transport kick model has been extended to include non-Alfvénic low-frequency perturbations, reproducing observations of (i) large fast ion losses due to synergistic effects when TAE and fishbones/kink instabilities occur simultaneously, and (ii) enhanced fast ion loss due to NTMs when the island width exceeds a threshold [5]. The stability and scaling of global Alfven eigenmodes (GAEs), previously correlated with central Te flattening in NSTX, has been predicted using hybrid MHD/kinetic-fast-ion simulations [6] and also newly derived analytic instability conditions [7], revealing a previously unidentified instability regime necessary to explain observed GAE excitation and stabilization. The chirping and avalanche behavior of Alfven eigemodes that can influence fast-ion losses in NSTX-U (and burning plasma $\alpha$ losses) has been predicted using a guiding center code with a delta-f formalism [8]. New analysis from NSTX and NSTX-U data has provided a detailed picture of ion cyclotron emission (ICE), being considered as a possible diagnostic of confined $\alpha$’s in burning plasma experiments [9]. A self-consistent resonance-broadened quasi-linear (QL) model has been developed for relaxation of fast ion distribution function by Alfvenic modes [10].
Transport physics: Analysis in moderate $\beta$ NSTX scenarios using gyrokinetic simulations coupled with a novel “synthetic diagnostic” identify conditions where both electron thermal transport and turbulence measured by high-k microwave scattering are explained entirely by short-wavelength electron-scale ETG turbulence [11]. Gyrokinetic analysis in high-$\beta_{pol}$ scenarios envisioned for high non-inductive fraction operation indicates the deep-core profiles (with relatively flat Te) sit very near KBM (or EPM) limits when including only thermal ions (or thermal + fast ion species), suggesting core profiles may ultimately be constrained by $\nabla p$-limited ballooning modes. Analysis of enhanced Pedestal (EP) H-modes demonstrates that this high confinement (H98$\leq$1.8), wide-pedestal ($\Delta\psi_{N,ped}\leq0.4$), ELM-free regime is accessed at low edge ion collisionality, e.g. via reduced wall recycling with lithium wall coatings. While MTM, TEM and ETG instabilities predicted in this region may account for electron thermal losses, it is hypothesized that peeling-ballooning and near-marginal KBM instabilities (predicted via MHD and gyrokinetic simulations, respectively, Fig. 2) may enhance edge particle transport during the evolution to the EPH phase that helps to sustain the low-density, low-collisionality state.
RF physics: A new 2D full wave code (FW2D) has been updated to predict the sensitivity of SOL losses to variations in the realistic boundary shape for high harmonic fast wave (HHFW) heating [12]. Losses are further predicted to be minimized when the SOL density is near the critical density for fast wave cutoff, as found in experiments. Calculations of HHFW deposition in the presence of NBI using AORSA identify a competition between electron and fast ions absorption [13]. Additional simulations show that a sufficient concentration of H+ minority species could open up new HHFW heating scenarios in NSTX-U without NBI.
Scenarios & Control: A unified, physics-based reduced model for direct inductive startup, which computes the timing of plasma initiation and duration needed for plasma density buildup, has been developed for both NSTX-U and MAST-U [14]. Physics-based control-oriented models have been used to develop advanced-control approaches to determine neutral-beam current-drive requirements and evolutions to track prescribed current profiles in closed-loop [15], as well as to iteratively optimize the access to high-performance scenarios [16]. Aided by this is the development of a neural network-based description of beam heating and current drive profiles, well suited for rapid calculations and real-time application [17]. Development of a physics-based algorithm for closed loop feedback control of snowflake divertor configurations will provide real-time tracking and control capabilities [18].
The NSTX-U Recovery planning is ongoing, with NSTX-U targeting resumption of operations in 2021 [19].
This work was supported by US Department of Energy Contract No. DE-AC02- 09CH11466.
References
1 Menard, J.E., et al., this conference.
2 Ferraro, N.M., et al., Nucl. Fusion 59 016001 (2019).
[3] Ferraro, N.M., et al., Nucl. Fusion 59 086021 (2019).
[4] Munaretto, S., et al., Nucl. Fusion 59 076039 (2019).
[5] Podesta, M, et al., this conference.
[6] Belova, E., et al., this conference.
[7] J. B. Lestz et al., Physics of Plasmas 27, 022512 & 022513 (2020);
[8] White, R., et al., this conference.
[9] Fredrickson, E., et al., this conference.
[10] Gorelenkov, et al., this conference.
[11] Ruiz Ruiz, J., et al., Plasma Phys. Cont. Fusion 61 115015 (2019)
[12] Kim, E.-H., et al., Phys. Plasmas 26 062501 (2019)
[13] Bertelli, N., et al., Nucl. Fusion 59 086006 (2019)
[14] Battaglia, D.J., et al., Nucl. Fusion 59 126016 (2019)
[15] Ilhan, Z.O., et al., Fusion Eng. and Design 146 555 (2019)
[16] Wehner, W.P., et al., Fusion Eng. and Design 146 547 (2019)
[17] Boyer, M.D., et al., this conference.
[18] Vail, P.J., et al., Plasma Phys. Cont. Fusion 61 035005 (2019)
[19] Gerhardt, S.P., et al., this conference.Speaker: Walter Guttenfelder (Princeton Plasma Physics Laboratory) -
In the recent deuterium experiment on the Large Helical Device (LHD), we have succeeded to expand the temperature domain to higher region both in electron and ion temperatures as shown by the red region in Fig.1. We found a clear isotope effect in the formation of Internal Transport Barrier (ITB) in high temperature plasmas. In the deuterium plasmas, we have also succeeded to realize the formation of the Edge Transport Barrier (ETB) and the divertor detachment, simultaneously. It is found that the Resonant Magnetic Perturbation (RMP) has an important role in the simultaneous formation of the ETB and the detachment. A new technique to measure the hydrogen isotope fraction was developed in LHD in order to investigate the behavior of isotope mixed fuel ions. The technique revealed that the non-mixing state and the mixing state of hydrogen isotopes can be realized in plasmas.
One of the objective of deuterium experiment is the achievement of the reactor relevant high temperature plasmas in order to investigate the problems arising in the helical type fusion reactor. As shown in Fig.1, the ion temperature of 10keV was achieved at the electron temperature of around 4keV. This was reported at the last IAEA-FEC(FEC2018). The operational domain was extended to the direction of higher electron temperature range. The electron temperature of 6.6keV was achieved keeping the ion temperature of 10keV. For lower ion temperature plasmas (Ti0=~7keV), the electron temperature range was extended over ~12keV. It is found that the extension in higher electron temperature region of high ion temperature plasmas is an effective way to suppress the Energetic particle driven InterChange mode (EIC), which often prevents the access of high ion temperature plasmas in LHD. This is preferable feature for the future reactor scenario development 1.
The isotope effect is a long underlying mystery in plasma physics because the most of experimental observations in tokamak plasmas show favorable effect by the ion mass on the energy confinement time (tau_E) while the theoretical prediction based on the gyro-Bohm model shows unfavorable effect by the mass, i.e., tau_E~ M^-1/2. A series of dedicated experiment using dimensionally similar L-mode hydrogen and deuterium plasmas in LHD showed that the energy confinement time scaling for L-mode plasmas had a non-significant dependence on mass (M^0.01) 2. On the other hand, a clear isotope effect was observed for the formation of Internal Transport Barrier (ITB) in LHD as shown in Fig.2. Here, the profile gain factor G1.0 is defined as
and TLref is the expected temperature profile for L-mode plasmas with same heating condition 3. As shown in the figure, the ITB intensity, i.e., G1.0, is clearly larger for deuterium at the density range below 2x10^19 [m^-3].
The realization of both divertor heat load mitigation and the good core confinement property is an important issue in developing future reactor scenarios. For divertor detached deuterium plasmas which are realized by the RMP application, we have succeeded to improve the confinement of 38% with a formation of ETB. Thus, we call this improved confinement mode as the RMP induced H-mode 4. As shown in Fig.3, a steep pressure gradient was formed in deuterium plasmas at the edge region inside the m/n=1/1 island which was produced by the RMP. This formation of the steep edge pressure gradient with the detachment has never been observed in hydrogen plasmas. The thermal transport analysis based on TASK3D-code shows significant reduction in the thermal diffusivity at the edge region (r/a=~0.8) which indicates the formation of ETB.
In the future fusion reactor, the deuterium (D) and tritium (T) ions co-exist in plasmas and are assumed to be uniformly mixed with the ratio of D/T=1. On the other hand, it is not clear whether this assumption is always valid. To clarify the assumption, experiments to investigate the isotope mixing state were performed using Deuterium and Hydrogen mixture plasmas. It was found that the non-mixing state of isotopes can be realized when the centrally fuelled ion species are different from the ion species fuelled at the edge 5. Figure 4 shows a typical example of the non-mixing state where the tangential NBI fuels Hydrogen ions in the central region while the Deuterium is supplied at the peripheral region by a wall recycling (black symbols and curves). It was found this state can be changed to mixing state by an injection of ice pellet of hydrogen isotopes, i.e. either of Hydrogen or Deuterium. It was also found that a formation of hollow density profile due to the pellet injection plays an important role for changing the state.
1 H. Takahashi et al., this conference
2 H. Yamada et al., Phys. Rev. Lett. 123 (2019) 185001
3 T. Kobayashi et al., this conference
4 M. Kobayashi et al., this conference
5 K. Ida et al., this conferenceSpeaker: Prof. Masaki Osakabe (NIFS)
-
-
-
Operating a full tungsten actively cooled tokamak:
overview of WEST first phase of operation
J. Bucalossi and the WEST Team (http://west.cea.fr/WESTteam)
CEA, IRFM, F-13108 St-Paul-Lez-Durance, France.E-mail: jerome.bucalossi@cea.fr
WEST is a MA class superconducting, actively cooled, full tungsten (W) tokamak. Equipped with two up-down symmetric divertors, it operates at 3.7T, up to 1MA, with a plasma volume of 15 m3 and an aspect ratio between 5 and 6. CW RF power is installed: up to 9 MW of ICRH power and 7 MW of LHCD.
In support of ITER operation and DEMO conceptual activities, WEST aims at power exhaust studies in long and steady-state pulses, in various divertor configurations (LSN, USN, DN), in a full W environment. The lower divertor, partially made of ITER-grade Plasma Facing Units (PFU) complemented with inertially cooled W-coated elements in phase 1 (2017-2019, see Fig. 1), will be replaced by a complete ITER-grade divertor in phase 2, starting autumn 2020. This paper reports on the main findings from WEST phase 1, in terms of operational domain, plasma performance achieved, and first tests of the ITER grade PFU.Initial phase of operation was hindered by the production of runaway electron beams, when the ohmic current failed to rise quickly enough. Interestingly, start-up runaway electrons have been avoided by reducing the prefill pressure. Severe damages on PFC were observed [Diez] and one runaway beam impact induced radiation even quenched one superconducting coil [Reux].
Additionally to 200°C baking and glow discharge cleaning, boronizations have been performed in the second campaign leading to long lasting improved breakdown conditions and higher density operational domain.
Apart from the few post-boronization pulses, the fraction of radiated power in LHCD, ICRH or LHCD+ICRH pulses remains high, around 50%. Tungsten is, in most cases, the major radiating species [Goniche]. Remarkably, in presence of W antenna limiters, the fraction of radiated power using LHCD or ICRH is similar [Colas], as long as the ICRH coupling conditions are optimized [Hillairet].
In the ohmic phase, W radiation can lead to central cooling hence deleterious (2,1) MHD modes [Maget]. Nitrogen injection during this early phase, by increasing the edge resistivity, leads to more peaked electron temperature, hence reduced MHD [Manas], allowing for higher performance of the RF-heated phase. Up to 9.2 MW of combined ICRH and LHCD power has been achieved and up ~5MW/1s separately [Hillairet, Liang].
In L-mode, the stored energy, WMHD, increases according to the ITER96 L-mode scaling law up to 350kJ. L-H transitions are observed after fresh boronization, when combining 4MW of LHCD with 1MW of ICRH [Goniche], for a power crossing the separatrix of the order of the Martin 2008 scaling law [Martin J. Phys. Conf. Ser. 2008]. It results in a significant increase of the particle confinement time (30% increase of plasma density with gas injection turned off). The Doppler reflectometry ExB velocity well gets deeper, reaching -5km/s [Vermare]. But, in most cases, the plasma radiation increases leading to an oscillatory regime.
On the actively cooled upper W divertor, long pulses lasting up to 55 s have been routinely achieved (see Fig. 2), with a loop voltage down to 90 mV [Goniche]. In these L mode, electron heated, torque free plasmas, no W accumulation is reported despite peaked density profiles attributed to dominant TEM turbulence [Manas].
The heat flux level and pattern on the lower and upper divertors have been characterized thanks to embedded thermal measurements, IR and flush-mounted Langmuir Probes. The maximum heat flux currently reported on the W-coated graphite components is slightly above 5 MW.m-2 [Gaspar]. This was obtained with a conducted power on the divertor of ~2MW in two different scenarios: 1/ with combined LH and ICRH power after boronization at low X-point height (dX = 40 mm), 2/ with LH power only and high X-point height (dX = 120 mm).
In SOLEDGE2D simulations, target temperature profiles measured by Langmuir probe, as well as the radiated power measured by bolometry, were well reproduced with 3% of Oxygen as effective medium Z charge [Ciraolo]. The simulated asymmetry of O between the inner and outer targets is in qualitative agreement with UV measurements. The force balance analysis shows that friction dominates over thermal gradient forces at the inner target, while, at the outer target, the repelling thermal gradient forces dominate.
On the ITER-grade PFU, cracking and local melting have been observed for misaligned PFU. In addition, optical hot spots, which have been predicted to occur in ITER at the projection of the toroidal gaps on the subsequent PFU, have been observed experimentally, even for PFU aligned within specifications [Diez].
Finally, a He campaign has been run to investigate interactions between He plasmas and W PFC, in particular the formation of W fuzz [Tsitrone, Pegourie, Douai]. More than a hundred of 20-30 s pulses were repetitively performed in LSN. The conditions for W fuzz formation have been reached in the outer strike point area on the inertial W divertor (Einc > 20 eV, fluence > 1e24 He.m-2, Tsurf > 700 °C). Articulated Inspection Arm inspections before and after the He campaign have shown no macroscopic sign of surface modification. Post mortem analysis of the W components is ongoing to characterize the He induced nanostructures formed.
WEST phase 2 will start in autumn 2020, to address long pulse / high fluence operation on the newly manufactured ITER-grade actively cooled divertor, up to 10 MW/1000 s.
References:
[Diez], [Reux], [Goniche], [Colas], [Hillairet], [Maget], [Manas], [Liang], [Gaspar], [Ciraolo], [Tsitrone], [Pegourié], [Douai], this conference. [Martin 2008] J. Phys.: Conf. Ser. 123 012033Speaker: Jerome Bucalossi (CEA) -
KSTAR$^{1,2}$ program has been focused on resolving the key physics and engineering issues for ITER and future fusion reactors utilizing unique capabilities of KSTAR. First of all, a new advanced scenario was developed targeting steady-state operation based on the early diverting and heating during the ramp-up phase of plasma current and significant progress has been made in shape control to address the MA level of plasma current and stationary ITER-similar shape (ISS). It is demonstrated effective use of the H&CD with instrumented plasma control and shaping parameters became a key to access to the advanced operation scenarios such as high $β_p$, high $l_i$, high $q_{min}$, hybrid, internal transport barrier (ITB) and low $q_{95}$ operation. The examples of advanced scenarios are shown in Figure 1. The stationary ITB (fig 1a) is successfully reproduced with comparable confinement as H-mode level ($H_{89}$ ~ 2) both in limited and USN configuration, a low qmin scenario (fig 1b) is developed based on early diverting and delayed core heating approach and finally stable long-pulse H-mode operation (fig 1c) was extended upto 88 sec.
Recent KSTAR 3D experiments have focused on several ITER-relevant issues, such as divertor heat flux broadening in 3-row vs 2-row resonant magnetic perturbations (RMPs) on ELM-crash suppression, RMP-driven ELM-crash-suppression on ITER-like low $q_{95}$ (~3.2-3.4) and the characterization of ELM-crash suppression window in terms of normalized electron collisionality ($\nu^*_e$) and plasma toroidal rotation ($V_{tor}$) at pedestal top. Strong up-down asymmetry in 3-row configuration was identified and effect of the kink/anti-kink configuration was also clarified for ELM suppression in LSN plasmas. We have demonstrated the ISS-compatible RMP control in KSTAR using n=2, +900 phasing RMP, although the ISS has been more vulnerable to mode-locking than typical KSTAR configuration. A detailed study of the KSTAR database (where RMP configuration of all the discharges belongs to n=1, +900 phasing) showed that the ELM-crash suppression phase in KSTAR is in the range of 0.2 < $\nu^*_e$ < 1.2 and $V_{tor}$> 40 km/s. During the ELM suppression phase, coexistence of filamentary mode and smaller scale turbulent eddies at pedestal with broad-range of wave number ($k_θ$<1.1 $cm^{-1}$ and frequency (f<100 kHz) is identified by ECE imaging (ECEI) and strong energy exchange of the filamentary and turbulent modes was measured. The bicoherence analysis of the edge harmonic oscillations (EHOs) at natural ELM-less mode shows that there is a strong nonlinear interaction between EHOs, and the nonlinear interaction of EHOs has a significant effect on the ELM structure and dynamics.
Cross-validation between the advanced diagnostics and the modeling provides new insight on the basic transport process at KSTAR. For example, in the recent MHD-quiescent KSTAR plasmas non-diffusive avalanche-like electron heat transport events are observed by the ECEI and these observations have been successfully reproduced by gyrokinetic simulations indicating the broad range of spatial scales up to the minor radius. In addition, various studies utilizing the KSTAR fluctuation diagnostics demonstrated the importance of the turbulence characteristics in plasma rotation and confinement. The extensive study of the intrinsic rotation in Ohmic plasmas found a clear link between the counter-current toroidal rotation direction and the quasi coherent mode (QCM) which is measured by the Microwave Imaging (MIR). The improved confinement in the low rotation experiment was correlated with the suppression of the broadband (~200 kHz) ECEI fluctuations, and Collective Thomson Scattering provides a detailed measurement on the high-k density turbulence which is suppressed during the typical LH transition. Finally, strong interaction between fast-ion and EP driven MHD mode was identified with Fast ion $D_α$ (FIDA) diagnostics.
KSTAR provided unique demonstration on the performance of symmetric multiple Shattered Pellet Injections (SPIs) which is the main strategy of ITER for disruption mitigation. It was shown successfully the current quench rate changes proportionally as the time difference varies from several percent to several tens of percent of the thermal quench (TQ) duration (1~2 ms) and it was demonstrated that peak density was increased twice with dual SPIs compared with a single SPI and energy can be radiated when multiple SPIs are injected simultaneously, as planned in ITER.
Lastly, the research plan in near term will be addressed with the machine upgrades. KSTAR will focus on the development of the DEMO/ITER relevant operational scenario, i.e., high-beta steady-state operation with benign MHD activities which will require robust plasma control in strong shaping, control of MHD modes and thorough analysis of the fundamental physics processes. In these regards, KSTAR upgrades will includes extensive NBI (off-axis, 6MW) & RF (Helicon CD, 4MW) heating & current drive capabilities and the installation of new tungsten divertors with active cooling.References:
$^1$G.S. Lee et al, Nucl. Fusion 40 575 (2000) 575
$^2$H. K. Park et al, Nucl. Fusion 59 (2019) 112020 (13pp)Speaker: Si-Woo Yoon (National Fusion Research Institute) -
Construction of JT-60SA is progressing on schedule towards completion of assembly in March 2020 and the first plasma in September 2020. As of January 2020, manufacture and assembly of all the main tokamak components have been successfully completed satisfying technical requirements including functional performances and dimensional accuracies. Development of plasma actuators and diagnostics is also going well such as achievement of long sustainment of high energy intense negative ion beam. Commissioning of the power supply and the cryoplant has also satisfied requirements. Development of all the control systems and evaluation procedures of tokamak operation has been completed towards the Integrated Commissioning starting in April 2020, and plasma operation scenarios in the first plasma phase have been established. Unique importance of JT-60SA for H-mode and high-beta steady-state plasma research has been confirmed using advanced integrated modellings. These experiences of assembly, integrated commissioning and plasma operation of JT-60SA contribute to ITER risk mitigation and efficient implementation of ITER operation.
Introduction
The JT-60SA (R/a =3m/1.2m, Ip-max =5.5MA, heating power = 41MW x 100s) project [ref.1] was initiated in 2007 under the framework of the Broader Approach agreement by EU and Japan for early realization of fusion energy by conducting supportive and complementary works for ITER towards DEMO. Construction of JT-60SA is progressing successfully towards completion of assembly in Mar. 2020 and the first plasma in Sep. 2020 by the very close collaboration between QST in Japan, F4E in Europe, EU Voluntary Contributors and EUROfusion. The JT-60SA Research Plan [ref.2] covering its machine lifetime of ~ 20 years coordinated with ITER and DEMO schedules has been established with variety of plasma prediction using integrated modeling codes [ref.3]. Recently in Nov. 2019, a new collaboration arrangement between ITER and JT-60SA was signed which covers assembly, integrated commissioning and operation/experiments for finalization of ITER component design, risk mitigation and efficient implementation of ITER operation.Tokamak Construction
After the last IAEA FEC [ref.1], manufacture of all remaining tokamak components has been completed successfully including, superconducting Centre Solenoid (CS), thermal shields, Cryostat Top Lid, Cryolines, etc. As of Dec. 2019, the closure of the vacuum vessel has been accomplished, and the tokamak has been covered by the Cryostat Vessel Body (Fig.1). All the tokamak components have been assembled with excellent dimensional accuracy of ±1mm thanks to careful and smooth positioning using specially designed jigs, high accuracy measurement by Laser trackers, and fine adjustment utilizing sims. The magnetic field error is now expected below 10-4 Bt as designed. Commissioning operation of all large power supply systems, the Quench Protection Cirquit, the Switching Network Units and Super Conducting Magnet Power Supplies, has also been progressed with few residual commissioning activities still ongoing. The commissioning operation of the Cryoplant (equivalent refrigeration capacity of 9 kW at 4.4K) has also been successfully completed by satisfying the required performances.Plasma Hating Systems
For the heating systems, Positive-ion source NBs (85keV, 100sec, 20MW by 12 unit), Negative-ion source NBs (500keV, 100sec, 10MW by 2 units), ECH with multiple frequency Gyrotron (110GHz & 138GHz for 100s and 82GHz for 1 sec) and movable launchers, R&D have been steadily progressing and the targets of their development have been achieved. In particular, high energy intense hydrogen negative ion beams with 500 keV, 154 A/m2 for 118 s, which exceeds the requirement for JT-60SA, has been demonstrated by using a semi- cylindrical negative ion source with a three-stage accelerator. This result was realized by integration of i) stable voltage insulation by suppression of arching, ii) precise beam control and iii) stable negative ion production by maintaining the temperature balance in the negative ion source.
Integrated Commissioning and Control Systems
From April 2020 to Feb. 2021, the integrated commissioning is planned with the first plasma in Sep. 2020 and subsequent 5 months of machine commissioning with plasmas (‘the first plasma phase’). In this phase, the goal of plasma operation is to demonstrate equilibrium controllability of MA-class (<2.5MA) diverted plasmas with the full performance superconducting coil systems, 1.5MW ECH and upper divertor. For such tokamak operations, the Supervisory Control System and Data Acquisition System (SCSDAS) has been developed having the roles of (a) plant monitoring and machine state management, (b) discharge sequence management, (c)real-time plasma control, (d) device protection and human safety, (e) data storage, archive, and database management etc. For plasma controls, we have simulated operation scenarios of the first plasma phase with a newly developed advanced codes with control logics, such as pre-magnetic optimization scheme, plasma equilibrium control with iso-flux control method, control gain optimization method, and strategies for accessing stable operational regimes. Figure 2 shows the discharge scenario at Ip=2.5MA. EC wall cleaning operations and EC-assisted breakdown are also explored with optimized EC injection and toroidal / horizontal field. These results in the JT-60SA first plasma phase will contribute to highly valued subjects in ITER first plasma/subsequent operations.![Feedback-controlled plasma current wave form at Ip=2.5MA with upper divertor.
Scenario Development for ITER and DEMO and Risk Mitigation of ITER
After machine enhancements in 2021-2022, physics experiments will start in 2023 using in-vessel coils, particle fueling and pumping with lower divertor, enhanced diagnostics and high heating power of 26MW at Ip up to 5.5 MA. Toward this phase, variety of predictions of H-mode and high-beta steady-state plasmas covering divertor-SOL-Pedestal-Core have been progressing using advanced integrated modellings including newly-developed globally optimized steady-state transport solver GOTRESS coupled with turbulence models and pedestal models, Gyrokinetic theory based neural-network transport modeling DeKANIS, etc. These studies have confirmed the unique and important characteristics of JT-60SA (highly-shaped, high-beta, 500keV high energy ions, electron heating, controllable rotation etc.) for study of fusion plasma physics such as impacts of fast ions and plasma shape on microturbulence. As for operation scenario development of high-beta steady-state with controlled divertor heat load, an important result has been achieved using the integrated divertor code SONIC upgraded to treat multiple impurity species simultaneously. The result has shown that ‘mixture-seeding of Ar with small amount of Ne’ can keep the peak heat load below allowable 10MW/m2 together with smaller Ar concentration in the SOL and core plasmas than an Ar-only case. These studies have also confirmed significant roles of JT-60SA for ITER risk mitigation (disruption and ELM mitigation) including magnetic perturbation effect on both transient and stationary heat load, vertical displacement event, plasma response to massive gas injection, pedestal and ELM stability and control with Pellet and RMP.[ref.1] P. Barabaschi, Y. Kamada, H. Shirai and JT-60SA Intrgrated Project Team,
Nucl. Fusion 59 (2019) 112005.
[ref.2] JT-60SA Research Plan - Version 4.0, Sept. 2018,
http://www.jt60sa.org/pdfs/JT-60SA_Res_Plan.pdf
[ref.3] G. Giruzzi, M. Yoshida et al., Plasma Phys. Control. Fusion, 62 (2020) 014009.Speaker: Yutaka Kamada (National Institutes for Quantum and Radiological Science and Technology) -
A new era of predictive integrated modeling has begun. The successful validation of theory-based models of transport, MHD stability, heating and current drive, with tokamak measurements over the last 20 years, has laid the foundation for a new era where these models can be routinely used in a "predict first" approach to design and predict the outcomes of experiments on tokamaks today. The capability to predict the plasma confinement and core profiles with a quantified uncertainty, based on a multi-machine, international, database of experience, will provide confidence that a proposed discharge will remain within the operational limits of the tokamak. Developing this predictive capability for the first generation of burning plasma devices, beginning with ITER, and progressing to tokamak demonstration reactors, is a critical mission of fusion energy research. Major advances have been made using this predict first methodology. Extensive predictive modeling has informed the planning for the JET D-T campaign. This includes integrated modeling of JET hybrid regimes with newly upgraded heating sources, for various concentrations of deuterium D and tritium T. The self-consistent profiles of tungsten, ion and electron temperature, toroidal rotation and densities, have been predicted using theory-based turbulence and neoclassical transport models. The EPED model predicts it is possible to access the, higher pressure, super-H pedestal regime for JET achievable shapes. This prediction has been confirmed with DIII-D experiments. Super-H experiments on JET are planned. A new high accuracy neural network fit to the QuaLiKiz transport model has been completed, opening the way to time dependent predictions, at near real time speed, of complete tokamak discharges. Neural network fits to the TGLF and Multi-Mode models are progressing. The EAST tokamak is using predictive modeling to optimize the high bootstrap fraction regime for fully non-inductive operation and to plan future upgrades of power and current drive systems. A new integrated modeling workflow called TRIASSIC is being developed and tested on the KSTAR tokamak. Predictive modeling of CFETR is informing the design activity. ITER is using predictive modeling to simulate phases of the experimental operations plan. An overview of several of these recent advances will be presented, providing the integrated modeling foundations of experimental successes, as well as progress towards the goal of integrated predictive modeling for experimental design. Two examples, selected from the many advances in the prediction of tokamak experiments, are summarized in this synopsis.
1st example: The fast response of cold pulses due to impurity injection in tokamaks, with an inversion of the inward electron temperature pulse from decrease to increase, has long been argued to be inconsistent with a local transport paradigm [1]. The first demonstration that the cold pulse temperature response could be captured by a local turbulence transport model (TGLF [2]) was performed for the C-MOD tokamak [3]. Only electron and ion temperatures were predicted in these cases, with the density profile being evolved in a prescribed way. It was found that the inversion of the electron temperature pulse from decrease to increase was caused by the stabilization of the trapped electron mode (TEM) by the flattening of the electron density profile. In discharges where the TEM mode was not dominant there was no inversion in agreement with experiment. The transport model was then used in predict first method to simulate the cold pulse response [4] in the DIII-D tokamak. The very fast, high spatial resolution, density profile data on DIII-D confirmed the speed of the prescribed density response and the electron temperature response predictions were confirmed. The final step was to prove that the TGLF model could predict the fast density response to the impurity injection. This required adding the injected impurity density to the transport modeling. This integrated modeling was performed for experiments on the ASDEX Upgrade tokamak [5]. The predicted electron temperature response is compared with data in. It was found that the destabilization of the ion temperature gradient mode (ITG) by the transiently hollow impurity profile increased the speed of propagation of the electron density pulse into the core. Thus, the speed of the combined electron, ion, and impurity, temperature and density pulses were accurately modeled and new physics insights were discovered. This is a convincing proof that local turbulence transport can account for the paradoxical cold pulse phenomenon.
2nd example: The new upgrades to off-axis NBI current drive capability on DIII-D were preceded by state of the art integrated modeling [6] illustrated by the advanced tokamak predictions in. The profiles in Fig. 2 are a steady state self-consistent solution of the pedestal structure (height and width), core transport, MHD equilibrium and heating and current drive using validated theory-based models. An iterative high performance workflow IPS-FASTRAN was developed to find the integrated optimum solution [6]. Well validated theory-based models for MHD equilibrium (EFIT) and stability (DCON), turbulent transport (TGLF), pedestal structure (EPED1), neutral beam heating and current drive (NUBEAM) and electron cyclotron heating and current drive (TORAY-GA) were integrated. The IPS-FASTRAN modeling predictions have been confirmed with experiments showing good agreement that will be reported at the FEC 2020 conference. Verification of the accuracy of these predict first method simulations are a valuable test of the new capabilities. The same integrated modeling workflow is being used in the design of the CFETR, CAT Fusion Pilot Plant and SPARC tokamaks and to predict ITER plasmas.
This work was supported by the US Department of Energy under DE-FG02-95ER54309, DE-FC02-04ER54698, DE-SC0019736
[1] K. W. Gentle, W. L. Rowan, R. V. Bravenec, G. Cima, T. P. Crowley, et al., Phys. Rev. Lett. 74 (1995) 3620
[2] G. M. Staebler, J. E. Kinsey, and R. E. Waltz, Phys. Plasmas 14, (2007) 055909.
[3] P. Rodriguez-Fernandez, A. E. White, N. T. Howard, B. A. Grierson, G. M. Staebler, et al., Phys. Rev. Lett. 120 (2018) 075001.
[4] P. Rodriguez-Fernandez, A. E. White, N. T. Howard, B. A. Grierson, L. Zheng, et al., Phys. Plasmas 26, (2019) 062503.
[5] C. Angioni, E. Fable, F. Ryter, P. Rodriguez-Fernandez, T. Putterich, and the ASDEX Upgrade team, Nuclear Fusion 59 (2019) 106007.
[6] J. M. Park, J. R. Ferron, C. T. Holcomb, R. J. Buttery, W. M. Solomon, et al., Phys. Plasmas 25 (2018) 012506.Disclaimer-This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
Speaker: Gary Staebler (General Atomics)
-
-
-
Achieving ignition and high fusion yield in the laboratory is a central goal of the U.S. Inertial Confinement Fusion (ICF) Program. Three major and credible approaches are currently being pursued: laser indirect-drive (LID), laser direct-drive (LDD), and magnetic direct-drive (MDD). While the three approaches use very different means for driving a spherical or cylindrical implosion that can compress and heat a mass of deuterium-tritium (DT) fuel (by laser-generated radiation drive, direct laser irradiation, or direct magnetic acceleration, respectively), they share many common challenges, including how to efficiently couple the kinetic energy of the implosion to internal energy of the fuel at stagnation, and how to assemble the fuel at the necessary conditions, in pressure, temperature, and areal density to initiate self-heating and ignition.
Significant progress has been made in each of these approaches in recent ICF experiments on the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory, the OMEGA laser facility at the Laboratory for Laser Energetics, and the Z pulsed power facility at Sandia National Laboratories. This includes progress in both advancing the absolute fusion output and, as importantly, improving our knowledge and understanding of the implosion behavior and causes of deviations from ideal or theoretical performance.
In this overview, we will review some of the major results from each facility, with a particular focus on recent advances in diagnostic measurement techniques and analysis that have improved our understanding of the states of the assembled fuel and confining shell at stagnation. Particular emphasis is placed on recent advances in determining the 3D spatial morphology and thermodynamic properties of the fusion fuel, including pressure, temperature, areal density, and mix, that determine the degree of alpha-particle production, confinement, and self-heating, and proximity to ignition.
As an example, Fig. 1 shows the hot spot temperatures and areal densities inferred from some of the principal LID experimental campaigns performed on the NIF since 2011. Improved performance can be observed as the implosion design evolved from the original low-adiabat† (a~1.6) CH ablator design (2011-2013), to a mid-adiabat (a~2.3) CH design (2013-2015), and then to a mid-adiabat (a~2.7) high-density-carbon (HDC) ablator design (2016-2019) [1-3]. The principal gains resulted from improved hydro-instability at higher adiabat, and improved energy coupling and implosion symmetry with HDC designs. As of today, the highest performing HDC implosions have achieved hot spot temperatures of ~4.5 keV, hot spot areal densities of ~0.3 g/cm2, hot spot pressures of ~350 Gbar, and fusion energy output of >50 kJ. At these conditions, the self-heating of the hot spot by alpha-particle deposition is estimated to be amplifying the total fusion output by a factor of ~2.5-3x. As can be seen in Fig. 1, these conditions are now quite close to the static self-heating boundary, dT/dt>0, where for a static hot spot at the time of peak compression, the time derivative of temperature is positive due to alpha-particle heating exceeding energy losses from radiation and thermal conduction. Whilst reaching this boundary will be a very noteworthy physics achievement, it is not a sufficient condition for ignition. For ignition, the alpha-heating rate must exceed all losses including mechanical work from expansion of the hot spot after peak compression. This condition, equivalent to the requirement d2T/dt2>0, depends on the shell confinement and necessitates on the order of a ~1 keV higher hot spot temperature [4,5].
In the LDD approach, remarkable progress has been made using a data-driven statistical mapping model to optimize input laser and target parameters that has led to a 3x increase in neutron yield over the past two years on OMEGA [6]. The best-performing spherical direct drive (SDD) implosion has a hydrodynamically-scaled predicted fusion yield in the 0.5 MJ range at NIF laser energy, approaching the burning plasma regime. Polar direct drive (PDD) implosions and laser-plasma coupling research are performed on OMEGA and the NIF for the LDD approach [7,8]. In the MDD approach, recent magnetized liner inertial fusion (MagLIF) experiments on Z have produced record yields and stagnation parameters through steady enhancements in the target initial magnetization, laser preheat energy delivered to the fuel, and electrical current delivered to the imploding shells [9]. In addition to record fusion performance, MagLIF implosions have also demonstrated significant potential for DT alpha-particle confinement and stopping via the strong magnetic fields entrained in the fuel [10,11], a necessary requirement for self-heating and ignition.
This work was performed under the auspices of Lawrence Livermore National Security, LLC (LLNS) under Contract DE-AC52-07NA27344, the Sandia National Laboratories, a multi-mission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. DOE NNSA under contract DE-NA0003525, and upon work supported by the U.S. DOE NNSA under Award Number DE-NA0003856, the University of Rochester, and the New York State Energy Research and Development Authority.
† The adiabat is the ratio of the pressure of the DT fuel to the Fermi degenerate pressure.References
1. J. Lindl et al., Phys. Plasmas 21, 020501 (2014)
2. O. A. Hurricane et al., Nature 506, 343 (2014)
3. S. Le Pape et al., Phys. Rev. Lett. 120, 245003 (2018)
4. P. T. Springer et al., Nucl. Fusion 59, 032009 (2019)
5. P. K. Patel et al., Phys. Plasmas (submitted, 2020)
6. V. Gopalaswamy et al., Nature 565, 581 (2019)
7. F. J. Marshall et al., Phys. Plasmas 23, 012711 (2016)
8. P. B. Radha et al., Phys. Plasmas 23, 056305 (2016)
9. M. R. Gomez et al., manuscript in preparation (2020)
10. P. F. Schmit et al., Phys. Rev. Lett. 113, 155004 (2014)
11. P. F. Knapp et al., Phys. Plasmas 22, 056312 (2015)Speaker: Pravesh Patel (Lawrence Livermore National Laboratory) -
We present here recent highlights from Wendelstein 7-X (W7-X), the most advanced and largest stellarator in the world, in particular stable detachment with good particle exhaust, low impurity content, and energy confinement times exceeding 100 ms, maintained for tens of seconds, as well as proof that the reduction of neoclassical transport through magnetic field optimization is successful. W7-X, which has a magnetic field strength of 2.5 T and a plasma volume of 30 m$^3$, started operation in 2015 [1-5]. Following the installation of a full set of in-vessel components, in particular 10 passively cooled fine-grain graphite test divertor units, it was operated again in 2017 and 2018. Plasma pulses up to 100 s were successfully sustained [6], despite the lack of active cooling. Stable and complete detachment was achieved routinely. The pumping efficiency was initially relatively low [7-9] but it significantly improved later. Detachment with high pumping efficiency was achieved for up to 28 seconds at a heating power of 5 MW with a very low impurity content [10] (Figure 1), indicating control of divertor-heat-flux, plasma density, and impurity content, and giving confidence for reaching the foreseen high-performance, quasi-steady-state (30 minutes) discharges in the future [11]. The performance of the W7-X divertor, and the behavior and parameters of the edge- and scrape-off-layer plasma are now understood in quite some detail [eg. 12], thanks to measurements from a suite of diagnostics [eg. 13-17].
The earlier-reported stellarator triple-product record discharge [18] has now been shown to provide proof that the optimization for reduced neoclassical transport in W7-X was successful, Figure 2 [19]: The high temperature (appr. 3.5 keV for both ions and electrons in the center) and high hydrogen ion density (appr. 7x1019 m-3 in the core) were achieved with 5 MW of heating, and an energy confinement time of 0.22 s corresponding to about 1.4 times the energy confinement time expected from the ISS04 stellarator scaling [20]. For other, less optimized stellarators scaled to the W7-X size and magnetic field strength, similar plasma temperature and density profiles would have required significantly higher heating power to balance neoclassical transport, in particular in the mid-radius (strong gradient) region.
A number of discharges with similar performance to the triple-product-record discharge have since been achieved. These are generally characterized by core density peaking, and a reduction of turbulent density fluctuations. Without such turbulence reduction, the central ion temperature appears to be clamped to appr. 2 keV [21]. These findings are consistent with W7-X transport usually being dominated by ITG turbulence, but stabilized by strong density gradients in a so-called stability valley [22], as exemplified in Figure 3 for the W7-X standard configuration.
During turbulence-dominated phases, impurity confinement times are low (of order the energy confinement time) and no impurity accumulation is seen, but they can be very large if the turbulence is suppressed, and this then leads to impurity accumulation [23]. Recent findings from the Large Helical Device (LHD) show that ITG-dominated discharges readily mix hydrogen isotopes, whereas electron-scale (trapped-electron mode) turbulence does not [24]. It is tentatively concluded that a non-negligible amount of ITG turbulence is beneficial for impurity control as well as for fuel (isotope) exchange and helium exhaust in a stellarator fusion reactor, whereas too much ITG turbulence could potentially clamp the ion temperature below the burn point. These and other recent results [see eg. 25-31] will be put into the context of future goals for the W7-X, the world stellarator program, and the magnetic confinement fusion program in general.
References
1 T. Klinger et al, Plasma Phys. Controlled Fusion 59(1) 014018 (2017)
2 H.-S. Bosch et al, Nuclear Fusion 57, 116015 (2017)
3 R. C. Wolf et al, Nuclear Fusion 57 102020 (2017)
[4] T. Sunn Pedersen et al, Physics of Plasmas 24 055503 (2017)
[5] A. Dinklage et al, Nature Physics (2018) https://doi.org/10.1038/s41567-018-0141-9
[6] T. Klinger et al, Nuclear Fusion 59 112004 (2019)
[7] T. Sunn Pedersen et al, Nuclear Fusion 59 096014 (2019)
[8] D. Zhang et al, Phys. Rev. Lett. 123, 025002 (2019)
[9] D. Zhang et al, this conference (2020)
[10] M. Jakubowski et al, submitted to Phys. Rev. Lett. (2020)
[11] M. Jakubowski et al, this conference (2020)
[12] F. Reimold et al, this conference (2020)
[13] V. Perseo et al, Nuclear Fusion 59 124003 (2019)
[14] V. Perseo et al, this conference (2020)
[15] T. Barbui et al, JINST 14 C07014 (2019)
[16] C. Killer et al, Plasma Phys. Control. Fusion 61 125014 (2019)
[17] C. Killer et al, this conference (2020)
[18] T. Sunn Pedersen et al, Plasma Phys. Control. Fusion 61 014035 (2019)
[19] C. Beidler et al, in preparation (2020)
[20] H. Yamada et al Nuclear Fusion 45 1684 (2005)
[21] M. Beurskens et al, this conference (2020)
[22] J. A. Alcusón et al, Plasma Phys. Control. Fusion 62 035005 (2020)
[23] A. Langenberg et al, this conference (2020)
[24] K. Ida et al, Phys. Rev. Lett. 124, 025002 (2020)
[25] J. Geiger et al, this conference (2020)
[26] Y. Feng et al, this conference (2020)
[27] G. Fuchert et al, this conference (2020)
[28] K. Aleynikova et al, this conference (2020)
[29] H. Laqua et al, this conference (2020)
[30] S. Lazerson et al, this conference (2020)
[31] A. Dinklage et al, this conference (2020)Speaker: Prof. Thomas Sunn Pedersen (Max Planck Institute for Plasma Physics) -
In the recent deuterium experiment on the Large Helical Device (LHD), we have succeeded to expand the temperature domain to higher region both in electron and ion temperatures as shown by the red region in Fig.1. We found a clear isotope effect in the formation of Internal Transport Barrier (ITB) in high temperature plasmas. In the deuterium plasmas, we have also succeeded to realize the formation of the Edge Transport Barrier (ETB) and the divertor detachment, simultaneously. It is found that the Resonant Magnetic Perturbation (RMP) has an important role in the simultaneous formation of the ETB and the detachment. A new technique to measure the hydrogen isotope fraction was developed in LHD in order to investigate the behavior of isotope mixed fuel ions. The technique revealed that the non-mixing state and the mixing state of hydrogen isotopes can be realized in plasmas.
One of the objective of deuterium experiment is the achievement of the reactor relevant high temperature plasmas in order to investigate the problems arising in the helical type fusion reactor. As shown in Fig.1, the ion temperature of 10keV was achieved at the electron temperature of around 4keV. This was reported at the last IAEA-FEC(FEC2018). The operational domain was extended to the direction of higher electron temperature range. The electron temperature of 6.6keV was achieved keeping the ion temperature of 10keV. For lower ion temperature plasmas (Ti0=~7keV), the electron temperature range was extended over ~12keV. It is found that the extension in higher electron temperature region of high ion temperature plasmas is an effective way to suppress the Energetic particle driven InterChange mode (EIC), which often prevents the access of high ion temperature plasmas in LHD. This is preferable feature for the future reactor scenario development 1. The isotope effect is a long underlying mystery in plasma physics because the most of experimental observations in tokamak plasmas show favorable effect by the ion mass on the energy confinement time (tau_E) while the theoretical prediction based on the gyro-Bohm model shows unfavorable effect by the mass, i.e., tau_E~ M^-1/2. A series of dedicated experiment using dimensionally similar L-mode hydrogen and deuterium plasmas in LHD showed that the energy confinement time scaling for L-mode plasmas had a non-significant dependence on mass (M^0.01) 2. On the other hand, a clear isotope effect was observed for the formation of Internal Transport Barrier (ITB) in LHD as shown in Fig.2. Here, the profile gain factor G1.0 is defined as
and TLref is the expected temperature profile for L-mode plasmas with same heating condition 3. As shown in the figure, the ITB intensity, i.e., G1.0, is clearly larger for deuterium at the density range below 2x10^19 [m^-3].The realization of both divertor heat load mitigation and the good core confinement property is an important issue in developing future reactor scenarios. For divertor detached deuterium plasmas which are realized by the RMP application, we have succeeded to improve the confinement of 38% with a formation of ETB. Thus, we call this improved confinement mode as the RMP induced H-mode 4. As shown in Fig.3, a steep pressure gradient was formed in deuterium plasmas at the edge region inside the m/n=1/1 island which was produced by the RMP. This formation of the steep edge pressure gradient with the detachment has never been observed in hydrogen plasmas. The thermal transport analysis based on TASK3D-code shows significant reduction in the thermal diffusivity at the edge region (r/a=~0.8) which indicates the formation of ETB.
In the future fusion reactor, the deuterium (D) and tritium (T) ions co-exist in plasmas and are assumed to be uniformly mixed with the ratio of D/T=1. On the other hand, it is not clear whether this assumption is always valid. To clarify the assumption, experiments to investigate the isotope mixing state were performed using Deuterium and Hydrogen mixture plasmas. It was found that the non-mixing state of isotopes can be realized when the centrally fuelled ion species are different from the ion species fuelled at the edge 5. Figure 4 shows a typical example of the non-mixing state where the tangential NBI fuels Hydrogen ions in the central region while the Deuterium is supplied at the peripheral region by a wall recycling (black symbols and curves). It was found this state can be changed to mixing state by an injection of ice pellet of hydrogen isotopes, i.e. either of Hydrogen or Deuterium. It was also found that a formation of hollow density profile due to the pellet injection plays an important role for changing the state.
1 H. Takahashi et al., this conference
2 H. Yamada et al., Phys. Rev. Lett. 123 (2019) 185001
3 T. Kobayashi et al., this conference
4 M. Kobayashi et al., this conference
5 K. Ida et al., this conferenceSpeaker: Prof. Masaki Osakabe (NIFS) -
Using its unique flexibility and advanced plasma diagnostics, the TJ-II stellarator is contributing to the understanding and solution of critical challenges in fusion plasmas. Next, we highlight some of the most relevant recent results in the framework of its research programme.
Towards validation of gyrokinetic and neoclassical simulations. Aiming at the validation of the instability properties predicted by gyrokinetic (GK) simulations and of the electrostatic potential variations on the flux surface, φ1, calculated by neoclassical (NC) codes, dedicated experiments have been carried out in TJ-II for a systematic characterization of turbulence wavenumber spectra and perpendicular rotation velocity measured by Doppler Reflectometry (DR) at poloidally separated positions on the same flux surface [1]. Poloidal asymmetries in the intensity of the wave number spectrum that depend on plasma conditions have been characterized and compared with global linear GK simulations by the code EUTERPE. Model and experiment qualitatively agree in the radial dependence of the turbulence intensity, in the turbulence dispersion relation and in showing a poloidal asymmetry that depends on the magnetic configuration. Recent experiments exploring configurations with different magnetic ripple have shown a reduction in the turbulence asymmetry at configurations with reduced ripple. Besides, the influence of base ion mass has been investigated in hydrogen and deuterium plasmas. The ion mass in TJ-II plasmas does not affect the properties of the turbulence, neither the amplitude nor the spectral shape or the poloidal asymmetry. The lack of dependence of the turbulence spectrum on the ion mass is also found in GK simulations. Model validation will also benefit from the effort of verification of GK simulations in different computational domains [2] as well as from the application of the recently developed GK code stella to multispecies turbulent transport calculations in TJ-II [3].
Poloidal asymmetries in radial electric field, Er, are found that depend on the plasma collisionality. These results have been compared with the contribution to Er arising from −φ'1 as calculated with the NC version of the code EUTERPE. These results show variations in Er comparable in size to those found in the experiments, but there is a disagreement regarding the sign of the Er correction. Recent simulations performed with the newly developed NC code KNOSOS [4] show that the effect of kinetic electrons on φ1 has to be taken into account due to the strong Te dependence of the electron contribution to φ1 when the electrons are in the 1/ν regime [Fig. 1]. KNOSOS (KiNetic Orbit-averaging SOlver for Stellarators) is a freely available, open-source code that calculates neoclassical transport in low-collisionality plasmas of three-dimensional magnetic confinement devices by solving the radially local drift-kinetic and quasineutrality equations.
Experiments in TJ-II with cryogenic and TESPEL pellets show that post-injection particle radial redistributions can be understood qualitatively from neoclassical predictions while also providing a means to benchmark the HPI2 code [5]. TJ-II measurements provide 2-D maps of plasma potential and fluctuations to address the question of how hollow density profiles, created by pellets, affect turbulent transport. Fluctuations are stronger in the negative density gradient than in the positive one in consistency with TEM linear simulations [6].
Towards the validation of fast ion induced stabilization and the identification of Alfvén Eigenmode actuators. An ambitious research programme is in progress to investigate the relation between zonal structures and Alfvén eigenmodes (AE) and its role on the nonlinear dynamics of AEs and transport as well as to develop and demonstrate AE control strategies using ECRH and ECCD in TJ-II. The unique TJ-II experimental set-up using a dual HIBP has shown that, in some conditions, long range correlations (LRC) are detected both at the AE frequencies and low frequencies (<10 kHz) [Fig. 2]. LRC are observed in plasma potential fluctuations but not in density fluctuations as expected in zonal flow structures [Fig. 2]. It is an open question whether those zonal structures are directly driven by fast particle effects or/and are the consequence of plasma scenarios with reduced damping of zonal flows. Experiments in TJ-II have demonstrated the effectiveness of ECRH and ECCD actuators to modify AE activity [7].
Towards the characterization of the interaction between neoclassical and turbulent transport mechanisms. We have investigated the impact of Er on turbulence propagation and the coupling between the plasma edge and the scrape-off layer (SOL) during electron–ion root transitions where Er is changed in a controlled manner from positive to negative values. It is shown that Er does not only affects the radial turbulence correlation length but it is also capable of reducing the propagation of turbulence from the edge into the SOL. This result was obtained using a technique based on the transfer entropy, which quantifies the propagation of information [8]. These observations are highly relevant for the understanding of the mechanisms that determine the SOL width.
The interplay between of NC radial electric fields, Reynolds stress gradients and LRC has been investigated in different plasma scenarios in TJ-II. Turbulent driven acceleration alone cannot explain the dynamics of zonal flows whose radial width is affected by the isotope mass [9]. These results are in line with the expectation that the interplay between turbulent and neoclassical mechanisms is an important ingredient of the dynamics of edge zonal flows.
Power exhaust physics: liquid metals. Solid and liquid samples of Li/LiSn/Sn, in a Capillary Porous System (CPS) arrangement, have been exposed to the edge plasma [10]. A simple 1D model was applied to the data, allowing for the evaluation of the kinetic energy (Ek) of ejected atomic species while their residence time at the edge was determined by monitoring the ratio of first ion/neutral emission light intensities. A clear evolution of Ek with sample temperature was deduced for Li atoms, this being associated to the different relative contributions of sputtered/evaporated atoms. SnI emission into the plasma has also been measured with radial and toroidal resolution. The deduced mean free paths for the ejected Sn atoms under sputtering conditions (low T) imply unrealistic high energies if the bibliographic data for the ionization rate constant of Sn are assumed. For the LiSn case, Li as well as Sn emissions were simultaneously detected and analysed. Plasmas under cut-off (collapsing) conditions were also investigated to check for the sensitivity of the recorder line intensities and ionization rates to the edge electron temperature.
[1] T. Estrada et al., Nuclear Fusion 59, 076021 (2019)
[2] E. Sánchez et al. Gyrokinetic simulations in stellarators using different computational domains, 28th IAEA Nice 2020
[3] J. M. García-Regaña et al., Turbulent transport of impurities in 3D devices, 28th IAEA Conf. Nice 2020.
[4] J. L. Velasco et al.,KNOSOS, a fast neoclassical code for three-dimensional configurations, 28th IAEA Nice 2020
[5] K. McCarthy et al., Pellet studies in TJ-II, 28th IAEA Nice 2020
[6] A. Melnikov et al., 2-D mapping of fluctuations and plasma profiles in TJ-II, 28th IAEA Nice 2020.
[7] A. Cappa et al., AE control strategies in TJ-II, 28th IAEA Conf., Nice 2020.
[8] G. Grenfell et al., Nucl. Fusion 60 014001 (2020)
[9] R. Gerrú et al., Nucl. Fusion 59 106054 (2019)
[10] F. Tabarés et al., Liquid metal studies in TJ-II, 28th IAEA Conf., Nice 2020Speaker: Prof. Carlos Hidalgo (CIEMAT)
-
-
-
-
-
For the first time, experiments on the DIII-D tokamak have demonstrated electron cyclotron current drive (ECCD) with more than double the efficiency of the conventional outside launch by using a novel top launch geometry (figure 1), as predicted by linear ray tracing and quasi-linear Fokker-Planck simulations. Studies have shown that off-axis current drive is a requirement for a steady-state reactor in the Advanced Tokamak (AT) regime$^{1,2}$; however, driving current off-axis efficiently remains a challenge. Launching electron cyclotron waves from the high field side of the plasma, but the low field side of the resonance, with large toroidal steering in a plane nearly parallel to the resonance layer (illustrated in figure 2) is found to greatly increase the ECCD efficiency at mid-radii compared to conventional outside launch. The higher ECCD efficiency is due to 1) selective EC wave damping on higher v$_{||}$ electrons, and 2) longer absorption path lengths to compensate for inherently weaker absorption at higher v$_{||}$. DIII-D experiments using a prototype top launch system with a fixed mirror have established these two tenets through scanning v$_{||}$ of the wave-particle interaction by varying the magnetic field B$_T$. Power deposition measurements show that the absorbed EC power decreases for higher v$_{||}$ interaction (lower B$_T$), giving rise to a “sweet spot” (optimal B$_T$) for maximum ECCD efficiency at $\rho$~0.5 (figure 3) where the higher current drive efficiency for higher v$_{||}$ is balanced by sufficient absorption. Simulations of ‘top launch’ ECCD for FNSF, DEMO and CFETR support it as an improved efficiency off-axis current drive technique for future fusion reactors$^{3-5}$.
Top launch ECCD with a long wave-electron interaction zone and a large Doppler shift ensures strong damping on tail electrons leading to higher ECCD. Top launch ECH experiments have been done previously on TCV using radial launch and 3$^{rd}$ harmonic X-mode to heat high density plasmas; the top launch experiments on DIII-D have the different goal of efficient off-axis current drive, and the launch scheme has been uniquely optimized for this purpose. As illustrated in figure 2, EC wave from this top launch 1) propagates nearly parallel to the resonance plane and only gradually approaches the resonance, resulting in a longer absorption path; 2) suffers less trapping effects by being on HFS of the axis; and 3) allows a larger Doppler shift, thus interacting with higher energy (less collisional) electrons. To experimentally validate and characterize this approach, a prototype top launch system is installed on DIII-D with a fixed mirror angle utilizing 2$^{nd}$ harmonic damping of either a single 110GHz or 117.5GHz gyrotron with injected power between 0.5-0.6MW.The longer absorption path of top launch predicted by the TORAY ray-tracing code is measured by modulating the ECCD power and observing the electron temperature oscillations with an ECE radiometer. The measured power deposition location generally agrees with TORAY. The vertical path is verified via comparison of X-mode and O-mode deposition, where the predicted location shift between X and O (reflecting the different vertical paths due to the different absorptions) is confirmed. A much longer (i.e., three times) absorption zone for top launch ECCD compared to outside launch is also measured, consistent with TORAY. The longer interaction zone usually results in a broader deposition profile when mapped onto $\rho$ space but not always. The more parallel the EC ray is to the flux surface, the narrower the deposition profile.
Selective damping on electrons with different v$_{||}$ via top launch ECCD geometry is evidenced by the reduced absorption measured with lower B$_T$ in DIII-D experiments, as predicted by TORAY. The cold resonance moves to higher v$_{||}$ at lower Bt and the wave-electron interaction follows. Damping on tail electrons that are less collisional and drive current more efficiently is crucial for high ECCD efficiency; however, because the electron population decreases with increasing energy, the total absorption can drop far below 100%. Reduced total absorption at extreme low B$_T$ (i.e., high v$_{||}$) is observed in both L-mode and H-mode (figure 3(a)) plasmas in DIII-D.
The highest top launch ECCD efficiency is predicted and achieved when balancing higher v$_{||}$ interaction and sufficient total absorption. The ECCD profile is determined from the change in the magnetic field pitch angles measured by motional Stark effect (MSE) polarimetry. As illustrated in figure 3(b), at high Bt (low v$_{||}$) the ECCD efficiency is low despite full absorption. The measured ECCD efficiency increases with decreasing B$_T$ (increasing v$_{||}$) until the curve rolls over when too little wave energy is absorbed by too few high v|| electrons. In H-mode plasmas, the ‘sweet spot’ for highest top launch ECCD is predicted and measured at Bt~1.55T, where the driven ECCD at $\rho$~0.5 is double for top launch compared to outside launch (figure 1), consistent with the predictions from TORAY and quasi-linear Fokker-Planck code CQL3D.
Studies are underway to evaluate whether the high-beta, steady-state goals of the AT program on DIII-D can be achieved using four top-launch gyrotrons and four outside-launch gyrotrons. Initial FASTRAN simulations with self-consistent transport/pedestal/current profile modeling shows that 3MW top launch can drive as much ECCD at $\rho$~0.65 as 6+ MW outside launch in the "high qmin" AT regime owing to the near doubling in ECCD efficiency, allowing access to the highest stable $\beta_N$ (~4.5) with a non-inductive current fraction of 1. These results suggest that the combination of 3 MW top launch and 3 MW outside launch is a reasonable optimum to achieve the high-beta, steady-state goal of the DIII-D AT program.
Top launch ECCD is a promising off-axis current drive technique for future fusion reactors. Top launch ECCD shares the same reactor-relevant features of conventional outside launch ECCD, such as easy coupling to the plasma, no near-plasma antenna, and small port requirements, along with long experience in gyrotron development. Modeling for FNSF-AT shows >50% higher off-axis current drive efficiency for top launch ECCD compared to outside launch$^3$, similar to the predictions for DEMO$^4$. Greater than 35% improvement in ECCD at $\rho$~0.5 has already been found in modeling the CFETR baseline$^5$, reaching a current drive figure of merit of $\gamma$~0.16x10$^{20}$ A/m$^2$W for 14.5keV; or a dimensionless current drive efficiency of $\xi$~0.37. The experimental demonstration of doubling off-axis ECCD on DIII-D and the great enhancement found in simulations of FNSF-AT, DEMO and CFETR strongly support top launch ECCD as an exciting reactor-relevant and efficient off-axis current drive technique.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-FG02-97ER54415, DE-AC52-07NA27344, DE-AC05-00OR22725, and DE-SC0019352.
$^1$F. Najmabadi, et al, FED 38 (1997) 3; $^2$F. Najmabadi, et al, FED 80 (2006) 3; $^3$R. Prater, et al, APS-DPP (2012); $^4$E. Poli, et al, NF 53 (2013) 013011; $^5$Xi Chen, et al., EPJ Web of Conferences, 203* (2019) 01004Speaker: Xi Chen (General Atomics) -
Future DT operation in ITER and DEMO will face a significant number of challenges. From the physics point of view, the change from DD to DT plasmas is poorly understood. There are indications that the core confinement, the ELM behavior, the pedestal confinement and the Scrape of Layer behavior can significantly change from DD to DT. From the operational point of view, scenarios with high enough alpha power production are essential to demonstrate that efficient electrical power generation is possible but this also requires techniques for power exhaust, ELM and impurity sources and accumulation control in the presence of enough additional input heating power and a metallic wall environment.
In order to address the relatively scarce knowledge of DT plasmas and in support of safe future DT operations, JET has developed a scientific program in DD, TT and DT with the aim of supporting and minimizing the risks of the transition from DD to DT plasma operation in ITER 1. Such program helps to understand and document the physics characteristics of DT plasmas but also to provide scenario solutions integrating several key aspects such as core transport, impurity accumulation avoidance, power exhaust, enough fusion power generation to reveal alpha particle effects and potential difficulties for operation.
To this end, JET has gone through an upgrade of diagnostics and heating systems, in particular, Neutral Beam Injection (NBI) input power, which has delivered the record of 32MW. With such power, high performance scenarios have been developed following two main routes, i.e., the baseline scenario with $q_{95}$~3, β$_N$~2, H$_{98}$(y,2) ~1 at high current and magnetic field and the hybrid scenario with $q_{95}$~4, β$_N$~2.4, H$_{98}$(y,2) ~1.3 at reduced current and low central magnetic shear.
Compared to previous campaigns [2], a new baseline JET DD neutron rate record in the ITER Like Wall (ILW), R$_{NT}$=4.1x10$^{16}s^{-1}$, has been attained as shown in figure 1. This high performance was possible at current Ip=3.5MA and magnetic field B$_T$=3.35T through a complex non-linear interaction between edge and core plasma regions. From the pedestal side, high frequency ELM’s help to flush impurities. This is obtained by combining 50%/50% particle source with pacing pellets and neutral gas puff as it allows simultaneously good confinement and impurity flushing. The pedestal density is relatively low for such a high current, ~4x10$^{19}m^{-3}$, but the strong peaking, which is significantly high in these conditions of low core NBI fueling, leads to a core density ~1x10$^{20}m^{-3}$. Rotation and its shear are high both for the core and the pedestal with a Mach number reaching 0.6 at mid-radius. The ExB shearing of such plasmas, which usually are in the ITG regime, have been shown to be important for the decoupling of ions and electrons [3], with Ti/Te~1.7 in the best performing discharge at H$_{98}$(y,2)=1.0, and for increasing the density peaking through an increase of the inward pinch [4]. The radiation power is mostly localized in the low field side pedestal for the whole discharge and never penetrates in the inner core, which allows long and stable flat-top pulses up to 5s.
This new baseline scenario at high Ip and with relatively low density and high temperature at the pedestal offers an attractive alternative to hybrid scenarios as no specific q profile tailoring or core improved confinement is required in order to obtain high neutron rate. Although the Greenwald fraction is moderate, G$_{fr}$=0.65, a key point is the density peaking, which provides core densities compatible with high fusion power. Furthermore, the ion pedestal temperature, 1.3keV, provides a route for avoiding W accumulation through neoclassical screening. The relevance of this scenario for ITER is strong for the long pulse operation program [5] as it can soften some of the constraints for the hybrid scenario in terms of off-axis current.
The hybrid route has also stablished a new neutron record generation compared to previous campaigns, R$_{NT}$=4.7x10$^{16}s^{-1}$. This has been obtained in conditions of reduced β$_N$=2.4 by working at higher magnetic field, i.e. 3.45T and $q_{95}$=4.5 rather than 2.8T used in previous campaigns with $q_{95}$=4.0. Actually, temperature and density profiles are essentially identical for both magnetic field showing that, at least in this regime, high confinement appears to be largely independent of the magnetic field. However, it has been also obtained that β$_p$≥1 seems to be necessary to get good confinement. Such hybrid plasmas are type I ELMy H-modes with Ti/Te ~1.5. They are prone to suffer from impurity accumulation when the low gas used in order to obtain high confinement leads short phases of too low ELM frequency. Further techniques, such as pacing pellets, will be used to maintain an adequate level of ELM frequency, and hence impurity flushing, but at reduced gas injection.
A key element that has significantly improved the robustness of both scenarios is the use of real time control techniques which have been applied for ELM, fueling and impurity accumulation control. With such techniques, the disruptivity has been significantly decreased to ~9% for the hybrid and to ~25% for the baseline when previously reached 60%.
An extensive exercise of ‘predict first’ modelling has been carried out with the aim of predicting the DT fusion performance expected in the future JET-DT campaign. As a first step, DT equivalent fusion power from recent plasmas at high power has been compared to past predictions from lower power discharges showing a reasonable agreement when using validated models, such as TGLF or QuaLiKiz, for core transport [6,7].
DT extrapolations to the maximum available power at JET show that 11-16MW of fusion power are possible both for the baseline and hybrid routes. In particular, favorable core isotope effects are found for the hybrid scenario in conditions of low turbulent transport, such as strong rotating and high core thermal and fast ion beta plasmas.
Finally, in baseline plasmas, electron heating by alpha power can be dominant when ICRH is temporary removed. Such characteristic can be used to demonstrate alpha heating in ITER relevant conditions for the first time. These predictions will be compared to TT and DT plasmas when they will be obtained in the upcoming campaigns at JET.
References
- E. Joffrin et al 2019 Nucl. Fusion 59 112021
- L. Garzotti et al 2019 Nucl. Fusion 59 076037
- Hyun-Tae Kim et al 2018 Nucl. Fusion 58 036020
- J Garcia et al 2019 Plasma Phys. Control. Fusion 61 104002
- A.R. Polevoi et al 2015 Nucl. Fusion 55 063019
- J. Garcia et al 2019 Nucl. Fusion 59 086047
- F. Casson et al., submitted to Nucl. Fus.
Speaker: Jeronimo Garcia (CEA IRFM) -
Experiments on DIII-D support a new approach, confirmed by transport modeling, to achieving Q=10 in ITER using a scenario with low plasma current (~ 8 MA), high $\beta_{\rm p}$, and line-averaged Greenwald fraction ($f_{\rm Gw}$) above 1. At 8 MA the disruption risk and ELM challenge are greatly reduced, with the possibility that uncontrolled ELMs may be acceptable$^1$. Due to the need of sufficient fusion power and low plasma current, this approach requires high density with $f_{\rm Gw}$>1.0 simultaneous with high confinement quality ($H_{\rm 98y2}$>1). Using impurity injection, the recent DIII-D experiments achieve and maintain these simultaneous conditions. Previously, high $\beta_{\rm p}$ plasmas with $f_{\rm Gw}$ up to ~1.0 and $H_{\rm 98y2}$>1 were obtained in JT-60U, albeit transiently and usually operated at low absolute density$^{2,3}$, which is not favorable to reactor plasma. For the first time in a tokamak, experiments demonstrate that a stationary ITB at large radius ($\rho$~0.7) is compatible with $H_{\rm 98y2}$>1, at reactor-level absolute density ($n_{\rm e0}>1.0×10^{20}\;m^{-3}$), $f_{\rm Gw}$>1, and reactor-relevant $q_{95}$ as well (fig. 1). Such ITB is a key feature of the ITER 8 MA Q=10 modeling. Comparison between the experimental DIII-D profiles and the predicted ITER profiles shows also a good match of ITB location and profile shape (fig. 2). The DIII-D experiments confirm that the high density ITB in ITER modeling is achievable at similar $q_{95}$ using the high $\beta_{\rm p}$ scenario.
For the first time, neon injection is observed to trigger the formation of a large radius ITB in the density channel at reactor level density on DIII-D, which provides an effective experimental approach to achieve line-averaged Greenwald fraction well above 1. Pedestal density feedback control is used in the experiment. Therefore, the pedestal density is kept below the Greenwald limit, e.g. $f_{\rm Gw,ped}$<0.7. The large radius ITB strongly elevates the plasma density inside the pedestal. The density profile is fairly flat inside the ITB with $n_{\rm e0}>1.0×10^{20}\;m^{-3}$ (fig. 2). These high $\beta_{\rm p}$ experiments on DIII-D also show that the ITB is sustained long after (>8×the energy confinement time, $\tau_{\rm E}$) the neon injection is turned off. Impurity transport modeling based on the experimental data shows that neon provides electron source at $\rho$~0.75, which is the location where the foot of the ITB emerges. One of the possible mechanisms for the trigger of ITB formation could be that the important electron source creates a seed of local density gradient. The locally increased density gradient strengthens the effect of $\alpha$-stabilization of turbulence by reducing the non-adiabaticity of electron response$^4$, and starts a positive feedback leading to ITB formation. The same technique may also work for ITER high $\beta_{\rm p}$ plasma, when higher Z impurity is employed.
Using this technique, new high $\beta_p$ experiments on DIII-D demonstrate the stationary large radius ITB at reactor level density and reactor relevant $q_{95}$. Fig. 1 shows that the stationary phases of $f_{\rm Gw}$~1.0-1.1 and $f_{\rm Gw}$~1.3 at $q_{95}$~8 are sustained for 21 and 8 $\tau_{\rm E}$’s, respectively. These plasmas have even higher density than ITER at its 9 MA Greenwald limit. With ITB, the energy confinement is well above standard H-mode in these plasmas, e.g. $H_{\rm 98y2}$ up to 1.4. The degradation of confinement in one of the discharges shown in fig. 1 ($f_{\rm Gw}$~1.3 case) is due to excessive neon injection causing very high core radiation. The averaged neon injection rates are $1.3×10^{20}$ /s ($f_{\rm Gw}$~1.3) and $4.8×10^{19}$ /s ($f_{\rm Gw}$~1.1). With further optimization of the impurity injection waveform, sustained high confinement is also expected in the $f_{\rm Gw}$~1.3 case. These high $\beta$ ($\beta_{\rm N}$≤3.5, $\beta_{\rm p}$≤2.7) plasmas have $q_{\rm min}$>2.0 and quiet MHD behavior. Meanwhile, a non-inductive current fraction up to 0.9 is achieved simultaneously in the high density phase.
1D Transport modeling suggests the goal of Q=10 on ITER can be achieved at low plasma current (~ 8 MA) using the high $\beta_{\rm p}$ scenario. The simulations are performed using the STEP module in the OMFIT framework, integrating sub-modules for transport, heating, current drive and equilibrium calculations. Compared to previous ITER simulation works, the innovative features in this work include: 1) Plasma density, temperature, current profiles and equilibrium are all evolved in the simulations, while pedestal heights for density and pressure are prescribed to values slightly below the Greenwald limit (for density) and EPED prediction (for pressure); 2) Use TGLF model for transport prediction; 3) In TGYRO, E×B shear effect on turbulence suppression is turned off. Simulations use ITER “Day One” heating and current drive power: NBI≤33 MW, EC≤20 MW. Calculations predict that the following parameters can be achieved: Q=9.5±2.5, $I{\rm p}$=7.85±0.35 MA, $q_{95}$=7.54±0.39, $H_{\rm 98y2}$=1.74±0.13, $f_{\rm Gw}$=1.48±0.13, $f_{\rm NI}$≥98% and $P_{\rm fus}$=350±50 MW. A large radius ITB is shown in both temperature and density channels for electron and ion species (fig. 2).
Density profiles experimentally achieved on DIII-D match the density profiles simulated for ITER Q=10 in shape (ITB radius) and absolute core density value (fig. 2), while the core confinement quality is maintained well above standard H-mode levels. Although the experimental electron temperature is low compared with ITER simulation results, its shape is still a very good match to the ITER simulation results, if a multiplier of 9 is applied to the DIII-D data. The results confirm that the required large radius ITB at $f_{\rm Gw}$ above 1 and reactor level density in ITER Q=10 modeling is achievable experimentally at similar q95 using the high $\beta_{\rm p}$ scenario. These new high $\beta_{\rm p}$ experiments on DIII-D strongly support the ITER Q=10 simulations and pave the avenue to a new low plasma current (~ 8 MA) approach for ITER’s Q=10 Goal.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698 and DE-SC0010685.
$^1$R. A. Pitts, et al., Nucl. Mater. Energy, 20 (2019) 100696
$^2$R. C. Wolf, Plasma Phys. Control. Fusion, 45 (2003) R1
$^3$N. Oyama and the JT-60 Team, Nucl. Fusion, 49 (2009) 104007
$^4$M. Kotschenreuther, et al., “Regimes of weak ITG/TEM modes for transport barriers without velocity shear”, UP10.00020, 61st APS-DPP, Oct 21-25, 2019, Fort Lauderdale, USSpeaker: Siye Ding (ORAU/ASIPP) -
Recent EAST experiment has successfully demonstrated long pulse steady-state high plasma performance scenario with core-edge integration since the last IAEA in 2018 $[1]$. A discharge with a duration over 60s with $\beta_P$ ~2.0, $\beta_N$ ~1.6, $H_{98y2}$~1.3 and internal transport barrier on electron temperature channel is obtained with multi-RF power heating and current drive, i.e. ~2.5 MW LHW and 0.9 MW ECH, where the plasma configuration is the upper single null with the strike points on the tungsten divertor (shown in figure 1). Loop voltage was well controlled to be zero which indicates the fully non-inductive current drive condition. Small ELMs (frequency ~100-200Hz) were obtained in this long pulse H-mode discharge. In the operation, the optimization of X-point, the outer gap and local gas puffing near LHW antenna were investigated to maintain RF power coupling and to avoid formation of hot spot on the 4.6 GHz LHW antenna. Global parameters of toroidal field $B_T$ and line averaged electron density <$n_e$> were optimized for high current drive efficiency of LHW and for on-axis deposition of ECH. The on-axis ECH was applied not only for the core electron heating, but also for the control of high $Z_{eff}$ impurities in the core plasmas.
Meanwhile, a higher $\beta_N$ ~1.8 with a duration of 20s is achieved by using the modulated neutral beam. Several normalized parameters of $\beta_P$ ($\beta_P$ ~2.0), $\beta_N$ ($\beta_N$ ~1.8), $H_{98y2}$ (~1.3), $n_e/n_{GW}$ (~ 0.75) are close or even higher than the phase III 1GW scenario of CFETR steady-state $[2]$. Other features such as metal wall (tungsten divertor), low torque injection ($\Gamma_{inj}$~1.0Nm), electron dominated heating ($T_e$>$T_i$), moderate bootstrap current fraction ($f_{bs}$~50%), broaden current density profile with the central q(0)>1.0 and good energy confinement, have also been demonstrated in this scenario. Note that high-Z impurity accumulation in the plasma core was well controlled in a low level by using the on-axis ECH and reducing the fast ion losses through beam energy optimization.
More recently, EAST has demonstrated a compatible core and edge integration in high $\beta_P$ scenarios: high confinement $H_{98y2}$>1.2 with high $\beta_P$ ~2.5/$\beta_N$~2.0, $f_{bs}$~50% is sustained with reduced heat flux by active divertor heat flux at high density $n_e/n_{GW}$ ~0.7 and moderate $q_{95}$~6.7 (shown in figure 2). The energy confinement quality was almost maintained with $H_{98y2}$>1.2 during the radiation feedback control. By active impurity seeding through radiative divertor feedback control via radiated power, the peak heat flux is reduced by ~30% on the ITER-like tungsten divertor, here a mixture of 50% neon and 50% $D_2$ is applied. Note that EAST has developed a number of heat flux control techniques to reduce heat load in separate experiments.
In summary, recent EAST experiments demonstrated long pulse steady-state high plasma performance scenarios and heat flux feedback control. Detailed physics basis to investigate stability and particle transport will be presented for the understanding of fully integrated core-edge solutions on EAST. As a test bed for ITER and CFETR, the EAST upcoming experiments will further exploit additional heating power and demonstrate the core-edge integration of steady-state long pulse high performance scenarios with full metal walls.
This work was supported in part by National Natural Science Foundation of China under Grant No. 11975274,11975276, US Department of Energy under DE-SC0010685 and DE-FC02-04ER54698.
$[1]$ X. Gong et al 2019 Nucl. Fusion 59 086030
$[2]$ J. Huang et al 2020 Plasma Phys. Control. Fusion 62 014019Speaker: Xianzu Gong (Insititute of Plasma Physics, Chinese Academy Sciences)
-
-
-
Off-axis Neutral Beam Current Drive (NBCD) physics has been tested on DIII-D for Advanced Tokamak (AT) operation with increased off-axis injection power ($P_{OANB}\simeq7$ MW) by using the newly available, toroidally steerable co/counter off-axis neutral beam (CCOANB) injection capability. DIII-D experiments confirm that the new CCOANB drives current as predicted by the classical model NUBEAM for MHD quiescent plasmas (Fig. 1). Compared to on-axis injection, substantial broadening of the current and pressure profiles has been achieved with dominant OANB heating by injecting both the new CCOANB and the previous vertically steerable OANB. This is consistent with predictions of the theory-based IPS-FASTRAN integrated modeling that has guided the DIII-D beam system upgrade for the development of reactor relevant $\beta_N>4$ steady-state scenarios. Projecting to the Compact Advanced Tokamak (CAT) fusion pilot plant shows that off-axis NBCD aligns well with the high bootstrap current $f_{BS}>0.8$ operation, maintaining a broad current profile with $q_{min}>2$ and excellent off-axis NBCD efficiency.
The NBCD profile driven by the new CCOANB was measured in H-mode plasmas and compared with modeling. The NBCD measurement$^1$ is based on the local measurement of the magnetic field pitch angles from Motional Stark Effect (MSE) diagnostics. These pitch angles are converted to the flux-surface-average current density ($J_\parallel$) and parallel electric field ($E_\parallel$), either using kinetic EFIT equilibrium reconstruction or a more direct MSE analysis. This allows the beam driven current to be determined by $J_{NB} = J_\parallel – \sigma_{NEO}E_\parallel -J_{BS}$, where $\sigma_{NEO}$ and $J_{BS}$ are the neoclassical conductivity and bootstrap current density calculated by the Sauter model using the measured kinetic profiles as input. Differential NBCD measurement reduces model dependencies ($\sigma_{NEO}$ , $J_{BS}$) and systematic uncertainties of measurements. In this study, we compare two discharges, either i) on- and off-axis or ii) “Left” (more tangent) and “Right” (more perpendicular) off-axis NBCD in otherwise similar discharge conditions. The measured NBCD profiles driven by the new CCOANB in the co-current direction agree reasonably well with classical model of Monte Carlo beam ion slowing down calculation NUBEAM (Fig. 1). The NUBEAM modeling employs an accurate beam injection model validated against fast visible image and neutron measurement$^2$. The toroidal field direction (+BT) was chosen for better alignment of NBI to the local B, leading to good off-axis NBCD efficiency. The minimum value of q was maintained at $q_{min}>1$ without any significant core MHD activities. The measured $J_{NB}(\rho)$ in Fig. 1 shows a clear hollow NBCD profile with the peak at about half the minor radius $\rho\sim0.5$. The net driven current normalized to the total injection power is $I_{NB}/P_{NB} = 14.9$ kA/MW, which is as good as on-axis NBCD since the increased fraction of trapped electrons reduces the electron shielding in the outer radius region. The measured NBCD ($I_NB$) increases with the off-axis NB power. Higher NB power operation produced mild low-frequency MHD modes (n=2 and 3), resulting in the reduced NBCD compared with the classical model prediction. The measured NBCD profiles at the highest inject power matches the NUBEAM modeling with a modest anomalous beam ion diffusion $D_b = 0.3 ~{\rm m^2/s}$. However, even including the effect of finite $D_b$, the measured NBCD efficiency ($I_{NB}/P_{NB}$) does not decrease with $P_{NB}$.A low pressure peaking factor has been obtained with dominant off-axis NB heating. Figure 2 compares two discharges with different ratio of the OANB power to the total NB power ($f_{OANB} = P_{OANB}/P_{NB}$) at ~same injection power for elevated $q_{min} > {\sim1.5}$ discharges. The pressure peaking, $f_p = p_0/\langle p \rangle$ decreases substantially when the on-axis NB power (blue traces) is replaced by the new CCOANB (red traces), while maintaining $\beta_N\sim3$ with high total ($H_{89}\sim2.3$) and thermal ($H_{98}\sim1.25$) energy confinement. The lower pressure peaking also results in the increased low-n ideal $\beta_N$ stability limit$^3$. Both discharges inject full available power from the vertically steerable OANB at the maximum tilt angle. This profile broadening has been reproduced by the IPS-FASTRAN modeling$^4$ that integrates theory-based models of core transport (TGLF), edge pedestal (EPED1), equilibrium (EFIT), stability (DCON), heating and current drive (NUBEAM, TORAY) self-consistently to find steady-state ($d/dt = 0$) solutions. Figure 3 compares the pressure profile between the measurement (kinetic EFIT equilibrium reconstruction) and the IPS-FASTRAN modeling, where all transport channels ($n_e$, $T_e$, $T_i$, rotation, and current) are predicted without any significant free input parameters except the density and rotation values at the pedestal top.
The measured off-axis NBCD does not lose CD efficiency by going to a larger radius, which is beneficial for future AT reactors. The IPS-FASTRAN modeling predicts significant improvement of energy confinement time for broad current profile with flat or weak negative magnetic shear, compared with monotonic $q_0\sim1$, for the AT reactors. Projecting to the CAT $^5$ Fusion Pilot Plant with R = 4 m, B = 7 T shows an excellent off-axis NBCD efficiency for high Greenwald density fraction operation leading to high fusion performance. Off-axis NBCD aligns well with the high bootstrap current $f_{BS}>0.8$ operation maintaining broad current profile with $q_{min}$>2.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC05-00OR22725, DE-FC02-04ER54698, DE-AC02-09CH11466, DE-FG02-07ER54917, DE-SC0012656
$^1$ J.M. Park, et al., Phys. Plasmas 16, 092508 (2009).
$^2$ B.A. Grierson, et al., submitted to this conference.
$^3$ B.S. Victor, et al., submitted to this conference.
$^4$ J.M. Park, et al., Phys. Plasmas 25, 012506 (2018).
$^5$ R.J. Buttery, et.al., IAEA FEC, FIP-P3/26 (2018).Speaker: Jin Myung Park (Oak Ridge National Laboratory) -
The DIII-D tokamak has developed a new regime for high-beta hybrid plasmas where the broad current profile is achieved with strong off-axis electron cyclotron current drive (ECCD) rather than anomalous poloidal magnetic flux pumping. The high-beta hybrid regime with $q_{min}$ slightly above 1 and without sawteeth is a candidate for the $Q=5$ steady-state scenario on ITER$^{1-3}$, but the anomalous flux pumping mechanism that maintains $q_{min}>1$ despite strong central current drive is not yet understood$^4$. Experiments on DIII-D have found that high performance with $\beta_N=3.7$ and $H_{98y2}=1.6$ is maintained (Fig. 1) in high-density hybrids when 3.4 MW of ECCD is moved from $\rho<0.2$ to $\rho\sim 0.5$. The good agreement between the experimental $q_{min}$ evolution and TRANSP simulations (Fig. 2) differs from the usual hybrid situation where the simulation predicts $q_{min}<1$ but experimentally $q_{min}>1$, showing there is no evidence for anomalous flux pumping in this new hybrid regime with off-axis ECCD. Transport analysis finds higher density leads to weaker Alfven eigenmode (AE) activity (Fig. 3) and lowers the electron thermal diffusion, but there is little change to thermal transport in moving ECCD from on-axis to off-axis.
A reproducible high-beta hybrid regime has been developed on DIII-D$^{1-3}$ with stationary performance as high as $\beta_N=3.7$ and $H_{98y2}=1.6$ using 11.2 MW of NBI and 3.4 MW of co-ECCD (Fig. 1), with a total injected energy of up to 56 MJ. Stationary hybrid plasmas always exhibit a $m=3/n=2$ or higher order mode. While these discharges are fully non-inductive at moderate densities ($\sim4\times 10^{19}\mbox{ m}^{-3}$), at higher densities (up to $6.1\times 10^{19}\mbox{ m}^{-3}$) this long-duration regime is well suited for radiative divertor studies with the ECCD aimed at $\rho\sim 0.5$ to avoid the right hand density cutoff$^5$. Off-axis ECCD results in a broader current profile than on-axis ECCD, as measured by motional Stark effect (MSE) polarimetry. In these experiments the six gyrotrons use identical aiming, giving very localized ECCD profiles. Perhaps because of this, deleterious $m=2/n=1$ modes are destabilized when the ECCD location is between $\rho=0.25–0.40$, which corresponds to the region of the $q=2$ surface. Furthermore, for off-axis deposition, radial ECH injection is found to be more stable to $m=3/n=1$ mode onset than current drive injection.
While hybrids with strong off-axis ECCD have a broad current profile with $q_{min}>1$, this behavior is not anomalous and there is no evidence of poloidal magnetic flux pumping. It is well established that hybrid plasmas with on-axis current drive have anomalously broad current profiles with a measured $q_{min}$ well above simulations of the current profile evolution from transport codes like TRANSP$^{2-3}$. When the ECCD is moved off-axis to $\rho\sim 0.5$, the measured $q_{min}$ value increases (as expected for off-axis current drive) and TRANSP simulations of the expected $q_{min}$ evolution are in reasonable agreement with the minimum safety factor from MSE-constrained equilibrium reconstructions (Fig. 2). The measured and simulated loop voltage profiles are also in good agreement, with central values around 50 mV and edge values near zero; the peaked loop voltage profile shows that the current profile is continuing to broaden with time. (While the loop voltage profile rests at zero for hybrids with $n_e\sim 4\times 10^{19}\mbox{ m}^{-3}$ and on-axis ECCD, these hybrids with $n_e\sim 6\times 10^{19}\mbox{ m}^{-3}$ and lower off-axis ECCD efficiency retain positive loop voltage.) While steady-state hybrids with on-axis ECCD and $q_{min}\sim 1.1$ generate strong fishbones, the absence of the fishbone instability for hybrids with off-axis ECCD confirms the higher $q_{min}$ values ($\sim 1.5$).
An additional advantage of off-axis ECCD in hybrids is that higher density plasmas can be investigated without encountering the right hand density cutoff, which increases the confinement time and allows higher $\beta_N$ to be achieved due to lower electron thermal transport and reduced AE activity. Moving the ECCD deposition from on-axis to off-axis by itself has little measurable effect on local thermal transport, although the smaller electron temperature gradient inside the ECCD deposition zone does decrease the global confinement time. By raising the density, however, high confinement and high $\beta_N$ can be recovered for off-axis ECCD. About half of the confinement improvement at higher density is due to ~30% lower electron thermal transport. The remaining confinement improvement at higher density is due to reduced beam ion transport. In TRANSP simulations that adjusted an anomalous beam ion diffusion coefficient to match the experimental neutron rate, $D_{beam}$ decreased from $\sim 1.5\mbox{ m}^2/\mbox{s}$ to less than $1.0\mbox{ m}^2/\mbox{s}$ with higher density (Fig. 3). While high $D_{beam}$ does not directly affect thermal transport, it reduces the neutral beam heating effectiveness and thus lowers the heat flux and temperature gradients. The lower $D_{beam}$ at higher density is consistent with the weaker AE activity observed on the density interferometer. It is also found in these hybrid plasmas that particle transport, but not thermal transport, is strongly affected by the onset of a $m=3/n=1$ mode, such that higher and broader density profiles can be achieved only in the absence of the $m=3/n=1$ mode.
Since the mechanism of poloidal magnetic flux pumping in hybrid plasmas is still under investigation, there is interest in developing an ITER scenario with comparable performance that does not rely on anomalous flux pumping to maintain the target safety factor profile. The 'classical' behavior of the current profile with off-axis ECCD reported here demonstrates the existence of a new hybrid regime that does not rely on anomalous flux pumping to maintain $q_{min}>1$. It is interesting to note that this regime has similar characteristics to experiments in the "high $q_{min}$" regime with broad off-axis ECCD between $\rho\approx 0.3–0.6$ in which $q_{min}$ dropped to $\approx 1.4$, although the $m=3/n=2$ mode that seems important in hybrids was not present in these "high $q_{min}$" plasmas$^6$. This indicates that different paths (starting from low or high $q_{min}$) can be taken to access this regime.
This material is based upon work supported by the Department of Energy under Award Number(s) DE-FC02-04ER54698, DE-FG02-04ER54761, and DE-AC52-07NA27344.
$^1$F. Turco, et al., Phys. Plasmas 22 (2015) 056113.
$^2$C.C. Petty, et al., Nucl. Fusion 56 (2016) 016016.
$^3$C.C. Petty, et al., Nucl. Fusion 57 (2017) 116057.
$^4$C.C. Petty, et al., Phys. Rev. Lett. 102 (2009) 045005.
$^5$T.W. Petrie, et al., Nucl. Fusion 57 (2017) 086004.
$^6$T.C. Holcomb, et al., Nucl. Fusion 54 (2014) 093009.Speaker: C. Craig Petty (General Atomics) -
The radial width of heat flux flowing into the DIII-D divertor is found to expand beyond that of the established empirical scaling (1) for conditions of high input power and high plasma density. This expansion is consistent with a scrape-off-layer (SOL) radial pressure gradient limited by the MHD ballooning stability limit, but does not inherently result in a degradation of edge pedestal pressure or core confinement due to additional edge turbulence. This result has favorable implications for access to dissipative divertor regimes in future reactor-scale tokamaks.
At low heating power, ~3 MW, the DIII-D SOL heat flux width remains consistent with the empirical scaling law (1), dependent only on the midplane poloidal field. The low power midplane separatrix normalized pressure gradient, $\alpha _{MHD}$, increases with the higher density required for divertor detachment. At high heating power, ~ 13 MW, a higher separatrix density, $n_{e,sep}$, and resulting higher separatrix pressure, are required to achieve divertor detachment. For $n_{e,sep}$ approaching half of the Greenwald density limit, $n_{GW}$, the separatrix pressure gradient saturates, consistent with previous studies (2). Further increases in density or input power result in a broadening of the SOL and divertor temperature and density profiles, maintaining the pressure gradient near the MHD limit. The increase and saturation of the separatrix pressure gradient is summarized in Fig. 1, where the pressure gradient is normalized to the MHD ideal ballooning limit, $\alpha _{crit}$.
The saturation in $\alpha _{MHD}$ occurs at the same separatrix density even for attached divertor conditions at high power indicating the saturation is not due to divertor detachment.
The separate components of the midplane pressure profile are measured with Thomson scattering for the electron pressure and Charge-Exchange Recombination spectroscopy (CER) of the CVI impurity emission for the ion temperature and density contributions to the pressure profile. The separatrix normalized pressure stability limit, $\alpha _{crit}$, is evaluated with the ideal MHD code BALOO based upon magnetic equilibria across the data set at $\alpha _{MHD}\approx2.2-2.7$. As shown in Fig. 1, the measured pressure gradient, $\alpha _{MHD}$, saturates at about 50% above the MHD limit. The high pressure gradient is likely due to the high $T_{i}>>T_{e}$, at the separatrix taken from CVI CER measurements. Recent measurements from main ion CER indicate a separatrix $T_{i}$ much closer to $T_{e}$, resulting in a pressure gradient closer to the stability limit.
The saturation of the SOL pressure gradient results in an expansion of the SOL width as power and density are increased, as shown in Fig. 2. The SOL $T_{e}$ width, remains constant at ~1.8 times the ITPA $\lambda_{q}$ scaling for $n_{sep}∕n_{GwW}\leq0.3$, but then increases for higher density. The SOL $\lambda _{Te} $ is ~40% below that implied by the ITPA scaling given that $\lambda_{Te}\sim\frac{7}{2} \lambda_{q}$ at these collisionalities. For detached plasmas, shown by solid symbols in Fig. 2, as the input power is increased from 2 MW to 13 MW while the density required for detachment increases, the SOL $\lambda_{Te}$ and implied $λ_{q}$, increase $\geq$ 50%. Expansion of the SOL density width, $\lambda_{ne}$, is even stronger with a factor of 2.5 increase in width for the high-power case with divertor detachment. Implications for the SOL expansion at high power and density can be seen in the divertor plasma as well.
Shown in Fig. 3 are radial profiles at low and high power through the divertor leg halfway between the target and the X-point. These are both detached plasmas with near complete exhaust power dissipation. At high power the divertor plasma is ~3 times broader than the lower power case with both at $T_{e} \sim5-10\,eV$ at the same vertical location. The broader profile allows for greater total radiated power without a significant increase in the divertor density.
The increased turbulence and radial transport at high power and density might be expected to degrade the edge pedestal and resulting core confinement. However, no degradation of the pedestal is found with increased SOL width at high power and detachment onset. For low and high-power the density is increased with deuterium injection to achieve divertor detachment with the intrinsic carbon impurity radiation. While the SOL heat flux width increases by 50% for the high-power case compared to the low power case which remains at the empirical width scaling (Fig. 2a), the pedestal pressure is maintained at that expected from the EPED model. The normalized confinement at high power divertor detachment also remains similar to that for detachment at low power.
These results are encouraging for the compatibility of divertor heat flux control with core operational scenarios in future high-power density tokamaks. The expansion of the SOL width due to MHD stability can reduce parallel heat flux density allowing for divertor detachment at lower plasma and seeded impurity density than implied by simple scaling arguments (3). However, these results also imply that for study of high-power density, and resulting high divertor plasma density, in existing or future divertor test tokamaks, will require similar magnetic field values to those planned for reactor-scale tokamaks.
(1) T. Eich, et al., Nucl. Fusion 53 (2013) 093031.
(2) T. Eich, et al., Nucl. Fusion 58 (2018) 034001.
(3) M.L. Reinke, et al., Nucl. Fusion 57 (2017) 034004.
This material is based upon work supported by the Department of Energy under Award Number(s) DE-FC02-04ER54698.Speaker: Anthony W. Leonard (General Atomics) -
Detachment control tests at DIII-D and EAST have expanded to new sensors and integration with high confinement ($H_{98,Y2}$≈1.5, $\beta_N$=3) core scenarios (see $^1$ for details on core performance). Active detachment control protects the divertor target from extreme heat fluxes and temperatures which might otherwise cause melting and erosion while minimizing fuel or impurity seeding commands to what is required and thus mitigating core performance degradation. Using Langmuir probes (LPs), saturation current normalized to its value at rollover can be controlled in the range $J_{sat}/J_{roll}$ > 0.3, including tracking of target values that step or ramp. Additionally, triple probe tips at EAST measure Te, which is controllable for $T_e$ ≳ 3 eV, with the exception of dithering across the $T_e$ cliff.$^2$ Identification of the system by fitting the response of $J_{sat}$ or Te to gas puffing leads to two things: estimates for initial proportional-integral-derivative (PID) gains for the controller, and a discrepancy between the classic system models tested and the actual behavior. The inability of classic system models to adequately describe the response to gas puffing suggests that future work in detachment control should focus on a control-oriented reduced model and Model Predictive Control (MPC).
LPs have been demonstrated to be effective sensors for detachment control systems at DIII‑D and EAST, with a variety of puffing gas species as the control actuators. These results go beyond the original demonstration of LP control at JET$^3$ by using Te from triple LPs at EAST, testing the controller with D$_2$, CD$_4$, N$_2$, Ne, and Ar puffing, and integrating detachment control with high confinement core scenarios on both devices.$^1$ The $J_{sat}$ rollover detection logic at DIII-D also goes beyond previously reported results by allowing detection of roll-back-over or reattachment; the controller should always know which side of rollover it is on. This is important for coping with off-normal events because the system gain changes sign as $J_{sat}$ crosses the rollover point. An advantage of using LPs over Divertor Thomson Scattering (DTS) as the sensor as in previous DIII-D detachment control demonstrations$^2$ is that LPs typically have much greater spatial coverage. The $J_{sat}$ controller has demonstrated ability to track targets in the range $J_{sat}/J_{roll} \geq$ 0.3, including ramps, steps, and constant values (an example is shown in Figure 1). The EAST $T_e$ controller has been successfully operated at a variety of $T_e$ values (Figure 2) and has been used as a tool in detachment studies.
Detachment control is useful for supporting core-edge integration since it keeps gas puffing down to the minimum needed to reach the target. Detachment control experiments have achieved $H_{98,y2}$ > 1 at EAST and $\approx$1.5 at DIII-D. It is typically not challenging to avoid overshooting the target with this controller with appropriate tuning of the full PID system (PI-only can be effective as well) and a realistic target trajectory. That is, best results are found when abrupt steps in the target are replaced by transitions over a time period similar to the system dead time plus response timescale, or about 100-200 ms, as in Figure 1.
System identification of the $J_{sat}$ controller (which has been tested more aggressively) with a First Order Plus Dead Time (FOPDT) or second order (SOPDT) model is sensitive to the absolute value of $J_{sat}/J_{roll}$, meaning the optimal tuning for going from 0.8 to 0.5 is not the same as the optimal tuning for going from 0.5 to 0.3. Figure 1 shows a case where the target $J_{sat}/J_{roll}$ decreases in two steps and then is held at a constant value after each. When the whole history is fit to the SOPDT model (which extends FOPDT by adding damping), the response is underfit with reduced $\chi^2$=12. When the fit is restricted to a time window surrounding a single step, the model provides a much better description of the response with reduced $\chi^2$ of 1 or 2. The separate fits have different parameter values. The problem can be linearized locally over a small enough step, but a small enough step appears to be $\Delta J_{sat}/J_{roll} \approx$ 0.2, whereas the controller should be able to handle attachment down to deep detachment, or $J_{sat}/J_{roll} \approx$ 1. Additionally, classic formulae for calculating PID gains from FOPDT identification, such as the Ziegler-Nichols rule, consistently need fine tuning by hand for best performance. This is consistent with FOPDT and SOPDT being inadequate descriptions of the system. So, PID is not the optimal method for controlling this system if the PID gains need to be functions of the target value of $J_{sat}/J_{roll}$. Future work should implement MPC to produce a more effective and more general tool, but the difficulty is in obtaining a suitable reduced model for the system.
ITER will need an active detachment control system to protect its divertor plates. Detachment control tests on present-day devices aim to demonstrate ITER-relevant control schemes or identify critical research needs. Work so far on DIII-D and EAST shows that LPs, DTS, and radiated power measurements are adequate sensors for managing fuel or impurity puff commands in a narrow operating space, and that the algorithms must be improved in order to handle more general conditions. This work was supported in part by the US Department of Energy under DE-FC02-04ER54698.
$^1$Liang Wang, presentation at IAEA FEC 2020
$^2$D. Eldon, Nucl. Fusion 57, 066038 (2017); doi: 10.1088/1741-4326/aa6b16
$^3$C. Guillemaut, Plasma Phys. Control fusion 59, 045001 (2017)Speaker: Dr David Eldon (General Atomics) -
Recent tungsten (W) divertor experiments in the DIII-D tokamak have made significant progress elucidating key mechanisms responsible for high-Z erosion, re-deposition, leakage, and scrape-off-layer (SOL) transport. These results have important implications for ITER and other next-step fusion devices, including insight into W sourcing during mitigated edge localized modes (ELMs), diagnosing and understanding the net erosion of W in steady-state and transient phases, and developing a predictive capability for both local and global high-Z material migration.
The free-streaming plus recycling model (FSRM), a recently developed analytic model for intra-ELM, high-Z gross erosion$^1$, was validated in ITER-relevant, ELM-mitigated regimes achieved via pellet pacing and resonant magnetic perturbations (RMPs) in DIII-D (Fig. 1). ELM control via pellet pacing reduces W sputtering by suppressing impurity build-up in the pedestal (by increasing the ELM frequency), resulting in less physical sputtering of W caused by free-streaming, fully-stripped impurities expelled from the pedestal. Pellet pacing also decreases the divertor target electron temperature (due to increased edge neutral fueling), leading to less tungsten sputtering by low C charge states recycled from the divertor.
During the application of RMPs for ELM control, however, the peak W gross erosion rate actually increases. FSRM calculations indicate this is due to a decrease in the intra-ELM magnetic connection length, $L_\parallel$, relative to the length of the ELM filament, $L_{ELM}$. The ratio of these two quantities is the only free parameter in the FSRM. RMPs may cause the formation of a direct path between the strike-points and a reservoir of hot impurities deep into the plasma core$^1$, increasing the effective physical sputtering yield of tungsten from ELM impacts.
Leveraging newly calculated atomic rate coefficients and a recently developed ultraviolet spectroscopy system$^2$, we also present the first detailed, in-situ analysis of the net erosion of high-Z material in a fusion device. Tungsten gross and net erosion are measured spectroscopically via the ionizations per photon, or S/XB, method$^1$. W net erosion is nearly equal to gross erosion when the neutral ionization length, $\lambda_{iz}$, is large relative to the W$^{^+}$gyro-radius, $\rho_W$, and magnetic pre-sheath width, $\lambda_{sh}$, due to a lack of W prompt re-deposition. As $\lambda_{iz}$ decreases, the rate of W net erosion also decreases, with a strong inflection point near $\lambda_{iz}/\rho_W \approx \lambda_{iz}/\lambda_{sh}$ ~ 2-3. At the lowest ionization lengths achieved, $\lambda_{iz}/\rho_W \approx \lambda_{iz} /\lambda_{sh}$ ~ 0.4, net erosion drops to as little as ~20% of the gross erosion. Measurements are lower than calculations from analytic prompt re-deposition models$^{3,4}$ when $\lambda_{iz}$ is large. This is attributed to non-local W re-deposition, which is not included in the models. When $\lambda_{iz}$ is small relative to $\rho_W$ and $\lambda_{sh}$, more net erosion occurs than calculated by the models. ERO simulations indicate that W re-deposition from charge states W$^{^{2+}}$and higher (not included in the simple analytic models) becomes important in this regime.
Mixed-material, high-Z impurity migration models have also been validated by DIII-D tungsten divertor experiments. Simulations conducted using the coupled DIVIMP-WallDYN codes indicate $E$×$B$ drifts are the dominant driver of material migration in the divertor region in L-mode discharges with the ion $B$×$\nabla B$ drift away from the strike-points. Incorporating $E$×$B$ drifts, calculated via integrating Ohm's law along parallel field lines, is essential to match the W migration features observed experimentally$^5$. A peak in the W deposition profile appears many ionization lengths radially outboard of the ring of W-coated divertor tiles (Fig. 2). The WallDYN model agrees with experimental data within a factor of two over the entire radial extent of the W re-deposition pattern when $E$×$B$ drifts are adjusted to 60% of their calculated value (Fig. 2). This suggests (a) high-Z material migration involves multiple erosion/re-deposition events before reaching equilibrium, and (b) effects beyond Ohm's law, such as SOL currents, may play an important role setting the electric field in the divertor region.
Finally, the net force on high-Z impurities in the far scrape-off-layer is observed to change dramatically with ion $B$×$\nabla B$ drift direction in DIII-D, suggesting strong high-Z entrainment in SOL flows. Tungsten flow patterns in the low-field side, far-SOL (deduced via a midplane collector probe system$^7$) are observed to be primarily towards the outer target with the ion $B$×$\nabla B$ drift away from the outer strike point (OSP), and vice versa for $B$×$\nabla B$ drifts towards the OSP. An ad-hoc SOL flow model added to DIVIMP simulations (based on experimental flow profiles) produced reasonable agreement with measured high-Z flow patterns. This suggests that background plasma flow patterns may be instrumental in causing the long-hypothesized near-SOL, high-Z impurity accumulation, which sources impurities into the far-SOL via cross-field diffusion.
Because the atomic physics and surface physics of carbon and beryllium are quite similar, experimentally validated PMI models on DIII-D significantly advance the field towards a predictive capability for the high-Z divertor erosion, re-deposition, and leakage in the mixed-material environment of ITER. Such models are essential to develop/optimize mitigation strategies for minimizing high-Z core contamination and maximizing fusion gain.
This work is supported by the U.S. Department of Energy under DE-FC02-04ER54698.
- T. Abrams et al., Phys. Plasmas 26 (2019) 062504
- C.A. Johnson et al., Plasma Phys. Control. Fusion 61 (2019) 095006
- D. Naujoks et al., Nucl. Fusion 36 (1996) 671
- J. Guterl et al., Contrib. Plasma Phys. (2019) 1-7
- W.R. Wampler et al., Phys. Scr. T170 (2017) 014041
- E.A. Unterberg et al., Nucl. Fusion 60 (2020) 016028
- D.C. Donovan et al., Rev. Sci. Instrum. 89 (2018) 10I115
Speaker: Tyler Abrams (General Atomics) -
We report the achievement of a world unique capability of high power co/counter steerable off-axis neutral beam injection on a major tokamak, which widens the broad pressure and current profile parameter space for high beta steady-state advanced tokamak (AT) scenarios on DIII-D, while retaining the ability to balance the injected torque for low rotation studies. The unique steering capability of co/counter off-axis neutral beam (CCOANB) is being used to validate physics-based energetic particle and thermal transport models that are utilized in designing next-step facilities based on the steady-state AT approach. Prior to meaningful validation, however, a careful assessment of the transmitted power and energetic ion population produced by this novel heating and current drive system is critical. In this work, the off-axis beam injection is assessed through visible imaging (Fig. 1), neutron measurements and rotation profile measurements at balanced torque (Fig. 2), and used to broaden the pressure profile in steady-state advanced tokamak scenarios.
DIII-D has undergone a major upgrade and successfully injected high power off-axis neutral beam power (~4 MW) using the CCOANB in both co-current and counter-current directions. The total off-axis neutral beam power is approximately 7.3 MW at nominal operating voltage of 75 kV. Achieving an off-axis and co/counter steering capability necessitated significant modifications to the ion sources, internal neutral beam components and adaptor seal at the tokamak vessel connection. DIII-D is equipped with four neutral beam lines injecting through midplane ports at toroidal angles of 30, 150, 210 and 330 degrees. Each beamline houses two ion sources individually capable of ~ 2 MW injected power at ~ 80 kV, labeled left (LT) and right (RT) as viewed from behind facing the torus. For the CCOANB upgrade to the 210 degree beamline, each ion source has been modified to achieve a stronger vertical focus by modifying the accelerator grid modules to aim towards the beam centerline in a manner similar to Ref. 1, but the width of the ion source plates have not been reduced to retain high power. Tilting of the ion source plates is required for the beam to pass through the smaller effective aperture when the beam is off-axis. Minimal losses of neutral beam power have been achieved by optimizing the strong focus ion sources and optimization of the gradient grid voltage, enabling maximum power transmission by minimizing the amount of power “scraping off” on internal beamline components. Tilting of the ion source has been guided by fast visible imaging in Ref. 2, as shown in Fig. 1 (a,b), and resulted in neutral beam injection along the design centerline, as shown in Fig. 1 (c,d), with empirical characterization of each beam’s divergence derived from the beam vertical profiles, as shown in Fig. 1 (e,f). These measured beam characteristics are used in the parameterization of the beam in the NUBEAM Monte-Carlo heating and current drive package.
Through exclusive power injection of each source into MHD quiescent plasmas across a range of neutral beam voltage, perveance and plasma current in the same manner as Ref. 3, we conclude that a modest reduction of transmitted power (compared to on-axis, standard focus) has occurred. Prior to the beamline modification the 210LT and 210RT sources operating at 75, 81 kV and 2.60, 2.56 $μ$-perv respectively, each produced 2.0 MW. After commissioning, optimizing the neutral beam aiming, and implementing the off-axis injection geometry in NUBEAM, an approximate 15% lower than expected neutron yield has been observed. We attribute this reduced neutron yield to interaction with internal beamline components and reionization in the drift duct, as indicated by thermocouple measurements and photoemission detected by photodiodes in the drift duct.
Good ability to balance the neutral beam torque has been demonstrated by injecting the new (210 degree) off-axis counter injected beam against the existing (150 degree) off-axis co-injected beam in 2.0 T, 1.0 MA, MHD quiescent L-mode plasmas. L-mode at 1.0 MA has been chosen to minimize the edge intrinsic torque, retain good beam ion confinement with off-axis injection and minimize MHD. Verifying the ability to operate with balanced injection is critical for achieving “low torque” and “low rotation” operation for physics studies and in ITER demonstration discharges. This torque balancing exercise has been performed in a matrix using both tangential and perpendicular injection (2x2) and a three point scan in counter-injected voltage, as shown in Fig. 2. In these conditions, the intrinsic rotation profile has a positive edge offset with a slightly hollow core rotation, and therefore we do not expect to achieve a flat zero velocity profile at 0.0 Nm. Nevertheless, at zero injected torque the observed velocity shear is very weak and much smaller than a single co-injected off-axis neutral beam injecting +0.62 Nm of torque, confirming that low toroidal rotation shear can be achieved by balancing torque and controlling the beneficial effects of $E \times B$ shear on confinement. This submission supports DIII-D papers by C.S. Collins, J.M. Park and B.S. Victor.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-FC02-04ER54698, DE-AC02-09CH11466, DE-AC05-00OR22725, DE-FG02-07ER54917, DE-SC0020337
(1) C.J. Murphy, et. al., lEEE/NPSS 24th Symposium on Fus. Eng. SP3-21 (2011)
(2) M.A. Van Zeeland, et. al., Plasma Phys. Control. Fusion 52, 045006 (2010)
(3) W.W. Heidbrink, et. al., Nucl. Fusion 52, 094005 (2012)Speaker: B.A. Grierson (Princeton Plasma Physics Laboratory) -
Introduction: Disruptions due to tearing mode locking are one of few potential obstacles remaining for successful ITER fusion reactor and beyond. Here, we report the experiments of locking avoidance, but also contributing to the H-mode recovery and sustainment by slowly rotating edge-localized tearing mode (TM) layers by applying 3D external field. The process is expected as a non-linear response to applied external 3D field, static error field and/or their combinations. Preliminary NIMROD simulation are qualitatively consistent with experiment.
Localized edge tearing layer response: It has been discovered that the H-mode recovery with a slowly rotating external 3D field is accompanied by an edge localized tearing layer synchronized with the 3D field. The amplitude of the tearing layer is a few hundred Gauss, suggesting that the perturbed current density must be comparable to the equilibrium current density. In contrast, at q~2 and 3 the response to 3D external field is minimal. One possible hypothesis is that the 3D field facilitates the H-mode recovery and the stable edge tearing layers assist the sustainment of H mode by reducing the influence of error field (EF) and other 3D fields around core including the applied external 3D field. This is consistent with the concept of “screening out” EF and other 3D fields, proposed by refs.$^{1,2}$. These results are with similar dependence as previously reported$^{3}$. We also discuss preliminary results of NIMROD simulation based on experimental conditions.
Figure 1 shows the H-mode sustainment by the NTM locking avoidance with rotating external 3D field and also the successful H-mode recovery after the preprogrammed reduction of NBI torque input terminated the H-mode phase ( the total NBI power input level was kept same). In the recovery phase, the ratio of the average density (Fig.1(f)) to the edge density (Fig.1(g)) is near unity, re-establishing the good edge confinement property. The beta_n recovered to the level comparable in the initial H-mode sustainment period (Fig.1(c)). However, The further gradual increase of the 3D field by upper/lower I-coils current ~3.2 kA (Fig.1(d)) reduced the density gradient at the edge together the beta_n decrease (Fig.1(e,g)), indicating that the fine optimization of the external 3D fields pattern is the key for sustaining high plasma confinement performance.
Reconstructed plasma response: Figure 2 shows the time-evolution of the localized TM layer response amplitude by the perturbed vertical field component dBz measured with Motional Stark Effect (MSE) in the H-mode recovery period (3300-3520ms). Two cycles of oscillation were decomposed by taking the covariance with external magnetic sensor signals (n=1, 2, 3). Although the radial coverage by MSE measurements are limited, the localized tearing layer expanded from edge to q=3 domain with I-coil current increase(Fig.2(b)). The magnetic response around q=2 (Fig. 2(c,d)) is minimal with lower 3D field just after H-mode recovery up to beta_n maximum time period. With increasing 3D field, the response below q=2 became finite and is synchronized from the core to the edge (Fig.2(e)). Thus, the higher 3D fields impacts on the structure from edge to core rather than functioning as the screening-out. (Similar dependence was observed on the ion rotation and ion temperature perturbation, not shown).
NIMROD simulation: Figure 3 shows two results of slow and fast (Ω = 3×10$^{4}$, 9×10$^{4}$ rad/s) rotation of the external 3D field by plotting the n=1 perturbed dBz field along the outboard mid plane. The calculation was performed using experimental equilibrium profile parameters (at 3445 ms), including the near-zero plasma rotation due to the 3D field viscous damping. The external 3D field pattern includes multi-poloidal components, but was simplified to a mix of 2/1, 3/1, 4/1 and 3/2. The EF was not included in this calculation. The plasma resistive timescales are roughly 100 times faster than in the experiment, with S = 3.5×10$^{6}$ the Lundquist number (ratio of the resistive to Alfven time). The radial derivative of dBz is the indicator of the appearance of a perturbed toroidal current at the rational surface. For slow 3D field rotation, the radial derivative of dBz increases around q=2 (rho~0.67). On the other hand, for fast 3D field rotation, the variation of radial derivative of dBz is minimal. This result is similar with the expectation from the screening-out m=2 component of 3D field coupled to the plasma response near the edge, but without error field, which can be explained by taking into account of the relative velocity of the rotating external 3D field and the static error field EF on the mode-moving frame (1,4). The relative frequency can be large enough to induce the screening effect simultaneously for both fields. The detail analysis has just begun.
Summary: The slowly-rotating 3D external magnetic field, by inducing the edge localized TM layers, can not only avoid tearing mode locking, but also screen both the static error field EF and external rotating 3D field. The present operational frequency range is several tens-Hz. If the frequency limit is related to the resistive-wall frequency (this seems to be the case in the DIII-D experiment), in the ITER and fusion reactors the effective resistive-wall frequency with blanket is expected lower than one-Hz. Thus, this slowly-rotating 3D field application seems to become very practical. Furthermore, this slowly rotating 3D field approach can serve as a tool to control actively the plasma flow velocity at various rational surfaces utilizing multi-3D field patterns. This could assist the plasma performance optimization from the plasma core to the edge.
(1) Inoue, S., et al., Nucl. Fusion 57 116020 (2017). (2) Inoue, S., et al., Plasma. Phy. Control. Fusion 60 025003 (2018). (3) Okabayashi, M., et al., Nucl. Fusion 59 (2019) 126015. (4) Finn, J., et al., Phys. Plasmas 22 (2015) 120701.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC02-09CH11466, DE-FC02-04ER54698, DE-AC05-06OR23100, DE-FG02-04ER54761, DE-AC52-07NA27344 and DE-SC-0017992.Speaker: michio Okabayashi (Princeton Plasma physics Laboratory) -
Initial experimental evidence shows that the L-H transition power threshold P_th can be reduced via Neoclassical Toroidal Viscosity (NTV) associated with applied n=3 non-resonant magnetic fields (NRMF) [Figs. 1 and 2], and, independently, via a fast reduction or reversal of intrinsic edge toroidal (co)rotation [Fig. 3]. It is also demonstrated that a small/moderate increase in lower triangularity reduces P_th [Fig. 4]. These actuators are shown to enhance local E×B flow and E×B shear just inside the LCFS. Rotation reversals reduce the L-H transition power threshold with unfavorable grad-B drift direction to values similar to P_th in favorable configuration.
Techniques for reducing P_th, or mitigating its increase with applied magnetic perturbations are crucially important for ITER, particularly in view of marginal auxiliary power in hydrogen plasmas during the PFPO-1 campaign. Several approaches are demonstrated here that increase the L-mode E×B shear preceding the L-H transition, and potentially open up a path to reduce or mitigate P_th in burning plasma experiments. Control of L-mode edge shear flow is accomplished using either externally applied NRMF torque, or by changing the edge rotation boundary condition. In addition, P_th has been reduced via small plasma triangularity changes (within ITER-relevant margins).
Initial evidence shows that a large NRMF torque applied via the C-coil (1) can reduce P_th at low ion edge collisionality [ν_i*(ρ=0.95)~0.15] [Fig.1]. It is obvious from Fig. 1 that the NBI power required to access H-mode is reduced with applied NRMF. The power threshold (expressed approximately as P_th ~ P_NBI+P_OH-dW/dt where the Ohmic power P_OH and the time derivative of the stored energy W have been taken into account) is reduced by ~25% (from ~4.3 MW without NRMF to ~3.15 MW with applied NRMF/NTV). The core radiated power is not separately measured but is estimated to be smaller than 80 kW. This observation of a reduction in power threshold with NRMF differs fundamentally from the power threshold increase observed with applied Resonant Magnetic Perturbations (RMP) due to edge stochasticity and increased edge toroidal co-rotation (2). In contrast, CER measurements here show a substantial reduction of the L-mode edge toroidal (Carbon) co-rotation with applied NRMF/NTV just before the L-H transition, qualitatively consistent with the direction of the expected NTV counter-torque [Fig.2]. As a consequence, E×B flow shear near the LCFS is found to increase via increased radial shear in the toroidal rotation term in the radial ion momentum balance. The (partial) ITER 3-D coil set available during the PFPO-1 campaign can be used to generate large NTV in the edge plasma layer, favored by the low collisionality expected in the ITER L-mode edge.
We have also investigated in detail the effect of magnetic geometry on P_th, and attribute the increased L-H power threshold with unfavorable grad-B drift direction clearly to reduced E×B flow shear due to higher edge toroidal co-rotation [Fig. 3(a)]. The L-H power threshold is shown to decrease monotonically with increasing E×B shear in the outer L-mode shear layer near the LCFS (shown via NBI torque scan, Fig. 3c), and does not depend on shear in the inner shear layer.
However, it has been observed that spontaneous local rotation transitions increase E×B edge shear flow preceding the L-H transition in balanced/low torque ITER-similar shape (ISS) plasmas in DIII-D. Fig. 3(b) shows radial profiles of toroidal edge rotation approaching the L-H transition. The rotation reduction/ reversal commences locally near ρ=0.95; the layer of reduced rotation then expands radially over several ms, providing sufficient E×B flow shear to induce the L-H transition. As a result, P_th in ISS plasmas can be reduced to values similar to those observed with favorable grad-B drift direction [Fig. 3(a)]. The threshold reduction occurs over a range in NBI torque but is most effective for balanced torque or with low counter-Ip torque. Reversal/ reduction of intrinsic edge co-rotation appears to be triggered by transient increases in edge power flow (via sawteeth crashes or radial transport avalanches).
In ISS plasmas, increasing the lower triangularity by ~12% results in a clear reduction in L-H power threshold (by ~15%, shown here for hydrogen ISS plasmas (Fig. 4). A smaller data set in deuterium plasmas also demonstrates a ~10-12% reduction of P_LH at increased triangularity at low torque (≤ ±0.2 Nm). In the higher triangularity shape the outer divertor strike point was moved slightly further away from the lower divertor cryopump, which could also have affected divertor neutral pressure.This work was supported by the US Department of Energy under DE-FG02-08ER549841,
DE-AC05-00OR227252,DE-FG02-08ER 549993, DE-FC02-04ER546985,DE-FG02-07ER549176,
and DE-AC02-09CH114667.
(1) K.H. Burrell et al., Nucl. Fusion 53 (2013) 073038.
(2) L. Schmitz et al., Nucl. Fusion 58 (2019) 126010.
Speaker: L. Schmitz (University of California Los Angeles) -
A novel control architecture for simultaneous regulation of several plasma scalar variables, such as the thermal stored energy ($W$), the bulk toroidal rotation ($\Omega_\phi$), and the safety factor at various spatial locations (e.g., $q_{95}$, $q_0$), and for active suppression of Neoclassical Tearing Modes (NTMs) by means of ECCD {1} is shown to improve plasma performance even in the presence of NTMs on DIII-D. The Off-Normal Fault-Response (ONFR) system is employed as a supervisor to monitor the NTM development and assign the gyrotron authority to the competing control tasks (profile control vs NTM suppression). Instead of following the common previous approach based on attempting to regulate whole profiles {2}, a different control solution is utilized that consists in regulating such profiles only at specific spatial locations and/or by related volume-average magnitudes. Such point-wise/volume-average approach for regulating profiles in a tokamak may be more realistic in some cases due to the under-actuated nature of these devices. The integrated-control scheme has been tested in one-dimensional simulations using the Control Oriented Transport SIMulator (COTSIM), as well as in experiments in the DIII-D tokamak. The results obtained suggest that integrated control techniques may help to improve the plasma performance even in the presence of NTMs (see Fig. 1). In addition, good qualitative agreement is found between simulation and experimental results, showing that COTSIM can be a very powerful tool to advance the design of integrated control architectures {3}.
Successful regulation of $W$ has been demonstrated in DIII-D scenarios targeting $\beta_N \approx 3$ and $q_{min} \approx 1.4$ in both simulations (Fig. 1.A) and experiments (Fig. 1.B), despite the existence of a 2/1 NTM within the plasma. The IPB98(y,2)-scaling confinement factor, $H_{98(y,2)}$, is substantially improved when feedback control is employed (Fig. 1.C), possibly due to a close tracking of the $W$ target during the ramp-up and early flat-top phases of the discharge (which delays the appearance of the NTM) and during the flat-top phase (by increasing the injected power once the NTM develops). In addition, tight regulation of $W$ simultaneously with $q_{95}$ (see Fig. 2), which is equivalent to regulating $\beta_N$, and the use of preemptive NTM stabilization (see Fig. 3), also help to delay the NTM appearance and its effect on the plasma confinement, and allow for systematically achieving relatively high $\beta_N$ values ($\approx 3$) which could not be sustained without feedback. Simultaneously with $W$ regulation, control of $q_{95}$ and $q_0$ (see Fig. 2.A), as well as $\Omega_\phi$ (not shown in this synopsis), has been achieved in one-dimensional simulations for the same steady-state high-$q_{min}$ scenario using COTSIM. During experiments in DIII-D, $q_{95}$ was successfully controlled (see Fig. 2.B) under feedback by varying $I_p$ (Fig. 2.C), but the unavailability of counter-$I_p$ neutral beam injectors did not allow for $\Omega_\phi$ regulation. Moreover, the controller's performance to regulate $q_0$ was significantly worsened in experiments (see Fig. 2.B) as a result of having substantially limited off-axis power (the co-$I_p$, off-axis 210 beamline could only be employed for 2 s). When the 210 beamline was turned on at 3.25 s (see Fig. 2.C), $q_0$ gets close to its target for a short period of time. Later development of the 2/1 NTM caused a slight decrease in $W$ (with the associated reduction in the electron temperature and increase in the plasma resistivity), which relaxed the $q$ profile and made it difficult for the controller to raise $q_0$ in order to reach the desired target.
Active NTM suppression techniques by means of ECCD deposition at the applicable rational surface were also tested at the same time that regulation of the aforementioned individual scalars was carried out. Simulation results obtained with COTSIM estimated that the 2/1 NTM could be totally suppressed with about 3 MW of EC power in these steady-state high-$q_{min}$ scenarios (see Fig. 3) for initial seed islands of width 10 cm. During experiments, lack of EC power (1.5 MW maximum using the gyrotrons “Luke”, “Leia” and “Tinman”) and defective poloidal-mirror steering did not allow for total NTM suppression, although a delay in its development and a lower MHD amplitude were achieved.
These initial results suggest that feedback-control techniques that integrate individual-scalar-variable regulation algorithms and NTM suppression may substantially improve the plasma performance in steady-state high-$q_{min}$ scenarios in DIII-D. However, actuator availability is critical to achieve successful regulation of the different scalars. Further efforts will focus on developing control algorithms for scalars (possibly adding more scalars such as $l_i$, $\beta_p$ or $q_{min}$) and MHD stability that have greater levels of integration (e.g., with proximity controllers), as well as actuator management schemes. The final goals are to increase the effectiveness of the NTM suppression mechanisms, improve the controller’s performance, and achieve higher $\beta_N$ values.
Work supported in part by the U.S. Department of Energy under DE-SC0010661 and DE-FC02-04E54698.
{1} LA HAYE, R. J., GUNTER, S., HUMPHREYS, D. A., et al., Physics of Plasmas 9 (2002) 2051.
{2} WEHNER, W., BARTON, J., LAURET, M., et al., Fusion Engineering and Design 123 (2017) 513 , Proceedings of the 29th Symposium on Fusion Technology (SOFT-29) Prague, Czech Republic, September 5-9, 2016.
{3} HUMPHREYS, D., AMBROSINO, G., DE VRIES, P., et al., Physics of Plasmas 22 (2015) 021806.Speaker: Dr Andres Pajares (Lehigh University) -
Impurity seeding studies were performed for the first time in the slot divertor at DIII-D, showing that with suitable use of radiators, full detachment is possible without degradation of core confinement. First ever multi species SOLPS-ITER simulations including full cross-field drifts and neutral-neutral collisions in DIII-D demonstrate the importance of target shaping and plasma drifts on reducing impurity leakage. In addition to advancing our understanding of SOL impurity transport, these results show that neutral and impurity distributions in the divertor can be controlled through variations in strike point locations in a fixed baffle structure. This leads to enhanced divertor dissipation and improved core-edge compatibility, which is essential for ITER and for the design of any future fusion reactor.
This work demonstrates the impact of both target shaping and impurity species on the detachment onset, impurity leakage and pedestal characteristics. Experiments comparing nitrogen and neon seeding highlight that Ne dissipates further upstream than N, consistent with their difference in ionization potential and with neon leading to a higher pedestal pressure gradient. The results from these experiments are also crucial to show that the detachment front is not a purely local phenomenon, but it extends through the entire slot divertor. This study has been enabled by unprecedented diagnostic coverage in the closed divertor, which enabled multiple independent observation of plasma cooling evolution. The experimental data show strong dependence of detachment and impurity leakage on strike point location. For a database which includes power and density scans, the detachment onset consistently requires half the nitrogen amount when the outer strike point (OSP) is on the slanted inner surface, compared to the outer corner of the slot. Relative nitrogen contamination levels are reduced by 15-20% in the core under these conditions, as measured by the Charge Exchange Recombination Spectroscopy and independently confirmed by core SPRED measurements of N-IV, when the OSP is on the slanted inner surface (fig. 1a and 1b).
Insight on these experiments has been obtained by employing for the first time in DIII-D SOLPS-ITER simulations with D+C+N, drifts and n-n collisions activated. The modeling results highlight the impact of target shaping on the ionization source distribution. When the OSP is at the inboard surface, the resulting wetted target is different than with the OSP at the outer corner, thus resulting in different distributions of the ionization source.The inclusion of the drifts in the simulations enabled to study the behavior of these flows in a highly closed divertor showing the relevant role of convection on divertor asymmetry and divertor detachment. With the ion BxB directed into the divertor as in this work, drifts shift the recycling source radially towards the inner target. As such the recycling source is affected by the superposition of the closure effect and plasma drifts. This results in a redistribution of plasma flow in the SOL and divertor plasma, altering plasma profiles and shifting the parallel pressure balance between upstream and downstream. Flow reversal is found for both main ions and impurities affecting the SOL impurity transport and explaining the dependence on strike point location of the detachment onset and impurity leakage found in the experiments (fig. 1c).In addition to target shaping, the effect of different radiative species on power dissipation has been evaluated experimentally by replacing nitrogen with neon. The results show that Ne dissipates further upstream than N and thus removes the capability of the divertor to dissipate as confirmed by calculations using the 2-point model. The crucial implication is that for a fixed Te at the target, less SOL density is required with Ne which could be attractive from a core-edge integration point of view.
The two routes for dissipation we identified in the work (using N through divertor radiation and with Ne radiating mantle upstream) lead to different pedestal responses. While N does not significantly affect the profiles, Ne injection leads to an increased pedestal gradient (fig. 2).
These results are linked to the interesting observation that while neon readily enters the pedestal, nitrogen remains compressed in the divertor. This different leakage behavior between the two impurities is consistent with the higher ionization potential for Ne (21.6 eV) compared to N (14.5 eV), resulting in ionization further upstream from the target. Neon injection leads to an improved peeling-ballooning stability due to increased diamagnetic stabilization. Moreover, a self-enhancing mechanism of Ne build up has been identified as due to the increased pedestal stability and the radiative mantle. The understanding of SOL and pedestal changes associated with Ne injection as shown here is crucial to evaluate the use of neon as radiative impurity in future reactors. The findings of this work demonstrate that mitigated divertor heat flux with impurity seeding balancing core contamination, can be obtained by choosing appropriate radiative species for pedestal conditions, as well as by optimizing divertor geometry and tailoring drifts for particle entrainment.
Work funded by the US DOE under Awards DE-FC02-04ER54698 (DIII-D), DE-AC52-07NA27344 (LLNL), DE-AC02-09CH11466 (PPPL), and DE-AC05-00OR22725 (ORNL), DE-NA0003525 (SNL) and LDRD project 17-ERD-020.Speaker: Dr Livia Casali (General Atomics) -
Internal magnetic fluctuation measurements identify magnetic turbulence in the DIII-D ELMy H-mode pedestal as micro-tearing modes (MTM) and mode growth accompanied by degraded plasma confinement is observed. This work provides the first direct measurement of internal magnetic fluctuations supporting the prediction of gyro-kinetic simulations$^1$ that MTM exist in the H-mode pedestal. Using a Faraday-effect polarimeter, magnetic turbulence is observed in the edge of ELMy H-mode DIII-D plasmas. The turbulence amplitude correlates with confinement degradation in ELMy H-mode plasmas during a slow density ramp. Line-averaged fluctuation amplitude indicates the turbulence originates from electromagnetic instability. Frequency, poloidal wavenumber and propagation direction of the magnetic turbulence all agree with expectations for MTM. Magnetic turbulence amplitude non-monotonically correlates with collision frequency, peaks off mid-plane and correlates with temperature gradient evolution between ELMs, consistent with MTM features identified from theory and simulation. These internal measurements provide unique constraints towards validating models and developing physics understanding of H-mode pedestal in future devices.
Using a newly-developed high-speed Faraday-effect polarimeter, magnetic turbulence is observed to correlate with confinement degradation (Fig. 1). The polarimeter has been verified capable of directly measuring the absolute amplitude of line-averaged radial magnetic field fluctuations with wave number $k_\theta<1/cm$ and frequency up to 1 MHz. In an ELMy H-mode plasma with slow density ramp, the measured internal magnetic turbulence grows and saturates and is correlated with a ~20% drop in pedestal-top pressure and global plasma confinement. The line-averaged magnetic fluctuation amplitude after confinement degradation (>3.5 s) is ~20 Gauss over the full bandwidth, which could lead to the onset of a stochastic magnetic field and account for the observed confinement degradation, as expected for MTM.
Magnetic turbulence is observed in the edge of a wide range of ELMy H-mode DIII-D plasmas by the Faraday-effect polarimeter. Typically, the magnetic turbulence ranges from 100-500 kHz, peaking at 250 kHz. Magnetic fluctuations first appear ~2 ms after the ELM crash, grow quickly and then saturate before the next ELM event, correlating most closely with the evolution of pedestal electron temperature gradient. Density fluctuations in the same frequency range are observed simultaneously with similar temporal evolution between ELMs. BES shows the density fluctuation peaks in the pedestal steep gradient region, has wave number $k_\theta\sim0.3/cm$ ($k_\theta \rho_s\sim0.06$) and propagates in electron diamagnetic direction in plasma frame. Magnetic and density fluctuations measured by polarimeter at the same location show strong coherence up to 0.45 from 100-500 kHz, indicating they arise from the same instability.
Observations indicate the magnetic turbulence originates from electromagnetic instability. In a typical discharge, the line-averaged amplitude (averaged over the total chord length) of magnetic turbulence is 0.8 Gauss at the peak frequency and 15 Gauss integrated over the entire bandwidth (150-500 kHz). Corresponding lower-bound $|\delta b/B|$ is $4×10^{-5}$ and $8×{10}^{-4}$, respectively, and lower-bound $|\delta b/B|/|\delta n/n|$ is 0.08 and 0.15, respectively. These estimated values are comparable to that of electromagnetic instability from gyro-kinetic simulations (e.g. ref. 2).
The magnetic turbulence has been characterized and identified as MTM (Fig. 2). Theory predicts MTM are electromagnetic, propagate in electron diamagnetic direction with $k_\theta \rho_s\ll1$, driven by temperature gradient and saturate as $|\delta b/B|\simρ_e/L_{T_e},$ i.e. $\sim 1×10^{-3}$, all consistent with observations. Linear GENE calculation under the same experimental conditions find unstable MTM peak in the steep gradient region of the pedestal, with wavenumber and frequency quantitatively agreeing with the observed magnetic turbulence. The GENE simulation also shows non-monotonic collision frequency dependence of MTM growth rate, qualitatively consistent with the collision frequency dependence of turbulence amplitude in multiple shots (Fig. 3). The magnetic turbulence is always observed to peak off mid-plane, similar to MTM results from global GENE simulation$^3$.
This work provides the first direct measurement of internal magnetic fluctuations supporting the prediction of gyro-kinetic simulations that MTM exist in the H-mode pedestal. In addition, these measurements give critical constraints needed for model validation of pedestal predictions for future fusion devices. Work supported by the US Department of Energy under DE-FG03-01ER54615, DE-FC02-04ER54698 and DE-SC0018287.
References
1. M. Kotschenreuther et al., Nucl. Fusion 59, 096001 (2019)
2. W. Guttenfelder et al., Phys. Rev. Lett. 106, 155004 (2011)
3. D.Hatch et al., Nucl. Fusion 56, 104003 (2016)Speaker: Dr Jie Chen (University of California Los Angeles) -
One of the major challenges in magnetic confinement thermonuclear fusion research concerns the confinement, inside the reaction chamber, of the energetic particles (EPs) produced by fusion reactions and/or by additional heating systems, as, e.g., electron and ion cyclotron resonant heating, and neutral beam injection. In such experiments, EPs, having their velocities of the order of the Alfven velocity, can resonantly interact with the shear Alfven waves. In order to predict and, eventually, minimize the Energetic Particle (EP) transport in the next generation fusion devices, several numerical models, based on different theoretical approaches, have been developed. In this respect, it is crucial to cross verify and validate the different numerical instruments available in the fusion community. For this purpose, in the frame of the Enabling Research project MET [1], a detailed benchmark activity has been undertaken among few of the state-of-the-art codes available to study the self-consistent interaction of an EP population with the shear Alfven waves, in real magnetic equilibria in regimes of interest for the forthcoming generation devices (e.g., ITER [2], JT-60SA [3], DTT [4]). The codes considered in this exercise are HYMAGYC [5], MEGA [6], and ORB5 [7, 8], the first two being hybrid MHD-Gyrokinetic codes (bulk plasma is represented by MHD equations, while the EP species is treated using the gyrokinetic formalism), the third being a global electromagnetic gyrokinetic code (both bulk and EP species are treated using the gyrokinetic formalism). The so-called NLED-AUG [9] reference case has been considered, both for the peaked off-axis and peaked on-axis EP density profile cases, using its shaped cross section version. This case poses an exceptional challenge to the codes due to its high EP pressure, the rich spectrum of experimentally observed instabilities and their non-linear interaction [10].
Particular care has been devoted to consider plasma and numerical parameters as close as possible among the three codes: the same input equilibrium file (EQDSK) has been considered, ion density profile has been obtained by imposing quasi-neutrality ($Z_{i} n_{i} + Z_{H} n_{H}=n_{e}$), as required by ORB5 (here $n_{i}, n_{e}, n_{H}$ are the bulk ions, electrons, and EP densities (both bulk ion and EPs are assumed to be Deuterons), respectively, and $Z_{i}, Z_{H}$ their electric charge numbers); finite resistivity and the adiabatic index, $\Gamma=5/3$, have been assumed for both the hybrid codes (this is the usual choice used in MEGA, where also some viscosity is considered to help numerical convergence; note that HYMAGYC do not include viscosity).
Only finite orbit width (FOW) effects has been retained for now, and an isotropic Maxwellian EP distribution function of Deuterons with $T_{H}$=93 keV, constant in radius, has been considered.
Perturbations with toroidal mode number n=1 will be considered; the results of simulations considering both off-axis and on-axis peaked EP density profiles will be presented. First, simulation results referring to the linear growth phase will be considered.
For the peaked off-axis EP density profile case the three codes give very similar results (note that for MEGA, two MHD models are available, the 'Standard MHD' model and the 'Hazeltine-Meiss MHD'[11] model). The dominant drive comes from the positive gradient portion of the EP density profile, $0. \leq s \leq 0.4$ ($s \propto \sqrt{\psi}$ is the normalized poloidal flux function). The radial profile of the poloidal components of the eigenfuncion (see Fig.1), as obtained by the three codes compare quite well, the most unstable mode being located radially close to the magnetic axis, around $s \approx 0.2$, and in frequency within the toroidal gap.
Moreover, HYMAGYC and MEGA (both models) show a very similar growth-rate dependence on the ratio of EPs to bulk ion densities, $n_{H}/n_{i}$, while ORB5 exhibits some stronger dependence. Also, the results of varying the EP temperature will be considered.
Similar analysis have been performed for the peaked on-axis EP density profile case. Frequencies of the most unstable mode found by all codes have opposite sign, w.r.t. the off-axis case; eigenfunctions for HYMAGYC, MEGA-'Standard MHD' model and ORB5 are quite similar (they correspond to a mode located at $s \approx 0.4$, slightly inside the radial position where the $q$-profile has its minimum, $s \approx 0.5$), whereas the one shown by MEGA-'Hazeltine-Meiss MHD' model differs somehow. Growth-rates of MEGA are typically lower than the ones found by HYMAGYC and ORB5 (which are in reasonable agreement among them), and a more detailed analysis to understand such less favorable results are required. Weakly driven modes (with lower $n_{H0}/n_{i0}$ w.r.t. the nominal value) are also observed by HYMAGYC and MEGA-'Standard MHD' model, located within the toroidal gap, where the throat corresponding to $q(s)=2.5$ occurs, as already observed by HYMAGYC for the peaked off-axis EP density profile case.
Results of runs extending to the non-linear, saturation regime will also be shown, in order to benchmark these codes also in regimes where the EP transport can become relevant.Acknowledgment: This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. The computing resources and the related technical support used for this work have been provided by EUROfusion and the EUROfusion High Performance Computer (Marconi-Fusion).
References
[1] MET Enabling Research Project, https://www.afs.enea.it/zonca/METproject/index.html
[2] Aymar R. et al. 1997, FEC 1996, IAEA, Vol. 1, p.3
[3] JT-60SA Research Plan: http://www.jt60sa.org/pdfs/JT-60SA Res Plan.pdf
[4] DTT Interim Design Report, https://www.dtt-project.enea.it/downloads/DTT IDR 2019 WEB.pdf
[5] G. Fogaccia, G. Vlad, S. Briguglio, Nucl. Fusion 56 (2016) 112004
[6] Todo Y. and Sato T. 1998 Phys. Plasmas 5 1321–7
[7] Jolliet S. et al., 2007 Comput. Phys. 177 409
[8] Bottino A. et al., 2011 Plasma Phys. Control. Fusion 53 124027
[9] Ph. Lauber,“The NLED reference case”, ASDEX Upgrade Ringberg Seminar (2016), (Ph. Lauber et al.,
NLED-AUG reference case, http://www2.ipp.mpg.de/ pwl/NLED AUG/data.html)
[10] Ph. Lauber et al, EX1/1 Proc. 27th IAEA FEC (2018)
[11] R. D. Hazeltine and J. D. Meiss, “Plasma Confinement” (Addison-Wesley, Redwood City) p.222 (1992).Speaker: Gregorio Vlad (Associazione Euratom-ENEA sulla Fusione) -
Using modern deep neural network architectures, accurate disruption predictions on the DIII-D tokamak are made possible with the raw data from a single, high temporal resolution diagnostic which contains multi-scale, multi-physics sequences of events. This work illuminates a path forward to meet the disruption prediction requirements of devices such as ITER by using raw diagnostic data containing rich physics information without having to hand-craft features from these complex datasets generated by multiple tokamak diagnostics.
Excellent time-slice disruption prediction performance of 94% accuracy, and an F1-score of 91% (a metric indicating low false positives and false negatives rates) [Churchill2019] is achieved using a neural network architecture for multi-scale time-series, the Temporal Convolutional Network (TCN) [Bai2018] (see Fig. 1). This builds on other work using machine learning for disruption prediction [Kates-Harbeck2019, Rea2018, Ferreira2018], but in our case we show accurate performance training exclusively on a diagnostic previously unused due to its high sampling rate: the Electron Cyclotron Emission imaging (ECEi) diagnostic [Tobias2010] (see Fig. 2). The TCN is trained on a cluster of 16 GPUs over 2 days, on a subset of the available data, suggesting the need to incorporate High Performance Computing (HPC) resources to achieve the most accurate results.
These neural network disruption predictions were designed to be sensitive to pre-disruption markers up to 600ms before a disruption, giving sufficient time on DIII-D to avoid, not just mitigate, the disruption. This can then be tremendously beneficial as an accurate predictor for use in control systems to actively modify plasma actuators for avoiding disruptions. The TCN architecture can be adjusted and expanded depending on the required time sequence sensitivity, to adapt to newer machines, or different ranges of timescales.
The ECEi diagnostic is particularly well suited for disruption predictions due to its sensitivity to multiple time- and spatial-scales. There are a number of different root causes for disruptions, including edge radiation, too high density, and MHD instabilities [Boozer2012]. ECEi systems acquire data at such a high sampling rate (1 MHz) and spatial resolution ($k_\theta$ up to 2.1 cm-1) under ideal conditions that well-behaved signals span spatiotemporal scales to reflect the dynamics of turbulent fluctuations, Alfvén eigenmodes, tearing modes, sawteeth, ELMs, and other potential pre-disruption markers. Non-ideal conditions also impact the signal in ways that can be difficult for a human diagnostician to interpret, but are rich in information. For example, a sudden loss of signal can be the result of density cutoff. Alternatively, a sudden spike in signal can be the result of a non-thermal electron distribution. A wide range of other conditions can impact the signal, producing fluctuations or other features that machine learning techniques might become sensitive to, even when the human data analyst finds them to be troublesome or ambiguous. By using raw, un-processed ECEi data, we make full use of all these signal features.
These deep learning techniques show promise to greatly enhance the current machine learning worldwide effort being done on disruption predictions, one of the most pressing issues for the success of the tokamak concept. They open up the possibility of expanding the number and types of diagnostics used in disruption predictions, accurately including many different physics processes in these predictions. In the future, the correlations between channels and diagnostics that these networks learn on one machine can potentially be transferred and fine-tuned for a newer machine, requiring much less data from the newer machine then needed to originally train the models.
This work was supported in part by the US Department of Energy under AC02-09CH11466, DE-FC02-04ER54698, and FG02-99ER54531
[Churchill2019] R.M. Churchill, the DIII-D team, 2nd Workshop on Machine Learning and the Physical Sciences (NeurIPS 2019), arXiv:1911.00149
[Bai2018] S. Bai, et al., arXiv:1803.01271
[Kates-Harbeck2019] J. Kates-Harbeck, A. Svyatkovskiy, W. Tang, Nature 568 (2019) 526–531.
[Rea2018] C. Rea, R.S. Granetz, Fusion Sci. Technol. (2018) 1–12.
[Ferreira2018] D.R. Ferreira, ArXiv E-Prints ArXiv:1811.00333 (2018).
[Tobias2010] B. Tobias, et al., Rev. Phys. Instrum., 81 (10) (2010) 10D928
[Boozer2012] A. Boozer, Phys. Plasmas, 19 (5) (2012) 058101.Speaker: Randy Churchill (Princeton Plasma Physics Laboratory) -
Recent analysis leveraging the broad array of measurable plasma parameters on the DIII-D tokamak has been used to elucidate the physics underlying detachment processes in the divertor and reveal the 2D nature of detachment important for design of detachment scenarios for next step devices. The dominant role of EUV/VUV radiation for radiative power exhaust has been established experimentally with accompanying spectroscopy leveraged alongside collisional radiative modeling to calculate the impurity density and charge-state distribution in the divertor. 2D measurement of critical plasma parameters for power exhaust studies ($n_e$, $T_e$, P$_\text{rad}$, $E_\text{VUV/EUV}$) reveal a greater radial emission extent compared to UEDGE fluid modeling simulations. This larger extent provides opportunity for greater dissipation volume, but also further demonstrates that fully-2D simulations including cross-field drifts are required for detachment studies working towards a predictive capability of divertor heat loads.
A combination of EUV/VUV-VIS-IR spectroscopy, ColRadPy collisional radiative modeling (1), and 2D $T_e$ and $n_e$ measurements from Divertor Thompson Scattering has been used to infer impurity densities in the divertor. This analysis primarily uses the EUV/VUV resonance lines that make up the vast majority of radiative emission ($\sim$>95% (2)) and are particularly well suited for determining ground-state densities. Inter-ELM intrinsic carbon impurity fraction was found to be $\sim$5% in attached H-mode conditions, falling to $\sim$0.5% in detached conditions while maintaining about the same total carbon density. UEDGE modelling with a full physics drift model similarly shows a reduction in impurity concentration in detachment but limited to a $2.8 \times$ drop. Using the same set of calibrated EUV/VUV spectroscopy measurements, the carbon population in unseeded discharges is inferred to be dominated by C$^{4+}$ in the divertor whereas detached plasmas show highly radiating narrow bands of C$^{2+}$ and C$^{3+}$ at the detachment front. Figure 1 shows these charge state distributions for a partially detached plasma with $\mathbf{B} \times \nabla B$ drift towards the primary X-point in DIII-D’s ‘shelf’ open divertor (4.5MW, 1.8T, 0.9MA) alongside the associated UEDGE predictions. In nitrogen seeded detached cases the additional available charge state results in a slightly increased range of radiating species (N$^{2+}$ to N$^{4+}$ with $\sim$2eV of additional $T_e$ range) in a regime dominated by N$^{3+}$ and N$^{4+}$ ions. The charge state distribution comparison with UEDGE modeling displays quantitatively similar 2D profiles to those experimentally inferred albeit with an additional charge-state mixing caused by the finite lifetime of ions and transportation via parallel flows that are not accounted for in the CR model. Quantitative 2D comparison between UEDGE-predicted and measured flows has recently been achieved using velocity imaging (3). An excellent agreement of He$^+$ velocities in a pure helium L-mode plasma is achieved near the divertor target where He is the main-ion species and electron physics dominates. Further upstream where ion-dominated physics plays a more important role, a factor of 2–3 underestimation of the velocity is observed indicating an underestimation of the role of ions in determining local plasma characteristics near the X-point that impacts our ability to predict impurity transport via parallel flows in the divertor, estimate convective power fraction in detached conditions, and establish total pressure dissipation.
The radial extent of the radiative volume in detached H-mode discharges has been shown to display much broader features in detached H-mode discharges compared to UEDGE fluid modeling (5.2MW, 0.9MA, 1.8T, $\mathbf{B} \times \nabla B$ drift towards the primary X-point) (4) with an increasing level of broadening observed at higher powers (5). This is observed in both charge-state resolved line emission (Figure 2) as well as total radiated power (bolometry), Divertor Thomson Scatting, and 2D visible imaging. UEDGE simulations with drifts and currents show that in these conditions the poloidal $\mathbf{E} \times \mathbf{B}$ drift can dominate the poloidal heat transport in the radiative front, expanding the poloidal extent of the radiation front as well as increasing the total radiative power. This indicates drift flows leading to a larger volume for dissipation and enhanced ability for divertor radiation than predicted by more commonly used 1D and 0D modeling approximations, or 2D modeling without drifts. This directly impacts our ability to predict detachment onset, detachment stability, the impurity fraction required to achieve detachment, and the heat flux mitigation that can be expected in planned divertors.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-AC52-07NA27344, and DE-NA0003525.
(1) Johnson et al., 2019 Nuclear Materials and Energy 20 100579
(2) Mclean A.M. et al., 2018 IAEA FEC 2018 EC/PC-15; Mclean A.M. et al., 2020 Plasma Surface Interactions Conference (upcoming)
(3) Samuell C.M. et al., Phys. Plasmas, 25 056110
(4) Jaervinen A.E. et al., 2019, Contrib. Plasma Phys; Jaervinen A.E. et al. 2020 NF (submitted)
(5) Leonard A.W. et al, 2020, IAEA 2020 (this meeting)Speaker: Adam McLean (Lawrence Livermore National Laboratory) -
In tokamak discharges there are often saturated Alfven modes (1). They are destabilized by gradients of the high energy particle population, so are to be expected in discharges with a significant alpha particle population, such as expected in ITER (2) or any fusion device. These modes may produce only a small local flattening of the particle distribution, or if the number and amplitude be large enough, cause particle loss when the resonances associated with the modes overlap sufficiently to cause chaotic pathways for redistribution (3).
There are two types of events in which the \alfven modes have rapid significant effect on the high energy particle distribution, chirping modes and avalanches. A chirp is a complex frequency modulation of a single mode, causing also local particle distribution modification (4,5). An avalanche is a combination of a few modes to produce a large scale cascade of particles, leading to particle loss and even to discharge termination (6). These phenomena have been simulated using the guiding center code ORBIT (7) using a $\delta f$ formalism (8,9,10), which is able of capturing details of large phase-space structures. While not able to follow the large scale modification of the particle distribution or nonlinear mode coupling dynamics involved in the final stages of these events, this formalism is adequate to investigate the conditions for the onset of these phenomena and the initial behavior, and hence provide information as to how they can be avoided. Simulations of an NSTX (11) chirp and an avalanche are shown in Figs. 1 and 2.
An energetic particle driven \alfven mode has energy transferred from the particle distribution to the mode through distribution flattening within a resonance, producing the linear growth rate $\gamma_L$, accomplished by the fine scale phase mixing in the resonance. Once the distribution has been flattened, without collisional replenishment of the density gradient, mode drive ceases. In addition, a mode can be strongly damped due to the continuum, trapped particle effects, Landau damping and radiation, with total damping $\gamma_d$. We find that chirping occurs when a mode experiences small collisions and strong damping, such as could be due to a small equilibrium change causing contact with the continuum. A chirp of an Alfven mode involves complex particle organization so as to extract energy from the particle density gradients at the edge of the resonance island, producing a particle clump and a hole and frequency sidebands departing from the original mode frequency.
Avalanches depend on the overlap of resonances located so as to provide a path for large scale particle redistribution. Thus the possibility of avalanche depends on the nature and magnitude of the resonances associated with the modes. If this path extends to the plasma edge, particle loss can occur. During a discharge, the drive of the modes can be increasing due to changes in the high energy particle distribution. In addition, we find that the size of the resonances increases with decreasing Alfven mode frequency, inversely proportional to the square root of the plasma density. Thus the increase in plasma density and the increase in high energy particle density can both contribute to the onset of an avalanche. We show that the location and nature of the existing resonances can be determined using the equilibrium and particle distribution data, and hence the possibility of avalanche and possible means of avoidance, without extensive numerical simulation.
Acknowledgement: This work was supported by the US Department
of Energy (DOE) under contract DE-AC02-09CH11466.References:
(1) C. Cheng, L. Chen, and M. Chance, Annals of Physics 161, 21 (1985).
(2) The iter project, http://www.cad.cea.fr/gb/activites/fusion/ITER.php.
(3) C.S.Collins,W.W.Heidbrink,andetal(DIII-Dteam),Phys. Rev. Lett. 116, 095001 (2016).
(4) E. Fredrickson, R. Bell, D. Darrow, and et al, Phys. of Plasmas 13, p056109 (2006).
(5) W. W. Heidbrink, Plasma Phys. Control. Fusion 37, 937 (1995).
(6) E. Fredrickson, N. Crocker, D. Darrow, and et al, Nuclear Fusion 53, 013006 (2013).
(7) R. B. White and M. S. Chance, Phys. Fluids 27, 2455 (1984).
(8) R. B. White, The Theory of Toroidally Confined Plasmas (Imperial College Press, 2014).
(9) R. B. White, V. N. Duarte, N. N. Gorelenkov, and et al, Phys. of Plasmas 26, 092103 (2019).
(10) R. B. White, V. N. Duarte, N. N. Gorelenkov, and et al., Phys. of Plasmas x, x (2020).
(11) S. M. Kaye, M. Bell, R. Bell, S. Bernabei, J. Bialek, and et al, Nucl. Fusion 45, S168 (2005).
(12) H. L. Berk, B. N. Breizman, and V. Petviashvili, Phys Lett A 6, 3102 (1999).Speaker: Roscoe White (Princeton University) -
A portable and interpretable data-driven algorithm for disruption prediction has been developed and installed in DIII-D and EAST Plasma Control Systems (PCS). The Disruption Prediction via Random Forest (DPRF) algorithm guarantees explainable predictions in real-time thanks to the feature contribution analysis [Rea2019], which is able to identify the main drivers of the disruptivity – the probability of an impending disruption – potentially informing the PCS on the proper actuators. DPRF was tested in closed-loop mitigation experiments on both DIII-D and EAST tokamaks. On DIII-D, indicators of plasma temperature, density and radiation based on profile peaking factors [Pau2018] were successfully implemented in real-time and used as input features in DPRF during recent experiments.
DPRF is based on the Random Forest machine learning algorithm, which estimates the probability of an impending disruption, i.e. disruptivity, by developing a large number (~ hundreds) of independent, de-correlated base learners (models), thus collecting a parallel set of predictions. The final prediction from the ensemble is aggregated by averaging this very large number of models’ predictions. A calibrated threshold on the disruptivity is then found to maximize the success rate and warning time [Montes2019]. While it is necessary to provide enough warning time to predict an impending disruption, it is crucial to also diagnose what type of event in the disruptive chain is causing the deviation from the safe operational space. This allows the PCS to respond adequately to different instabilities by directing it to trigger different actuators. 1D/2D profiles of plasma temperature, density and radiation contain highly predictive information that can be used for example to diagnose impurity accumulation events leading to radiative collapse and therefore disruptions. Reducing profiles to their peaking factors (following what documented in [Pau2018]) handles the curse of dimensionality and represents a manageable and real-time compatible way for a data-driven algorithm to augment predictive capabilities. In addition to these synthetic diagnostics, other 0D (where time is their only dependence) and mainly dimensionless input features were used to train the machine learning-based algorithm on DIII-D, consisting in data coming from equilibrium reconstruction codes or magnetic probes.
On DIII-D, DPRF was tested in ITER Baseline Scenario plasmas, where closed-loop performances were verified by successfully integrating the disruptivity in the Off Normal Fault Response algorithm [Eidietis2018] for emergency response. In different discharges, the Massive Gas Injection (MGI) was triggered in response to dangerous tearing mode activity, the Electron Cyclotron Heating (ECH) system was aimed at q=3/2 surface when such precursor emerged, and an early shutdown via fast ramp down was triggered also due to the presence of a tearing mode (see Fig. 1a). Additionally, the peaking factors were verified to represent a good metric for impurity accumulation events, when probing via Argon injection – see Fig. 1b. The successful integration of DPRF in DIII-D PCS is part of a broader approach to qualify advanced disruption prevention strategies to address ITER’s critical needs. More details about the “Disruption Free Protocol” can be found in the contribution by J. Barr et al IAEA-FEC 2020.
On EAST, DPRF was tailored to predict high-density scenarios and closed-loop performances were tested during recent mitigation experiments, where several disruptions were mitigated via MGI. In Fig. 2 we show one example of a real-time experiment in January 2020, where predictions and interpretations were provided in around 200 microseconds computing time. EAST discharge 94520 shows symptoms of a density-limit disruption (high contribution from the Greenwald density fraction) but eventually disrupts due to a radiative collapse caused by impurities in the plasma (dominance of loop voltage contribution at the end). The developed algorithm still presents some residual flaws as it shows to be prone to false alarms when H-L back-transitions occur during the discharge. More work will be required to also implement analogous peaking factors synthesized from EAST diagnostic systems, crucial to define a device-independent framework for properly diagnosing disruption precursors.
Current results represent a step forward in data-driven control for scenario optimization and disruption avoidance for ITER and next generation devices. This work establishes the importance of developing tools capable of identifying and informing in real-time the PCS on the dangerous plasma parameters deviations to the disruptive space to enable the proper actuators’ response. Effort is ongoing to map the feature contributions to the actions of the available actuators, but also to define a robust framework to transfer the knowledge gathered using data from existing experiments to unseen devices.
This work was supported in part by the US Department of Energy under DE-FC02-99ER54512, DE-SC0014264, DE-SC0010720, DE-SC0010492, DE-FC02-04ER54698 and by the National MCF Energy R&D Program of China, Grant No. 2018YFE0302100.
[Rea2019] C. Rea et al 2019 Nucl. Fusion 59 096016
[Pau2018] A. Pau et al 2018 IEEE TPS 46 7 2691
[Montes2019] K.J. Montes et al 2019 Nucl. Fusion 59 096015
[Eidietis2018] N.W. Eidietis et al 2018 Nucl. Fusion 58 056023Speaker: Cristina Rea (Massachusetts Institute of Technology) -
New studies identify the critical parameters and physics governing disruptive neoclassical tearing mode (NTM) onset. A m/n=2/1 mode in DIII-D begins to grow robustly only after a seeding event (ELM Fig. 1, or sawtooth precursor and crash Fig. 2) causes the mode rotation to drop close to that of the plasma’s Er=0 rest frame; this condition opens the stabilizing ion-polarization current “gate” and destabilizes an otherwise marginally stable NTM. Our new experimental and theoretical insights and novel toroidal theory-based modeling are benchmarked and scalable to ITER and other future experiments. The nominal ITER rotation at q=2 is found to be stabilizing (“gate closed”) except for MHD-induced transients that could “open the gate.” Extrapolating from the DIII-D ITER baseline scenario (IBS) discharges, MHD transients are much more likely to destabilize problematic 2/1 NTMs in ITER; this makes predictions of seeding and control of both ELMs and sawteeth imperative for more than just “simply” minimizing divertor pulsed-heat loading.
While nearly steady state in betaN, li and rotation, the classical tearing stability index Delta’r0 in DIII-D may evolve slowly. IBS discharges (ITER similar shape, H-mode with ELMs and sawteeth, betaN~2 and q95~3) were run with 1 msec faster resolution CER (standard 5 msec) of toroidal and poloidal rotations. Discharges can exhibit a rotating m/n=2/1 magnetic perturbation, in response to sequential ELMs and sawteeth, that evolves from marginal to robust growth into an eventual locked mode and disruption. Key conditions for algebraic linear temporal growth include an MHD-induced transient with a large enough magnetic perturbation Brms and mode rotation change. Figures 1 and 2 show analyses of NTM seeding events with the fast CER that have been examined in greatest detail, among the multiple discharge broader database.The key stabilizing factor depends on the relative rotation [parameterized by a gate function F(fm)≤1] between the mode rotation frequency fm and the Er=0 frame of the plasma fE being non-zero but not more than that of the (positive in DIII-D) ion diamagnetic mode frequency f*i ~ 1 kHz in Figs. 1 and 2. The mode island width growth rate dw/dt is modeled with a modified Rutherford equation (MRE) along with models for changes in both mode and plasma rotation:
$\begin{equation} \frac{dw}{dt}=D_{\eta}\left[\frac{d_{NTM}}{w}-\frac{w^2_{pol}}{w^3}F(f_m)\right],\;\;F(f_m) \equiv -4\frac{(f_m-f_E)(f_m-f_E-f_{*i})}{f^2_{*i}} \;\;\;\;\;\;\;\;\;\; \end{equation}$
Here, the classical tearing stability index is negligible, i.e., Delta’r0=-0.1. The critical island for growth is w0=wpol(F/dNTM)1/2 where dNTM is bootstrap drive and wpol is 2X the ion banana width wib. A toroidal adaptation of recent slab-model theory (Beidler 2018) predicts MHD transients abruptly induce a 2/1 tearing response, radially local torque δJ_∥ δB_x, radial electric field and flows that reduce the relative mode frequency. The mode rotation dynamics is described by
$\begin{equation} \frac{df_m}{dt}\cong-\frac{f_m-f_t}{\tau_\zeta}-\frac{f_m-f_E}{\tau_w}+\delta J_\parallel\delta B_x,\;\; \tau_\zeta\cong\frac{a^2_{eff}}{4D_\mu},\;\;\frac{1}{\tau_w}\cong C_w\frac{v_{Ti}}{Rq}\frac{w}{r_0} \;\;\;\;\;\;\;\;\;\; \end{equation}$
Fits to experimental data (for an ELM-induced NTM in Fig. 1 as an example) capture the experimental behavior (pink lines in Fig. 1) of evolution from marginal to robust 2/1 growth. The stabilizing gate factor F depends on the relative rotation and plunges during an ELM, either recovering or remaining down for large enough mode amplitude. In addition to theory, the NIMROD code is used to study evolution at the 2/1 surface in response to an ELM. The code uses an extended MHD model with heuristic closures to model the electron and ion neoclassical parallel stresses. NIMROD indicates the Fig. 1 DIII-D equilibrium is stable to classical tearing modes but a pulsed MHD perturbation at the computational boundary can kick off a 2/1 mode.
Applying a MHD transient torque to the situation in Fig. 3 can drive relative rotation down and open the gate, as in Fig. 4. The predicted island growth rate in ITER is slower due to its much smaller magnetic field diffusivity but shifted to smaller w0 and very much smaller (0.1X) relative size w0/r0. The IBS equilibrium in DIII-D is very similar to what is modeled (Polevoi 2006, 2019) for ITER. Similar j and q profiles imply comparable classical tearing stability Delta’r0=-0.1 in ITER which is neglected in the MRE of Eq. 1. The ratio of q=2 bootstrap to equilibrium current density (~d_NTM) is also similar. At critical island w0 for F≅1, DIII-D mode rotation fm is about 0.5Xfi=0.6 kHz above fE (Fig. 3); ITER is similar with fi of 0.165 kHz.
This new work gives experimental and theoretical insights, as well as novel benchmarked toroidal theory-based modeling, to a longstanding uncertainty in projecting how NTMs are triggered (Buttery 2007, Hender 2007) for scaling to ITER and beyond. It also provides the framework (e.g., for real-time monitoring) to develop criteria for transient-MHD-induced excitation and robust growth of 2/1 NTMs that can lead to problematic locked modes and disruptions in burning plasma tokamaks, and for which experimental data is limited.This work was supported in part by the US DOE under DE-FC02-04ER54698.
Speaker: Dr Robert La Haye (General Atomics) -
The dynamics of fast electrons driven inductively or by resonant interactions with radio- frequency waves is known to be highly sensitive to the presence of impurities in hot magnetized hydrogen plasmas. The possibility to use tungsten for the ITER divertor, thanks to its low tritium retention and high melting temperature, has raised the question of the impact of partially ionized high-Z atoms on current drive efficiency by enhancing pitch-angle scattering but also collisional slowing-down. Pioneering work on the impact of the partial screening effect in kinetic calculations was carried out primarily for the problem of runaway electron mitigation in very cold post-disruptive plasmas [1]. In this paper the approach is adapted and extended to regular plasma regimes, allowing to take into account any type of high-Z metallic impurity in the plasma core on the fast electron dynamics. In addition, the enhancement of non-thermal bremsstrahlung by partially ionized high-Z atoms in the plasma is calculated.
Effect of partial screening is investigated in the framework of the Born approximation by calculating the usual form factor to account for the spatial extent of the high-Z ion using atomic electron densities deduced from simplified models (Thomas-Fermi, Yukawa potential) or from the Density Functional Theory (DFT) describing accurately many-body exchange and correlation interactions. In order to reduce computational effort either for kinetic calculations or bremsstrahlung emission, analytical formulas of the form factor deduced from Thomas- Fermi and Yukawa potential atomic models are used, with effective ion sizes determined for each ionization state by a best fit of the form factor deduced numerically from quantum relativistic calculations using the GAUSSIAN chemistry software package [2]. While Thomas- Fermi model form factor gives better results when high-Z atoms are weakly ionized, Yukawa potential model turns out to be more appropriate when the screened ion charge is greater than Z/3, where Z is the atomic number, a condition encountered for tungsten in standard core tokamak plasma conditions where electron temperature reaches few keV.
From the spinless relativistic Rutherford elastic scattering cross-section, the modified pitch-angle collision operator is used in Fokker-Planck calculations, taking into account of the partial screening for each species and all corresponding ionization states. It is proportional to the factor (Z-N)2lnΛ(p)+g(Z-N,p), where lnΛ(p) is the usual momentum-dependent Coulomb logarithm, N, the number of bound electrons, and g an analytical function describing the enhanced pitch-angle scattering by inner populated atomic shells. When g is small as compared to the usual Coulomb logarithm term, screening effect on pitch-angle is small. Inelastic scattering resulting from the mean excitation of partially-ionized high-Z atoms is also considered from the Bethe formula for the electron relativistic stopping power.
For non-thermal bremsstrahlung, Yukawa potential model is usually preferred for describing screening effect of bounded electrons, since inner atomic shells contribute significantly, whatever the ionization state [3]. In this case, an original semi-analytical formula is derived for the doubly differential quantum relativistic cross-section in photon energy and in photon emission angle from the most general Bethe-Heitler bremsstrahlung cross-section [4], which greatly enhances calculation speed, while keeping a high numerical accuracy. It is shown that bremsstrahlung scales like Z^2, with an enhanced reduction factor as the ratio k/Ec decreases, where k is the photon energy and Ec the incoming fast
electron kinetic energy (Fig. 1). Screening effects tends to disappear progressively when the angle between the photon emission and the incoming electron velocity increases.
Consequences on the current drive efficiency have been investigated using the kinetic solver LUKE of the 3-D linearized relativistic bounce-averaged electron Fokker-Planck equation [5] and on the fast electron bremsstrahlung using the quantum relativistic radiation code R5-X2 [6]. Thermal ionization states for all species are determined by ADAS code [7]. A full simulation of the high-power WEST tokamak discharge #55539 is investigated, where most of the plasma current is driven by the Lower Hybrid wave, taking into account of the tungsten level deduced from radiative power losses using the METIS tokamak code [8]. With an estimated concentration of tungsten of 4x10-4, it is shown that the reduction of the LH driven current is about 4%, while conversely, bremsstrahlung is increased by a factor 3 approximately as compared to the fully screened ion case. From simulations of high-power WEST tokamak Lower Hybrid discharges [9], the general impact of partially ionized metallic impurities on RF current drive is discussed, as well as on fast electron bremsstrahlung diagnosis capability.
Acknowledgements. This work has been partially funded by National Science Centre, Poland (NCN) grant HARMONIA 10 no. 2018/30/M/ST2/00799. We thank the PLGrid project for computational resources on the Prometheus cluster.
References
1 L. Hesslow, et al., Phys. Rev. Letter 118 (2017) 255001
2 M. J. Frisch, et al., Gaussian 09, Revision E.01, Gaussian, Inc., Wallingford CT, 2016. 3 M. Lamoureux and N. Avdonina, Phys. Rev. E, 55 (1997) 912
4 H. W. Koch and J. W. Motz, Rev. Mod. Phys., 31 (1959) 920
5 Y. Peysson and J. Decker, Fusion Science and Technology, 65 (2014) 22
6 Y. Peysson and J. Decker, Phys. Plasmas, 15 (2008) 092509
7 H. P. Summers, the ADAS User Manual, version 2.6 http://www.adas.ac.uk (2004)
8 J.F. Artaud, et al., Nuc. Fusion 58 (2018) 105001
9 C. Bourdelle, et al., Nucl. Fusion 55 (2015) 063017Speaker: Yves Peysson (CEA) -
ELM control and mitigation efforts are of highest priorities for ITER and beyond. However, often such efforts with RMPs and pellets do not focus on the pedestal turbulence and transport and their effect on ELM characteristics. Detailed analysis of pedestal turbulence as a function of the heating mix may provide crucial insights as well as additional handles for ELM control. It is found that recovery of pedestal pressure gradient in the inter-ELM phase is delayed and magnetic and density fluctuations increase along with excitation of several quasi-coherent modes in the ECH dominated discharges in DIII-D. As a result, ELM frequency decreases by 40% when heating is changed from pure NBI to predominantly ECH. This would be consistent with our hypothesis that turbulence driven transport increases with these quasi-coherent modes and thus keeping the pedestal away from ELM threshold for a longer duration of time, as shown in Fig. 1(a). In this study, localization of these modes at the pedestal is confirmed. Growth of these modes has some correspondence with the steepening of
$\nabla T_{e}$at the pedestal with ECH injection. That these mode activities may lead to enhanced transport at the pedestal steep gradient region in the inter-ELM period is supported by the baseline of the D$_{\alpha}$signal, which indicates increased particle flux.
Fig. 1 (right) shows the spectrogram of density fluctuations from a Doppler Backscattering (DBS) diagnostics channel at the steep gradient region ($\rho$= 0.95) from a candidate ECH shot. It can be seen that there are two modes evolving in the inter-ELM period: following the ELM crash, a low frequency (400 kHz) quasi-coherent mode (LFQC; black oval) is observed in the$\nabla T_{e}$phases #1b and #4. It is apparent that the LFQC emerges and survives in a narrow range of values of$\nabla T_{e}$, roughly bounded by the two horizontal broken blue lines on$\nabla T_{e}$. On the other hand, high frequency (~2 MHz) broadband fluctuations (HFB; white oval) dominates in phases #2 and #5 leading towards the small Dα spike and the large type-I ELM respectively. Whenever, the LFQC is present (phases #1b and #4), the D$_{\alpha}$baseline is enhanced indicating increased particle flux. Based on this evolution of$\nabla T_{e}$the occurrence of these two modes appears consistent with affecting transport and thereby frequency and amplitude of the ELMs. In phase #2, the gradients have reached the threshold for ELMs to occur, while the pedestal heights, especially$n_{e}$and hence,$p_{e}$pedestal heights, are still low and evolving. Hence, one or two small spikes in D$_{\alpha}$are observed in between two consecutive large type-I ELMs.A similar correlation is observed between magnetic fluctuations and
$\nabla T_{e}$, as seen in the DBS data. A distinct group of three modes (13~116 kHz) is seen in magnetic fluctuations in Fig. 2. ELM-synced analysis shows that these modes briefly grow and die down in the first 2-7 ms of the inter-ELM period. After that the modes grow again from 13 ms onwards and saturate at ~25 ms. The initial growth bump at ~5 ms and the growth from 13 ms are in$\nabla T_{e}$phases #1b and #4, similar to the growth patterns of the LFQC in DBS, as shown in Fig. 1.
The heat and particle diffusivities are calculated from the measured profiles in TRANSP. The electron particle diffusivity ($D_{e}$) is much less than the electron heat diffusivity ($\chi_{e}$) in the steep gradient region. As per the recently proposed fingerprint analysis approach [M. Kotschenreuther et al., Nucl. Fusion 59, 096001 (2019)], this suggests that the transport is possibly of the nature of$\nabla T_{e}$driven, dominated by TEM and MTM. TGLF simulations show that the linear growth rate of the most dominant mode peaks at$k_{\theta}$ρs ~ 0.4 at the steep gradient region. Corresponding frequency is in the electron direction over the entire pedestal for$k_{\theta}$ρs < 1.5, indicating that TEM and/or MTM could be important. This study may provide vital inputs towards understanding inter-ELM pedestal recovery in varied transport regimes and predicting/optimizing pedestal performance and ELM behavior in future reactor grade plasmas. This work is supported by US DOE under DE-SC0019302, DE-FG02-08ER54999, DE-FG02-08ER54984, DE-AC02-09CH11466 and DE-FC02-04ER54698.Speaker: Dr Santanu Banerjee (College of William and Mary) -
We present novel techniques for fast-ion modelling that allow more extensive studies and support orbit-following modelling [1], and we apply those to study fast-ion transport in ITER.
While orbit-following Monte Carlo simulations are frequently used to make predictive estimates for fast-ion losses and wall loads, these simulations have the drawback that they are computationally expensive to perform. Furthermore, orbit-following tools are based on first principles and, as such, it is difficult to interpret which processes are responsible for the losses seen in simulations. This contribution addresses these issues.
It is shown that the collisionless transport of fast ions can be treated as an advection-diffusion process, where the transport coefficients can be evaluated with an orbit-following model. This results in a significant (a factor of 50 - 100) decrease in time that is required to estimate fast ion losses during the slowing-down process with an orbit-following model alone as the coefficients can be evaluated within just a few bounce times. With this approach it becomes possible to make an extensive study of fast-ion losses due to ITER ELM control coils using different current configurations. We vary the mode and poloidal phase of the coils and find the configurations where the losses have their minimum and maximum as shown in Fig.3. However, the study performed here is done in vacuum approximation while plasma response is needed for more accurate studies [2].
Another technique that is presented is the so-called loss maps where the birth position of particles, which are lost due to a 3D magnetic field, are mapped to a constants of motion phase space. We show that this mapping allows one to make the connection between different loss-mechanisms and the losses, thus increasing the confidence in one's results. Furthermore, we show that the particle birth position also predicts to which location on the wall it will be lost to. However, the main benefit of using the loss maps is that they can be constructed solely using the analytical formulas for different transport processes as exemplified in Fig.4. This provides a basically instant way of estimating losses, thus avoiding the need to resort to more time-consuming orbit-following simulations. The use of loss maps is demonstrated for ITER baseline and reduced field scenarios in the presence of various stationary magnetic field perturbations.
[1] Särkimäki, K. (2020). Efficient and rigorous evaluation of fast particle losses in non-axisymmetric tokamak plasmas. Nuclear Fusion, 60(3), 036002.
[2] Varje, J., et al (2016). Effect of plasma response on the fast ion losses due to ELM control coils in ITER. Nuclear Fusion, 56(4), 046014
Speaker: Konsta Särkimäki -
In this paper we report on experimental and modeling results concerning the energetic particle (EP) dynamics in plasma scenarios with off-axis neutral beam (NB) injection at ASDEX Upgrade (AUG). The tools validated in this processes are applied to selected scenarios at JT-60SA and ITER pre-fusion plasmas.
Off-axis NB injection is an important tool to control and optimise the current profile in both conventional and advanced tokamak scenarios. Via tailoring the safety factor profile, rational surfaces can be avoided, the local magnetic shear can be changed or reversed shear regions can be established. Whereas in present devices the beam energies are typically 10-20 times larger than the plasma thermal energies and smaller than the Alfvén velocity (vNBI /vA ∼ 0.3 − 0.4), in future devices such as JT-60SA and ITER these ratios will go up to 100 for vNBI/vthermal and to vNBI/vA ~ 1. Thus, it is expected that the related EP-driven instabilities and the relaxation of the spatial EP pressure gradients will be different (e.g. mode number spectrum, non-linear saturation) than in present-day experiments.The related EP transport is directed to deplete the gradients, i.e. inwards in the positive gradient region and outwards in the negative gradient region. Since the redistribution of the EP beam will affect the background plasma properties through various channels, it is of interest to analyse the EP redistribution and to test if stability predictions and related EP transport calculations are able to catch the experimental signatures and thus can be used with confidence in future comprehensive scenario simulations.
In 2017 a new scenario on ASDEX Upgrade has been established for the dedicated investigation of energetic particle (EP) physics [1] that is optimised to maximise βEP/βth and the ratio vNBI/vthermal. This scenario has been recently further developed into both an L-mode and an H-mode scenario with stable flat-top phases and with more complete diagnostic coverage. As in the previous discharges, we let impurities (mainly tungsten) accumulate in the core. Due to strong radiation losses the background temperatures and pressures of both ions and electrons stay low, despite 2.5 − 5 MW NB heating. In order to avoid transient q = 2 sawtooth-like crashes, as seen previously with a total plasma current of 800kA, the current has been reduced to 700kA. Under these conditions we reach an EP-β comparable to the background β and ratios of 100 and larger for vNBI/vthermal, whereas vNBI /vA ∼ 0.3 − 0.4. A rich spectrum of modes is destabilised: EP-driven geodesic acoustic modes (EGAMs), beta-induced Alfvén eigenmodes(BAEs), reversed shear Alfvén eigenmodes (RSAEs) and toroidal Alfvén eigenmodes (TAEs). In particular the EGAM onset is triggered by TAE bursts indicating a coupling of these modes via EP phase space transport. Bicoherence analysis also reveals non-linear coupling signatures between various frequency bands. Comparisons of FIDA-measured EP profiles and neoclassical calculations are shown, and indications for anomalous background ion heating via EP-driven instabilities are investigated. Using the linear gyro-kinetic code LIGKA [2], the onset conditions for various instabilities are analysed, as the experiment provides excellent data in this respect. Experimental mode properties are determined and compared to the LIGKA results. In particular, symmetry breaking signatures of non-perturbative mode structures [4] are investigated. Non-linear hybrid simulations (HAGIS/LIGKA [2, 3]) are carried out in order to quantify the difference between experimental measurements and this model. By comparing to other models such as non-linear hybrid kinetic MHD models and gyrokinetic codes, the importance of e.g. non-linear wave-wave coupling processes can be inferred. This analysis is linked [5] to a benchmark and validation exercise [6] including the codes HYMAGIC, MEGA and ORB5, that is based on this AUG case (see also refs in [6]).
The exploration of scenarios with off-axis NB deposition leading to non-inductive steady-state op- eration at high β is one of the main missions of the JT-60SA project starting operation in 2020 [7]. The high-energy negative ion sources (~500 keV) at JT-60SA deposit exclusively off-axis. An exhaustive kinetic-hybrid MHD analysis using the MEGA code has been performed in refs. [8]. Building on these results further gyro-kinetic analysis using the LIGKA/HAGIS tool is carried out (higher mode numbers, EGAM/BAE thresholds) in order to investigate the scaling of various parameters connected to linear onset and non-linear EP transport compared to the AUG case discussed above.
During the lifetime of ITER various scenarios with off-axis NB injection are forseen. Although a change of the beam geometry from on-axis to off-axis will be possible, it cannot be performed frequently since the cycles are limited due to the mechanical stress it induces on the various compo- nents connected to the beam source. For this reason a good understanding of the expected heating characteristics and deposition properties including a possible deposition broadening due to EP- driven instabilities can help to optimise the planning and operation of the experiments. Based on the fully IMAS [9] integrated heating and current drive workflow [10] the stability of pre-fusion H-plasmas is investigated for different beam deposition locations (off-axis, mixed on/off-axis). The analysis will also serve as a verification test of the recently IMAS-integrated LIGKA/HAGIS workflow (see fig. 1). The results will be related and compared to the findings on AUG and JT-60SA.Acknowledgements: This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014- 2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. ITER is the Nuclear Facility INB no. 174. The views and opinions expressed herein do not necessarily reflect those of the ITER Organisation.
References:
[1] Ph.Lauber et al, EX1/1 Proc. 27th IAEA FEC (2018)
[2] Ph.Lauber et al, J.Comp.Phys.,226/1 (2007)
[3] S.D.Pinches, Comp.Phys.Comm.,111 (1998)
[4] Z.Lu et al, PPCF 25 012512 (2018)
[5] Eurofusion Enabling Research Projects ’NAT’ and ’MET’ (2017,2019)
[6] G.Vlad et al, this conference
[7] JT-60SA Research Plan v4.0, September(2018)
[8] A.Bierwage et al, PPCF 59 125008 (2017), PPCF 61 014025 (2019)
[9] S.D.Pinches et al, TH/P6-7 Proc.27th IAEA FEC(2018)
[10] S.D.Pinches et al, this conference
[11] ARTAUD,J.F. et al., Nucl.Fusion 58 105001 (2018)Speaker: Philipp Lauber (IPP Garching) -
Controlled mitigation of heat load has been demonstrated for the first time by doping a closed divertor$^{1}$ plasma at DIII-D with low Z impurity powders. Injection of low-mid Z impurities is a technique under investigation to address the issue of power exhaust in ITER and next-step fusion reactors$^{2,3}$. The use of non-recycling impurities in powder form provides a new capability extending the limited number of impurity species and mixtures usable with conventional gas injection$^{4}$. At DIII-D, it has now been shown that local doping of the outer strike point (OSP) with lithium (Li), boron (B), and boron nitride (BN) powders can be used to enhances radiative cooling of the divertor plasma. In particular, using B as a radiator has the advantage of concentrating the radiation zone more rooted in the scrape-off layer (SOL), allowing for high core performance and high dissipation in the boundary and divertor at the same time$^{5}$. Moreover, Li and B have beneficial effects in terms of wall conditioning and ELM mitigation and suppression$^{6,7}$.
The experiments focused on ELMy H‑modes (I$_{p}$~1 MA, B$_{t}$=2 T, P$_{NB}$~6 MW, f$_{ELM}$~80 Hz, n$_{e}$~3.6-5.0$\cdot$10$^{19}$ m$^{-3}$), confined in upper-single-null geometry with the OSP positioned in the small angle slot (SAS)$^{1}$. The geometry of the SAS closed divertor, the location of the OSP at the SAS target, the position of the Langmuir probes, the injection location, and the magnetic equilibrium are shown in figure 1 a, b. In this configuration, powders were dropped with rates of 1-50 mg/s directly into the OSP region, allowing to study near-target power dissipation and divertor leakage. For all impurity species, injection caused substantial drops in downstream electron temperature (T$_{e}$), particle fluxes (J$_{sat}$), and heat fluxes q$_{||}$, as measured by a densely spaced array of Langmuir Probes in the SAS (figure 1 a). The time traces of these downstream parameters measured poloidally along the SAS wall, are shown in figure 1 c-f for a discharge with BN injection at t~3 s, causing a sustained strong drop of heat and particle fluxes at the OSP.
In this and similar cases, features characteristics of detachment, such as a strong increase of near-target neutral pressure, were also observed. In general, the high divertor radiation state could be achieved with a 2-15% degradation in H-mode energy confinement.To study the impact of the impurities on SAS dissipation and examine the effect of different impurities’ radiation efficiency and spatial distribution, these scenarios have been analyzed with the 3D plasma-fluid and kinetic neutral edge transport code EMC3-EIRENE. In a first step, transport simulations have been conducted in an axisymmetric geometry, for upstream densities of 1-3$\cdot$10$^{19}$ m$^{-3}$ with B, N, Li as main impurities separately to survey the distribution of the radiated power of each impurity individually. The 2D emissivity patterns for Li and B are shown in figure 2 for an upstream density of 2$\cdot$10$^{19}$ m$^{-3}$. The Li radiation is concentrated near the separatrix and at the strike line while the B peak emission front is concentrated in the far SOL and outer tail of the main recycling region. These characteristic features, qualitatively consistent with measurements of spectral divertor imaging, are a result of the different cooling efficiencies of these two impurities combined with transport effects.
The combined set of experimental results and modeling indicates that injection of low Z impurities in powder form can be a novel effective method to enhance and control of radiative dissipation in closed divertor configurations. This first-time assessment of the effects of different impurity powder species supports the development of new divertor solutions avoiding impurity leakage from the divertor and maintaining high performance.
(1) H.Y. Guo et al 2017 Nucl. Fusion 57 044001
(2) A. Kallenbach et al 2013 Plasma Phys. Control. Fusion 55 124041
(3) L. Casali et al. Contributions to Plasma Physics 2018; 58: 725– 731
(4) A. Nagy et al Rev. Sci. Instrum., 10K121 (2018) 89
(5) A. Yu. Pigarov 2017 Phys. Plasmas 24, 102521
(6) A. Bortolon et al., Nucl. Mater. and Energy, 384-389 (2019) 19
(7) R. Maingi et al., Nucl. Fusion, 024003 (2018) 58- This work is supported by the US Department of Energy under DE-AC02-09CH11466, DE-FC02-04ER54698, DE-AC05-00OR22725, DE-SC0020357, DE-SC0020284, DE-AC52-07NA27344 and DE-FG02-07ER54917.
Speaker: Dr Florian Effenberg (Princeton Plasma Physics Laboratory, Princeton, NJ, 08543 USA) -
Energetic particles (EP) represent the main source of heat and momentum for burning plasmas. However, EPs can drive instabilities that, in turn, can cause redistribution and loss of EPs. The reduced physics, energetic particle kick model for EP transport enables interpretive and predictive capabilities for time-dependent integrated tokamak simulations including the effects of EP transport by instabilities [Podestà 2017]. Over the last two years, the model has been extended include effects of low frequency instabilities in addition to Alfvénic modes, thus providing a common framework to simulate the possible synergy between different types of instabilities. The model is based on transport probability matrices describing the effects of instabilities in the Monte Carlo module NUBEAM of TRANSP. Matrices are computed for each scenario through particle following codes. The kick model has demonstrated the importance of phase-space resolved simulations of EP dynamics to unravel details of EP transport for detailed theory/experiment comparison [Podestà 2017][Heidbrink 2018] and for scenario planning, e.g. based on the optimization of NBI parameters [Podestà 2019]. Enhancements to TRANSP via the inclusion of the kick model enables scenario development and predictions including realistic treatment of EP transport by instabilities, which is required for reliable and quantitative projections of present results to ITER and next-step devices.
The kick model has been mostly used to simulate the effects of Alfvénic instabilities (AEs) in NSTX/NSTX-U and other tokamaks, enabling detailed studies of the mechanisms driving the instabilities and of the resulting EP transport [Podestà 2017][Podestà 2019]. Recently, the model has been extended to include perturbations such as Neoclassical Tearing Modes (NTMs) [Heidbrink 2018][Bardóczi 2019], sawteeth [Kim 2019] and kink/fishbones [Podestà 2019]. Work is also in progress to extend the model outside the separatrix to include externally imposed 3D perturbations and toroidal field ripple. These developments have resulted in a unique tool to study the synergistic effect of multiple instabilities from the plasma core out to the edge, which can greatly enhance EP transport $\&$ loss and degrade plasma performance.
Figure 1 shows a spectrum of magnetic fluctuations and simulation results from a Neutral Beam heated NSTX-U plasma. Multiple instabilities are destabilized, including toroidal AEs (TAEs), fishbones and kink modes [Podestà 2019]. In the simulations, TAE radial structure and damping rates are obtained from NOVA-K [Cheng 1992][Gorelenkov 1999]. Amplitude of TAEs is inferred by balancing the damping rate from NOVA-K with the drive from NUBEAM [Podestà 2017]. Simple analytic expressions are used for the kink/fishbone radial structure. Their amplitude is adjusted to match the measured neutron rate and stored energy.
‘Classical’ TRANSP simulations that do not include EP transport by instabilities over-estimate the measured neutron rate (Fig. 1b). When both TAEs and fishbones/kink are included, the simulation agrees well with the measurements. By enabling only specific types of modes, their relative role in causing EP transport and loss can be studied (Fig. 1c). Fishbones/kink modes appear to be the primary cause of EP losses. Notably, the loss rate for the case including all perturbations does not equal the sum of loss rates obtained with specific modes only. This indicates a synergy between different modes. The phase space resolution implemented in the kick model enables a detailed analysis of the mechanisms leading to enhanced EP losses. Figure 2 shows examples of root-mean-square energy kicks vs EP phase space for a $n=1$ kink and a $n=4$ TAE mode. Because of their spatial localization, the kink and TAE modes cause larger kicks near the axis and from mid-radius to the plasma edge, respectively. The combined effect of the two modes is therefore to open up an efficient loss channel that connects the plasma core to the edge.
In addition to the effects of AEs and kink/fishbones, the kick model has also been extended to NTMs [Heidbrink 2018][Bardóczi 2019][Yang 2018].
Figure 3a shows magnetic fluctuations for a NSTX discharge in which a 2/1 NTM is destabilized towards the end of the pulse. Analysis of 2D soft X-ray (SXR) emission provides the width of the NTM island (Fig. 3b). Mirnov coil data are then rescaled based on results from SXR analysis, which provides the time-dependent island width information for the simulations. The measured and simulated neutron rate (Fig. 3c) shows that the NTM causes a maximum decrease in neutron rate of $\sim 10\%$ with respect to ‘classical’ simulations. Significant deviation from the classical run is only observed for island width exceeding $\sim 8$ cm, indicating a possible threshold in mode amplitude for EP transport. Below that threshold, the1 mode causes a local EP redistribution with no adverse impact on the overall EP confinement [Yang 2018]. Future work will extend these results to cases with other instabilities in addition to NTMs.
This work is supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences under contract number DE-AC02-09CH11466.References:
[Podestà 2017] M. Podestà et al., Plasma Phys. Control. Fusion 59 (2017) 095008
[Heidbrink 2018] W. W. Heidbrink et al., Nucl. Fusion 58 (2018) 082027
[Podestà 2019] M. Podestà et al., Nucl. Fusion 59 (2019) 106013
[Bardóczi 2019] L. Bardóczi et al., Plasma Phys. Control. Fusion 61 (2019) 055012
[Kim 2019] D. Kim et al., Nucl. Fusion 59 (2019) 086007
[Cheng 1992] C. Z. Cheng, Phys. Rep. 1 (1992) 211
[Gorelenkov 1999] N. N. Gorelenkov et al., Phys. Plasmas 6 (1999) 2802
[Yang 2018] J. Yang et al., Characterization of Tearing Modes in NSTX, 61st APS-DPP Meeting (2019)Speaker: Mario Podesta (Princeton Plasma Physics Laboratory) -
The understanding and control of runaway electrons (RE) is a top priority of the nuclear fusion program because, if not avoided or mitigated, RE can severely damage the plasma facing components of a tokamak. Two key open problems are the generation and the impurity-based mitigation of RE. The first problem requires the computation of the production rate of RE. That is, given a plasma state, determine how many RE are produced during the thermal and current quench phases of a disruption. The second problem requires the simulation of the interaction of RE with partially ionized impurities in the presence of spatio-temporal evolving electric and magnetic fields. The study of these problems is motivated by the practical challenges of finding the optimal impurity injection protocol for the controlled shutdown of a plasma (avoiding or minimizing the generation of RE) and the design of impurity injection strategies for the effective dissipation of the RE beam once it is formed.
The main goal of this paper is to advance the current understanding of RE generation and mitigation by presenting a numerical study focusing on the role played by usually neglected, or highly approximated, spatio-temporal effects. Of particular interest is the dependence of the RE production rate on general dynamic scenarios exhibiting time dependent plasma temperature and electric fields as well as RE beam losses due to radial transport. The study of spatio-temporal effects in the mitigation of RE by impurity injection is also an important goal of this paper. Most of the results will focus on prescribed electric and magnetic fields on a given plasma state. However, preliminary results on self-consistent effects involving coupling of the kinetics of RE to temperature and electric field evolution models will also be presented.
The study of these problems has been a topic of significant interest in the fusion community and many important results have been obtained over the last decade, see for example the recent review paper [1] and references therein. What distinguishes the study presented here from previous work is the focus on spatial effects and the unique approach followed. This approach is based on the combination of a probabilistic Backward-Monte-Carlo method (BMC) [2] and kinetic simulations using the Kinetic Orbit Runaway electron Code (KORC) [3]. The BMC method is based on a direct numerical evaluation of the Feynman-Kac formula that establishes a link between the solution of the adjoint Fokker-Planck problem for the probability of runaway, PRE, and the stochastic differential equations describing the dynamics of RE in the presence of collisions. Computationally, the BMC is a deterministic algorithm that reduces the problem to the evaluation of Gaussian integrals that can be efficiently computed with high accuracy using Gauss-Hermite quadrature rules. Reference [4] discussed how the BMC can be used to efficiently compute the coupling between a fluid and a kinetic description of RE dynamics.
Figure 1 shows an example of a BMC computation of the PRE, incorporating spatial dependence. Going beyond our previous study [2], that limited attention to the computation of the PRE for a given momentum and pitch angle, we extended the BMC to account for radial transport, which in this example is modeled as a diffusive process. The numerical implementation of this 3D extension uses hierarchical sparse-grid interpolation methods with piece-wise polynomials to approximate the map from the phase space to the runaway probability. Also, to handle the sharp transition layer between the runaway and non-runaway regions we use adaptive refinement techniques [5]. Another important extension of the BMC that will be discussed in this paper is the computation of the PRE in time dependent scenarios incorporating models for the temperature and electric fields evolution during the thermal quench.
Figure 2 shows an example of a KORC simulation of a RE beam in time-dependent magnetic and electric fields in the presence of impurities. For these simulations, KORC has been extended by incorporating state-of-the-art collisional models with partially ionized impurities and spatio-temporal models of impurity transport [6]. As indicated by the vertical dashed line, in this simulation, the evolution of the RE beam can be divided in two stages. For t<0.015, as shown in panel (b), there are no RE lost to the wall. However, as shown in panel (a), the energy of the beam actually increases. Interestingly, this increase of the energy is accompanied by a decrease of the parallel (zero pitch angle direction) current due to strong pitch angle scattering or the RE with the impurity. For t>0.015, the fraction of RE lost to the wall significantly increases. This happens because, in this simulation, the time evolution of the magnetic field exhibits a strong horizontal displacement and the RE “peel off” at the high-field side. This loss of confinement leads to a decrease of the RE energy (shown in panel (a)) but this is mostly due to the deconfinement of high energy RE and not because impurity-driven dissipation which, as seen in the green curve in panel (b), is small. The main message of this simulation is that the assessment of the effectiveness of impurity-based RE mitigation is a complex problem involving the competition of different physics mechanisms (magnetic confinement, electric field evolution, coulomb drag, pitch angle scattering, and impurity transport and ionization) with different times scales. In particular, if as in the simulation shown in Fig.2, the time scale of the magnetic field evolution is faster than the time scale of the stopping power of the impurity, then the RE might hit the plasma facing components of the tokamak before they can be significantly slowed down. Simulations exploring this general scenario in the context of DIII-D experiments will be discussed in Ref.[6].
[1] B. Breizman, P. Aleinikov, E. Hollmann and M. Lehen, Nucl. Fusion 59 083001 (2019).
[2] G. Zhang and D. del-Castillo-Negrete, Phys. Plasmas 24, 092511 (2017).
[3] L. Carbajal et al., Phys. Plasmas 24, 042512 (2017).
[4] E. Hirvijoki, C. Liu, G. Zhang, D. del-Castillo-Negrete and D. Brennan, Phys. of Plasmas 25, 062507 (2018).
[5] M. Yang, G. Zhang, D. del-Castillo-Negrete, M. Stoyanov, and M. Beidler, https://arxiv.org/abs/2001.05800 (2020).
[6] M. Beidler et al., 28th IAEA Fusion Energy Conference (FEC 2020).*Research sponsored by the Office of Fusion Energy Sciences of the US Department of Energy at Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Department of Energy under contract DE- AC05-00OR22725.
Speaker: Diego del-Castillo-Negrete (Fusion Energy Division. Oak Ridge National Laboratory) -
This work reports on a breakthrough on the way to a comprehensive modelling of burning fusion plasmas. For the first time, global electromagnetic gyrokinetic PIC simulations of Alfvénic modes have been successfully performed for a high beta ITER plasma.
This finally gives us the ability to model alpha particle driven modes self-consistently in the non-linear regime and to predict the related alpha particle transport with a high level of confidence.
The ITER 15 MA scenario [1], with significant alpha particle pressure, is a scenario in which a broad range of Alfvén eigenmodes may be present. More specifically, a large number of possible of toroidal Alfvén eigenmodes (TAEs) can exist in the plasma, and multiple of these may be driven by the alpha particle gradient. Whilst the induced transport is expected to be small at nominal parameters, it is of significant interest to consider the sensitivity and borders to enhanced transport regimes. Previous modelling of this discharge has been performed (see, for example reference [2] and references therein), most notably by perturbative hybrid (MHD or gyrokinetic eigenvalues and drift- or gyro-kinetic energetic particles), nonperturbative hybrid MHD-kinetic, or local gyrokinetic. Although we know that kinetic, nonperturbative, and global effects are important for the TAEs in ITER, due largely to the difficulties caused by the large size and high plasma β of ITER, it has previously not been possible to apply nonlinear models containing all of these effects consistently. In parallel to this, significant progress [3] has recently been made in global, electromagnetic gyrokinetics, which has yielded e.g. detailed nonlinear studies of energetic particle driven Alfvén eigenmodes in simplified conditions [4]; benchmarks of Alfvén eigenmodes in experimental conditions [5, 6]; and the nonlinear interaction between electromagnetic turbulence and Alfvén modes [7].
The simulations in this work are performed with the ORB5 code [8, 9], a global electromagnetic particle-in-cell (PIC) code using spectrally filtered finite elements for the representation of the fields, with all species considered kinetically, and using the “pullback scheme” [3] to mitigate the cancellation problem.In this contribution, we present the first application of a global nonlinear electromagnetic gyrokinetic code to address the issue of TAE stability in ITER, focussing first on the linear stability of the modes over a range of mode numbers and gap positions, before looking at the saturation observed when retaining the wave-particle nonlinearity. This is expanded by considering also multiple modes present simultaneously, observing a significant increase of the mode saturation levels by approximately one order of magnitude, and accordingly increased alpha particle redistribution, in a case with double the nominal alpha particle density, and when neglecting the gyro-average on the energetic particles. This is consistent with the finding of previous hybrid modelling [10].
In the linear regime, we observe a range of mode numbers for which there exist multiple eigenmodes, with global mode structures spanning multiple gaps, reflecting what is found with linear eigenvalue calculations. We show that these cases require a global domain to describe the linear mode properties. For more localized mode structures, with higher mode numbers, considering a reduced domain is justified, and we see close agreement, even for nonlinear saturation levels between full and reduced radial domains.
Figure 1: A snapshot of the absolute magnitude of the poloidal harmonics of the electrostatic potential of an n=12 TAE in the linear phase.In linear simulations without energetic particles present, where we allow an initial perturbation to decay until eigenmodes are observed, we find also different modes present, for example those from the higher frequency gaps (e.g. EAE and NAE), as well as odd-parity TAEs from the upper part of the gap, as these modes (although barely driven in the presence of energetic particles) are weakly damped. An interesting observation is that in the post-saturation phase of the nonlinear multi-mode TAE simulations, we also observe odd-parity TAEs present. This observation motivates a reconsideration of which modes to include for hybrid perturbative simulations of the same scenario, where previously the odd-parity modes were neglected a priori due to the weak drive.
Figure 2: Time evolution of the peak radial values of the different toroidal mode numbers of the electrostatic potential in a nonlinear simulation.To make simulations feasible, we have allowed several simplifications of the physics problem, but we have considered the impact of each, which we will discuss, in particular, we show the effect of reducing the ion/electron mass ratio.
We also present the application of the ORB5 code to an experimental case taken from an ASDEX Upgrade discharge with particularly interesting energetic particle physics [11], and this result is benchmarked against hybrid codes. The details of this benchmark will be presented in reference [12].
Acknowledgement: This work has been carried out within the framework of the EUROfusion
Consortium and has received funding from the Euratom research and training programme 2014-
2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein
do not necessarily reflect those of the European Commission.References
[1] A. Polevoi et al., JPFR-S, (2002)
[2] S. Pinches et al., Phys. Plasmas (2015)
[3] A. Mishchenko et al., Comp. Phys. Comm. (2019)
[4] M. Cole et al., Phys. Plasmas (2017)
[5] S. Taimourzadeh et al., Nuclear Fusion (2019)
[6] F. Vannini et al., Submitted to Physics of Plasmas
[7] A. Biancalani et al., this meeting
[8] S. Jolliet et al., Comp. Phys. Comm. (2007)
[9] E. Lanti et al., Comp. Phys. Comm. (2020)
[10] M. Schneller et al., Plasma Phys. Control. Fusion (2015)
[11] Ph. Lauber et al, EX1/1 Proc. 27th IAEA FEC (2018)
[12] G. Vlad et al, this meetingSpeaker: Thomas Hayward-Schneider (Max Planck Institute for Plasma Physics) -
Recent experiments on DIII-D have utilized the new off-axis neutral beam injection (NBI) power to achieve $\beta_N$ = 3.8 with n = 1 ideal stability limits up to $\beta_N$ = 6. The NBI upgrade adds two additional co-current, off-axis, beams giving a total of 8 MW of on- and 7 MW of off-axis NBI power for advanced tokamak (AT) scenario development in these experiments. In addition, 1.6 MW of electron cyclotron (EC) power is used as an additional off-axis heating and current drive source. These off-axis current drive sources broaden the current and pressure profiles to better couple to the vessel wall thereby raising the ideal-wall, low-n kink stability $\beta_N$ limits.
Despite higher ideal stability limits with the additional off-axis current drive capabilities, a majority of these high-q$_{min}$ discharges are limited by tearing modes. Past analysis indicates discharges with higher ideal stability limits have higher tearing mode stability limits$^1$. However, these recent experiments, with the additional off-axis beam power, have increased the ideal-wall limit without apparent improvement in tearing mode stability. The DCON stability code$^2$ is used to calculate the ideal-wall and no-wall stability limits. At the time of tearing mode onset, the ideal-wall $\beta$ limits range between $\beta_N$ = 4.2-5.7, no-wall $\beta$ limits between $\beta_N$ = 2.8-3.5, and achieved experimental $\beta_N$ = 2.5-3.8, Fig. 1. In three of the discharges, tearing modes form with an ideal-wall $\beta$ limit of $\beta_N$ > 5 and experimental $\beta_N$ < 3. Clearly, increasing ideal stability limits has not been sufficient for preventing the frequent appearance of tearing modes in these discharges. Furthermore, it is observed that tearing modes frequently form when $\beta_N$ is near the no-wall stability limit. None of the discharges exceed the no-wall $\beta$ limit by more than 10% despite ideal-wall stability limits that exceed the no-wall limits by 50%. With the additional beam power available, these discharges are stability, not power, limited. A better understanding of tearing mode onset physics and avoidance requirements is needed for this regime.
Large tearing modes in these plasmas are m/n = 3/1 and result in a confinement reduction of $\approx$50%. A majority of the tearing modes have a 5/2 tearing mode precursor, which causes a relatively minor reduction in confinement. Tearing modes occur in 10 of 12 discharges shown in Fig. 1 with the tearing mode onset indicated by a circle. After the tearing mode forms, the plasma confinement does not recover in a majority of the discharges.
q$_{min}$ > 2 operations eliminate the 2/1 rational surface from the plasma and avoid deleterious fast-ion modes. However, confinement reduction from 3/1 tearing modes have been significant enough to prevent higher $\beta_N$ operation in this regime. Timing and onset of tearing modes do not show a clear relationship to broader current (higher q$_{min}$) or pressure (lower pressure peaking factor) profiles for operations near q$_{min}$ = 2.
The highest $\beta_N$ with the new off-axis NBI capabilities compared to a similar discharge with all on-axis NBI power was achieved with q$_{min}$ = 1.1-1.5, Fig. 2. This discharge achieved ideal $\beta_N$ stability limits near 6, significantly higher than the reference discharge with only on-axis beam power. Feedback control with 3D fields was applied in both discharges to maintain optimal error field correction and resistive wall mode stabilization. In addition, this result was achieved despite a reduction in available EC power from 2.9 to 1.6 MW.
Predictive TRANSP simulations aided the development of these discharges with increased off-axis NBI power. TRANSP runs of past discharges with majority on-axis beam power were modified to inject power with the new off-axis beam geometry. TGLF was used to evolve the temperature, density, and current profiles with the increased off-axis NBI power. These predictive simulations showed that early application of EC power raises q$_{min}$ and increases the non-inductive current fraction of the plasma, which was observed in subsequent experiments. Broadening of the NBI current density profile with the new beam geometry was also accurately predicted using TGLF in TRANSP.
High fusion gain steady-state tokamaks are based on broad current and pressure profiles to achieve wall-stabilization of ideal-MHD kink modes at high $\beta_N$. The results discussed show that obtaining a high ideal-wall limit, while necessary, is not sufficient, as tearing modes still appear at lower $\beta_N$, usually around the no-wall limit. The relationship between the ideal- and no-wall stability limits and the current density and pressure profiles that determine tearing mode stability will be explored to better sustain higher $\beta_N$ plasmas.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC52-07NA27344, DE-FC02-04ER54698, and DE-FG02-04ER54761.
$^1$F. Turco, et. al, Physics of Plasmas 19, 122506 (2012).
$^2$A. H. Glasser, Physics of Plasmas 23, 072505 (2016).Speaker: Brian Victor (Lawrence Livermore National Laboratory) -
In this paper the integrated modelling for the steady-state regime of a fusion neutron source DEMO-FNS (R/a=3.2m/1m, B=5T, Ip=4-5MA) [1] is complemented by the helium balance in the divertor and core plasma. The model describes the power and particle balances consistently in the divertor and core plasmas according to approach [2] and finds the condition to fulfill the global requirements on: neutron source value Sneut >1019/s, maximal peaking heat load is limited by qpk <10MW/m2, keeping up-down symmetry and avoidance of deep detachment, μ<1, condition for plasma current overdrive Ipl ≤ IBS+ICD , when the plasma current Ipl is generated by the noninductive methods from neutral beam ICD and from the bootstrap current IBS and on a source by pellets Spel>fT/(1-fT)SNB, that control the tritium fraction fT =nT/(nT+nD) in a core, where SNB is the D source from neutral beam, Spel is the pellet source of D and T.
The energy and particle balances in the core plasma are modelled with transport equations in the ASTRA code [3] using ITER IPB(y, 2) [4] with prescribed H-factor, ratio of main ions and helium confinement times to the energy confinement time τp/τE, τНе/τE , tritium fraction, fT, electron averaged density <ne> (maintained by the pellet source Spel) and density at the pedestal (see Table 1). Model for the pedestal poloidal beta and the pedestal width is taken from [5,6].
The state of divertor plasma is determined by the values of the incoming heat flux PSOL, the neutral pressure in divertor volume pn, the concentration of the neon impurity CNe and the pumping speed Cpump to pump out the helium ash produced by the fusion reaction (see Table 2). Results of SOLPS4.3 [7] series runs are approximated by analytical dependencies (scalings) in order to install the boundary conditions and heat and particle fluxes through the separatrix into the core plasma.
Self-consistent modeling of central and divertor plasma for DEMO-FNS [8] allows to determine the acceptable window of the divertor parameters pn and CNe where the global requirements are fulfilled (see Figure). It should be noted that in comparison to the modelling without helium balance [1] the acceptable parameter window become significantly constricted. The neutron yield is weakly reduced but remained to be at the reasonable level Sneut= 1.1·1019/s. As one of the neutron reduction factor is the decrease of the tritium fraction in the core plasma in order to fulfill the requirement to tritium fraction control with pellet source (see Table 3). The up-down symmetry requirement (μ<1 red line in Figure) become a strong factor in the divertor neutral pressure pn limit. We need to make an accurate investigation of a single X-point configuration as it can be more efficient to mitigate the power load to divertor plates.
References
[1] A.Yu. Dnestrovskiy et al., Nuclear Fusion 2019 59 (9) 096053
[2] H. D. Pacher, et al., J. Nucl. Mater. 313 (2003) 657
[3] G.V. Pereverzev and P.N. Yushmanov, IPP-Report IPP 5/98 (2002),
[4] ITER Physics Expert Group on Confinement and Transport, 1999 Nucl. Fusion 39 2175
[5] P. B. Snyder et al. Nucl. Fusion, 2011, vol. 51, p. 103016
[6] S.Yu. Medvedev., et al., 2012 Probl. At. Sci. Technol. Ser. Thermonucl. Fusion 35 21 (in Russian)
[7] A.S. Kukushkin, et al., Fusion Eng. Des. 86 (2011) 2954
[8] Yu.S. Shpanskiy and DEMO-FNS Team 2019 Nucl. Fusion 59 076014Table 1 Set of core plasma parameters
τp/τE, τНе/τE 4
Н-factor 1.3
fT 0.5
Ipl, MA 4.5
nped/<ne> 0.7
<ne> 1019/m3 7Table 2 Set of divertor plasma parameters
PSOL, MW 37.5
pn, Pa 2
Cpump, m3/s 20
СNe=ΣNNe/ne 0.025
qpk, MW/m2 9.7
μ 0.87Table 3 Result global values
Sneut 1019/с 1.1
IBS, MA 1.7
ICD, MA 2.8
PDT МW 31
PDT_pp /PDT_bp 1.8
SpelT,1019/s 58
SpelD,1019/s 21Speaker: Alexey Dnestrovskiy (NRC Kurchatov Institute) -
Contrary to previous thinking, recent experiments on DIII-D suggest that the low-frequency instability known as the beta-induced Alfvén-acoustic eigenmode (BAAE)$^1$ will not degrade high-energy fast-ion confinement on future devices. On the other hand, another low-frequency instability, the beta-induced Alfvén eigenmode (BAE)$^2$, interacts strongly with the high-energy fast-ion population and remains a threat.
Deuterium neutral beam injection (NBI) drives BAAE, BAE, and reversed-shear Alfvén eigenmode (RSAE)$^3$ instabilities (Fig. 1a). When the NBI and electron cyclotron heating (ECH) patterns are held constant, the AE activity (and q profile evolution) is highly reproducible at the time of interest. Transient changes in heating isolate the driving gradients for the instabilities. When the NBI heating is altered, the BAE activity rapidly changes (Fig. 1b) in a time much shorter than the ~100 ms slowing-down time, indicating that high-energy ($E_{beam}\sim80$ keV) beam ions drive the BAEs unstable. In contrast, BAAE activity persists even in the absence of NBI. Even though the neutron rate decays to <5% of its previous value, the BAAEs remain unstable, indicating that fast ions with energies above ~30 keV are not driving the BAAEs. In other discharges, the angle of beam injection was changed from tangential to perpendicular at approximately constant NBI power. The frequency of BAE activity increased but BAAE activity was hardly affected. Calculations of the resonance condition $\omega=n\omega_\phi+p\omega_\theta$, where $\omega_\phi$ and $\omega_\theta$ are the toroidal and poloidal orbital frequencies of recently injected beam ions, show that the increase in frequency $\omega$ allows perpendicular beam ions to resonate with the BAE. (Here, n is the toroidal mode number and p is an integer.)
The intermittent appearance of RSAE, BAE, and BAAE activity in Fig. 1 occurs because all three modes are sensitive to the q profile, which continuously decreases during the illustrated time. Electron cyclotron emission (ECE) and beam emission spectroscopy (BES) diagnostics show that the radial eigenfunction for all three modes peaks near $q_{min}$. Unlike RSAEs and BAEs, the BAAEs are undetectable on Mirnov coils, suggesting less electromagnetic polarization for these modes.
In addition to a database of the 20 shots in the dedicated experiment of Fig. 1, a database of over 1000 DIII-D discharges with ~ 2500 entries has been compiled. For each entry, RSAEs, BAEs, and BAAEs (as well as TAEs and EAEs) are classified as stable, marginal, or unstable. In both databases, BAAE activity occurs most often in plasmas with large electron temperature $T_e$ but relatively modest beta (Fig. 2a). Since the BAAE is expected to have an acoustic component to its polarization, these dependencies may reflect an underlying dependence on ion Landau damping, which is enhanced at low $T_e$ and in plasmas with large ion pressure. As expected from the dedicated experiment, the dependence on NBI parameters in both databases is very weak for the BAAE.
In contrast, BAEs occur more frequently as the poloidal beta $\beta_p$ increases (Fig. 2b) and in plasmas with more beam power and parallel beam beta. For $\beta_p>1.5$, the occurrence of BAEs is comparable to the occurrence of TAEs, highlighting the importance of predicting BAE stability in Advanced Tokamak reactor scenarios.
Initial comparisons of gyrokinetic (GTC and LIGKA) and gyrofluid (FAR3D) codes against the data of Fig. 1 are encouraging. GTC finds a linearly unstable mode with frequency and mode structure similar to the experimental BAE; an unstable mode with BAAE frequency is only found in simulations without beam ions. Preliminary nonlinear simulations indicate that zonal flows cause the saturation of the unstable BAE. After benchmarking, these codes will predict BAAE and BAE stability in ITER.
Work supported by US DOE under DE-FC02-04ER54698 and DE-SC0020337.
$^1$N.N. Gorelenkov et al., Phys. Lett. A 370 (2007) 70.
$^2$W.W. Heidbrink et al., Phys. Rev. Lett. 71 (1993) 855.
$^3$S.E. Sharapov et al., Phys. Lett. A 289 (2001) 127.
Speaker: william heidbrink (UC Irvine) -
Unstable Alfvén eigenmodes (AEs) are a key issue in magnetically confined fusion, both for currently operating machines (JET, AUG, etc) and for next step devices such as JT60-SA and ITER, due to their potential to cause energetic-ion (heated by NBI/ICRH or fusion alphas) redistribution and losses [1,2]. Toroidicity-induced AEs (TAEs), resulting from the coupling of two shear-Alfvén waves, are one of the most extensively studied Alfvénic instabilities in tokamaks [1,2]. However, lower-frequency AEs (i.e., $\omega<\omega_{tae}$) have been observed in DIII-D high-beta plasmas under intense NBI, associated with large energetic-ion loss levels, much similar to those caused by TAEs [3,4].
In this contribution, we report and discuss Alfvénic activity observed at JET, at about half the TAE frequency (figure 1), in plasmas heated by NBI and ICRH [5]. AEs with frequencies lower than the TAEs have been previously explained by the beta-induced coupling of shear-Alfvén and acoustic waves (BAAEs), via the lowest-order harmonic of the field-line geodesic curvature [6,7]. However, their expected frequency $\omega_{baae} \sim 2 \beta^{1/2} \omega_{tae}$ is too low to explain the measured eigenfrequencies on JET. Here, we show that the experimental measurements can be explained by the previously unexplored gaps in the frequency of shear-Alfvén and acoustic continua. In the vicinity of these gaps, several acoustic waves couple to a single shear-Alfvén wave via higher-order harmonics of the geodesic curvature caused by the plasma shaping and, in particular, the elongation of JET plasmas [8]. In the limit of plasmas with circular magnetic surfaces, the proposed model reduces to the well-known lowest-order coupling. New frequency gaps are predicted at integer multiples of $\omega_{baae}$, but only under certain conditions imposed by the local shaping parameters and a limiting value of the safety factor [8].
The continuous spectra computed with the compressible ideal-MHD code CASTOR display the predicted frequency gaps, which open where acoustic branches cross with the up-shifted shear-Alfvén branch (figure 2). Although quite narrow, such gaps allow the existence of global AEs whose radial structure and frequency is also computed with CASTOR (figure 2). In this contribution, the computed AE radial location and frequency are shown to be in fair agreement with experimental data, the latter within a few percent of measured values and the former matching soft x-ray observations.
The proposed high-order geodesic acoustic eigenmodes (HOGAEs) were found to be driven unstable by energetic-ion populations with characteristics similar to those accelerated by ICRH (figure 3), with a temperature around 1 MeV and on-axis density about 1% of the value corresponding to the thermal ions [8]. In this contribution, their stability in JET plasmas will be evaluated with the hybrid MHD/drift-kinetic code CASTOR-K, taking into account energetic-ion populations heated by NBI and ICRH, their distributions functions being computed, respectively, by the codes ASCOT and PION. Wave-particle resonances, along with drive/damping mechanisms, will also be discussed. Amplitude saturation levels and mechanisms will be addressed with the non-linear code POLARIS-K. Overall, the presented results will allow a better understanding of HOGAEs role in energetic-ion redistribution and losses and their potential impact on the operation of next step devices like JT60-SA and ITER.
Acknowledgments:
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014- 2018 and 2019-2020 under grant agreement No 633053. One of the authors (FC) was supported by FuseNet from the Euratom research and training programme under Grant Agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. IPFN activities were also supported by “Fundação para a Ciência e Tecnologia” (FCT) via project UID/FIS/50010/2013.References:
1 A. Fasoli et al., Nucl. Fusion 47, S264 (2007).
2 N. Gorelenkov et al, Nucl. Fusion 54, 125001 (2014).
3 W. Heidbrink et al, Phys. Rev. Lett. 71, 855 (1993).
[4] W. Heidbrink et al, Phys. Plasmas 6, 1147 (1999).
[5] R. Dumont et al, Nucl. Fusion 58, 082005 (2018).
[6] B. Holst et al, Phys. Plasmas 7, 4208 (2000).
[7] N. Gorelenkov et al, Phys. Lett. A 370, 70 (2007).
[8] F. Cella and P. Rodrigues, “Shaping effects on the interaction of shear-Alfvén and slow sonic continua”, Proc. 46th EPS Conference on Plasma Physics 8-12 July 2019.Speaker: Paulo Rodrigues (Instituto de Plasmas e Fusao Nuclear) -
Energetic particle physics is a crucial issue in burning plasmas such as the International Thermonuclear Experimental Reactor (ITER). Instabilities driven by energetic particles, such as fishbones and various Alfvén eigenmodes, can induce the transport and loss of energetic particles, degrade fast particle confinement, and even lead to serious wall damage. A non-monotonic safety factor profile with a reversed magnetic shear configuration has been proposed as an advanced scenario for future ITER operation. For the consideration of the fishbone instability, there are two different conditions: the minimum value of safety factor qmin is less or a little larger than unity. There are few simulations to investigate the fishbone instabilities with reversed safety factor profile. As a result, in this work, linear stability and nonlinear dynamics of the fishbone instabilities with reversed safety factor profile have been investigated by the hybrid code M3D-K[1,2], including both the non-resonant type with qmin larger than unity and the type with dual q = 1 surfaces, which we will infer as non-resonant fishbone (NRF) and dual resonant fishbone (DRF) in the following.
Based on EAST-like parameters, the linear simulation results of the n = 1 mode with double q = 1 rational surfaces are firstly presented. With fixed total pressure, the linear growth rate and mode frequency as a function of beam ion pressure fraction Phot,0/Ptotal,0 are shown in figure 1, where Phot,0 is the central fast ion pressure, and Ptotal,0 is the central total pressure. At zero beam ion pressure with Phot,0/Ptotal,0 = 0, the ideal internal kink mode is unstable. The corresponding mode structure is shown in figure 1 (a). This mode has an up-down symmetric structure with zero mode frequency, and it exhibits splitting feature due to double q = 1 surfaces. The dominant mode number is n = m = 1, where n is the toroidal mode number and m is the poloidal mode number. When the value of Phot,0/Ptotal,0 is small and increases from 0 to 0.4, the mode is stabilized due to the kinetic effects of beam ions. However, when Phot,0/Ptotal,0 is larger than 0.4, the DRF is excited, which is an energetic particle mode driven by trapped beam ions. Figure 1 (b) shows the mode structure with Phot,0/Ptotal,0 = 0.3. Compared to the ideal internal kink mode, the mode structure shows a twisted feature with finite mode frequency. The mode structure of the DRF with Phot,0/Ptotal,0 = 0.6 is shown in figure 1 (c), and it becomes more twisted with much larger frequency. When qmin increases from below unity to above unity, the fishbone instability transits from the DRF to the NRF, and the mode frequency of the NRF is higher than the DRF as the NRF is resonant with fast ions with larger precessional frequency. The mode structure of the NRF is shown in figure 1 (d).
Nonlinear simulations show that the saturation of the DRF with Phot,0/Ptotal,0 = 0.6 is due to MHD nonlinearity with a large n = 0 component. Figure 2 shows the reason why the DRF mode cannot saturate just with the nonlinearity of energetic particles. Without MHD nonlinearity, as shown in figure 2 (a), it is found that the distribution of fast ions becomes flattened in the core region during the nonlinear phase. However, near the magnetic axis there still exists the steep radial gradient of the fast ion distribution, which could drive the instability. Correspondingly as shown in figure 2 (c), the inner m/n = 1/1 DRF mode structure shrinks in the central region at t = 700 τA, where τA is the Alfvén time. As a result, the DRF mode still tap the free energy associated with the fast ion radial gradient during the nonlinear phase, and the DRF mode does not saturate without MHD nonlinearity. However, the saturation of the NRF is mainly due to the nonlinearity of fast ions. Figure 3 (a) shows the time evolution of the n = 1 mode amplitude of the NRF. It is observed that the NRF amplitude firstly grows and then saturates from t ≈ 1000 τA. Correspondingly as shown in figure 3 (b), the NRF frequency starts to chirp down at t ≈ 1500 τA from ω ≈ 0.12 ωA to ω ≈ 0.04 ωA, where ωA is the Alfvén frequency. Finally, the redistribution of beam ions due to the DRF and NRF with MHD nonlinearity is discussed, and it is found that the distribution of the fast ions become flattened in the core region. By comparing the fast ion redistribution induced by the DRF, the redistribution level of the fast ions due to the NRF is weaker, and the flattening region of the beam ions is located more centrally in the radial direction, which is consistent with the resonant analysis. Because the equilibrium profiles and parameters are chosen based on EAST-like conditions, the simulation results will provide guidance for and can be verified by future EAST experiments.
References
1 W. Park, E. V. Belova, G. Y. Fu, X. Z. Tang, H. R. Strauss, and L. E. Sugiyama, Phys. Plasmas 6, 1796 (1999).
2 G. Y. Fu, W. Park, H. R. Strauss, J. Breslau, J. Chen, S. Jardin, and L. E. Sugiyama, Phys. Plasmas 13, 052517 (2006).Numerical computations were performed on the ShenMa High Performance Computing Cluster in Institute of Plasma Physics, Chinese Academy of Sciences. This work is supported by National Key R&D Program of China under Grant No. 2017YFE0300400, the National Natural Science Foundation of China under Grant Nos. 11605245, 11975270, 11705236, 11575249, 11805239 and 11505022.
Speaker: Wei Shen (Institute of Plasma Physics, Chinese Academy of Sciences) -
Hybrid simulations for energetic particles interacting with a magnetohydrodynamic (MHD) fluid were conducted using the MEGA code [A, B] to investigate the spatial and the velocity distributions of lost fast ions due to the Alfvén eigenmode (AE) bursts in the Large Helical Device (LHD) [C, D]. It is found that the spatial distribution of lost fast ions in the divertor region during the AE burst is helically symmetric and peaks along the divertor location. Affected by the direction of grad-B and curvature drifts, the distribution of the co-going (counter-going) lost fast ions has two peaks in the outboard (inboard) side depending on fast-ion energy and pitch angle. The numerical fast-ion loss detector “numerical FILD” was constructed in the MEGA code. The velocity distribution of lost fast ions detected by the numerical FILD during AE burst is in good agreement with the experimental FILD measurements. This demonstrates that the MEGA is a useful tool for the prediction and the understanding of the fast-ion transport and losses brought about by AEs.
The LHD is one of the largest helical devices with non-axisymmetric 3-dimensional magnetic configuration. In the LHD, the fast-ion confinement has been investigated by using three tangentially injected neutral beams (NBs) with energy 180 keV and/or two perpendicularly injected NBs with energy 40-80 keV. The recurrent AE bursts were observed during the tangentially injected NB [C, D]. The fast-ion driven instabilities enhance the fast-ion transport and losses. It is important to identify the instabilities and clarify the properties of the fast-ion transport due to the instabilities. A hybrid simulation code for nonlinear MHD and energetic-particle dynamics, MEGA, has been developed to simulate recurrent bursts of fast-ion driven instabilities including the energetic-particle source, collisions and losses [A]. Since the equilibrium magnetic field in the real coordinates are used in MEGA, fast ion can be traced even in the peripheral region including the divertor region. The multi-phase simulation, which is a combination of classical simulation and hybrid simulation for energetic particles interacting with an MHD fluid, was applied to the LHD experiments #47645 [C] and #90090 [D] in order to investigate the AE bursts with beam injection, collisions, losses, and transport due to the AEs [A, B]. In the classical simulation, fast-ion orbits are followed in the equilibrium magnetic field with NBs and collisions while the MHD perturbations are turned off. The fast-ion loss rate brought about by the AE burst is proportional to the square of AE amplitude, which is consistent with the quadratic dependence of fast-ion loss observed in the LHD experiment [B].
In this work, the spatial and the velocity distributions of lost fast ions due to the AE bursts are investigated and compared with the experimental measurements. The multi-phase simulation was conducted for the LHD experiment #90090 where the 2-dimensional velocity distribution of lost fast ions is measured by scintillator-based FILD [D]. The AE bursts occur recurrently and then the fast ions are significantly lost during the AE bursts. Figure 1 shows the spatial distribution of lost fast ions in the divertor region during the AE burst. We see in Fig. 1 that the spatial distribution of lost fast ions in the divertor region during the AE burst is helically symmetric and peaks along the divertor trace. The lost fast ions reach the divertor region following the divertor magnetic field. There is a helical symmetry for the lost fast ion location even during the AE burst while some peaks are present in the number of lost fast ions in the poloidal direction. The time evolution of the spatial distributions of lost fast ions along the divertor trace is shown in Fig. 2. Affected by the direction of grad-B and curvature drifts, the distribution of the co-going (counter-going) lost fast ions has two peaks in the outboard (inboard) side depending on fast-ion energy and pitch angle. For the comparison with the lost fast-ion velocity distribution measured with the FILD, we have constructed the “numerical FILD” in the MEGA code. In the MEGA simulations, the guiding-center orbit is followed for fast ions. In the numerical FILD, when a fast ion approaches the FILD, we split the guiding-center particle into 64 particles around the guiding center with corresponding Larmor radius and 64 particles with different gyration phase are followed with Newton-Lorentz equation. The aperture of the numerical FILD is a circle with radius 6 mm. Only the fast ions passing through the aperture are detected by the numerical FILD. Figure 3 compares the pitch angle and energy distribution of the lost fast ions in the MEGA simulation and the FILD measurements in the experiment. Before the AE burst, fast ions with energy close to the injection energy are mainly detected by the numerical FILD. During the AE burst, we see in Fig. 3(b) that fast ions of 100-150 keV and 35-50 degree are detected by the numerical FILD. The velocity space region of the lost fast ions due to the AE burst is in good agreement with that observed in the experiment shown in Fig. 3(c) [D], although the two peaks observed in the experiment are not well resolved in the numerical FILD. The numerical FILD measurement is consistent with the experiment for the lost fast ions with pith angle = 30-40 degree which increased during the AE burst.
![Spatial distribution of lost fast ions in the divertor region during the AE burst. Color represents the number of lost fast ions [arb. unit]. The promptly lost particles are not included. The star indicates the installation position of FILD in the LHD.](http://workshop.nifs.ac.jp/fec2020/image/71-SEKI-image-fig_1.jpg)
![Comparison of pitch angle and energy distribution of lost fast ions among (a) MEGA simulation before AE burst, (b) simulation during AE burst, and (c) FILD measurements in the LHD experiment [D].](http://workshop.nifs.ac.jp/fec2020/image/71-SEKI-image-fig_3.jpg)
[A] Y. Todo, et al., Phys. Plasmas 24, 081203 (2017).
[B] R. Seki et al., Nucl. Fusion 59, 096018(2019).
[C] M. Osakabe et al., Nucl. Fusion 46 S911 (2006).
[D] K. Ogawa et al., Nucl. Fusion 52, 094013 (2012).
Speaker: RYOSUKE SEKI (National Institute for Fusion Science) -
Experiments on DIII-D and C-Mod show that high neutral opacity is compatible with a steep density gradient at the plasma edge [1,2]. Future reactors, including ITER, will operate at high neutral opacity, which will strongly limit direct fueling of the pedestal structure inside the Last Closed Flux Surface (LCFS) through ionization from edge sources in comparison with current existing experimental devices. At the highest opacities, the electron density pedestal structure is stiff and is not degraded by increases in fueling and Scrape-Off Layer (SOL) density. These experiments directly question integrated modeling predictions for ITER that show high opacity is incompatible with a steep pedestal density gradient inside the LCFS when particle transport is considered purely diffusive [3].
High neutral opacity means that neutrals will ionize predominately inside the SOL, limiting the impact of edge fueling and peaked density profiles, under the assumption of purely diffusive particle transport. To test whether limited edge fueling in high neutral opacity regimes results in the collapse of the pedestal density profile, as predicted by simulations for ITER [3], we conduct a set of experiments on DIII-D and C-Mod in which we increase the neutral opacity. In our experiment we obtained opacity values that are only a factor 2 lower than those on ITER, while current day machines operate typically at opacity values that are about a factor 10 lower than those expected on ITER. Opacity can be approximated using values of the electron density and the minor machine radius, $n \times a$. The experiments on DIII-D and C-Mod ranging in opacity range from $1.5 \times 10^{19} \; m^{-2}$ to $5.5\times 10^{19} \; m^{-2}$ by operating at various plasma currents as well as adding an additional gas puff. Within this span of opacities, both DIII-D and C-Mod have robust density pedestal structures and increases in fueling and opacity did not affect the maximum density gradient on either machine. The max($1/L_{ne}$) on DIII-D remained close $\sim 8 \times 10^{21} \; m^{-4}$ in absolute value for a fixed IP when the opacity was raised using gas fueling. This is also reflected in the fact that $ne_{ped}/ne_{sep}$ remains constant for DIII-D as well as C-Mod over a wide range of opacities, see figure 1b). Thus, suggesting that changes in neutral penetration has little effect upon the pedestal density structure.
Using the filterscope measurements of $D_{\alpha}$ light at the mid plane inside the LCFS as well as in the SOL, we observe that with increasing opacity, there is a decrease in the neutral density especially inside the pedestal structure on DIII-D. Even though direct fueling through ionization of edge neutrals sharply decreases in the pedestal structure, the pedestal density still increases with increasing opacity, see figure 1a). This agrees with SOLPS-ITER modeling of these DIII-D discharges, which also shows a decrease in neutral density, not just at the midplane, when the neutral opacity increases in these experiments.
While opacity does not exclude a robust pedestal density structure, it does reduce our ability to fuel the core plasma using gas puffing. With increasing opacity, the increase in the pedestal density as a function of the additional gas fueling becomes smaller. This observation is again counter to the integrated predictive ITER H-mode simulations, which indicated that there was no saturation observed in the pedestal or separatrix density with increasing gas fueling. These models rely on fixed transport coefficients in the SOL and use a Bohm/GyroBohm approximation for the transport on closed flux surfaces. The increase in fueling in these simulations increases the SOL and separatrix density and the choice of the transport model for the core results in this increase propagating up to the pedestal density.
In DIII-D at the highest fueling levels, when the outer strike point detaches, we observe a saturation in the SOL density increase due to the formation of a density shoulder. A density shoulder is associated with a strong change in radial transport, which cannot be captured by these integrated models or SOLPS-ITER self-consistently [4]. The strong increase in radial outward transport associated with the formation of a density shoulder effectively impacts the ability to increase the separatrix density. As such, fueling an opaque plasma using a gas puff will be very inefficient or even completely ineffective. It does however not result in the collapse of the steep pedestal density.
To complicate transport models even further, for the first time, an up-down asymmetry in the electron density inside the LCFS was observed in DIII-D H-mode plasmas. The electron density was higher on the LFS close to the X-point, when compared to measurements taken on the LFS close to the crown of the plasma inside the LCFS. This asymmetry has been predicted by neoclassical models [5]. These results are different from prior research where a high electron density region in the HFS SOL region close to the X-point is observed in AUG H-mode plasmas [6].
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-AC02-09CH11466, DE-SC0014264, DE-SC0019302, DE-SC0007880
[1] S. Mordijck et al. Nucl. Fus. (2020) submitted to Nuclear Fusion
[2] J.W. Hughes et al. Nucl. Fus. 58 (2018) 112003
[3] M. Romanelli et al. Nucl. Fus. 55 (2015) 093008
[4] D. Carralero et al. Nucl. Fus. 57 (2017) 056044
[5] R.M. Churchill et al. Nucl. Fus. 53 (2013) 122002.
[6] F. Reimold et al. NME 12 (2017)193–199
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.Speaker: Saskia Mordijck (College of William and Mary) -
Improving Energetic Particle Confinement in Stellarator Reactors
A. Bader$^1$, M. Drevlak$^2$, D.T. Anderson$^1$, C.C. Hegna$^1$, S.A. Henneberg$^2$,
T.G. Kruger$^1$, A. Ware$^3$
1: University of Wisconsin-Madison, WI, USA,
2: Max-Planck Institut fur Plasmaphysik, Greifswald, Germany,
3: University of Montana, MT, USAEnergetic particle confinement is a key issue for the scalability of stellarators to fusion power plants. Prompt losses of alpha particles born from fusion reactions can cause significant material damage. Analytically derived proxies for collisionless energetic particle confinement (1) have been used for the first time to optimize quasihelically symmetric stellarator equilibria (2). This paper will expand on recently published results, along with inclusion of analysis to account for collisional alpha particle transport with reactor relevant alpha sourcing profiles.
A proxy for energetic particle transport, $\gamma_c$, accounts for both the net bounce-averaged radial particle drifts, a quantity to be minimized, and poloidal drift, a quantity to be maximized. Minimization of 𝛾c corresponds to aligning contours of the second adiabatic invariant, $J_\parallel$, to flux surfaces. This metric has been included in the ROSE optimization code (3) and used to optimize equilibria for good energetic and neoclassical particle transport. Previous results indicate that two classes of stellarators, quasihelically-symmetric stellarators, and maximum-J stellarators should have better energetic particle transport than other configurations (4). This paper focuses on optimizations of quasihelically symmetric stellarators. Results indicate the existence of equilibria which nearly eliminate all collisionless losses within the plasma mid-radius (figure 1a) for an ARIES-CS scale reactor (450 m3, 5.8 T). These configurations are obtained by optimizing simultaneously for $\gamma_c$ and a metric for quasihelical symmetry. Interestingly, configurations with improved energetic particle transport did not correlate with improvements to neoclassical transport in the 1/𝜈 regime as measured by $\epsilon_\mathrm{eff}$.
It is well known that the ripple associated with realistic plasma coils can negatively affect alpha particle confinement. However, using coil optimization codes REGCOIL (5) and FOCUS (6), we show that it is possible to reproduce configurations with high enough fidelity that the alpha particle confinement is not significantly degraded (figure 1b). A key feature for the coil optimization is the realization of the equilibria with coils placed farther from the plasma, thus reducing high order harmonics associated with the toroidal mode numbers equivalent to the coil number. These high order modes have been previously found to be deleterious to energetic particle confinement on both stellarators and tokamaks.
New results will be presented regarding alpha particle transport that include collisions with the background plasma. Collisional calculations require additional assumptions about configuration parameters, most importantly the density and temperature profiles which govern the collisional equations and the alpha particle source distribution. We show that when including collisions, configurations exist at the ARIES-CS scale and with ARIES-CS parameters with total energy loss below 4% (figure 2a) with most of the losses occurring from particles born outside the midradius (> r/a = 0.55) (figure 2b). Results will be presented that show energetic particle transport under a variety of density and temperature profile assumptions.
In light of these results, the outlook for energetic particle optimization for quasihelical stellarators is bright. The results presented here represent only a first attempt and with improved optimization algorithms better configurations may yet be found. Additionally, further improvements in stellarator coil design can help gain confidence that such configurations are realizable.
References:
(1) V. V. Nemov et al. Physics of Plasmas 15 052501 (2008)
(2) A. Bader et al. Journal of Plasma Physics 85 5 (2019)
(3) M. Drevlak et al. Nuclear Fusion 59 016010 (2018)
(4) W. Lotz et al. Plasma Phys. Control. Fusion 34 1037 (1992)
(5) M. Landreman Nuclear Fusion 57 046003 (2017)
(6) C. Zhu et al. Nuclear Fusion 58 016008 (2018)Work supported by DE-FG02-93ER54222, DE-FG02-99ER54546 and UW2020 135AAD3116
Speaker: Aaron Bader (University of Wisconsin-Madison) -
New DIII-D results may explain why achieving ELM suppression by resonant magnetic fields (RMPs) remains elusive in double null (DN) diverted configurations: the lack of ELM suppression in DN correlates with a damped high-field side response of field-aligned structures that could be indicative of a missing resonant tearing needed to stop inward growth of pedestal. This is found despite favorable conditions for RMP suppression in lower single null (LSN): low $\Omega_{E\times{B}}$ aligned with a resonant surface at the pedestal top at low $n_{e,ped}$. The DN configuration is advantageous for future machine design as it allows improved divertor power handling and particle control, but still needs ELM handling solutions and may not be compatible traditional RMP ELM suppression driven from low-field side coils.
In experiments where the magnetic balance is varied from LSN toward DN, ELM suppression was obtained for $dR_{sep}<-1.7$ cm, where $dR_{sep}$ is defined as the separation between the separatrices from the lower null and upper null at the outboard midplane. In discrete steps of $dR_{sep}$, $q_{95}$ is scanned to find a window in ELM suppression under the model that aligning a resonant surface in a region of low $\Omega_{E\times{B}}$ results in resonant tearing inhibiting the inward growth of the pedestal otherwise leading to an ELM {1}. Results of these scans are shown in Figure 1a where the values of $\Omega_{E\times{B}}$ at resonant surfaces are within $\pm 2\%$ of $\psi_{ped,top}$ (to account for uncertainty in profile fitting). Figure 1b shows $n_{e,ped}$ where ELM suppression was achieved (with the largest value of $n_{e,ped}\sim2.4×10^{19} m^{-3}$). This shows for each ELM-suppressed discharge, a resonant surface is near the pedestal top with $\left|\Omega_{E\times B}\right|<20$ krad/s.
In near balanced DN ($dR_{sep}\sim-0.1$ cm), similar scans show that nominal ELM suppression conditions are demonstrated while still ELMing. ELM suppression is not achieved over a range of $q_{95}$ from 3.4 to 4.1 where it was observed in LSN. This is shown in Figure 1c where $\Omega_{E\times{B}}$ at resonant surfaces aligned within $\pm 2\%$ pedestal top are $\left|\Omega_{E\times B}\right|<10$ krad/s—a tighter range than in LSN. These discharges also achieve a lower value of $n_{e,ped}$ than the highest value suppressed in LSN. The pedestal temperature width is consistently wider in ELMing DN plasmas compared to the ELM suppressed LSN plasmas. This leads to a wider total pedestal pressure, consistent with lacking a mechanism inhibiting the pedestal inward growth.
The 3D plasma response to applied RMPs measured on the high-field side (HFS) drops in plasma shapes transitioning from LSN to DN and recovers in upper single null (USN) as shown in Figure 2. The plasma response on the low field side (LFS) remains relatively constant during the shape transition. The reduced HFS response is found similarly for $n=2,3$ over a range of $q_{95}$ from 3.4 to 5. This feature is found broadly across a range of $\left|dR_{sep}\lt0.1\right|$ cm indicating it is not restricted to exactly balanced DN or specific pedestal conditions.
Linearized single-fluid resistive MHD modeling with M3D-C1 shows relatively good agreement with plasma response measurements transitioning from LSN to DN for both the HFS and LFS indicative of a strong damped of perturbations on the HFS in DN. This is further illustrated in the modeled $T_e$ perturbations in Figure 2 where the perturbations are strongly damping on the HFS in DN. This can be partially understood using a simple geometric model assuming field-aligned resonant perturbations. Field-aligned modes driven from the LFS (as is the case with I-coils in DIII-D) are connected to the HFS through a region of low poloidal field in the presence of a secondary null. This can lead to interference of radially-separated resonant modes on the HFS. In balanced double null, this leads to the strongest interference on the HFS.
Results presented here are consistent with using HFS response as a proxy for local tearing drive responsible for ELM suppression by stopping inward growth of pedestal. This is consistent with previous results showing correlation of HFS response and tearing drive needed for ELM suppression {2}. The benefit of DN in power handling resides in a narrow region of $dR_{sep}$ where exact splitting of the heat flux depends on cross-field drifts and has been shown to balance at $dR_{sep}\sim0.25$ cm (near double null) with the ion $\nabla B$ drift directed to the lower divertor {3}. This region of $dR_{sep}$ lies within the damped HFS response and lack of demonstrable ELM suppression. If this model is correct, we can use it to optimize shape and coil positions to better attempt ELM suppression.
This work was supported in part by the US Department of Energy under DE-AC05-00OR22725, DE-FC02-04ER54698, and DE-AC02-09CH11466.
{1} Snyder P.B. et al 2012 Phys. Plasmas 19 056115; Nazikian R. et al 2015 Phys. Rev. Lett. 114 105002; Paz-Soldan C. et al 2019 Nucl. Fusion 59 056012.
{2} Paz-Soldan C. et al 2015 Phys. Rev. Lett. 114 10500
{3} Leonard A.W. et al 2005 Fusion Sci. Technol. 48 1083Speaker: Morgan Shafer (Oak Ridge National Laboratory) -
Improved understanding of the mechanisms that govern thermal transport in the pedestal region is crucial for determining the fundamental processes behind the L-H transition and pedestal structure, and providing a foundation for predicting and optimizing the pedestal and performance of future devices such as ITER. We report world first inferred ion and electron heat fluxes in the pedestal region of deuterium plasmas using direct measurements of the main-ion temperature for power balance across ion collisionalities between 0.1 and 1.2. The ion heat flux (Qi) in the narrow H-mode pedestal, evaluated in the last 10% of the ELM cycle on DIII-D H-modes is compared with neoclassical transport simulations from the NEO code (1), showing agreement within a factor of 2 at higher collisionality (v*~1.2), but increasing anomalous transport as the collisionality is decreased (Fig 1a). This discrepancy suggests that turbulent transport becomes increasingly important for ion thermal transport in the pedestal at lower collisionalities.
Accurate assessments of the ion heat flux on several devices has historically been impeded by anomalies in the impurity charge exchange recombination spectroscopy (CER) temperature measurements near the plasma edge and relying on the assumption that the main-ion temperature is equal to the impurity based measurements. This can lead to a large ion-electron collisional exchange term that produce physically suspicious negative ion heat fluxes near the plasma edge, with a corresponding attribution of excessive transport through the electron channel. The development of main-ion CER (MICER) in the pedestal region on DIII-D (2,3) has allowed direct measurements of the D+ temperature which is free from these anomalies and improves the interpretation of both ion and electron transport inferred from power balance calculations. The main-ion measurements show significant differences compared with the impurities in the H-mode transport barrier. In particular, there can be large (>2x) differences between the main-ion and impurity temperatures measured in the steep gradient region. These differences are contrary to expectations based on equilibration time. Many of the discrepancies at the pedestal top are resolved by including Zeeman and fine structure broadening of the impurity measurement. In the steep gradient region increasingly accurate calculations of the ion motion may be needed to understand the charge exchange signals, this is being accessed using the SPIRAL code. Use of direct main-ion measurements for the temperature profile in the steep gradient region can have large impacts on the ion heat flux when assuming the main-ion temperature is the same as the impurities (Fig 2a) as opposed to using the direct measurements (Fig 2b). Additionally, the use of the steeper main-ion temperature gradient has the potential to impact the stability of microturbulence calculated in gyrokinetic simulations.
A dedicated experiment was run to acquire these new measurements across a range of collisionalities, providing the inputs to interpretive TRANSP calculations of the ion (and electron) heat flux using conditionally averaged profiles from 90-100% of the ELM cycle. The inferred ion thermal transport in the steep gradient region of the H-mode pedestal rises above the neoclassical level as the collisionality is decreased (Fig 1a) suggesting that anomalous ion thermal transport becomes increasingly important at low collisionality conditions relevant to the range expected in ITER. BES measurements show increasing ion scale fluctuations in the pedestal region as the collisionality is reduced (see Fig 1b) suggesting increased ion scale turbulence may be responsible for the anomalous ion thermal transport.This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-FC02-04ER54698, DE-AC02-09CH11466, DE-SC0019352, DE-FG02-08ER54984,DE-FG02-08ER54999, DE-AC05-00OR22725, DE-FG02-07ER54917, DE-SC0020337.
(1) E. A. Belli, et. al., Plasma Phys. Control. Fusion, 50, 095010 (2008)
(2) B. Grierson, et. al., Rev. Sci. Instr. 83, 10D529 (2012)
(3) S. R. Haskey, et. al., Rev. Sci. Instr. 87, 11E553 (2016)Speaker: Shaun Haskey (Princeton Plasma Physics Laboratory) -
New power law scalings of the error field (EF) penetration thresholds across a wide range of tokamaks have been developed for toroidal mode numbers n=1 and 2 and project values for ITER that the construction tolerances and correction coils satisfy. This paper presents a multi-variable n=2 threshold regression across a wide range of densities, toroidal fields, and pressures in 3 machines (DIII-D, EAST, and COMPASS) using a common metric to quantify the EF in each device. It compares this new n=2 scaling to updated n=1 scalings using a larger 6 machine ITPA database. The results validate nonlinear single-fluid MHD simulation scalings, which are used to lend confidence to the projected scalings to ITER. These projections set the tolerances for non-axisymmetric components (like Test Blanket Modules) and the corresponding requirements for EF correction coil arrays in ITER.
Nonaxisymmetric fields four orders of magnitude smaller than the axisymmetric field can drive islands that cause disruptions in tokamaks; and the GPEC overlap metric, $\delta$, provides a way of identifying and quantifying the most dangerous of these asymmetries. The metric uses a resonant field spectrum determined from a combination of the applied field and the plasma response (amplification and/or shielding of various components)$^{1, 2}$, surpassing the robustness of the old “3-mode” vacuum model$^{3}$. Tolerances for the design and optimization of tokamak coils have been projected to ITER using fit scalings of this overlap metric with macroscopic 0D parameters that are easily identified both in current and for future devices.
The n=2 database, consisting of 3 devices and a parameter range shown in Figure 1, reveals tolerances of a similar order of magnitude to those for n=1 in current devices. Experimental thresholds are determined by ramping up artificial EFs using 3D field coil arrays until a core island penetration event is observed and recording the corresponding amplitude in EF overlap $\delta$. The predictions of a power law fit to this data using a kernel density weighted regression are compared to the true experimental thresholds in Figure 2. The full regression used has a density scaling exponent of 1.07±0.09, a toroidal field exponent of -1.52±0.2, a major radius exponent of 1.46±0.09 and a normalized pressure ($\beta_N/\ell_i$) exponent of 0.36±0.11. This fit projects very high n=2 thresholds for ITER due to the strong size scaling.
Although single parameter scans across regimes accessible by a single machine can reveal varied behaviors, the multi-variable, multi-machine scaling provides the most robust projection to new devices. Figure 3 shows individual density scans in DIII-D can exhibit a wide variety of trends depending on the initial target plasma. The second panel, however, shows how the seemingly discrepant DIII-D experiments are unified when normalizing by the general toroidal field and pressure scalings. Observations at even higher density show that ohmic confinement regime transitions in any given device can drastically alter the density scaling in that particular experiment. The ratio of non-resonant to resonant 3D field applied in a given experiment also alters the dynamics through neoclassical toroidal viscosity (NTV) braking. Recent advances in optimization of nonresonant fields for robust quasi-symmetry minimizing such secondary effects are beyond the scope of this dominant-order resonant field analysis. The local and secondary phenomena are purposely smoothed out in the multi-machine scaling presented here, in an intentionally analogous manner to the treatment of single-machine variations in the international confinement scalings that have proven so useful for the fusion community.
The nonlinear, single fluid MHD code TM1 has been used to model the experimental scalings, providing confidence in some scalings and insight into experimental needs. The model reproduces the experimentally observed toroidal field and $\beta_N/\ell_i$ scaling, and shows the scalings hold out to ITER values. The density scaling exponent calculated by TM1 falls below the experimental n=2 fit, closer to the better constrained n=1 empirical fit. This, and a large discrepancy between n=1 and 2 size scalings, identifies what n=2 data is needed to improve ITER projections. The code projects n=2 thresholds in ITER roughly 2-3 times that of the projected n=1 thresholds, consistent with observations in existing devices to date.
This combination of robust, cross-regime experimental scalings and tightly coupled modeling project EF tolerances of above a Gauss for both n=1 and 2 EFs in ITER, which are criteria the ITER construction tolerances and correction coils easily satisfy.
This work was supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC02-09CH11466 and DE-FC02-04ER54698.
- J.-K. Park, A.H. Boozer, et. al, Physics of Plasmas 16, 056115 (2009).
- N.C. Logan, C. Paz-Soldan, et. al, Physics of Plasmas 23, 056110 (2016).
- J.-K. Park, N. C. Logan, et. al, ITPA MHD, ITER Report #ITER_D_UMLSUW (2017).
Speaker: Nikoas Logan (Princeton Plasma Physics Laboratory) -
Energetic particle (EP) instability models based on gyro-Landau closure techniques (1) have addressed important nonlinear simulation and linear stability survey challenges that will be critical for the understanding and control of burning plasmas in ITER and the next generation of fusion systems. The long-term intermittency and frequency spreading characteristics of saturated Alfvén instabilities (Fig. 1) are important features that will determine peak wall heat loads from escaping energetic ions as well as play a central role in regulating energetic particle anomalous transport levels. Reduced models using gyrofluid closures such as the TAEFL (2) and FAR3d (3) models, allow routine simulation of Alfvénic instabilities far into the nonlinear regime and demonstrate the dependence of the nonlinear dynamics on source-sink balances, zonal flow/current damping, and turbulence levels in the thermal plasma (modeled in Fig. 1 using diffusivities). The efficiency of this approach also facilitates linear instability surveys as profiles/parameters/plasma shapes are varied (Fig. 2); this capability is essential for simulating the dynamical changes that occur in realistic simulations of tokamak discharge evolution.
Reduced physics models for EP instabilities based on gyro-Landau closures (2,3) offer a computationally efficient tool for understanding effects of macroscopic parameters that control the saturated turbulent state of these instabilities. These models are based on an optimized set of closure coefficients that provide good fits to response functions derived from kinetic theory. Wave-particle resonances that exist in phase space are effectively mapped into real space while preserving growth/damping effects associated with EP energy distribution functions. These reduced models have been checked against more complete models (4). In the nonlinear regime, mode-coupling nonlinearities that drive zonal flows, zonal currents, localized flattening in the EP pressure profile, and couple to damped modes are included. The viability of simulating EP modes far into the nonlinear limit with these models is based on using a third order predictor-corrector time stepping algorithm, and the fact that there is no particle noise or small perturbation ordering, as typically limits simulation times for particle-based models. In regimes with moderate drive, balanced sources/sinks and where zonal flows/currents dominate, predator prey phenomena (5) are evident (black waveform in Fig. 1). As diffusivities are lowered or instability drive is increased, bursting phenomena and frequency chirping are observed (green waveform in Fig. 1). Saturation into a steady level can be achieved as EP drive is diminished by EP profile changes (Fig. 3). EP nonlinear transport fluxes (Fig. 4) and 3D convection cell phenomena can also be predicted.
The reversed shear/high bootstrap current tokamak regime offers the possibility of steady-state operation but is often unfavorable for EP-driven instabilities. Reversed shear/high bootstrap current discharge formation is a dynamical process and EP instability evaluation requires consideration of evolving profiles/parameters and plasma shapes. The gyrofluid model provides a unique, efficient multiple eigenmode solver approach that has been applied to these regimes. Low frequency Alfvén-acoustic (BAE/BAAE) instabilities are of particular concern since they can lead to larger radial transport levels than higher frequency AE modes (6). The survey capability of FAR3d has been applied to a DIII-D discharge with evolving q-profiles where multiple low frequency modes were observed (Fig. 2) and the results have shown similarities with the mode frequency evolution from the experiment.This work was supported by the U.S. Department of Energy under DE-AC05-00OR22725, DE-FC02-04ER54698, and the U.S. DOE SciDAC ISEP Center.
(1) G. W. Hammett, F. W. Perkins, Phys. Rev. Lett. 64, 3019 (1990).
(2) D. A. Spong, et al., Nuclear Fusion 53, 053008 (2013).
(3) J. Varela, et al., Nuclear Fusion 59, 046017 (2019).
(4) S. Taimoirzadeh, et al., Nuclear Fusion 59, 066006 (2019).
(5) D. A. Spong, Plasma and Fusion Research 9, 3403077 (2014).
(6) W. Heidbrink, et al., 16th IAEA Technical Meeting on Energetic Particles (EPPI 2019).Speaker: Dr Donald Spong (Oak Ridge National Laboratory) -
Energetic-particle (EP)-driven instabilities such as Toroidal Alfvén Eigenmodes (TAEs) can be responsible for the effective ion heating via collisionless EP energy channeling. Although the quantitative estimation of the EP transport by instabilities has been actively conducted [1–3], studies on the energy channeling have been limited [4,5]. It is important to estimate collisionless energy transfer from EP to ions for the development of burning plasma scenarios with effective nuclear reaction, and to compare this beneficial effect against EP transport due to TAEs.
For high-n TAEs which is expected to be in ITER, the modes strongly overlap with slightly different radial locations and eigenfrequencies. It was suggested that the interaction between the modes such as ion Compton scattering (ICS) [5–7] could be a dominant mechanism for the saturation of high-n TAEs, while the nonlinear wave-particle interaction by EP redistribution is dominant for low-n TAEs [8].
The wave-kinetic equations derived in [5,6] which describe the evolutions of the TAE mode energy lead to the predator-prey system between the linearly unstable modes in the TAE gap and the stable modes near the continuum. Figure 1 (a) shows the time-evolution of the TAE energy and the ion heating, and (b) shows the predator-prey like energy transfer by ICS in a single burst. Since the TAE gap structure depends on the radial profiles, the mode amplitude evolves in real space as well. Figure 1 (c) shows the mode energy propagates in a radial direction by the nonlinear mode transfer, which can lead to an avalanching EP transport.
EP transport by TAEs changes the linear growth rate, so that makes the TAEs eventually saturated. At the saturated phase over the transport time-scale when the nonlinear effect balances with the linear growth, we can estimate the net ion heating rate by TAEs from the linear growth rate.
In order to estimate the energy channeling as well as the EP transport by TAEs, we used TGLF to calculate the stability of TAEs [3 ]. Figure 2 (a) and (b) show the TGLF calculation of TAE stabilities. We see that there is a critical gradient of EPs to destabilize TAEs. Figure 2 (c) shows the EP profile transported by TAE which is calculated by critical gradient model (CGM) [3 ]. We also calculated the energy channeling through TAEs by using a property held during the wave-particle interaction, (dE/dt)/(dP_ϕ/dt)=ω/n which is valid for single-n TAE. We implemented an integrated simulation with ASTRA for ITER baseline scenario. We used TGLF calculation for EP transport, and BgB model for thermal plasma transport. In order to maximize EP effect, we set NBI-only external heating with P_NB=50 MW.
Figure 3 (a) and (b) show the thermal plasma profile and EP density profile for Q=10 ITER reference plasma without TAE effect. Figure 3 (c) shows the ion temperature profiles for 3 cases, (1) without considering TAE, (2) considering EP transport by TAE, and (3) considering EP transport and energy channeling. If we consider the TAE-induced EP transport, Q drops 18 %, and increases by 4 % if the energy channeling is considered. Figure 3 (d) shows the heating profile by EPs and the energy channeling. The ion heating by energy channeling is ~6 MW, about 10 % of the external heating, but the temperature difference is not much significant due to the profile stiffness.
References:
- L. Chen and F. Zonca, Rev. Mod. Phys. 88 015008 (2016)
- M. Podestà et al., Phys. Plasmas 23 056106 (2016)
- E.M Bass and R.E Waltz, Nucl. Fusion 60 016032 (2019)
- A. Bierwage, N. Aiba and K. Shinohara, Phys. Rev. Lett. 114 015002 (2015)
- T.S. Hahm, Plasma Sci. Technol. 17 534 (2015)
- T.S. Hahm and L. Chen, Phys. Rev. Lett. 74 266 (1995)
- Z. Qiu, L. Chen and F. Zonca, Nucl. Fusion 59 066024 (2019)
- H.L. Berk and B.N. Breizman, Phys. Fluids B 2 2246 (1990)
Speaker: Jaemin Seo (Seoul National University) -
In this work we show that nonlinear MHD plasma response simulations are essential for understanding and predicting accurate heat and particle flux striations in DIII-D during ELM suppression. Understanding the nature of heat and particle distributions on the divertor plates due to splitting of the separatrix by 3D magnetic perturbation in RMP ELM suppressed discharges is an important issue for preventing the premature degradation of divertor components in ITER. Our simulations now match the experimental observations in DIII-D. We further show that in cases with no observable heat flux splitting during ITER-like shaped DIII-D discharges with RMP ELM suppression the loss of heat flux splitting may be due to an increase in carbon radiation above the surface. Since carbon will not be present in ITER, these results raise a question concerning whether a similar effect will be observed in ITER during divertor detachment operations with medium Z impurities such as neon or nitrogen.
In DIII-D, strike point splitting is routinely observed in the divertor particle flux during operation with RMPs (1). We use small modulations of the in-vessel I-coil n=3 RMP current amplitudes or toroidal rotation of n=2 RMP fields in order to modify the poloidal spectrum of applied perturbation fields. These modulations affect the position and the size of the divertor footprints, and provide useful information to validate different numerical models for non-axisymmetric footprint lobes by comparing to high-precision visible imaging and IR tomography experimental measurement. The observed splitting is consistent with the toroidal mode number n of the applied perturbation but the measured separation of the divertor particle flux footprint lobes exceeds predictions of vacuum (TRIP3D) and linear plasma response (MARS) models by factors of 3-5 (2). The plasma response to the RMP in ITER-like conditions using linear, resistive MHD simulations (M3D-C1, NIMROD) with both single-fluid and two-fluid models is dominantly a screening response that reduces the divertor footprint splitting below the vacuum model predictions. The nonlinear MHD simulations using JOREK, a fluid model for the main plasma and the neutrals, show a much better match to the measured separation of the lobes (Figure 1). The time evolution of the neutrals in JOREK is described by a diffusion model combined with a boundary condition that reflects outgoing ions as incoming neutrals (3). In this work we also examine the possibility of near-SOL field lines affecting the formation of the outer-most lobes in the measured particle flux footprints.
At the same time, the heat flux to the divertor often does not show significant splitting in DIII-D ITER-like RMP ELM-suppressed discharge, as displayed in Figure 2, which is potentially good for ITER. One hypothesis suggested by observations is that the lack of splitting in heat flux may be related to the C-III volumetric radiation immediately above the surface that obscures the lobe structure. We performed fully 3D plasma-fluid and kinetic edge neutral transport Monte-Carlo EMC3-EIRENE simulations using vacuum and linear plasma response MHD solutions, and the results highlight the effect of different levels of carbon impurity radiation near the strike point on the divertor footprints. We also examined a possibility of heat flux lobe smearing partially due to the ion grad-B drifts as MAFOT simulations suggest. MAFOT results also indicate minor effects on heat flux footprints due to ExB fields and plasma sheath near the divertor surfaces.
(1) Moyer et al., RSI 89, 10E106 (2018)
(2) Orlov et al., IAEA FEC 2018 (Ahmedabad, India)
(3) Hujismans et al., 46th EPS, P1.1059 (2018)This work is supported by the US Department of Energy under DE-FG02-07ER54917, DE-FG02-05ER54809, DE-FC02-04ER54698, DE-SC0012706, DE-AC52-07NA27344, DE-NA0003525, DE-AC02-09CH11466 and DE-AC04-94AL85000.
Speaker: Dmitri Orlov (University of California San Diego) -
Novel internal measurements and analysis of ion cyclotron frequency range fast-ion driven modes in DIII-D are presented which advance understanding of the dynamics controlling mode stability and thereby the physics basis for prediction of fast-ion (e.g. alpha) transport in burning plasmas. Observations, including internal density fluctuation ($\tilde{n}$) measurements obtained via Doppler Backscattering (DBS) [1], are presented for two classes of instability: modes at $f \sim 0.6 f_{ci}$ (Fig. 1) and modes at low harmonics of $f_{ci}$. The former are identified as global Alfvén eigenmodes (GAE) (never before reported in a conventional tokamak) and the latter as coherent Ion Cyclotron Emission (ICE—so named because until now observations have always been via external RF antennae [2]). Both modes are excited by Doppler-shifted cyclotron resonance with fast-ions and observation of these modes in an experiment reveals the creation or movement of velocity-space inversion features in the distribution of fast-ions. Multiple aspects of the theory for this instability mechanism are validated. Highlights include: 1) a fast-ion density stability threshold is demonstrated for the GAEs [3] utilizing a unique DIII-D beam capability, consistent with a fundamental theoretical expectation that mode stability is controlled by competition between damping from various processes (e.g. Landau damping) and the drive from fast-ions; 2) recently developed analytic theory [4] predicting the frequency range for unstable GAEs is supported; 3) simulations [5] show unstable GAEs with frequencies and toroidal mode nubmers similar to experiment; 4) observations of ion-cyclotron harmonic $\tilde{n}$ in the core (Fig. 2) challenge a prevalent theory (see [6] and references therein) predicting the spatial localization of modes, and 5) simulations of edge ICE observations [7] from a DIII-D H-mode plasma agree well with experimental measurements from edge magnetics. Analysis of the ion-cyclotron harmonic $\tilde{n}$ observations is also presented leveraging the theory for the instability mechanism to advance understanding of ELM-induced fast-ion transport.
A fundamental element of stability theory for fast-ion driven modes is validated in an experiment demonstrating a fast-ion density threshold for fast-ion driven modes [3]. The modes, with frequencies $\sim 5.5 - 5.6 \text{ MHz} \sim 0.6 f_{ci}$ (Fig. 1a), are excited by off-axis nearly perpendicular beam injection. Using a unique DIII-D beam capability, beam current is slowly ramped down (Fig. 1b) at constant voltage in order to change the resonant beam ion density without significantly changing the resonance itself. The slow ramp is designed to create an approximately self-similar fast-ion distribution (neglecting initial transients at beam turn-on) changing only by a scale factor over time. The mode is abruptly stabilized as beam current crosses below a threshold of ~ 49 A (Fig. 1b), consistent with the expectation for a stability threshold set by competition between fast-ion drive and mode damping processes. (Note that a sawtooth briefly re-excites the mode after the initial stabilization. Note also that the mode frequencies sweep downward slowly as beam fueling increases density, and sawtweeth modulate the frequencies on a ~ 25 ms time-scale.)
The modes in Fig. 1 are identified as Doppler-shifted cyclotron resonant GAEs by analysis of the fast-ion distribution for destabilizing resonances, taking into account measurements of the toroidal mode number (via external magnetics) and internal mode structure. Fig. 1c shows a spectrum of core measurements (via DBS in reflectometry mode) with a narrow peak corresponding to the dominant GAE. Fig. 1d shows the time-dependent amplitude and radial mode structure, which is broad, obtained from a spatial array of simultaneous measurements. The identification is consistent with recently developed analytic theory [4] which, for these experimental conditions, predicts GAEs to be unstable in a narrow range of frequencies, $0.5 < \left. f \middle/ f_{ci} \right. < 0.7$. The identification is also consistent with simulations using the Hybrid MHD code (HYM, [5]) for the experimental conditions. The simulations found unstable, core-localized Doppler-shifted cyclotron resonant GAEs with mode numbers close to the measured value and $f ≳ 0.6f_{ci}$.
Unique observations of fast-ion driven modes with frequencies above $f_{ci}$ ($f \sim 10 - 50$ MHz $\sim 1 - 5 f_{ci}$) test numerous aspects of theory. Measurements of $\tilde{n}$ are obtained in a variety plasma conditions via DBS inside the plasma across a broad spatial range, giving mode amplitude and structure. Observed frequencies can match low harmonics of edge and core $f_{ci}$. Fig. 2a shows an example from an H-mode plasma which poses a challenge to a prevalent theory that predicts the modes to be radially localized close to where mode frequency matches a harmonic of $f_{ci}$ (see [6] and references therein). Fig. 2a shows a mode at $f \sim 15$ MHz, matching $f_{ci}$ on axis as expected, but it also shows modes at $f \sim 10$ and 20 MHz, unexpectedly matching the 1st and 2nd $f_{ci}$ harmonics at the edge. Simultaneous measurements at multiple radial locations show that the 10 MHz mode is actually core localized. Simulations with the EPOCH code test theoretical and computational analysis of ICRF modes, which indicates the central role of velocity-space population inversion for fast-ions for which $v_\perp \sim v_{Alfvén}$. Simulations of related edge ICE ($f ≥ f_{ci}$) observations from a DIII-D H-mode plasma agree well with experimental measurements from the magnetics, down to the level of the relative amplitudes of the first four ICE spectral peaks [7].
The observations in Fig. 2 also test other aspects of theory. The mode amplitudes strongly increase with total power from two beams injecting (nearly) tangentially in the co-current direction (Fig. 2b) at ~ 80 keV. However, the amplitudes are insensitive to power from beams with other injection geometries (e.g. co-current nearly perpendicular). This sensitivity to injection pitch and direction is consistent with expectation for the Doppler-shifted cyclotron resonance mechanism. The rapid switching on/off of the beams reveals a dynamic consistent with another aspect of the theory. In the ~ 10 ms period after each step-up in beam power, the mode amplitudes surge and then decay. This time scale is comparable to a fast-ion slowing down time, during which the high energy part of fast-ion distribution evolves from an initial bump-on-tail, with a positive energy gradient favorable to mode excitation, to a slowing down distribution, with a negative gradient. Fig. 2b also shows that ELMs transport fast-ions in the resonant region of phase space. In particular, when an ELM occurs during injection, the modes are transiently excited, consistent with the expectation that these modes can be excited by fast-ions ejected by MHD events [2].
In conclusion, these results strengthen the physics basis for predicting fast-ion driven mode activity and concomittant fast-ion tranport in future burning plasmas. Comparisons of experiment with analytic theory and simulation show areas of agreement, building confidence in the predictive power of these tools. Observations of ICE advance understanding of ELM-induced fast-ion transport.
This work was supported in part by the US Department of Energy under grant and contract numbers DE-FC02-04ER54698, DE-FG02-99ER54527, DE-SC0011810, DE-SC0019352, DE-SC0020337, and DE‐AC02‐09CH11466. It was supported in part from the RCUK Energy Programme grant no. EP/T012250/1, and carried out in part within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053.
[1] WA Peebles, et al., RSI $\bf81$ (2010) l10D902.
[2] KG McClements, et al., NF $\bf55$ (2015) 043013.
[3] SX Tang, et. al., APS DPP Meeting 2019, BP10.43.
[4] JB Lestz, et al., PoP $\bf27$ (2020) 022513.
[5] EV Belova, et al., PoP $\bf26$ (2019) 092507.
[6] KE Thome et al. NF $\bf59$ (2019) 086011.
[7] A Zalzali, R O Dendy, K E Thome et al, in preparation.Speaker: Dr Neal Crocker (University of California Los Angeles) -
Runway electrons (REs) [1] in a tokamak is of great concern irrespective of the size of the machine. Such runway electrons carries significant amount of plasma energy of several MeV can severely damage the first wall and in-vessel components of the tokamaks [2] as well as can interfere with the complex plasma phenomena like plasma equilibrium, MHD instabilities and plasma disruption. Therefore the control of runway electrons are of great concern for ITER and DEMO like reactors. Similarly for reliable operation of smaller tokamaks such runway electrons should be suppressed or extracted without affecting the plasma operation. Several techniques exists viz., massive gas injection (DIII-D, TEXTOR), magnetic field perturbation (JT-60U, Versator I), resonant magnetic field perturbation (TEXTOR), additional gas-puff (JET, ASEDEX, ADITYA) [3] , ECH heating and LHCD (FTU) [4] for suppression/extraction the runaway electrons in tokamak. Another magnetic field perturbation technique similar to Versator I tokamak [5], for runway electron extraction has been applied in the ADITYA tokamak experiment. In ADITYA local vertical field coils (LVF coils) to generate local magnetic field perturbation has been successfully applied to extract runway electrons during early phase of discharge time (0.5 - 15 ms) which improved the plasma performance as reported in ref [6]. In this work, an attempt has been made to numerically model the runway electron extraction experiment using LVF coil in ADITYA. All coil systems to generate magnetic field configuration and the runway electron dynamics without radiation loss of the RE has been taken into consideration in the numerical simulation to explain the experimentally observed extraction of the RE in ADITYA tokamak. The magnetic field generated by all coil systems of ADITYA are numerically modeled and calculated by EFFI code [7]. The field calculated by the EFFI code then used by the PARTICLE3D code [8] to track the runway electrons following relativistic dynamics without radiation loss. PARTICLE3D stops following the electron when it hits the vessel wall so as to record that hit point which can be used to match the experimental observations in case of localized hard X-ray emission or physical damage of the wall.
The Ohmic coil system of ADITYA consists of one single air-core solenoid and 4 pair of vertical field coils. Toroidal coil systems of ADITYA has 20 TF coils to generate magnetic field of 0.75 - 1.5 T at the major radius 0.75 m. LVF coils are a pair of up-down symmetric coils placed on top and bottom TF coil I-Beam of ADITYA tokamak. The dimensions of the coils are ID $\approx$ 42.4 cm, OD $\approx$ 51 cm, $\Delta$r $\approx$ 4.3 cm, $\Delta$Z $\approx$ 8.5 cm, conductor dia. 0.85 cm, number of turns (n) $\approx$ 50 turns/coil and the distance from the equatorial plane (Plasma centre) to coil center (Z) is $\approx$ $\pm$ 80 cm. The coils are connected in series to produce LVF perturbation in the direction opposite to the actual equilibrium field at that location.
All coils of ADITYA tokamak as detailed above has been numerically modeled to generate the magnetic field configuration of the initial phase of the discharge. The toroidal field considered in this preliminary numerical simulation study is ~ 1.5 T at the major radius 0.75 m. Local vertical field used in the simulation corresponds to 4.2 kA of input current. A single runway electron of 3MeV energy has been traced using PARTICLE3D code with and without LVF coils. The simulated orbits of the 3 MeV runway electron for three cases (I) no LVF (II) +ve LVF and (III) -ve LVF are shown in figure 1(a).
It can be seen that under the influence of LVF coil the trajectories of the runway electron for three different cases differs significantly. While for + LVF coil the electron takes lesser time to be extracted than the without LVF coil, for _ve LVF the time taken is higher than with out LVF coil case. The time scale of the deconfinement is ~$10^{-8}$ sec for 3 MeV electrons which is very small as compared to the experimental time scale of $10^{-3}$ sec. The phenomenon can be easily understood from the plot of Z-position of the particle vs time as depicted in figure 1(b).
Therefore it can be concluded from the simulation results that under the influence of the local vertical field applied in the opposite direction to the equilibrium vertical field (at the mid plane of LVF coil location) the runway electrons gets deconfined. The simulation study shows the capability of the simulation methodology to model the RE deconfinement experiment and to explain the experimental observations. Though simulation of RE dynamics related to other phenomena in the tokamak have been reported earlier, however the simulation of RE deconfinement experiment using local magnetic field perturbation presented here is the first of its kind. Details of modelling and simulation results of the ADITYA RE deconfinement experiments with corresponding input variables would be presented. As RE mitigation is of great concern for larger tokamaks, the discussed simulation of RE de-confinement technique has the potential to be adopted for RE mitigation in such machines.Reference:
[1] H. Knoepfel, D.A Spong, “Runaway Electrons in Toroidal Plasma”, Nucl. Fusion, Vol. 19, No.6, (1979).
[2] C. Rasouli, D. Iraji et al, “Runaway electrons energy measurements using hard X-ray spectroscopy in Damavand tokamak, Rev. Sci. Instrum. 80 013503 (2009).
[3] Yu. K. Kuznetsov et al, “Runaway discharges in TCABR”, Nucl. Fusion 44 (2004), pg. 631 –644.
[4] S A Bozhenkov, M Lehnen et al, “Generation and suppression of runaway electrons in disruption mitigation experiments in TEXTOR”, Plasma Phys. Control. Fusion 50 (2008) 105007 (18pp).
[5] D.S Stone and B. Richards, “Magnetic extraction of runaway electrons from a tokamak plasma”, Phys. Fluids 21(1), January 1978.
[6] Tanna R.L. et al 2015 Novel approaches for mitigating runaway electrons and plasma disruptions in ADITYA tokamak Nucl. Fusion 55 063010
[7] Sackett, S. J., EFFI: A code for calculating the electromagnetic field, force, and inductance in coil systems of arbitrary geometry. 1978.
[8]. Dutta et al 2019 Plasma Sci. Tech. https://doi.org/10.1088/2058-6272/ab2947Speaker: Someswar Dutta (Institute For Plasma Research) -
Tokamak devices aim for magnetically confined burning plasmas in order to reach steady state operations and produce economically exploitable fusion energy. One of the main issues are the strong levels of transport due to the highly nonlinear turbulent plasma behaviour, which causes an increment of heat fluxes with respect to neoclassical theory. It is believed that microturbulence, characterized by small-scale and local resonant excited instabilities, is one of the principal drivers for the confinement degradation. Therefore, controlling and reducing the microturbulent transport is of paramount interest towards the exploitation of future devices.
A possible mechanism for such reduction could be identified in the fast ion population which is generated by neutral beam injection (NBI) and ion cyclotron resonance frequency heating systems in present day tokamaks. Recently, an enhancement of the thermal confinement with respect to the IPB98(y,2) scaling law has been experimentally detected in several devices, such as JET, ASDEX-Upgrade (AUG) and JT-60U, among others, in the presence of a significant amount of fast ions. Later, extensive dedicated gyrokinetic analyses demonstrated that fast ions beneficially impact on the ion-temperature gradient (ITG)-driven transport, reducing and partially suppressing the main ion heat fluxes [1,2]. The suprathermal species account for both stabilization of the linear growth rate and for an extra reduction of the nonlinear heat transport. Linearly, a wave-particle resonant effect - between the frequency of the instability and the magnetic drift frequency of the energetic species - has been established to be the major mechanism for the reduction of the ITG growth rate in JET L-mode scenario studies [3]. Subsequently, a complex multi-phase interaction between the ITG-scale turbulent transport and the zonal flow generation has been shown to affect beneficially the intensity of the heat transport [4]. The same nonlinear positive effect is found also in JET, DIII-D and AUG H-modes [5] and advanced scenarios, all of them being dominated by the ITG instability. In addition, a combination of NB-injected fast ions and α particles have been shown also in ITER predictive hybrid scenario to reduce the nonlinear saturated ion heat fluxes [6]. The latter results is significantly relevant since in ITER the E×B shearing turbulence reduction mechanism is expected to be weak; thus, fast ions could provide a valid alternative for such reduction.
Nevertheless, previous dedicated studies lack of generality, since multiple turbulence regimes, or even alternative ones, can usually be dominant in plasmas beyond the ITG paradigm. As a matter of fact, another relevant source of core microturbulence is represented by the trapped electron mode (TEM), often excited by the efficient electron heating systems or high density peaking, which can be subdominant to ITG modes.
In this paper, extensive gyrokinetic numerical studies have been performed with the local version of the GENE code [7] for a NBI-heated JT-60U hybrid scenario [8], previously identified with TEM dominated core turbulence by linear analyses [9], which share common characteristics with the already analysed JET hybrid scenario 2. The multi-species simulations also include collisions, magnetic fluctuations (both in the parallel and perpendicular directions), and plasma magnetic equilibrium computed by the CRONOS integrated modelling suite of codes [10].
Spectra from linear analyses reveal the destabilization of dominant fast-ion-driven beta alfvénic eigenmodes (FI-BAEs) at low binormal wavenumbers k_y, for the case labelled standard which is set with nominal input parameters computed by CRONOS. Subsequent nonlinear simulations for the same case show that significant electromagnetic fluctuations (identified through the value of the electron-β) drive even more unstable the FI-BAEs, leading to a drastic increase of the thermal and fast ion energy transport with respect to experimental range of values – as it is displayed in Figure 1 for the thermal Deuterium heat diffusivity. Hence, in order to evaluate the fast ion impact on dominant TEM-induced transport, FI-BAEs had to be stabilized. As a result, tuning both the thermal and suprathermal input physical parameters (principally the density and temperature gradients of the main and fast ion species, and also the electron-β), a stabilization of the fast-ion-driven mode has been achieved. Eventually, in this new configuration, the TEM has been found to be the dominant driver of the nonlinear saturated transport. Thus, it is shown that fast ions do not affect the TEM-induced turbulent transport [11], demonstrating that the turbulence suppression due to fast ion presence is not universal – in Figure 2, the heat flux time-traces are shown for comparison between the with and the without fast ions cases. This lack of impact is established up to the definitive excitation of fast-ion-driven modes that leads to a complex scenario in which strong electromagnetic effects, fast ion pressure gradient and thermal turbulent transport are intimately related. Therefore, in contrast to what occurs for ITG-dominated systems, the fast ions do not affect the TEM-induced heat fluxes in this JT-60U hybrid scenario.
A possible explanation for the different impact of fast ions on ITG and on TEM is related to the different saturation mechanism. Indeed, for the ITG instability, zonal flows (ZFs) are well-established to play a significant role in the saturation and also, as already stated, in the beneficial interplay among fast ions. On the other hand, dissipative TEMs, driven mainly by the strong electron temperature gradient, do not saturate through ZFs [12]. Deeper analyses performed for the same JT-60U discharge highlight that no energy exchange occurs between the TEM wavenumber and the zonal component of the flux, even when fast ions are introduced as an active species in the simulations. Furthermore, a local conservation relation for the particle toroidal momentum is derived and then applied to the GENE code in order to study the main physical parameters underlying the ZF excitation in the presence of fast ions, for which a competition among neoclassical damping and turbulent Reynolds-Maxwell stress effects is undergoing.
The results achieved opens the way to a more detailed physical view about the fast-ion effect on the microturbulence saturation process. In this sense, the universal physical mechanism governing the interaction among fast ion pressure gradient, electromagnetic fluctuations and microinstability-induced transport could be unveiled, leading to a better tailoring of the fast-ion species through the control of the external heating systems.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions express herein do not necessarily reflect those of the European Commission.
![Deuterium heat diffusivity for different fast ion pressure gradient cases scanned over the electron-β parameter. ‘Standard’ input parameters case is displayed here. The dashed vertical line represents the nominal value of β [11].](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/807/1.png)
![Timetrace comparison between the electrostatic heat fluxes with and without fast ions of electron and thermal deuterium species for the ‘modified’ case [11].](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/807/2.png)
References
- J. Citrin et al, Phys. Rev. Lett. 111, 155001 (2013)
- J. Garcia et al, Nucl. Fusion 55, 053007 (2015)
- A. Di Siena et al, Nucl. Fusion 58, 054002 (2018)
- A. Di Siena et al, Nucl. Fusion 59, 124001 (2019)
- H. Doerk et al, Nucl. Fusion 58, 016044 (2018)
- J. Garcia et al, Phys. Plasmas 25, 055902 (2018)
- F. Jenko et al, Phys. Plasmas 7(5), 1904-1910 (2000)
- N. Oyama et al, Nucl. Fusion 49, 065026 (2009)
- J. Garcia et al, Nucl. Fusion 59, 093010 (2014)
- J.F. Artaud et al, Nucl. Fusion 50, 043001 (2010)
- S. Mazzi et al, Nucl. Fusion, doi.org/10.1088/1741-4326/ab74a1 (2020)
- D.R. Ernst et al, Phys. Plasmas 23, 056112 (2016)
Speaker: Samuele Mazzi (Aix-Marseille Université, CNRS PIIM, UMR 7345 Marseille, France) -
Introduction
A powerful method for diagnosing runaway electrons in tokamak experiments is to measure the synchrotron radiation that the relativistic electrons emit. The radiation intensity depends on the electron energy, pitch angle as well as the background magnetic field, and using a spatially resolved diagnostic, such as a camera, allows the spatial distribution of electrons to be studied. Synchrotron radiation is therefore sensitive to the details of certain parts of the electron distribution function, making it ideal for validating kinetic models for runaway electrons. The detailed predictions obtained from kinetic runaway models can otherwise be difficult to test experimentally, as one must then usually rely on signals which are averages of the full distribution function.Principles
Synchrotron radiation is the ultra-relativistic extension of cyclotron radiation, and is as such emitted almost exactly along the particle velocity vector due to relativistic beaming effects. This means that while usual cyclotron radiation would give rise to a familiar torus of radiation on a camera image, a camera observing synchrotron radiation would only see one or a few distinct ``patches'' of radiation from one side of the torus, corresponding to the locations in the device where the electrons' velocity vectors are pointing at the synchrotron camera. This complicated dependence of the synchrotron patches on the background magnetic field means that a simulation tool taking the magnetic field geometry into account must be used, and for this reason we have developed the synthetic synchrotron diagnostic SOFT [1]. With SOFT, it is possible to simulate all types of synchrotron diagnostics used in present-day experiments, including visible- and IR cameras [2], spectrometers [3] and MSE polarimeters [4].Dominant particles
Generally, electrons with larger pitch angles and higher energies emit more synchrotron radiation. Conversely, most runaway electrons tend to have very small pitch angles, and if the population has been sufficiently multiplied through large-angle collisions, the number of particles with energy $\gamma$ decreases exponentially with $\gamma$. As a result, only a small subset of the runaway electron population contributes significantly to any synchrotron radiation measurement, corresponding to the particles that dominate emission of synchrotron radiation. This small region in momentum space is therefore the only region which we can hope to explore in true detail, although with the help of state-of-the-art runaway models, this is often sufficient to gain some understanding for the kinetic processes affecting the majority of electrons in the population.A synchrotron image or spectrum usually shows clear similarities to the image or spectrum that would result if all electrons had the same energy and pitch angle as the particle emitting the most synchrotron radiation. Hence, one can often characterize a synchrotron measurement with the ``dominant particle'' parameters. While one should be careful not to conclude that all particles have the same parameters from this, determing the dominant particle parameters can greatly help in modelling the scenario appropriately.
Figure 1 Examples of SOFT simulations for a JET-like ($B_0 = 3.45\,\mathrm{T}$, $R_{\rm maj} = 2.96\,\mathrm{m}$, $a = 1.25\,\mathrm{m}$) circular plasma with a parabolic current profile and on-axis safety factor $q_0 = 1$. The figures show synchrotron (a) polarization fraction and (b) polarization angle measured by a single MSE polarimeter line-of-sight [5] as a function of the runaway electron momentum $p$ and pitch angle $\theta_{\rm p}$.
Polarized synchrotron emission
The light emitted by runaway electrons is linearly polarized, and the polarization degree and direction of this light can be measured using for example MSE polarimeters [4,5]. These polarimeters are usually equipped with a number of narrow lines-of-sight, viewing across the plasma and observing light near the Balmer-$\alpha$ spectral line ($\lambda\approx 656\,\mathrm{nm}$). This is beneficial for observing synchrotron radiation, since the continuous spectrum of synchrotron radiation well covers the polarimeter spectral line, and oftentimes dominates over background radiation.The direction of the synchrotron radiation polarization vector is determined by the local magnetic field direction as well as the direction of motion of the emitting electron. Most of the radiation is observed to have either vertical or horizontal polarization, with very little light consisting of a clear super-position of the two. Hence, in a fixed magnetic field, a radially varying threshold is found in pitch angle across which the observed polarization sharply transitions between the two directions, as is shown in Figure 1. The sudden transition of the polarization vector also gives rise to a significant drop in the polarization fraction around the threshold. These signatures can help in identifying the pitch angle of the dominant particles, which are key to understanding experimental observations.
References
[1] M. Hoppe et al, Nucl. Fusion 58 026032 (2018).
[2] R. A. Tinguely et al, Plasma Phys. Control. Fusion 60 124001 (2018).
[3] R. A. Tinguely et al, Nucl. Fusion 58 076019 (2018).
[4] R. A. Tinguely et al, Nucl. Fusion 59 096029 (2019).
[5] N. C. Hawkes et al, Rev. Sci. Instrum. 70 894 (1999).Speaker: Mathias Hoppe (Chalmers University of Technology) -
Measurements of pellet-triggered edge localized mode (ELM) heat fluxes are presented here from experiments in ITER-relevant low collisionality pedestals (normalized pedestal collisionality $\nu^*_{ped}$ < 1) on DIII-D. These measurements demonstrate a reduction of peak ELM energy fluence at the inner strike point as compared to natural ELMs by as much as $\sim$50%. The inner strike point is typically the limiting location for material heat flux limits due to ELMs, so reducing the ELM heat flux there enables a higher allowable pedestal height before an ELM would need to be triggered. This in turn reduces the pellet pacing frequency required to protect materials and increasing the time-integrated pedestal pressure.
Pellet ELM pacing will be investigated in ITER in its pre-fusion power operation phases while access to RMP ELM suppression and/or other ELM-free operational regimes are assessed. Although previous experiments have demonstrated a reduction in peak ELM heat flux using pellet pacing for higher collisionality pedestals {1}, the capabilities and limits of this actuator for mitigating ELM heat flux and core impurity accumulation in low collisionality conditions are poorly understood. At low collisionality, the edge bootstrap current is increased for a given pedestal pressure, leading to pedestal growth being limited by the current density (kink/peeling instability) rather than the pressure gradient (ballooning instability). Exploring pellet ELM triggering in low-collisionality regimes in DIII-D tests the applicability of this ELM control actuator in this new regime and enables model validation so that we may more confidently extrapolate to ITER and other future devices. This pellet ELM triggering can then be extended to ELM pacing at a rate several times that of the natural ELM frequency, but the foundational experiments presented here use only infrequent pellets.
By shifting the location of the outer strike point away from the pumping duct during pellet ELM triggering while simultaneously attempting to access low collisionality conditions, heat flux measurements were obtained during pellet ELM triggering experiments in DIII-D. Figure 1 gives the ELM energy fluence profiles (the heat flux profiles integrated over 2 ms in this case) from one of these discharges for two natural ELMs (in blue and green), an ELM triggered by a large pellet (in red), and the inter-ELM heat flux integrated for the equivalent time frame, for reference (in dashed black). All four profiles are within the same $\sim$150 ms of DIII-D discharge 178555, with normalized pedestal collisionality < 0.7. The x-axis in Figure 1 follows the lower divertor target from the inner wall to the floor to the far scrape-off layer, with the inner strike point (ISP) and outer strike point (OSP) indicated around 0.6 m and 0.9 m, respectively. An area near the OSP is obstructed from view of the infrared camera used here, resulting in the shadowed region of Figure 1 between 0.98 m and 1.1 m where no heat flux measurements are available.
As seen in Figure 1, for large pellet–triggered ELMs, the relative amplitude of the integrated heat flux at the inner strike point is reduced by about a factor of 2, while a smaller increase in heat flux is observed in the far SOL. Because the ISP is the location of largest energy fluence for low collisionality ELMs in this dataset, the reduction of heat flux at this location would be beneficial for minimizing material damage for a given ELM size. The far scrape-off layer (on the shelf, where distance > 1.1 m) is also shown to have an additional heat flux lobe when large pellets trigger ELMs. This feature appears to be a three-dimensional structure that propagates in time as the ELM evolves, and also appears in the D-alpha fast camera images.
One goal of a pellet ELM triggering or pacing scheme would be to keep the peak ELM energy fluence values below the melting limit of the divertor material. The peak ELM energy fluences are plotted in Figure 2 for the ISP (star symbols) and OSP (circle symbols) for a series of natural and triggered ELMs from DIII-D discharge 178555. For natural ELMs, plotted as small unfilled symbols, the ELM energy fluence is almost always larger at the ISP than the OSP. This finding that the limiting value of ELM energy fluence occurs at the ISP is consistent with previous studies for naturally ELMing discharges with low collisionality pedestals in DIII-D {2}. $\nu^*_{ped}$ is $\sim$0.5 before pellets are injected and increases throughout the discharge, to $\sim$1.0 at 5.0 s and $\sim$1.5 at 5.2 s.
When large pellets (1.8 mm) are injected from the low field side of the device and trigger ELMs, as shown with large red symbols in Figure 2, the peak ELM fluence at the ISP is observed to be greatly mitigated. For ELMs triggered by smaller (1.3 mm) pellets plotted in blue, however, the heat flux distribution is comparatively unmodified with respect to the natural ELMs. Assuming the allowable ELM size determines the allowable pedestal height, these results suggest that reducing the peak ELM energy fluence at the ISP would increase the allowable pedestal height before an ELM triggering event is necessary. This could enable lower frequency pellet pacing and higher peak and time-averaged pedestal heights.
This work was supported in part by the US Department of Energy under award numbers DE-FC02-04ER54698, DE-AC05-00OR22725, DE-AC02-09CH11466, DE-FG02-07ER54917 and DE-AC52-07NA27344.
{1} L.R. Baylor et al., Phys. Rev. Lett. 110, (2013) 245001; Phys. Plasmas 20, (2013) 082513.
{2} M. Knolker et al., Nucl. Fusion 58, (2018) 096023.Speaker: Robert Wilcox (Oak Ridge National Laboratory) -
Reliable whole device modeling (WDM) of present and future burning plasmas critically depends on correct treatment of the auxiliary heating by energetic particles (EP) introduced into the plasma externally or by the fusion alpha particles. These energetic, or super-thermal, ions can effectively resonate with Alfvénic plasma oscillations, lead to EP losses and modify the profiles of the current drive. For burning plasma experiments, such as in ITER, it is important to not only predict the plasma operational regimes when those instabilities occur, but to learn how to operate in the regimes when the effects of the Alfvénic mode (AM) instabilities are benign. This is possible when all the essence of EP distribution function are accurately evaluated, which is envisioned in recently developed resonance broadened quasi-linear (RBQ) code [1].
We report on the RBQ code development with two dimensional (2D) computational capabilities to compute the distribution function evolution in time. This is an essential element of WDM simulations for self-consistent evaluation of EP distribution function. The RBQ code employs both the ideal MHD computed mode structures and their kinetic growth and damping rates. The velocity space diffusion coefficients in the constant of motion space are the result of RBQ simulations. They are transferred to the whole-device modeling code TRANSP.
RBQ utilizes a self-consistent quasi-linear (QL) theory in the presence of multiple unstable Alfvénic modes in the near-threshold regimes when the resonances between the EP and AM are broadened by a specific value prescribed by the analytic theory [2]. The formulation includes the discrete-resonance collisional functions to broaden the resonance regions for both Krook and Fokker-Planck scattering collisions. These functions are shown in Fig.1(b), which replace a simple resonance delta function that appears in the diffusion coefficient for the case of no broadening. For considered DIIID plasma the resonances are broadened by of the plasma minor radius independent on the mode amplitude. These functions remove a major arbitrariness with respect to previous resonance broadening approaches, which consisted of tuning broadening parameters to match the expected saturation levels [2]. The resonance functions are essential components of the QL model of the code to describe the dynamics of EP distribution evolution, where particle diffusion occurs in both canonical toroidal momentum and particle energy. We verify with the help of the guiding center code ORBIT that the resonances are indeed broadened by the amount determined by the pitch angle scattering frequency. The broadening enhanced by scattering is shown in left Fig.1(a), which shows that the broadened region is insensitive to the resonance island and fully determined by the scattering frequency.
The RBQ simulations find constant of motion diffusion rates for subsequent computation by TRANSP’s NUBEAM package using the probabi-lity density functions. The RBQ code has the capability to efficiently evolve the mode amplitudes simultaneously (in regimes of both overlapping and isolated resonances) while self-consistently relaxing the fast ion distribution function in the presence of collisions in both interpretive and predictive regimes. RBQ employs realistic eigenstructures, damping rates and wave-particle interaction matrices pre-computed by the NOVA-K code. The code in its 1D version was applied to DIII-D critical gradient experiments in predictive mode and reproduced the observed hollow fast ion radial profiles within the experimental uncertainties [1]. Examples of EP distribution function with and without canonical momentum diffusion computed by RBQ1D are shown in Fig.2.
The EP diffusion dependence across the resonance region allows the RBQ to compute the amplitude evolution which can exhibit oscillating, intermittent and saturated state behavior depending on the pitch angle scattering rate. This can be clearly observed in experiments when only few Alfvénic modes are excited such as in DIIID [3] and TFTR [4].
Our work reports on time-efficient, realistic simulations of fast ion redistribution and losses in tokamaks in presence of Alfvénic instabilities and their implementation in WDM using TRANSP. The RBQ code is upgraded to include 2D constant of motion diffusion with the goal to include simulate and design the burning plasma experiments, such as ITER. An important element of such simulations is to understand how tokamak devices can operate with unstable but benign Alfvénic instabilities. Another goal is to predict the current drive by beam ions and fusion products.
References
[1] N.N. Gorelenkov, V. N. Duarte, C.S. Collins et al., Phys. Plasmas 26 (2019) 072507.
[2] V.N. Duarte, N.N. Gorelenkov, R.B. White, H.L Berk, Phys. Plasmas 26 (2019) 120701.
[3] M.V. Van Zeeland et al., “Modification of Alfvén Eigenmode Drive and Nonlinear Saturation Through Variation of Beam Modulation in DIII-D”, IAEA TCM 2019, Shizuoka, Japan.
[4] K.L. Wong, R. Majeski, M. Petrov et al., Phys. Plasmas 4 (1997) 393.
Speaker: Nikolai Gorelenkov (PPPL, Princeton University) -
The Kinetic Orbit Runaway electrons Code (KORC) {1} has been extended to model post-disruption runaway electron (RE) dissipation by impurity injection incorporating state-of-the-art collisional models for partially ionized impurities {2} and models of thermal electron and impurity spatiotemporal dynamics. We fit these models to data from the DIII-D and JET tokamaks, exploring the role of spatial effects as compared to previous studies (e.g. Ref. {3}). This work presents model validation through detailed comparisons of shattered pellet injection (SPI) and massive gas injection (MGI) dissipation of REs and a detailed study of different collisional dissipation models. We find that the evolution of the spatiotemporal electron and impurity density profiles due to impurity injection and interaction with REs plays a significant role in the dissipation process and is an outstanding and critical modeling need. Additionally, we find that the dissipation timescale is fundamentally sensitive to the initial RE energy distribution. Beyond validation, this work allows for comparing and optimizing different impurity-based mitigation scenarios and their scaling towards ITER.
KORC has been extended to serve as a general framework for simulating RE physics, including validation and verification of the theoretical models needed to understand RE dissipation by impurity injection. To make calculations of RE dissipation numerically feasible, the RE guiding center orbit equations of motion have been implemented and are evaluated from experimental fields and plasma profile information via interpolation. The magnetic field configuration is held fixed and taken from experimental reconstructions, with EFIT used for JET and JFIT for DIII-D, as seen in overlaid (black) contours in the rightmost panels of Fig. 1 for JET pulse 95128 and DIII-D shot 164409. The toroidal electric field uses a 1/R spatial model fit to multiple experimental loop voltages. Initial RE distributions are initialized using a flexible sampling algorithm to initialize desired multidimensional distributions. The electron and impurity density profiles are taken from ad hoc models with a variable impurity charge state ratio that are fitted to experimental data using a synthetic line-integrated electron density (LID) diagnostic at multiple locations. The solid traces in the left panels of Fig. 1 correspond to experimental data from JET pulse 95128 with Ar SPI (top) and DIII-D shot 164409 with Ne MGI (bottom), and dashed traces correspond to synthetic LID signals for the profile shown in the right panels of Fig. 1 that are used in KORC calculations. A linearized, relativistic, Coulomb collision operator has been implemented via a Monte Carlo approach, where the effects due to bound electrons of partially ionized impurities are implemented through various models {2}. Additionally, synchrotron and bremsstrahlung radiation are included through additions to the guiding center orbit equations of motion.
The flexibility of KORC allows for simulations employing a hierarchy of models. Fig. 2 shows experimental plasma current from DIII-D shot 164409 with Ne MGI (top panel) and corresponding modeling results from KORC (bottom panel). The canonical case (black trace) includes bound electron physics, spatiotemporal electron and impurity density profiles fitted according to Fig. 1, a mix of singly and doubly ionized Ne, and an initial 10 MeV monoenergetic RE distribution. The critical importance of including bound electron collisional effects is evident from the different timescales of the without bound electrons (dark blue trace) and canonical cases. This physics is not only important for capturing accurate dissipation rates, but also the difference in dissipation rates for different injected impurities (not shown) and impurity charge state ratios (green trace). The KORC simulations also show a significant importance on the inclusion of a spatiotemporal density profile, as indicated by the difference between the constant and uniform density (red trace) and canonical cases. The initial energy distribution also plays an important role in the simulated dissipation timescale, shown by the difference between the canonical case and a case using an energy distribution inferred from a separate DIII-D study {4} (cyan trace). Interestingly, in the present work we find that for RE dissipation, as opposed to RE generation, the toroidal electric field plays a smaller role even when scaled (not shown), albeit a role that increases as the dissipation timescale increases. We additionally find that synchrotron and bremsstrahlung radiation have a comparatively minor effect (not shown). These results indicate the importance of including bound electron physics and spatiotemporal density evolution for assessing the efficacy and optimization of RE dissipation strategies for ITER.
{1} Carbajal et al., Phys. Plasmas 24, 042512 (2017)
{2} Hesslow et al., Phys. Rev. Lett. 118, 255001 (2017)
{3} Martin-Solis et al., Nucl. Fusion 57, 066025 (2017)
{4} Hollmann et al., Phys. Plasmas 22, 056108 (2015)*This work was supported by the US DOE under contracts DE-AC05-00OR22725 and DE-FC02-04ER54698 and by the ITER Organization (TA C18TD38FU) and carried out within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission or the ITER Organization. This research used resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02-05CH11231.
**See the author list of E. Joffrin et al. accepted for publication in Nuclear Fusion Special issue 2019,
https://doi.org/10.1088/1741-4326/ab2276Speaker: Matthew T. Beidler (Fusion Energy Divison; Oak Ridge National Laboratory) -
Analysis of “super H-mode” experiments on DIII-D shows that high rotation, not high pedestal, plays the essential role in achieving very high energy confinement quality ($H_{98y2}$ > 1.5) $^1$. While the stored energy increases as expected with higher pedestal, the energy confinement quality mainly depends on the toroidal rotation (figure 1). At moderate rotation, similar to levels expected in ITER, very high pedestal height only yields marginal confinement quality improvement relative to standard H-mode. The dependence of the thermal energy confinement quality on the toroidal rotation is similar to that observed in the DIII-D advanced inductive scenario (also called hybrid scenario), which is typically characterized by high normalized fusion performance at high rotation, but suffers strong confinement degradation at low rotation $^2$. Linear gyrofluid and nonlinear gyrokinetic transport modeling confirm that the effect of E×B turbulence stabilization is far larger than other mechanisms, including EM stabilization and so-called hot-ion stabilization ($T_i/T_e$), as well as the fast ions effect.
Good confinement is critical for an economical fusion reactor design. Understanding the mechanisms leading to good confinement quality in a particular experimental scenario is essential to the ability to extrapolate that confinement quality to a future reactor. Super H-mode plasmas exploit strong plasma shaping to achieve high edge pedestal pressure at high density $^3$, and often exhibit excellent confinement quality exceeding standard H-mode, i.e. $H_{98y2}$ >> 1.0. It was initially believed that the very high pedestal was responsible for the very high energy confinement quality $^4$. However, careful analysis of the experimental data and transport modeling shows that the reason for the very high H-factor values is the rapid DIII-D toroidal rotation, which does not extrapolate to a reactor.
Figure 1 shows trajectories of several super H-mode discharges illustrating how the confinement quality is linearly correlated with the toroidal rotation in the core, while the pedestal pressure often increases as both the rotation and the confinement quality decrease. Very high confinement is reached early on in the H-mode phase of these discharges, when the pedestal is still low, but after the toroidal rotation has built-up to very high levels in the core. As the discharge evolves, the confinement quality is linearly correlated with the core toroidal rotation, which varies according to different levels of injected neutral beam torque per particle. Note that, even beyond the set of discharges analyzed in this paper, there is no experimental observation of very high confinement achieved at low rotation in any DIII-D super H-mode experiments to date. Also note that, as shown in figure 1, for given rotation the same $H_{98y2}$ is obtained with significantly different pedestal heights (the electron pedestal pressure is used as proxy of total pedestal pressure, since the ratio of electron to ion pedestal pressure stays fairly constant throughout the H-mode phase). Consistent with what one might expect, the stored energy is indeed higher with higher pedestal (not shown in figure), but other parameters (like plasma current, auxiliary power, etc.) must change to yield a different pedestal height, resulting in a constant $H_{98y2}$ for given rotation.
Figure 2 shows that linear gyrofluid modeling achieves a good match of the experimental observations by including the toroidal rotation and related E×B shear. The temperature reduction that results from removing the E×B effect in the simulation is calculated to decrease $H_{98y2}$ from 1.7 to 1.3. Similar analysis was repeated for several time slices in the same discharge, and all the calculated $H_{98y2}$ with and without ExB effect are reported in Fig.1 (blue circles), showing a similar prediction of the energy confinement quality without rotation, for time slices with different experimental confinement values. Note that all cases without ExB shear are slightly higher than the empirical “low rotation” $H_{98y2}$ (the dashed line shown in Fig.1) and this over-prediction may be due to having a fixed boundary condition at the pedestal in the modeling. Figure 3 shows an example of nonlinear gyrokinetic modeling exploring both the effect of the ExB shear and of the $T_i/T_e$ ratio. A large change in the normalized ion energy flux is predicted when the E×B effect in the simulation is removed. Little effect is predicted when the $T_i/T_e$ ratio is increased from the experimental value during the stationary confinement phase, ~1.16, to the peak value achieved in the discharge, ~1.32.
In conclusion, super H-mode plasmas can achieve very high energy confinement quality when accessed following a large neutral beam torque injection that transiently drives very rapid plasma toroidal rotation during and shortly after the L- to H-mode transition. As the density increases during the H-mode and super H-mode phases, the rotation decreases. With rotation closer to reactor-relevant levels, most of the confinement improvement over standard H-mode is lost, despite sustained very high pedestal pressure. Further studies on how to improve the energy confinement at low rotation will be key to the development of super H-mode scenario for a high performance fusion reactor.This material is based upon work supported by the US DOE under contracts DE-FC02-04ER54698, DE-AC02-09CH11466, DE-SC0010685, DE-SC0018287, DE-FG02-94ER54235
$^1$ Ding S. et al, 2020 Nucl. Fusion 60 034001
$^2$ Solomon W.M. et al, 2013 Nucl. Fusion 53 093033
$^3$ Solomon W.M. et al, 2014 Phys. Rev. Lett. 113 135001
$^4$ Snyder P.B. et al, 2015 Nucl. Fusion 55 083026Speaker: Andrea M. Garofalo (General Atomics) -
The high-power helicon antenna has been extensively tested at low power with excellent results, and the full installation in the DIII-D vessel was successfully completed in February 2020, as shown in Fig. 1. High-power striplines feed the antenna from either end (not visible in Fig.1). Commissioning of the 1.2 MW 476 MHz klystron source and associated high voltage power supply and switching network is scheduled to be complete by the end of Spring 2020, enabling the first high-power helicon current-drive experiments in the Summer of 2020. Helicon current drive, or fast wave current drive in the lower hybrid range of frequencies, has long been predicted to be a promising current drive tool for reactor grade plasmas. The system at DIII-D will be the first test of high-power helicon current drive in reactor-relevant plasmas, where full single-pass absorption is expected [a]. Substantial off-axis driven current is predicted for DIII-D (~60 kA/MW).
The antenna is a traveling wave antenna of the comb-line type with 30 modules, fed from either side to allow both co and counter current drive. Power is transferred down the antenna through mutual coupling from one module to the next with low losses. All 30 antenna modules were tuned individually to the same resonant frequency to within ±0.5 MHz, and show low losses with a quality factor in excess of 1000. The linear electromagnetic characteristics of the unloaded module array has been extensively tested on the bench and on the DIII-D mock-up wall at instrumentation power levels, with sub-arrays of 10, 15, 20 and 30 modules on curved backplates. The six backplates are coupled capacitively to minimize disruption forces while providing a good RF ground plane. The placement of the modules on the backplates in a 2-module poloidal stagger pattern, which increases the opacity of the Faraday shield to field line penetration, resulted in a periodicity in the mutual coupling between modules that can drive large reflections (~20% power reflected). Adjustment of module placement compensates for this, resulting in a low 1-2% power reflection coefficient and 98.7% power transfer from module to module in air, as shown in Fig. 2 for the 30-module installation in the DIII-D vessel. Also shown in Fig. 2 is data obtained with a pick-up coil in front of the straps just outside the Faraday screen bars. The phase progression and probe amplitude are steady along the length of the array, as expected for a traveling wave antenna.A 1.2 MW klystron, obtained from SLAC, and the associated high-voltage power pulsing network is being installed and commissioned. The klystron with circulator and dummy load feeds power into a 4-port switching network. The 4-port switch allows feeding of a stripline on either end of the antenna, with the reflected power and the remaining transmitted power each dumped into the dummy load. To satisfy the vessel constraints, the stripline geometry was 3D printed out of Inconel and stainless steel, and copper-coated. The striplines act as a 180-degree phase splitter combined with two 3dB splitters. The four equal-power out-of-phase signals are fed to the coupling straps on the end-modules of the antenna. The antenna and striplines are instrumented with pick-up coils, arc detectors and thermocouples.
Several new and upgraded diagnostics are planned for the helicon program. An edge-density diagnostic will provide density profiles at the antenna. A laser-based technique known as Doppler-free saturation spectroscopy (DFSS) will be implemented to measure the helicon wave electric field vector over a 2D region of space in the edge plasma near the antenna. The wave electric field vector is obtained by fitting the Schrodinger equation to Doppler-free Dβ spectra. This technique results in an accuracy of <5 V/cm, two orders of magnitude greater than that of previous work. For far-field wave measurements, the existing Phase Contrast Imaging (PCI) diagnostic is being upgraded with a heterodyne detection system able to detect fluctuations at 476 MHz. Using values for the wave electric field estimated by full-wave modeling work, it is expected that the PCI will be able to measure the spatial structure of perturbed density with favorable signal-to-noise ratio.Helicon wave current drive will soon be investigated in high density, high electron beta plasmas where full single-pass absorption is expected – conditions similar to that in steady-state AT burning plasma devices.
This work was supported in part by the US Department of Energy under DE-FC0204ER54698, DE-AC05-00OR22725, DE-FG02-94ER54235, DE-SC0013911 and DE-AC02-09CH11466.
[a] R. I. Pinsker et al, Nuclear Fusion 58, 106007 (2018)Speaker: Bart Van Compernolle (General Atomics) -
Understanding impurity transport within fusion research plasmas is of critical importance as progress is made toward burning plasmas. High core impurity concentration can have a deleterious effect on plasma performance: fuel dilution (for fully stripped low Z impurities), triggering of MHD instabilities, or increased core radiated power (for high Z impurities), thus hindering the achievement of stable long-pulse, high energy confinement discharges. A full understanding of impurity transport conditions is especially crucial when observations from multiple devices display significantly different results to what is presumably a similar set of conditions. The effects of evaporated lithium on the NSTX lower divertor and the effects of lithium powder injection were primarily evidenced in the edge/pedestal region where changes in the electron pressure profiles resulted in ELM free operation1. This state however did not lead to an enhanced lithium presence in the core with concentrations rarely rising above 0.1%. Lithium aerosol injections into DIII-D2 discharges utilized 90% less material, but were also successful in generating 300 ms ELM free periods and enhanced pedestal pressures. However these discharges saw substantial penetration of the lithium into the core plasmas, with measured concentrations up to 15-20%. New experiments combining the introduction of lithium and boron powders counterposed with carbon and lithium granule injections have shown the ability to manipulate the core carbon concentration to levels both higher and lower than those observed in reference discharges allowing a comparison with neoclassical impurity transport theory.
According to neoclassical theory, carbon will act as the dominant impurity and screen out lower-z impurities above a threshold concentration. To determine if a carbon threshold effect can mitigate core lithium penetration, powders and granules were introduced into DIII-D ITER baseline discharges with IP = 1.3 MA, BT = -1.7 T, ne = 8x1019 m-3, PNBI = 9.5 MW and N = 2. Powders are gravitationally dropped into the upper edge of a lower single null H-mode discharge by means of an impurity powder dropper (IPD)3 and granules were horizontally injected into the discharge utilizing an impurity granule injector (IGI)[4]. A reference discharge shown in Figure 1 displays a rapid increase in core carbon concentration during the transition to H-mode which stabilizes to a near constant signal intensity that is maintained during the discharge.
To modulate core carbon concentration the two actuators were utilized separately. Core carbon levels were increased by injection of 400 micron vitreous C granules 2.5 – 4.5 s at mass injection rates of 6 and 4 mg/s, respectively. Measurements of core carbon impurities saw the signal levels start to rise at approximately t = 3 s with the higher injection rate reaching a saturation point after 300 ms. Saturation occurs at an intensity twice that of the baseline level as seen in panel 1 of Figure 2. These results are contrasted with those seen from Li powder injection. For these cases, the introduction of lithium powder caused a prompt reduction of the carbon signature by a factor of 4x which was maintained for the duration of the injection event before returning again to the baseline level. The introduction of Li powder also had a marked effect upon the ELM cycle, with the higher level of injection resulting in extended periods of ELM free behavior as seen in figure 3. These results are consistent with previous reports of Li powder injection2.
To determine if the lithium buildup within the DIII-D core plasma during Li dropper operation can be flushed by carbon microgranule injection a combination of the two actuators was used. As is shown in panel 2 of figure 2, Li powder is utilized to drop the core carbon concentration prior to the introduction of carbon granules. Carbon levels are seen to increase during the period of C injection from 3.5-4.5 s. However, once the granule injection is stopped the C signal returns to the Li suppressed baseline level rather than remaining at the standard carbon level, thus demonstrating a substantial pumping effect by the lithium leading to the observation that this carbon level does not provide an impurity screening effect. In contrast, panel 3 shows the results when the two actuators are temporally swapped. The long carbon granule pulse is active from 2.5-6 s while the short lithium powder pulse is active from 2.5-5 s. At these elevated carbon levels the Li aerosol does not seem to be able to have a substantial effect on the core carbon intensity, thus there is the possibility that a carbon transport barrier has been established. While increasing the Li powder rate may overcome this barrier, these discharges are already at a substantial impurity fraction and it was observed in subsequent injections with boron powder that larger impurity injections lead to a premature discharge termination event.
As can be seen from these results, there is a complex interplay between the impurities present within the plasma core dependent upon both the species of impurities introduced as well as the order in which they are deposited into the discharge. These measurements provide benchmarking data for neoclassical transport codes such as NEO and XGC to help determine if the corresponding variations in impurity transport can be explained by simulation and will inform favorable transport conditions in future tokamaks.This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-FC02-04ER54698 and DE-AC02-09CH11466
1 F. Scotti et al., Nucl. Fusion 53 (2013) 083001
2 T. H. Osborne et al., Nucl Fusion 55 (2015) 063018
3 A. Nagy et al., Rev. Sci. Instr. 89 (2018) 10K121
[4] M. Vorenkamp et al., Fusion Sci. & Tech. 72:3 (2017) 488Disclaimer: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
Speaker: Robert Lunsford (Princeton Plasma Physics Laboratory) -
We present an overview of theory and simulation of low-frequency drift Alfvén waves (DAW) in toroidal fusion plasmas based on the framework of the general fishbone like dispersion relation (GFLDR) [1, 2, 3]. In addition to recovering various limits of the kinetic MHD energy principle, this approach can also be applied to general e.m. fluctuations characterized by a broad range of spatial and temporal scales consistent with gyrokinetic descriptions of both core and supra-thermal plasma components. Therefore, the GFLDR provides a unified description of DAW excited by energetic particles (EPs) as either Alfvén eigenmodes (AEs) or energetic particle modes (EPMs). Furthermore, the GFLDR is given in the following generally variational integral functional form
$\begin{equation} 2 \pi^2 \int_0^a dr \left[ \frac{k_\vartheta^2 c^2 (d\psi/dr)}{{\cal J} B_0^2} \right]_{\vartheta = 0} \left| A(r)\right|^2 \left[ \delta \bar W_f (r) + \delta \bar W_k (r) - i \lambda (r) \right] = 0 \; . \;\;\;\;\; \end{equation}$
For radially localized fluctuations, the GFLDR can then be written as
$\begin{equation} i \lambda (r) = \delta \bar W_f (r) + \delta \bar W_k (r) \; , \;\;\;\;\; \end{equation}$
where $\Lambda$ represents a generalized inertia, while $\delta \bar W_f$ and $\delta \bar W_k$ describe the potential energy of the fluctuations accounting, respectively, for the fluid and kinetic plasma response [1, 2, 3]. From the structure of Eqs. (1) and (2), the present approach allows extracting spatial and temporal scales of the considered fluctuation spectrum as well as the underlying physics. For example, radial singular structures of the shear Alfvén wave (SAW) continuous spectrum are shown in FIG. 1 using the analytic as well as the numerical solution for the generalized inertia, Λ, in the MHD fluid limit [4, 5]. Meanwhile, analytic results on mode frequency, damping and finite orbit width effects can be verified numerically with very good agreement for both SAW continuum accumulation points [6, 7] as well as for geodesic acoustic mode oscillations [8]. Thus, the GFLDR as a unified theoretical framework can help understanding experimental observations as well as numerical simulation and analytic results with different levels of approximation.
The gyrokinetic analysis of low-frequency DAW is necessary for a proper description of short wavelengths and/or accounting for finite parallel electric field and wave damping as well as for EP excitations [3, 6, 7]. Here, we derive the corresponding GFLDR theoretical framework in general tokamak geometry as a joint effort of ongoing research projects [9, 10]. As first application of the general theory, we extend to gyrokinetic analysis the application to the Alfvén - acoustic fluctuation spectrum of Ref. [4] in the MHD fluid limit. This analysis, presented in detail in a dedicated contribution [11], confirms the general prediction that fluctuations with acoustic polarization are unfavored because of the stronger Landau damping and the weaker response to EPs [4], in agreement with recent observations in DIII-D [12].
As further application, motivated again by recent experimental observations [12] consistent with earlier theoretical predictions [4], we apply the GFLDR theoretical framework to simple (s, α)-equilibrium tokamak geometry. In particular, we use the analytic expression of the generalized inertia, Λ, derived earlier for well circulating/deeply trapped thermal ions [13, 14]. Our results, presented in detail in a separate work [15], show, again, that the acoustic branch is strongly suppressed by Landau damping and that EP excitation is generally weak.
Numerical results by the Drift Alfvén Energetic Particle Stability (DAEPS) code will also be shown and discussed [10]. DAEPS belongs to a long-term project launched to better relate experimental observations and theoretical/physical understandings of DAW (AEs/EPMs) interacting with thermal plasma as well as EPs (including alpha particles). DAEPS is a gyrokinetic eigenvalue solver, which can properly represent kinetic compression and wave-particle resonances of core and energetic plasma components. Compared to initial value codes, the DAEPS approach is a valuable method to analyze the whole stable/unstable spectrum besides the most unstable mode. As demonstration, the beta-induced AE (BAE) is used to illustrate the relevant aspects of the code.Acknowledgments
This work was supported by EUROfusion Consortium and ITER-CN grants.[1.] Zonca F and Chen L 2014 Phys. Plasmas 21 072120
[2] Zonca F and Chen L 2014 Phys. Plasmas 21 072121
[3] Chen L and Zonca F 2016 Rev. Mod. Phys. 88 015008
[4] Chen L and Zonca F 2017 Phys. Plasmas 24 072511
[5] Falessi M V, Carlevaro N, Fusco V, Vlad G and Zonca F 2019 Phys. Plasmas 26 082502
[6] Bierwage A and Lauber Ph 2017 Nucl. Fusion 57 116063
[7] Lauber Ph and Lu Z 2018 J. Phys. Conf. Ser. 1125 012015
[8] Biancalani A, Bottino A, Ehrlacher C et al. 2017 Phys. Plasmas 24 062512
[9] MET Enabling Research Project, https://www.afs.enea.it/zonca/METproject/
[10] Li Y et al. 2020 “Gyrokinetic analysis and simulation of Alfvén Eigenmodes” To be submitted
[11] Chen L et al. 2020 “On energetic particle excitations of low-frequency Alfvén eigenmodes in toroidal plasmas: gyrokinetic theory” This Conference
[12] Heidbrink W W 2019 Private communication
[13] Chavdarovski I and Zonca F 2009 Plasma Phys. Contr. Fusion 51 115001
[14] Chavdarovski I and Zonca F 2014 Phys. Plasmas 21 052506
[15] Chavdarovski I et al. 2020 “Effects of core plasma on the low frequency Alfvén and acoustic eigenmodes” This Conference.Speaker: Dr Fulvio Zonca (ENEA C.R. Frascati) -
Energetic particle (EP) physics in fusion research are excepted to play crucial roles in the next generation of tokamak burning plasma experiments, e.g., ITER. Energetic fusion alpha particle heating of fuel ions through collisional and collisionless channels is crucial for achieving self-sustained burning. On the other hand, free energy associated with EPs pressure gradient, can drive collective instabilities, e.g., shear Alfvén waves, via wave-particle resonance, and induce EP anomalous transport due to the breaking of toroidal symmetry. One of such Alfvén wave instabilities, which has received considerable interest, is the beta-induced Alfvén eigenmode (BAE)
[1].
$\quad$BAEs have been recently observed in HL-2A during the Ohmic and ECRF heating[2]. Moreover, gyrokinetic simulations using HL-2A parameters[3]show that the precessional resonance of trapped energetic electrons (EEs) can drive BAE (e-BAE) instabilities and induce the typically observed croissant-like up-down asymmetric mode structures. To our knowledge, detailed theoretical understanding of the e-BAE physics, including the excitation mechanism, the global stability and corresponding radial mode structures as well as the e-BAE nonlinear saturation, are still lacking. On the other hand, the EE finite orbit width normalized to the minor radius of the present-day tokamaks, could be comparable to that of the alpha particles characterized by small dimensionless orbits in reactors, e.g., ITER. Thus, the in-depth understanding of the e-BAE physics based on first-principle-based theory[4]is needed, and this constitutes the main motivation of the present work.
$\quad$In this work, employing the WKB-ballooning mode representation along with the generalized fishbone like dispersion relation[5], the two-dimensional (2D) global dispersion relation of the high-n e-BAEs excited by precessional resonance of magnetically trapped EE is derived in large aspect-ratio, low-$\beta$ and low magnetic shear tokamaks with shifted circular flux surfaces. It is found that[6]the contribution of the trapped EEs to the global e-BAE dispersion relation is limited to the ideal MHD structure of the BAE due to the EE bounce averaging dynamic being governed by normal curvature. Moreover, our numerical results show that, (i) for the local properties of e-BAE: the mode can be destabilized by EEs using the typical equilibrium parameters in HL-2A, and the mode frequency is consistent with the experimental observation. Varying the background plasma parameters can lead to transitions between e-BAEs and energetic particle modes. Moreover, the dependence of the e-BAE frequencies and growth rates on energetic electron parameters shows that the growth rates monotonically increase(decrease) with the energetic electron density(the normalized energetic electron density gradient scale length), and the frequencies are not much affected. The frequency and growth rate are sensitive to the energetic electron temperature, and there exists a maximum growth rate. (ii) For the global properties of e-BAE[6], the mode is radially localized in the potential well generated by the pressure gradient of EEs, and the most unstable mode gradually changes from the highly excited state to the lowest radial bound state with the increase of the EE density. The anti-Hermitian contributions due to wave-energetic particle resonance give rise to the twisting radial mode structures, shown in Fig. 1, and the direction of deformation is opposite to that of the mode structure caused by energetic ions, which is consistent with the simulation results of GTC[3]. Moreover, the symmetry breaking of the e-BAE mode structure with respect to parallel wave-number, as shown in Fig. 1, has a potential impact on toroidal momentum transport, and is analyzed. Finally, generation of zonal flow, as a potential channel for the e-BAE nonlinear saturation, is also investigated[7].Acknowledgment
$\quad$This work is supported in part by National Natural Science Foundation of China under Grant No. 11705050, the ITER-CN under Grant Nos. 2018YFE0304103, 2017YFE0301900 and National Natural Science Foundation of China under Grant Nos. 11875024 and 11875021.References
[1]M.S. Chu et al., Phys. Fluid B 1992, 4, 3713.
[2]W. Chen et al., Phys. Rev. Lett. 2010, 105, 185004.
[3]Y. Chen et al., Phys. Plasmas 2019, 26,102507.
[4]L. Chen et al., Journal of Geophysical Research 1991, 96, 1503.
[5]F. Zonca et al., Phys. Plasmas. 2014, 21, 072121.
[6]R.R. Ma, et al., Nucl. Fusion 2020, https://doi.org/10.1088/1741-4326/ab7f15. R.R. Ma, et al., Global theory of beta-induced Alfvén eigenmodes by energetic electrons, in preparation.
[7]Z.Y. Qiu, et al., Zero frequency zonal flow excitation by energetic electron driven beta-induced Alfvén eigenmode, Submitted to Plasma Physics and Controlled Fusion, 2020.Speaker: Dr Ruirui Ma (Southwestern Institute of Physics) -
Energetic particles (EP) are ubiquitous in fusion plasmas and need to be well-confined in order to transfer their energy to thermal particles and thus achieve self-sustained fusion reactions. However, a fusion plasma is a complex system where micro- and macro-instabilities develop. These instabilities can dramatically reduce the EP confinement and therefore limit the performance of future fusion devices such as ITER. This is the reason why understanding and controlling EP transport in the presence of different instabilities is of prime importance on the route towards steady-state scenarios. Here we focus on the analysis, understanding and quantification of the EP transport and losses induced by electro-magnetic modes. To simplify the analysis, we study the impact of single-helicity modes on the EP transport and losses. Single-helicity electro-magnetic modes are characterized by one poloidal (m) and one toroidal (n) mode numbers. Although such modes have been usually believed not to result in any chaotic transport, significant losses of EP have been observed, both experimentally and numerically. The purpose of this work is to shed light on these observations and provide a reduce model to understand and predict the transport and losses of EP. For this purpose, we use an analytical gyro-kinetic model describing the impact of the EP on electro-magnetic modes. This model is combined with a recently developed 5D Guiding-Centre Tracking (GCT) code to obtain a reduced model for the self-consistent interaction between EP and electro-magnetic modes. The transport and losses are quantified and compared to analytical predictions based on quasi-linear theory.
The emphasis is put on two paradigmatic single-helicity modes: (a) the energetic geodesic acoustic modes (EGAMs) [1-4], excited by EP and exhibiting mainly an electrostatic component, and (b) the tearing modes (TMs), excited by the radial gradient of the parallel current and characterized by a magnetic component.
Because EGAMs are not only single-helicity but also axi-symmetric modes, they have been assumed in the past to play little role in transport. However, experiments in DIII-D provided puzzling evidence that particles can be de-confined in the presence of EGAMs [1]. Using GCT and a potential obtained from full-f gyro-kinetic simulations we elucidate the underlying mechanism and show that even if the EGAM is axisymmetric and non-turbulent, a chaotic channel from the inner region to the edge of the tokamak is created (FIG. 1) leading to losses of counter-passing EP [5]. We provide for the first time evidence that EP dynamics is governed by trapping-induced super-diffusion, leading to asymmetric transport and a net toroidal torque.
The analysis is extended to the TMs, known to induce EP losses [6-9], which is extensively analysed here by means of GCT in the context of alpha particle transport in ITER-like scenarios. As for EGAMs, counter-passing particles are predominantly lost (FIG. 2), whose impact on ITER plasma confinement and tokamak performance is assessed. For this purpose, a reduced model has been developed based on the evolution of the thermal and EP populations in the presence of single-helicity modes using a transport coefficient calculated from the previous numerical simulations. We present parametric analyses for EP transport and losses in ITER-like scenarios.
The theoretical explanation of the observed transport is as follows. EP are characterized by large orbit widths. In the presence of one single (m,n) mode, the large orbit width generates multiple resonances in phase-space, parametrized by a generalized mode number n_2. If these resonances overlap, transport can occur. As an example, we plot in FIG. 3 some of the resonances created by a single helicity (m,n) mode. The plotted resonances are represented by the blue, red and green curves, ranging from smaller to higher mode numbers n_2. On the same figure, the boundary between the trapping and passing particles is given by a dashed-dotted line. An example of passing EP trajectory is given by the magenta curve. Such an EP is initially confined and exhibits a perturbed motion in the presence of the resonances, approaching the edge of the tokamak where it is lost. The width of the island corresponding to each of the resonances is calculated analytically and a criterion for the overlap is derived. This criterion is used to quantify the radial transport of EP in the presence of single-helicity modes, which will be compared to predictions obtained using the reduced model based on GCT and gyro-kinetic theory.
[1] R. Nazikian et al 2008 Phys. Rev. Lett. 101 185001
[2] G. Fu 2008 Phys. Rev. Lett. 101 185002
[3] D. Zarzoso et al 2013 Phys Rev Lett 110 125002
[4] D. Zarzoso et al. 2012 Phys. Plasmas 19 022102
[5] D. Zarzoso et al. 2018 Nucl. Fusion 58 106030
[6] H. E. Mynick 1993 Phys. Fluids B 5 2460
[7] E. M. Carolipio et al 2002 Nucl. Fusion 42 853-862
[8] M. Garcia-Muñoz et al 2007 Nucl. Fusion 47 L10-L15
[9] W.W. Heidbrink et al 2018 Nucl. Fusion 58 082027Speaker: David Zarzoso (CNRS) -
The (2,1) neoclassical tearing mode (NTM) has been proposed as a candidate to explain the larger than expected losses of high energy ions produced by neutral beam injection observed in DIII-D and ASDEX-U [1-4]. Although the numerical simulations performed so far to study the effect of NTMs on energetic ions have reproduced some features of the experimental results, the situation is not completely satisfactory. In particular, it has been difficult to reproduce the total amount of losses, which are affected by the details of the perturbation in the edge region [3-4]. We are studying the effect of NTMs on the confinement of energetic ions produced by NBI injection using a full orbit code [5] that includes the time dependent perturbed electric and magnetic fields.
The main result of this study is that when the frequency of the NTM matches the precession frequency of the trapped particles (f ~ ), the losses significantly increase. According to our simulations, the main losses correspond to trapped particles (with average pitch at the loss point of 0.53). This is in accordance with experiments performed in ASDEX U [2], where the lost ions had a defined energy and pitch. The perturbed electric field increases the losses of initially passing particles by changing the nature of their orbit, from passing to trapped and so on.
The fields employed to calculate the trajectories are the sum of a 2D equilibrium magnetic field plus the 3D electric and magnetic fields produced by the NTM. To calculate the perturbed fields we employ the experimental information available (i. e. width of the magnetic island, temperature profiles, etc) and the method proposed in [6]. The perturbed magnetic field is obtained from the perturbed poloidal flux, which is calculated by solving Ampere's equation with a parametrized perturbed current (see eq. (5) in Ref. [6]). The amplitude of the perturbed current is adjusted to obtain the desired island width. The perturbed electric field is calculated with the resistive Ohm's law, where the velocity is obtained from the displacement, which is parametrized as in eq. (2) of Ref. [6].
We simulate the NBI with a set of 250000 particles initially distributed uniformly inside the plasma, with fixed energy and uniformly distributed pitch between 0.2-0.9. Figure 1 shows the trajectory of the guiding center (calculated from the exact orbit) of a 70 keV D ion. The left frame shows the unperturbed case, the middle one the case with a static magnetic perturbation and the right one the case with time dependent electric and magnetic fields. In the las two cases the exact orbit crosses the last closed magnetic surface (green line).
Figure 2 shows the percentage of trapped and passing particles that are lost as a function of the normalized NTM frequency (Ω is the cyclotron frequency of the ions) for two energies (70 and 35 keVs). The precession frequency distribution of the trapped particles is also showed, with full boxes (70 keV in brown and for 35 keV in cyan). It is clear that the losses of trapped ions peak when the precession frequency of the trapped particles is similar to the mode frequency. Passing particle losses are lightly affected by the rotating mode. When the perturbation is reduced the losses are reduced.
Finally, figure 3 shows ion losses as a function of time in a narrow interval of toroidal angle (between 0 and 0.2) for an NTM with frequency 3.75 e-4 Ω. The losses produce a periodic signal with the frequency of the NTM mode.
1 E. M. Carolipio et al. Nucl. Fus. 42, 853 (2002).
2 M. García-Muñoz et. al. Nucl. Fus. 47, L10 (2007).
3 E. Strumberger et al. N. Jour. Phys. 10, 023017 (2008).
4 M. Gobbin et al. Nucl. Fus. 49, 095021 (2009).
5 C. F. Clauser et al. Comput. Phys. Comm. 234, 126 (2019).
6 V. Igochine et al, Nucl. Fus. 43 1801 (2003).Speaker: hugo ferrari (Comision Nacional de Energia Atomica CNEA) -
Turbulence driven shear flow through Reynolds stress associated with the coexistence of multiple edge instabilities lowers the L-H power threshold ($P_{LH}$) across multiple parameters on DIII-D: $q_{95}$ (Fig. 1), ion ∇B drift direction (Fig. 2), plasmas with and without resonant magnetic perturbation (RMP) (Fig. 3), as well as ion isotope mass {1}. Application of RMP raises turbulence decorrelation rates and reduces Reynolds stress driven flow and flow shear, hence increasing the L-H power threshold {2}. These results demonstrate the importance of the turbulence and turbulence driven flow in lowering the L-H transition power threshold. They support current L-H transition theories {3} but suggest a complex behavior that can inform a more complete physics-based model of the L-H transition power threshold for ITER and beyond.
Dual mode density fluctuations are consistently observed when the L-H transition power threshold is low at higher $q_{95}$ and thus appear to at least partially explain the lower threshold. Long wavelength density fluctuations are measured across the L-H transition at DIII-D using an 8×8 array of Beam Emission Spectroscopy (BES) channels at 0.85<ρ<1. The plasmas were operated in favorable geometry (ion ∇B drifts towards the X-point) in ITER similar shape with balanced torque neutral beam injection. The L-H transitions were obtained with the heating power just above the threshold. As seen in Fig. 1 (a), $P_{LH}$ is reduced at higher $q_{95}$=4.9 at $n_e$~3e19 $m^{-3}$, which is near the $P_{LH}$ density minimum on DIII-D. Ion and electron diamagnetic directed modes are observed to coexist at ρ~0.95-1.0 propagating in different directions as indicated in the wavenumber spectrum (Fig. 1c): a low frequency broadband mode (<10 kHz) propagates in the ion diamagnetic drift direction and a higher frequency mode (>10 kHz) in the electron drift direction (labeled accordingly). A single mode is seen at lower $q_{95}$=3.5 (Fig. 1b).
The two modes observed at $q_{95}$=4.9 appear to be associated with enhanced turbulence Reynolds stress and larger poloidal turbulence flow shear {4}. Linear gyro-kinetic CGYRO simulations with experimental profiles have shown a transition from one mode to two modes as $q_{95}$ is increased consistent with the experimental observations. It is found that $T_i$ profiles strongly impact the ion mode growth rate, even in the pedestal region. On ASDEX-U and C-Mod, a critical ion heat flux has been reported to be necessary for L-H transition {5,6}. This critical ion heat flux may be related directly to the edge ion temperature gradient, which is observed here to strongly impact the ion mode.
An increase of turbulence Reynolds stress and turbulence flow shear is observed when the ion ∇B drift changes from unfavorable to favorable by scanning the dRSEP parameter, the radial distance between the upper and lower divertor separatrices at the outboard midplane. As seen in Fig. 2, three 20 ms time windows are analyzed during the dRSEP scan approaching the L-H transition. At earlier time around 1500 ms, which is in unfavorable direction, turbulence Reynolds stress measured with BES is very small. However as it changes towards the favorable direction approaching the transition (1960 ms-1980 ms), Reynolds stress is significantly increased. Consistently, poloidal turbulence flow increases and such increase is mainly in the plasma edge region at ρ~0.95-1.0. Toroidal field, plasma current and input power are kept constant during the process and there is little change in the edge plasma profiles of density and temperature. The turbulence amplitude decreases approaching the transition.
Application of RMP at a magnitude required to suppress ELMs has been observed for the first time to raise the turbulence decorrelation rates and reduce Reynolds stress driven flow shear. It has been observed from multiple experiments and across multiple devices that RMP increases L-H transition power threshold. The turbulence decorrelation rate (Δω$_D$) measured with BES in the plasmas with and without MP application is shown in Fig. 3(a).
It is found that Δω$_D$ increases by 50% across ρ~0.9-1.0 when RMP is applied. It is simultaneously observed that the Reynolds force and turbulence flow shear are significantly reduced when RMP is applied (Fig. 3(b) and (c)). These increases in Δω$_D$ and reductions in the flow shear rate disrupt the turbulence suppression mechanism. It is also found that the application of RMP reduces transient kinetic energy transfer from turbulence to the flow approaching the L-H transition {2}. When non-resonant MP is applied, little change in the decorrelation rate has been observed, and the turbulence Reynolds stress and flow shear exhibit less reduction, which is consistent with the much reduced impact on the L-H transition power threshold with non-resonant MP compared with RMP.
These experimental results advance our understanding of how turbulence and flow dynamics impact the L-H transition power threshold across multiple parameters, and suggest techniques to reduce the power threshold for ITER. This understanding is crucial to develop a physics-based model of the L-H transition power threshold in ITER/burning plasmas. This work was supported by the US DOE under DE-FG02-08ER54999, DE-FG02-89ER53296, DE-FC02-04ER54698, DE-FG02-08ER54984, DE-SC0018287, DE-SC0019352 and DE-AC02-09CH11466.
{1} Z. Yan, P. Gohil, et al., Nucl. Fusion 57, 126015 (2017)
{2} D.M. Kriete, G.R. McKee, et al., Phys. Rev. Lett. Submitted, 2019
{3} K. Miki, P.H. Diamond, et al., Phys. Plasmas, 19, 092306, (2012)
{4} Z. Yan, G.R. McKee, et al., Phys. Plasmas 26, 062507 (2019)
{5} F. Ryter, L. Barrera Orte, et al., Nucl. Fusion 54, 083003 (2014).
{6} M. Schmidtmayr, J. W. Hughes, et al., Nucl. Fusion 58, 056003 (2018).Speaker: Dr Zheng Yan (University of Wisconsin - Madison) -
New theoretical study of Alfvén eigenmodes (GAEs) in the sub-cyclotron frequency range explains the observed GAE frequency scaling with beam parameters in experiments across devices (1-3). Global Alfvén eigenmodes are frequently excited during neutral beam injection (NBI) in the National Spherical Torus Experiment (NSTX/NSTX-U) (4,5), as well as other beam-heated devices such as MAST and DIII-D (6). These modes are driven unstable through the Doppler shifted cyclotron resonance with the NBI ions, and can be excited in ITER due to super-Alfvénic velocities and strong anisotropy of the beam ions. They can also be excited by alpha particles near the outer edge of the ITER plasma due to anisotropies in the alpha particle distribution. Observations link these modes to flattening of electron temperature profiles and anomalously low central temperature at high beam power in NSTX (7), therefore, the ability to control them will have significant implications for NSTX-U, ITER, and other fusion devices where Alfvénic and super-Alfvénic fast ions might be present.
Numerical simulations using the HYM code have been performed to study the excitation of GAEs in NSTX, NSTX-U and most recently for DIII-D. The HYM code is an initial value 3D nonlinear, global stability code in toroidal geometry, which treats the beam ions using full-orbit, delta-f particle simulations, while the one-fluid resistive MHD model is used to represent the background plasma. Nonlinear HYM simulations show unstable counter-rotating GAEs with toroidal mode numbers, frequencies and saturation amplitudes that match the experimentally observed unstable GAEs in NSTX-U and NSTX (1,2).
New simulations performed for typical DIII-D plasma and beam parameters demonstrate that high-frequency modes with $\omega/\omega_{ci}\sim 0.6$, previously identified as compressional Alfvén eigenmodes (CAEs) (6), have shear Alfven polarization and are in fact the GAEs (Fig.[1]). Simulations show unstable counter-propagating GAEs with $\delta B_\|<0.1 \delta B_\perp$, high toroidal mode numbers $|n|>20$, and frequencies close to the observed $\omega/\omega_{ci} \sim 0.6-0.7$ for NBI injection velocity $V_0/V_A=0.9$. The unstable modes have $k_\perp \rho_b\sim 0.5$, and growth rates $\gamma/\omega_{ci} \sim 0.002-0.003$. These simulation results combined with growth rate calculations (Fig.[2]) based on the local dispersion relation (1,3) explain some of the puzzling DIII-D observations. Namely, high-frequency modes (identified as CAEs in Ref.(6)) observed in DIII-D had small values of $k_\perp \rho_b$ ($k_\perp \rho_b \sim$ 0.8, in some cases $<$0.5), i.e. smaller than previously predicted (5,8) for the unstable CAEs ($1 < k_\perp \rho_b <2 $) or GAEs ($2 < k_\perp \rho_b <4$); the ratio $\omega/\omega_{ci}$ remained approximately constant when the toroidal field was varied; the most unstable mode frequency scaling with $n_e$ was weaker than the Alfven speed scaling, i.e. weaker than $1/\sqrt{n_e}$; the observed frequency splitting was not consistent with theoretical predictions for CAEs.
Simulation results and growth rate calculations (1,3) show that previously derived and widely cited instability conditions for counter-propagating GAE: $2< k_\perp \rho_b <4$ and for CAE: $1< k_\perp \rho_b <2$ (5,8) are valid only for higher frequency modes with $\omega\sim \omega_{ci}$, when the beam injection velocity is very large compared to the resonant velocity $V_0 \gg V_{res}$. The new theory predicts that the GAE linear growth rate is largest for small values of $k_\perp$ ($k_\perp \rho_b<1$) (Fig.[2]). For a beam ion distribution function $f(\lambda) \sim exp[-(\lambda-\lambda_0)^2/\Delta \lambda^2]$, a sufficient condition for the instability is: $1-V_{\|res}^2/V_0^2 ≤ \lambda_0$ , where $\lambda=\mu B_0/\varepsilon$ is a pitch related parameter. The most unstable counter-GAEs have frequencies in the range (1):
$ (1+V_0/V_A)^{-1} < \omega/\omega_{ci} ≤ (1+V_0/V_A\sqrt{1-\lambda_0})^{-1} , (1) $
where the most unstable frequency roughly corresponds to the upper bound in Eq.(1). For DIII-D parameters ($V_0/V_A \sim 1$, $\lambda_0 \sim $0.5-0.7), this gives $0.5<\omega/\omega_{ci} ≤0.65$, consistent with observations and the simulation results. The observed scaling with $B_{tor}$ , $\lambda_0$ and weak $n_e$ scaling also agrees with Eq.(1). For example, “left beam sources” in DIII-D, have $\lambda_0 \approx$ 0.5 and excite modes with $\omega/\omega_{ci} \approx$ 0.56, whereas for “right sources” $\lambda_0 \approx$ 0.78 and $\omega/\omega_{ci} \approx$ 0.69 (6). Our theory predicts $\omega/\omega_{ci}\approx$ 0.58 and 0.68 respectively. In addition, it is shown that CAEs have the same instability condition and range of unstable frequencies as the GAEs in the limit $k_\perp \ll k_\|$, otherwise CAE’s growth rates are much smaller than that of GAEs, consistent with simulation results and the NSTX(-U) observations.Expression (1) correctly predicts the scaling of the most unstable GAE frequencies with NBI parameters for NSTX, NSTX-U and DIII-D. Thus, for $\lambda_0 \sim$0.6 and large normalized injection velocities in NSTX ($V_0/V_A\sim$3-5), the predicted frequency range $\omega/\omega_{ci}\sim$0.2-0.3 agrees with observations. Due to stronger toroidal field in NSTX-U, and smaller relative injection velocities ($V_0/V_A≲$2), higher frequency GAEs were observed $\omega/\omega_{ci}\sim$0.4 (4). For DIII-D conditions with $V_0/V_A\sim$1, a much higher frequencies predicted consistent with the observed $\omega/\omega_{ci}\sim$0.6 (6).
References
1. Belova E.V., E. Fredrickson, J. Lestz, N. Crocker, Phys. Plasmas 26, 092507 (2019).
2. Belova, E.V., Gorelenkov, N.N., Crocker, N.A., et al., Phys. Plasmas 24, 042505 (2017).
3. Lestz, J., N.N. Gorelenkov, E.V. Belova, et al., Phys. Plasmas 27, 022513 (2020).
4. E. D. Fredrickson et al., Phys. Rev. Lett. 118, 265001 (2017).
5. Gorelenkov, N.N., Fredrickson, E, Belova, et al., Nucl.Fusion 43, 228 (2003).
6. Heidbrink, W., E. Fredrickson, N. Gorelenkov, et al., Nucl. Fusion 46, 324 (2006).
7. D. Stutman, et al., Phys. Rev. Lett. 102, 115002 (2009).
8. Kolesnichenko, Ya.I., White R.B., Yakovenko, Yu.V., Phys. Plasmas 13, 122503 (2006).Speaker: Elena Belova (PPPL)
-
-
-
Nonlinear two-fluid MHD simulations reveal the role of resonant field penetration in ELM suppression and density pump-out in low-collisionality ITER-Similar-Shape (ISS) plasmas in the DIII-D tokamak$^1$. The operational window for ELM suppression in DIII-D ISS plasmas coincides with calculations for magnetic island formation at the pedestal top ($n_e<3\times10^{19}m^{-3}, B_r/B_t >10^{-4}, \nu_e^*<0.3$) based on nonlinear MHD simulations using the TM1 code$^2$. Key phenomenology in experiment are reproduced in the simulations including density pump-out for field penetration at the foot of the pedestal and pedestal pressure and width reduction due to magnetic island formation at the top of the pedestal. TM1 simulations also reproduce the observed $q_{95}$ width of ELM suppression windows in DIII-D. Analysis indicates that wide $q_{95}$ windows of ELM suppression may be accessible in DIII-D and ITER by operating at higher toroidal mode number.
For our analysis we use the cylindrical initial value nonlinear two-fluid MHD code TM1$^2$ with the helical magnetic field boundary condition provided by the ideal MHD code IPEC$^3$. The fully toroidal ideal MHD code IPEC calculates the perturbed 3D magnetic equilibrium (vacuum field plus ideal MHD kink/peeling response to the I-coils) in the actual magnetic geometry. TM1 takes the measured kinetic profiles from DIII-D before the RMP is applied, including initial transport coefficients and neoclassical resistivity from TRANSP. Multiple helical field harmonics from IPEC are applied at the simulation boundary for a given toroidal mode number (e.g. $m/n=6/2, 7/2, 8/2, 9/2, 10/2, 11/2, 12/2$ harmonics for $n=2$ RMP). The TM1 simulation then predicts the penetration or screening of these resonant fields in the plasma interior and the effect of these fields on the electron density and temperature profile. The TM1 model solves the torque balance that governs the bifurcation from screening to penetration of resonant fields including diamagnetic drifts that are important in the pedestal. TM1 also solves the electron continuity and energy transport equations, taking into account enhanced collisional (parallel) transport across magnetic islands.
TM1 simulations show that RMPs readily penetrate into the collisional foot of the DIII-D pedestal near the separatrix ($\psi_N>0.98$), generating a narrow region of edge stochasticity and enhanced collisional transport. The enhanced transport leads to density pump-out, reproducing the magnitude of the observed density reduction observed in experiment$^{1,4}$. While pump-out is ubiquitous at low collisionality, special conditions are required for ELM suppression in DIII-D ISS plasmas, including low plasma density ($<3\times10^{19}m^{-3}$), high co-$I_p$ toroidal rotation and high RMP amplitude ($B_r/B_t >10^{-4}$)$^5$. These conditions correspond to the requirements for resonant field penetration at the top of the pedestal from TM1 simulations. Figure 1 (a) shows the similarity in the TM1 predicted (white) and measured electron pressure profile (yellow) during pumpout and ELM suppression using $n=2$ RMPs in a DIII-D ISS plasma with $q_{95}\approx4.1$, together with the Poincaré plot showing penetration of resonant fields at the top and foot of the pedestal. ELMs are suppressed when the pedestal height and width fall $\approx15\%$ below the EPED model prediction$^6$. Several hundred non-linear TM1 simulations were performed to derive the following scaling relation for the field penetration threshold at the pedestal top in DIII-D ISS plasmas
$\ \ \ \ \ \ \ \ \ \ \ \ \ \ B_{r,th}^{scale}/B_t=3.5\times10^{-2}n_e^{0.7}|\omega_E+\omega_{*e}|^{0.9}B_t^{-1}, \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ $(1)
where $\omega_E$ and $\omega_{*e}$ are the $E\times B$ and electron diamagnetic drift frequency. This scaling closely resembles analytic estimates for error field penetration in the plasma core$^7$ and is quantitatively consistent with the threshold for ELM suppression in DIII-D. A 2D contour plot of the penetration threshold $B_r/B_t$ at the pedestal top versus density and $\omega_E$ is shown in Figure 1 (b). The RMP amplitude for penetration decreases (yellow to blue) as the density and the magnitude of $\omega_E$ decrease, consistent with experiment for ELM suppression$^5$. A database of $n=2$ RMP discharges are overlaid onto Figure 1 (b), with ELM suppression (yellow stars) and ELMing (black stars). The boundary contour between suppression and ELMing corresponds to the maximum available RMP amplitude $B_r/B_t\sim3.5\times10^{-4}$ in the experiment (white dashed line), representing a remarkable level of agreement between experiment and nonlinear MHD theory for the ELM suppression threshold. Analysis of a model ITER $Q=10$ discharge$^8$ predicts that the threshold for $n=3$ penetration at the top of the pedestal will be significantly lower ($B_r/B_t\sim2\times10^{-5}$) than in DIII-D due to the lower rotation and diamagnetic frequency expected in ITER.
TM1 simulations also reveal the conditions required for wide $q_{95}$ windows of ELM suppression. Wide $q_{95}$ windows are essential for operational flexibility in ITER. However, near the threshold of resonant field penetration at the top of the pedestal, only narrow $q_{95}$ windows of ELM suppression are observed in DIII-D. These narrow windows are reproduced using TM1 simulations and EPED model predictions. Figure 2 shows the predicted pedestal pressure reduction (color contour) versus $q_{95}$ from TM1 simulation and its intersection with the $n=2$ and $n=3$ RMP amplitude on the plasma boundary obtained from IPEC (horizontal dashed lines). For similar RMP amplitudes on the plasma boundary, $n=3$ produces wider $q_{95}$ windows of ELM suppression than $n=2$, consistent with DIII-D experiment$^5$. Figure 2 (c) shows a TM1 simulation with $n=4$ RMPs in DIII-D, indicating window overlap for the same ISS plasma condition. The TM1 simulations predict access to wide operational window of ELM suppression with relatively weak pressure reduction at higher-$n$. Analysis of model ITER equilibria also reveals a similar trend with toroidal mode number, suggesting that operating with dominant $n=4$ RMP, accessible to the ITER ELM coil design, may be advantageous for ITER operation.
To conclude, TM1 simulations quantitatively reproduce ELM suppression windows and density pumpout in the DIII-D tokamak. ITER simulations suggest that the threshold for penetration will decrease relative to DIII-D due to the expected lower $E\times B$ and diamagnetic frequency expected in ITER. Our simulations account for the observed width of $q_{95}$ windows of ELM suppression in DIII-D and indicate that $n=4$ RMPs may be effective to produce wide windows of ELM suppression in ITER.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC02-09CH11466 and DE-FC02-04ER54698.
Reference:
1 Q.M. Hu et al., Phys. Plasmas 26, 120702 (2019).
2 Q. Yu and S. Günter, Nucl. Fusion 51, 073030 (2011).
3 J.K. Park and N.C. Logan, Phys. Plasmas 24, 032505 (2017).
4 Q.M. Hu et al., Nucl. Fusion submitted (2019).
5 C. Paz-Soldan et al., Nucl. Fusion 59, 056012 (2019).
6 P.B. Snyder et al., Phys. Plasmas 19, 056115 (2012).
7 R. Fitzpatrick, Phys. Plasmas 5, 3325 (1998).
8 F.M. Poli et al., Nucl. Fusion 58, 016007 (2018).Speaker: Dr Qiming Hu (Princeton Plasma Physics Laboratory) -
Edge Localized Modes (ELMs) comprising of repetitive plasma eruptions from the edge of a tokamak plasma are a very common feature of high confinement mode (H-mode) operation in advanced tokamak devices. Each ELM outburst is associated with an expulsion of a large amount of energy and particles in a very short time that can potentially cause serious damage to plasma facing components and hence their control and mitigation are a major concern for safe tokamak operation. One of the most successful approaches for ELM control is by the use of resonant magnetic perturbations (RMPs) introduced from the edge of the plasma. A large number of experimental studies as well as theoretical investigations devoted to this topic are currently in progress worldwide.
The dynamics of ELMs is quite complex. The linear characteristics of the mode are well modeled in terms of the peeling/ ballooning instabilities$^1$ but their nonlinear evolution is not yet fully understood. Introduction of RMPs further complicates the dynamics and are best studied using numerical simulations. In an actual operational scenario additional physical factors such as plasma rotation can introduce significant modifications in the dynamics of ELMs. A few past studies$^{2,3}$ have shown that toroidal and/or poloidal rotation can significantly influence the stability properties of ELMs. In particular a sheared toroidal rotation is found to stabilize a Type-I ELM whereas a poloidal rotation can be either stabilizing or destabilizing depending on the direction of the flow. Some of these effects have also been experimentally identified. In the light of the above findings it is important to investigate what happens to the efficacy of RMPs in the presence of plasma rotation. Our present study is devoted to such an investigation.We have carried out detailed nonlinear simulations to study the dynamics of ELMs in the presence of a n=2 resonant magnetic perturbation. The simulations have been done on the Culham Transporter of Ions and Electrons (CUTIE) code$^{4,5}$ - a two-fluid initial value electromagnetic nonlinear global code which solves the full set of model two fluid equations at a scale intermediate between the device size and the ion gyro-radius and takes account of classical and neoclassical transport effects. A periodic cylinder model of the tokamak geometry is adopted with the toroidal curvature effects of the magnetic field lines kept to first order in the inverse aspect ratio. Poloidal mode coupling is implemented through additional curvature terms. ELMs are simulated by introducing a particle source in the confinement region and a particle sink in the edge region. CUTIE is capable of simulating multiple edge relaxation periods and has in the past also successfully modeled the L-H transition behavior in COMPASS-D$^4$. More recently CUTIE has been used to study ELM dynamics in the presence of RMPs$^6$ as well as pellet injection$^7$.
To study the effect of plasma rotation, we have introduced a toroidal momentum source to generate an equilibrium plasma rotation. An external magnetic perturbation of n=2 is also present to simulate the effect of RMPs. Our principal findings are as follows. The presence of flows significantly impacts both the magnitude and frequency of the ELM modes. In Fig.1 we show the time evolution of the ion temperature fluctuations (dTi/Ti0) for various situations e.g. with or without the presence of RMPs and flows. We see that the presence of a toroidal flow in the co-current direction has a positive impact on the mitigation process. In particular, the amplitudes of the ELM bursts are seen to be significantly diminished and their frequencies increased. This also leads to an overall improvement of the energy confinement and an increase in the plasma beta. This is shown in Fig. 2 for the various cases. The flow values that we have used at the axis are M= 0.011, 0.027 and 0.042 where M the Mach number is defined in terms of the Alfven velocity. Such flows are likely to be present in tokamaks either due to external driving from neutral beams or spontaneous generation from turbulence effects. Other noticeable changes (not shown) are in the profile modifications at the edge and the spectra of energy transfer. In general we also find that a co-current flow is more effective than a counter-current flow – in fact a large counter-current flow actually reduces the stabilizing influence of the RMP. Our results could be of direct relevance to present day experiments on RMP control of ELMs on tokamaks like DIII-D where substantial sheared flows can exist due to unbalanced NBI injection$^8$.References
$^1$P.B. Snyder, H.R. Wilson, T.H. Osborne et al, Plasma. Phys. Contr. Fus. 46 (2004) A131.
$^2$F. Orain, M. Becoulet, J. Morales et al, Plasma Phys. Contr. Fus. 57 (2015) 014020.
$^3$N. Aiba, M. Furukawa, M. Hirota et al, Nuclear Fusion 51 (2011) 073012.
$^4$A. Thyagaraja, Plasma Phys. Contr. Fus. 42 (2000) B255.
$^5$A. Thyagaraja, M. Valovic and P.J. Knight, Phys. Plasmas 17 (2010) 042507.
$^6$D. Chandra, A. Thyagaraja, A Sen et al, Nuclear Fusion 57 (2017) 076001.
$^7$D. Chandra, A Sen and A. Thyagaraja, Plasma Phys. Contr. Fus. 61 (2019) 085019.
$^8$T.E. Evans, M.E. Fenstermacher, R.A. Moyer et al, Nuclear Fusion 48 (2008) 024002.
Speaker: Dr Debasis Chandra (Institute for Plasma Research) -
EAST is typically operated on the radio-frequency (RF) waves heating scenarios, which is also highly related to ITER and CFETR. For the recent experiments of EAST H-mode operations, RF waves are found to be efficient to mitigate and suppress the edge instabilities, such as ELMs [1,2]. From the simulation point of view, the RF heating and driving effects can be studied from 2 aspects: directly on the equilibrium and directly on the instabilities. For the first aspect, the heating effects by RF waves are presented as the increase of the temperatures, and the current driving as the change of the current density profile, then those profiles are considered into the reconstruction of the magnetic equilibrium. In the previous work, the equilibriums including RF effects have already been used for the studies of the ELM behaviors [3,4]. Therefore, this presentation only focuses on the physics understanding for the active control of the edge instabilities by RF waves directly.
It is considered that there are several different mechanisms which are able to achieve the suppressions and mitigations on the edge instabilities by RF waves directly. For example, the low-hybrid wave (LHWs) and ion-cyclotron wave (ICW) show different capabilities on the edge instability control due to their different physics in the plasmas. LHW can excite the forced mode, and the strong nonlinear mode coupling between the forced and spontaneous modes leads to the mitigation of the edge turbulence. For ICW, the sheath potential of the ICW antenna is essential to change the radial electric field and able to suppress the linear growing of ELMs. Therefore, the six-field two-fluid module in BOUT++ framework [5] is extended based on the requirements to simulate the interactions between edge instabilities and RF waves [6].
The LHW is considered to drive the helical filamentary current (HFC) in the SOL region which can change the boundary topology by the radial magnetic field[1]. This radial magnetic field is treated as the initial perturbation in the simulations. The nonlinear evolution of the different modes of Te fluctuations are shown in figure 1 (a) and (b). The most obvious difference is the linear growing phase of the fluctuations which are changed by HCF. Figure 1(b) shows that all the modes of Te grow much more slowly than the case without HCF. Although the radial magnetic field induced by HCF could be much smaller than the perturbed field of the edge turbulence, it is still able to excite the perturbations with n = nHCF to grow up at the start of the linear phase. As shown by the red curve in figure 1(b). This forced mode is effective to compete with the spontaneous fluctuations and changes the spectrum of the eigenmodes even in the linear growing phase. There are strong nonlinear wave–wave interactions happened between those modes, and the phase coherence time gets decreased dramatically [7]. Therefore, none of the modes can grow to its amplitude in the no-HCF case. As shown in figure 1(c), the averaged total fluctuations of all the modes for the HCF case is able to be mitigated by 35% at least. This implies the edge turbulence is able to be mitigated by HCF. The mechanism gives a potential explanation to the active control on ELMs with LHW. This mitigation mechanism by HCF from LHW is quite similar to RMP coils, and could provide the more flexible method compared to the fixed RMP coil systems.
![(a) The time evolution of the different toroidal modes without HCF. (b) The same figure for the case with HCF. The nHCF = 5 case is exhibited as the reference. (c) The time evolution of the fluctuations for the original case without HCF (black), with n = 5 HCF (red) and n = 4 HCF (blue). The dashed lines represent the time averages of the fluctuations in the nonlinear saturation phases, respectively[6].](http://theory.ipp.ac.cn/wp-content/uploads/fig1.jpg)
The ELMs are observed fully suppressed by the ICW heating during the H-mode discharge #77741 in EAST, as shown in Fig 2(a). The simulations with BOUT++ reveals that the key factor is the RF sheath on the ICW antenna. The RF sheath potential is able to increase the radial magnetic field inside the SOL and introduces a large flow shear across the separatrix [8]. The linear growth rate can be decreased to less than 0.02 with RF sheath effects in Fig. 2(b), and the ELM size is decreased from over 10% to 0.66%, which are mainly consistent with the experimental results in Fig. 2(c). The large flow shearing rate also enhances the nonlinear wave-wave interactions between the different modes, so the free energy is shared by these modes, which can decrease the energy loss and suppress the ELMs effectively. In the experiments, the probability of the ELM suppression by ICRF is low. In the simulations, it is also found that if the ELM is fully suppressed by ICRF, the requirements for the radial electric field Er is relatively high. A small sheath potential range are found for the good ELM suppression by ICRF, as shown in Fig. 2(d).
As the conclusion, this presentation exhibits two different mechanisms on the mitigation and suppression of the edge instabilities by two different waves. LHW excites the forced mode through HCF, and the strong nonlinear mode coupling between the forced and spontaneous modes leads to the mitigation of the edge turbulence. The sheath potential of the ICW antenna is essential to change the radial electric field and suppress the linear growing of ELMs. Both mechanisms can be explained by the enhancement of the nonlinear wave-wave interactions. Therefore, both LHW and ICW show a potential to be an effective method for the active ELM control besides RMP coils and pellet injections for the future tokamaks, such as CFETR and ITER.
[1] Liang Y.F. et al 2013 Phys. Rev. Lett. 110 115002
[2] Wan B.N. et al 2015 Nucl. Fusion 55 104015
[3] Wu Y.B. et al 2018 Plasma Phys. Control. Fusion 60 055007
[4] Huang Y.Q. et al 2020 Nucl. Fusion 60 026014
[5] Xia T.Y. et al 2013 Nucl. Fusion 53 073009
[6] Xia T.Y. et al 2019 Nucl. Fusion 59 076043
[7] Xi P.W. et al 2014 Phys. Rev. Lett. 112 085001
[8] Gui B. et al 2018 Nucl. Fusion 58 026027Speaker: Tianyang XIa (Institute of Plasma Physics, Chinese Academy of Sciences) -
E-mail: ysna@snu.ac.kr
Edge Localized Modes (ELM) are rapid MHD events occurring at the edge region of tokamak plasmas, which can result in damages on the divertor plates. Therefore, to fully suppress ELM via resonant magnetic perturbation (RMP) [1-4] is of great help to reach and sustain high-performance H-mode plasmas. It was found that certain conditions must met for the RMP-driven ELM crash suppression [5], so understanding its mechanism is crucial for reliable ELM control using RMP. The initial understanding of its mechanism was a stabilization of linear edge instability due to the pedestal gradient degradation by RMP. However, experimental observation showed that peeling-ballooning mode (PBM)-like mode structures remained in the ELM suppression phase [6]. In addition, the bifurcation of mode rotation in the edge region was found to be closely related to mode suppression [7]. Therefore, the initial understanding may have difficulties in explaining these experimental findings and it indicates that additional physics properties should be included to understand the mechanism. For this purpose, we have carried out nonlinear MHD simulations with 3D reduced MHD code, JOREK [8] for a recent n=2 RMP-driven ELM-crash-suppression in KSTAR [7]. We successfully reproduced [9] the natural ELM without RMP (Fig.1(a)), mode mitigation with small RMP strength ($I_{RMP}=2kA$, Fig.1(b)), and mode suppression by experimental RMP strength ($I_{RMP}=4kA$, Fig.1(c)). Also, such ELM-crash-suppression is attributable not only to the degraded pedestal but also to direct coupling between peeling-ballooning mode (PBM) [10] and RMP-driven plasma response. The coupling between PBM and RMP can 1) enhance the size of the island at the pedestal reducing the instability source by further pedestal degradation, and 2) increase the spectral transfer between edge harmonics preventing catastrophic growth and crash of unstable mode. Because of these effects, PBMs are nonlinearly saturated, and they persist during the suppression phase without a mode crash. This outcome is consistent with the previous studies [11,12]. In addition, the locking (or rotation bifurcation) of PBMs has been numerically simulated during the suppression phase. This mode-locking is a distinguishing feature of the mode suppression as rotating mode structure remains for the natural ELM and mode mitigated case (Fig. 1(d)). PBM locking may enhance the interactions between PBMs and RMP, and therefore, it is favorable to RMP driven ELM suppression. Here, slowly rotating PBM before RMP application can be easily locked by RMP. As $V_{E×B}$ is approximately equal to the initial PBM rotation in our case, $V_{E×B}\approx0$ in the pedestal will be advantageous to onset of ELM-crash-suppression. To test our hypothesis, we conduct additional RMP-ELM simulation with modified $V_{E×B}$, and confirm that the mode suppression is not achieved with enlarged $V_{E×B}$ on top of the pedestal, which may support the importance of $V_{E×B}\approx0$ on ELM-crash-suppression.
References:
- T. Evans et al., PRL 92, 235003 (2004)
- Y. Jeon et al., PRL 109, 035004 (2012)
- Y. Sun et al., Phys. Rev. Lett. 117, 115001 (2016)
- W. Suttrop et al., Plasma Phys. Control. Fusion 59, 014049 (2017)
- J. K. Park et al., Nature Physics 14 (2018), 1223
- J. Lee et al., Physics Review Letter 117 (2016), 075001
- J. Lee et al., Nucl. Fusion 59, 066033 (2019)
- G. T. A Huysmans et al., Plasma Phys. Controlled Fusion 51, 124012 (2009)
- S. K. Kim et al., Nucl. Fusion 60, 026009 (2020)
- J. W. Connor et al., Plasma Phys. Controlled Fusion 27, 621 (1985)
- M. Becoulet et al., PRL 113, 115001 (2014)
- F. Orain et al., Phys. Plasmas 26, 042503 (2019)
Speaker: SangKyeun Kim (Seoul National University) -
Highlight of this work: This work predicts the optimal coil phasing, semi-empirical threshold coil current and ‘favorable’ $q_{95}$ window for ELM mitigation for HL-2M 1MA discharge scenario. It is found that pressure gradient may play an important role on determining the peeling-tearing displacement near X-point, due to the curvature effect (GGJ effect) of equilibrium magnetic field.
Resonant magnetic perturbation (RMP) generated by external coils is an effective method to suppress or mitigate edge localized mode (ELM) in H-mode toroidal plasma. Extensive efforts have been devoted to understand the mechanism of controlling ELM. It is demonstrated that edge-peeling response to RMP fields plays an essential role. The linear single fluid model (employed in MARS-F code) predicted results are in good agreement with experiential measurements in many cases $[1]$. MARS-F is widely applied to interpret the experimental observations and to optimize of RMP coil configuration.
This work focuses on the optimization of coil phasing for ELM control for the coming HL-2M tokamak device and on the influence of pressure profile at pedestal region on RMP fields, using MARS-F code. In the computations, plasma rotation and plasma resistivity are included. The former induces the screening of the applied RMP fields, while the latter yields the penetration of field. Moreover, the strong parallel sound wave damping term is also included, which moderately damps the core-kink response. The resonant radial perturbed magnetic field component b$^1_{res}$ at plasma edge or the plasma displacement ($\xi_X$) near the X-point is taken as the indicator to optimize coil phasing here. Actually, these two criteria are basically equivalent $[2]$.
We consider a equilibrium of HL-2M with 1.0 MA current. The key parameters are: major radius $R_0=1.75$ m, minor radius $a=0.65$ m, $B_T=1.8$ T, $q_0=1.07$, $q_{95}=3.25$ and plasma normalized pressure $\beta_N=1.63$ being much smaller than the no wall beta limit ($\beta^{no-wall-limit}_N \ \sim 3.6$). There will be two off-midplane rows of coils. Each row includes 8 coils, which allows the configurations with the maximum toroidal harmonic being $n=1, 2$ and $4$.For HL-2M, the coil basic parameters were already determined, such as the the coil width ($\Delta \theta =15^o$) and radial location ($\theta_c=\pm 40^o$). However, the off-midplane coils have one degree of freedom to choose : the coil phasing $\Delta \Phi$. The current on upper and lower coils is simply expressed as $I_{upper}\propto cos(n\phi)$ and $I_{lower}\propto cos(n\phi+\Delta\Phi)$, respectively. The numerical results indicate that the optimal coil phasing for n=1, 2 and 4 are $\Delta\Phi_{opt}=\pm 180^o , 100^o$ and $-50^o$, respectively. At the optimal phasing $\Delta\Phi_{opt}$, the edge-peeling response is dominant over the so-called core-kink response $[3]$. The maximum of $\xi_X$ for n=1 is about 1.5 and 10 times larger that of n=2 and 4, respectively. While the optimal phasing is not sensitive to the choose of toroidal rotation profile and pressure profile. During the variation of pressure profile, the normalized beta $\beta_N$ and q profile are fixed. More interesting, it is found that the amplitude of $\xi_X$ is generally reduced when the pressure gradient at edge increases. This is likely due to that the pressure gradient (GGJ effect) makes kink-tearing mode more stable. It is implied that the required minimum coil current for suppressing/mitigating ELM is enhanced when plasma pressure profile becomes more sharp at edge.
The comparison between linear response modeling and experiments in MAST $[4]$ yields a critical X-point displacement $\xi_X\sim 1.5 $ mm for achieving ELM mitigation. We simply assume the critical value $\xi_X\sim 2 $ mm as the guideline for controlling ELM on HL-2M, although there are difference in plasma configuration, coil geometry, and the actual threshold coil current between these two machines. In fig.1, the solid white curves represent the 2 mm level of X-point displacement. Clearly, at the designed coil geometry ( $\Delta \theta=15^o$ , $\theta_c=\pm 40^o$ ), the required coil current depends on the choice of coil phasing. With the bad choice of coil phasing (e.g. $0<\Delta \Phi<\sim 50^o$ ), the required coil current exceeds the the allowed maximum RMP coil current (=10 kAt) as designed. On the other hand, there is a wide region of ‘good’ coil phasing, which needs $I_c< 5$ kAt for achieving ELM mitigation based on the 2 mm X-point displacement criterion. Similar study will be carried out for other toroidal mode number.
Usually, the ELM mitigation/suppression is sensitive the $q_{95}$ value $[1]$. We predict the effective $q_{95}$ window for HL-2M as shown in fig.2. For $n=1$ case, the most effective $q_{95}$ window is in 3.1<$q_{95}$<3.2, in which the maximum (e.g. at the optimal coil phasing) of b$^1_{res}$ amplitude is about 8 times larger than that outside of this window. For $n=2$ case, the best window exists near $q_{95}~3$. Another ‘favorable’ $q_{95}$ window is 3.4<$q_{95}$<3.5. It is noted that the optimal coil phasing is not sensitive to the variation of $q_{95}$ for the studied equilibrium. Here, during scan of $q_{95}$, $\beta_N=1.63$ is fixed.
References
$[1]$ Y.Q. Liu, et.al, Phy. Plasmas 24, 056111, (2017);
$[2]$ L.N. Zhou et al. Nucl. Fusion, 58,076025, (2018);
$[3]$ Y.Q.Liu et.al. Nucl. Fusion 51, 083002, (2011);
$[4]$ A. Kirk et al. Nucl. Fusion, 55,043011, (2015)Speaker: Dr G.Z. Hao (Southwestern Institute of physics)
-
-
-
A. Bhattacharjee(a), B. Allen(g), C.-S. Chang(a), H. Chen(f), Y. Chen(f), J. Cheng(f), E. D’Azevedo(b), P. Davis(e), J. Dominski(a), M. Dorf(c), M. Dorr(c), S. Ethier(a), A. Friedman(c), K. Germaschewski(h), R. Hager(a), A. Hakim(a), G. Hammett(a), J. Hittinger(c), S. Janhunen(d), F. Jenko(d), S. Klasky(b), S. Ku(a), R. Kube(a), L. LoDestro(c), N. Mandell(a), G. Merlo(d), A. Mollen(a), M. Parashar(c), J. Parker(c), S. Parker(f), F. Poli(a), L. Ricketson(c), A. Scheinberg(a), M. Shepherd(i), A. Siegel(g), S. Sreepathi(b), B. Sturdevant(a), E. Suchyta(b), P. Trivedi(a), G. Wilkie(a), and M. Wolf(b)
(a) Princeton Plasma Physics Laboratory, Princeton, NJ, (b) Oak Ridge National Laboratory, Oak Ridge, TN, (c) Lawrence Livermore National Laboratory, Livermore, CA, (d) Institute for Fusion Studies, University of Texas at Austin, Austin, TX and Max Planck Institute for Plasma Physics, Garching, (e) Rutgers University, New Brunswick, NJ, (f) University of Colorado, Boulder, CO (g) Argonne National Laboratory, Chicago, IL
(h)University of New Hampshire, Durham, NH (i) Rensselaer Polytechnic Institute, Troy, NYWhole Device Modeling (WDM) is generally described as assembling physics models that provide an integrated simulation of the plasma. All components that describe a magnetic confinement device, from macroscopic equilibrium to micro-turbulence and control systems, are included in WDM, which describes the evolution of a plasma discharge from start-up to termination. Economical and safe operation of burning plasma devices requires predictive WDM with a confidence level established by validation and uncertainty quantification. Simulations covering the whole device, while certainly not a substitute for experiments, are much more cost-effective than building multiple billion-dollar facilities to test new ideas or concepts, similar to how aircraft manufacturers used simulations to reduce the number of physical wings they needed to build in designing superior aircraft (1).
The High-Fidelity Whole Device Model of Magnetically Confined Fusion Plasma is an application (hereafter referred to as WDMApp) (2,3) in the DOE Exascale Computing Project (ECP). The ECP is a DOE 413.3b project---the largest in the DOE Office of Science--- and is governed by the same rigorous rules of operation as major experimental facilities. The ultimate problem target of the project is the high-fidelity simulation of whole device burning plasmas applicable to an advanced tokamak regime (specifically, an ITER steady-state plasmas with ten-fold energy gain), integrating the effects of energetic particles, plasma-material interactions, heating, and current drive. The most important step in this project, and one that involves the highest risk, is the coupling of two existing, well-established, extreme-scale gyrokinetic codes – the GENE continuum code for the core plasma, and the XGC particle-in-cell (PIC) code for the boundary plasma. We have accomplished this challenging milestone for the first time in the magnetic fusion community. Fig. 1 demonstrates a coupled GENE-XGC in a WDMApp simulation for nonlinear ITG turbulence.These developments would not be possible without the remarkable advancements in edge turbulence simulation codes for which XGC is an exemplar, along with COGENT and GKEYLL.
COGENT is a continuum gyrokinetic code for edge plasmas (4, 5). The code is distinguished by its use of fourth-order conservative discretization and mapped multiblock grid technology to handle the geometric complexity of a tokamak edge. It solves full-f gyrokinetic equations for an arbitrary number of plasma species, which can also be coupled to a set of lower-dimensionality fluid equations in cases where a reduced fluid model is adopted to describe electrons or neutrals. The code offers a number of collision models, ranging from the simple Krook operator to the fully nonlinear Fokker-Plank operator, and includes an ad-hoc anomalous transport model that can be utilized for the case of 4D axisymmetric transport calculations. Recent applications of the COGENT code to the analysis of cross-separatrix edge plasma properties include (a) 4D calculations, which demonstrate the values of radial electric field and toroidal rotation comparable to those observed on the DIII-D facility, and (b) 5D calculations of ITG turbulence, which elucidate the role of magnetic shear stabilization in the X-point region.
The GKEYLL project (6) is developing a continuum gyrokinetic capability that can evolve the electromagnetic gyrokinetic equations in the tokamak edge. The code uses a Hamiltonian form of the full-f equations, and for electromagnetic terms, uses a symplectic formulation. A novel version of a high-order discontinuous Galerkin scheme is used, ensuring that total energy (particles plus fields) is conserved by the spatial discretization. GKEYLL has performed the first fully nonlinear full-f continuum simulations of electromagnetic gyrokinetics in the scrape-off layer (SOL), including sheath boundary conditions. In Fig. 2 we show a snapshot of the turbulence profiles when statistical steady-state has been obtained. Shown are density and temperature contours near the midplane. Intermittent blob-like structures are seen ejected from the source region as they propagate outwards. Comparisons with electrostatic simulations show that the turbulence is larger amplitude and much more intermittent in the electromagnetic case. Also shown are magnetic field lines between the top and bottom divertor plate being stretched by blobs. Full details of the scheme and detailed description of the results are described in (7).References
(1) “Case Study: Boeing Catches a lift with High Performance Computing,” Report by Council on Competitiveness, 2009
(2) G. Merlo, J. Dominski, A. Bhattacharjee, C.-S. Chang, F. Jenko, S. Ku, E. Lanti, and S. Parker, Phys. Plasmas 25, 062308 (2018)
(3) J. Dominski, S.-H. Ku, C.-S. Chang, J. Choi, E. Suchyta, S. Parker, S. Klasky, and A. Bhattacharjee, Phys. Plasmas 25, 072308 (2018)
(4) M. Dorf and M. Dorr, Contr. Plasma Phys. 58, 434 (2018)
(5) M. Dorf and M. Dorr, Contr. Plasma Phys. (2020); available online DOI: 10.1002/ctpp.201900113
(6) A. Hakim, N. Mandell, T. Barnard, M. Francisquez, G. Hammett, and E. Shi, Continuum Electromagnetic Gyrokinetic Simulations of Turbulence in the Tokamak Scrape-Off Layer and Laboratory Devices, submitted to Phys. Plasmas. (2020)
(7) N. Mandell, A. Hakim, G. Hammett, and M. Francisquez, J. Plasma Physics 86, 149 (2020)Speaker: Mr Amitava Bhattacharjee -
The TCV tokamak continues to leverage its unique shaping capabilities, flexible heating systems and modern control system to address critical issues in preparation for ITER and a fusion power plant. For the 2019-20 campaign its configurational flexibility has been enhanced with the installation of divertor gas baffles and its diagnostic capabilities with an extensive set of upgrades. Experiments are performed in part by topical teams, under the auspices of the EUROfusion medium-size tokamak programme and local teams at the Swiss Plasma Center together with international collaborators, resulting in a rich and focused scientific programme.
Auxiliary heating is provided by NBI and ECRH. An improved acceleration grid for the NBI system reduced power losses in the duct allowing for a 2.5x increase in injected energy with an injected power up to 1.3MW. The legacy ECRH system of 1.4MW from two 83GHz gyrotrons (X2) and 0.9MW from two 118GHz gyrotrons (X3) was enhanced by a 1MW dual frequency (84/126GHZ) gyrotron for X2 or X3 heating. The new gyrotron performs as designed validating the numerical models used for its development.
The most conspicuous upgrade was the installation of removable gas baffles that separate the vessel into main and divertor chambers, Fig. 1. The baffles seek to increase the divertor neutral pressure and thereby facilitate the extrapolation to future devices, such as ITER that will rely upon operation with high divertor neutral pressure [1]. They allow for various divertor and limited configurations. Experimentally, Ohmically heated diverted discharges confirm SOLPS-ITER predictions with up to a 5x increase in divertor neutral pressure. Optional fuelling into the divertor or the main chamber can disentangle the effects of fuelling rate, divertor or main chamber neutral pressure, $p_\mathrm{n,div/main}$, and plasma density, $n_\mathrm{e}$.
As predicted, increasing $p_\mathrm{n,div}$ in the baffled divertor facilitates access to detachment, which commences at ~30% lower $n_\mathrm{e}$. Scanning the plasma plugging by displacing the X-point with respect to the baffles indicates that the installed divertor closure may be close to optimal. Experiments also provide further evidence that the onset of detachment is determined by $p_\mathrm{n,div}$ rather than $n_\mathrm{e}$. These experiments ultimately seek to validate edge models. Reversing the toroidal field direction reveals changes in the target currents and the formation of a potential well below the X-point in reverse field as predicted by SOLPS-ITER including drifts.
All previously obtained alternative configurations were achieved in the baffled divertor with detailed investigations first focusing on the Super-X divertor [1]. Specific configurations designed to disentangle the effects of a large target radius, $R_\mathrm{t}$, and the angle between the divertor leg and the target surface confirm an expected strong dependence of the detachment onset on the angle, whereas a predicted dependence on $R_\mathrm{t}$ remains elusive.
Two new gas-puff imaging (GPI) systems, diagnosing the X-point region and the outboard midplane, have greatly increased the ability to investigate scrape-off layer (SOL) transport. Characterisation of the SOL turbulence was also extended into H-mode, linking $p_\mathrm{n,div}$ to the formation of the density shoulder [2].
H-mode studies were facilitated by an apparent reduction in the power threshold, $P_\mathrm{LH}$, for the baffled divertor. Here, the inter-ELM target temperatures were lower and nitrogen seeding led to detachment. At the H-mode density limit, a full MARFE develops as a dense strongly radiating region at the X-point that subsequently moves up the HFS edge.
Particular attention was dedicated to the pedestal in type-I ELMy H-modes, where the baffles lead to significantly higher $T_\mathrm{e,ped}$ and, hence, higher $p_\mathrm{e,ped}$, Fig. 2. The pedestal degrades with fuelling that increases $n_\mathrm{e,sep}$, but decreases $T_\mathrm{e,ped}$, consistent with previous findings. The role of $n_\mathrm{e,sep}$ in the pedestal is further highlighted in discharges with a range of $R_\mathrm{t}$ that require different fuelling rates to obtain the same $n_\mathrm{e,sep}$, but then display the same pedestal characteristics, highlighting alternative divertors’ weak effect on pedestal and core properties.
In a continued effort to extrapolate ELMy H-mode performance to the ITER baseline (IBL) scenario, NBI and X3 heated H-modes succeeded in matching the ITER targets of $\kappa$=1.7, $\delta$=0.4, $\beta_\mathrm{N}$=1.8 and $q_{95}$=3.0 whilst retaining good confinement ($H_\mathrm{98y2}$~1) [3]. The Greenwald fraction reached 0.6, but at those densities X2 ECH absorption becomes unreliable and ELM triggered NTMs were only avoided by lowering $I_\mathrm{P}$ with a stationary demonstration at $q_\mathrm{95}$=3.6.
In the quest to ELM-free regimes, negative triangularity (NT), Fig. 1 (c), is confirmed as an attractive scenario with an L-mode confinement matching that of H-modes for positive triangularity. NBI heating extended the operating space of NT plasmas to the IBL value of $\beta_\mathrm{N}$. However, similarly to the low-$q_{95}$ IBL in TCV, these discharges are prone to NTMs, but inaccessible to X2 ECH control. Reciprocating probe plunges past the LCFS of Ohmic NT plasmas confirm a reduced turbulent E×B flux extending from the core into the plasma edge. The measurements are corroborated by measurements from the new mid-plane GPI system.
TCV continues to address critical aspects of the discharge evolution that may limit plasma performance or even pose a danger to components in ITER and future power plants. This includes NTMs, with a successful validation of a new analytical model of the classical island stability, which will facilitate NTM pre-emption in future devices. Enabled by NBI, fast-ion studies are gaining prominence with the development of robust scenarios that display rich, fast-ion driven, MHD spectra and the commissioning of a fast-ion loss detector. Further experiments aim at understanding and controlling runaway electrons (REs) [4], whether created at low density or by mitigated disruptions, as they may cause severe damage in larger devices. Filtered imaging for multiple wavelength ranges provided the first measurements of synchrotron radiation on TCV, Fig. 3, revealing information about location, size and energy distribution of REs and even allowing to detect pre-disruption seed distributions. RE scenarios were extended from Ne and Ar to He, Kr and Xe injection, and to NT and diverted configurations, increasing the space for model validation. In addition, strategies to purge the impurities after the RE beam formation with further $\mathrm{D}_2$ injection are being explored.
TCV also continues to employ its flexible digital control system to enhance available control solutions [5]. With a view to future long-pulse tokamak discharges, a generic plasma control framework has been developed, implemented and applied to avoid density limit disruptions by controlling the NBI power based upon an estimated proximity to the disruptive boundary. An ability to re-assign EC sources to $\beta_\mathrm{N}$ or NTM control was demonstrated. Plasma exhaust control, for future reactors, was explored using an estimate of the C-III radiation profile along the divertor leg, indicative of the local $T_\mathrm{e}$, with a feedback control of the distance of the radiation from the target to the X-point demonstrated using gas injection as actuator for both L- and H-mode scenarios.
Various short baffled and un-baffled campaigns are planned for 2020 highlighting yet another dimension in TCV’s signature flexibility.[1] C. Theiler, et al., this conference, [2] N. Vianello, et al., this conference, [3] O. Sauter, et al., this conference, [4] G. Papp, et al., this conference, [5] F. Felici, et al., this conference.
Speaker: Holger Reimerdes (Ecole Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC)) -
The ADITYA Upgrade (ADITYA-U) is a medium sized (R0 = 75 cm, a= 25 cm) tokamak having toroidal graphite limiter, configured to attain shaped-plasma operations with an open divertor in single and double-null configurations [1]. The foremost objective of ADITYA-U is to prepare the physics and the technological base for future larger tokamaks by expanding the ADITYA-U operating space and by performing dedicated experiments for validation of physics models. Since the 2018 IAEA-FEC Conference, ADITYA-U operations have been mainly devoted on realizing the plasma parameters close to the design parameters of circular plasmas in limiter configuration and also the initiation of shaped plasma operation. Emphasis has been given to novel experiments on runaway electrons (REs) and disruption control in the ADITYA-U [2]. Furthermore, experiments on radiative-improved modes using Ne, Ar gas injection, modulation of MHD modes [3] and edge turbulence using periodic fuel gas-puffs, density dependence of plasma toroidal rotation reversal [4], fuelling using SMBI etc.
For the typical discharges in ADITYA-U, in absence of any strong pre-ionization, the gas breakdown and successful plasma start-up is normally achieved with peak loop voltages of ~ 18 – 20 V (Electric field ~ 4.5 V/m). 42 GHz ECR [5] assisted low loop voltage (~10 – 12 V, Electric field ~2.1 V/m) start-up was successfully achieved with wave launched in fundamental O-mode from low field side. The toroidal magnetic field is ~ 1.4 T. Plasma discharges having plasma current ~ 170 kA, plasma duration ~ 330 ms, chord-averaged electron density ~ 2 – 6 x 10^19 m^-3 and central electron temperature ~ 300 – 500 eV has been achieved. The time evolution of typical high current, longer duration discharge of ADITYA-U is shown in Figure 1 and the overall progress of plasma current and duration enhancement during the year 2018-2019 are shown in Figure 2(a) and 2(b) respectively. Repeated cycles of vacuum vessel baking up to 135° C, followed by extensive wall conditioning using novel techniques [6] along with lithium coating [7] resulted in substantial reduction in partial pressures of various mass species and achievement of lower base vacuum of ~ 6 x 10^-9 Torr. Successful recovery of volt–sec along with adequate control of real-time horizontal plasma position and multiple gas puffing led to the achievement of longer discharge duration discharges in ADITYA-U.
Over the last two years, significant progresses have been achieved in ADITYA-U experimental research including, 42GHz ECR assisted low loop voltage start-up and heating experiments, electromagnetically driven pellet impurity injector for injecting micron size particles at high velocity (~220m/s) to understand disruption mitigation, wall conditioning by using different techniques of lithium coating, formation of runaway beam and its avoidance, Neon and Argon impurity injection for radiative improved modes, suppression of electrostatic fluctuations using hydrogen gas puff and its correlation with hard X-rays, effect of positive edge electrode biasing on drift tearing modes and runaway transport and effect of SMBI with edge safety factor (qedge), toroidal rotation reversal threshold studies etc. Dependence of current quench time (CQT) on qedge during disruptions and relation of CQT on the prevailing MHD activities prior to disruption has been studied in detail. Discharges with low qedge showed a high CQT compared to those observed at high qedge. Exploring the RE formation mechanisms, controlled RE generation experiments have been carried out by lowering the plasma density and adjusting the vertical magnetic field as shown in Figure 3 (#33061).
Fast visible imaging video camera, used for 2D tangential viewing, captured images which showed features of the RE beam formation with high spatial and temporal resolution as shown in Figure 4. This RE beam generation in controlled fashion is very useful in studying the mitigation techniques of REs using different techniques as the prevention of such RE beam is of a vital importance in future tokamaks, especially in the ITER, because of its potential danger to the plasma facing components.
In another significant RE experiment carried out in ADITYA-U, correlated suppression of RE loss, evident from hard X-rays intensity measurements, has been observed with suppression of edge electrostatic fluctuations in discharges where magnetic fluctuation amplitudes are not sufficient for affecting REs. Multiple periodic gas puff, which are used for plasma fuelling, suppresses the electrostatic fluctuations in the floating potential as well as the density fluctuations (measured with Langmuir probes) in the edge region. Figure 5 shows the suppression of edge turbulence due to gas puff. This is also observed clearly in the spectrogram of density fluctuations. The hard X-rays flux is also seen to be highly correlated with the edge turbulence. The decrease in edge turbulence is accompanied by a decrease in HXR flux. This observation of correlated effect of electrostatic fluctuation on RE transport throws light on new mechanism of RE loss and may be exploited to design novel RE mitigation methods.
In another interesting experiment, the REs are confined by applying a voltage to an electrode placed at the edge region of the plasma prior to the disruption. Figure 6 shows the multiple pulses of biasing applied to an electrode placed ~ 2.5 cm inside the LCFS for the shot #33336. The last pulse of biasing coincides with the plasma disruption as shown in Figure 7, describes that in presence of biasing during the disruption, the HXR intensity as well as HXR flux persists for ~10 ms even after the termination of the plasma current. A possible mechanism of confinement is the Er x BΦ motion of REs after the disruption, where the Er is generated by electrode biasing.
For the first time, ADITYA-U has experimentally demonstrated the use of electromagnetically driven payloads for particle injection into the tokamak plasmas for disruption mitigation studies. The impurity particles reached the core of the plasma within ~ 1.25 ms and causes fast termination of plasma current and radiate the whole plasma energy in ~ < 2 ms. Furthermore, the preliminary experiments related to plasma shaping by charging the divertor coils during plasma current plateau, confinement improvement with ion cyclotron resonance (ICRH) assisted auxiliary heating and deuterium injection is undergoing and the results of the same will be presented. This paper summarizes the experimental research of ADITYA-U tokamak in the key areas of thermo-nuclear fusion over the last two years.References
[1] Tanna R. et al 2018 Plasma production and preliminary results from the ADITYA Upgrade tokamak Plasma Sci. Technol. 20 074002.
[2] Tanna R. et al 2019 Overview of operation and experiments in the ADITYA-U tokamak Nucl. Fusion 59 112006.
[3] Raj Harshita et al 2020 Effect of periodic gas-puff on drift-tearing modes in ADITYA/ADITYA-U tokamak discharges Nucl. Fusion 60 036012.
[4] Shukla G. et al 2019 Observations of toroidal plasma rotation reversal in the ADITYA-U tokamak Nucl. Fusion 59 106049.
[5] Shukla B.K. et al 2019 Commissioning of Electron Cyclotron Resonance Heating (ECRH) system on tokamak ADITYA-U Fusion Eng. Des. 146 2083 – 86.
[6] Jadeja K.A. et al 2019 Novel approach of pulsed-glow discharge wall conditioning in the ADITYA Upgrade tokamak Nucl. Fusion 59 086005.
[7] Jadeja K.A. et al Lithium Wall Conditioning Techniques in ADITYA-U Tokamak for Impurity and Fuel Control (In this conference).Speaker: RAKESH TANNA (Institute for Plasma Research) -
Experiments on ST40 towards burning plasma conditions
M. P. Gryaznevich for TE.Ltd team
Tokamak Energy Ltd, 173 Brook Drive, Milton Park, Abingdon, OX14 4SD, UK
e-mail: mikhail.gryaznevich@tokamakenergy.co.ukSpherical Tokamak (ST) path to Fusion has been proposed in R Stambaugh et al, Fus. Tech. 33 (1998) 1, and experiments on STs have already demonstrated feasibility of this approach. Advances in High Temperature Superconductor (HTS) technology (M Gryaznevich et al, Fus. Eng. & Design 88 (2013) 1593) allows significant increase in the Toroidal Field (TF) which was found to improve confinement in STs. The combination of the high beta, which has been achieved in STs, and high TF that can be produced by HTS TF magnets, opens a path to lower-volume fusion reactors, in accordance with the fusion power scaling ~ beta^2Bt^4V. High field spherical tokamak ST40 (design parameters: R=0.4-0.6m, R/a=1.6-1.8, Ipl=2MA, Bt=3T, k=2.5, pulse~1-2sec, 2MW NBI, 2MW ECRH/EBW, DD and DT operations) is the first prototype on this path and is now operating, Fig.1.
Plasma current > 0.5MA at 2T TF, electron and ion temperatures in a several-keV range produced using merging-compression formation, solenoid-assisted ramp-up and 1MW of 25kV NBH, and densities up to 2x10^20m^-3 have been achieved in the first experimental campaigns in 2018-2020. At the flat-top, measured Ti increases with TF, in agreement with observations on other STs. However, on ST40, at TF > 1 – 1.1T we observe sharp increase in Te, Ti and W(EFIT), Fig.2.
TF Cu magnet in ST40 and some PF coils are LN2 cooled and research is on-going on development of full-HTS magnets. HTS prototype magnet with 24.4 T (at 21 deg K) has been built, Fig.3, and now we are planning to increase the field. LN2 cooling of present Cu magnets, installation of the second 1MW 50kV beam and upgrades of power supplies are on-going and will allow increase of TF to 3T and the pulse duration from ~0.3sec (at present) to 1-2sec.
Experiments are carried out to study transport properties in ST at higher TF, higher heating power (up to 4MW), low collisionality, in aim to bring plasma parameters close to burning conditions. Transport simulations with ASTRA, NUBEAM and TSC codes have been performed to model ST40 parameters and to support the physics basis of the compact high field ST path to Fusion. We show that high confinement regimes with just collisional (neoclassical) transport can be expected even when only ohmically heated. In an auxiliary heating regime, we find a hot ion mode with Ti in the 10keV range to be achievable in ST40 with as low as 1MW of absorbed power, Fig.4.
Issues connected with specific features of the high field ST are discussed, i.e. limitations of applicability of confinement scalings for prediction of performance of ST40. However, we show that if the performance achieved on other spherical tokamaks can be extended to ST40 conditions, up to 1 MW of Fusion power can be expected in DT operations. Studies of fast ions and alpha particle transport, heating and current drive, torque deposition and momentum transport have been performed using ASCOT, NUBEAM, Monte Carlo code NFREYA and the Fokker - Planck code NFIFPC. Different NBI energies and launch geometries have been studied and optimised. The confinement of thermal alphas in ST40 3T/2MA scenario is studied with full orbit following (which is necessary because of the large, compared to the plasma size, alpha particle gyro radius). The first orbit losses are seen to be above 60% even in the high-performance scenario illustrating that the alpha confinement in a small device is very difficult even at the highest available fields and plasma currents. However, DT experiments on ST40 will provide useful information for verification of such simulations.
We are intensively working on the design of our next tokamak, ST-F1, with plasma volume ~0.5 of JET. This device is aimed to demonstrate Q>3, will have HTS magnets, tritium blanket, and the goal is to build it by 2025.
Speaker: Mr Mikhail Gryaznevich -
Spherical tokamak (ST) research in Japan 1 is being conducted as a nationally coordinated program of university-scale ST devices under the ST Research Coordination Subcommittee organized by National Institute for Fusion Science (NIFS). The roles of university ST research include: (1) unique and challenging research through creativity and innovation which might be considered too risky for large ST devices, (2) establishment of the scientific basis for achieving ultra-high beta and ultra-long pulse (Fig. 1),
(3) contribution to the scientific basis for practical and economically competitive fusion power, complementing the mainline tokamak research (JT-60SA, ITER, etc.), and (4) development and training of a future generation of world-leading tokamak scientists. Specific research topics include: (a) development of start-up, current drive, and control techniques without the use of the central solenoid (CS), (b) formation and sustainment of very high beta plasmas, and (c) demonstration of steady-state operation and the study of steady-state issues such as heat and particle control, divertor physics, and plasma-wall interaction.
(a)-1. Plasma current ($I_p$) start-up by RF waves: Electron cyclotron wave (ECW) at 2.45 GHz and 5 GHz are used to excite the electron Bernstein wave (EBW) via O-X-B mode conversion on LATE (Kyoto U.). Highly overdense ST plasmas (up to 7 times the plasma cutoff density) are formed when the fundamental EC resonance layer is located in the plasma core, and EBW is excited in the 1st frequency band ($\omega_{ce} < \omega < 2\omega_{ce}$). Whereas EBW in the 1st frequency band heats the bulk electrons, EBW in the 2nd frequency band is absorbed by high-energy electrons and drives $I_p$. Intermittent plasma ejections across the plasma boundary synchronized with poloidal field decrement were observed in highly overdense plasmas. Oscillations in the Alfven frequency range and potential increase were observed, suggesting the loss of high-energy electrons. The 28 GHz RF injection system on QUEST (Kyushu U.) can regulate wave polarization and parallel index of refraction ($N_{||}$) with a beam radius focused down to 50 mm. $I_p > 100\ kA$ was achieved by injecting X-mode with $N_{||} = 0.78$, assisted by poloidal field induction. The RF power is likely absorbed by energetic electrons. Electron temperature of up to 0.5 keV was obtained by injecting X-mode with $N_{||} = 0.1$, indicating that effective bulk heating is possible. In $I_p$ start-up experiments using the lower hybrid wave (LHW) on TST-2 (U. Tokyo), top-launch was found to be more efficient for Ip ramp-up than outboard-launch. A 2-dimensional phase space model explaining X-ray emission shows that LHW driven radial transport is the dominant loss mechanism of fast electrons, and higher density is preferable in the present situation [2]. This is the first clear demonstration of RF driven electron transport.
(a)-2. CS-less $I_p$ start-up by non-RF methods: In transient CHI, magnetic reconnection plays an important role in the formation of closed flux surfaces during $I_p$ start-up. Plasmoid-driven reconnection following the tearing instability of the elongated current sheet and associated ion heating in the presence of the toroidal guide field were investigated on HIST (U. Hyogo) [3]. Several small-scale plasmoids generated during the injection phase merge with each other to form one or two large-scale closed flux surfaces during the decay phase. Transient CHI is also investigated on QUEST (US-JA collaboration). Investigation of synergistic effects of electron beam injection and EBW current drive in overdense plasmas has begun on LATE [4].
(a)-3. Optimization of inductive plasma start-up: In low voltage inductive $I_p$ start-up with ECW pre-ionization on TST-2, application of a weak vertical field with positive decay index during breakdown was found to be beneficial at low pre-fill pressure and high ECW power. Application of ECW power extended the low pressure limit for breakdown as well as the high pressure limit for burn-through. An MHD equilibrium model with fast electron orbits taken into account, and a model to simulate electron diffusion in both velocity space and real space are being developed.
(b) Access to high temperature and/or beta regime: Reconnection heating for direct access to burning plasmas, being investigated on TS-3U, TS-4U, UTST (U. Tokyo), will be reported in Ref [5].
(c) Demonstration of steady state operation by high-temperature wall: A high-temperature wall plays an essential role in reducing wall-stored hydrogen and facilitates hydrogen recycling. A clear extension of pulse duration at the wall temperature of 473 K was observed on QUEST by water cooling, indicating that recycling can be controlled by wall temperature. During long duration discharges, a high concentration of neutral particles was achieved behind the bottom divertor plate [6].
Stabilization using helical field coils: Suppression of the oscillation and the outer displacement in the radial position was observed by applying the helical field to the tokamak plasma on TOKASTAR-2 (Nagoya U.), which is an ST-helical hybrid device equipped with parallelogram-shaped partial helical field coils [7].
ST research in Japan has produced many innovative results including (i) $I_p$ start-up by LHW (TST-2), EBW (LATE), ECW/EBW (QUEST), CHI (HIST, QUEST), electron beam (LATE); (ii) optimization of ECW-assisted inductive start-up and pre-ionization by AC operation of the Ohmic coil (TST-2); (iii) extension of ion heating by plasma merging (TS-3U, TS-4U, UTST); (iv) hydrogen recycling control with high-temperature wall (QUEST); and (v) radial position stabilization by superposed helical field (TOKASTAR-2).
1 Y. Takase et al., Nucl. Fusion 57, 102005 (2017).
[2] A. Ejiri et al., this conference; N. Tsujii et al., this conference.
[3] M. Nagata et al., this conference.
[4] H. Tanaka et al., this conference.
[5] Y. Ono et al., this conference.
[6] T. Onchi et al., this conference; K. Hanada et al., this conference.
[7] K. Yasuda et al., Plasma Fusion Res. 13, 3402072 (2018).Speaker: Yuichi Takase (University of Tokyo) -
The report provides an overview of the results obtained at the upgraded Globus-M2 spherical tokamak 1 since the last IAEA conference. The tokamak was designed to reach the toroidal magnetic field as high as BT =1 T and the plasma current Ip = 0.5 MA having a small plasma minor radius a = 0.22-0.23 m. Currently 80% of highest magnetic field and plasma current value are reached, so during the reported period the experiments were performed with the toroidal magnetic field up to 0.8 T and plasma current up to 0.4 MA. The plasma breakdown conditions were improved noticeably with regard to the Globus-M ones, 30% breakdown loop voltage decreasing was achieved. The discharge duration was increased due to higher central solenoid volt-second consumption. The plasma column magnetic configuration explored was the divertor lower null with the aspect ratio A = R/a = 1.5-1.6, triangularity up to δ~0.35 and elongation up to κ~2.2.
The first neutral beam heating experiments on Globus-M2 have demonstrated an increased efficiency, comparing with the Globus-M ones, at the same NBI parameters (deuterium beam with particle energy 28 keV and the heating power 0.8 MW). The electron and ion central plasma temperatures exceeded 1 keV at the central density as high as 1×10^20 m-3. The diamagnetically measured plasma thermal energy increased up to 10 kJ, which is nearly triple as high as in Globus-M (BT =0.4, Ip = 0.2 MA). NPA spectra demonstrating improved fast particle confinement are presented. The energy confinement time increased more than two times that is significantly higher than the IPB98(y,2) scaling predicts. The effect is due to the strong dependence of the energy confinement time on the toroidal magnetic field in accordance with the Globus-M experimental scaling that is found to be valid for a wider range of BT. The regression fit of the Globus-M/Globus-M2 data yields the following scaling for energy confinement time:
τE ~ Ip^(0.58)BT^(1.23)Pabs^(-0.66)ne^(0.63)
where Pabs is the absorbed heating power and ne is the line average density. The scaling confirms weak τE dependence on Ip that emphasizes the major role of BT on heat perpendicular transport in spherical tokamaks, Enhanced plasma parameters allowed us to obtain regimes with much lower collisionality. That make possible investigation of dependence of the normalized energy confinement time (BTτE) on collsionality (ν~ne/T^2) in the wide range of plasma collisionalities 0.018<ν< 0.23. This dependence turned out to be rather strong BTτE ~ ν^(-0.8) for a fixed values of safety factor q ~ BT/Ip, normalized ion gyroradius ρ ~ T^(0.5)/BT and parameter βT ~ W/BT^2. The power balance analysis carried out using ASTRA transport code indicates the reduction of both electron and ion heat diffusivity with collisionality decrease while the ion heat diffusivity remains near the neoclassical level.
Important results are related to non-inductive current drive. About 30% of the loop voltage drop was recorded during the NB injection, which indicates a noticeable amount of non-inductively (mainly bootstrap) driven current. For the first time in spherical tokamaks a non-inductively driven current was recorded during the launch of the electromagnetic waves of the lower hybrid (LH) range (2.45 GHz) with the help of toroidally oriented grill. The fraction of noninductively driven current has exceeded 30% in the discharge with the total current of 0.2 MA. The modelling results of the experimental data by means of the ASTRA transport code and Fast Ray Tracing Code incorporated to ASTRA 2 are presented.
Plasma scrape of layer (SOL) and divertor characteristics were investigated in new experimental conditions of enhanced magnetic field and plasma current. Heat and particle fluxes together with currents and potentials in SOL and divertor plate vicinity were measured with a divertor Langmuir probe array and movable Langmuir probe. The plasma parameters in SOL were also modelled with the fluid version of the SOLPS-ITER code. Currents and drifts were included in the simulations. Comparison of experimental and simulated heat flux power density decay length (λqt) in SOL with the well-known scalings is presented.
The study of Alfvén modes (АМ) was continued during the reported period. An increase in plasma parameters led to a change in the nature of AM and the expansion of their frequency spectrum (50–300 kHz). Together with the toroidal Alfvén eigenmodes (TAE), observed earlier on Globus-M, the so-called Alfvén cascades (AC or RSAE) were identified. Observation of ACs made it possible to apply the method of MHD spectroscopy to determine the evolution of qmin in a discharge. In experiments on current drive by the LH waves, modes with a frequency of about 1 MHz, excited by fast electrons, were detected. To study the spatial structure of AM, Doppler backscattering diagnostics was used [3] with application of a multi-channel microwave scheme. Using the neutral particle analyzer and a neutron detector, we studied the dependence of fast particle losses initiated by TAEs on the magnetic field and plasma current. It was shown that losses decrease significantly with increasing field and current, demonstrating dependence favorable for compact neutron sources.
Also presented are new diagnostics designed to fill in the missing data on plasma parameters and improve the quality of the simulation, such as: diagnostics Z eff, laser interferometer, charge-exchange resonance spectroscopy (CXRS), etc.
1. V.B. Minaev et al 2017 Nucl. Fusion 57 066047
2. A.D. Piliya, A.N. Saveliev, JET Joint Undertakin Abingdon, Oxfordshire, OX14 3EA, 1998
3. V.V. Bulanin et al 2019 Tech. Phys. Lett., v.45, 11 p.p. 1107-1110Speaker: Yury Petrov (Ioffe Institute) -
The mission of the spherical tokamak NSTX-U is to advance the physics basis and technical solutions required for optimizing the configuration of next-step tokamak fusion devices, and to advance the development of the ST concept towards a compact, low-cost Pilot Plant 1. NSTX-U will operate at up to 2 MA and 1 T with up to 10 MW of Neutral Beam Injection (NBI) power for 5 seconds and 4 MW of High Harmonic Fast Wave (HHFW) power. NSTX-U has three main objectives: to explore confinement and stability at low aspect ratio and high beta at low collisionality, to develop the physics understanding and control tools to ramp-up and sustain high performance plasmas in a fully-non-inductive regime, and to develop and evaluate conventional and innovative power and particle handling techniques to optimize plasma exhaust in high performance scenarios. Following the initial 2016 NSTX-U run campaign, analysis has continued during NSTX-U Recovery to address physics issues to develop understanding and capabilities that, once operation commences, will aid in achieving these three objectives.
Stability and 3D physics: The resistive DCON model for calculating tearing mode stability ($\Delta'$) has been developed, benchmarked against extended-MHD (M3D-C1) simulations, and used to identify regions in $q_{95}-\beta_N$ that are simultaneously stable to both 2/1 tearing modes (Fig.1) and n=1 ideal kink.
M3D-C1 has been extended to predict Te and Ti independently in the presence of impurities during a thermal quench 2. The simulations show a contraction of the current channel that is sensitive to temperature, and a fast stochastization of the B-field, highlighting the difficulty in cooling the plasma while avoiding a thermal quench using impurity injection. Continued analysis on 3D error fields have shown that misalignment of the toroidal field (TF) was the largest source of the error field on NSTX-U during initial operation in 2016, and these calculations, guided by constraints on the field line pitch at the divertor plates in order to mitigate potentially high heat fluxes, were used to drive the engineering tolerances for TF shift and tilt for NSTX-U Recovery [3]. Additional calculations show that misalignments in other PF coils lead to extended divertor footprints but which are contained within the divertor region designed to handle high heat fluxes [4]. Energetic particle (EP) physics: The phase-space-resolved reduced EP transport kick model has been extended to include non-Alfvénic low-frequency perturbations, reproducing observations of (i) large fast ion losses due to synergistic effects when TAE and fishbones/kink instabilities occur simultaneously, and (ii) enhanced fast ion loss due to NTMs when the island width exceeds a threshold [5]. The stability and scaling of global Alfven eigenmodes (GAEs), previously correlated with central Te flattening in NSTX, has been predicted using hybrid MHD/kinetic-fast-ion simulations [6] and also newly derived analytic instability conditions [7], revealing a previously unidentified instability regime necessary to explain observed GAE excitation and stabilization. The chirping and avalanche behavior of Alfven eigemodes that can influence fast-ion losses in NSTX-U (and burning plasma $\alpha$ losses) has been predicted using a guiding center code with a delta-f formalism [8]. New analysis from NSTX and NSTX-U data has provided a detailed picture of ion cyclotron emission (ICE), being considered as a possible diagnostic of confined $\alpha$’s in burning plasma experiments [9]. A self-consistent resonance-broadened quasi-linear (QL) model has been developed for relaxation of fast ion distribution function by Alfvenic modes [10].
Transport physics: Analysis in moderate $\beta$ NSTX scenarios using gyrokinetic simulations coupled with a novel “synthetic diagnostic” identify conditions where both electron thermal transport and turbulence measured by high-k microwave scattering are explained entirely by short-wavelength electron-scale ETG turbulence [11]. Gyrokinetic analysis in high-$\beta_{pol}$ scenarios envisioned for high non-inductive fraction operation indicates the deep-core profiles (with relatively flat Te) sit very near KBM (or EPM) limits when including only thermal ions (or thermal + fast ion species), suggesting core profiles may ultimately be constrained by $\nabla p$-limited ballooning modes. Analysis of enhanced Pedestal (EP) H-modes demonstrates that this high confinement (H98$\leq$1.8), wide-pedestal ($\Delta\psi_{N,ped}\leq0.4$), ELM-free regime is accessed at low edge ion collisionality, e.g. via reduced wall recycling with lithium wall coatings. While MTM, TEM and ETG instabilities predicted in this region may account for electron thermal losses, it is hypothesized that peeling-ballooning and near-marginal KBM instabilities (predicted via MHD and gyrokinetic simulations, respectively, Fig. 2) may enhance edge particle transport during the evolution to the EPH phase that helps to sustain the low-density, low-collisionality state.
RF physics: A new 2D full wave code (FW2D) has been updated to predict the sensitivity of SOL losses to variations in the realistic boundary shape for high harmonic fast wave (HHFW) heating [12]. Losses are further predicted to be minimized when the SOL density is near the critical density for fast wave cutoff, as found in experiments. Calculations of HHFW deposition in the presence of NBI using AORSA identify a competition between electron and fast ions absorption [13]. Additional simulations show that a sufficient concentration of H+ minority species could open up new HHFW heating scenarios in NSTX-U without NBI.
Scenarios & Control: A unified, physics-based reduced model for direct inductive startup, which computes the timing of plasma initiation and duration needed for plasma density buildup, has been developed for both NSTX-U and MAST-U [14]. Physics-based control-oriented models have been used to develop advanced-control approaches to determine neutral-beam current-drive requirements and evolutions to track prescribed current profiles in closed-loop [15], as well as to iteratively optimize the access to high-performance scenarios [16]. Aided by this is the development of a neural network-based description of beam heating and current drive profiles, well suited for rapid calculations and real-time application [17]. Development of a physics-based algorithm for closed loop feedback control of snowflake divertor configurations will provide real-time tracking and control capabilities [18].
The NSTX-U Recovery planning is ongoing, with NSTX-U targeting resumption of operations in 2021 [19].
This work was supported by US Department of Energy Contract No. DE-AC02- 09CH11466.
References
1 Menard, J.E., et al., this conference.
2 Ferraro, N.M., et al., Nucl. Fusion 59 016001 (2019).
[3] Ferraro, N.M., et al., Nucl. Fusion 59 086021 (2019).
[4] Munaretto, S., et al., Nucl. Fusion 59 076039 (2019).
[5] Podesta, M, et al., this conference.
[6] Belova, E., et al., this conference.
[7] J. B. Lestz et al., Physics of Plasmas 27, 022512 & 022513 (2020);
[8] White, R., et al., this conference.
[9] Fredrickson, E., et al., this conference.
[10] Gorelenkov, et al., this conference.
[11] Ruiz Ruiz, J., et al., Plasma Phys. Cont. Fusion 61 115015 (2019)
[12] Kim, E.-H., et al., Phys. Plasmas 26 062501 (2019)
[13] Bertelli, N., et al., Nucl. Fusion 59 086006 (2019)
[14] Battaglia, D.J., et al., Nucl. Fusion 59 126016 (2019)
[15] Ilhan, Z.O., et al., Fusion Eng. and Design 146 555 (2019)
[16] Wehner, W.P., et al., Fusion Eng. and Design 146 547 (2019)
[17] Boyer, M.D., et al., this conference.
[18] Vail, P.J., et al., Plasma Phys. Cont. Fusion 61 035005 (2019)
[19] Gerhardt, S.P., et al., this conference.Speaker: Walter Guttenfelder (Princeton Plasma Physics Laboratory)
-
-
-
Repetitive buildup and collapse of the edge confinement barrier (called pedestal) in H-mode plasmas can seriously damage the plasma-facing components in tokamak fusion devices [1]. Hence, an accurate understanding of the underlying mechanism of the collapse is essential for the safe operation of fusion devices. We have proposed the formation of a solitary perturbation (SP) in the edge as a candidate trigger of the pedestal collapse, based on our observation of SP within $\sim100~\mu \mathrm{s}$ before the onset of the collapse in the KSTAR [2]. Figure 1 shows an example of SP detected by a toroidal array of Mirnov coils (MCs). Low-pass filtering ($<50~k\mathrm{Hz}$) of the MC signals reveals a temporally and spatially localized perturbation of high amplitude in proximity to the pedestal collapse, as highlighted in blue color in figure 1(d), which is not easily discernible in raw signals. The SP has distinct features in the spatial structure, rotational speed, and mode pitch, distinguished from the edge localized mode (ELM), which is widely known to cause the pedestal collapse.
However, a statistical study on extensive data would be a prerequisite before ascertaining the SP as the collapse trigger. We have constructed a machine learning (ML) model based on a convolutional deep neural network architecture to automate check on the concurrency of the pedestal collapse and the SP. The ML model takes sequential signals detected from 19 toroidal MCs as input and predicts whether each temporal frame corresponds to an SP. As it is required to identify the temporally shifted solitary patterns across the MCs, we adopted 2D convolution operations that capture local context patterns within the input matrix. We trained the network in a supervised manner on a training set consisting of signals with manually annotated SP locations and synthetic signals that imitate pedestal collapses without SPs. The inclusion of the synthetic data prevents the network from falsely recognizing the signals from the collapse as SPs due to their consistent co-occurrences.
The trained model achieves $95.9~\%$ of per-frame accuracy on the test set with the final score threshold of $0.7$. The average precision (AP) of the model is about $0.88$. The AP measures the reliability of the per-frame prediction as a weighted mean of precisions in the precision-recall curve, where the precision and recall describe the ratio of correct SP predictions to all positive predictions and to all true SP cases, respectively. When applied to data sequences, our trained model achieves $100~\%$ of per-sequence accuracy for all the test data sequences. Here, the per-sequence accuracy measures the existence of SP based on the overall prediction of the model in a sequence. We further demonstrate the reliability of the model by visualizing the discriminative parts of the input signals that the model recognizes.
Finally, we performed a statistical analysis on the concurrence of pedestal collapse and SP using the developed model. The data set used for the analysis consists of a total of 16244 sequences, and each sequence is labeled with the existences of pedestal collapse and SP. For the collapse labeling criterion, we manually inspected the occurrences of the collapse in the sequences and found that all clear pedestal collapses have the maximum H-alpha intensity, $H_\mathrm{\alpha}^{max}>1×10^{21}$. For the SP labeling, we applied the ML model to the individual sequences and obtained the sequence prediction $y_\mathrm{s}$, which is the prediction score of the model accumulated the range of $100~\mu \mathrm{s}$ centered on the midpoint of three consecutive positive predictions in a sequence. We chose $y_\mathrm{s}>60$ as the criterion for the presence of SP based on our cross-validation with the training data set. Figure 2 plots the values of $H_\mathrm{\alpha}^{max}$ and $y_s$ to visualize the correlation between the two variables. We identify four regions, and a table in figure 2 presents the number of sequences in each region. Region 1 ascertains that the pedestal collapse of high amplitude accompanies SP. Region 2 contains one minor exception, but it is not against the overall trend because the SP prediction score $y_\mathrm{s}$ is near the decision boundary. Region 3 and 4 represent the sequences without a clear signature of pedestal collapse. Most of these sequences have zero $y_\mathrm{s}$ and are located in Region 3. It is worth mentioning that many sequences in region 4 correspond to small pedestal collapse. Therefore, the most conservative interpretation of the statistical analysis is that a large pedestal collapse always coexists with SP.
Our results strongly support the SP as a pedestal collapse trigger. This conclusion suggests that studying the effect of SP on the collapse is essential to understand and control the edge pedestal collapse. Besides, our result is a meaningful example that illustrates the possibility of applying deep learning techniques for expedient analysis of complex MHD phenomena.This work is supported by the Center for WISET grant funded by the MIST under the Program for Returners into R&D and NRF of Korea under grant no. NRF-2019M1A7A1A03088456.
References:
[1] A. Loarte et al, Plasma. Phys. Controlled Fusion 44, 1815 (2002)
[2] J. E. Lee et al, Sci. Rep. 7, 45075 (2017).Speaker: Jieun LEE (Pohang University of Science and Technology) -
For future fusion devices including ITER and DEMO, it is crucial to handle the power flux to the divertor targets in order not to exceed the steady-state material limit, q$_{t,⊥}$≤ 10 Wm$^{-2}$ [Ref. 1]. One of the viable effective methods to achieve it under development is divertor detachment via impurity gas seeding or deuterium injection in the divertor region in order to dissipate the excessive power flux. In KSTAR, L-mode detachment studies have been performed extensively, and the in/out divertor asymmetry was found where the detachment occurred at a lower upstream electron density at the outer target than at the inner target [Ref. 2], as opposed to the observations in other tokamaks. In a recent dedicated experiment, the first step toward divertor detachment in H-mode was successfully done by pure deuterium fueling and the study on the asymmetric characteristics of detachment has been extended to more complicated H-mode plasmas. This paper reports on the results of the experiment and numerical modelling using the SOLPS-ITER code to get better understanding of physical processes behind divertor detachment.
The discharges in the H-mode detachment study had a plasma current of I$_{p}$ = 0.7 MA and a forward toroidal magnetic field of B$_{T}$ = 2.0 T with 8 s flattop (1.5 s – 9.5 s). Shot traces of plasma parameters for the experiment (#22849) are shown in figure 1. The total heating power was about 3 MW by neutral beam injection. The outer strike point was swept for 2 – 4 cm during the current flattop phase in order to obtain target ion saturation current (j$_{sat}$) profiles by the tile-embedded Langmuir probe array. D$_{\alpha}$ line emissivity near the outer target (Fig. 1(e)) gradually increased with some periodic fluctuations which are due to the strike point sweeping. The deuterium fueling rate was linearly increased, such that the line averaged density $\bar{n}_{e}$ kept increasing up to 6x10$^{19}$ m$^{-3}$ (f$_{GW}$ ≈ 0.7) during the discharge before the H-L back transition. Consequently, the divertor regime was changed from low recycling to high recycling and then detached states. A partial detachment was achieved on both inner and outer divertor targets during the inter-ELM phase and the similar behavior to L-mode was found in terms of the in/out divertor asymmetry. The peak particle flux at the outer target reduced by a factor of 4 compared to the maximum value shown in the attached divertor regime, although there was no significant reduction at the inner target as illustrated in figure 2. This asymmetry that was generally found in the detachment experiment in KSTAR was explained by the neutral particle dynamics and geometrical effect through the SOLPS-ITER modelling without particle drifts [Ref. 2].
Meanwhile, drifts are considered as one of the driving factors of the asymmetry along with the geometry. Thus, we have performed the SOLPS simulations of L-mode plasmas including drifts to check how the drifts influence the asymmetry. The modeling setup is basically same as in [Ref. 2] but with diamagnetic drift and ExB drift turned on. The multiplying factor to the drift terms was scanned in parallel to the separatrix density scan. According to the preliminary result, the drifts modify the target profiles including particle flux density, electron density and temperature, and their effects are larger at the inner target than at the outer target.
Simulations using the SOLPS-ITER code have been performed for the H-mode discharges with pure deuterium injection. The input power into the simulation domain was 2.4 MW by taking into account that 20% of the heating power is radiated inside the core plasma. The fueling rate was scanned from 2.5x10$^{21}$ s$^{-1}$ to 1.0x10$^{22}$ s$^{-1}$ to mimic the density ramp experiment. Figure 3 presents the parallel particle flux density onto the targets obtained from the modelling, and it indicates that the modelling successfully reproduces the anomalous asymmetry as observed in the experiment. Larger reduction of the particle flux is shown at the outer target than at the inner target. However, contrary to the experiment, the attached heat flux to the outer target well exceeds 1 MWm$^{-2}$ which is rarely seen from the divertor infrared television (IRTV) system in KSTAR. The analysis on the total radiation profiles in the divertor region was also demonstrated. As the separatrix density increased, the maximum radiation front moved upstream from the targets to the X-point and the radiation finally became concentrated at the X-point in the detached state. Refinement of the radial profiles of the perpendicular transport coefficients is underway which will give better quantitative agreement between the simulation results and the experimental data.
Acknowledgements
This work was supported by the National R&D Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2019M1A7A1A03087560).References:
(1) A. Loarte $\textit{et al}$, Nucl. Fusion 47, S203-63 (2007).
(2) J.-S. Park $\textit{et al}$, Nucl. Fusion 58, 126033 (2018).Speaker: Junghoo Hwang (Korea Advanced Institute of Science and Technology) -
Simultaneous control of large ELMs and divertor heat load in a metal wall environment is crucial for steady-state operation of a tokamak fusion reactor. A new scenario for ELM suppression compatible with radiative divertor has been demonstrated, for the first time, in the EAST superconducting tokamak. An n = 1 mode [FIG 1(f)] along with its harmonics, initiating from the oscillation of a radiation belt in the high-field-side SOL near the X point [FIG 1(h)&(i)], is excited during impurity seeding with CD4 from the upper divertor near the outer strike point in the H-mode plasmas at a sufficiently high impurity concentration. ELM suppression [FIG 1(b)] has been achieved robustly with the presence of this mode in a wide q95 range from 4.5 to 6.5 (Ip = 400-600 kA and Bt = 2.25, 2.47 T) and a wide heating power range with source power ranging from 3 MW up to 9 MW. Along with ELM suppression, divertor detachment has been achieved with target electron temperature Tet < 10 eV [FIG 1(d)] and significant reduction in ion saturation current at the inner target plate. The plasma stored energy keeps nearly constant [FIG 1(c)], indicating that good energy confinement is maintained. Active feedback control of either Tet or divertor radiation with impurity seeding has been demonstrated in this regime. Furthermore, high-Z impurity concentration, as indicated by tungsten (W) 27 line emission intensity from the plasma core [FIG 1(e)], is suppressed and maintained at a low level when the mode appears. The estimated W impurity confinement time is < 200 ms, which is significantly shorter than that in ELM-free H-mode plasmas, but close to that in typical ELMy H-mode or EDA H-mode plasmas on EAST. These may suggest that sufficient particle transport is driven across the pedestal so that a quiescent H-mode with a low impurity concentration can be maintained. The n = 1 mode drives particle transport as evidenced by oscillations in divertor Dalpha and ion saturation current signals and an increased pedestal foot density, which may be responsible for the sustainment of ELM suppression. However, the dominant pedestal fluctuations, Edge Coherent Mode (ECM) [1], in the ELMy phase are significantly reduced during CD4 seeding, as shown by polarimeter-interferometer measurements [FIG 1(g)], suggesting that the quiescent pedestal is not sustained by the ECM-driven transport.
This n = 1 mode has been observed only in H-mode and disappears as the heating power decreases down to near the L-H transition threshold power. Excessive impurity seeding also leads to mode suppression. These may suggest that the excitation of this mode requires a sufficiently high local temperature or radial temperature gradient. A clear threshold in the carbon impurity concentration for the mode excitation has been identified, as shown in FIG 2. Shots #93373 & #93374 are two adjacent discharges with the same operation conditions except for the CD4 injection valve voltage, 3V for #93373 and 4V for #93374. The mode appears in the time window when the CVI line radiation intensity exceeds a threshold, as indicated by the magenta dashed line in FIG 2(b). In addition, statistical analysis indicates that the CVI intensity threshold increases with heating power.
In addition to CD4 puffing, a similar n = 1 mode and its harmonics have been observed in EAST with helium, lithium, boron, neon and argon seeding, given a sufficiently high impurity concentration, although the mode excitation with argon and neon appears to be more difficult, especially for neon. We have only observed this mode with neon seeding in very limited high-power cases. Furthermore, the mode excited by CD4 puffing exhibits a broader frequency band (deltaf/fpeak > 50%), as shown in FIG 1(f) and FIG 2(c)&(d), with weak magnetic oscillations, while in other cases the mode is more coherent (deltaf/fpeak < 30%) with stronger magnetic oscillations. The oscillation is mostly localized near the main X point. However, it can extend to the midplane when the mode is strong, and especially at lower Ip. In the latter case, the mode is also detectable near the secondary X point. In addition to the mode in the high-field side SOL near the X point, another n = 1 mode sometimes appears in the low-field side SOL between the outer target and the X point when the outer target is partially detached.
A model based on impurity-radiation-driven drift instability has been developed to explain the excitation mechanism of this n = 1 mode. It is noted that there is a negative-slope Te range in the impurity radiation loss function, as shown in FIG 3(a) for carbon. In this Te range, radiation loss increases with decreasing Te as plasma is cooled by impurity radiation, leading to a further lower Te. This gives rise to the so-called ‘radiation condensation instability’ as the impurity concentration is sufficiently high [2]. In a strongly magnetized toroidal plasma, it couples to drift waves, which makes it more easily unstable. In tokamak divertor region, a localized cold, high-density, radiating plasma region (a radiation belt as seen toroidally) usually forms in the high-field-side SOL near the X point with impurity seeding. Before the radiating region intrudes into the closed flux surfaces, forming a confinement-degrading ‘X-point MARFE’, it can be maintained below the X point without degrading the pedestal confinement. The impurity-radiation-driven drift instability is destabilized as the local Te in the radiating region is reduced down to the negative-slope Te range in the impurity radiation loss function, given a sufficiently high impurity concentration. For carbon, the Te range is roughly 7.6-20 eV in the EAST case [FIG 3(c)]. The model predicted mode frequency and excitation conditions are consistent with the experiments quite well. In general, many impurity species have this negative-slope Te range, such as helium, lithium, carbon, boron, neon and argon, but their Te ranges are different. Neon and argon have relatively high Te ranges, 34-110 eV for neon and 19-54 eV for argon, which are more difficult to access in EAST with relatively low plasma density and impurity concentration, but is readily accessible in future large devices, such as ITER, with higher Te and plasma density near the X point. The model suggests that the mode could be even stronger in future large devices with a much higher divertor plasma density, as the mode is mainly damped by parallel electron thermal conductivity, which will be significantly reduced at a higher density.
In summary, this new ELM control scenario discovered in EAST appears to be insensitive to impurity species, q95, Bt direction, heating power and plasma toroidal rotation, while being compatible with divertor detachment, thus offering a promising alternative solution to the control of both ELM-induced transient and steady-state divertor heat loads for long-pulse H-mode operations.
Acknowledgements
This work was supported by National Natural Science Foundation of China under Grant No. U19A20113, the Key Research Program of Frontier Sciences, CAS under Grant No. QYZDB-SSW-SLH001, the CASHIPS Director’s Fund under Grant No. BJPY2019A01 and the K C Wong Education Foundation.
References:
[1] H. Q. Wang, G. S. Xu*, et al., PRL 112, 185004 (2014).
[2] B. Meerson, Rev. Mod. Phys. 68, 215 (1996)Speaker: Guosheng Xu (Institute of Plasma Physics, Chinese Academy of Sciences) -
KSTAR has a mission such as achieving a pulse length for more than 300 seconds and achieving a high-performance plasma1. The pulse length of the KSTAR discharge has increased each year gradually. Assigned research on the long pulse operation in KSTAR has been conducted since 2015. The pulse length of 90 seconds is achieved in the 2018 KSTAR experimental campaign.
The high $\beta_{P}$ operation mode which helps to be in high fNI is preferred to explore the long pulse discharge. The high $\beta_{P}$ operation mode in KSTAR is achieved by optimizing the operating conditions such as the reduction of plasma current and the precise control of the ECCD deposition location. Operating conditions of the long pulse discharge are usually $I_{P}$=400-450 kA, $B_{T}$=1.8-2.5 T, $P_{NBI}$=2.0-4.3 MW, and $P_{EC}$=0.7 MW. Plasma parameters of the long pulse discharge are presented as $n_{e,core}$~2.5-5.5x$10^{19}$ $m^{-3}$, $T_{e,core}$~2.0-6.0 keV, $\beta_{P}$~1.3-2.7, and $V_{loop}$≤0.1 V.
The operating limits which do not allow to go the longer pulse are as follows; The first is the heat control on the PFCs. KSTAR has a limited PFC cooling capacity compared with heat sources at present. The temperature on the PFCs is kept rising, especially the poloidal limiter and the divertor, and any steady-state condition of temperature is not observed yet. The second is the significant non-linear drifting signal in the magnetic diagnostics such as Rogowski coil, magnetic probes, and flux loops. If magnetic signals used in the real-time control of plasma shape are suffered from significant non-linear drift during the long pulse discharge, the plasma shape has distorted and deteriorated control accuracy and reliability. The third is the plasma performance degradation for long time scale. KSTAR routinely observes gradual performance degradation in the long pulse discharges.
In the 2018 long pulse experiment, the plasma shape and the operating condition of NBIs have been firstly optimized to reduce the heat on PFCs such as divertor and poloidal limiter, which it is allowed to conduct the longer pulse length. And the real-time EFIT operation utilized for the real-time plasms shape control has been optimized by excluding signals showing the significant non-linear drift. The performance degradation for a long time scale is under investigation and resolving.
Figure 1 shows the overview of the longest discharge #21735, a pulse length of ~90 seconds, achieved in the 2018 KSTAR experimental campaign. Plasma features are plasma current of 400 kA, toroidal magnetic field strength of 2.44 T, NBI power of 2.8 MW, and ECH power of 0.7 MW. High $\beta_{P}$ operation mode is sustained until ~50 s with high electron temperature reaching ~6 keV with electron density of ~2.5x$10^{19}$ $m^{-3}$ and ion temperature of ~2 keV at the plasma center. After ~50 s, the plasma performance $\beta_{P}$ is degraded and accelerated with the change of the X-point location that is helpful to be the longer pulse length by reducing the burden of PF3 and PF4.
As shown in Fig. 2, the discharge #21735 is suffered from significant non-linear drifting signals in the magnetics due to the hot plasma with the long pulse length. Figure 2(a) shows that plasma shape analyzed with non-linear drift corrected signals is much different from one with un-corrected signals, especially in Rout and RX. Also, the rate of $\beta_{P}$ degradation is evaluated to be reduced a little bit when it analyzed with non-linear drift corrected signals. Figure 2(b) shows that the influence of drifting signals in magnetics on the temperature of the poloidal limiter which is a region showing the highest temperature among PFCs in the long pulse discharge at present. By excluding significant drifting signals of the magnetics in the analysis of the real-time EFIT, the plasma shape is relatively controlled further to the real target. Controlled and stable plasma shape in terms of Rout especially leads to stable and low poloidal limiter temperature further.
For the longer pulse discharge in KSTAR, the heat to the poloidal limiter as well as the divertor must be reduced and controlled. KSTAR is upgrading its H&CD with NBI of ~12MW in 2020 and the divertor material and shape to control the heat load on the plate in 2021. The fast-ion loss driven by the KSTAR NBI system is the leading cause of overheating the poloidal limiter. The correlation between plasma shape and temperature on the poloidal limiter is related to the behavior of fast-ion particles. Ref. 2 discusses that the extended plasma shape to the outward leads the fast-ion loss through a bad orbit loss process. Furthermore, the unintentional change in plasma shape by non-linear drifting signals is likely to cause plasma performance degradation.
The fast-ion loss is also enhanced by MHD activities in the core and transport regions such as NTM and the Alfven mode. The increase of plasma current, i.e., from 400 kA to 600 kA, is expected to reduce the bad orbit loss of the fast ion. However, the temperature growth rate of the poloidal limiter varied little with the increase of plasma current in the KSTAR long pulse experiment. In these discharges, the fast-ion loss due to the enhanced transport by MHD activities is considered more dominical than the fast-ion loss through the bad orbit. The enhancement transport of fast-ion loss was analyzed by using TRANSP and NuBDeC 2. Thus, the presence of MHD activities in the core and transport regions reduces the absorption of the NBI power, which results in the degradation of plasma performance and increases the temperature of the poloidal limiter, as well as makes worse plasma performance in itself.
At the discharge #21735, PF1-PF4 coils reach their limit of the induced current set to 15 kA. Therefore, additional H&CD and higher plasma performance are required for reducing the burden of PF coils and for conducting the longer pulse length up to over 102 seconds. Based on existing KSTAR long pulse discharges, it is analyzed whether the H&CD upgrade will allow going for the longer pule operation within the operating limits. Mainly we focus on how to control the heat on the PFCs and the significant non-linear drifting magnetic signals. Also, we suggest the practical guidelines to achieve the longer pulse length with the high-performance plasmaReference
1 Lee, G.S., et al., Nucl. Fusion 40 (2000) 575.
2 Rhee, T., et al., Phys. Plasmas 26 (2019) 112504
Speaker: Dr Hyun-Seok Kim (National Fusion Research Institute) -
Disruption prediction and avoidance is critical for ITER and reactor-scale tokamaks to maintain steady plasma operation and to avoid damage to device components. The present status and results from the disruption event characterization and forecasting (DECAF) research effort (1) are shown for multiple tokamak devices. Access to the full KSTAR, MAST, NSTX, AUG, TCV, and DIII-D databases is presently available. The DECAF paradigm is primarily physics-based and provides quantitative disruption prediction for disruption avoidance. DECAF automatically determines the relation of events leading to disruption and quantifies their appearance to characterize the most probable and deleterious event chains, and also to forecast their onset.
Present DECAF analysis of KSTAR, MAST, and NSTX databases shows low disruptivity paths to high beta operation. The disruptivity does not increase at high normalized beta, $\beta_N$, as is often mistakenly expected. Automated analysis of rotating MHD modes now allows the identification of disruption event chains for several devices including coupling, bifurcation, locking, and potential triggering by other MHD activity. DECAF can now provide an early disruption forecast (on transport timescales) allowing the potential for disruption avoidance through profile control. The first hardware to allow real-time evaluation of this activity on KSTAR was installed in 2019. The first real-time data and offline test results are shown below. DECAF event characterization and event chain analysis shows that disruption forecasting analysis often starts during plasma operational states and at parameters that appear safe. This fact is completely missed by “disruption database” studies that only process data near the disruption time.
Disruption prediction research using DECAF also allows quantifiable figures of merit (i.e. the plasma disruptivity) to provide an objective assessment of the relative performance of different models. This allows an assessment of how well the predictor performs compared to ITER needs. Figure 1
shows a progression of DECAF disruption forecasting models. The earliest models included about 10 events and were run on databases for which the events that led to the disruption were known and yielded very high performance (e.g. 100% true positives). A next evaluation of models focused on earlier forecasting once the first physics forecasting model was implemented in the code. True positives were found to be ~84%, which was a measure mainly of the single forecasting model. Forecasting models continue to be added to improve that performance. Recent code testing has been on large databases ~ 10,000 shot*seconds of plasma run time tested (Figure 1). This was done with a smaller number of events due to computer limitations. With 5 events, applied to all plasma shots from an NSTX database, DECAF shows performance levels of over 91% true positive disruption predictions. False positives in this analysis reached 8.7% which is fairly high. Present code development that allows the events to poll each other will improve the false positive tally.There is an extensive physics research effort supporting DECAF model development. Two recent studies are summarized here. First, analysis of high performance KSTAR experiments using TRANSP shows non-inductive current fraction has reached 75%. Resistive stability including $\Delta$’ calculation by the Resistive DCON code is evaluated for these plasmas using kinetic equilibrium reconstructions with magnetic field pitch angle data to determine capability for instability forecasting. (2) To design experiments for the 2019 KSTAR run, “predict-first” TRANSP analysis was performed showing that with the newly-installed 2nd NBI system (assuming usual energy confinement quality and Greenwald density fraction), 100% non-inductive plasmas scenarios are found in the range $\beta_N$ = 3.5–5.0. These plasmas will provide a unique long pulse (~20s) database for disruption forecasting studies. Second, new analysis of MAST plasmas has uncovered for the first time global MHD events at high $\beta_N$ identified as resistive wall modes (RWMs). A stability analysis of MAST plasmas shows a significant ballooning shape of the three-dimensional RWM eigenfunction that compares well to fast camera images (Figure 2).
This new result shows that the MAST RWM eigenfunction shape and growth rate appear significantly altered by the location of conducting structure compared to results from NSTX, which shows a much more spherical shape due to close-fitting copper plates. (3) The conducting wall stabilizing effect on the kink mode is computed to be relatively small in MAST. The plasma with an unstable RWM has a computed no-wall $\beta_N$ limit of 5.0 and a with-wall limit of 5.16. Another new result of this analysis shows that kink mode stabilization was primarily due to the vacuum vessel in MAST. In contrast, design equilibria of MAST-U plasmas show a significant increase in kink stabilization due to the addition of stainless steel divertor plates. The analysis including 3D conducting structure shows an increased stabilized range of $\beta_N$ = 3.8 – 5.7. MAST-U design equilibria with closer plasma-plate coupling show higher with-wall $\beta_N$ limits.
.
Real-time DECAF analysis is now being constructed for KSTAR. The first of several real-time computers and diagnostic interfaces has been installed to detect and decompose rotating MHD activity in the device. Sixteen channels of a toroidal array of magnetic probes have been acquired in real-time during MHD activity (Figure 3).
Offline DECAF analysis of the real-time signals shows that the mode decomposition and DECAF object decomposition replicates the local KSTAR spectrogram analysis (Figure 3). The DECAF object decomposition is directly used to produce a warning level for MHD activity shown to provide an early warning forecast (~ 300 ms) for mode locking and subsequent disruption in KSTAR, potentially allowing active profile control to avoid the mode. The real-time hardware includes a field-programmable gate array chip to allow the computation of FFTs used for mode decomposition in real time.*Supported by U.S. DOE grants DE-SC0016614, DE-SC0018623, and DE-FG02-99ER54524
(1) S.A. Sabbagh, J.W. Berkery, Y.S. Park, et al., IAEA FEC 2018 (Gandhinagar, India), paper EX/P6-26.
(2) Y. Jiang, S.A. Sabbagh, Y.S. Park, et al., submitted to this conference.
(3) S.A. Sabbagh, J.W. Berkery, R.E. Bell, et al., Nuclear Fusion 46 (2006) 635.Speaker: Steven Sabbagh (Columbia University) -
Tearing modes with poloidal/toroidal mode numbers $m$/$n$ = 2/1 have been routinely observed in KSTAR [1] with modes having notably large amplitude leading to significantly reduced both normalized beta, $\beta_N$, and plasma stored energy. Global kink/ballooning or resistive wall modes (RWMs) are observed to be stable at high $\beta_N$ above the $n$ = 1 ideal MHD no-wall stability limit, $\beta_N^{no-wall}$. Global MHD stability modified by kinetic effects can explain the observed RWM stable plasma operation above $\beta_N^{no-wall}$ through stabilization by kinetic effects [2]. The observed 2/1 modes onset at relatively low $\beta_N$ ~ 1.4 in the early phase of the discharge with electron cyclotron heating (ECH) applied to the core region of the plasma. Neighboring MHD activity that could trigger the mode by providing a seed island is not observed. The destabilized 2/1 mode initially has a small amplitude and growth rate, and by considering its ‘triggerless’ nature, the mode at low $\beta_N$ could be driven by an island resulting from a current density profile yielding instability to the 2/1 mode which can be represented by a positive tearing stability index, $\Delta'$, at the $q$ = 2 surface. In contrast, a comparable 2/1 mode which onsets at high $\beta_N$ ~ 3.5 above $\beta_N^{no-wall}$ coincides with large ELM crashes observed that can trigger the mode, and has significantly faster mode growth. Although their onset conditions and mechanisms appear to be different, once these islands grow above the critical island size of the neoclassical tearing mode (NTM), they commonly could grow dominantly by neoclassical pressure-driven effects. The resistive MHD stability examined by the resistive DCON code [3] is compared between these 2/1 modes destabilized at different $\beta_N$ and onset conditions. The $\Delta'$ from resistive DCON indicates that once the 2/1 mode onsets and saturates, the computed $\Delta'$ is commonly unstable ($\Delta'$ > 0). A stable $\Delta'$ is computed in the high $\beta_N$ > $\beta_N^{no-wall}$ phase where the ‘triggered’ 2/1 mode onset is observed indicating that the stability is apparently consistent with a linearly stable NTM.
The NTM island growth rate is described by the modified Rutherford equation (MRE) [4], $\tau_R$ $\partial w$/$\partial t$ = $\Delta'$($w$) + $k_{NC}$$\Delta'_{NC}$ − $k_{pol}$$\Delta'_{pol}$ − $\Delta'_{GGJ}$ − $\Delta'_{CD}$. To identify the governing destabilization mechanism in the mode stability, the neoclassical drive term $\Delta'_{NC}$ in MRE which is given by $\Delta'_{NC}$($w$) = 16$J_{BS}$$/$($sw \langle J \rangle$) is computed by the TRANSP code. The magnitude of the $\Delta'_{NC}$ is computed to be large with values in the range 2.4–2.8 for the equilibria at high $\beta_N$ in which the NTM stability is expected to be unstable suggesting the dominant effect of the pressure-driven bootstrap current in the mode destabilization. The $\Delta'_{NC}$ becomes significantly lower in the range 0.6-0.9 after the island saturates and the confinement is consequently degraded. The TRANSP computed bootstrap current ($J_{BS}$) profiles for the ‘triggered’ 2/1 mode in the high $\beta_N$ discharge are shown in FIG. 1. In the fairly broad $J_{BS}$, the distinctive edge $J_{BS}$ is created by the steep edge pressure pedestal. The $J_{BS}$ at the $q$ = 2 surface is reduced almost by a factor of 3 after the mode saturation. The total non-inductive current fraction reached 64% in the high $\beta_N$ equilibrium before the mode onset. The remaining terms in the MRE are computed to develop a preliminary NTM stability model for the high $\beta_N$ KSTAR plasma. In the MRE, the critical island width, $w_d$, which accounts for the transport threshold due to small island size is calculated by assuming somewhat reduced transport anisotropy, $\chi_\perp$$/$$\chi_\parallel$, in the range 1e$^{-11}$-1e$^{-9}$ inside the island. The $w_d$ is computed to be ~1 cm for the anisotropy values examined.
For high beta plasma operation, active stabilization of NTMs by compensating the perturbed $J_{BS}$ by utilizing the electron cyclotron current drive (ECCD) has been studied on KSTAR. The EC wave can be launched from the two real-time steerable launchers located at above and below the midplane. The EC system can produce a wave frequency of 105 or 140 GHz, and under the operating toroidal field in the range 1.8-2.0 T used in typical H-mode discharges, the 105 GHz, second harmonic ECCD is used to drive localized current in the island. Figure 2 shows the ray trajectories of the 105 GHz ECCD launched from the upper midplane launcher (EC-2) in the experiment which drive current on the resonance intersection at Z = +20, +26, +28 and +30 cm along the second harmonic resonance layer with $B_T$ = 1.9 T computed by the TORAY-GA ray-tracing code. The trajectory to the resonance intersection on the same Z > 0 along the resonance layer is used since it can generate a relatively narrow current profile width ($\delta_{EC}$) with a higher current density ($J_{EC}$) beneficial for NTM stabilization. The EC deposition location is varied in steps around the estimated $q$ = 2 surface location using the 2D ECE imaging diagnostic in the experiment. By using a total EC power of 0.75 MW, a localized $J_{EC}$ having a peak magnitude of ~7 A/cm$^2$ on the mode rational surface is computed when the 2/1 mode amplitude is decreased by ~80% due to the ECCD suitably aligned on the island shown in FIG. 3. The computed peak $J_{EC}$ magnitude somewhat smaller than the $J_{BS}$ of ~10 A/cm$^2$ at $q$ = 2 (FIG. 1) could be responsible for the partial stabilization of the mode amplitude. The results indicate that active stabilization of the 2/1 NTM by ECCD is plausible by using a total available EC power of 1.5 MW from the two EC systems on KSTAR. To analyze the NTM stabilization by the effect of ECCD ($\Delta'_{CD}$), the MRE model is examined by fitting the model by using the experimental data. The electron profiles and equilibrium conditions maximizing the control performance are investigated. This study will provide the required foundation for active stabilization of NTMs which is a critical need for high beta plasma operation in KSTAR.Supported by US DOE Grant DE-SC0016614
[1] Y.S. Park, S.A. Sabbagh, W.H. Ko, et al., Phys. Plasmas 24 (2017) 012512
[2] J.W. Berkery, S.A. Sabbagh, R. Betti, et al., Phys. Rev. Lett. 104 (2010) 035003
[3] A.H. Glasser, Z.R. Wang and J.K. Park, Phys. Plasmas 23 (2016) 112506
[4] E.D. Fredrickson, M. Bell, R.V. Budny, et al., Phys. Plasmas 7* (2000) 4112
Speaker: Young-Seok Park (Columbia University) -
The configuration of reversed shear q(r)-profile is existence for the burst of sawteeth-like events, and the collapse is triggered periodically by the magnetic reconnection of double tearing modes (DTM) [1], with abbreviation of DTRC (double tearing reconnection crash). The excitation conditions of DTRC has strong relationship with the impurity ions, the power threshold of ECRH, the influence of resonant magnetic perturbation (RMP), and the beam direction of NBI, et al. One example is given in Fig. 1, and the energetic ions (neutron yield) is collapse swiftly when the density of copper impurity ions achieves the threshold condition. The internal transport barrier of electron temperature (e-ITB) [2-4] is formed before the DTRC, and the confinement of energetic ions (thermal particles) is improved dramatically, where one kind micro-instability is excited accordingly before the formation of e-ITB (the thermal particles are transport in the ion $B\times\nabla B$ drift direction that is similar to turbulence [5]).
Series of Alfven Eigenmodes are aroused during the oscillation of DTRC after the injection of energetic ions, i.e. Beta-induced Alfven Eigenmodes (BAEs), Reversed Shear Alfven Eigenmodes (RSAEs) and Beta-induced Alfven-Acoustic Eigenmode (BAAE) as shown in Fig. 2. The BAAE coexists and possesses similar characteristics as the pair of BAEs-RSAEs, e.g. the radial structures of BAAE and BAEs are deformed into the analogous triangle shapes along poloidal direction, and the frequencies of BAAE and RSAEs are sweeping upward synchronously for the decreasing of $q_{min}$ ($q_{min}\leq1$), and the mode structure of BAAE is achieved experimentally with m = 4/n $\approx$ 4. The excitation conditions of BAAE in EAST are also discussed in Fig. 3, and the BAAE is damped easily for $\tau_o \leq$ 150 ms, where the $\tau_o$ is the period of DTRC oscillation. It needs to be stressed that the gradient of electron temperature at vicinity of $q_{min}$ and the portion of energetic ions in central region are increasing with $\tau_o$ [4], and the BAAE should be excited by the special distribution of energetic ions, where the gradient of electron temperature might reduce the excitation threshold of BAAE.
Reference
[1]. Chang, Z., et al. "Off-Axis Sawteeth and Double-Tearing Reconnection in Reversed Magnetic Shear Plasmas in TFTR." Physical Review Letters 77(17): 3553 (1996).
[2]. Ming Xu, et al. "Characteristics of off-axis sawteeth with an internal transport barrier in EAST." Nuclear Fusion 59(8): 084005 (2019).
[3]. Ming Xu, et al. “Excitation of Alfven Eigenmodes and Formation of ITB during off-axis Sawteeth in EAST”, Oral on 16th IAEA technical meeting on Energetic Partciles.
[4]. Ming Xu, et al. " Excitation of Beta-induced Alfven-Acoustic Eigenmode during Sawteeth-like Oscillation in EAST." Accepted in Nuclear Fusion.
[5]. Brower, D. L., et al. "Multichannel Scattering Studies of the Spectra and Spatial Distribution of Tokamak Microturbulence." Physical Review Letters 54(7): 689-692 (1985).Speaker: Dr Ming Xu (Institute of Plasma Physics, Chinese Academy of Sciences) -
Leading edge induced material damages are very critical in future fusion devices which may have cassette structure for plasma facing components 1. The upper tungsten divertor in EAST is the first application of active cooled ITER-like W/Cu monoblocks modules in a tokamak 2. The misalignment between neighboring monoblocks was formed inevitably during fabrication and assembly processes, providing the possibility to investigate the leading edge induced damages. The in situ leading edge induced damages, i.e. cracking and melting on W/Cu monoblocks during long pulse operations in EAST were characterized and discussed, which are of important references for ITER and future fusion devices.
Many in-situ tungsten melting phenomenons were observed by means of distinguishing the droplets ejection from CCD video during plasma discharges. Correspondingly, the post mortem inspection also detected the melting vertical length up to 14 cm around strike point on both inner and outer targets [3]. It is found that all tungsten melting occurred only at the edges of cassette modules where larger misalignment up to 3 mm could be formed during assembly process. Due to low misalignment below 0.3 mm between neighoring monoblocks within each cassette module, no any damage was occurred on the edges of monoblocks within cassette module. The melted layer which was mainly driven by the electromagnetic force was moved either up or down in different conditions, and induced only a few bridge connections along polodial direction. Such tungsten melting ejected a large number of droplets into the core plasma, and resulted in a sharp increasing of tungsten impurity and power radiation and could eventually lead to disruptions. With droplets ejection and melted layer removal, the melted corner was leaved with a moderate chamfer structure as depicted in Fig.1. From thermal simulation predictions [3], such beveled structures may mitigate the temperature rise in the following shots. Indeed, no evidence showed that the melted monoblock could be melted again in the following plasma discharges. Such melting does not have a big impact on the EAST operations.
Fig.1. Morphology of a typical leading edge induced melting at G2 cassette module in EAST. (a) the in-situ picture of melting in vacuum vessel; (b) the removed melted monoblocks; (c) the beveled leading edges after melted layer movement and droplets ejection; and (d) the metallic inspection on the near surface.Meanwhile, many new kinds of macro cracks were also found universally on the edges of cassette modules. They were generally horizontal and along toroidal and radial direction. The cracks were initially generated at the edge corners by leading edge thermal loading, and then propagated into deeper and central areas. Fig. 2 is the typical morphology of cracks generated from leading edges on W/Cu monoblocks. The width of such cracks was about 70 µm, while the depth of cracks can reach up to 4 mm. Since the melting occurs, the temperature can reach up to thousands of oC at leading edges. Accordingly, the thermal stress and strain are in a high level with such high temperature gradient, which could induce such macro cracks. It is noted that such new kinds of cracks were different from the macro cracks along radial direction and poloidal directions during high heat load tests especially under multiple cycles of 20 MW/m2 only on surface of such kinds of W/Cu monoblocks [4]. Such cracks seemed to divide W surface at leading edges equally to similar lamellar structures. Because such cracks were parallel to the heat propagation direction and the elongated grains, they probably have little effect on heat transfer. So far, no clear influence on plasma discharges by such kinds of cracks was found in EAST.
Fig.2. Morphology of cracks at the leading edge of W/Cu monoblocks. (a) the top view of a crack; (b) the side view of a crack; (c) the typical multiple cracks; and (d) the micromorphology of a crack.
Currently, such leading edge induced damages in EAST may be acceptable to some extent. However, with gradually increasing of auxiary heating power, such leading edge induced damages are foreseen to become more seriously, will be likely to restrict the long pulse and high performance operations in future. Thus, further efforts should be performed to mitigate such damages, such as optimization of the chamfering structure to protect the leading edges.
This work was subsidized by the programs of National Natural Science Foundation of China (Nos. 11775260, 11675219, 11675218, 11861131010).
References:
1 Missirlian M, et al., Fusion Eng. Des. 88 (2013) 1793-1797
2 Damao Yao, et al., Fusion Eng. Des. 98-99 (2015) 1692-1695
3 Dahuan Zhu, et al., Nucl. Fusion, 2020, 016036
4 T. Hirai, et al., , Nucl. Mater. Energy. 9 (2016) 616-622.Speaker: Mr Dahuan Zhu (Institute of Plasma Physics, Chinese Academy of Sciences) -
Plasma-wall interaction (PWI) processes are important in long-pulse plasma operation of fusion devices due to the main issues of increased fuel retention, material erosion and redeposition which are induced by a large increase in particle fluence to the wall compared to present experiments. The in situ approach is an urgent requirement to be utilized to real-time measure the fuel content on the first wall during the tokamak operation. Meanwhile, the characteristic of fuel retention during PWI processes is strongly linked to edge particle flux and fluence. Therefore, the key challenge for fuel retention studies is to monitor not only the elemental composition on the PFCs but also the local plasma condition during experimental processes in fusion device in situ and real-time. Experimental Advanced Superconducting Tokamak (EAST), which is the world’s first fully superconducting magnetic confinement facility with ITER-like magnetic field configuration, provides a unique platform to investigate the physical and engineering issues of PWI research for the next H-mode long-pulse fusion devices such as ITER and beyond. Laser-induced breakdown spectroscopy (LIBS) as a well-established elemental composition analysis method has been demonstrated as one of the most promising approaches for in situ PWI studies on the first wall of fusion devices [1,2]. The in situ LIBS system in EAST 3 provide information about the elemental composition of the first wall on the high-field side (HFS) near mid-plane in EAST. There is a need to address the link between the fuel retention on the first wall and local plasma conditions. For this purpose, an in situ LIBS system combined with optical emission spectroscopy (OES) is used to not only measure the fuel retention on the first wall but also simultaneously monitor the edge conditions during long-pulse discharges of EAST.
The in situ LIBS diagnosis system is located at the H-port, which is one of the 16 main horizontal ports along the toroid in EAST. The schematic of the system is illustrated in Fig. 1. The detail of the optical system can be found in our previous works 3. The intensity of D retention on the first can be achieved and analyzed by the spectrometer. During the EAST long-pulse plasma discharges, the spectrometer is also used for measuring the edge OES spectra for analyzing the local plasma conditions and providing the photon flux of the Dα line emission from the region of scrape off layer (SOL) and edge plasma. The typical spectra of LIBS from the TZM first wall and edge OES in EAST are shown in Fig. 2. The LIBS spectrum is measured after Li wall conditioning and D discharge using single pulse LIBS on the surface of the mid-plane TZM first wall. It clearly shows Li and Dα spectral lines from fuel retention as well as Mo I emission from the first wall substrate. The H content in the core discharge plasma is very low due to the Li wall conditioning, thus the fuel retention on the wall is dominated by D implantation and codeposition. The H content is lower than the limit of detection of the spectrometer at these experimental conditions. The lines of Li I, Li II and Dα are observed by edge OES measurement. Some spectral lines of Mo impurity due to first wall erosion may be observed in some cases of edge localized modes burst or fast plasma shutdown. The spectral intensity ratio of Li II and Li I in the edge OES spectra is much higher than that in the LIBS spectra.
As shown in Fig. 3, the relationship between the edge OES Dα photons fluence and flux and LIBS Dα photons fluence from the unit area on the first wall could qualitatively reflects the correlation between the edge D particles fluence and D retention on the first wall. Five typical long-pulse operations (#73341, #73342, #73930, #73977, and #73978) with duration time from 20 s to 80 s are used to illustrate the correlation between the OES and LIBS measurements. The Dα photons fluxes from the SOL and edge plasma are from 2×10^21 photons/sm^2 to 8×1021 photons/sm^2 in five long-pulse plasma discharges. The edge Dα photons fluences are from 0.5×10^23 photons/m^2 to 4.0×10^23 photons/m^2. These result in that the Dα photons fluences from LIBS plasma on the D retention first wall are from 0.3×10^17 photons/m^2 to 1.2×10^17 photons/m^2. The Dα photons fluences from the first wall almost linearly increase as Dα photons fluence from OES. By calculating S/XB of the edge plasma and LIBS plasma, the edge D particle fluences and D retention amounts on the first wall are estimated from 0.67×10^24 m^-2 to 4.83×10^24 m^-2 and from 1.88×10^19 m^-2 to 5.59×10^19 m^-2, respectively. Thus, the correlation between the intensities of Dα from LIBS spectra on the first wall and the Dα intensity from the edge spectra is clearly demonstrated. This indicates that the D retention on the first wall strongly related to the edge D particles fluence. We conclude that the fuel retention on the HFS first wall is dominantly influenced by the local D particles implantation. The results could improve the understanding of fuel retention on the first wall in EAST and demonstrate the prospect of LIBS approach to in situ investigate PWI in the upcoming fusion device like ITER.
In the future, the fuel retention on the first wall as a function of the distance between HFS and LCFS will be studied to search for the features for the retention behavior in the transition from net erosion to net deposition zone. In addition, it is planned to use an endoscope system to upgrade and extend the LIBS and OES detectable area to the tungsten upper divertor region which is a strong PWI interaction zone. The wall Langmuir probe and infrared camera observation systems will be used to characterize the edge condition and wall surface temperature. Deposition and erosion in divertor zones will be in situ investigated using the upgrade LIBS system in EAST.
1 V. Philipps et al, Development of laser-based techniques for in situ characterization of the first wall in ITER and future fusion devices, Nucl. Fusion 53, 093002 (2013).
2 C. Li et al, Review of LIBS application in nuclear fusion technology, Front. Phys. 11, 114214 (2016).
3 D. Zhao, C. Li et al, Remote in situ laser-induced breakdown spectroscopic approach for diagnosis of the plasma facing components on experimental advanced superconducting tokamak, Rev. Sci. Instrum. 89, 073501 (2018).Speaker: Dr Cong Li (Dalian University of Technology) -
Hybrid scenarios are under development in KSTAR which are defined as “stationary discharges with β_N ≥ 2.4 and H_89 ≥ 2.0 at q_95 < 6.5 without or very mild sawtooth activities”. β_N≲3.0, H_89≲2.4 and G-factor (≡β_N H_89/q_95^2) ≲0.46 has been obtained simultaneously at ne/nGW~0.7 and sustained for ≳40 τ_E during the main heating phase as shown in figure 1.
The hybrid scenarios are established by several approaches; early heating, plasma current overshoot, and late heating approach. Fully non-inductive current drive has been obtained in the plasma current overshoot recipe and more stable discharges have been established with the early heating scenario by adjusting the timing of the 3rd NBI. MHD analyses show that fishbones are the main instabilities in KSTAR hybrid scenarios. Internal kink modes are appeared while ECH is applied around the on-axis region. In relatively low and high q_95 ranges, fishbones appear frequently. On the other hand, n = 2 mode, probably NTM appears dominantly in the intermediate q_95 range.
The origin of confinement enhancement is investigated in a slow transition period from the standard H-mode to the hybrid mode with 0-D power balance, 1-D kinetic profiles, linear gyro-kinetic, and pedestal stability analysis. The thermal energy confinement enhancement is thought to be mainly due to increase of the both ion and electron pedestal and some off-axis ion energy confinement improvement via stiffness weakening. The 0-D power balance analysis exhibits that the fast particle confinement is improved up to the 3rd phase of the slow transition period, whereas the thermal energy confinement only in the 3rd phase. The ion heat diffusivity is globally increased in the 3rd phase but some increase of R/L_Ti is observed in the off-axis region. Linear GKW [1 ] simulations show that the dominant turbulent mode is changed from TEM to ITG as the thermal confinement enhances. This results in increase of the core electron temperature. The finite β stabilisation effect plays a role together with the fast particle stabilisation effect around the core region ρ_tor=0.35. ω_(E×B) can reduce the linear growth rate of ITG in the off-axis region, ρ_tor=0.50 and 0.70 where the toroidal rotation contribution is crucial. The alpha stabilisation effect is also found at ρ_tor=0.5. ETG is estimated to sit in ρ_tor= 0.5 and 0.7 from linear gKPSP [2 ] simulations. The pedestal is improved due to increase of β_p and subsequent Shafranov shift. The EPED model [3 ] could reproduce the height of the pedestal if the feature of hybrid scenarios, a small Ohmic current fraction, and a rather flat q-profile in the core, consequently much smaller l_i than the standard EPED equilibria, is considered to calculate the Shafranov shift properly. The diamagnetic effect turns out to boost the pedestal growth.
A hypothesis to explain the confinement enhancement mechanism is suggested as shown in figure 2. The primary effect of NBI is to increase β_th, β_fast, and V_tor and alter the q- and magnetic shear profile. This increase of β can improve the pedestal stability owing to Shafranov shift which can increase the core temperatures via profile stiffness. This is the secondary effect. If the q- and magnetic shear profiles are changed to be hybrid-like, q(0)~1 with low magnetic shear, the sawtooth and the core turbulence can be mitigated or stabilised. This results in an increase of core β. If β increases larger than β_th enough to stabilise or alleviate ITG but lower than β_crit to avoid triggering an EM mode, the β stabilisation effect together with the alpha stabilisation effect could increase the core further probably via stiffness mitigation. Fast particle stabilisation can contribute as well while the fast particle confinement is improved. Increase of the toroidal rotation can increase ω_(E×B) which can contribute to stabilise ITG. All these effects come up with increase of pedestal through Shafranov shift. This is the tertiary effect, causing transition to the hybrid regime. Correlation between the NB coupling and the core and pedestal plays as a background engine to booster these effects.
References
[1 ] Peeters A. G. et al 2009 Comput. Phys. Commun. 180-12 2650–2672
[2 ] Kwon J.-M. et al 2017 Comp. Phys. Communications 215 81
[3 ] Snyder P.B. et al 2009 Phys. Plasmas 16 056118Speaker: Yong-Su Na (Seoul National University) -
$\quad$Plasmas with a high runaway electron (RE) current fraction, $f_{RE}$ > 0.5, have been achieved during the flat-top of EAST Ohmic discharges with both a circular limited and an X-point diverted configuration. Low toroidal mode number Alfvén eigenmodes (AE) in the frequency range of 100-300kHz including TAE, KTAE and GAE, which are excited by low-energy REs, are clearly identified in the quiescent regime. Operation in the quiescent regime including accurate measurement of all key parameters related to REs provides a suitable experimental platform for RE excitation and dissipation, which could potentially have beneficial implications to the post-disruption RE regime $^{[1]}$.
$\quad$Extremely low-density operation ($n_e$ < 4*10$^{18}$ m$^{-3}$) free of error field penetration supports the excitation of fruitful quiescent RE populations $^{[2]}$. By slowly letting the density ramp down during the flattop, REs are firstly confirmed by visible hard X-rays (HXRs) and electron cyclotron emission (ECE) and then the signals of HXRs and ECE grow fast, indicating that amount of REs are generated due to the avalanche process, as shown in figure 1. At a lower density, a transition from growth of HXRs and ECE to saturation are simultaneously observed. Meanwhile, a large drop of the surface loop voltage (down to <50% of the loop voltage value before this transition) is found, indicating the replacement of the resistive plasma current by that carried by the REs.
$\quad$ After the transition, continuing to ramp down the density does not raise the toroidal electric field ($E_{loop}$) and the amplitude of HXRs and ECE keeps constant, supporting that the stable characterizations of the RE current fraction and the energy distribution in the regime. Also, the saturated electric field is ~8 times above the theoretical critical electric field for avalanche growth ($E_C$) but lower than the threshold electric field for Dreicer generation (12-20 * $E_C$).
$\quad$ During low-density ohmic discharges, a bump on energetic electron energetic distribution is formed. Low toroidal mode number Alfvén eigenmodes (AE) in the range of 100-300kHz, including TAE, KTAE and GAE, are excited by energetic electrons and detected by magnetic signals and reflectometry, as shown in figure 2. These activities strongly correlate with evolution of energetic-electron energetic distribution.
$\quad$Besides, Multiple KTAEs excited by energetic electrons $^{[3]}$ and geodesic acoustic mode have been simultaneously observed. Three-wave interactions between these modes are conclusively identified, indicating fixed phase relationship. This nonlinear coupling, may lead to SAW instability nonlinear saturation, as well as EP energy channeling through KTAEs into the GAM, and eventually, bulk plasma heating through GAM collisionless damping $^{[4]}$].This material is based upon work supported in part by the National Key R&D Program of China under Contract No. 2017YFE0301205, the National Natural Science Foundation of China under Contract No. 11775267, and Youth Innovation Promotion Association of Chinese Academy of Sciences under Contract No. 2017480.
References
[1] P. Aleynikov and B. Breizmann 2015 Phys. Rev. Lett. 114 155001
[2] L. Zeng et al 2016 43rd EPS conference on Plasma Physics, Leuven, Belgium, O4.112
[3] F. Zonca and L. Chen 1996 Phys. Plasmas 3 323
[4] Z. Qiu et al 2018 Phys. Rev. Lett. 120 135001Speaker: Long Zeng (Institute of Plasma Physics, Chinese Academy of Sciences) -
Recent control advances at KSTAR enabled us to not only sustain the ITER-similar shape (ISS) in a stationary manner but also experimentally demonstrate the ISS-compatible RMP-ELM control in KSTAR for the first time, using the n=2, +90-deg phasing RMP, matching the ITER-like dimensionless parameters in lower single null (LSN) configuration with vastly contrasting upper/lower triangularities. The kinetic parameters of the ISS are different from the typical KSTAR configurations seen in ELM-RMP suppression experiments [Y.M.Jeon PRL 2012, Y.In NF 2019].
Achieving ITER relevant parameters in KSTAR has been a challenging target. The main difficulty lies in the fact that low edge safety factor requirement ($q_{95}$ ~ 3.1 - 3.4) would result in kink-driven mode-locking easily, considering fundamental challenges of the independent control of highly up/down asymmetric triangularities with constraints by a large up/down symmetric central solenoid (CS). However, the establishment of the ITER relevant parameters in a superconducting device like KSTAR allows us to explore various ITER-related physics and engineering constraints, prior to the ITER-era.
In order to accomplish this ISS target, various advanced magnetic controllers have been implemented and executed in the real experiments, including enhanced vertical stabilization using inboard flux loop differences [Mueller FED 2019], a multi-input multi-output (MIMO) shape control design suitable for LSN shape and the real-time feedforward algorithm [Walker CCA 2016] that would minimize accumulation of integral gain errors.
Figure 1: ISS discharges obtained in KSTAR (a) major parameters in time: Used 2 NB ion sources of total 3.0 MW (#22889) or 1 beam + 2 gyrotrons (#21368). From top to bottom, Ip [MA], $q_{95}$, $\beta_N$, mid-RMP current [kA/turn], edge Thomson [1e19 $m^{-3}$], and $D_{\alpha}$. (b) Obtained shape at the shot #22889, t=5.5065s.The experimental setup of the typical ISS discharge examples is shown in Figure 1. Using available combinations of 1-2 neutral beam sources and 2 ECH 105GHz gyrotrons, the ISS with 3 different levels of the toroidal rotation were created at the total power level of 3.0-3.6 MW. Two types of power combinations were frequently used in order to create different toroidal rotation levels: 1) Two neutral beam (NB) ion sources (up to 3.6 MW) or 2) one NB + 2 gyrotrons (up to 3.3 MW). As shown in Figure 1(a), the discharges match major ITER-like parameters, i.e. $q_{95}$= 3.2 - 3.4, $\beta_N$= 1.6 ~2.0, at the plasma current flattop Ip = 780 - 850 kA at $B_T$ = 1.7~1.8T. The measured upper triangularity ($\delta_u$) is around 0.5~0.55, and lower triangularity ($\delta_l$) is fixed at 0.45, matching $\delta = (\delta_u + \delta_l)/2$ ~ 0.52 in the ITER shape: a typical ISS example is shown in Figure 1(b). The kinetic properties are, however, very different from the typical RMP-driven, ELM control experiments reported in similar $q_{95}$ range in the ref [Y.In NF 2019]. For instance, the ISS shows relatively higher line averaged density $n_e$ ~ 5-7 e19 $m^{-3}$, accompanied by high edge density pedestals with edge Thomson $n_e$ ~ 2-3e19 $m^{-3}$. The edge electron collisionality is estimated as $\nu^*$ ~0.4-0.5 at the pedestal top $\rho$~ 0.89, using the definition at [Sauter PoP 1999] with the assumption of $Z_{eff}$~ 2.
At first the ELM control on the ISS by resonant magnetic perturbation (RMP) coils was attempted using n=1, +90-degree RMP configuration, but it disrupted the plasma instantly, as frequently observed in the typical low-q95 KSTAR configurations. However, from previous observations, the application of n=2 RMP is expected to be more manageable both for the typical low-$q_{95}$ and for the ISS. In fact, we have found the n=2 RMP-driven, ELM-crash control was greatly affected by the absolute density level: In the highest electron density range, with line-averaged $n_e$ ~ 7e19 $m^{-3}$, we were only able to observe a very low mode-locking threshold. Meanwhile, there were strong ELM-crash mitigations where the ELM frequency from 6-10 Hz to 100-120 Hz, even at relatively low level of RMP current = 1.3-1.5 kA/turn.
On the other hand, the ISS experiments with moderate level of density, at line-averaged $n_e$~5e19 $m^{-3}$, expecting lower edge collisionality than the cases above, showed more promising results: The RMP with n=2, +90-degree phasing showed a locking threshold of 3.1-3.3 kA/turn. Application of the RMP current below the locking threshold onto the ISS successfully accessed a marginal but clear ELM-crash suppression window for the first time at $q_{95}$=3.3-3.4, as shown in Fig. 2, for a short time of 200-600 milliseconds. The ELM-crash suppression was accompanied by 1) a global electron density pumpout (as measured at Thomson scatting density profiles and the line-averaged density), and 2) a characteristic $\beta_N$ drop at the level between that of H- and L-mode. The experimental results suggest that the sustainable ELM-crash suppression window for this ISS is likely to be at a level of RMP current = 2.5 – 2.9 kA/turn.
In the near future experimental exploration is planned to robustly access the sustained RMP-driven, ELM-crash suppression on the ISS in KSTAR, helping us to articulate the main advantages of a superconducting device that is capable of a much longer pulses over hundreds of energy confinement time.
Speakers: Sang-hee Hahn (National Fusion Research Institute), Yongkyoon In (Ulsan National Institute of Science and Technology) -
A significant advancement of RF modeling was achieved by realizing for the first time the entire 3D full torus plasma simulation with detailed 3D realistic antenna and SOL plasma. Many experiments in different fast wave (FW) heating regimes, such as hydrogen minority heating and high harmonic fast waves (HHFW), have found strong interactions between RF waves and the SOL region. In a fusion
device powered significantly by RF, however, the loss of RF power in the SOL can be a real plasma material interaction (PMI) issue for possible RF impurity generations and plasma facing component damage besides reducing the RF core performance. A significant interaction between FW and energetic ions generated by neutral beam injection (NBI), also, plays an important role in the current
experiments and with fusion alphas for future experiments such as ITER. Commonly, previous RF simulations in the plasma core are neglecting the SOL plasma and they compute the RF field in a 2D domain assuming one single toroidal wave number. State-of-art RF SOL/antenna simulation is yet limited to a relatively small volume in front of the antenna, and it involves significant physics simplification such as stratifying antenna strap structure and/or treating the antenna front volume as vacuum. This paper, instead, examines the full 3D device geometry including realistic antenna geometry in order to capture the experimental situation including the 3D effects and the antenna plasma interaction in the SOL plasma and, at the same time, the core wave propagation. In particular, in this work we employ the Petra-M code (1, 2), which is a newly developed state-of-the-art generic electromagnetic simulation tool for modeling RF wave propagation based on MFEM [http://mfem.org], open source scalable C++ finite element method library. Furthermore, we report for the first time the interaction between FW and fast ions in a 3D geometry including both the SOL plasma and the full-orbit effects, which are very important for NSTX-U and future experiments such as ITER by using the wave field evaluated by Petra-M in the SPIRAL full-orbit following particle code (3).
3D full wave simulations of HHFW regime in NSTX-U plasmas are analyzed (4). In Figure 1, one can see the realistic geometry employed in the simulations: the HHFW antenna geometry (figure (a)) and the 3D mesh for NSTX-U (figure (b)). Unlike the 2D simulations where a single wave number is employed, the 3D simulations allow us to study the 3D effects of the realistic antenna geometry and the role of the antenna spectrum, which can be modified by employing different antenna phasing between straps as done experimentally. Unlike the previous 2D full wave simulations (5, 6, 7, 8), 3D full wave simulations do not show strong cavity modes in the SOL plasma emphasizing the need to consider the 3D geometry. A scan of the antenna phasing shows a strong interaction between FWs and the SOL plasma for lower antenna phasing, which is consistent with previous NSTX HHFW experimental observations (9) (see Figure 2). This strong interaction for lower antenna phasing can lead to a potential RF power loss in the SOL plasma. Figure 2 shows also that (i) the wave field tends to propagate in the toroidal direction more that the radial direction when the antenna phasing is
increased. This is expected because the ratio of the parallel and perpendicular group velocity goes like the parallel refractive index ($N_{\parallel}$), namely, for higher $N_{\parallel}$ the waves should travel toroidally more; (ii) some wave field propagation appears on the planes above and below the mid-plane for all antenna phasing. Finally, the effect of the 3D wave field on the fast ion population from both the radial and tangential NBI beams in NSTX-U is quantified by using the 3D field obtained from the Petra-M simulations in the SPIRAL full-orbit following particle code. Figure 3 shows the $E_z$ component of the wave electric field in a toroidal cross-section on the mid-plane of NSTX-U overlaid with a single fast ion orbit as obtained by SPIRAL. From this figure it is clear that the fast ions should be affected mainly in the region in front of the antenna where the wave electric field is strong. In order to show this point we used an ensemble of 40k particles in SPIRAL assuming an initial Maxwellian distribution with a fast ion temperature $T_{\rm FI}\sim 25$ keV and a central fast ions density of $n_{\rm FI}\sim 2 \times 10^{18}$ m$^{-3}$. At the same time, we used the three components of the wave electric field evaluated by Petra-M in the geometry shown in Figure 1 as a perturbation of the equilibrium field in the Lorentz equation. Figure 3 shows the contour plot of the fast ions power deposition evaluated by SPIRAL including the full 3D RF wave field evaluated by Petra-M. It clearly appears that the interaction between fast waves and fast ions occurs mainly in front of the antenna, as expected. This result demonstrates how 3D effects are important in these simulations for fast ions studies in which RF is included. Generally, a 2D wave field obtained with only a single toroidal wave number is used assuming toroidal symmetry. This approximation results in acceleration/deceleration of the fast ion due to RF everywhere in the torus and not only in the region where the RF wave field is localized. Further studies with realistic fast ion distribution functions from the NBI beams in NSTX-U will be discussed.
This work is supported by U.S. DOE Contract # DE-AC02-09CH11466
References
(1) S. Shiraiwa et al., EPJ Web of Conferences 157, 03048 (2017).
(2) S. Shiraiwa et al., this conference.
(3) G. J. Kramer et al., Plasma Phys. Control. Fusion 55, 025013 (2013).
(4) N. Bertelli et al., Invited talk at the 23rd RF Power in Plasmas Conference (China, 2019).
Accepted to be published in AIP Conf. Proceeding (2019).
(5) D. Green et al., Phys. Rev. Lett. 107, 145001 (2011)
(6) N. Bertelli et al., Nucl. Fusion 54, 083004 (2014).
(7) N. Bertelli et al., Nucl. Fusion 56, 016019 (2016).
(8) E.-H. Kim et al., Phys. Plasmas 26, 062501 (2019).
(9) J. C. Hosea and et al., Phys. Plasmas 15, 056104 (2008).Speaker: Nicola Bertelli (Princeton Plasma Physics Laboratory) -
Access to high current (IP >~ 1 MA) relativistic electron (RE) beams in the DIII-D and JET tokamak reveal excitation of current-driven (low safety factor) kink instabilities that promptly terminate the RE beam on an Alfvenic time-scale [Ref. 1], a phenomenon first observed during the early JET carbon wall operation period. Unlike past results however, this phenomenon when combined with the injection of D2 is found to minimize heating of the first-wall, offering an unexpected alternate pathway to RE beam mitigation without collisional dissipation. Minimized heating is explained by two synergistic effects. First, during large-scale kink events experiments find [Ref. 1] and MHD orbit-loss modeling calculates [Ref. 2] a significant increase in the wetted area of the RE loss. Second, as previously identified at JET and DIII-D, the fast kink loss timescale precludes RE beam regeneration and the resulting conversion of magnetic energy to kinetic energy. During the termination the runaway kinetic energy is lost to the wall but the current transfers to the cold bulk, enabling benign Ohmic dissipation of the magnetic energy on longer timescales. In addition, this novel approach exploits D2 secondary injection, which is found to be a necessary ingredient. Practically, D2 enables access to high IP (via reduced collisionality & resistivity) and directly inhibits RE beam regeneration. Sufficient D2 purity to avoid heating is achieved by both shattered pellet secondary injection and conventional massive gas valves. While unexpected, this path scales favorably to fusion-grade tokamaks and offers a novel RE mitigation scenario in principle accessible with the planned actuators of ITER.
The baseline strategy for mitigating a fully-formed RE beam in ITER is the injection of massive quantities of high-Z material (Argon) using shattered pellet injection (SPI). This collisional approach is challenged by low rates of high-Z assimilation into the RE beam as well as the acceleration of the vertical instability. Recent observations on DIII-D [Refs. 1,2] and JET reveal a possible alternate pathway for RE mitigation by exploiting current-driven kink instabilities naturally occurring at low safety factor (q_a). Predictions (by the DINA code) of ITER’s RE beam evolutions mostly cross the low qa stability boundary as IP is decreasing, with some variation based on the size and localization of the RE seed [Ref. 3]. In present tokamaks, increasing the RE current is utilized to controllably access low qa. RE position control is also used, and the relevance to a vertically unstable ITER beam will be discussed.
After a period of rising IP, both DIII-D and JET observe the RE beam to be promptly terminated (at t=tloss) by MHD activity with Alfvenic (sub-ms) growth rate, as shown in Fig. 1(a-b). Injection of D2 reduces the RE resistivity and allows the solenoid to push to low qa. DIII-D magnetic measurements and MHD stability modeling with the MARS-F code find that the observed MHD is consistent with current-driven resistive external kinks. The kink stability is found to be similar to regular plasmas (Fig. 1[c]). Surprisingly, while the potential for damage is substantially larger due the higher IP (Fig. 1[d]), wall heating during these fast terminations is negligible and below the noise floor of the JET infrared cameras (Fig. 1[e]).
Modeling and experiment support the absence of wall heating as being due to both an increase of the wetted area during the MHD-driven RE loss and an inhibition of the conversion of magnetic energy into RE kinetic energy normally observed during RE loss events. Figure 2(a)-(c) presents orbit loss modeling of DIII-D with MARS-F. As the magnitude of the kink instability increases all possible RE orbits are eventually deconfined at a level consistent with magnetic measurements (dBmeas). Tracing the location of the wall strike for these orbits (Fig. 2[d-e]) finds an increase in the spatial extent of the possible RE loss, enabling the energy flux per unit area to be significantly decreased as compared to the conventional very localized RE strike. Spatially distributed HXR measurements (Fig. 2[f-g]) confirm a wider spatial extent of the HXR loss pattern during low qa loss events. The computed complete de-confinement of all orbits also supports the absence of RE beam regeneration, as the electric field induced by the RE loss cannot accelerate further REs. The RE current instead transfers to the cold bulk population, which then dissipates the magnetic energy resistively on a longer timescale. Further studies in JET reveal that absence of regeneration requires a high purity of D2, for reasons that are not yet fully understood.
To conclude, observations reveal a benign mitigation of high IP, fully-formed RE beams. While further experimentation and modeling is needed, these results offer an unexpected and novel pathway for RE mitigation in ITER: inject D2 and excite large MHD by crossing the kink stability boundary.
Work supported by the US DOE under DE-FC02-04ER54698, DE-SC0020299 and by the ITER Organization (TA C18TD38FU) and carried out within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission or the ITER Organization. [Ref. 1] C. Paz-Soldan et al, Plasma Phys. Control. Fusion 61 054001 (2019). [Ref. 2] Y.Q. Liu et al, Nucl. Fusion 59 126021 (2019). [Ref. 3] K. Aleynikova et al, Plas. Phys. Rep. 42 486 (2016).
Speaker: Carlos Paz-Soldan (General Atomics) -
Joint research on the tokamaks DIII-D and EAST demonstrates a successful integration control of divertor detachment with excellent core plasma performance, a milestone towards solving the critical Plasma-wall-interaction (PWI) issues for ITER and future reactors. DIII-D has achieved actively controlled fully detached divertor with low plasma temperature ($T_{e,div} \le $ 5 eV across the entire divertor target), low particle flux (DoD > 3) and heat flux reduction > 85%, simultaneous with very high core performance ($\beta_N$ ~ 3, $\beta_P$ > 2 and $H_{98}$ ~ 1.5) in the high $\beta_p$ scenario developed for ITER steady state operation. EAST has achieved actively controlled detachment with $T_{e,div}$ ~ 5 eV and $H_{98} \ge$ 1.1 in various H-modes, i.e., normal ELMy H-mode, grassy ELMy H-mode and high $\beta_P$ H-mode scenarios. The divertor detachment with $T_{e,div} \le $ 5 eV is highly desirable for suppression of the erosion in reactor-grade devices.
In DIII-D, full detachment was achieved, for the first time, with very high $H_{98}$ ~ 1.5 by utilizing feedback controlled impurity seeding in the high $\beta_P$ scenario $[1]$. The self-organized with co-existence of internal transport barrier (ITB) and edge transport barrier (ETB, or pedestal) benefits access to detachment without degradation of core performance. The scenario makes detachment easier by optimizing impurity seeding at relatively higher $q_{95}$ and lower separatrix density, with very high core confinement enabled by a large-radius ITB. Fig. 1 shows a fully detached plasma with excellent core confinement in the high $\beta_P$ scenario on DIII-D with nitrogen seeding. In the full detachment phase, the steady-state peak heat flux measured by infra-red (IR) camera is reduced by > 85%, and the particle flux measured by divertor Langmuir probes is reduced by > 80%. Nitrogen seeding is more efficient for full detachment than neon-seeding experiments and leads to a small reduction of ETB. The demonstration of compatibility with divertor detachment adds to the well-known benefits of high bootstrap current fraction and low disruption risks, positioning the high $\beta_P$ scenario as a prime candidate for steady-state fusion reactor operation.
In EAST, actively feedback controlled H-mode detachment with simultaneous $T_{e,div}$ ~ 5 eV and $H_{98}$ > 1.1 was achieved using either divertor neon or argon seeding in USN configuration, i.e. divertor detachment operating on an ITER-like tungsten divertor, was maintained with energy confinement quality higher than the ITER baseline scenario, as shown in Fig. 2. The feedback detachment waveforms follow closely the preset targets. Experiments using injection of different impurity species for radiative divertor and detachment were performed to understand the extrapolation to future burning plasmas on ITER and CFETR. Neon exhibits good core-divertor integration, while argon is more efficient for detachment with a slight loss of confinement. In addition, different detachment feedback controllers including degree of detachment (DoD) via Langmuir probe measured divertor particle flux $[2]$, divertor $T_{e,div}$ $[3]$, plasma radiative power $P_{rad}$ $[4]$, $T_{e,div}$ + $P_{rad}$ have all been developed successfully in EAST.
The success of actively controlled detachment is achieved with different divertors and benefits from the divertor closure and pumping. A closed divertor is beneficial for neutral trapping and thus detachment access $[5-6]$, as well as the drift effect in the SOL and divertor volume. Furthermore, the sustained detachment compatible with core high performance is independent of plasma heating schemes and divertor materials, i.e., EAST with ITER-like tungsten wall + RF heating and DIII-D with carbon wall + NBI heating, respectively.
The compatibility of efficient divertor detachment with a high-performance core is vital to the realization of magnetically controlled fusion energy. The significant progress in DIII-D and EAST show that the potential for a high-performance core plasma suitable for long pulse operation of fusion reactors can be achieved with controlled PWI. The present joint DIII-D/EAST methodology will be used for minute-time-scale long pulse operation on EAST in near future and thus offers a useful solution for ITER, CFETR as well as future fusion energy power plants.
This work was supported by National Key R&D Program of China under 2017YFE0301300 and the U.S. DOE under DE-FC02-04ER54698, DE-AC04-94AL85000, DE-NA0003525, DE-AC52-07NA27344.
$[1]$ L. Wang, H. Q. Wang$^*$, A. M. Garofalo et al, “First Observation of Fully Detached Divertor Compatible with Improved Core Confinement in Tokamak”, to be submitted to PRL
$[2]$ L. Wang$^*$ et al., Nucl. Fusion 59, 086036 (2019)
$[3]$ D. Eldon, …, L. Wang et al., this conference (Poster submission)
$[4]$ K. Wu, …, L. Wang$^*$ et al., Nucl. Fusion 58, 056019 (2018)
$[5]$ J. B. Liu, L. Wang$^*$ et al., Nucl. Fusion 59, 126046 (2019)
$[6]$ X. J. Liu, L. Wang$^*$ et al., Phys. Plasmas 16, 102510 (2019)Disclaimer: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
Speaker: Liang Wang (Institute of Plasma Physics, Chinese Academy of Sciences (ASIPP)) -
Novel disruption prevention solutions spanning the range of control regimes have been developed and tested on DIII-D to enable ITER success. First, the disruption risk during fast, emergency shutdown after large tearing and locked modes can be significantly improved by transitioning to a limited topology during shutdown. More than 50% of limited shutdowns reach a final normalized current $I_N$ < 0.3 before terminating (Fig. 1), scaling to the 3 MA ITER requirement. This is in contrast to diverted shutdowns, the majority which disrupt at $I_N$ > 0.8. Despite improvements, these results highlight the critical importance of early prevention. Second, a new control algorithm has been developed and tested for regulating nearness to stability limits in real-time. Its first application reliably prevented vertical disruptions (VDEs) by adjusting plasma elongation ($\kappa$) and the inner-gap between the plasma and inner-wall in response to a neural-network based VDE open-loop growth-rate ($\gamma$) estimator (Fig. 2). Third, a novel emergency shut down method has been developed which excites MHD instabilities to form a warm, helical core post-thermal quench {1}. The current quench extends to ~100ms and avoids VDEs and runaway electron generation (Fig. 3). Novel real-time machine learning disruption prediction {2} has been integrated into the DIII-D plasma control system (PCS), and a multi-mode MHD spectroscopy technique has been developed which is real-time compatible.
The disruption risk during emergency shutdown in response to large tearing and initially-rotating locked modes is significantly improved by using a limited topology during shutdown. In over 100 discharges, an early, rapid shutdown was triggered on large n=1 rotating mode amplitude or non-rotating locked-mode amplitude to assess its effectiveness under these dangerous conditions. Fast $I_p$ ramp-down rates of 2-3 MA/s and sustained, modest NBI heating were applied, which minimized disruptivity in a previous large-scale study of pre-programmed shutdowns {3}. As the stored magnetic energy is proportional to $I_p^2$, the appropriate metric for emergency shutdown success is the final value of Ip reached before current quench. ITER 15 MA discharges must reach $I_p$ < 3 MA before disrupting, corresponding to $I_N$ < 0.3. Rapidly transitioning to a limited topology in shutdown can dramatically reduce the final $I_N$ reached before terminating, with as much as 50% of shots below the ITER limit (Fig. 1). This is in contrast to maintaining a diverted topology, which results in less than 20% of discharges reaching safe final currents, and the majority disrupting above $I_N$ > 0.8. ITER tolerance to a limited topology is an open question, and likely depends on whether an unintentional H-L back-transition has already occurred before attempting an emergency shutdown. Nonetheless, the risks must be weighed against the danger of high-current disruption. While significant improvement is found with limited emergency shutdowns, early prevention and reliable mitigation will be critical to meet ITER disruption tolerances.
A new proximity-to-instability regulation algorithm (the “Proximity Controller”) has been developed for real-time disruption prevention on DIII-D, and first applied to prevent VDEs in experiment. The VDE open-loop growth-rate ($\gamma$) was robustly regulated at safe levels of 300-400/s for > 2s, despite a pre-programmed ramp in $\kappa$ intended to induce VDEs (Fig. 2). Without the controller active, these discharges ended in disruption. However, disruptions were prevented in all discharges which used VDE protection via the Proximity Controller, until the time the controller was intentionally disabled. The open-loop-$\gamma$ was estimated using a novel, real-time, neural-network-based reduced model trained on tens of thousands of time slices, and included uncertainty estimation. The growth-rate was used to continuously modify targets for the plasma elongation and inner-gap. The controller is being extended to additional real-time stability calculators and control targets for 2020 experiments.
A new soft-landing alternative has been developed that routinely enables high-current ($I_p$ ~1.7 MA) discharges to survive major disruptions (Fig. 3). The technique destabilizes core MHD instabilities in the post-TQ plasma with the aid of applied 3D fields, after which flux surface healing is observed. A helical core is promptly formed and warmed by neutral beam and ohmic heating {1}. In combination with impurity puffing, the plasma current quench duration greatly extends to ~100 milliseconds with a warm core, avoiding the generation of runaway electrons for optimal mitigation. This points to a safe emergency termination alternative. The underlying physics mechanism is under active investigation.
Advanced machine learning and multi-mode MHD spectroscopy (MMS) techniques for plasma stability monitoring have been developed for real-time use. A random-forest machine learning algorithm for disruption prediction has been implemented into the DIII-D PCS {2} and used to trigger fast, emergency shutdowns for disruption prevention (see C. Rea et al this IAEA FEC). MMS based on subspace system identification has been demonstrated as feasible for detecting plasma eigenmodes in real-time, and is being implemented in the DIII-D PCS.
ITER’s low disruption tolerances demand comprehensive solutions, including prevention, suppression, and emergency responses for likely physics problems. Results presented here were enabled by a focused effort, the Disruption Free Protocol, in DIII-D’s 2019-20 campaign to complement disruption prevention experiments with a large piggy-back effort. In addition to testing novel techniques, it’s estimated to have directly prevented 33+ disruptions in 2019.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698.
{1} X.D. Du et al 2019 Nucl. Fusion 59 094002
{2} C. Rea et al 2019 Nucl. Fusion 59 096016
{3} J.L. Barr et al 2018 Proc. 27th IAEA FEC EX/P6-21Speaker: Jayson Barr (General Atomics) -
An optimized pedestal regime called the Super-H Mode (SH-mode) is leveraged to simultaneously couple a fusion relevant core plasma with a scrape-off layer appropriate for realistic reactor exhaust solutions. Recent DIII-D experiments have expanded the operating space from previous studies of the SH regime and investigated optimization of impurity seeding, deuterium gas puffing, 3D magnetic perturbations, and plasma shape. Experiments demonstrate gas puffing and impurity seeding lead to a radiative mantle and low divertor temperatures (< 15eV) that are compatible with maintaining SH-mode and have marginal impact on pedestal and core pressure. An important recent result is that access to the SH-mode has been achieved in shapes matching JET plasmas with moderate plasma triangularity ($\delta_{avg} \sim 0.4$), providing a pathway for increased performance for the JET D-T campaign as well as increased confidence in the EPED predictions for SH-mode access in ITER.
Plasma shape is a key parameter impacting pedestal stability, and when SH-mode access is marginal, small changes in triangularity and aspect ratio can lead to an increase in global metrics like plasma stored energy through pedestal optimization. Previous experiments maximized plasma triangularity and volume in the SH regime in order to maximize pedestal and core performance; however, recent experiments show SH access can still be obtained at moderate plasma shaping. Figure 1 shows the pedestal electron pressure and density in two JET similar shapes, with one having an increased plasma triangularity from 0.3 (gray) to 0.4 (red). This change in triangularity opens access to the SH-mode channel, allowing a higher pedestal at the same density, and higher stored energy even with a slightly reduced plasma volume. The relatively modest plasma triangularity compared to the double null SH experiments leads to pedestal pressures which are farther from ideal $\beta$ limits, allowing plasma trajectories deep in the SH channel to be maintained in a stationary state. Robust SH-mode access in lower single null shapes with intermediate levels of triangularity implies applicability for potential use as a target scenario in both JET and ITER. Figure 1 indicates that SH-mode is compatible with JET plasma shapes and could increase plasma stored energy in the upcoming D-T campaign at the same engineering parameters.
By employing a dynamic density trajectory and shape control, peeling and ballooning physics can be decoupled in the pedestal {1}. The peeling-limited pressure pedestal reaches ~20-30kPa on DIII-D and up to ~80kPa on C-Mod {2}, even with strong gas puffing. The pedestal maintains low collisionality with a high separatrix density {3}, which is important for achieving a low heat flux to the divertor plate without degradation of the pedestal pressure. Recent experiments on DIII-D have employed co-current beam injection at full magnetic field ($B_t=2.1-2.2T $) and current ($I_p=1.4-2.0 MA$) in both closed and open divertor configurations. Initial indications show that a slightly larger divertor volume with a longer leg between the x-point and strike points allows more power to be radiated in the scrape off layer and pedestal regions, and to be excluded from the core more effectively. Advanced control algorithms {4,5} simultaneously optimize the line average density and divertor radiative power. Introducing 3D magnetic perturbations that pump out particles actively controls the line average density and allows the SH-mode plasmas to enter an extended stationary phase for 2s with sustained pedestal pressure and controlled impurity content. Dual seeding with deuterium and nitrogen radiate power near the separatrix and reduce the divertor heat flux and temperature to promote integration with requirements of plasma facing components, as shown in Fig. 2 for a lower biased double null ($\delta_{avg} \sim 0.57$). The SH-mode plasma has a pedestal electron temperature of ~1keV with divertor temperature <15eV. Feedback control on the divertor radiation was employed for optimal nitrogen seeding to maintain a steady dissipation of ~42% of injected power contained to the divertor and pedestal regions while maintaining a pedestal with $\beta_N^{ped}~0.8$ and core plasmas with $\beta_N^{core}=2.2$ and $W_{MHD} \sim 2MJ$.
{1} W.M. Solomon, et. al., Phys. Rev. Lett. 113, 135001 (2014)
{2} J. Hughes, et. al., NF 58, 112003 (2018)
{3} P.B. Snyder, et. al., Nucl. Fusion 59, 086017 (2019)
{4} D. Eldon et al., Nucl. Mater. and Energy 18, 285-290 (2019)
{5} F. Laggner et al., submitted Nucl. Fusion 2019Speaker: Theresa Wilks (UsMIT) -
New scaling laws and modeling, developed at DIII-D and benchmarked with data from JET and KSTAR, provide a path for projecting Shattered Pellet Injection (SPI) performance to ITER, while improved understanding of higher-order effects such as asymmetries better constrain the expected behavior. In the limit of radiative shutdown by high-Z impurity injection, the volume-averaged performance of disruption mitigation by SPI can be largely understood through global energy balance, without consideration of magnetohydrodynamic (MHD) effects. Particle assimilation rates are quantitatively predicted, with good agreement between simulation, empirical scalings, and a large database of experimental measurements (Figure 1). Dual-SPI shutdown characteristics (with two injectors at two different ports) are also well captured by simulations based on the energy-balance model, including additional density rise due to added deuterium. Radiation asymmetries due to the rapid and localized SPI particle source are experimentally observed, but current estimates of the peaking factor remain below ITER limits. For runaway electron (RE) mitigation, pure argon SPI is able to completely dissipate fully-avalanched post-disruption RE beams. These results provide a basis for optimizing injection scenarios for the ITER disruption mitigation system.
New simulations show that the global evolution of the SPI shutdown is primarily governed by energy balance, rather than by MHD effects, with good agreement found with data from DIII-D, JET, and KSTAR experiments. The SPI shutdown is modeled through modification of the 0D KPRAD code {1}, which tracks volume-averaged energy balance during the disruption, incorporating a particle source determined by the shielding-limited species-dependent ablation {2} of the shattered pellet plume. A large collection of thermal quench (TQ) and current quench (CQ) characteristics observed in experiments are explained through robust agreement with simulations based on this model (Figure 1). Particle assimilation rates are quantitatively predicted by these simulations, and compare well with measured densities across a large database of SPI shots. The resulting CQ rates in all three devices are accurately predicted, as the mixture of high-Z and low-Z injection species is systematically varied (Figure 2).
Empirical regression scalings, which are a function of pre-SPI plasma parameters, also successfully reproduce measured CQ densities in DIII-D (Figure 1, right). The dependence of the assimilation primarily on global plasma parameters is consistent with the modeling assumption that energy-balance plays a dominant role in the process (although possible hidden variables, evidenced by a small number of outliers, are under study). Electron temperature is found to be the dominant dependence for early assimilation of the shattered pellet, while Ohmic dissipation of the poloidal magnetic energy is important later for sustaining the ionization of the CQ plasma. Both of these quantities are expected to increase favorably for ITER (towards higher density), although scaling with device-size has not been considered (and will tend towards lower density).
The behavior of dual-SPI shutdowns (with two injectors from two different ports) {3} is also well captured by the 0D KPRAD simulations, indicating that 3D effects related to the multiple injection locations play a lesser role. In particular, the ability of additional deuterium to further raise the density (over pure neon SPI) is quantitatively matched by simulation, as is the measured impact on neon assimilation {3}. These results suggest that 3D effects do not have a significant impact on the globally averaged mitigation, and are instead important primarily for heat load asymmetry effects.
Such peaking of heat loads due to TQ radiation asymmetries is detected experimentally in DIII-D, but current estimates of the peaking remain below ITER limits (Figure 3). The asymmetries are broadly centered on the injection port, and are due to the rapid and localized SPI particle source (in contrast to MHD-driven asymmetries during massive gas injection). The toroidal peaking factor (TPF) is estimated from comparisons with ray tracings of various TQ emissivity profiles, and the current estimated value of 1.33 remains below ITER limits (TPF<2). However, refinements of these estimates by separately considering the TQ and CQ heating may lead to higher values.
For the mitigation of REs, pure argon SPI (enabled by a new mechanical punch) is able to completely dissipate centered post-disruption RE plateaus in DIII-D, in contrast to previous neon SPI which was accompanied by a deuterium ‘shell/grease’ layer {4}. This supports the picture that injection species is the dominant factor determining the effectiveness of ‘second injection’ for RE dissipation (rather than gas/pellet form), and suggests that upper port plugs in ITER with a higher-bend-angle shatter tube may be effective for this application. RE current dissipation due to argon and neon SPI, albeit partial, is also observed at JET.
This DIII-D research, in collaboration with the JET and KSTAR SPI programs, continues to expand SPI physics understanding and to improve projection of these results to ITER. The newly developed scalings and simulations provide guidance for future experimental and leadership-class-computational studies, and provide a path for extrapolating SPI performance beyond a single device, towards ITER.
{1} D.G. Whyte, et al., Journ. Nucl. Mater. 313 (2003) 1239
{2} P.B. Parks, 7th Annual Theory and Simulation of Disruptions Workshop (2017) Princeton, USA
{3} J.L. Herfindal, et al., Nucl. Fusion 59 (2019) 106034
{4} D. Shiraki, et al., Nucl. Fusion 58 (2018) 056006Work supported the US DOE under DE-FC02-04ER54698, DE-AC05-00OR22725, DE-FG02-07ER54917, and DE-AC52-07NA27344, and by the ITER Organization (TA C18TD38FU) and carried out within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. Views and opinions expressed herein do not necessarily reflect those of the European Commission.
Speaker: Daisuke Shiraki (Oak Ridge National Laboratory) -
ITER adopts a strategy that distributes radiated power evenly during the disruption mitigation and reduces the time to prepare pellets, using simultaneous multiple shattered pellet injections (SPIs)$^1$. However, since there were no existing devices with perfectly symmetric SPIs, as planned in ITER$^2$, sufficient studies have not been conducted on the effects of simultaneous multi-injections. To verify the feasibility of the disruption mitigation strategy of ITER, KSTAR installed two SPIs with exactly same design at toroidally opposite locations as shown in figure 1$^3$. Each SPI can use three barrels of different diameters to control the number of injected particles, selectively. The species used can vary deuterium, neon, argon, or their mixture depending on the mitigation purpose such as thermal load mitigation or runaway electron suppression/mitigation.
In 2019, we mainly examined the difference in disruption mitigation by intentionally changing the arrival times of two SPIs to assess possible jitter effect among multiple SPIs. As shown in figure 2a), the current quench rate changes proportionally as the time difference varies from several percent to several tens of percent of the thermal quench (TQ) duration (1~2 ms). Through this, it was experimentally demonstrated that more energy can be radiated when multiple SPIs are injected simultaneously, as planned in ITER. The result resolved an ambiguity about the simultaneous multi-injections observed in previous experiments performed with two SPIs only 120 degree apart$^4$. Moreover, the sensitivity to time difference identified by KSTAR experiments provided guidance in designing the ITER disruption mitigation system (DMS). The effect of multiple injections observed in KSTAR experiments is being analyzed using KPRAD, a radiation cooling code.
On the other hand, in the disruption mitigation process, it is also important to form a high plasma density to prevent the transfer of magnetic energy toward runaway electrons. For this study, a dispersion interferometer with short wavelength in 1064 nm was developed and installed to measure the density during the mitigation process that is one or two order higher than that of conventional plasma. In the case of dual SPIs, it was measured the peak density $1.2 \times 10^{21} m^{-3}$ near TQ end, which is almost twice the value of single SPI, as desired.
Excessive particle injection of SPI and subsequent radiation create a strong MHD mode in the plasma. Conversely, this MHD mode has a significant impact on the behavior of the injected particles. As shown in figure 3, the well-synchronized dual SPIs exhibited much mild MHD mode than the asynchronized SPIs. Preliminary numerical analysis of the SPI-induced MHD mode (not shown here) indicated that ideally symmetric injection of two SPIs causes negligible odd perturbations (e.g., $n=1$) and causes significant even perturbations (e.g., $n=2$).
The disruption mitigation with SPI is complex phenomena depending on the plasma and SPI parameters. The study of the interaction with pre-existing MHD mode such as the cause of disruption is also important for establishing a realistic mitigation strategy. Among the various themes of DMS, we plan to focus firstly the multi-injections from different toroidal positions with varying the above-mentioned parameters as well as the multi-barrel injections from same poloidal/toroidal position in accordance with the plan of ITER DMS. For the purpose, the largest size barrel will be changed to middle size one to simulate ITER SPIs which have all same size barrels. It is expected to provide the data that underlie the design of the ITER DMS.
References:
$^1$ L.R. Baylor et al., 2009 Nucl. Fusion 49 085013.
$^2$ M. Lehnen et al., 2018 IAEA fusion energy conference.
$^3$ S.H. Park et al., 2020 Fusion Eng. Des.154 111535.
$^4$ J.L. Herfindal et al., 2017 APS division of plasma physics meeting.Speaker: Jayhyun Kim (National Fusion Research Institute) -
Diverted discharges at negative triangularity on the DIII-D tokamak (figure 1a) sustain normalized confinement and pressure levels typical of standard H-mode scenarios ($H_{98,y2}\simeq 1$, $\beta_N\simeq 3$) without developing an edge pressure pedestal (figure 1b), despite the auxiliary power far exceeding the L→H power threshold expected from conventional scaling laws. The power degradation of confinement is substantially weaker than the ITER-89P scaling (figure 2), resulting in a confinement factor that improves with increasing auxiliary power. The absence of the edge pedestal is beneficial in several aspects, such as eliminating the need for active mitigation or suppression of Edge Localized Modes (ELMs), low impurity retention and a reconstructed Scrape-Off Layer (SOL) heat flux width at the mid-plane that exceeds the measured H-mode value by approximately 50%.
High confinement L-mode edge plasmas at negative triangularity previously obtained in inner-wall limited plasmas on DIII-D$^1$ have been recently extended to diverted configurations by creating a novel lower single null equilibrium. This plasma shape features top-bottom averaged triangularity that is negative and is compatible with operations at high auxiliary power on DIII-D. In particular, the X-point was placed at the outermost radial location that allowed strike points not to impinge on the outer wall, thereby making a near zero lower triangularity, while the separatrix above the mid-plane features strong shaping with upper triangularity $\delta_u$ = -0.4. The successful creation of this unusual shape was streamlined by the flexibility of the DIII-D control system, which employed recently installed bipolar power supplies for the creation and control of the X-point as well as a full predictive simulation of the discharge that minimized the time allocated for scenario development.
Plasmas maintained an approximately constant line averaged density and the effective charge radial profile, $\rm Z_{eff}$, was in the range 1.4-2; this is a significant improvement over previous inner-wall limited experiments$^2$ for which the line averaged density increased with beam fueling and $\rm Z_{eff}$ was as large as three. Discharges were executed at full field, 2 T, and plasma current of 0.9 MA, yielding a safety factor of 4.5 at 95% of the normalized poloidal flux.
The lack of an H-mode transition postulates that the L→H power threshold increases strongly with increasing negative triangularity. Neither inner-wall limited nor diverted negative triangularity plasmas transitioned to a standard H-mode regime despite the fact that the power flow crossing the separatrix greatly exceeded the expected threshold. In one diverted case, when the upper triangularity was accidentally relaxed to $\delta_u$ = -0.2, the standard L→H transition was observed with 2 MW of net auxiliary power, whereas for $\delta_u$ = -0.4 plasmas maintained L-mode edge pressure profiles despite 10 MW of net injection. Previous experiments on the TCV tokamak using a similar shape observed H-mode transitions with $\delta_u$ ~ -0.2, although more extreme values of the upper triangularity were not attempted$^3$. This result needs further study and could make L-mode edge operation a viable candidate for operation in future reactors, provided that the L→H power threshold will remain large enough to prevent plasmas from developing an edge pedestal.
Negative triangularity discharges sustained H-mode grade confinement and pressure levels despite maintaining relaxed edge pressure radial profiles typical of L-mode scenarios. More specifically, plasmas routinely featured relatively high normalized pressure levels up to $\beta_N$ = 3, or $\beta_N/l_i$ = 3, and confinement factor $\rm H_{98,y2}\geq$ 1 while being intrinsically free of ELMs. The confinement factor progressively improved with increasing auxiliary power as a consequence of a much lower power degradation than that predicted by the ITER-89P scaling law, as displayed in figure 2, in substantial agreement with previous inner-wall limited experiments.
The ratio of the impurity confinement time to the energy confinement time was measured to be lower than that in standard H-mode regimes. The particle confinement time of Aluminum was computed by laser ablating suitable targets and resulted in $\tau_P/\tau_E$ of order unity. As opposed to H-mode regimes, for which such ratio typically reaches or exceeds three, this scenario makes impurity retention less problematic but may make fueling more difficult.
The heat flux width in the Scrape-Off Layer was measured to exceed values measured during H-mode phases. The radial profile of the heat flux was measured by infrared cameras on the divertor plate and mapped at the mid-plane, where direct measurements were not available. While the heat flux width in inter-ELMs phases of the only negative triangularity H-mode plasma obtained matches almost exactly the value predicted by the Eich scaling law, all the L-mode discharges featured heat flux widths larger by about 50%, consistent with increased fluctuation levels in the SOL compared to the H-mode regime.
These L-mode edge plasmas, in view of their H-mode grade confinement and normalized pressure levels, weak power degradation, intrinsic ELM free nature, larger than standard SOL heat flux width as well as low impurity retention, appear to be a viable candidate for operations in future magnetically confined fusion reactors. Technical advantages regarding the manufacturing and maintenance of the divertor and internal poloidal field coils are additional attractive features of this scenario$^4$.Work supported by US DoE under DE-FG02-94ER54235, DE-FC02-04ER54698, DE-FG02-97ER54415, DE-FG02-04ER54761.
$^1$ M.E. Austin et al 2019, Phys Rev. Lett. 122 115001
$^2$ A. Marinoni et al. 2019, Phys. Plasmas 26 042515
$^3$ A. Pochelon et al. 2012, Plasma Fusion Res. 7 2502148
$^4$ M. Kikuchi et al. 2019, Nucl. Fusion 59 056017Speaker: Alessandro Marinoni (Massachusetts Institute of Technology) -
For the first time, experiments on the DIII-D tokamak have demonstrated electron cyclotron current drive (ECCD) with more than double the efficiency of the conventional outside launch by using a novel top launch geometry (figure 1), as predicted by linear ray tracing and quasi-linear Fokker-Planck simulations. Studies have shown that off-axis current drive is a requirement for a steady-state reactor in the Advanced Tokamak (AT) regime$^{1,2}$; however, driving current off-axis efficiently remains a challenge. Launching electron cyclotron waves from the high field side of the plasma, but the low field side of the resonance, with large toroidal steering in a plane nearly parallel to the resonance layer (illustrated in figure 2) is found to greatly increase the ECCD efficiency at mid-radii compared to conventional outside launch. The higher ECCD efficiency is due to 1) selective EC wave damping on higher v$_{||}$ electrons, and 2) longer absorption path lengths to compensate for inherently weaker absorption at higher v$_{||}$. DIII-D experiments using a prototype top launch system with a fixed mirror have established these two tenets through scanning v$_{||}$ of the wave-particle interaction by varying the magnetic field B$_T$. Power deposition measurements show that the absorbed EC power decreases for higher v$_{||}$ interaction (lower B$_T$), giving rise to a “sweet spot” (optimal B$_T$) for maximum ECCD efficiency at $\rho$~0.5 (figure 3) where the higher current drive efficiency for higher v$_{||}$ is balanced by sufficient absorption. Simulations of ‘top launch’ ECCD for FNSF, DEMO and CFETR support it as an improved efficiency off-axis current drive technique for future fusion reactors$^{3-5}$.
Top launch ECCD with a long wave-electron interaction zone and a large Doppler shift ensures strong damping on tail electrons leading to higher ECCD. Top launch ECH experiments have been done previously on TCV using radial launch and 3$^{rd}$ harmonic X-mode to heat high density plasmas; the top launch experiments on DIII-D have the different goal of efficient off-axis current drive, and the launch scheme has been uniquely optimized for this purpose. As illustrated in figure 2, EC wave from this top launch 1) propagates nearly parallel to the resonance plane and only gradually approaches the resonance, resulting in a longer absorption path; 2) suffers less trapping effects by being on HFS of the axis; and 3) allows a larger Doppler shift, thus interacting with higher energy (less collisional) electrons. To experimentally validate and characterize this approach, a prototype top launch system is installed on DIII-D with a fixed mirror angle utilizing 2$^{nd}$ harmonic damping of either a single 110GHz or 117.5GHz gyrotron with injected power between 0.5-0.6MW.The longer absorption path of top launch predicted by the TORAY ray-tracing code is measured by modulating the ECCD power and observing the electron temperature oscillations with an ECE radiometer. The measured power deposition location generally agrees with TORAY. The vertical path is verified via comparison of X-mode and O-mode deposition, where the predicted location shift between X and O (reflecting the different vertical paths due to the different absorptions) is confirmed. A much longer (i.e., three times) absorption zone for top launch ECCD compared to outside launch is also measured, consistent with TORAY. The longer interaction zone usually results in a broader deposition profile when mapped onto $\rho$ space but not always. The more parallel the EC ray is to the flux surface, the narrower the deposition profile.
Selective damping on electrons with different v$_{||}$ via top launch ECCD geometry is evidenced by the reduced absorption measured with lower B$_T$ in DIII-D experiments, as predicted by TORAY. The cold resonance moves to higher v$_{||}$ at lower Bt and the wave-electron interaction follows. Damping on tail electrons that are less collisional and drive current more efficiently is crucial for high ECCD efficiency; however, because the electron population decreases with increasing energy, the total absorption can drop far below 100%. Reduced total absorption at extreme low B$_T$ (i.e., high v$_{||}$) is observed in both L-mode and H-mode (figure 3(a)) plasmas in DIII-D.
The highest top launch ECCD efficiency is predicted and achieved when balancing higher v$_{||}$ interaction and sufficient total absorption. The ECCD profile is determined from the change in the magnetic field pitch angles measured by motional Stark effect (MSE) polarimetry. As illustrated in figure 3(b), at high Bt (low v$_{||}$) the ECCD efficiency is low despite full absorption. The measured ECCD efficiency increases with decreasing B$_T$ (increasing v$_{||}$) until the curve rolls over when too little wave energy is absorbed by too few high v|| electrons. In H-mode plasmas, the ‘sweet spot’ for highest top launch ECCD is predicted and measured at Bt~1.55T, where the driven ECCD at $\rho$~0.5 is double for top launch compared to outside launch (figure 1), consistent with the predictions from TORAY and quasi-linear Fokker-Planck code CQL3D.
Studies are underway to evaluate whether the high-beta, steady-state goals of the AT program on DIII-D can be achieved using four top-launch gyrotrons and four outside-launch gyrotrons. Initial FASTRAN simulations with self-consistent transport/pedestal/current profile modeling shows that 3MW top launch can drive as much ECCD at $\rho$~0.65 as 6+ MW outside launch in the "high qmin" AT regime owing to the near doubling in ECCD efficiency, allowing access to the highest stable $\beta_N$ (~4.5) with a non-inductive current fraction of 1. These results suggest that the combination of 3 MW top launch and 3 MW outside launch is a reasonable optimum to achieve the high-beta, steady-state goal of the DIII-D AT program.
Top launch ECCD is a promising off-axis current drive technique for future fusion reactors. Top launch ECCD shares the same reactor-relevant features of conventional outside launch ECCD, such as easy coupling to the plasma, no near-plasma antenna, and small port requirements, along with long experience in gyrotron development. Modeling for FNSF-AT shows >50% higher off-axis current drive efficiency for top launch ECCD compared to outside launch$^3$, similar to the predictions for DEMO$^4$. Greater than 35% improvement in ECCD at $\rho$~0.5 has already been found in modeling the CFETR baseline$^5$, reaching a current drive figure of merit of $\gamma$~0.16x10$^{20}$ A/m$^2$W for 14.5keV; or a dimensionless current drive efficiency of $\xi$~0.37. The experimental demonstration of doubling off-axis ECCD on DIII-D and the great enhancement found in simulations of FNSF-AT, DEMO and CFETR strongly support top launch ECCD as an exciting reactor-relevant and efficient off-axis current drive technique.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-FG02-97ER54415, DE-AC52-07NA27344, DE-AC05-00OR22725, and DE-SC0019352.
$^1$F. Najmabadi, et al, FED 38 (1997) 3; $^2$F. Najmabadi, et al, FED 80 (2006) 3; $^3$R. Prater, et al, APS-DPP (2012); $^4$E. Poli, et al, NF 53 (2013) 013011; $^5$Xi Chen, et al., EPJ Web of Conferences, 203* (2019) 01004Speaker: Xi Chen (General Atomics) -
Recent EAST experiment has successfully demonstrated long pulse steady-state high plasma performance scenario with core-edge integration since the last IAEA in 2018 $[1]$. A discharge with a duration over 60s with $\beta_P$ ~2.0, $\beta_N$ ~1.6, $H_{98y2}$~1.3 and internal transport barrier on electron temperature channel is obtained with multi-RF power heating and current drive, i.e. ~2.5 MW LHW and 0.9 MW ECH, where the plasma configuration is the upper single null with the strike points on the tungsten divertor (shown in figure 1). Loop voltage was well controlled to be zero which indicates the fully non-inductive current drive condition. Small ELMs (frequency ~100-200Hz) were obtained in this long pulse H-mode discharge. In the operation, the optimization of X-point, the outer gap and local gas puffing near LHW antenna were investigated to maintain RF power coupling and to avoid formation of hot spot on the 4.6 GHz LHW antenna. Global parameters of toroidal field $B_T$ and line averaged electron density <$n_e$> were optimized for high current drive efficiency of LHW and for on-axis deposition of ECH. The on-axis ECH was applied not only for the core electron heating, but also for the control of high $Z_{eff}$ impurities in the core plasmas.
Meanwhile, a higher $\beta_N$ ~1.8 with a duration of 20s is achieved by using the modulated neutral beam. Several normalized parameters of $\beta_P$ ($\beta_P$ ~2.0), $\beta_N$ ($\beta_N$ ~1.8), $H_{98y2}$ (~1.3), $n_e/n_{GW}$ (~ 0.75) are close or even higher than the phase III 1GW scenario of CFETR steady-state $[2]$. Other features such as metal wall (tungsten divertor), low torque injection ($\Gamma_{inj}$~1.0Nm), electron dominated heating ($T_e$>$T_i$), moderate bootstrap current fraction ($f_{bs}$~50%), broaden current density profile with the central q(0)>1.0 and good energy confinement, have also been demonstrated in this scenario. Note that high-Z impurity accumulation in the plasma core was well controlled in a low level by using the on-axis ECH and reducing the fast ion losses through beam energy optimization.
More recently, EAST has demonstrated a compatible core and edge integration in high $\beta_P$ scenarios: high confinement $H_{98y2}$>1.2 with high $\beta_P$ ~2.5/$\beta_N$~2.0, $f_{bs}$~50% is sustained with reduced heat flux by active divertor heat flux at high density $n_e/n_{GW}$ ~0.7 and moderate $q_{95}$~6.7 (shown in figure 2). The energy confinement quality was almost maintained with $H_{98y2}$>1.2 during the radiation feedback control. By active impurity seeding through radiative divertor feedback control via radiated power, the peak heat flux is reduced by ~30% on the ITER-like tungsten divertor, here a mixture of 50% neon and 50% $D_2$ is applied. Note that EAST has developed a number of heat flux control techniques to reduce heat load in separate experiments.
In summary, recent EAST experiments demonstrated long pulse steady-state high plasma performance scenarios and heat flux feedback control. Detailed physics basis to investigate stability and particle transport will be presented for the understanding of fully integrated core-edge solutions on EAST. As a test bed for ITER and CFETR, the EAST upcoming experiments will further exploit additional heating power and demonstrate the core-edge integration of steady-state long pulse high performance scenarios with full metal walls.
This work was supported in part by National Natural Science Foundation of China under Grant No. 11975274,11975276, US Department of Energy under DE-SC0010685 and DE-FC02-04ER54698.
$[1]$ X. Gong et al 2019 Nucl. Fusion 59 086030
$[2]$ J. Huang et al 2020 Plasma Phys. Control. Fusion 62 014019Speaker: Xianzu Gong (Insititute of Plasma Physics, Chinese Academy Sciences) -
Next step fusion devices such as ITER will need a reliable method for controlling the quasi-periodic expulsion of a large amount of heat and particles onto the plasma-facing components caused by edge-localized modes (ELMs). Several options are currently being considered to achieve the required level of ELM-crash control in ITER; this includes operation in plasma regimes which naturally have no or very small ELMs and suppression of ELM-crashes by active control of edge pedestal with resonant magnetic perturbations (RMPs). Many experiments at different tokamak devices have been dedicated to providing a solution that can be applied to ITER by the application of different approaches. However, due to a lack of understanding of the ELM control mechanism, a reliable ELM control is far from perfect yet.
The characteristics of two different non-ELM-crash plasmas have been studied using 2D fluctuation imaging systems on the KSTAR; (1) RMP-driven ELM crash suppressed H-mode, (2) ELM-free (or ELM-less) H-mode plasmas.
The ELM-crash suppressed H-mode plasma by the RMP is characterized by the coexistence of the filamentary mode and smaller-scale turbulent eddies at the plasma edge $[1]$. We have identified filamentary mode similar to ELM filament that still maintained, even when the ELM crash has been completely suppressed on $H_{\alpha}$ signal by RMP. Correlation analysis among the ECEI channels showed that the RMP would keep enhancing turbulent fluctuations at the plasma edge toward the ELM-crash suppression. A cross-phase analysis showed that such edge turbulence has a rather broad dispersion with a wide range of wavenumber ($k_\mathrm{\theta}<1.1$ cm$^{-1}$) and frequency ($f<100$ kHz). A detailed analysis suggests that energy exchange between filamentary mode and RMP-driven turbulent fluctuations would be responsible for the ELM-crash suppression (Fig. 1). Also, the plasma perpendicular rotation and its local shear estimated by the movement of the turbulent eddies decreases rapidly at the transition of the ELM-crash suppression $[2]$. Such reduction of the perpendicular rotation and its local shear could affect the turbulent fluctuation level by RMP.
In contrast, the ELM-free H-mode plasma was accompanied by benign edge harmonic oscillations (EHOs) near the separatrix. The EHOs had long-wavelength (toroidal mode number, $n\le4$) and remained stable during the ELM-free phase, and the ELM crashes rarely occurred at this stage. The observed EHOs appeared discontinuously and synchronized with quasi-periodic RF ($f\sim500$ MHz) spikes, providing an enhanced particle transport from the core plasma. This could avoid an increase in plasma edge pressure and prevent ELM crashes. The bicoherence analysis revealed that there is a strong nonlinear interaction between EHOs, and the nonlinear interaction of EHOs has a significant effect on the ELM structure and dynamics (Fig. 2). In addition to the EHOs, rotational shear found to play a significant role in the ELM-free phase. If the rotational shear is sufficiently large enough compared to the typical ELMing H-mode plasma, the ELM crashes disappeared, and the fluctuation level from the EHO increased, resulting in an intense transport event.
* This work supported by the Korea Ministry of Science and ICT under NFRI R&D programs (NFRI-EN2001-11) and National Research Foundation of Korea (NRF) (No. 2019R1F1A1057545).References:
$[1]$ J. Lee et al, Phys. Rev. Lett. 117, 075001 (2016).
$[2]$ J. Lee et al, Nucl. Fusion 59, 066033 (2019).Speaker: Dr Jaehyun Lee (National Fusion Research Institute (NFRI)) -
Full suppression of Edge Localized Modes (ELMs) by using n=4 resonant magnetic perturbations (RMPs) has been demonstrated for ITER for the first time (n is the toroidal mode number of the applied RMP). This is achieved in EAST plasmas with low input torque and tungsten divertor, thus also addressing significant scenario issues for ITER. In these conditions energy confinement does not drop significantly (<10%) when ELM suppression is achieved compared to the ELMy H-mode conditions, while core plasma tungsten concentration is clearly reduced. The target plasma for these experiments in EAST is chosen as close as possible to the ITER type-I ELMy H-mode operational scenario with low torque input. In these experiments the NBI torque is around $T_{\mathrm{NBI}}$ ~ 0.3-0.4Nm, which extrapolates to 11 – 16 Nm in ITER (compared to a total torque input of 35 Nm when 33 MW of NBI are used for heating). The observed ELM suppression window is consistent with the peak in the modeled edge stochasticity using the MARS-F code. ELM suppression is maintained up to 60% $N_{\mathrm{GW}}$ (the Greenwald density) by feedforward gas fueling after suppression is achieved. These results expand physical understanding and demonstrate the potential effectiveness of RMP for reliably controlling ELMs in future ITER high Q plasma scenarios.
Plasma energy confinement during ELM suppression in these EAST plasmas decrease very slightly (< 10%) compared to the Type I ELMy H-mode with no RMPs applied, as shown in Fig. 1. This is very different from previous observations using low n (n=1 and 2) RMPs in EAST with high $q_{95}$ (≥4) [Ref.1]. As shown in Fig. 1, ELMs are completely suppressed by using n =4 RMPs with odd parity (opposite phases in the upper and lower rows of coils current) in EAST but not when even parity is applied to a type-I ELMy H-mode plasma with $q_{95}$ ≈3.65 and low input torque (TNBI ~ 0.3-0.4Nm) leading to a plasma beta $\beta_N$ ~ 1.5-1.8, similar to that in the ITER Q =10 scenario. The electron and ion temperature are very similar with $T_{i0}$ ≈ $T_{e0}$ ≈2keV. Significant density pump out (20% reduction) takes place during ELM suppression, while the drop of stored energy is negligible (5%). The tungsten concentration (lower subgraph) is also reduced by a factor of 2 compared to type I ELMy H-mode when ELMN suppression is achieved, despite the lower plasma density. The threshold n=4 RMP current for full ELM suppression is around 2kA (⨉ four turns). Striations of the heat and particle fluxes at the divertor target are observed during ELM suppression and are consistent with the modeled magnetic footprint by TOP2D.
Suppression windows in both $q_{95}$ and plasma density are observed; in addition, lower plasma rotation favours access to ELM suppression. ELM suppression is achieved in a narrow $q_{95}$ window in [3.6, 3.8] with odd n=4 RMP configuration with a continuous q95 ($I_p$ ramp-up). ELM suppression can be maintained up to 60% of Greenwald density by feedforward gas fueling after suppression. It is interesting to note that there is not only an upper density but also a lower density threshold for ELM suppression of 40% $N_{\mathrm{GW}}$. Outside the $q_{95}$ and density window only ELM mitigation, not suppression, is observed.
The modelled magnitude of edge stochasticity taking into account the linear MHD plasma response evaluated with MARS-F [Ref.2] is found to be strongly linked to the observed ELM suppression effect. Although vacuum modeling by the MAPS code shows that the edge resonant harmonics are stronger for the even coil configuration, MARS-F shows that the plasma shielding effect is stronger for the even coil configuration than for the odd one. The edge resonances with plasma response shown in Fig. 2 are stronger for the odd coil configuration; this applies not only the resonant harmonic just near the pedestal top (m = -13, n =4) but to the other edge resonant harmonics as shown in Fig. 2. Therefore, the edge stochasticity, i.e. edge Chirikov parameter ($\sigma$) or stochastic layer width ($\delta_{\mathrm{ergodic}}$) might be a better figure of merit for ELM control optimization (previous studies for this parameter [Ref. 3] did not account for plasma response). Including plasma response leads to edge magnetic field stochasticity for odd parity configuration being higher than that for the even one, which is consistent with the experimental results in Fig. 1.
The observed ELM suppression window in this experiment provides a good opportunity to test the effectiveness of different figure of merits for ELM control optimization. MARS-F modelling shows that both the Chirikov parameter near the pedestal top and the normalized edge stochastic layer width have a peak at $q_{95}$~3.6-3.7 for the odd parity configuration as shown in Fig. 3. The resonant window modeled by $\delta_{\mathrm{ergodic}}$ agrees well with the suppression window observed in this experiment. Comparison between different criteria including pedestal top island width and edge X-point displacement to this experimental observation will also be presented. Further detailed modelling studies and experimental analysis of the 3D physics processes in these experiments will be addressed in this presentation, with the objective to expand physics understanding for ELM suppression and to provide a solid physics basis for its extrapolation to future burning plasma devices such as ITER.
This work is supported by the National Key R&D Program of China under Grant No. 2017YFE0301100, the National Natural Science Foundation of China under Grant No. 11875292 and U.S. DOE under DE-SC0020298.References:
(1) Sun Y et al, Phys. Rev. Lett. 117, 115001 (2016).
(2) Liu Y et al, Phys. Plasmas 17, 122502 (2010).
(3) Evans T et al, Phys. Rev. Lett. 92, 235003 (2004).Speaker: Prof. Youwen Sun (Institute of Plasma Physics, Chinese Academy of Scienses) -
Injection of boron powder in the EAST X-point showed edge-localized mode (ELM) suppression with no confinement degradation over a wide range of operations. The work shows that ELM suppression was achieved by making the pedestal marginally "leaky” via an edge-localized GAM-induced particle transport. This approach potentially opens up novel methods for increasing the ELM suppression toolbox for future devices, such as ITER.
The power load onto plasma-facing components caused by type-I ELMs is a critical issue for the lifetime of in-vessel components in fusion reactors such as ITER. Significant effort has been invested in controlling ELMs, and the mainstream method to suppress ELMs is with the use of edge resonant magnetic perturbations (RMPs). A characteristic of ELM suppression by RMPs is the existence of a narrow q95 window of ELM suppression. Furthermore, such ELM suppression generally occurs with confinement degradation. ELM suppression coupled with optimum confinement over a wide range of operations are highly desirable for operational flexibility in ITER and future reactors. In addition to demonstrating ELM suppression without confinement degradation, this work investigates the physical mechanism leading to ELM suppression using boron powder injection on EAST.
Experimental results - ELM suppression was demonstrated on the EAST tokamak by injecting boron impurity in the upper X-point to investigate the effects of boron powder on tungsten core accumulation. Figure 1 displays the time history of relevant discharge parameters without and with boron injection (20 mg/s - a fraction of which produces localized density perturbation). Here, a proxy of boron injection is shown using the B-V emission (Fig. 1a). The presence of ELMs is indicated using the D$_{\alpha}$ signal (Fig.1b). It is clear from this figure that ELM suppression is associated with a slight increase in the stored energy (Fig. 1c) due to an overall increase of electron temperature for the same input power, as well as a reduction of carbon impurity (Fig. 1d). While in this discharge the core tungsten (W) is maintained albeit at a higher level than without B injection (Fig. 1f), other discharges have shown a reduction of core W accumulation.
ELM suppression is not caused by a cumulative/compound effect of boron on the divertor but induced by the interaction of boron powder with the background plasma. To rule out the cumulative effects of boron, a causality test was performed. Specifically, the injection timing was varied for repeated discharges to show that ELM suppression occurs after the boron emission reaches a threshold on the B-V monitor. In addition, we also showed that when boron injection is halted, ELMs return promptly. The direct effect of boron on ELM suppression provides a clear contrast from lithium induced ELM suppression, which is characterized by cumulative conditioning effects and a small reduction in stored energy [R. Maingi et al. 2018 Nucl. Fusion 58 024003].
We then turn to the interaction between the ablated boron powder and the background plasma by examining the fluctuations. ELM suppression is correlated with the onset of modes with multiple harmonics (see Fig 2). These n=1, 2,3 modes are detected in upper X-point (localization of the boron-induced density perturbation) using XUV (proxy for impurity and electron density fluctuations) detectors as well as on magnetic probes. Using multiple complementary diagnostics (BES, interferometers), we showed that these modes are radially localized in the edge near the separatrix.
A clear onset of the mode is observed when the boron emission reaches a certain threshold. Subsequently, the mode’s amplitude increases until saturation is reached. We observe that the mode onset is associated with a reduction of the core W signal, suggesting that the mode drives particles out. Evidence of the fluctuation at this mode’s frequency has been observed on Langmuir probes and divertor D$_{\alpha}$ signals corroborating the fact that particles are kicked out (though orbit loss) from the confined region to the SOL. In addition, the reduction of W is associated with the increase of mode amplitude, which is reminiscent of the edge-harmonic modes in Q-H mode being responsible for particle transport. A question remains: What is the nature of the mode?
Theory - The injection of boron occurs at the upper X-point and the ablated boron creates a local density perturbation. The driving mechanism of the GAM-like observations is described below.The local perturbation induced by the ablated boron leads to a poloidal asymmetry of charges. Such asymmetry will generate an oscillation akin to a Geodesic Acoustic Mode (GAM). While GAMs do not usually cause a net particle flux on their own, it is conceivable that this might change due to orbit loss for oscillations near the separatrix. Simulations are being performed to assess the existence of GAMs and kinetic calculations are in progress to investigate the GAM effectiveness of driving particles out if localized near the separatrix. The GAM frequency for radial wavenumber k$_r$=0 is calculated by solving a two-fluid eigenmode problem [K. Hallatschek, PPCF 49, B137 (2007), R. Hager et al., PPCF 55, 035009 (2013)] in realistic geometry. Because of the complex coupling between the GAM and the sound wave spectrum close to the separatrix, several GAM candidates exist that are classified according to their ratio of perpendicular to parallel kinetic energy E$_\perp$/E$_{\parallel}$ (Fig. 3). The mode with the largest fraction of perpendicular energy is the GAM with fGAM≈10 kHz at normalized poloidal flux $\psi_n$=0.995. But other (more soundwave-like) modes with non-negligible coupling to the radial electric field (E$_\perp$/E$_{\parallel}$∼10%) exist at lower frequencies between 2.8 and 6.6 kHz at $\psi_n$=0.995. The initial two-fluid calculations are consistent with the modes observed in the experiment. However, calculations of the accurate kinetic GAM frequency with nonlocal effects, in the absence of turbulence, are being performed using XGC to compare with the two-fluid calculations. Results presented could open new research opportunities in controlling ELMs by making the pedestal marginally “leaky” via edge-localized GAM-induced particle transport. Work by US DoE Work DE-AC02-09CH11466.Speaker: Ahmed Diallo (PPPL) -
Recent experiments in the DIII-D tokamak have shown that a broadened fast-ion pressure profile enables better control of Alfvén Eigenmodes (AEs), improves fast-ion confinement, and allows access to new regimes. New discharges reach 15% higher normalized plasma beta ($\beta_N$) than previously achieved in steady-state scenarios with negative central shear and $q_{min}>2$ at high field ($B_T=2.0~T$) and $q_{95}=6.0$. Reverse shear, $q_{min}>2$ scenarios are attractive candidates for fully non-inductive tokamak operation and are among those envisioned for compact fusion power plants, with high normalized beta limits, elevated confinement, and avoidance of low-order tearing modes. One potential drawback of these scenarios, however, is their susceptibility to AE-induced fast-ion transport which has been shown to significantly reduce performance, resulting in measured neutron rates that are typically half of the classically expected values {1}. Understanding regimes in which AE control strategies work and charting a path to improved fast-ion confinement is important for ITER and essential for optimization of Advanced Tokamak relevant scenarios.
Experiments show that Reversed Shear Alfvén Eigenmodes (RSAEs) can be reduced using DIII-D’s recently upgraded off-axis beams, which have the effect of both reducing the fast-ion pressure gradient at $q_{min}$ and altering the q-profile while maintaining $q_{min}>2$. In the current ramp (t<2.5 sec), replacing on-axis beams with new off-axis beams at equivalent total power resulted in fewer AEs, with ~24% higher ratio of measured neutrons to calculated classical neutrons (Fig. 1c, blue), and ~8% higher neutron ratio in the flattop. The neutron fraction in the current ramp was further improved using off-axis beams along with Electron Cyclotron Current Drive (ECCD) aimed on-axis (instead of mid radius), with ~36% higher neutron ratio than the reference shot (Fig. 1c, red).
Analysis suggests RSAEs were reduced in these discharges because ECCD moved the $q_{min}$ location ($\rho_{qmin}$) inward to a location of reduced beam pressure gradient and higher plasma pressure (Fig. 2). In general, AEs can be suppressed by manipulating equilibrium profiles to alter or remove eigenmodes, increase mode damping by changing background plasma parameters and profiles, or decreasing the fast-ion drive by reducing geometric or velocity-space gradients. The reduced beam pressure gradient reduces drive for the modes, and increased thermal pressure/gradient has been associated with increased RSAE continuum damping and even removing the eigenmode altogether {2}.
![Neutron ratio from Fig. [1] shows improvement (green to red) when $\rho_{qmin}$ was located at lower beam pressure gradient (reduces RSAE drive) and increased pressure (increases RSAE damping).](https://fusion.gat.com/conference/event/104/attachments/161/1581/Collins.Cami.IAEA2020.Fig2.png)
The fast ion confinement improved as the maximum in the classical beam pressure gradient was decreased (Fig. 3a). In previous experiments with predominantly on-axis beam power, AE activity was altered by transiently increasing the density, broadening the density profile with injected deuterium pellets, radially scanning electron cyclotron current drive or heating, or reducing neutral beam voltage. Some of these techniques resulted in decreased performance as well as inability to maintain steady-state. Recent experiments repeated these AE control methods but increased the off-axis beam power fraction from 30% to 70% using DIII-D’s upgraded beams, enabling operation at decreased beam pressure gradient and ~25% increase in fast ion confinement while maintaining the high-qmin steady-state scenario.
Record parameters were achieved for this scenario at high-field ($B_T=2.0~T$) and $q_{95}=6.0$ (Fig. 3b). The highest performing discharges were produced at relatively higher flattop density using pre-programmed, early beam and ECCD heating and density feedback control in order to raise the plasma temperature early and slow the inward diffusion of poloidal flux to maintain elevated $q_{min}$. Improvements to fast-ion confinement in the current ramp made beam heating more efficient and allowed access to higher flattop density while still maintaining $q_{min}>2$ and reverse shear.
These experiments mark significant progress in advancing the high-qmin scenario and increasing understanding of AE control. This advances our ability to determine how proper plasma parameters and current profile in alpha-dominated plasmas can avoid AE-induced fast-ion redistribution, loss, reduced heating efficiency, and limits to the achievable $β_N$. This is important to enable ITER to reach its peak performance goals, as well as for optimization of the Advanced Tokamak approach to fusion energy.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698, DE-FG03-94ER54271, and DE-AC02-09CH11466.
{1} C.T. Holcomb, et.al., Physics of Plasmas 23, 062511 (2016)
{2} M.A. Van Zeeland, et al., Nucl. Fusion 56, (2016) 112007.Speaker: Cami Collins (General Atomics) -
ITER high Q operation requires the integration of high performance plasmas with high plasma density, good fast ion confinement with acceptable stationary and transient power fluxes to plasma facing components (1). To control transients associated with high performance plasmas (Edge Localized Modes or ELMs), ITER is equipped with a set of in-vessel coils that modify the edge magnetic field structure from 2-D to 3-D (so called Resonant Magnetic Perturbations or RMPs) thereby achieving the required control. While ELM control using this approach has been demonstrated in many tokamak experiments, there remain open issues regarding its integration with other requirements essential to achieve high Q operation in ITER, such as its effect on the achievable energy, particle and fast ion confinement, compatibility with semi-detached/radiative divertor operation, low input torque operation, pellet fueling, etc.
To address these integration issues a series of experiments has been carried out at the EAST tokamak in ITER-like H-mode plasmas with q_95 ~ 3.7, normalized plasma beta β_N ~ 1.7 and with NBI injected torque T ~ 0.3-0.4 Nm, which is similar (somewhat lower) to the normalized torque input of ITER plasmas with 33 MW of NBI. Control of ELMs has been explored with n = 2 and n = 4 symmetry for the toroidal waveforms of the currents applied to the EAST in-vessel ELM control coil set, which is composed of two rows of 8 coils above and below the midplane at the low field side of the tokamak (2). The magnitude of the applied current and the toroidal phase of the applied perturbation (for n = 2) has been scanned and their effects on energy and particle confinement, NBI fast ion losses and divertor power fluxes has been characterized.
As shown in Fig. 1 ELM suppression can be achieved with both n = 2 and n = 4 with similar levels of current in the RMP coils (12 kAt). However, the decrease of particle and energy confinement associated with the achievement of ELM suppression is much larger for n = 2 (30%) than for n = 4 (< 10%). This supports the use higher n RMP perturbations to optimize ELM suppression with regards to energy and particle confinement as required for ITER, where n = 3 or 4 are considered. It is important to note that ELM suppression with n = 4 RMPs can be achieved in a range of density 〈n_e 〉=0.4-0.6 n_Greenwald thereby having an upper and a lower density level.
Because of the lower effects on energy and particle confinement of n = 4 RMP, as required for ITER, the compatibility of pellet fueling with ELM suppression was investigated in this configuration. As shown in Fig. 2 the injection of pellets leads to sudden increases of the plasma density and to spikes in the D_alpha signal from the divertor. However, unlike ELMs, these are not accompanied by an increase in the divertor heat fluxes, unlike the ELMs that are triggered when no RMPs are applied. This is consistent with non-linear MHD simulations of pellet injection in H-modes that always show triggering of MHD activity when pellets are injected but not growing exponentially to trigger ELMs in all cases (3). When successive pellets are injected (10 Hz was used in experiments) and the plasma density increases beyond 0.65 n_Greenwald, ELM suppression between pellets is lost in a similar fashion as with gas fueling described below.
Studies have also been performed to investigate the integration of ELM suppression with radiative/semi-detached divertor operation for n = 2 and 4 RMPs. It has been found that divertor power flux reduction can be achieved while maintaining ELM suppression either by gas fueling or increasing intrinsic impurity radiation by Ne seeding. The reduction of the power flux is larger for the former approach as shown in Fig. 3, although ELM suppression is lost when the density increases to high values.
Interestingly, the divertor power flux and temperature decrease in both near separatrix and off-separatrix lobes, which is in contrast with previous EAST experiments in which ELM control was achieved by LHCD (4). This is consistent with the field line mapping shown in Fig. 3 (with TOP2D in the vacuum approximation) indicating that the outer lobes field lines in this n = 4 are not connected to plasma regions deep inside the separatrix (the pedestal plasma corresponds to 1-ρ_min ~0.05).
The paper will describe the detailed analysis of the wide range of experimental measurements obtained (including NBI losses) and will discuss implications for ITER.
This work is supported by the National Key R&D Program of China under Grant No. 2017YFE0301100 and the National Natural Science Foundation of China under Grant No. 11875292. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization.
References
(1) A. Loarte et al., Nucl. Fusion 54, 033007.
(2) Y. Sun et al. this conference.
(3) S. Futatani et. al. Nucl. Fusion 54, 073008.
(4) J. Li et. al. Nature Phys 9 (2013) 817.
Speaker: Alberto Loarte (ITER Organization) -
In the study of burning plasmas it is important to understand multi-scale interactions between energetic-particle-driven MHD mode and drift-wave turbulence for establishing good confinement of both energetic particles and bulk plasmas simultaneously. We investigate nonlinear multi-scale interactions between TAE, which is unstable at low $n$, and drift-wave turbulence, which is driven by micro-instabilities at high $n$, by means of a global gyrokinetic simulation code (GKNET [A]). We have revealed that TAE suppresses the most unstable drift-wave mode by violating the ballooning structure of the drift-wave mode, and the TAE transfers the energy from the most unstable drift-wave mode to lower $n$ modes to modulate turbulence, because the TAE has a finite $n$ in contrast to zonal flows ($n=0$). This modulation of drift-wave turbulence by TAE leads to an enhancement of both energy flux of bulk ions and particle flux of energetic ions. Hence, TAE and drift-wave turbulence synergistically enhance the transport of both bulk plasma and energetic particles.
Introduction: In order to realize good confinement of burning plasmas it is necessary to reduce both energetic particle transport and bulk plasma transport simultaneously. In burning plasmas drift-wave turbulence (DWT) and toroidal Alfven eigenmode (TAE) driven by energetic particles coexist, and they interact each other by nonlinear mode coupling, and thus the interaction may result in new transport phenomena, for instance, the transport of energetic particles may be influenced by turbulence and zonal flows which are active even in finite $\beta$ plasmas [B]. Since TAE is an MHD mode and drift-wave turbulence is electromagnetic at finite $\beta$, magnetic perturbations play an important role in the interaction and can increase turbulent transport by intensifying electrostatic potential perturbations [C]. In addition, interactions between TAE and drift-wave turbulence may excite stable low $n$ modes and increase turbulent transport as shown by the study of multi-scale interactions between magnetic islands and drift-wave turbulence [D].
Simulation model and linear stability: We investigate nonlinear interactions between TAE and drift-wave turbulence by means of a global $\delta$f gyrokinetic simulation code (GKNET) [A]. We consider a normal magnetic shear tokamak plasma which has energetic particle pressure gradient and bulk plasma pressure gradient with $\beta=1.28\%$, $T_f/T_i=25$, $m_i/m_e=100$, and $\rho_*=1/100$. Figure 1 (a) shows that this plasma is unstable against a TAE at low toroidal mode number $n=2$, which has real frequency in the gap of Alfven continuum indicated by yellow color. On the other hand, a drift-wave instability (kinetic ballooning mode: KBM) is unstable at high toroidal mode number $n \geq 6$. The TAE has global structure (Fig. 1 (b)), while the KBM has ballooning structure characterized by micro-scale (Fig. 1 (c)).
Turbulence modulated by TAE: We have performed a nonlinear simulation of the plasma which is unstable against both the TAE and drift-wave instability (KBM) (referred as "TAE+DWT"). In addition, we carried out nonlinear simulations of a plasma with flat energetic particle pressure profile to obtain drift-wave turbulence ("only DWT") and a plasma with flat bulk pressure profile to obtain TAE ("only TAE"), and then we compare them with "TAE+DWT" to understand the influence of interactions between TAE and drift-wave turbulence.
In "TAE+DWT", drift-wave turbulence (DWT) with zonal flows is established at first ($t=12$ and 16 in Fig. 2) because the growth rate of the drift-wave mode (KBM) is much higher than the TAE as shown in Fig. 1 (a). Then, in this turbulent state, the TAE ($n=2$) grows slowly and violates the ballooning structure of the turbulence to reach a quasi-steady turbulent state ($t=22$ and 36 in Fig. 2). We here compare time evolution of some main toroidal modes of perturbations in "TAE+DWT" and "only DWT" in Fig. 3 (a). The most unstable drift-wave mode ($n=12$) gets saturated by producing zonal flows ($n=0$) at $t=13$ for both "TAE+DWT" and "only DWT". Then, at $t=20$, the TAE ($n=2$) grows in "TAE+DWT", while $n=2$ mode decreases in "only DWT", resulting in much higher amplitude of $n=2$ mode in "TAE+DWT" as indicated by the red arrow. Following the growth of TAE ($n=2$) the most unstable drift-wave mode ($n=12$) further decreases in "TAE+DWT" compared to "only DWT" after $t=20$ as indicated by the blue arrow. Since the TAE has finite toroidal wavenumber in contrast to zonal flows ($n=0$), this nonlinear mode coupling between the TAE ($n=2$) and the drift-wave mode ($n=12$) enhances another lower toroidal wavenumber mode ($n=12-2=10$) as indicated by the green arrow. Hence, the TAE suppresses the most unstable drift-wave mode but enhances a lower toroidal wavenumber mode to modulate the drift-wave turbulence. Due to this modulation of turbulence by TAE, the energy flux of bulk ions $Q_i$ in "TAE+DWT" is enhanced at middle wavenumbers ($4 \leq n \leq 10$), and the peak of $Q_i$ in "TAE+DWT" is shifted from $n=12$ to $n=10$ compared to "only DWT" (Fig. 3. (b)). In addition, the particle flux of energetic ions $\Gamma_f$ is enhanced in "TAE+DWT" compared to "only TAE" (Fig. 3. (c)). Thus, the interaction between TAE and drift-wave turbulence enhances the transport of both bulk plasma and energetic particles.
[A] K. Imadera, Y. Kishimoto, K. Obrejan, T. Kobiki and J. Q. Li, IAEA-FEC, TH/P5-8 (2014).
[B] A. Ishizawa, K. Imadera, Y. Nakamura, and Y. Kishimoto, Phys. Plasmas, 082301 (2019).
[C] A. Ishizawa, D. Urano, Y. Nakamura, S. Maeyama, and T.-H. Watanabe, Phys. Rev. Lett., 025003 (2019).
[D] A. Ishizawa, Y. Kishimoto, and Y. Nakamura, Plasma Phys. Control. Fusion, 054006 (2019).Speaker: Akihiro Ishizawa (Kyoto University) -
Detailed new H-mode pedestal measurements in the inter-ELM periods of DIII-D discharges find that ion temperature gradient (ITG)-scale density fluctuations (ñ) can explain anomalous ion heat flux (Q
$_{i}$) during the ELM (Fig. 1), followed by Q$_{i}$becoming neo-classical$^{1}$until the next ELM, and with trapped electron mode (TEM) and microtearing mode (MTM) like ñ and magnetic fluctuations ($\tilde{B}$) able to explain anomalous electron heat flux (Q$_{e}$) between ELMs (Fig. 1). The observed ITG-scale, TEM-scale, and MTM-like turbulence (ñ,$\tilde{B}$) behaviors are in clear contrast to the often assumed suppression or near suppression of these modes in the pedestal during inter-ELM periods$^{2}$. This new information impacts our fusion power goals as it will lead to improved predictive capability when existing (and new) models take into account the transport caused by these modes and their various temporal phases during the inter-ELM period. The goal of predicting and controlling pedestal pressure evolution and improving ELM control (or elimination) is believed necessary to optimize fusion performance and to protect plasma facing components of ITER and future burning plasmas.
Inter-ELM behavior of electron pedestal density, temperature, and pressure gradients ( ∇n$_{e,ped}$, ∇T$_{e,ped}$,and ∇P$_{e,ped}$) estimated from Thomson scattering measurements and pedestal localized turbulence were studied in ITER similar shape (ISS) H-mode plasmas. The gradients of electron temperature, density, and pressure remain approximately saturated for more than 75% of the inter-ELM period (Fig 2). Immediately after an ELM, evolution of ITG-scale low-k (k$_{\theta}\rho_{s}$~0.2, k$_{\theta}$is the fluctuation poloidal wavenumber and$\rho_{s}$is the ion Larmor radius) ñ (Fig 2b), measured by Doppler backscattering at the foot of the pedestal, correlates strongly with a rapid increase and subsequent saturation of ∇n$_{e,ped}$(Fig 2d). This low-k ñ increases strongly when the pedestal profiles relax after an ELM crash, is correlated with a decrease in local E$\times$B shear and later is suppressed by increase in local E$\times$B shear (Fig 2c). The evolution of this ITG-scale ñ amplitude correlates well with the divertor D$_{\alpha}$light intensity (Fig 2a) and ∇T$_{e,ped}$(not shown) starts to increase after suppression of the ITG-scale ñ. These observations are consistent with ITG-scale ñ driving electron and ion heat flux as well as particle flux just after the ELM crash. Experimental results from ASDEX-U indicating the anomalous nature of both ion and electron heat fluxes right after an ELM crash, are consistent with our observations$^{1}$.
TEM-scale intermediate-k (k$_{\theta}\rho_{s}$~1) ñ measured by Doppler backscattering in the steep gradient region of the pedestal shows a critical ∇P$_{e,ped}$behavior with ñ increasing rapidly when a critical ∇P$_{e,ped}$is reached (Fig. 3). As the pressure gradient increases, increases in E$\times$B velocity shear (not shown here) leads to saturation of TEM-scale ñ (at ~120 kPa/m, Fig 3) indicating a balance between saturated values of gradient drive, shear suppression, and turbulence level. The approximate saturation of ∇P$_{e,ped}$that follows the large increase of TEM-scale ñ at critical ∇P$_{e,ped}$also suggests the related turbulence transport has a role in clamping ∇P$_{e,ped}$. Note however, although ∇P$_{e,ped}$becomes approximately constant within 25% of the ELM cycle there remains a slow increase in ∇P$_{e,ped}$as shown in Fig 2g and Fig 3. Magnetic turbulence$\tilde{B}$, measured by polarimetry and fast magnetic probes (Fig 2h), has been identified as MTM turbulence$^{3}$and increases with ∇T$_{e,ped}$(∇T$_{e,ped}$not shown but is very similar to the ∇P$_{e,ped}$shown in Fig. 2g ) in the inter-ELM period. Both TEM-scale ñ and MTM-like$\tilde{B}$can cause significant electron thermal transport and may explain the anomalous nature of Q$_{e}$in the pedestal as observed by transport calculations using experimental profiles (in ASDEX-U$^{1}$and DIII-D$^{4}$tokamaks).
Initial linear trapped-gyro-Landau-fluid (TGLF)$^{5}$calculations in the nearly saturated phase of the ELM cycle (60-99% of the ELM cycle, refer Fig 2) show that intermediate-k (k$_{\theta}\rho_{s}$~1) electrostatic fluctuations are the most unstable modes in the steep gradient region of the pedestal and propagate in the electron diamagnetic direction in the plasma frame. The growth rates are approximately 1.5-2 times more than the local E$\times$B shearing rates and so are relatively unaffected by the shearing. This is qualitatively consistent with the observation of these modes by the DBS system. These modes are still the most unstable modes when electromagnetic effects are included in TGLF calculations. Interestingly, scans of electron temperature and density scale lengths indicate that these modes are driven by both ∇T$_{e}$and ∇n$_{e}$in qualitative agreement with the observed ñ vs ∇P$_{e}$behavior (Fig 2 and 3). Further, the TGLF simulations indicate a critical gradient type response similar to experiment (Fig 3). Based on these above TGLF results we initially identify these modes as TEM type instabilities. The above TGLF results are consistent with TEM type instabilities leading us to initially identify the experimentally observed density fluctuations to be trapped electron modes.
The new scientific understanding reported here on the ELM and inter-ELM turbulence, thermal fluxes, and temporal behavior are critical to testing, validating, and improving current models and simulations. These improved models and simulations are in turn fundamental to predicting and controlling pedestal pressure and ELMs (or even ELM elimination) necessary for optimizing and protection of future fusion devices.This work was supported by the US Department of Energy under grants DE-FG02-08ER54984, DE-AC02-09CH11466, DE-FG02-08ER54999, DE-SC0019302 and DE-FC02-04ER54698.
$^{1}$E. Viezzer et al., Nucl. Fusion 57, 022020 (2017).
$^{2}$P.B. Snyder et al., Nuclear Fusion 51, 103016 (2011).
$^{3}$J. Chen et al., Submitted to IAEA FEC 2020.
$^{4}$S. Haskey et al., Submitted to IAEA FEC 2020.
$^{5}$G. Staebler et al., Phys. Plasmas 14, 055909 (2007).Speaker: Kshitish Kumar Barada (University of California-Los Angeles) -
An accurate and predictive model for turbulent transport fluxes driven by microinstabilities is a vital component of first-principle-based tokamak plasma simulation. However, tokamak scenario prediction over energy confinement timescales is not routinely feasible by direct numerical simulation with nonlinear gyrokinetic codes. Reduced order modelling with quasilinear turbulent transport models provides significant computational speedup, and is justified in many regimes. The justification of the quasilinear approximation for transport driving spatial scales is a consequence of the underlying structure of tokamak microturbulence, and is validated by comparison to nonlinear simulations. This approach has emerged as a successful tool for prediction of core tokamak plasma profiles. We focus on significant progress in the quasilinear gyrokinetic transport model QuaLiKiz [1,2], and its application within flux driven integrated tokamak simulation suites.
To model 1s of JET plasma on order of 24 hours with 10 CPUs, QuaLiKiz employs an approximated solution of the mode structures to significantly speed up the computation time compared to full linear gyrokinetic solvers. Additional approximations include maintaining shifted-circle $(\hat{s}-\alpha)$ geometry, and the electrostatic limit. These approximations, together with optimisation of the dispersion relation solution algorithm within integrated modelling applications, leads to flux calculations $10^{6-7}$ faster than local nonlinear gyrokinetic simulations. This allows tractable simulation of flux-driven dynamic profile evolution over multiple confinement times including all transport channels: ion and electron heat, main particles, impurities, and momentum. QuaLiKiz is open source and available at www.qualikiz.com.
In this contribution, we will summarize the justification of the quasilinear approximation [3,4], sketch the basis of the QuaLiKiz transport model and its validity in comparison to nonlinear simulations, and illustrate validation of the model against experimental measurements at JET through flux-driven simulations within the JINTRAC integrated modelling suite [5,6], see figure 1 for an example. This capability 1) enhances the interpretation of present-day experiments, 2) enables “Predict First” simulations to aid with experimental optimization, and 3) allows theory-based extrapolation to future machine performance, at least with respect to core turbulence physics. While we focus here on JINTRAC simulations, QuaLiKiz is also coupled to the ASTRA [7,8], CRONOS [9] and ETS [10] integrated modelling codes.
Recent QuaLiKiz applications within integrated modelling include: W-accumulation interpretation and optimization, where the QuaLiKiz prediction of background kinetic profiles is critical for setting the neoclassical heavy impurity transport level [11-13]; modelling of multiple-isotope experiments at JET, where fast isotope mixing in the Ion Temperature Gradient (ITG) regime is crucial for experimental interpretation and has important implications for potential scenarios in JET DT, as well as for reactor burn control [14]; development of Uncertainty Quantification methods using Gaussian Process Regression to enhance statistical rigour in model validation, providing avenues for error propagation within QuaLiKiz simulations in integrated modelling [15]; predictive modelling for ITER scenarios, which predict the target Q∼10 when using a theory-based pedestal boundary condition [16]; and predictive modelling for DTT scenarios [17].
Beyond standard application within integrated modelling, QuaLiKiz has been leveraged for the development of realtime calculation capability for scenario optimization and realtime-oriented applications. This is based on machine learning methods, where a large database of pre-calculated QuaLiKiz runs is used to train feedforward neural networks to accurately reproduce model predictions. The neural network transport model provides a further 6 orders of magnitude speedup, 1 trillion times faster than the anchoring nonlinear simulations [18]. By coupling to the RAPTOR [19] control-oriented fast tokamak simulator, realtime-capable transport predictions are possible. This opens up a plethora of possibilities and innovation in realtime controller design and validation, scenario preparation, and discharge optimization.
While QuaLiKiz has had significant predictive success, continuously challenging and improving the model is a crucial component for instilling validity in wide parameter space. Beyond its role in experimental interpretation and prediction, reduced models such as QuaLiKiz are a key player in the multi-fidelity model hierarchy due to its feasibility for systematic comparison with experiments and identifying trends in model validation. This spurs further research, also incorporating higher fidelity linear and nonlinear models, ultimately improving our understanding of core tokamak turbulence physics.
We thus conclude with an overview of recent work dedicated to testing and improving the underlying QuaLiKiz assumptions. This includes: modification of the collisionality model, critical for obtaining the correct parameter dependencies of Trapped Electron Modes (TEM); validating the QuaLiKiz Electron Temperature Gradient (ETG) model versus multi-scale nonlinear GENE simulations; testing validity of QuaLiKiz towards the L-mode edge, where the standard ITG/TEM/ETG paradigm breaks down at high collisionality, due to the onset of modes with a drift-resistive nature, currently out of QuaLiKiz scope; testing the impact of s-α geometry on the turbulence regime, compared to full geometry, particularly at more outer radii where shaping effects are more prominent. Future work will extend QuaLiKiz to electromagnetic regimes.
![Multi-channel predictive modelling with JINTRAC-QuaLiKiz of JET discharge #91227, corresponding to stationary state following simulation over 10 energy confinement times. Experimental profiles were fitted with Gaussian Process Regression corresponding to data averaging over the time window t=8.2-8.5s. Core boundary condition was set at $\rho=0.8$. From Ref [14].](https://gitlab.com/qualikiz-group/QuaLiKiz-documents/-/raw/master/conferences/conference_images/citrin_IAEA_2020.jpg)
References
[ 1] J. Citrin et al., Plasma Phys. Control. Fusion 59 124005 (2017), and http://qualikiz.com
[2] C. Bourdelle et al., Plasma Phys. Control. Fusion 58 014036 (2016)
[3] A Casati et al,. Nucl. Fusion 49 085012 (2009)
[4] J Citrin et al., Phys. Plasmas 19 062305 (2012)
[5] G. Cenacchi and A. Taroni, JET-IR , 84 (1988), eNEA-RT-TIB–88-5
[6] M. Romanelli et al., Plasma and Fusion Research 9 3403023 (2014)
[7] G. V. Pereverzev et al., IPP Report 5/42 (August 1991)
[8] E. Fable et al., Plasma Phys. Control. Fusion 55 124028 (2013)
[9] J.F. Artaud et al., Nucl. Fusion 50 043001 (2010)
[10] D. Kalupin et al., Nucl. Fusion 53 123007 (2013)
[11] S Breton et al., Nucl. Fusion 58 96003 (2018)
[12] F. Casson et al., submitted to Nucl. Fusion
[13] O Linder et al., Nucl. Fusion 59 016003 (2019)
[14] M. Marin et al., Nucl. Fusion 60 046007 (2020); and this conference
[15] A. Ho et al., Nucl. Fusion 59 056007 (2019)
[16] P. Mantica et al., Plasma Phys. Control. Fusion 62 014021 (2020)
[17] I. Casiraghi. P. Mantica et al., this conference
[18] K.L. van de Plassche et al., Physics of Plasmas 27, 022310 (2020) ; and this conference
[19] F. Felici et al., Plasma Phys. Control. Fusion 54 025002 (2012)Speaker: Jonathan Citrin (FOM DIFFER - Dutch Institute for Fundamental Energy Research) -
A predictive 3D optimizing scheme in tokamaks is revealing a robust path of error field correction (EFC) across both resonant and non-resonant field spectrum. The new scheme essentially finds a way to deform tokamak plasmas in the presence of non-axisymmetric error fields while restoring a quasi-symmetry in particle orbits as much as possible. Such a “quasi-symmetric magnetic perturbation” (QSMP) has been predictively optimized by general perturbed equilibrium code (GPEC) {1} and successfully tested in DIII-D and KSTAR tokamak plasmas (Fig. 1). The QSMPs in experiments demonstrated no performance degradation despite the large overall amplitudes of 3D fields, as clearly compared with a resonant magnetic perturbation (RMP) or a typical non-resonant magnetic perturbation (NRMP) (Fig. 2). The results indicate that the tokamak EFC can be improved beyond the present resonant-overlap EFC approach alone, if the residual non-axisymmetry can be further compensated to a level of QSMP. The studies also validate that the torque response matrix, a unique product of the GPEC formulation, can be used to assess the degree of not only resonant but also non-resonant EF correction – which has been desired by ITER for a long time.
Successful EFC is critical in tokamaks for preventing disruptions, especially in next-step devices like ITER due to unfavorable scaling with high $B_T$ and $\beta_N$. Improved understanding of plasma response in the last decade allowed the development of a more reliable approach than earlier ones, using the resonant-overlap field {2}, i.e. error field component that triggers the dominant RMP response. This led to the successful multi-machine scaling of resonant EF thresholds against disruptive locked modes across wide operational regimes and 3D field spectra, including the n=1 and n=2 toroidal mode numbers {3}. Recent highlights include the successful EFC against 2/1 locked modes due to the EF by the high-field-side (HFS) using the low-field-side (LFS) coils in COMPASS and NSTX-U, despite an overall increase of non-axisymmetry in both cases. These experiments, however, also posed an important question that must be addressed; how to quantify and compensate the next key mode, or simply NRMPs. The residual EFs after the dominant RMP correction remained disruptive during L-H transitions in COMPASS {4}, and significantly degraded performance in evolving discharges in NSTX-U {5}. Such residual EFs also generally drive rotational damping through uncorrected NRMPs as shown in DIII-D {6}.
It turns out that the residual EF effects can be almost entirely suppressed if the additional coils are available to minimize the neoclassical toroidal viscosity (NTV) simultaneously with the dominant RMP correction. To experimentally demonstrate this, it is necessary to have 3 rows of coils – one coil to generate a strong proxy EF and RMP response, another coil to compensate the dominant resonant response while leaving a NRMP response, and yet another to minimize the remaining NRMP-driven NTV and leave only QSMPs. Figure 1 shows the n=1 coil configurations designed to test QSMPs in DIII-D and KSTAR, with the predicted torque profiles compared to RMPs and NRMPs. One can see that the local torque near resonant layers in RMP is significantly reduced in NRMP, and the global non-resonant torque in NRMP is minimized in QSMP. These fields were applied with maximum coil currents to high performance ($\beta_N\sim3$) DIII-D plasmas with ~9MW NBI power. As shown in Fig. 2, there was no change observed in performance and confinement during the QSMP, compared to the clear rotation braking with the NRMP or the strong density pumping and rotation braking with the RMP which eventually led to a locked mode. Surprisingly, the NRMP remained disruptive during L-H transition when tested with marginal DIII-D H-modes, but the effect could be eliminated in QSMP, suggesting a potential resolution to the aforementioned issue in COMPASS with the dominant RMP correction alone. The QSMP optimization was also performed in KSTAR discharges ($\beta_N\sim2$), again showing no performance changes in contrast to the RMP and NRMP (Fig. 3). Note that KSTAR can make a pure NRMP with strong NTV {7} as can be seen by complete elimination of local resonant torque in Fig. 1, by taking advantage of it’s 3 rows of in-vessel coils and also low intrinsic EF.
EFC will never eliminate all small 3D fields in a tokamak and it, in fact, commonly increases them when the correction coils are shaped very differently from the intrinsic error field sources. Instead, one can identify a safe and robust 3D state such as a QSMP and see if it is accessible in the course of EFC. The self-consistent perturbed equilibrium calculations with neoclassical transport in GPEC offer a torque response matrix $T(\psi)$, from which one can immediately predict the torque profile by quadratic operation $\Phi^\dagger\cdot T\cdot\Phi$, where $\Phi$ is a field spectrum vector on a toroidal surface or coil vector representing its amplitude and phase by complex numbers {1}. The minimum eigenstate of $T(\psi)$ then represents the best possible way to deform the plasma while sustaining the minimum variation in the field strength or action variation ($\delta B_L \approx 0$, $\delta J \approx 0$) as well as minimum resonant parallel current and corresponding resonant torque, i.e. achieving quasi-symmetry.In summary, new experiments demonstrated that a QSMP could be an ideal EFC state without any performance degradation, offering a new and complementary EFC approach in addition to the present resonant-overlap method and also a possible resolution on non-resonant error field correction problem. The torque response matrix in GPEC will enable the QSMP optimization in more complicated 3D tokamak environments such as ITER with many more 3D coils and potential EF sources. This QSMP is also an interesting concept in and of itself, as it holds a sizable local perturbation at least near the divertor and its possible utility will be further discussed. *This research was supported by U.S. DOE contracts #DE-AC02-09CH11466 (PPPL), #DE-FC02-04ER54698 (DIII-D), and also by the Korean Ministry of Science, ICT and Future Planning (KSTAR).
{1} J.-K. Park and N. C. Logan, Phys. Plasmas 24, 032505 (2017)
{2} J.-K. Park, N. C. Logan et al., “Assessment of EFC criteria for ITER”, ITPA MDC-19 Report (2017)
{3} N. C. Logan, J.-K. Park et al., “Scaling of the n=2 error field threshold in tokamaks”, submitted to Nucl. Fusion (2020)
{4} T. Markovic et al., the 45th EPS DPP in Prague, Czech Rep. (2018)
{5} N. Ferraro, J.-K. Park et al., Nucl. Fusion 59, 086021 (2019)
{6} C. Paz-Soldan, N. C. Logan et al., Nucl. Fusion 55, 083012 (2015)
{7} Y. In, Y. M. Jeon et al, Nucl. Fusion 59, 056009 (2019)Speaker: Jong-Kyu Park (Princeton Plasma Physics Laboratory) -
A small angle slot (SAS) divertor concept 1 with a closed slot structure and appropriate target shaping in the near SOL has been developed in order to explore a potentially robust boundary solution with acceptable plasma surface interaction which is essential for fusion reactor plasma conditions, in particular for high-power steady-state operation. Recent experimental tests in DIII-D have demonstrated the advantages of the SAS divertor over the other divertor geometries on the power dissipation. Compared to either a vertically slanted divertor or a horizontal open divertor, the SAS divertor can strongly cool the plasma at the divertor target over a wider range of H-mode plasma conditions, while the performance of the core plasma is either preserved or the usual deterioration with increasing density is reduced. These benefits are realized by leveraging a strong synergy between the SAS geometry and drifts for ion grad-B drift away from the active divertor, while the effects are much smaller for the opposite $B_T$ direction. This is qualitatively reproduced by SOLPS-ITER simulations which include drifts indicating that, at least for DIII-D scale devices, divertor geometry and drifts have comparable effects on the divertor plasma. This finding is an important step in the understanding of the behavior of advanced divertors for resolution of power exhaust issues.
As can be seen in Fig. 1, experimental results show that for the ion $B \times \nabla B$ drift away from the X-point, a highly dissipative divertor plasma with $Te \leq 10$ eV across the entire SAS divertor target can be achieved at very low main plasma densities. In contrast, for the ion $B \times \nabla B$ drift toward the X-point, the SAS divertor plasma remains hot and attached across the entire divertor target plate until the eventual onset of detachment at significantly higher densities.
Initial SOLPS-ITER simulations including full $E \times B$ drifts produce similar trends and similar profiles to those observed in experiments. The analysis shows that for the ion $B \times \nabla B$ drift away from the X-point, a strong poloidal $E \times B$ drift moves particles from the inner divertor via the private flux region toward the outer divertor. The drift flow couples synergistically with the geometric effects due to the target shaping predicted in the original SOLPS design modeling without drifts, which strongly enhances neutral momentum and energy dissipation across the divertor SOL. Thus, the SAS divertor can achieve dissipative divertor regime at relatively low densities. However, for the opposite $B_T$ direction, a strong radial $E \times B$ drift moves particle from the outer SOL towards the private flux region and then a poloidal $E \times B$ drift moves particles toward the inner divertor 2. These flows offset the geometric effects of SAS, resulting in a much higher density to achieve detachment.
Experimental data also exhibit the effective cooling of the SAS divertor, compared with other target shaping. Both Langmuir probe measurement and newly installed SETC (surface eroding thermocouple) measurements show that as the strike point is moved across the slot, the peak heat flux drops when the strike point is located near the outer corner (as per the original SAS concept), compared to other strike point locations, i.e., near the inner corner, or on the outboard target in the slot, see Fig. 2. Langmuir probes also show that when the strike point is near the outer corner, the plasma temperature across the measured profile is lower than for other locations. For the ion $B \times \nabla B$ drift toward the X-point, the SETC measurements also show a similar trend that the heat flux is low when the strike point is near the outer corner. This is mainly due to the low particle flux (measured by the divertor Langmuir probes) and is consistent with the simulation, while the plasma temperature remains high until the onset of detachment.
It is notable that a ‘Te cliff’, i.e. Te suddenly drops to below 10 eV for increasing plasma density, is observed for both $B_T$ directions in SAS with the strike point either near the outer corner (Fig. 1) or at the inner slant surface. This is different from the situation for the open divertor where a Te cliff is only observed for the ion $B \times \nabla B$ drift toward the X-point 2. These results also demonstrate a stronger impact of drifts on the dynamics of divertor detachment in a closed slot divertor than in an open divertor.
Furthermore, the SAS divertor with the ion $B \times \nabla B$ drift away from the X-point also improves pedestal performance and core confinement. The confinement collapse associated with the onset of an X-point MARFE occurs at significantly higher pedestal densities for the SAS divertor than for the open divertor in DIII-D, thus widening the window of H-mode operation compatible with a dissipative divertor operation.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698 and DE-SC0010685.
1 H.Y. Guo et al 2019 Nucl. Fusion 59 086054; 2 A. E. Jaervinen et al 2018 Phys. Rev. Lett. 121, 075001Speaker: Dr Huiqian Wang (General Atomics) -
KSTAR has clarified a set of unresolved 3-D physics issues that could be addressed in the ITER-like in-vessel 3-row, resonant magnetic perturbation (RMP) configurations. In particular, considering that one of the most critical metrics of RMP ELM-crash control would require the compatibility with the divertor heat fluxes under the given material constraints, a series of intentionally misaligned RMP configurations (IMC)$^{1,2}$ have been explored to reveal the relationship between RMP ELM control and divertor heat fluxes. Specifically, taking advantage of the time-resolved IR camera, each rotating IMC in either 3-row or a combination of 2-row IMCs helped us diagnose the ‘wet’ area of divertor in the vicinity of ELM-crash-suppression; ELM-crash-mitigation, ELM-crash-suppression, and mode-locking.
First of all, we have articulated the contrasting effect of kink (i.e. “away” phasing) vs anti-kink (i.e. “toward” phasing) responses on the ELM-crash suppression, as shown in Figure 1.
Starting from a sub-marginal level of RMP current in a typical n=1, 90 deg phasing, the 3-row IMC in kink phasing (in red) becomes more kink-influenced, demonstrating the synergistic benefit of ‘kink’ phasing in ELM-crash-suppression. In contrast, the 3-row IMC in the anti-kink phasing (in green) becomes more insensitive to ELM-crashes at the sub-marginal level of RMP. In a way, this helps us recast the “away” and “toward” phasing as kink and anti-kink phasing respectively, as schematically shown in the lower left inset of Figure 1. Such experimental observation is in excellent agreement with what ideal MHD theory predicts $^{3}$. Previously, we had shown the divertor heat flux broadening with 3-row IMC-driven ELM-crash-suppression in both “away(kink)” and “toward (anti-kink)” phasings, while no such broadening was observed in the 2-row IMCs with top/bottom coils $^{1}$. Now, we have newly observed that the ‘wet’ area of ELM-crash-mitigation got more broadened than that of ELM-crash-suppression (not shown here), based on these 3-row IMC discharges.
Also, we have further investigated whether or not 2-row IMCs would be fundamentally deficient in divertor heat flux broadening during ELM-crash-suppression. In the earlier study, no mid-row was involved in the 2-row IMCs, although the mid-row is much more influential than the other off-mid rows. To clarify this issue, a set of 2-row IMCs, including mid-row, has been explored. Figure 2 shows the time evolutions of various plasma parameters and ‘wet’ area, where each phasing of 2-row IMC varies by the denoted angle in shades incrementally from a typical n=1, 90 degree phasing angle in the anti-kink direction. Throughout the whole IMC application period, the 2-row IMC-driven, ELM-crash-suppression has been accomplished, as shown at the bottom of Figure 2 (a). At the same time, no evidence of the divertor heat flux broadening can be found on the ‘wet’ area in this combination of middle/bottom row IMCs, as shown in Figure 2(b).
Even for kink phasing in the 2-row IMCs with middle/bottom coils (as shown in Figure 3 (b)), a similar outcome has been obtained. Thus, it is a fair conclusion that the divertor heat flux broadening would require a third row, suggesting that the dispersal of the divertor heat flux in 3-row IMCs cannot be driven by helically structured 2-row IMCs alone. Nonetheless, no physics mechanism of the 3-row IMC-driven, divertor heat flux broadening during ELM-crash-suppression has been understood yet, while several hypotheses are being assessed$^{1}$. Interestingly, we have found that middle/bottom rows are much more effective in suppressing the ELM-crashes than top/mid rows, revealing strong up/down asymmetry in lower-single-null (LSN) plasmas, as shown in Figure 3.
Considering that the 3-row IMC-driven, ELM-crash-suppression in kink phasing have been securely obtained at 2.3 kA, a set of 2-row IMCs have been designed to compensate a missing off-mid row current (mid: 2.3 kA, off-mid: 4.6 kA). Surprisingly, such conditions led to a vastly contrasting outcome, proving a much more effective coupling of middle/bottom rows in ELM-crash-suppression than that of top/middle rows. In fact, there was a much lower threshold of ELM-crash-suppression in the combination of middle/bottom rows, even suggesting no need of top row (not shown here). This is reminiscent of the critical influence of X-point on RMP ELM control studied in MAST$^{4}$, though it was related to ELM-crash-mitigation, rather than ELM-crash-suppression.
Overall, the KSTAR has established a new holistic understanding of ITER-like RMP ELM control, elaborating various subtle points in the vicinity of ELM-crash-suppression and ‘wet’ area on divertor. These new findings in 3-D physics is expected to help us further reduce the uncertainty associated with 3-row ITER RMP.Acknowledgements
This work was supported by the Korean Ministry of Science and ICT for National Research Fund (NRF-2020M1A7A1A03007919), the UNIST research fund (1.180056.01), and the KSTAR project (NFRI-EN1901-11).References
1. Y. In et al., Nucl. Fusion 59 (2019) 126045
2. Y. In et al., Nucl. Fusion 59 (2019) 056009
3. J.K. Park et al, Nature Physics 14 (2018) 1223
4. A. Kirk et al, Nucl. Fusion 55 (2015) 043011Speaker: Yongkyoon In (Ulsan National Institute of Science and Technology) -
For the first time, the progress in RF full wave modeling allows for simulating the wave field in arbitrary 3D antenna/first wall geometry together with the scrape-of-layer (SOL) and the entire tokamak/stellarator core plasmas in an integrated manner. Universal observation among many RF heating and current drive (H/CD) experiments in the ion cyclotron (IC), high-harmonic fast waves (HHFW), lower hybrid (LH), and electron cyclotron (EC) frequency ranges is that the wave interaction in SOL plasmas has a significant impact on the RF actuator performance. The simulation of the wave propagation in SOL, however, is challenging due to the fact that it is intrinsically a 3D problem. The computation domain is surrounded by geometrically complicated antenna structure and the background plasma lacks a symmetry due to the private flux regions and fluctuations. Indeed, state-of-art RF SOL/antenna simulation is yet limited to a relatively small volume in front of the antenna, involving significant simplification in the geometrical representation of antenna and/or wave physics. Lifting this limitation is critical to evaluate the RF waves field accurately, especially as the focus of RF experiments orient towards the performance optimization with many RF actuator components being operated simultaneously. Such a device includes, but not limited to, WEST, NSTX-U, SPARC, and ITER. Moreover, the wave propagation in SOL regions and the hot core region, where the wave is spatially dispersive, are not isolated each other. It is important to solve the Maxwell problem in two regions self-consistently. This self-consistency is crucial both for improving the predictable capability of H/CD performance and for better understanding the RF induced plasma-material interaction (PMI).
Our approach to above issues are in twofold: first, we overcome the limitation of present-day antenna/SOL RF simulation tools with a newly developed 3D RF fullwave simulation, Petra-M $[1]$, based on the scalable MFEM library developed by LLNL (http://mfem.org), and second, we combine the Petra-M simulation with our start-of-art core RF solvers in order to incorporate the finite temperature effects of RF waves. On the Petra-M platform (Figure 1), a user can perform the FEM analysis seamlessly from parametrized geometry generation to FEM linear solve on a computing cluster. This paper reports the analysis of RF wave including the 3D SOL plasmas and antenna structure in IC, HHFW, and LH frequency ranges.
Our advanced RF wave solver allows for solving the HHFW propagation including the entire NSTX-U plasma together with 3D RF antenna $[2]$. In Figure 2 and 3, we analyzed the RF electric field induced on the NSTX-U vacuum vessel and its dependence on the antenna phasing. The computation domain is generated directly from the CAD model of the 12 strap antenna structure and the NSTX-U EFIT equilibrium. Note that the simulation in Figure 2 automatically takes into account various effects including the curved antenna/wall structure, and the actual antenna spectrum with the plasma loads, which were either not considered or supplied via a separate calculation in the past. A key parameter for RF induced PMI is the wave electric field on the wall. In figure 3,the normal component ($E_{\perp}$) on vessel wall with the different antenna phasing are compared, showing that a stronger $E_{\perp}$ is induced in the low antenna phasing case. This result is consistent with the experiments in NSTX, where a lower phasing showed poorer core heating performance $[3]$. Further analysis of HHFW using Petra-M including the investigation of interaction between HHFWs and NBI beam fast ions is presented in $[4]$.
The core-edge coupling approach $[5]$ allows for including the spatially dispersive nature of hot core plasma to an FEM based wave simulation for SOL using the complicated geometry (figure 4 (a)). Importantly, it allows for separating the RF power absorbed in core and parasitically in edge. The comparison of D-(He3) and D-(H) heating simulation indicates that a significantly higher (by ~60%) fraction of injected power is absorbed in the core, when the single pass power absorption is strong $[6]$. We extended this approach to use the 3D geometry for the SOL region. The first ever simulation of 3D ICRF wave field combining the hot core, the cold SOL plasma, and 3D antenna structure is shown in figure 4. We used Petra-M for the SOL and antenna regions and the TORIC spectral solver $[7]$ for the core region. Here, the D-(H) heating scenario with the rotated ICRF antenna on Alcator C-Mod is modeled. The result shows that the wave field in the core is smoothly connected with the edge RF wave field, indicating that there is no significant wave reflection at the LCFS. Also, the excited RF wave field does not spread to the high field side of the tokamak, consistent with the high strong pass in this scenario.
The LH wave has a short wave length comparable to the scale length of density fluctuation typically found in the plasma edge. The wave scattering due to such fluctuations has been an issue when modeling the LH current drive. However, since the fluctuation stretches along the magnetic field line, this process cannot be captured well in 2D fullwave simulations. We investigated this issue with Petra-M by adding density profile perturbation which is imposed to follow the magnetic field line. In figure 5, (a) and (b) shows that such a density fluctuation can destroy the resonant cone wave field propagation pattern. The change in the wave field pattern results in a broadening of perpendicular wave number spectrum. In figure 5 (c), the impact of the spectral modification was studied using GENRAY/CQL3D raytracing Fokker-Planck package (http://www.compxco.com), showing that the LH current drive profile is altered dramatically.
In summary, we demonstrated that the fullwave RF simulation for fusion plasmas can scale to a large problem size needed to compute the wave field in the whole fusion device. Our approach allows for including the complicated SOL plasma and antenna structure, while using the hot plasma conductivity for the core. The international validation effort on the WEST tokamak is in progress. This work opens up the possibility to investigate physics issues such as slow wave generation and RF rectified potential $[8]$ in a realistic configuration.
Acknowledgements
Supported by U.S.DOE Contract # DE-AC02-09CH11466
References
$[1]$ S. Shiraiwa et al., EPJ Web of Conferences 157, 03048 (2017).
$[2]$ N. Bertelli et al., Invited talk at the 23rd RF Power in Plasmas Conference (China, 2019). Accepted to be published in AIP Conf. Proceeding (2019).
$[3]$ J. C. Hosea and et al., Phys. Plasmas 15, 056104 (2008).
$[4]$ N. Bertelli et al., this conference.
$[5]$ S. Shiraiwa et al., Nuclear Fusion 57 (8), 086048 (2017).
$[6]$ J. Wright et al., EPJ Web of Conferences 157, 02011 (2017).
$[7]$ M. Brambilla, Plasma Phys. Control. Fusion 41 1 (1999).
$[8]$ J. R. Myra and D. A. D’Ippolito, Phys. Plasmas 22 , 062507 (2015)Speaker: Dr Syun'ichi Shiraiwa (PSFC, MIT) -
ITB formation due to energetic particles. The performance of present-day and future fusion devices is largely determined by turbulent transport generated by plasma turbulence. Any mechanisms able to reduce the overall radial propagation of energy and particles is, therefore, crucial in view of scenario optimization. This contribution presents numerical results of turbulence suppression by supra-thermal ions and experimental results from ASDEX Upgrade, which support these numerical findings. More precisely, the simulations demonstrate, for the first time, the generation of an internal transport barrier (ITB) triggered purely by energetic particles in a monotonic safety factor configuration.
Physical mechanisms. These results are explained in terms of a resonant interaction between ion-driven turbulence and supra-thermal particles, recently identified via gyrokinetic flux-tube simulations [1]. Fast ions have been found to interact with the plasma micro-instabilities through a wave-particle resonance mechanism when the fast ion magnetic-drift frequency is close to the linear frequency of the ion temperature gradient (ITG) microinstability, thus amplifying an otherwise negligible interaction. A theoretical analysis and numerical simulations have shown that the flow of such a resonant energy exchange is determined by the fast particle temperature, density and their gradients and, in turn, sets the direction of the fast ion energy losses. Inward (outward) supra-thermal ion particle and heat fluxes are observed in correspondence with the strongest fast ion stabilization (destabilization). Therefore, in a radially global setup, stabilizing (inward fluxes) and destabilizing (outward fluxes) energetic particle effects on plasma turbulence occur at different radial positions - depending on the local values of the fast particle parameters - thus strongly affecting the bulk ion energy fluxes.
Gyrokinetic GENE simulations. In this contribution, we significantly extend the previous findings - based on flux-tube analyses - by means of global electromagnetic GENE [2] simula-tions with kinetic electrons (realistic proton-electron mass ratio) and a sophisticated bi-Maxwellian model for the energetic particle distribution. The main results are displayed in Fig.1, where the time evolution of the radial profile of the (surface averaged) ion heat flux is shown for the simulations without (left plot) and with (right plot) energetic particles.Fig.1 shows – to our knowledge - the first numerical evidence of an ITB generated solely by the energetic particles. In particular, a full suppression of ITG-driven turbulence is observed in the radial domain \rho_{tor}=[0.2,0.3]. As the turbulent heat flux drops in the ITB region, a corresponding increase of the neoclassical transport is observed. However, the resulting overall flux remains at significantly lower levels compared to the case without energetic particles. The same magnetic equilibrium and kinetic profile of the main plasma are employed in both the simulations regardless the presence of the supra-thermal ions, thus excluding e.g. any effects of the geometry on the generation of the ITB. These results are fully consistent with the physical picture of the wave-particle resonant interaction summarized above. More precisely, due to the local changes in the fast ion temperature and density profiles, the effect of supra-thermal particles on plasma turbulence turns from stabilizing in \rho_{tor}=[0.2,0.25] to destabilizing in \rho_{tor}=[0.25,0.3]. Thus, reverting the sign of the energetic particle energy flux from inward to outward. This sharp change in the direction of the fast ion heat flux strongly affects the shearing rate levels. Localized shearing layers are generated in correspondence of the negative inward and positive outward fast particle heat flux, thus tearing apart the radially elongated ITG eddies and resulting in the first ITB solely triggered by energetic particles.
Experimental evidence. These predict-first numerical results led to the design of an ASDEX Upgrade discharge [3] where this resonant interaction between ITG-driven turbulence and supra-thermal ions was maximized via gyrokinetic GENE simulations. The energetic particle profiles have been calculated with the code TORIC/SSFPQL [4]. A substantial increase of the peaking of the main ion temperature (Ti) profile of the order of ~80% is observed (as shown in Fig. 2) in the radial domain where the resonant interaction is predicted to be maximized. Such a relevant Ti is obtained via on-axis (\rho_{tor}=0) ion-cyclotron-resonant-frequency (ICRF) heating of the H minority species in D plasmas, with a large H concentration of n_H/n_e ≈ 0.11. Furthermore, no degradation of the energy confinement is observed during a ramp-up of the ICRF power, thus suggesting a substantial reduction of the anomalous turbulent transport. More specifically, a power balance analysis reveals a central region of improved confinement as the ICRH power is increased. The resulting radial heat flux profiles obtained with global GENE simulations exhibit the same features as in Fig. 1. The overall fluxes (including the neoclassical contribution) are in agreement with the experimental power balance only in the presence of energetic particles.Conclusions. This contribution provides, for the first time to our knowledge, numerical results that energetic particles can effectively trigger internal transport barriers in realistic tokamak configurations. These findings are supported by experimental evidence at ASDEX Upgrade and represent an essential step forward to access unique and still unexplored high confinement regimes.
References
-
A. Di Siena et al., NF 58, 054002 (2018); PoP 26, 052504 (2019).
-
F. Jenko et al., PoP 7, 1904 (2000); T. Görler et al., JCP 230, 7053 (2011); A. Di Siena et al., PoP 25, 042304 (2018).
- R. Bilato et al., to be submitted (2020).
- M. Brambilla and R. Bilato, NF 49, 085004 (2009); R. Bilato et al., NF 51, 103034 (2011).
Speaker: Alessandro Di Siena (Max Planck Institute for Plasma Physics) -
-
Most of high-performance discharges based on the advanced scenario have shown the active generation of the Alfvén eigenmodes (AE), driven by enhanced fast-ion pressure gradient and broad current density profile in the core region$^{1,2}$.
Among various AE control tools, it has been found that the ECCD and ECH are able to mitigate or suppress the toroidal Alfvén eigenmodes (TAE) and the reversed-shear Alfvén eigenmodes (RSAE) in DIII-D$^{2-4}$, ASDEX-Upgrade$^{5,6}$ and KSTAR tokamaks$^{7}$ as well as helical devices such as TJ-II$^{8}$ and LHD$^{9}$. ECCD tailors the current density profile near or inside the location of the TAEs so that continuum damping is enhanced effectively. This research shows the feasibility of the active control of TAE using the electron cyclotron current drive (ECCD) and the heating (ECH) in the high-performance discharges.
Previous work has shown the co-current directional ECCD mitigates and suppresses the TAEs in the elevated $q_0$ discharge$^{7}$, leading to the increase in $\beta_N$, $\beta_P$ and the reduction of neutron deficit. In addition to co-ECCD applications, the effect of counter-current directional ECCD was investigated. Experiments show that both co- and counter-current directional ECCD are able to mitigate and suppress the TAE activities while the ECCD deposition location stays in the specific range (–12 cm < $Z_{EC}$ < +3 cm), as shown in figures 1 and 2. Moreover, co-ECCD shows TAE mitigation effects in a wider deposition range than the counter-ECCD. It is found that the co-ECCD is more useful to enhance the overall performance, associated with TAE mitigations. Plasma performance in the co-ECCD application is higher than in the counter-ECCD (co-ECCD: $\beta_N$ ~ 2.4, $\beta_P$ ~ 2.1, neutron ~ 920 (a.u.); counter-ECCD: $\beta_N$ ~ 2.1, $\beta_P$ ~ 1.8, neutron ~ 830 (a.u.)). Maximum plasma performance in the TAE mitigation stage seems to be associated with the elevated plasma pressure (or $\beta_N$, $\beta_P$), change in internal inductance ($l_i$) and $q_0$, which affect the effective ECCD deposition window. Radial locations of both co- and counter-ECCD deposition affect the major damping channels such as the continuum damping (magnetic shear)$^{10}$ and the thermal ion Landau damping (central electron and ion temperatures). In particular, NOVA-k$^{11}$ modeling indicates that continuum damping is enhanced more in case of co-ECCD by increasing core magnetic shear. Total damping rate in the co-ECCD case is higher than the case of counter-ECCD, hence the TAE amplitude in the initial stage of counter-ECCD is higher than in the co-ECCD application. Radiative damping seems to be negligible because the observed TAEs are the low-n modes (n < 6). Another stabilizing effect by enhancing the plasma pressure is expected in both co- and counter-ECCD, which is beneficial to mitigate TAEs by making the normalized pressure gradient larger than the critical level$^{12,13}$.
To extend the operational space of the ECCD-assisted TAE control experiment, empirical scanning in the parameter space, represented by the central safety factor ($q_0$)$^{14}$, core $T_e$ (or fast-ion slowing-down time) and the internal inductance ($l_i$), has been carried out. Based on the parameter scan, experimental condition with high $q_0$, low $l_i$ and high $T_e$ is prone to excite the TAEs in the core due to reduction of continuum and beam-ion damping rates and the enhanced fast-ion pressure gradient (drive). From this parametric study, mild off-axis deposition of ECCD ($\rho_{pol,ECCD}$ ~ 0.15 – 0.2 for high $\beta_{P}$ scenario, $\rho_{pol,ECCD}$ ~ 0.25 – 0.35 for high $q_{min}$ scenario) is quite beneficial to reduce or suppress the core TAE activities, hence the plasma performance is elevated significantly in the advanced scenarios. Experimental database provides the best option for ECCD/ECH-assisted active TAE control in the variety of advanced scenarios. Co-ECCD has shown TAE mitigation and suppression with elevated performance, however, additional consideration for keeping high $q_{min}$ in the TAE control phase should be focused on the optimization of counter-ECCD application as demonstrated in shot# 22937 (figure 2), which shows suppression of n = 3, 4 TAEs under the conditions of high $q_0$ ~ 1.8 and m / n = 2/1 tearing-mode avoidance.References
$^{1}$ Holcomb C.T. et al., 2015 Phys. Plasmas 22 055904
$^{2}$ Heidbrink W.W. et al., 2014 Plasma Phys. Control. Fusion 56 095030
$^{3}$ Van Zeeland M.A. et al., 2009 Nucl. Fusion 49 065003
$^{4}$ Kramer G.J. et al., 2017 Nucl. Fusion 57 056024
$^{5}$ Sharapov S.E. et al., 2018 Plasma Phys. Control. Fusion 60 014026
$^{6}$ Garcia-Munoz M. et al., 2019 Plasma Phys. Control. Fusion 61 054007
$^{7}$ Kim J. et al., 2019, 16th IAEA Technical Meeting on Energetic Particle Physics (Shizuoka-City, Japan), ID 49.
$^{8}$ Cappa A. et al., 2019, 16th IAEA Technical Meeting on Energetic Particle Physics (Shizuoka-City, Japan), ID 58.
$^{9}$ Nagaoka K. et al., 2013 Nucl. Fusion 53 072004
$^{10}$ Zonca F. and Chen L. 1993 Phys. Fluids B 5 3668
$^{11}$ Cheng C.Z. 1992 Phys. Rep. 211 1
$^{12}$ Fu G.Y. 1995 Phys. Plasmas 2 1029
$^{13}$ Sharapov S.E. et al., 1999 Nucl. Fusion 39 373
$^{14}$ Strait E. et al., 1993 Nucl. Fusion 33 1849
Speaker: Dr Junghee Kim (National Fusion Research Institute) -
The neutron and tritium production during the pre-DT, Pre-Fusion Plasma Operation, (PFPO) phase of ITER need to be quantified in view of the plans for commissioning and operation of the heating systems in hydrogen and helium plasmas, as discussed in the ITER Research Plan (IRP) 1. In the assessment presented here, we consider a number of tritium and neutron sources of different origins. In hydrogen plasmas with Ohmic and ECRH heating, neutrons and tritium appear due to the presence of residual deuterium nD/nH ~ 1.5 10-4 caused by the finite purity of the hydrogen fuel. Application of hydrogen NBI (Ep = 500-870 keV) and ICRH H and 3He minority heating schemes in He and H plasmas respectively, creates and accelerates fast protons and 3He ions which produce neutrons interacting with Be impurities. In addition, fast protons produce fast D and alphas, which in turn produce neutrons interacting with Be impurities (figure 1). The evidence for the importance of nuclear fusion reactions of fast particles produced by auxiliary heating with Be impurities has been demonstrated in JET experiments [2]. Besides stationary plasma conditions, we also estimate the possible sources that can lead to in-vessel component activitation in this phase produced during plasma disruptions should runaway electrons be formed. This includes both bremsstrahlung emission and secondary neutrons. The impact of runaway electrons on the divertor or first wall will yield different bremsstrahlung and secondary neutron production levels.
An assessment of neutron production has been carried out for a wide range of plasma scenarios foreseen in the ITER Research Plan 1 for the PFPO phase, including the application of H0-NBI, ECRH, ICRH H and 3He minority heating, and with the 3 ion ICRH heating scheme with 3He minority as a function of Be fraction, fBe= nBe/ne. Fast deuteron, alpha and neutron production by the interaction of fast protons originating from NBI and hydrogen minority ICRH heating with Be impurity: 9Be(p,d)2a, 9Be(p,a)6Li, 9Be(p,n)X, is calculated using the ASTRA-NBI [3,4] and TORIC-SSFPQL[5] codes. On the basis of the calculated sources of fast deuterons and alphas with these codes, the neutron sources due to secondary reactions with Be impurities: 9Be(d,n)X, 9Be(a,n)X are derived. Neutron sources produced by fast 3He minority ions accelerated by ICRH in the reaction 9Be(3He,n)X (4) are calculated for the 3-ion heating scheme [6] and the 3He minority scheme. The impact of the synergy between the H0-NBI ions and hydrogen minority ICRH heating on neutron production is assessed for 7.5 MA/2.65 T Helium H-mode plasmas with 33 MW (NBI), 20 MW (ECRH) and 20 MW (ICRH) heating power levels to be performed in PFPO-2. [7]. In these plasmas, the main source of neutrons is due to the interaction of Be impurities with suprathermal protons produced by ICRH and NBI. Secondary fusion reactions occur due to the interaction with beryllium of the fast deuterium and alphas from the 9Be(p,d)2a , 9Be(p,a)6Li reactions as well but the magnitude of this neutron source is much lower. Our modelling shows that the synergy between the H0-NBI and the hydrogen minority ICRH noticeably reduces the neutron yield (Sn,ICRH + Sn,NBI > Sn,ICRH+NBI), which is due to widening of the ICRH absorption and reduction of the fast proton tails due to the preferential absorption of ICRH waves by fast NBI protons in the plasma (figure 2).
Local fractions of the NBI and ICRH fast ion pressure and their gradients during PFPO-2 are higher than those of fast alphas in the ITER Q = 10 baseline scenario, making AEs unstable over a wide range (x < 0.5) of the low magnetic shear. For instance, the injection of H0-NBI with Ep=0.87 MeV in helium plasmas is superalfvenic for the whole range of magnetic fields to be explored in ITER. The stability of TAE modes has been analysed in these plasmas which typically have both a high pressure of suprathermal particles and a Weak Reversed Shear (WRS) current profile due to the large current drive produced by the NBI [8,9]. The possible impact of saw-tooth oscillations and TAEs on neutron production in these plasma scenarios has been assessed based on a linear instability analysis and it has been found that they can noticeably reduce the local neutron density source and beam-driven current drive but have moderate impact on the integral neutron production.
Disclaimer: ITER is the Nuclear Facility INB no. 174. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization. This publication is provided for scientific purposes only. Its contents should not be considered as commitments from the ITER Organization as a nuclear operator in the frame of the licensing process.
References
1 ITER Research Plan, ITER Technical Report ITR-18-03 (1 “ITER Research Plan within the Staged Approach” (Level III – Provisional Version). ITER Technical Report ITR-18-003, 17 September 2018)
[2] A.V. Krasilnikov, et al, Nucl. Fusion 58 (2018) 026033
[3] A. Polevoi, et al “Benchmarking of the NBI block in ASTRA code versus the OFMC calculations”, JAERI-DATA-Code 97-014, March 1997,
https://inis.iaea.org/collection/NCLCollectionStore/_Public/28/052/28052091.pdf
[4] A.R. Polevoi et al, “Reassessment of Steady State Operation in ITER with NBI and EC Heating and Current Drive“, P5.1012, 46th EPS Conference on Plasma Physics, 8 - 12 July 2019, Milan, Italy, hppt://ocs.ciemat.es/EPS2019PAP/pdf/P5.1012.pdf
[5] R. Bilato, M. Brambilla, et al., Nucl. Fusion, 51 (2011) 103034
[6] Ye.O. Kazakov et al., Nature Physics 13, 973-978 (2017)
[7] R. Bilato, et al, “Synergies between H-NBI fast-ions and ICRF heating in the non-activated operational phase of ITER”, 45th EPS Conference on Plasma Physics, July 2-6, Prague, Check Republic, http://ocs.ciemat.es/EPS2018PAP/pdf/P1.1070.pdf
[8] G.Y. Fu, C.Z. Cheng, and K.L. Wong, Phys. Fluids B 5, 4040 (1993)
[9] N.N. Gorelenkov, C.Z. Cheng, and G.Y. Fu, Phys. Plasmas 6, 2802 (1999)
Speaker: Alexei POLEVOI (ITER Organization) -
The primary goal of Tokamak based fusion reactors is to achieve a self-sustained plasma satisfying ignition criterion by maximizing the product of plasma density, temperature and energy confinement time. There are limits on achievable plasma density and temperature due to current driven or pressure driven instabilities. Also, the Tokamak performance can be enhanced by operating with advanced configurations and it is necessary to study them extensively to maintain specific profile of plasma parameters. The understanding and controlling of the energy confinement time and plasma profiles fall in the domain of plasma transport. In self-sustained Tokamak based fusion reactors, the plasma heating will be mainly carried out by the alpha-particles that are produced in the fusion of deuterium and tritium. It is thus necessary to study the burning plasma performance with plasma heating due to alpha-particles. This study is aimed at finding the optimal power deposition due to alpha-particles required to maintain a particular plasma profile, for which we have developed a 1.5-dimensional self-consistent transport model. We present a qualitative comparison of the results of this model with experimental results and results of other models of TFTR {1}, JET {2} and ITER-like {3} cases. We also present our simulation results for SST-2-like {4} case.
To perform a burning plasma transport simulation, it requires to solve continuity equations, energy equations, Ohm’s Law and Maxwell equations. The steady state momentum equation to simulate plasma equilibrium that includes alpha-particle physics is coupled with this model. The plasma equilibrium is considered as fixed-boundary equilibrium and the boundary is defined by a set of points that define the plasma shape. In this study, transport equations are considered in the flux co-ordinates and the plasma configuration is assumed to be toroidally axisymmetric. The continuity and energy equations used in our model are {5}:
$ \frac{\partial\langle n_i\rangle}{\partial t}=-\frac{1}{V'}\frac{\partial}{\partial\rho}(V'\Gamma_i)+\left[\frac{dn}{dt}\right]_{iz}+\left[\frac{dn}{dt}\right]_{fus} $
$ \frac{\partial\langle n_n\rangle}{\partial t}=-\frac{1}{V'}\frac{\partial}{\partial\rho}\left(V'\Gamma_n\right)-\left[\frac{dn}{dt}\right]_{iz} $
$ \frac{\partial p_i}{\partial t}=-\frac{2}{3V'}\frac{\partial}{\partial\rho}\left(V'\left(q_i+\frac{5}{2}T_i\Gamma_i\right)\right)+P_{\alpha i}+P_{coll} $
$ \frac{\partial p_e}{\partial t}=-\frac{2}{3V'}\frac{\partial}{\partial\rho}\left(V'\left(q_e+\frac{5}{2}T_e\Gamma_e\right)\right)+P_{\alpha e}-P_{coll}+P_{ohm}-P_{rad}+P_{aux} $
Here, $\rho$ is flux coordinate, $\langle f\rangle$ denotes flux-surface average of $f$, $t$ is time, $V$ is volume within a flux-surface, $\Gamma_i$ is ion flux, $\Gamma_e$ is electron flux, $\Gamma_n$ is neutral particle flux, $\langle n_i\rangle$ is flux-surface averaged ion density, $\langle n_n\rangle$ is flux-surface averaged neutral particle density, $p_i$ is ion pressure, $p_e$ is electron pressure, $T_i$ is ion temperature, $T_e$ is electron temperature, $q_i$ is ion heat flux, $q_e$ is electron heat flux, $[dn/dt]_{iz}$ is rate of change of density due to ionization, $[dn/dt]_{fus}$ is rate of change of density due to fusion, $P_{\alpha i}$ is power per unit volume due to alpha-particle heating for ions, $P_{\alpha e}$ is power per unit volume due to alpha-particle heating for electrons, $P_{coll}$ is power per unit volume due to ion-electron collisions, $P_{ohm}$ is Ohmic power per unit volume, $P_{rad}$ is power per unit volume due to radiation, $P_{aux}$ is power per unit volume due to auxiliary heating. We assume quasi-neutrality for electron density.
We model alpha-particle heating $P_{\alpha i}$ and $P_{\alpha e}$ by solving the Fokker-Planck equation for alpha-particles. We find that energy-loss characteristic time for 3.5 MeV alpha-particles is ~1s, which corroborates solving for its energy-loss equation along with transport simulation. We assume Bremsstrahlung radiation loss to be the dominant mechanism among radiation losses and ignore other mechanisms.
The transport equations are solved using LCPFCT {6}, which uses the explicit algorithm FCT (Flux-Corrected Transport). Important features of FCT are that it is conservative and maintains positivity and monotonicity. For our purpose we use a modified version of LCPFCT that is compatible with flux co-ordinates. To solve Grad-Shafranov equation, we use VMOMS {7}, which uses a variational moment method that expands radial co-ordinates as Fourier series in flux co-ordinates and solves the resulting equations using Shooting method for a closed boundary system in 2-dimensions. We use an empirical mixed Bohm/gyro-Bohm model {8} to find the particle and heat fluxes for ions and electrons.
In our simulations, total runtime consists of several energy confinement times. We start with ramp-up of plasma by evolving its LCFS and heat with auxiliary heating until alpha-particle heating dominates the plasma heating. Plasma is then maintained at a plateau region, where its LCFS is not further evolved. We observe formation of edge pedestals, which are characteristics of H-mode Tokamaks. For a snapshot of a typical ITER-like case at its plateau region, alpha-particle heating profile obtained from our alpha-particle heating equation is shown in figure below.
References:
{1} Zweben S.J. et al. 2000. “Alpha particle physics experiments in the Tokamak Fusion Test Reactor” Nuclear Fusion 40, 1
{2} Thomas P. R. et al. 1998. “Observation of Alpha Heating in JET DT Plasmas” Physical Review Letters 80, 25
{3} Schneider M. et al. 2019. “Modelling one-third field operation in the ITER pre-fusion power operation phase” Nuclear Fusion 59, 126014
{4} R Srinivasan and the Indian DEMO Team, Fusion Engineering Design, 112 (2016) 240
{5} Hinton, F.L.; Hazeltine R.D. 1976. “Theory of Plasma Transport in Toroidal Confinement Systems.” Reviews of Modern Physics, 48, 2, Part-1
{6} Boris, J.P.; A.M. Landsberg; E.S. Oran; and J.H. Gardner. 1993. "LCPFCT - A Flux-Corrected Transport Algorithm for Solving Generalized Continuity Equations." NRL Memorandum Report 93-7192
{7} L. Lao et al. 1982. “VMOMS – A Computer Code for Finding Moment Solutions to the Grad-Shafranov Equation”, Computer Physics Communications 27, 129-146
{8} Erba M et al. 1997. “Development of a non-local model for Tokamak heat transport in L-mode, H-mode and transient regimes” Plasma Physics and Controlled Fusion, 39, 2Speaker: Mr Udaya Maurya (Institute for Plasma Research) -
The steady-state superconducting tokamak (SST1) [ref1], having major and minor radius of 1.1m and 0.2 m respectively, equipped with a Copper-based central solenoid to provide the required loop voltage. Like in other superconducting tokamaks [ref2-3], SST1 tokamak also relies on ECR breakdown technique to form Ohmic plasmas. Plasma preionization and current start-up is assisted by the 42 GHz ECR system and the LHCD system [ref4-5] is used for driving plasma current non-inductively to achieve continuous wave (CW) discharges. In SST-1, loop voltage amplitude is not only smoothen out inside the vessel but also, delayed by ~12ms due to the continuous vacuum vessel. Moreover, the equilibrium field is provided by a pair of Vertical field coils located outside of cryostat that required a perfect feed-forward current profile and hence, poses challenges for having repeatable Ohmic breakdown. With recent modifications in the magnetic null field configuration [ref6], good repeatable Ohmic plasma discharges could be produced with narrow EC pulse (~50ms) and that allowed in carrying out LHCD current drive experiments on SST1 machine.
The schematic of the major configuration of the machine is displayed in Figure-1. The direction of the toroidal magnetic field is in anti-clockwise direction and the plasma current is also driven in the same direction by flowing current in Ohmic transformer in anti-clockwise direction. The current in the vertical field coils is driven in clockwise direction to provide equilibrium to the plasma. In order to drive plasma current non-inductively using lower hybrid waves (LHW’s) in the same direction as of plasma current, the grill launches LHW’s in clockwise direction so that momentum from the waves is transferred to the resonant electrons through collisionless Landau damping mechanism. To launch LHW’s in the clockwise direction, the grill is phased from left to right. LHCD system for SST1 machine comprises of klystrons, each rated for 0.5MW-CW rf power at a frequency of 3.7 GHz. The grill antenna having two rows, each row accommodating 32 waveguide elements, is placed 3mm behind the limiter. As the periodicity of the grill antenna is 9mm, it launches waves having parallel refractive index (N//) centered at 2.25 when the adjacent waveguide elements are relatively phased at 90^o. Each waveguide elements are connected to adjustable phase-shifters which provides flexibility in launching the waves with desired N//. The spectrum with different phasing of the grill antenna is shown in figure-2. In the experiments, the phasing of the grill antenna is varied from 90^o to 130^o to launch LH waves with N// centered at 2.25 to 3.25. The toroidal magnetic field is kept at 1.5 T and plasma is formed with plasma current up to ~75kA at [0.9 – 2] x 10-5 mbar pressure of hydrogen gas. Up to ~200 kW of LH power is injected in to the plasma. A typical plasma shot with LH driven plasma current is shown in figure-3. Initially ECR assisted Ohmic discharge is formed and once a stable target plasma at ~75kA is achieved, ~125kW of LH power is injected at ~250 ms. Thereafter, the plasma is maintained with LHW’s up to ~650ms. Figure-3 also shows the temporal evolution of loop voltage and vertical magnetic field. The zero loop voltage current drive indicates complete non-inductive current drive fully supported by LH power. The estimated efficiency of current drive, in worst case, is ~80kA/125kW ~0.67 A/W assuming the entire injected power is absorbed for CD. The line averaged density of similar shots is estimated to be ~[6-8]x10^18 m^{-3} and yields figure of merit for LH current drive as, η_LH = (n_e R_o I)/P= (0.7x〖10〗^191.180)/120 ~ 0.05 x 〖10〗^20 A〖/W/m〗^2.
The beneficial effect of suppression of hard x-ray with LH driven current is demonstrated on SST1 machine also as shown in figure-4. The spikes in hard x-rays after LH injection and before ~400ms is attributed to plasma hitting the walls. To further establish that population of suprathermal electrons are generated when injected LH power interacts with tail electrons of the distribution function, the CdTe detectors were employed to measure the photons coming out from plasma lying in the energy range from 20keV to 200 keV. The PHA data shown in figure-5 clearly shows that the counts increases significantly in the presence of LH power and confirms generation of suprathermal electrons when LHW’s interact with tail of the distribution function of the electrons. In experiments with modulated LH power, the entire plasma current could be supported with LH power alone corresponding to nearly zero loop voltage signal with reduced plasma-wall interaction. The effect of modulation could also be observed in 2nd harmonic ECE signals, confirming generation of suprathermal electrons with LH.
References:
[ref1] Saxena, Y. C., SST1 Team, Nucl. Fusion, 40 (6), 1069, 2000.
[ref2] J. Bucalossi et al, Nucl. Fusion, 48 (2008) 054005.
[ref3] Journal of the Korean Physical Society, Vol. 51, No. 4, October 2007, pp. 1313-1319
[ref4] Bora, D., Sharma, P. K., Rao, S. L., Trivedi, R. G., et. al., Fusion Engg. Design, 82, 141, 2007.
[ref5] Sharma, P. K., Ambulkar, K. K., Parmar, P. R., Virani, C. G., et. al., Jour. of Phys. Conf. Series, 208, 012027, 2010.
[ref6] Sharma, P. K., Jain, Y. M., Ambulkar, K. K., Parmar, P. R., et. al., Plasma and Fusion Research, 13, 3502100, 2018.Speaker: Dr Promod Kumar Sharma (Institute for Plasma Research) -
Regulation of the $q$ profile via feedback control has been recently demonstrated in EAST in both L-Mode and H-mode experiments. Extensive studies have shown that the $q$ profile, which is closely related to poloidal magnetic flux profile, is a key factor to achieving advanced-tokamak operating scenarios that are characterized by improved confinement and the non-inductive sustainment of the plasma current necessary for steady-state operation. A general framework for real-time feedforward + feedback control of magnetic and kinetic plasma profiles as well as scalars has been implemented in the EAST Plasma Control System (PCS). Moreover, a first-principles-driven, control-oriented model of the poloidal magnetic flux profile and internal energy evolutions has been used to design feedforward {1} and feedback {2} model-based control algorithms and to tune non-model-based control algorithms in closed-loop simulations. Several of these proposed controllers have been tested successfully in reference tracking and disturbance rejection experiments in EAST. These experiments constitute the first time ever closed-loop $q$-profile regulation has been successfully achieved in EAST.
The controllers have the capability of either regulating several points of the $q$ profile, minimizing the integrated squared error between actual and target profiles, or controlling integral properties such as the internal plasma inductance $l_i$. Moreover, the controllers can simultaneously regulate $\beta_N$, which is another plasma property playing a key role in the access to high-performance, MHD-stable modes of operation. This level of controllability has been achieved in EAST by placing several critical actuators under the PCS command, namely the ohmic coils, two Lower Hybrid Wave (LHW) sources, and four Neutral Beam Injection (NBI) sources. The so-called Profile Control category recently implemented in the EAST PCS, which houses the $q$-profile+$\beta_N$ control algorithms, does not directly control these actuators. Instead, it sends requests for the needed total plasma current and non-inductive powers to the different actuator categories. In order to enable the use of NBI power for $q$-profile+$\beta_N$ control, a pulse-width-modulation algorithm was also implemented in the EAST PCS to convert the power requests from the Profile Control category into on/off time commands for the beams. While this actuator mechanism proved to be too powerful for smooth $\beta_N$ regulation in L-mode discharges, it is anticipated that it can play an important role in H-mode plasmas. Both the $q$ profile at 11 points and $\beta_N$ are passed to the Profile Control category by pEFIT {3}, an equilibrium reconstruction code exploiting the massively parallel processing cores of graphic processing units (GPUs).
Fig. 1 shows feedback regulation of the $q$ profile at three points in space, namely $q(\hat{\rho}=0.1)$, $q(\hat{\rho}=0.5)$, $q(\hat{\rho}=0.9)$, by using the ohmic coils, the low-frequency (2.45 GHz) LHW source, and the high-frequency (4.60 GHz) LHW source as actuators for EAST shot #95183. The target evolutions for the $q$ profile at these three points, which have been obtained from a previous shot to ensure feasibility, are shown in dashed red lines in Fig. 1(a), Fig. 1(b) and Fig. 1(c). The solid magenta lines show the evolutions of the $q$ profile at these points for feedforward-only EAST shot #95176. The feedforward actuation for the plasma current, the 2.45 GHz LHW power, and the 4.60 GHz LHW power, shown respectively in Fig. 1(d), Fig. 1(e) and Fig. 1(f) by the solid magenta lines, is not adequate enough to track the desired targets. This can be appreciated from Fig. 1(a), Fig. 1(b) and Fig. 1(c), where the associated feedforward-only evolutions of the $q$ profile are also shown by solid magenta lines. When feedback control is turned on during shot #95183 for 2s$
As the number of actuators is increased (by testing the use of NBI power modulation in H-mode plasmas and incorporating the command of ECRF and Ion Cyclotron Range of Frequency (ICRF) H&CDs under the PCS), the quality of the real-time reconstruction is improved (by constraining pEFIT with POlarimeter-INTerferometer (POINT) measurements {4}), and the prediction accuracy of the control-level models used for control design is enhanced (by further developing control-physics understanding and continuing validation efforts), the capability of tightly regulating the $q$ profile and $\beta_N$ will be further augmented in order to routinely enable the access to long-pulse, disruption-free, high-performance operation in the EAST tokamak.
This work has been supported in part by the U.S. Department of Energy under DE-SC0010537.
{1} WANG, H., SCHUSTER, E., RAFIQ, T., KRITZ, A., and DING, S., Fusion Eng Des 123 (2017) 569.
{2} WANG, H. and SCHUSTER, E., Fusion Eng Des 146 (2019) 688.
{3} HUANG, Y., XIAO, B., LUO, Z., et al., Fusion Eng Des 112 (2016) 1019.
{4} HUANG, Y., XIAO, B., LUO, Z., et al., Fusion Eng Des 120 (2017) 1.Speaker: Prof. Eugenio Schuster (Lehigh University) -
A novel integrated model GOTRESS+ has been developed, which consists of the iterative transport solver GOTRESS as a kernel of the integrated model$^{1,2}$, the equilibrium and current profile alignment code ACCOME and the neutral beam heating/current-drive code OFMC. GOTRESS is able to find out an exact steady-state solution using global optimization techniques, enabling us to robustly deal with even stiff transport models like TGLF. Due to numerical algorithms implemented yielding a large amount of data, a surrogate model can readily be established with Machine-Learning (ML) techniques employed, which makes computation much faster$^2$. This unique characteristic of GOTRESS in terms of affinity to ML techniques distinguishes GOTRESS+ from other similar codes/frameworks$^3$ and makes it very competitive. GOTRESS+ is then suitable for developing operation scenarios of JT-60SA as well as ITER and validating the feasibility of the scenarios. In conjunction with the EPED1 model, GOTRESS+ has gained the capability of predicting an entire plasma with pedestals, essential for achieving high-performance plasmas.
GOTRESS+: A novel integrated model for developing and validating scenarios
Eliciting the performance of JT-60SA as high as possible, a variety of typical operation scenarios have been proposed, having their specific target parameters. An assessment of their feasibility must be performed by an integrated transport code with a transport model. For this purpose, the integrated time-dependent code TOPICS has been employed with the CDBM transport model used and some of the scenarios have been confirmed to meet the target parameters. However, the advanced transport model TGLF is difficult to adopt in conventional transport codes because it takes long time to compute and it sometimes causes wiggled profiles that make a simulation result unphysical due to its stiff nature.
The iterative transport solver GOTRESS has been developed to manage such stiff transport models by overcoming these difficulties$^{1,2}$. GOTRESS makes use of global optimization techniques such as genetic algorithms and the Nelder-Mead method to solve steady-state transport equations and is free from evaluating gradients of temperature, which are crucial inputs to stiff transport models. This unique characteristic avoids the potential problem associated with numerical differentiation. GOTRESS has successfully been executed together with TGLF on an MPMD framework$^2$.
Also, a method for constructing a surrogate model of transport models based on an artificial neural network (NN) model was established$^2$, taking advantage of the GOTRESS feature that produces a large amount of data in a straightforward fashion. Due to the excellent hyperparameter optimization technique, the surrogate model is capable of reproducing the behavior of the original model with high precision, as shown in Fig. 1. Evaluation of the surrogate model of TGLF is about $10^4$ times faster than the original TGLF. The ion temperature $T_i$ profile predicted by GOTRESS with the NN surrogate model perfectly matches that by GOTRESS with TGLF, which successfully reproduced experimental observation in JT-60U H-mode discharge, as shown in Fig. 2. This characteristic realizes the frequent use of a high-fidelity transport model as a surrogate model in developing operation scenarios.
With GOTRESS utilized as a kernel of the model, an integrated model GOTRESS+ has been developed, in conjunction with ACCOME and OFMC. The main advantages of GOTRESS+ over conventional time-dependent codes including TOPICS are:
- GOTRESS+ takes less time to finish computation to obtain a steady-state solution.
- GOTRESS can deal with stiff transport models and
- can rigorously find out a steady-state solution if it exists even in the cases where time-dependent transport codes are hard to do.
GOTRESS+ with CDBM is applied to JT-60SA scenario 5-1, what is called a fully-current-driven high-$\beta$ scenario. Even in this plasma with negative magnetic shear, the $q$ profile converges well. This scenario has already been assessed by TOPICS, which makes it possible to compare both results for benchmarking purposes. The GOTRESS+ prediction of $T_i$ excellently conforms to the TOPICS one, albeit very different numerical schemes adopted. The agreement is also obtained for the electron temperature $T_e$. Regarding computation time, GOTRESS+ runs about six times faster than TOPICS for this case. This comparison revealed the effectiveness and the validity of GOTRESS+.
Coupling of GOTRESS+ and the EPED1 model for entire plasma predictions
In the GOTRESS+ simulations so far, the boundary condition has typically been set to $\rho=0.8$ so as not to include the pedestal, because the physics governing the formation of the pedestal is different from that in the core region. The EPED1 model has widely been used to predict the pedestal height and width$^4$. GOTRESS+ is now extended to include our EPED1 model that uses MARG2D, the ideal MHD stability code covering from low to high $n$ modes, replacing the ELITE code originally used in Ref. 4, a fact which makes it possible to examine the ideal linear stability over the entire plasma. This feature is of critical importance for operation scenario development. A flow chart of cooperative calculation with EPED1 is shown in Fig. 3. Since EPED1 needs to be given profiles in functional form, the profiles computed by GOTRESS must be fitted in tangent hyperbolic form by means of generic algorithm$^1$. In EPED1, temperature profiles are scaled up/down if the initial ones render a plasma MHD stable/unstable and then an equilibrium is reevaluated using ACCOME. In this procedure, the density pedestal is subsequently changed even though GOTRESS does not currently calculate particle transport. This process is repeated until the marginal stability condition is met. A whole procedure of GOTRESS+ illustrated in Fig. 3 is then repeated until convergence. Figure 4 shows a converged $T_e$ profile by GOTRESS+ in JT-60SA after 5 iterations, which is a self-consistent solution between transport, the equilibrium, the safety factor, the heatings and pedestal stability.
$^1$M. Honda, Comput. Phys. Commun. 231, 94 (2018).
$^2$M. Honda and E. Narita, Phys. Plasmas 26, 102307 (2019).
$^3$O. Meneghini et al., Nucl. Fusion 57, 086034 (2017).
$^4$P. B. Snyder et al., Phys. Plasmas 16, 056118 (2009).
Speaker: Dr Mitsuru Honda (National Institutes for Quantum and Radiological Science and Technology) -
Experiments have been carried out to explore quiescent H-mode (QH-mode) $[1]$ scenario under the condition of low torque and pure RF injection (true zero torque) with ITER like tungsten divertor on EAST, which will facilitate the eventual use of this scenario on ITER and other reactors. EAST achieved the stationary QH-mode over 50-80 energy confinement time or 3-6 current relaxation time with the energy confinement close or slightly below $H_{98,y2}$ scaling ($H_{98,y2}$ ≤ 1.0) by using counter and co-Ip neutral beams (-0.5 Nm ≤ $T_{inj}$ ≤ -1.0 Nm) or pure RF power with non-resonant magnetic fields (NRMF) in separated experiments. Characteristics of coherent modes or broadband MHD fluctuations with enhancing the edge turbulence transport and replacing ELMs during QH-mode phase are observed from the spectrum analysis of Mirnov probes, microwave reflectometer and other diagnostics.
EAST first QH-mode was accessed by using the counter NB to produce plasma rotation in the counter direction. Results show that stationary ELM-free regime obtained more easily with lower recycling and stronger counter NBI torque. Figure 1 shows an example of QH-mode discharge ($I_p$~0.5MA, $B_T$=2.5T, $q_95$~5.0) with ELM free from 3.5s to 7.0s. The plasma configuration is an upper single null (USN) with the strike points on the tungsten divertor. The two counter NBs were used to access QH-mode and then one counter NB was schemed to maintain this regime, where the beam energy was optimized to mitigate the prompt fast ion losses for reduced impurity influx, together with zero gas feed-forward after 2.2s. The spectrum analysis shows that coherent multi-harmonic modes ($f$ ~60/120kHz $n$=-1/-2) appear to replace ELMs. Here n is the toroidal mode number.
Experiments on the exploration of lower torque have shown that QH-mode can be achieved by using near balanced NBI at low recycling high density regime. The coherent modes become weak and change to broadband fluctuations with more balanced NBI, similar to the previous DIII-D experiments $[2]$. The relative high density (<$n_e$>/$n_{GW}$~0.6) operation might be a knob in this experiment which can reduce the shine-through and fast ion loss because EAST NB has the small tangential tilt angle.$[2]$
Using the 3D fields, EAST recently achieved the QH-mode with the RF only heating, suggesting that QH-mode can be obtained from a start with zero injected torque rather than ramping the torque down from a high value $[3]$. Figure 2 presents a typical discharge of QH-mode ($I_p$~0.5MA, $B_T$=2.5T, $P_{LHW}$=3.5MW, $P_{EC}$=0.9MW, $q_{95}$~5.0, USN) with the application of n=2 non-resonant rotating field, where the upper array coils phase rotated 3 periods within 2.5s. ELMs completely disappeared and were replaced by the quasi-coherent/broadband MHD fluctuations from 3.5s to 6.5s. ELMs came back after 6.5s suggesting that the NRMF on the plasma edge is the dominant effect. Note that the RF power coupling was also a challenge with the reduced or zero gas puff.
In summary, recent EAST experiments demonstrated QH-mode scenarios with near balanced NBI and with RF-only heating. More analysis of experimental data from core and edge diagnostics is ongoing, to investigate edge stability and particle transport. With the upgrades $[4]$ in 2020, EAST will be capable to further extend QH regime towards more reactor relevant by exploiting two extra ECH gyrotrons and the condition of full metal walls.
This work was supported in part by National Natural Science Foundation of China under Grant No. 11975274, US Department of Energy under DE-SC0010685 and DE-FC02-04ER54698.
$[1]$ Greenfield C.M. et al 2001 Phys. Rev. Lett. 86 4544
$[2]$ Garofalo A. M. et al 2015 Phys. Plasmas 22 056116
$[3]$ Garofalo A.M. et al 2011 Nucl. Fusion 51 083018
$[4]$ Wan B. et al this conferenceSpeaker: Jinping Qian (Institute of plasma physics, Chinese academy of sciences) -
Introduction
Flexible and easy-to-use extrapolation tools are needed for extrapolating to, and planning for, the forthcoming JET DT campaign (DTE2) $[$1$]$. For this purpose and based on the needs of the scientific teams involved in scenario development, a streamlined, automated analysis workflow for the ASCOT heating and fast particle following code [2,3] has been developed. The input to ASCOT can include any combination of experimental data or modelled data, such as profiles, and are taken from the JETPEAK database [4,5]. As it does not resort to physics based transport modelling, this procedure is much faster and capable of handling much larger datasets than is possible with integrated modelling approches. This allows intershot analysis within ~10 minutes and DT extrapolation over larger datasets (~10$^2$ samples), presented in this paper, to be completed overnight.
Scaling D plasmas to DT target conditions
ASCOT has been coupled with the JETPEAK [4,5] database and analysis environment. This database currently includes some 8000 samples from stationary phases in all JET-ILW experiments and more than 1000 structured variables from diagnostics and a variety of modelling codes. From this database, subsets in baseline (q$_{95}$$\approx$3, 𝐼$_p$$\le$3.5 MA), hybrid (q$_{95}$$\approx$4-4.5, 𝐼$_p$$\le$2.7 MA) and AT (q$_{95}$$\approx$4, 𝐼$_p$$\le$2.8 MA) discharges have been selected for extrapolation to DT conditions at target power (40 MW for baseline and hybrid, 30 MW for AT) and target current (4 MA for baseline, ~2.5 MA for hybrid, ~2.8 MA for AT). The expected stored energy, ion temperature profiles, electron temperature and density profiles are empirically extrapolated on the basis of power law regressions of the large available datasets in D plasmas. An example is shown in fig.1. The regression is shown with coloured symbols for different classes of plasma current in MA. The black circles correspond to direct extrapolations from the regression results. For the best performing cases an alternative extrapolation is also shown with blue + signs. For these it was assumed that an increase to target parameters scales with the power laws from the regression as W$_{th}$=W$_{th0}$$\Pi$(X$_i$/X$_{i0}$)$^{\alpha i}$, where the X$_i$ are the regression variables, $\alpha_i$ is the exponent for variable $i$ and subscript 0 stands for the case to be extrapolated.The best performing discharges extrapolated to thermal stored energies are in the range of 12-14 MJ, whichever regression method is used. These plasma profiles and the NBI parameters for D and T are then used as inputs for ASCOT, together with similarly extrapolated plasma densities and temperature ratios (T$_i$/T$_e$). For ASCOT based extrapolations, a confinement enhancement consistent with the observed isotope scaling in hydrogen and deuterium plasmas, $\tau_E \propto$A$_{eff}^{0.4}$ where A$_{eff}$=$\sum$ A$_i$n$_i$/$\sum$ A$_i$n$_i$ is assumed [6].
DT extrapolation coupling ASCOT and JETPEAK
ASCOT has been used at JET for analysis actively since late 1990’s and during campaigns in 2016 (C36-37) it was extended with a complete fusion product module AFSI [7] and coupled with synthetic neutron diagnostics. The most comprehensive testing and validation against neutron production data was published in [4]. 𝐼$_p$$\le$2.5 Testing has been performed with the help of the Matlab interface for JETPEAK-ASCOT, in which ASCOT has been coupled to the JETPEAK database [4].
The latest step of development is a database-coupled interface and definition of new scalings for DT extrapolations are presented in this paper. The analysis has been carried out for sets of baseline, hybrid and AT plasma samples, with the highest DD neutron rates achieved so far. The target NBI powers for these simulations were 34 MW for baseline and hybrid and 30 MW for AT. For DT extrapolations, the input profiles of the samples are scaled consistently with the scalings for W$_{th}$, n$_e$ and T$_i$/T$_e$ in deuterium as described above, specifically for each scenario. The set-up of the NBI system was appropriately adjusted to correspond to D and T beams and the target plasma was assumed to be 50% D, 50% T.
Fig.2 shows the calculated DT neutron rates against the measured neutron rates in D plasmas. A DT neutron rate of 5.3$\times$10$^{18}$ n/s corresponds to 15 MW of DT fusion power. The initial results suggest that both the baseline and the hybrid scenarios may reach DT fusion powers near 15 MW, which is one of the objectives of the upcoming DT campaign. These results are consistent with, and complementary to, recent results from integrated modelling [8].
Ongoing work includes expanding the analysis with wider data sets and systematic comparison of the results with other extrapolation tools, such as TRANSP, as well as predictions for alpha particle effects.
References
$[$1$]$ E. Joffrin et al., Nuclear Fusion 2019 https://doi.org/10.1088/1741-4326/ab2276
$[$2$]$ J. A. Heikkinen et al. 2001 J. Comput. Phys. 173 527-48
[3] E. Hirvijoki et al. 2014 Comput. Phys. Commun. 185 1310-1321
[4] P. Sirén et al. 2019 JINST 14 C11013
[5] H. Weisen et al, 2020 Nucl.Fusion 60 036004
[6] C. F. Maggi et al. 2019Plasma Phys. Control. Fusion 60 014045
[7] P. Sirén et al. 2018 Nucl. Fusion 58 016023
[8] J. Garcia et al. 2019 Nucl. Fusion 59 086047This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
Speaker: Paula Siren -
The Divertor Tokamak Test facility (DTT) [1-3] is a D-shaped superconducting tokamak (R=2.14 m, a=0.65 m, BT≤6 T, Ip≤ 5.5 MA, pulse length ≤ 100 s, auxiliary heating ≤ 45 MW, W first wall and divertor), whose construction is starting in Frascati, Italy. Its main mission is to study the controlled exhaust of energy and particle from a fusion reactor, which is a top priority research item in the European Roadmap [4] towards thermonuclear fusion power production. This will be possible in DTT by achieving large PSEP/R values (where PSEP is the power flowing through the last closed magnetic surface) using 45 MW of auxiliary heating in a high performance machine characterised by high flexibility in the choice of the divertor and of the magnetic configurations. The characteristics of the machine will allow to address many ITER and DEMO relevant physics issues besides plasma wall interaction in a fusion relevant range of plasma parameters. The heating mix foresees the use of 170 GHz ECRH, 60-90 MHz ICRH and 400 keV negative ion beam injectors, with ECRH being the main system, although the precise sharing between the three systems has still to be optimised.
In order to help with the heating system definition, and to provide scenarios for the design of diagnostics and pellet injector, or for the evaluation of issues such as ripple losses or neutron shields, it is a key priority to achieve multi-channel integrated modelling of DTT scenarios based on state-of-art first principle quasi-linear transport models, whose reliability stems from an extensive validation work against experiments and high fidelity gyrokinetic simulations carried out within the EUROfusion and ITPA frameworks (see e.g. the recent overview [5] and references therein). It is also important that the integrated modelling results for some cases are validated against gyrokinetic simulations with the specific DTT parameters, to corroborate the validity of the reduced models in the particular case of DTT. In this paper, we summarise the first results of this activity, which extends the preliminary predictions reported in 1.
The integrated modelling of DTT has been carried out with the JINTRAC suite [6] and covers the region inside the separatrix, whilst the values of temperature and density at the separatrix are taken consistently with the scrape-off layer simulations described in 1. The pedestal has been determined with the EPED1 model [7] implemented in the Europed code [8], and core-edge coupling has been taken into account on an iterative basis. The pedestal density has been set to achieve a volume averaged density <ne>~ 0.43 nGW (Greenwald limit). The region in-side the top of the pedestal has been modelled using the QuaLiKiz [9] or the TGLF [10] turbulent transport models and NCLASS [11] for the neoclassical transport. The simulations pre-dict steady-state profiles of ion and electron temperature, density, rotation, current density, impurity (Ar, W) density, and calculate a self-consistent equilibrium starting from a fixed boundary taken from [12]. The heating has been modelled self-consistently using PENCIL[13] for NBI, PION[14] for ICRH and GRAY[15] for ECRH. SANCO [16] has been used to calculate impurity ionisation and recombination and radiation. The rotation has been predicted using a semi-empirical estimate of Prandtl and pinch numbers [17] due to numerical issues using the turbulent momentum transport from the quasi-linear models.
Fig.1 shows profiles obtained for the SN full power H-mode scenario with 32 MW ECRH, 15 MW NBI and 3 MW ICRH using QuaLiKiz for turbulent transport, which is mainly driven by ion-scale ITG/TEM. The strong central ECRH peaks Te far above Ti in the central part. Ions are rather stiff and Ti stays below Te also in most of the outer region, in spite of a large amount of thermal exchange power from electrons to ions. The ne profile is moderately peaked. A peaked rotation profile with central value of 50 krad/s does not provide a significant ExB stabilisation of the ion heat transport. Global plasma parameters are βN=1.6, τE=0.28s, total DD neutron rate~1.4 1017 s-1 (30% thermal). Total radiation is 15 MW. Both Ar and W show peaked profiles, with hints of W central accumulation, however a better treatment of W neoclassical transport using NEO [18] is in plan to check this prediction. In this simulation MHD has not been included, but some considerations on MHD stability will be discussed. Similar simulations for the half-power DAY1 heating configuration yield very similar profiles and double confinement time, which is also an indication of the high stiffness. Simulations with TGLF are being finalised with the latest release of the model [19], and differences between the predictions by the two models will be assessed against gyrokinetic simulations using GENE [20].
1 DTT interim design report (2019) https://www.dtt-project.enea.it/downloads/DTT_IDR_2019_WEB.pdf
2 Special Section of Fusion Engineering and Design Vol. 122, 2017, 253-294, e1-e25
[3] P.Martin et al., Proc. 46th EPS Conference on Plas-ma Physics, Milano, 2019, ECA Vol. 43C, O2.102
[4] A.J.H.Donné, Phil. Trans. R. Soc. A 377 20170432 (2019)
[5] P.Mantica et al., Plasma Physics Controlled Fusion, 62 014021 (2020)
[6] M.Romanelli et al., Plasma and Fusion Research, 9 3403023 (2014)
[7] P.B. Snyder et al., Phys. Plasmas 16 056118 (2009)
[8] S. Saarelma et al., Plasma Phys. Control. Fusion 60, 014042 (2018)
[9] J.Citrin et al., Plasma Phys. Control. Fusion 59, 124005 (2017)
[10] G.M.Staebler et al., Phys. Plasmas 23 062518 (2016)
[11] W A Houlberg et al., Phys. Plasmas 4 3230 (1997)
[12] R.Ambrosino et al.,Fusion Engineering and Design 146 1246 (2019)
[13] C.D. Challis et al., Nucl.Fusion 29 563 (1989)
[14] , L-G. Eriksson et al Nucl.Fusion 33 1037 (1993)
[15] D. Farina, Fusion Sci. Technol. 52, 154 (2007)
[16] L.Lauro Taroni et al., Proc. 21st EPS Conf. on Contr. Fus. and Plasma Phys. (Montpellier, 1994) 1, 102
[17] A.Peeters et al., Phys.Rev.Lett. 98, 265003 (2007)
[18] E. A. Belli and J.Candy, Plasma Phys. Control. Fusion, 50 95010 (2008)
[19] G.M.Staebler et al., to be presented at 47th EPS Conference on Plasma Physics, Sitges, 2020
[20] F.Jenko et al., Phys. Plasmas 7 1904 (2000)Speaker: Paola Mantica (Istituto per la Scienza e Tecnologia dei Plasmi, Consiglio Nazionale delle Ricerche (CNR), 20125 Milan, Italy) -
Due to edge opacity, future tokamaks will rely on injection of cryogenic pellets for plasma fuelling. Maintaining the desired density and isotope composition crucially depends on the transient plasma response to pellet injection, motivating dedicated studies for increased understanding and predictive capabilities. In a recent experiment at the Joint European Torus (JET), Deuterium pellets were injected in a pure Hydrogen plasma, allowing multi-isotope transport analysis. The size of the pellets,scaled to the plasma volume, lead to shallow deposition and transient inverted density profile, similarly to what is expected in the International Thermonuclear Experimental Reactor (ITER).
The desired isotope composition was reached and the isotope particle transport coefficients were determined by interpretative modelling [1], using the semi-empirical Bohm-Gyrobohm anomalous transport model,and matching the transient response of the D-D neutron rates. The rapid increase in the neutron rate during the pellet train, particularly following the initial injection, led to an inference of relatively large $ D_{D}/\chi_i $ values, and isotope mixing timescales faster than the energy confinement time. Such findings are consistent with previous experimental observations of fast isotope mixing, attributed to large ($D_i/D_e>1$, $|V_i|>|V_e|$) ion particle transport coefficients due to Ion Temperature Gradient (ITG) modes [2]. While in a pure plasma this phenomena is impossible to observe due to ambipolarity, in a multi-ion plasma the different ions can interchange at different timescales to the electron particle transport. This is most prevalent during transient, non-stationary states, such as during pellet injections due to the significant modifications of the local density gradients. Modifying the pellet isotope ratio compared to the background isotope ratio leads to rapid mixing of ions, significantly modifying the core isotope mix without affecting the time averaged electron profile.
This contribution applies the JINTRAC [3] integrated modelling framework, already proven to be effective in capturing the fast isotope mixing effect in experiments with NBI and gas-puff fuelling [4]. The use of the quasilinear gyrokinetic model QuaLiKiz [5, 6] for turbulent transport ensures a first-principle-based approach, taken here for the first time in the modelling of pellet cycles. It is however crucial to stress that the predictive capabilities of this approach are limited to the core, while the transport in the pedestal region had to be prescribed. Specifically, the particle and heat transport coefficients in the pedestal region were adjusted to match the interferometer measurements. Given this constraint,the modelled and experimental pre-pellet $ T_{i} $ , $ T_{e} $ and $ n_{e} $ are well reproduced, as well as the neutron rate evolution. The deuterium transport timescale following D pellet injection was found to be on theorder of the energy confinement time. In particular, the rapid evolution of the neutron rate after the first pellet was correctly reproduced in the model (Figure 1). This timescale depends on the turbulent regime and the agreement is a validation of the fast isotope mixing and of both QuaLiKiz and the pellet ablation model, HPI2 [7].
The identification of the correct turbulent regime by QuaLiKiz underlines the results. Depending on the radial position and on the phase of the pellet cycle, different modes are excited. TEM was found by QuaLiKiz to be the dominant instability after the pellets for $ \rho > 0.8 $, in conjunction with a very large negative density gradient. This causes a large particle flux directed outwards, in line with expectations from previous works [8]. Both ITG and ETG are stable between $ 0.5 < \rho < 0.8 $, where the density gradient is positive, for $ \sim 4 ms $ following each pellet injection. Immediately after, the cooling caused bythe adiabatic ablation of the pellets results in a locally steeper $ R/L_{T} $ gradient, balancing the stabilizing impact of negative $ R/L_n $ which occurs for ITG modes with kinetic electrons. This is key since the fast mixing of the deuterium depends on the ITG drive. To verify this important observation, QuaLiKiz was compared with the higher fidelity code GENE [9] using as an input the parameters encountered in the integrated modelling simulation. Good agreement was observed between the two codes. Since the fast isotope mixing depends crucially on ITG turbulence, additional simulations were carried outwith artificially reduced collisionality by a factor 10, consistent with ITER and reactor collisionality regimes, to gain insight on whether ITG is maintained at lower collisionality when Trapped Electron Modes (TEM) are further destabilized. In these simulations the turbulent regime changes, as expected for future reactors [10], to a mixed ITG-TEM regime and the density peaking increases. However, ITG at lower wave numbers is still destabilized by the pellet and the timescale for the deuterium penetration is almost unchanged. These results are promising with regard to reactor fuelling capability and burn control.
References
[1] M. Valovic et al. “Control of the hydrogen:deuterium isotope mixture using pellets in JET”. In Nucl. Fusion 59 (2019), p. 106047.
[2] M. Maslov et al. “Observation of enhanced ion particle transport in mixed H/D isotope plasmas on JET”. In Nucl. Fusion 58 (2018), p. 076022.
[3] Michele Romanelli et al. “JINTRAC: A system of codes for integrated simulation of Tokamak scenarios”. In Plasma and Fusion Research 9 (2014).
[4] M. Marin et al. “First-principles-based multiple-isotope particle transport modelling at JET”. In Nucl. Fusion 60 (2020), p. 046007.
[5] C. Bourdelle et al. “Core turbulent transport in tokamak plasmas: Bridging theory and experiment with QuaLiKiz”. In PPCF 58 (2016), p. 14036.
[6] J. Citrin et al. “Tractable flux-driven temperature, density, and rotation profile evolution with the quasilinear gyrokinetic transport model QuaLiKiz”. In PPCF 59 (2017), p. 124005.
[7] Florian ̈Kochl et al. “Modelling of Pellet Particle Ablation and Deposition : The Hydrogen Pellet Injection code HPI2”. In Preprint EFDA at JET PR(12)5712 (2012), p. 82.
[8] C. Angioni, P. T. Lang, and P. Manas. “Density gradient driven micro-instabilities and turbulence in ASDEX Upgrade pellet fuelled plasmas”. In Nucl. Fusion 57 (2017), p. 116053.
[9] F. Jenko, W. Dorland, M. Kotschenreuther and B. N. Rogers. “Electron Temperature Gradient Driven Turbulence”. In Turbulent Transport in Magnetized Plasmas 1904 February 2000, pp. 338–375.
[10] E. Fable et al. “The role of the source versus the collisionality in predicting a reactor density profile as observed on ASDEX Upgrade discharges”. In Nucl. Fusion 59 (2019), p. 076042.Speaker: Mr Michele Marin (DIFFER) -
The toroidal rotation without any external momentum sources known as an intrinsic rotation has been focused an important topic since the most promising toroidal rotation source driven externally from a neutral beam injection may not successful for the future burning devices like ITER and DEMO. The toroidal rotation in pure ohmic plasmas is self-generated and it is considered as one of the most fundamental types of intrinsic rotation for magnetic fusion researches 1. There have been reported a wide range of magnitudes, directions and abrupt reversals for intrinsic rotation studies no clear physical mechanisms are concluded to explain the complicated intrinsic rotation behaviors. Hence, investigation for the generation mechanism and characteristics for ohmic plasmas provide critical information to understand the origin of intrinsic rotation. In general, the faster intrinsic rotation is the more beneficial to suppress resistive wall modes and neoclassical tearing modes which easily lead major plasma disruptions.
All ohmic L-mode discharges more than 10 years experimental campaigns in KSTAR consistently show distinctive behaviors in the beginning of plasma current ramp-up phase that the core toroidal rotation always starts in the counter-current direction and the toroidal rotation shows dynamic changes at early plasma current ramp-up as shown in Fig. 1. The (-) sign of the toroidal rotation means a counter-current direction, which indicates in the opposite direction to the plasma current. Although the core toroidal rotation in the beginning of the plasma current ramp-up phase usually starts in the counter-current direction its magnitude and direction are dramatically changed with the electron density during plasma current flat-top phase as shown in Fig. 2. The core toroidal rotation bifurcations like rotation reversals are clearly observed to the counter to co-current and co to counter-current direction when the electron density is changed to the lower and higher density regimes, respectively. It is suggested that the physics mechanisms at early plasma current ramp-up and current flat-top phase are separately investigated since the main plasma parameters are quite different between the two phases.
The generation mechanism at early plasma current ramp-up for the counter-current rotation in ohmic discharges shown in Fig. 1 is speculated from the momentum transfer between neutrals and plasma particles. It is suggested that the momentum transfer from electrons to neutrals is larger than that of the ions to neutrals and the momentum exchange of neutrals easily go back to plasma by the plasma-neutral interactions. The calculated toroidal rotation with continuous dotted lines based on the momentum transfer mechanism is sensitive to the ratio of electron density to the main ion density (ne/nD+) and temporal behavior of toroidal rotation agrees well with the experimental measurement with red dots as shown in the bottom box of Fig. 3 (a). The core ohmic rotation during plasma current flat-top phase has been investigated from scaling studies for many ohmic L-mode discharges in KSTAR [1,2]. As a result, the core intrinsic toroidal rotation scaling for ohmic plasmas strongly depends on the ion temperature over plasma current regardless of the toroidal rotation direction. However, the toroidal rotation shown in Fig. 3 (b) keeps almost constant after passing the dynamic rotation change at early plasma current ramp-up stage even the electron density increases beyond the saturated ohmic confinement (SOC) regime. The calculated SOC density is marked with a blue asterisk in Fig. 3 (b). Unlike to the slower rotation less than 20 km/s with higher plasma current discharge shown in Fig. 2, the core toroidal rotation reversal is not observed under the fast rotating discharge with lower plasma current discharge. This experimental evidence elucidates that there is a certain threshold for the toroidal rotation reversal phenomena from ohmic L-mode discharges.
In this presentation, we will discuss the generation mechanism and detailed characteristics of intrinsic rotation including rotation reversal and its threshold, and extended rotation scaling results.
*Work supported by the Korea Ministry of Science and ICT under the KSTAR project contracts.
References:
1 Lee S. G. et al., 2018 Phys. Plasmas 25 044502.
2 Yoo J. W. et al., 2017 Phys. Plasmas 24 072510.
Speaker: Sang Gon Lee (National Fusion Research Institute) -
The helium plasmas have been demonstrated for the first time on EAST under the condition of pure RF-heating and ITER-like tungsten divertor, which advances physical understanding in support of the ITER non-nuclear operational phase $[1]$. Concentration of helium ($C_{He}$) in the plasma is confirmed to play a critical role in H-mode operation, as higher concentration raises the H-mode threshold power and deteriorates the energy confinement in H-mode. At lower $C_{He}$, EAST achieves the stationary Type-I ELMy H-mode over 80 energy confinement time with the energy confinement slightly above $H_{98,y2}$ scaling ($H_{98,y2} \approx 1.1$) by using pure RF power.
EAST first H-mode in helium plasma is achieved by optimizing plasma shape to improve LHW coupling and particle exhaust. Results suggest that a higher possibility of stationary Type-I ELMy H-mode comes with lower $C_{He}$, pedestal electron collisionality and edge safety factor $q_{95}$. Figure 1 shows an example of high performance H-mode discharge ($I_P$ = 0.5MA, $B_T$ = 2.4T, $q_{95}$ = 5.5) with characteristic of Type-I ELMs behavior ($f_{ELM}$ ~ 5-30Hz) from 3.5s to 8.0s. The plasma configuration is an upper single null (USN) with the strike points located on the vertical targets of tungsten divertor. With growing $C_{He}$ under identical experimental conditions, the ELM behavior evolves into higher frequency ~100Hz, or even ELM-absent. ELM mitigation and suppression are demonstrated by n=1 resonant magnetic perturbation (RMP) coils and boron powder injection method, which shows little deleterious effect on core plasma performance for both.
Experiments on the exploration of high energy confinement operation have shown that the global energy confinement time in He is about 30% lower than in D, similar to AUG results $[2]$. For both ion species, the energy confinement time steadily elevate with central line-averaged density over the range of 2-5$\times10^{19}m^{-3}$. In fixed plasma condition ($I_P$ = 0.5MA, $B_T$ = 2.4T, $n_e$ = 4.3$\times10^{19}m^{-3}$, $P_{inj}$ = 4MW), $C_{He}$ lowered by active D gas puffing gradually improves the H-mode performance.
Using the density scan, EAST achieves H-mode over wide density range of 2-6$\times10^{19}m^{-3}$ with the RF heating, where H-mode threshold power ($P_{thr}$) indicates strong dependence on density. Figure 2 presents the statistic result of normalized threshold power (radiated power subtracted), versus density, exhibiting a minimum of $P_{thr}$ at $n_{e,min} \approx 4\times10^{19}m^{-3}$. When plasma densities are below ne,min, the required power to enter H-mode in He is 1.2-2.2 times higher than the ITPA scaling for D $[3]$, otherwise it follows the scaling. The scatter of threshold power at one certain density is attributed to the variation of $C_{He}$. It is not possible to access H-mode at $C_{He}$ > 70% with the available heating power level at the time of experiments.
In summary, recent EAST experiments demonstrates high performance H-mode operation in helium plasma with RF-only heating. More analysis of experimental data from core and edge diagnostics is ongoing, to investigate physical mechanism of $C_{He}$ effect on H-mode operation. With the upgrades $[4]$ in 2020, EAST will be able to further extend high performance helium H-mode towards ITER-relevant by exploiting two extra ECH gyrotrons and the condition of full metal walls.
This work was supported by the Foundation of President of Hefei Institutes of Physical Science CAS under Grant No. YZJJ2018QN8.
$[1]$ ITER Physics Basis Editors 1999 Nucl. Fusion 39 2137
$[2]$ Ryter F. et al 2009 Nucl. Fusion 49 062003
$[3]$ Martin Y.R. et al 2008 Journal of Physics: Conference Series 123 012033
$[4]$ Wan B. et al this conferenceSpeaker: Dr Bin Zhang (Institute of Plasma Physics, Chinese Academy of Sciences) -
Development of high performance operation regimes in the magnetic confinement fusion devices has been of great interest in the fusion community for decades as it is critical to accomplish efficient steady-state operations of fusion reactor. Since the first discovery of high confinement mode (H-mode) in tokamaks that claimed enhancement of energy confinement by more than a factor of 2 compared to low confinement mode (L-mode), a number of advances have been made to further expand the high performance confinement regimes. Examples of such efforts have been addressed as very high confinement mode (VH-mode), improved H-mode, internal transport barrier (ITB) discharges, enhanced pedestal H-mode (EP H-mode), Super H-mode, and so on. These regimes are typically identified by enhanced transport barrier either or both at the core and edge of plasmas, leading to higher energy confinement than the standard H-mode [1]. Toroidal rotation and shear flow have been greatly acknowledged for their roles in accessing the improved confinement regimes [2].
We report a new improved confinement discharge achieved in the relatively slow rotating plasmas in KSTAR experiment, triggered by non-axisymmetric (3D) magnetic field. A series of experiments for a purpose of control of toroidal plasma rotation has been performed utilizing 3D magnetic field and electron cyclotron heating (ECH). Plasma parameters are in the range of $B_T$=1.6T-1.8T, $I_P$=500kA-700kA, and $q_{95}$=3.7-5.4. Neutral beam heating of 4MW was injected using 3 beam sources, which supplies strong toroidal torque to generate fast rotating plasmas in the H-mode confinement. Then, the 3D magnetic field of $n=1$ was applied to reduce the rotation and modify the rotation profile.
Time history of one of those discharges (#22705) is illustrated in Fig. 1. The 3D field coil current is turned on at 4.5s after H-mode transition and ramped up to 4kA/turn. The toroidal rotation is significantly reduced in the whole plasma volume by 3D magnetic field driven magnetic braking. Increases of ion temperature are observed along with the reduction of toroidal rotation. Particle transport is increased by the 3D field as shown in the density pump-out. Mitigation of ELMs was also observed. Surprisingly, total stored energy (real-time EFIT) is increased by up to 15% during the 3D field phase in spite of decreased density, which indicates improved energy confinement. Increase of neutron rate is consistently observed during the same period under the 3D field. It is notable that the toroidal flow is instantly damped by the 3D field in the whole volume (vertical shade), however build-up of the ion temperature and the stored energy is followed relatively slowly after the rotation reaches near-minimum level. This implies the rotation reduction by the 3D magnetic field triggers the improvement of the energy confinement.
The ECH of 2 x 0.6MW power launched at 6.5s under the 3D field further reduces the toroidal rotation and increases the particle transport while the improved energy confinement is maintained at a similar level. The ion temperature at the core drops after the ECH, and the electron temperature is raised to sustain the enhanced stored energy. This implies modification of the transport channel by the ECH. The stored energy returns to the earlier level without the 3D field soon after the 3D field coil is turned off even though the ECH is still on, which evidences the 3D field is responsible for the improved energy confinement. Finally, the reduced toroidal rotation is recovered to the similar level before the magnetic braking after the ECH is turned off. Such key transport features are similarly observed in a series of discharges produced with different $I_P$ and 3D field coil current.
FIDA measurement indicates fast ion confinement is improved as well during the improved confinement phase under the 3D magnetic field, as presented in the time trace of FIDA intensity at the core in Fig. 2. The FIDA intensity is increased and sustained during the 3D and 3D+ECH phases. Such behaviors are consistent with the evolution of stored energy. However, the FIDA signal significantly drops after turning-off of the 3D field in spite of the same heating power, proposing the primary role of the 3D field for the improved fast ion confinement. Fast ion profiles will be analyzed using TRANSP and FIDASIM codes with realistic kinetic profiles, leading to a clearer understanding on the fast ion transport associated with the 3D magnetic field and ECH.
Profiles of toroidal rotation and plasma temperature are presented in Fig. 3 for a set of discharges of improved confinement. It is obvious that significant reduction of the toroidal rotation is strongly correlated to the build-up of ion and electron temperature (red & blue). The improved confinement can be identified by enhanced transport barrier around the edge pedestal, where pedestal heights are raised by up to 50%. Fluctuation measurement by ECEI finds high frequency turbulences of ~ 200kHz near the pedestal are suppressed in the improved confinement phase, which is consistent with the kinetic profile measurement. Profile evolution in Fig. 3 also shows the ECH modifies transport channel to decrease and increase the ion and electron temperature, respectively. The height of the pedestal is lowered to the standard H-mode level in the ECH-only phase after the 3D field is turned-off (green).
The newly achieved improved confinement mode can be characterized by modification of multiple transport channels by the 3D magnetic field, as represented by global rotation braking (momentum transport), limited density pump-out (particle transport), and enhanced stored energy (energy transport). Measurements consistently indicate that the 3D magnetic field and modified toroidal rotation play a crucial role in those processes. Interestingly, mitigations of ELMs were also achieved, which further benefits the improved confinement mode for future devices. We will carry out the gyrokinetic analysis [3] to reveal the origin of improved energy confinement associated with the toroidal rotation reduction and transport modification by the ECH.References
[1] F. Wagner, Eur. Phys. J. H 43, 523 (2018)
[2] K.H. Burrell, Phys. Plasmas 4, 1499 (1997)
[3] J.M. Kwon et al, Comput. Phys. Commun. 177, 775 (2017)
Speaker: Dr Kimin Kim (National Fusion Research Institute) -
The high-performance operation is one of the major missions of HL-2M [1, 2] for supporting ITER, CFETR and future fusion reactors. Notice that, high-performance scenarios with the large plasma current (2.5-3MA) and the high elongation (1.8-2.0) are normally accompanied by the potential VDEs risk. This requires an efficient and reliable feedback control system, which is under construction. This paper focuses on a standard single-null plasma with the favorable vertical stability (elongation $\kappa$=1.5, minor radius a=0.64, triangularity $\delta$=0.43), and two kinds of expected high-performance D-D scenarios (including the convertional inductive regime of Ip=1.8MA and the hybrid regime of Ip=1.4MA) are investigated, based on the integrated modelling suite—CRONOS [8].
For the conventional inductive H-mode regime of $I_p$=1.8MA / $B_0$=2.2T with the line-averaged density $N_{bar}$ of 9.0×10$^{19}$m$^{-3}$ (Greenwald density fraction $f_G$=0.69), NBI of 15MW combining with ECW of 8MW are implemented. In such condition, the O-mode with the frequency of 105/140 GHz allows the EC wave propagates deeper than the X-mode in the plasma, obtaining the deposition peak around $\rho$=0.4-0.5. Meanwhile, the NBI deposition power on ions is around 4 times of that on electrons in the core, and the total NBI deposition power profile is flat within $\rho$=0.65. As shown in Figure 1, the non-inductive current fraction is 0.34 with the bootstrap current fraction of 0.29. Due to the very low off-axis additional current drive, the current profile gets peaked in the center with $\beta_p$=1.2. The thermal energy of the plasma reaches 2.0MJ with the high $\beta_N$ of 3.0 which is seemed to be compatible with li(3)=0.9 of the peaked current profile for avoiding the resistive wall mode (RWM) instability. Similar to ITER baseline, the q95 can reach 3.0. Both the ion and the electron temperature of the center can reach around 5keV.
For the hybrid regime of $I_p$=1.4MA / $B_0$=2.2T with $N_{bar}$=4.9×10$^{19}$m$^{-3}$, NBI of 8MW combine with the equatorial ECW (X-mode + 105GHz) of 6MW and upper ECW (X-mode + 140GHz) of 2MW are implemented. In this case, power the deposition peak of NBI is on-axis, while the ECW deposition peak is off-axis. Comparing to the conventional inductive regime, both the bootstrap current fraction and the additional drive current fraction increase. The total non-inductive drive current fraction reaches 0.6. As shown in Figure 2, the substantial increased off-axis drive current allows the magnetic shear to get flat around the center. The center safety factor q0 increases to 1.2 and the minimum safety factor qmin increases to 0.94 with the location of $\rho$=0.35. Such weak reversed shear leads to an internal transport barrier (ITB) generated around the center, allowing the energy confinement to increase. The H98(y,2) reaches 1.3. Similar to the ITER hybrid scenario, q95 in this regime reaches 3.9. The ion and the electron temperature of the center can reach 6.4keV and 8.5keV, respectively. The thermal energy of the plasma reaches 1.4MJ with the high $\beta_N$ of 3.1. Further, the RWM instability of this regime will be thoroughly analyzed by MARS code.
It is worth to notice that the high-performance operation is normally accompanied by the potential heat load issue of the divertor. The power flowing to the SOL region of the conventional inductive regime and the hybrid regime reaches 17MW and 14MW, respectively. It may lead to the great challenge for the outer divertor baffle in the standard configuration. More completely integrated analysis of CRONOS + FEEQS + SOLPS is performing by coupling the boundary condition of the LCFS (shown as Figure 3). The effect, of the impurity injection from the divertor to mitigate the heat load, will be iteratively analyzed with the core operation. Further, high-performance scenarios compatible with the snowflake-minus divertor configuration will be explored.
Acknowledgement
This work is performed in the framework of SIFFER, supported by the CNNC Elite Project under grant nos. 2019JZYF-01 and the National MCF Energy R&D Program under grant nos. 2018YFE0301101References
[1] Duan X. R. et al 2017 1st Asia-Pacific Conference on Plasma Physic (Chengdu, China, 18-23 Sep. 2017) P29 [2] Li Q. et al 2015 Fusion Engineering and Design 96-97 338 [3] Zheng G.Y. et al 2016 Nucl.Fusion 56 126013 [4] Cao J.Y. et al 2013 Fusion Engineering and Design 88 872 [5] Huang M. et al, 17th Joint Workshop on Electron Cyclotron Emission and Electron Cyclotron Resonance Heating, EPJ Web of Conferences 147, 04006 (2017). [6] X.Y. Bai et al 2017 Proceedings 22nd Topical Conference on RF Power in Plasma [7] XUE L. et al 2020 Nucl. Fusion 60 016016 [8] Artaud J.F. et al 2010 Nucl. Fusion 50 043001Speaker: Prof. Lei Xue (Southwestern Institute of Physics) -
The Integrated Modelling & Analysis Suite (IMAS) is the software infrastructure that is being developed building upon the modelling expertise from across the research facilities within the ITER Members to support the execution of the ITER Research Plan 1. It is built around a standardised representation of data described by a Data Dictionary that is both machine independent and extensible. Machine independence is important since it allows tools and workflows developed in IMAS to be tested and refined on existing devices, whilst extensibility allows the Data Dictionary to grow and evolve over time as more Use Cases are addressed. The use of standardised Interface Data Structures (IDSs) fosters the creation of modular physics components and (sub-)workflows that can be flexibly re-used to address different needs.
One of the focal points driving development within IMAS is the creation of a high-fidelity plasma simulator that can be used to predict ITER plasma performance. The DINA code [2,3] has been extensively used to validate the capability of the ITER poloidal field system to support the plasma scenarios foreseen in the ITER Research Plan. It includes a free-boundary equilibrium evolution solver implementing feedback control of the plasma current, position and shape, taking into account eddy currents in the vacuum vessel, as well as numerous engineering limits imposed on the coils, their power supplies, and plasma-wall gaps. The JINTRAC code [4] refines the physics description in the plasma core and also couples its behaviour with that in the plasma edge. It can describe plasma heating, fuelling and transient behaviour.
Both the DINA and JINTRAC codes have been adapted to use the IMAS Data Model [5], and are now being further modularised to allow exchanging IDSs with additional external physics modules to enable the incorporation of other higher-fidelity physics models.
One such high-fidelity physics workflow that has been developed over the last year by a combination of ITER Staff, ITER internships, ITER Scientist Fellows and voluntary contributions, is a comprehensive heating and current drive (HCD) workflow that is capable of describing all of the ITER heating systems as well as synergistic effects between them [6]. This workflow exemplifies the IMAS integrated modelling paradigm and has driven further refinements in the IMAS infrastructure. The workflow builds upon the extensive work carried out within the EUROfusion Work Package for Code Development (WPCD). Whilst IMAS is independent of any particular choice of workflow engine, this HCD workflow has been implemented in Python to facilitate distributed development and portability.
The IDS-based database of ITER scenario simulations is continuously expanding and is used to support ITER design activities including assessments of the ITER heating systems and diagnostics. Recent additions include the set of SOLPS4.3 simulations of ITER edge conditions and JINTRAC simulations of ITER L [7] and H-mode [8] conditions.
In preparation for the Live Display of information during ITER operations, work has started on the creation of displays using the scenario information contained within the ITER scenario database together with synthetic diagnostics. Figure 1 shows a still frame of an evolving Live Display derived from data calculated with the METIS [9] and SOLPS-ITER [10] physics codes.
Development is also underway on a workflow to assess the energetic particle stability of IDS-based plasma scenarios on computational timescales that enable relatively extensive studies to be performed and for it to be embedded within other workflows. The workflow is based upon the LIGKA [11] and HAGIS [12] physics codes and is being validated against other results and predictions for ASDEX Upgrade and JT-60SA [13].
In preparation for ITER operations, work has started on the development of experimental data processing and analysis pipelines. These are essential since in contrast to existing machines upon which they can be tested, on ITER they will form the only way to process raw experimental data for subsequent processing, analysis and interpretation. A key element of this activity to allow testing is dynamic access to existing experimental data in the form of IDSs. This has been enabled by the creation of local plug-ins that handle the reading and mapping of experimental data into IDSs such as have been created by EUROfusion WPCD for the EU MST1 and JET devices, and Korea for KSTAR [14]. The above dynamically mapped data has been used as input in the form of IDSs to the European Transport Simulator (ETS) for modelling JET and KSTAR scenarios [15]. The (typically static) Machine Description metadata, which is also included in IDSs and enables the creation of device-generic workflows, must also be locally curated. At ITER, a separate Machine Description database has been set up and integrated into the Universal Data Access (UDA) data server architecture.The views and opinions expressed herein do not necessarily reflect those of the ITER Organization.
[1] ITER Research Plan within the Staged Approach: https://www.iter.org/technical-reports
[2] R. R. Khayrutdinov, V. E. Lukash, J. Comput. Phys. 109 (1993) 193
[3] V.E. Lukash et al., Plasma Devices and Operations 15 (2007) 283
[4] M. Romanelli et al., Plasma Fusion Res. 9 (2014) 3403023
[5] S.D. Pinches et al., TH/P6-7, IAEA FEC 2018, Ahmedabad, India
[6] M. Schneider, This conference, IAEA FEC 2020, Nice, France
[7] E. Militello Asp et al., TH/P2-23, IAEA FEC 2016, Kyoto, Japan
[8] L.Garzotti et al., Nucl. Fusion 59 (2019)
[9] J.F. Artaud et al., Nucl. Fusion 58 (2018) 105001
[10] S. Wiesen, D. Reiter, V. Kotov, M. Baelmans, et al., J. Nucl. Mater. 463 (2015) 480
[11] Ph. Lauber et al., J. Comp. Phys., 226/1 (2007)
[12] S.D. Pinches et al., Comp. Phys. Comm., 111 (1998)
[13] Ph. Lauber et al., This conference, IAEA FEC 2020, Nice, France
[14] M. Romanelli et al, Submitted to Fusion Science and Technology
[15] M. Romanelli et al., This conference, IAEA FEC 2020, Nice, FranceSpeaker: Simon Pinches (ITER Organization) -
Disruption prediction and avoidance is a high-priority challenge for tokamaks to sustain long pulse and high performance plasmas that are critical for ITER and next-step devices for fusion generation. Disruption-free, continuous operation of high performance plasmas over long pulse is a main goal of modern superconducting tokamak devices such as the Korea Superconducting Tokamak Advanced Research (KSTAR) facility. Stability analyses including ideal global magnetohydrodynamic (MHD) and resistive MHD instabilities have been developed to provide input to plasma disruption forecasting analysis. Existing research has shown that analyses spanning only limited time periods have proved difficult to create reliable time evolution of stability analysis with continuous and clear predictions using experimental data. Equilibrium reconstruction noise and apparent inconsistencies with experimental diagnostic data or modeling yielding poor convergence need to be resolved for successful model validation. High fidelity kinetic equilibrium reconstructions are an essential requirement for accurate determination of the plasma stability and disruption prediction analyses to support the goal of continuous, disruption-free operation.
In this work, equilibria with the required accuracy and fidelity are produced as input for a suite of stability analysis, including ideal MHD stability (codes PEST and DCON) (1,2), resistive MHD (resistive DCON) (3), and kinetic MHD (MISK) (4), which can then be used to determine correlations with experimental plasma stability. Compilations of such analyses can produce multivariate models for marginal stability to modes that can either directly or indirectly cause plasma disruption (5). Such analyses comprise an important part of the disruption event characterization and forecasting (DECAF) framework presently expanding in capability to characterize and forecast plasma disruptions in tokamaks (6).
Unlike analysis that uses only external magnetic measurements (magnetics-only), kinetic equilibrium reconstructions provide measurement constraints internal to the plasma including the plasma temperature and density profiles. The present kinetic equilibrium reconstructions include Thomson scattering (TS) data, charge exchange spectroscopy (CES) data, and allowance for fast particle pressure in addition to external magnetics and shaping field current data, and inclusion of vacuum vessel and passive plate currents following an approach used successfully in NSTX. In addition, up to 25 channels of motional Stark effect data are used to constrain the local magnetic field pitch angle to produce reliable evaluation of the q profile.
The present low-convergence error level (from $10^{-10}$ to $10^{-13}$ maximum Grad-Shafranov error) of the equilibria can provide clear and continuous trends in DCON and resistive DCON shown in Fig. 1, except when $\it{q_{min}}\geq $ 2 (no 2/1 surface precludes the mode). The analysis of KSTAR kinetic equilibria shows unstable ideal stability (above the no-wall limit) during the high $\beta_N$ period as shown in (c) and (b). The corresponding experimental plasmas do not show unstable global RWM instabilities indicating that kinetic effects need be added to the analysis to accurately determine the plasma stability. Although the ideal no-wall limit stability analysis is not by itself sufficient to predict disruption, the clear, full time evolution still provides important plasma stability information and can be used to produce reduced stability forecasting models in DECAF analysis for the reduced kinetic stability model in the code. The resistive DCON calculation (d) indicates an unstable tearing mode in the whole plasma evolution, but there were not any strong tearing modes observed in the experiment. The magnetic spectrogram (a) shows evidence of n = 2 mode activity when $\it{q_{min}}$ is below but close to $\it{q}$ = 2.
Kinetic equilibria with MSE have been improved to sufficiently high accuracy and validation compared to experiment by 1) optimizing the basis functions for the plasma current profile, 2) compensating magnetics from 3D fields produced by resonant magnetic perturbation coils, and 3) evaluating MSE error with systematic and statistical components. Moreover, the q profile is validated by electron cyclotron emission (ECE) radiometry and ECE imaging, as shown in Fig. 2. The ECE radiometers in (a) show 1D radial electron temperature $\it{T_e}$ profiles, which indicate the possible $\it{q}$ = 2 surface location at the inboard midplane. Meanwhile the 2D ECE imaging (ECEI) in (b) and (c) detects the local electron temperature fluctuations and provides a magnetic island position, evaluating the outboard $\it{q}$ = 2 surface at the midplane. The $\it{q}$ profile from kinetic equilibrium reconstructions with MSE (Fig. 2 (d), KSTAR 21520) compares well to the rational surface positions from the ECE and ECEI results.
*Supported by US DOE grant DE-SC0016614
(1) Grimm R. C. et al 1976 Methods in Comput. Phys. 16 253
(2) Glasser A. H. and Chance M. S. 1997 Bull. Am. Phys. Spc. 42 1848
(3) Glasser A. H. et al 2016 Phys. Plasmas. 23 112506
(4) Hu B. et al 2005 Phys. Plasmas 12 057301
(5) J.W. Berkery, S.A. Sabbagh, R.E. Bell, et al. 2017 Phys. Plasmas 24 056103
(6) S.A. Sabbagh, J.W. Berkery, Y.S. Park, et al. to be submitted 2020Speaker: Yanzheng Jiang (Columbia University) -
Control of impurity species in fusion plasmas is one of the main issues for long and stable plasma operation. The impurities can be generated by interaction of the plasma with its facing materials and/or by intentional injection for the purpose of maintaining radiative mantle. These impurities can cause detrimental radiation cooling and fuel dilution as they become accumulated inside the plasma. In this regard, impurity transport study should necessarily be conducted and the first step would be obtaining their local concentration distribution. The research framework that we have been developing consists of spectroscopic impurity diagnostics, tomographic reconstruction of the chord-integrated atomic emissions, and forward modelling using the KAIST Impurity Modelling (KIM) code. The main impurity diagnostics in this study are Compact Advanced EUV Spectrometer (CAES) [Ref. 1] in KSTAR and WEST, and the grazing incidence spectrometer [Ref. 2] in WEST. Based on the diagnostic measurements, a new tomography algorithm developed for limited viewing angle [Ref. 3] and forward modelling using the KIM code were utilized to deduce the local distributions of emission intensity and density of impurity species.
Previous research revealed that the centrifugal force due to the rapid toroidal plasma rotation brings about significant poloidally asymmetric distribution and also modifies neoclassical transport of heavy impurities. Since the toroidal rotation speed of KSTAR plasmas reaches several hundreds of km/s, low-Z impurity species such as carbon and oxygen ions also showed poloidally asymmetric distribution as shown in Figure 1. The top row images in the figure are space-resolved EUV spectra measured by CAES, and the bottom ones are the tomographically-reconstructed emission intensity profiles of the 3.37 nm C VI line (yellow), 3.80 nm O VIII line (green), and 4.03 nm C V line (blue) on the midplane by using our new reconstruction code for the limited viewing angle [Ref. 3]. The core rotation speeds in the figure are (a) 138 km/s, (b) 291 km/s, and (c) 318 km/s, respectively, and the effective Mach numbers of the impurities are larger than the unity when the toroidal speed is higher than 320 km/s. For example, the effective Mach numbers of C VI and O VIII at 320 km/s are 1.0 and 1.2, respectively. During the early phase of the plasma, which does not have a sufficiently high rotation speed, the carbon and oxygen line emission profiles show centrally-peaked and poloidally symmetric features as seen in figure 1(a). On the other hand, radially-hollow and poloidally asymmetric distributions are clearly seen in Figure 1 (b) and (c) as the plasma rotation becomes faster. This implies that the poloidal asymmetry effect can be non-negligible for the rapidly rotating plasma with the impurity Mach number exceeding the unity. From the view point of impurity transport, this is important because flux-averaged transport analysis without taking this effect into account may mislead the real phenomena.
In the WEST tokamak, where tungsten (W) is employed as a divertor material, analyses were performed to obtain the 2D distribution of W ions. Since there are a large number of W transition lines known as quasi-continuum radiation in the EUV range, the reconstruction method to find their local distributions is inappropriate due to the convoluted line emissions. However, the forward model using the KIM code enables deducing the local density profiles by comparing the synthetic data and the measured data by chi-square fitting. Figure 2 shows the comparison of the three W lines (4.86 nm W$^{28}$+, 6.09 nm W$^{44}$+, and 6.23 nm W$^{45}$+) measured by the grazing incidence spectrometer. Assuming that the plasma is in steady-state, the emission spectra at different time points can be treated as space-resolved spectra. The simulated W emission profiles are also analyzed for different impurity transport coefficients (diffusion coefficient and convection velocity) cases. The result suggests that W$^{28}$+ line emission seems a good candidate for studying impurity transport because it is more sensitive to the convection compared to the emissions from highly charged ions (W$^{44}$+ and W$^{45}$+) that are localized mostly in the core region. Thus, the convection characteristics of the W ions can be investigated through W$^{28}$+ lines under various plasma situations such as applying actuators to reduce the core accumulation of W. Although the sensitivity calibration is needed to determine the absolute value of the density, the calculated fractional abundance and Zeff profile can provide valuable information how W ions are distributed inside the plasma. The analysis result suggests that a considerable amount of tungsten exists in the core.
In summary, we observed that even the light impurity species under the fast toroidal plasma rotation show significant poloidal asymmetric distribution in KSTAR and tungsten impurity shows core accumulation even under the slowly rotating plasma in WEST. This work emphasizes the importance of the analysis of the local impurity density profiles in parallel to the development of impurity control methods. This study is expected to contribute to the impurity transport study in ITER with the ITER VUV spectrometer [Ref. 4] that also has a capability similar to CAES to resolve spatial distribution of the impurity line emissions.![The measured space-resolved EUV spectra (top) and the reconstructed profiles of their line emissions on the midplane (bottom) during KSTAR shot #16611: C VI – 3.37 nm (yellow), O VIII (second order) – 3.80 nm (green), and C V – 4.03 nm (blue). Although the density distribution of each impurity species may change with time as seen at each time point (a) 1.25 s, (b) 2.25 s, and (c) 3.25 s, the increase of poloidal asymmetry is still observed as the plasma rotates faster ((b) and (c)) [Ref. 3]](https://gdpl.kaist.ac.kr/files/attach/images/2433167/412/433/002/4a41c6ebc44efc01d8e097558e9e5b0e.jpg)
Acknowledgements
This work was supported by the National R&D Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science and ICT (NRF-2019M1A7A1A03087560), and also partly by the ITER Project Contract (NRF-2018M1A7A2019060).References:
(1) I. Song et al., Rev. Sci. Instrum 88, 093509 (2017).
(2) J. Schwob et al., Rev. Sci. Instrum 58, 1601 (1987).
(3) I. Song et al., Nucl. Fusion 60, 036013 (2020).
(4) C.R. Seon et al., Eur. Phys. J. D 71, 313 (2017).Speaker: Haewon Shin (Korea Advanced Institute of Science and Technology (KAIST)) -
Kinetic plasma control based on extremely simple data-driven models and a two-time-scale approximation has been developed and validated on non-linear plasma simulations in recent years. Both in these models and in the associated control algorithms, the fast component (kinetic time scale) of the plasma dynamics is considered as a singular perturbation of a quasi-static magnetic and thermal equilibrium, which is governed by the flux diffusion equation (resistive time scale). Combined with classical optimal control theory, the effectiveness of this approach to simultaneously control the plasma poloidal flux profile, ψ(x), and the normalized pressure parameter, βN, in non-inductive, high-βN discharges was demonstrated experimentally on the DIII-D tokamak [1]. Real-time control of the full safety factor profile, q(x), and of the normalized plasma pressure is far more difficult. An advanced model-predictive control (MPC [2]) algorithm, also based on the two-time-scale approximation, was therefore developed in this aim. It was validated in closed loop nonlinear plasma transport simulations, which showed excellent performance [3]. Here, we report on the first experiments using this new kinetic control algorithm in its simplest version to track time-dependent targets for the central safety factor, q0, and for the poloidal pressure parameter, βp (Fig. 1a-1b-2a-2b). The experiments were performed in a steady state H-mode operation scenario on the EAST tokamak.
It was already shown earlier on several tokamaks that the combination of extremely simple slow and fast plasma response models, identified in a given operation scenario either from experimental or simulated data [4], can satisfactorily approximate the coupled response of the plasma parameter profiles to relatively large random variations of the heating and current drive actuators. A dedicated system identification procedure has been developed and improved in recent years [3-4]. Here, the same procedure was applied to EAST experimental data in a typical H-mode scenario with electron cyclotron resonance heating (ECRH) and lower hybrid current drive (LHCD). When these experiments were performed, the only actuator available with enough real-time dynamics was the LHCD system at 4.6 GHz, with a coupled power between 1 MW and 2.5 MW. Additional LHCD power (0.5 MW) was injected at 2.45 GHz during the plasma current ramp-up, and 0.9 MW of ECRH power was injected during the 350 kA current flattop, from 2 gyrotrons at 140 GHz. The system identification data was obtained from two discharges, with chirping frequency and pseudo-random binary sequence modulations of the LHCD power, respectively. A linear state space model with 9 eigenmodes was found to reproduce satisfactorily the coupled evolution of the poloidal flux profile, ψ(x, t), of the inverse of the safety factor profile, i(x, t) = 1/q(x, t), and of the slow and fast components (βp,slow and βp,fast, respectively) of βp, with βp(t) = βp,slow(t) + βp,fast(t).
The full paper will describe details of the model identification and of the MPC control algorithm including an observer estimation of the model mismatch, and discuss the results of the first control experiments performed on EAST with this algorithm. In the first discharge, the target q0 was set at 2.4 from t = 2.7 s to t = 4.5 s and was raised to q0 = 2.8 at t = 4.52 s (the control cycle time was 20 ms). The evolution of q0 and the LHCD command are shown on Fig. 1a and Fig. 1b, respectively. The q0 targets are reached in about 1 s. To cope with the nonlinear response of the LHCD actuator to the command, a proportional plus integral actuator control module was added in cascade with the MPC module. The effectively coupled LHCD power is also shown on Fig 1b (blue trace). In the second discharge, a piecewise linear βp target waveform with 1.6 < βp < 1.9 was tracked. The evolution of βp was perfectly under control, as shown on Fig. 2a. The LHCD command and coupled power are shown on Fig. 2b.
Simultaneous control of the full q(x) profile and of βp (or any other kinetic variable or profile) was already achieved in non-linear simulations [3] and can now readily be tested experimentally with the algorithm implemented in the EAST plasma control system. For this task, more actuators will be used, such as the 2.45 GHz LHCD and the co-current and counter-current neutral beam injection systems, in addition to the 4.6 GHz LHCD system.
Acknowledgements: This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
[1] D. Moreau, et al., Nucl. Fusion 53 (2013) 063020.
[2] J. M. Maciejowski, “Predictive Control with Constraints”, Pearson Education Ltd (2002).
[3] D. Moreau, et al., Proc. 27th IAEA Fusion Energy Conf. (2018), Paper EX/P2-26.
[4] D. Moreau, et al., Nucl. Fusion 55 (2015) 063011.Speaker: Didier Moreau (CEA/DRF/IRFM) -
Introduction
Compass-Upgrade tokamak, which is being constructed in the Institute of Plasma Physics of the Czech Academy of Sciences in Prague, will replace existing Compass tokamak $[1]$. It will be a compact, medium-size ($R=0,89 m, a=0,3 m$), high-magnetic-field ($5 T$) device. The tokamak is expected to operate with plasma densities up to $n_e=10^{21} m^{-3}$. Plasma heating will be performed by $4 \times 1MW$ Neutral Beam Injectors (NBI) with $80 keV$ injection energy. Injectors are grouped by 2 vertically spaced pairs. In the future, heating system will be extended by at least 4 MW Electron Cyclotron Resonant Heating (ECRH) system. At the present design stage, much effort has been devoted to scenario development. NBI system, which is expected to be the main source of plasma heating, demands precise modeling. The high plasma density of Compass-U scenarios is limiting NBI core heating efficiency. To compensate for this effect, the injection tangency radius ($R_t$) can be reduced. On the other hand, more perpendicular injection leads to an increase of the trapped fast ion fraction, which depending on the fast ion deposition, and last closed flux surface (LCFS) position, can cause fast ion orbit losses. Another significant channel of losses can be charge-exchange losses. This research is dedicated to find an optimal geometry of the NBI system using set simulations with the NUBEAM code $[2]$.
Simulation setup
Compass-Upgrade scenarios were developed with the METIS code $[3]$. Magnetic equilibrium geometry was reconstructed by the FIESTA code $[4]$ using METIS output data. Using plasma profiles and magnetic equilibrium from METIS and FIESTA codes, static targets for NUBEAM NBI simulations were created. Three baseline H-mode discharge scenarios ($4MW$ of NBI and $2MW$ of ECRH power) with different toroidal magnetic field strength ($2.5T$,$4.3T$,$5T$) and plasma currents are considered. Each of scenarios was prepared with METIS for various plasma line averaged densities ($1.1\cdot10^{20}-3.2\cdot10^{20}m^{-3}$).
Plasma core heating efficiency
The variation of vertical alignment of NBI injectors in Compass-Upgrade is limited by injectors and port construction. However, the NBI tangency radius can be adjusted up to $65cm$. $R_t$ can be optimized for better plasma core heating performances. Fig.1 shows NBI power fraction that was delivered to both, plasma ions and electrons inside closed flux surface with an index $\rho<0.1$ as a function of $R_t$ and line averaged density for the scenario with $B_t=5T$. In this scenario, high plasma density can be reached. Simulation results show that $80keV$ beam heat deposition is located on the plasma periphery for $R_t=65cm$ when electron densities is higher than $2\cdot10^{20}m^{-3}$. The decrease of $R_t$ improves heat deposition to the plasma core. However very small $R_t$ leads to high trapped particle fraction and reduced number of fast ion trajectories that cross the plasma core. In simulations with densities $n_e>2\cdot 10^{19} m^{-3}$, in order to achieve efficient plasma core heating, injection tangency radius has to be set to around $40cm$.\section*{NBI power losses}
In a case of small electron density, some of the injected beam neutrals can shine through the plasma without ionization. With a low divergence beam of Compass-Upgrade and high NBI power, beam shine-through can cause damage to the first wall. Shine-through losses calculated for scenario with $2.5T$, as it is characterised by a lower densities fig.2(a). With $R_t<40cm$ injection is dangerous in a case of small electron density ($n_e<9\cdot10^{19}$). The first wall shine-through power deposition distribution was calculated using code NUR - Monte-Carlo code (developed by the authors). The code NUR can calculate fast ion deposition and wall shine-through loads using NUBEAM-like beam geometry and fast neutrals stopping rates presented by S. Suzuki $[5]$. Code NUR became a part of a particle following code EBdyna [6]. One can see on the figure Fig.2(b) example of shine-through power distribution on the central column created by a pair of vertically separated injectors. This result was obtained for density of $1.1\cdot10^{-20}m^{-3}$ with $R_t=40cm$. Shine-though, in this case, is equal to $2.4\%$ of NBI power.
In the case of perpendicular injection, most of the fast ions (more than $80\%$) are trapped on banana orbits. However, even in this case, fast ions are well confined and orbit losses are very small for simulated baseline scenarios ($<1\%$). High fraction of trapped ions, in this case, can lead to large ripple losses (not included into the simulation). It was found that orbit losses increases when the plasma is closer to the first outer wall. To investigate this, simulations were repeated using different magnetic equilibrium with the increased plasma minor radius (from $26.8cm$ to $28cm$). As the result, the orbit losses reaching up to $11\%$ of total NBI power in a case of perpendicular injection $R_t=0cm$, and relatively high plasma density $2\cdot 10^{20}m^{-3}$.
In the presence of background neutrals, charge-exchange losses can decrease NBI heating efficiency. Charge-exchange losses increase with fast ion slowing down time, which is usually higher in a case of low electron density. The neutral density profile for the NUBEAM simulations was obtained by METIS code. The result of simulations show that charge-exchange losses reaching values up to $10\%$ of injected power in a case of low-density plasmas ($n_e<8\cdot 10^{19}m^{-3}$). These results can describe general charge-exchange losses behavior, but for quantitative modeling, better simulation model of background neutrals has to be used.Conclusions
NUBEAM code was adopted to work with METIS and FIESTA codes output data, and predictive modelling of NBI heating on Compass-Upgrade tokamak has been performed. The simulations has shown good confinement of NBI fast ions in most cases. However, during injection with $65cm$ tangency radius in high-density plasmas the core heating is not efficient. The reduction of the injection tangency radius improves plasma core heating efficiency. But on the other hand, it increases the fraction of trapped fast ions. In the case of a smaller magnetic field (scenario with $B_t=2.5T$) when LCFS is close to the plasma wall it leads to the orbit losses that can reach $11\%$ from injected NBI power. High trapped ion fraction could also cause strong ripple losses. Based on our results we can conclude that in the first stages of the tokamak operation, when scenarios with a high magnetic field and densities will not be performed, the optimal NBI injection radius is equal to $65cm$. Setting $R_t$ this way will allow to use NBI for smaller plasma densities without possible wall damage by shine-trough neutrals. On the later operation phase, in order to improve core heating efficiency, NBI has to be performed at $R_t=40cm$. Injections at $R_t<30$ are becoming ineffective at the low-density plasmas due to shine-trough and charge-exchange losses.
References
$[1]$ Panek, R. et al.Fusion Eng. Des.123, 11, DOI: https://doi.org/10.1016/j.fusengdes.2017.03.002 (2017)
$[2]$ Pankin, A. et al.Comput. Phys. Commun.159 Issue 3, 157, DOI: https://doi.org/10.1016/j.cpc.2003.11.002 (2004)
$[3]$ Artaud, J. et al.Nucl. Fusion58, 105001, DOI: https://doi.org/10.1016/j.fusengdes.2017.03.002 (2018)
$[4]$ Cunningham, G. et al.Fusion Eng. Des.88 Issue 12, 3238, DOI: https://doi.org/10.1016/j.fusengdes.2013.10.001 (2013)
$[5]$ Suzuki, S. et al.Plasma Phys. Control. Fusion40, 2097, DOI: https://doi.org/10.1088/0741-3335/40/12/009 (1998)
$[6]$ Jaulmes, F. et al.Nucl. Fusion54, 104013, DOI: https://doi.org/10.1088/0029-5515/54/10/104013 (2014)Speaker: Dr Georgiy Zadvitskiy (Institute of Plasma Physics of the Czech Academy of Sciences, Prague, Czech Republic) -
Successful operation of ITER will require robust regulation of the plasma temperature and density despite the plasma's nonlinear dynamics and various uncertainties. In this work, a burn-control algorithm was designed to determine control efforts that will drive the plasma to desired targets. In order to effectively achieve the commanded control efforts, control allocation modules were developed to optimally manage ITER's heating and fueling actuators. Using Lyapunov techniques, the nonlinear controller was synthesized from a zero-dimensional two-temperature model. For burning plasmas in ITER, the accuracy of the widely used scaling laws for the energy confinement time is unsure because they are based on experiments without alpha particle heating. Therefore, the proposed controller was designed with an adaptive estimation scheme to handle uncertainties in the plasma confinement and other complex phenomena (i.e. the fuel recycling and impurity sputtering that results from plasma-wall interactions). The plasma density is regulated by two control laws that request injection rates of deuterium and tritium. In ITER-like burning plasmas, the majority of the alpha particle heating will be deposited into the plasma electrons (ions receive less than $25\%$ of it). Consequently, the electron and ion temperatures will be uncoupled. The proposed controller independently regulates the ion and electron temperatures with separate control laws for the external ion and electron heating. Because the radiation losses depend on the electron temperature and the fusion reactivity depends on the ion temperature, independent temperature regulation permits plasma scenarios with improved performance.
ITER will have access to an ion cyclotron (IC) heating system ($20$ MW), an electron cyclotron (EC) heating system ($20$ MW) and two neutral beam (NB) injectors ($16.5$ MW each). With two commanded heating efforts (for the ions and electrons) and four heating actuators, an optimal control allocation (CA) module is necessary to map the commanded heating efforts to the heating actuators. The map can be written as $v=Bu$ where the vector $v$ contains the controller's commanded efforts, the vector $u$ contains the actuator allocation (power supplied by each actuator), and the control effectiveness matrix $B$ includes efficiency factors ($0-100\%$) for each actuator and the fractions of neutral beam heating that is delivered to the plasma ions and electrons. The allocator was designed with modularity such that it can be readily swapped out for another allocator. At each simulation time step, the proposed CA module solves an optimization problem that considers the saturation and rate limits of the heating actuators. The optimization problem was formulated as a strictly convex quadratic program (QP) that always permits a unique optimal solution. Since ITER will have only two fueling pellet injectors to satisfy the proposed controller's deuterium and tritium injection rate commands, CA is not required for density control.
For comparison to the QP CA module, a simpler pseudo-inverse CA module is introduced: $u=B^+ v$. To demonstrate performance of the adaptive control and CA algorithms, three simulations with different control schemes were generated. To assess the adaptive burn controller, an adaptive controller with QP CA is compared to a non-adaptive controller with QP CA. To assess the QP CA module, the adaptive controller with QP CA is also compared to an adaptive controller with pseudo-inverse CA. Each simulation uses the same plasma parameters, initial conditions, initial estimates of uncertain parameters, desired targets and actuator efficiencies. At $90$s, the desired target was changed. At $180$s, an actuator fault scenario was simulated by setting the actuator efficiency of the second neutral beam injector to zero. Fig. $1$ (a, b, c, d) shows how well the three control schemes can track the desired targets for plasma density, ion energy ($E_i$), electron energy ($E_e$), and tritium fraction (density of tritium over density of both deuterium and tritium) using the actuation shown in Fig. $2$. The adaptive controller with QP CA robustly stabilizes the equilibria even when the target changes and the second neutral beam injector shuts down. The non-adaptive controller with QP CA fails to track the target due to the modeled uncertainties. Unlike the adaptive controller with QP CA, the adaptive controller with pseudo-inverse CA cannot hold the plasma at the desired energy targets when the actuator fault occurs. By plotting the allocation errors in Fig. $1$ (e, f), the QP and pseudo-inverse CA modules can be better compared. The allocation error is the absolute value of the difference between the controller's commanded efforts and the control efforts produced from the allocated actuators. Allocation errors occur because either the commanded efforts are unattainable due to actuator constraints or the CA module fails to find the solution of $v=Bu$. The QP CA module can find a solution with zero allocation error throughout the simulation. In both the first-third and last-third of the simulation, the pseudo-inverse CA module fails to generate the commanded efforts by a few megawatts. The pseudo-inverse CA module's failures generate steady-state errors in the plasma energies (at $270$s $\Delta E_i=-1.12\times10^{5} \text{J}$ and $\Delta E_e=-1.42\times10^{5} \text{J}$).
An adaptive CA module with uncertainty in the allocation mapping ($B$) was also designed for the proposed adaptive burn controller. Unlike the previous QP CA module, this CA module considers the controller's commanded fueling efforts so that uncertainties in the tritium fraction of the fueling pellets can be handled. Furthermore, the actuator efficiency factors and the neutral beam heating fractions for the ions and electrons were considered uncertain. This adaptive CA module takes into account actuator dynamics, which also contain uncertainties, such as actuation lag times resulting from the thermalization delay of injected neutral beam particles, the ablation time for fueling pellets, ect. A Lyapunov function was used to formulate a dynamic update-law for $u$ (more computationally efficient than solving an optimization problem online).
This work has been supported in part by the U.S. Department of Energy under DE-SC0010661.
Speaker: Vincent Graber (Lehigh University) -
Helium (He) operation has recently been successfully performed on EAST equipped with an upper ITER-like, water-cooled, tungsten (W) monoblock divertor. The main plasma-wall interaction issues in He plasmas have been studied and compared with those in deuterium (D) plasmas, such as divertor detachment, W erosion, material migration, ELM characteristics and control, etc. Studying the impact of He plasma operation on W plasma-facing components (PFC) and the general characteristics of He tokamak discharges with a W divertor is a high priority for development of the ITER Research Plan, since He plasmas are currently foreseen in the ITER non-active operation phases due to the generally observed lower H-mode power threshold in He compared to hydrogen. Furthermore, He is naturally present during D-T operations.
The He discharges obtained in EAST comprise an extensive set of pure RF-heated, H-mode plasmas with different types of ELMs. Higher power He discharges enhance deuterium (D) removal efficiency and lead to smooth changeover from D to He. The global recycling coefficient of He measured by particle balance increases with heating power, and is higher than in D. The He pumping speed is relatively low compared to D, and can be slightly improved by moving the strike point closer to the cryopump. It is found that the energy confinement and pedestal characteristics are strongly dependent on the He purity in the plasma, measured by an edge visible spectrometer observing HeII and D$_{β}$ lines simultaneously. High performance Type-I ELMy H-modes can only be achieved for He concentrations <60%. Both n=1 resonant magnetic perturbations (RMP) and boron (B) powder injection have been used to suppress ELMs effectively in He plasmas. With different RMP spectra, ELM suppression can be achieved at a similar n=1 threshold current. A flow rate threshold of boron power injection has also been found for ELM suppression with negligible impact on the core plasma parameters.
Divertor detachment in He plasma has been studied using density ramps. As shown in figure 1, the detachment threshold density is significantly increased with increasing heating power and is higher than for equivalent D discharges at the same power to the scrape-off layer (P$_{SOL}$). Low levels of neon (Ne) impurity seeding can help to achieve divertor detachment at a lower density. Active feedback control of radiation power by Ne seeding helps to approach divertor detachment without degradation of core plasma performance. The divertor heat flux width in He is similar with that in D for similar plasma conditions. The divertor heat load has been controlled by RMPs along with the Ne seeding, with W erosion reduced due both to ELM suppression and lower T$_e$ as a result of Ne seeding.
During He H-mode discharges, two peaks of W gross erosion rate during each ELM burst were often observed by a high time resolution WI spectroscopy indicating erosion caused by main ions and impurities respectively. The erosion processes can be quantitively reproduced by using the Free-Streaming model (FSM) $[1]$ taking into account different species in the plasma. As shown in figure 2 for the He discharge with a He concentration about 40%, the first W erosion peak is mainly caused by the He$^{2+}$ ions which is consistent with the peak of the ion saturation current measured by divertor probes. The second W erosion peak is due to the streaming C$^{6+}$ ions, for which the time delay can be as much as 0.5 ms for the He plasma discharge, and is caused by the discrepancy of the thermal speed between C$^{6+}$ and main ions. It is found that the ELM-averaged W erosion rate by He plasma increases nearly linearly with the heating power, similar to D plasmas. The intra-ELM W sputtering source also shows a strong positive correlation with the ELM frequency. The inter-ELM W erosion rate in He is about 3 times that in D with similar divertor conditions due to the higher W sputtering yield of He ions.
Three dedicated He experiments with sample exposures have been successfully carried out using the outboard midplane manipulator MAPES. Two sets of He pre-exposed W samples have been used to study material erosion and fuzz formation features. To understand material migration in magnetically shadowed regions at the first wall panels in ITER $[2]$, a proxy tile with two different material coatings, carbon and aluminum, on different sides of the plasma wetted area, have been exposed in L-mode He plasma, thus avoiding the effects of chemical erosion. Net material deposition in the shadowed regions is not found. A known quantity of $^{13}$CD$_4$ was injected and traced using graphite samples in both He and D plasmas to study material migration in the main chamber wall and the effect of chemical erosion on C redeposited layers. A redeposited layer is formed close to the injection point after the exposure and it can be seen that the direction of material migration is dominated by the E×B drift. Post-mortem nuclear reaction analysis finds the unexpected result that the $^{13}$C deposition fraction in D is about 50% higher than that in He, indicating a higher physical sputtering rate by He than the chemical erosion rate by D.
$[1]$ W, Fundamenski et al., Plasma Phys. Control. Fusion 48 (2005) 109
$[2]$ R. Ding et al., Nucl. Fusion 55 (2015) 023013Speaker: Dr Rui Ding (Institute of Plasma Physics, Chinese Academy of Sciences) -
Preparation of the ITER experimental campaigns will require use of verified and validated models for the prediction of plasma response to actuators and of fusion performances. Several simulation codes have been developed by the fusion community for modelling of plasma equilibrium, transport processes, MHD stability, heating and current drive and fusion reactions. The integration of these codes into flexible, verified and validated workflows 1 has been the subject of the work carried out by the EUROfusion consortium which has culminated in the release and exploitation of the European Transport Simulator (ETS) 2. EUROfusion is currently utilising five different Tokamaks JET, TCV, AUG, MAST-U and WEST to carry out its research plan. The analysis and modelling with ETS of data from the above set of different devices required a high degree of standardization and the development of a common data platform. The choice of EUROfusion is to fully embrace the ITER Integrated Modelling and Analysis Suite (IMAS) [3] based on the Interface Data Structure (IDS) for data standardization and code integration. In this paper we present the result of a coordinated activity started in 2019 to develop tools, such as UDA (Universal Data Access) plugins and the OMFIT [4] module IMASgo, for the mapping of experimental data in IMAS / IDS of all the EUROfusion Tokamaks along with a campaign for the validation of codes and models in ETS. One of the main results of the above study using JET and Medium Size Tokamaks (MST) data has been the assessment of the residual turbulent transport in the ETB (external transport barrier) region of an H-mode plasma which has been found to be linked to the level of turbulence at the top of the pedestal and the development of a simplified transport model applicable to the plasma edge. In this paper we will also report on the results of the study with ETS of plasma scenarios in multi-tokamaks exploring different operational regimes: simulation of DIII-D scenarios with variable NBI injection angles have been used to validate the beam deposition codes in ETS against measurements of torque, ion temperature and neutron yield. The impact of both rotation and heating profiles on confinement has been found to be large on the flux calculated with first principle models such as QualiKiz and TGLF. Long pulse, steady-state plasmas in KSTAR have been simulated to validate the equilibrium and current diffusion modules in ETS along with the impurity source and transport modules in non-transient conditions. The slow deterioration of confinement observed in KSTAR during long pulses has been addressed in the modelling by investigating the drift of the plasma equilibrium, by scanning ECRH power / resonance to control impurity accumulation and by controlling impurity influxes. The high-beta fully non-inductive advanced scenario of JT-60SA has been simulated with the CDBM transport model including self-consistent calculation of the NBI and ECRH power deposition.
The use of IMAS / IDS for the integration of modules (actors) in ETS makes it easy to test within this framework any physics code or module that has been adapted to IMAS. The adoption in ETS of the CDBM model developed in Japan demonstrates this concept. Preparation for an effective JET DT campaign requires extrapolation to DT of scenarios in deuterium in order to predict plasma confinement and fusion performance. An extensive validation of the heating and current drive / transport modules in ETS on JET deuterium discharges with both interpretative and predictive simulations has been carried out including the statistical benchmark with TRANSP on more than hundred JET discharges.
Finally, ETS has been used to simulate the ITER 15 MA, 5.3 T baseline scenario. The results presented in this paper indicate that IMAS is indeed an effective tool for facilitating the analysis of data and the exchange of physics modules. The use of IMAS allowed us to validate the models in ETS in various plasma conditions building confidence in the predictions for ITER scenarios. The extensive use of IMAS in the fusion community will in the longer term provide a database of fusion data that can be exploited for theory studies, model validation, advanced Machine Learning and Artificial Intelligence applications in support of ITER exploitation.
1 A Becoulet et al (2007), Computer physics communications 177 (1-2), 55-59
2 P. Strand et al (2018), 27th IAEA FEC, TH/P6-14, IAEA CN-258
[3] F. Imbeaux et al (2015), Nuclear Fusion, Volume 55, Number 12
[4] O. Meneghini et al (2015), Nuclear Fusion, Volume 55, Number 8Speaker: Prof. Michele Romanelli (Tokamak Energy) -
The practical and economic viability of tokamak fusion reactors depends, in a significant way, on the efficiency of radio frequency (RF) waves to deliver energy and momentum to the plasma in the core of the reactor. The RF electromagnetic waves, excited by antenna structures placed near the wall of a tokamak, have to propagate through the turbulent edge plasma along their path to the core of the fusion device. In present day experiments, the radial width of the edge region and scrape-off layer is of the order of a few centimeters. In ITER, and in future fusion reactors, this width will be of the order of tens of centimeters. Any modifications to the properties of RF waves in the edge region can have deleterious effects on the efficiency and spatial distribution of power coupled to the plasma. This could affect, for example, the stabilization of the neoclassical tearing mode in ITER by electron cyclotron RF waves. Thus, it is imperative to properly understand the propagation of RF waves through plasma turbulence. This paper is on a multi-pronged, theoretical and computational, approach that is being pursued to quantify the effect of edge plasma turbulence on the propagation of RF waves. The theoretical and analytical models are based on solutions of the Faraday-Ampere equation in a magnetized plasma (1-3) and on the Kirchhoff tangent plane approximation (4). An effective medium approach has been developed so as to approximate the permittivity of a turbulent plasma by analytical expressions (5). The computations are being carried out with a newly developed code ScaRF that solves Maxwell's equations in three-dimensions for an arbitrarily assigned plasma permittivity (6).
THEORETICAL MODELING OF SCATTERING OF RF WAVES
A full-wave theoretical model, using Maxwell's equations, has been developed for scattering of RF plane waves by blobs and filaments embedded in a magnetized plasma (1-3). The spatial structure of the electric and magnetic fields is given by the Faraday-Ampere equation, and mathematical techniques developed for Mie scattering are used to solve the wave equations for a cold plasma permittivity. We have formulated a complimentary analytical model based on the Kirchhoff's tangent plane approximation [4]. The Kirchhoff method has commonly been used for studying scattering of electromagnetic waves from rough surfaces. The basis of the Kirchhoff approximation is built upon the theory of physical optics. In this approximation, the scattered field due to planar turbulence is determined by the wave fields on the surface of the turbulence separating two different plasma densities. Each point on this surface is assumed to be part of an infinitely extended plane that is along the local tangent at that point. Consequently, the Kirchhoff approximation, along with physical optics, leads to the study of reflection, refraction, and side-scattering of RF waves.This not only simplifies the modeling of RF scattering but, more importantly, gives physical insight into several, experimentally significant, aspects of scattering. We find that the spectrum of the RF fields is modified along, and across, the direction of the magnetic field. We have derived analytical expressions for the changes in the wave spectrum based on the spatial variation of the density fluctuations. The theory also predicts linear coupling of RF waves due to turbulence. For example, in the electron cyclotron range of frequencies, an incident ordinary wave can couple some of its power to the extraordinary wave in the presence of fluctuations. The theoretical results are being validated by full-wave simulations of RF scattering by different forms of plasma turbulence.
COMPUTATIONAL STUDIES ON SCATTERING OF RF WAVES
A full-wave electromagnetic computational code ScaRF, based on the finite difference frequency domain (FDFD) method, has been developed to study the effects of density turbulence on the propagation of RF waves (6). The anisotropic plasma permittivity used in the code is that for a magnetized, cold plasma. The code was initially used to study the propagation of an RF plane wave through a spatially modulated, periodic, density interface. Such an interface arises in the edge region due to magnetohydrodynamic instability and/or drift waves. Figure 1 shows the effect of a sinusoidal density perturbation on the propagation of a plane wave (incident from the top of the figure). Besides the transmitted wave through the density perturbation, there is a reflected wave as well. The Fourier spectrum of the transmitted RF electric field, Fig. 2, shows a broadening of the spectrum as the wave propagates away from the density fluctuation. The results from ScaRF are being used to validate the analytical results obtained from our theoretical models and for determining the limitations of the Kirchhoff approximation. While ScaRF has been used to study a periodic density fluctuation, the code is general enough to include different varieties of density fluctuations in the edge region - such as blobs and filaments, and spatially random fluctuations.
MODELING THE PERMITTIVITY OF PLASMAS IN THE SCRAPE-OFF LAYER
The theoretical and computational studies on scattering, using either the full-wave approach or the physical optics model, require a proper description of the plasma permittivity in the turbulent edge region. The tenuous plasma in this region is a mix of fluctuations with a broad range of spatial correlations. Electromagnetic homogenization, or ``effective medium approximation'', is the process of estimating the effective electromagnetic properties of composite materials. We have generalized this approximation, following the Maxwell-Garnett approach to describe the plasma permittivity in the scrape-off layer (5). Our formulation is suitably adapted for magnetized plasmas and for RF waves that are commonly used for heating and current drive.The advantage of this approach is that we can describe the turbulent plasma as an effective homogeneous medium for which the permittivity is suitably approximated. This eliminates the need to have a detailed description of the edge plasma - an experimentally formidable task. The effective permittivity is being implemented in a computational code and will be validated against specific models of the plasma permittivity.
Acknowledgements
AKR is supported by the US Department of Energy Grant numbers DE-FG02-91ER-54109 and DE-FC02-01ER54648. All other authors are supported by EUROfusion and the National Program for Thermonuclear Fusion.
References
(1) A. K. Ram and K. Hizanidis, Phys. Plasmas 23, 022504 (2016).
(2) Z. Ioannidis, A. K. Ram, K. Hizanidis, and I. G. Tigelis, Phys. Plasmas 24, 102115 (2017).
(3) S.-I. Valvis et al., J. Plasma Phys. 84, 745840604 (2018).
(4) P. Beckmann and A. Spizzichino, ``The scattering of electromagnetic waves from rough surfaces,'' Artech House, Massachusetts (1987).
(5) F. Bairaktaris et al., J. Plasma Phys. 84, 45840501 (2018).
(6) A. D. Papadopoulos et al., J. Plasma Phys. 85, 905850309 (2019).
Speaker: Dr Abhay Ram (Massachusetts Institute of Technology) -
The quasioptical ray tracing code named PARADE (PAraxial RAy DEscription) [a,b], which can contribute to the modeling of propagation and absorption of wave beams in both tokamak and stellarator, is newly developed based on the eXtended Geometrical Optics (XGO) theory [c]. Transmitted power profile of Electron Cyclotron Waves (ECWs) experimentally observed in Large Helical Devise (LHD) [d,e] is numerically analyzed, as one of the applications of PARADE. Qualitative consistency between numerical and experimental results shown here is one of the validations of PARADE applied on realistic magnetic confinement system, and thus, suggests the usefulness of quasioptical codes for the predictions of Electron Cyclotron Resonance Heating (ECRH) in future reactors.
ECRH is one of the essential components to achieve the nuclear fusion. Due to capabilities of the high-power density and localized heating, ECRH is expected also to play a role of the high-functional plasma control system, for example, to suppress the Neoclassical Tearing Mode. However, since ECW beam can refract, and importantly, its profile can vary along the propagation in fusion plasmas, it is important to quasioptically evaluate the propagations and absorptions of ECW from both numerical and experimental sides.
PARADE can simulate the quasioptical propagation of wave beams in inhomogeneous anisotropic media within a reasonable computational resource. The code is based on the Schrödinger-type partial differential equation, that accounts for refraction, diffraction, non-uniform dissipation across the beam, and uniquely, mode-conversion. This Schrödinger-type equation is solved along the Reference Ray (RR) trajectory given by the ray equation [a,c]. One of the advantage of PARADE is a capability of treating mode-conversion[b,c], and another one is a capability to simulate an arbitrary beam profiles with non-uniform dissipation, whereas most quasioptical codes assume the Gaussian beam profile and uniform dissipation.
An example of the PARADE results, which calculates the O-mode beam propagation with partial absorption in plasma, is shown in Fig. 1. Wave beam is absorbed non-uniformly across the beam near the cyclotron resonance. The original intensity profile is distorted, and due to the diffraction, strongly diffused propagation is obtained. Also, the beam center, which is defined as a first-moment of the beam profile and should be considered as true-ray trajectory, is diverged from the RR trajectory, which has been considered as ray trajectory in conventional approaches. Such propagation will affect the ECRH power deposition/driven current profile and the beam profile of the passed through power that can cause deteriorating effect on the first wall or sensitive diagnostics. These situations can occur also in tokamaks when the beam crosses the resonance layer in shallow angle.
![PARADE simulation result of the O-mode wave beam with frequency $f = 77.0$ GHz, propagating along $x$ axis and partially passing through the cyclotron resonance (red-line) in the hot electron plasma with constant density $n_{e} = 1.0 \times 10^{19}$ m$^{-3}$, constant temperature $T_{e} = 2.0$ keV, and magnetic field $(B_x,B_y,B_z) = B_0 (\cos \theta, 0.0, \sin \theta)$ where $B_0 = 7.3 \exp[-(x/1.5 - 1.0)^2-(y/5.0 - 1.0)^2]$ and $\theta = 80.0^\circ$. Blue-line shows the beam width along the RR (black-arrow), orange-line shows same simulation result in non-absorbed case as a reference, and black-dashed-line shows beam center trajectory which should be considered as true-ray trajectory. The wave beam is injected initially as Gaussian beam, but due to the inhomogeneous resonance structure where the wave power is absorbed non-uniformly across the beam, distorted and strongly diffracted to non-Gaussian beam. Also, beam center position is diverged from the RR.](https://workshop.nifs.ac.jp/fec2020/image/32-Yanagihara-image-fig_1.png)
As one of the applications of the PARADE code, the experimental results from the transmission power profile measurement are compared with the calculated one. A target plate facing on the ECRH antenna via plasma was set inside the LHD. By measuring the temperature increase of the target plate, transmitted intensity profile of ECW beam, which is not absorbed by the LHD plasma in a single path, is obtained. This direct measurement system gives the experimental evaluation of refraction, diffraction, and partial absorption of ECW caused by plasmas. This first quantitative comparison come to be realized due to the development of the PARADE code, because conventional multi-ray methods used in the past cannot simulate a beam profile accurately, due to an ignoring of the diffraction.
Figure 2 shows intensity profiles of ECW on the target plate, experimentally obtained by the direct measurement system and numerically simulated by PARADE. By increasing the electron density of LHD plasma, the beam positions of both numerical and experimental results are gradually shifted to clock-wise- and negative R-directions in same scale. This qualitative agreement is a good verification that the PARADE certainly traces quasioptical propagations of ECW in LHD.
![Intensity profiles of ECW on the target plate facing on the ECRH antenna via plasma in LHD, that are experimentally obtained by the direct measurement system and numerically simulated by PARADE. The shift of beam profile position due to the density increase observed experimentally are captured by PARADE. Experimental figures are reproduced from Ref. [d].](https://workshop.nifs.ac.jp/fec2020/image/32-Yanagihara-image-fig_2.png)
As another application of PARADE, ECRH power deposition profile in a typical LHD experiment is simulated. Figure 3(a) shows the cross-section of $154$~GHz 2nd harmonics X mode ECW beam injected into the LHD, and Fig. 3(b) shows ECRH power deposition profile for same simulation with (a), where orange-line is the PARADE result, and red-line is the result from the conventional multi-ray tracing, which cannot account for the diffraction. The centers of power density of both results are very similar, but its profile of PARADE is smoother than that by multi-rays. This is because of the reproducibility of the envelope structure. The envelope with continuous and smooth amplitude profile by PARADE, which contains diffraction, gives more realistic power deposition profile than conventional multi-rays method.
Newly developed quasioptical ray tracing code ``PARADE'' can describe wave beams with inhomogeneous absorption across the beam, that should be treated carefully not only in LHD-type stellarator but in tokamak as well. Quasioptical propagations and absorptions of ECW in LHD, numerically performed by PARADE are in qualitative agreement with experimental results in LHD and show the superiority over the conventional multi-ray tracing, and note that, PARADE can apply to any plasma configuration such as tokamak plasmas. These confirm that PARADE is useful to predict the realistic ECRH experiments, and powerful tool for the detailed design of future tokamak and stellarator where ECRH is expected as the main heating.
[a] K. Yanagihara et al., Phys. Plasmas 26 (2019), 072111.
[b] K. Yanagihara et al., Phys. Plasmas 26 (2019), 072112.
[c] I. Y. Dodin et al., Phys. Plasmas 26 (2019), 072110.
[d] S. Kamio et al., Rev. Sci. Instrum. 85 (2014), 11E822.
[e] H. Takahashi et al., EPJ Web Conf. 87 (2015), 02019.Speaker: Kota Yanagihara -
The electron cyclotron heating (ECH) is one of the intense methods of non-inductive plasma current drive (CD). The ECH waves accelerated the electrons with the Doppler-shifted electron cyclotron resonance (ECR) interactions, and effectively ramped and sustained the plasma current non-inductively to achieve long discharge duration. The plasma sustainment with ECH waves is a key issue for the steady state operation in the tokamak configuration. The plasma was sustained for more than 2 hours by the ECH/CD non-inductive method in the Q-shu University experiment with steady-state spherical tokamak (QUEST) 1. The plasma current generation as well as sustainment of the plasma is strongly related to the ECR conditions and confinements of the electrons in the magnetic field. In this study, the relativistic electrons’ orbit trajectories were calculated and analyzed under equilibrium magnetic configuration of spherical tokamak in the ECH/CD to evaluate the ECR conditions and confinements of the electrons in the magnetic field. A framework with several sets of calculation codes accordingly for various theoretical models was developed with the EFIT (Equilibrium Fitting) plasma equilibrium to calculate and analyses orbit trajectories.
This study observed that a large number of passing resonant electrons with initial positive $v_\parallel$ starting from the high field side were maintained their orbits outside the last closed flux surface (LCFS), while all the passing resonant electrons with initial negative $v_\parallel$ starting from the high field side were maintained their orbits inside the LCFS. The trapped resonant electrons only being the 2nd harmonic ECR were maintained banana orbits at the low field side of the torus where most of the portions of the banana orbits of initial positive $v_\parallel$ were placed outside the LCFS.
In this study, the equilibrium magnetic configuration was obtained from the plasma equilibrium solution using 129x129 (129 grid points in the R direction and 129 grid points in the $Z$ direction) spatial resolutions EFIT code. The poloidal flux $\psi$ and toroidal current density of plasma $J_T$ were calculated by the EFIT code on the rectangular grid with the external magnetics for constraint of a discharge that satisfy the model provided by the Grad-Shafranov equation 2.
The orbit trajectories of the resonant electrons were calculated and analyzed for the 8.2 GHz electron cyclotron heating and current drive waves on the equilibrium magnetic configuration of the Q-shu University experiment with steady-state spherical tokamak (QUEST) in the relativistic Doppler-shifted ECR. The down-shifted and up-shifted fundamental and second (2nd) harmonic relativistic resonances were considered separately. Various parallel and perpendicular velocities to the magnetic field, $v_\parallel$ and $v_\perp$ were considered from the Doppler-shifted ECR condition for the electrons to be resonant with the ECH waves. The parallel refractive index $N_\parallel$ from -1 to +1 with step 0.1 was taken into account in the multiple-wall reflection model. The maximum energy of the resonant electrons was restricted at 100 keV and calculated the orbits of those electrons whose initial energies were between 1 to 100 keV. The orbit trajectories of the ECR electron were obtained as contour plot of the resonant electron’s energy under the equilibrium magnetic configuration on the poloidal cross-section. The energy was expressed in terms of magnetic moment and toroidal angular momentum. The energy, magnetic moment, and toroidal angular momentum were conserved in the orbit trajectories. All the resonant and confined electrons’ orbits trajectories were obtained for various positions of the coordinates $(R, Z)$, resonant pitch angles and parallel refractive index. The number of the step parameters was more than 5,200000.
The red and blue contours in figure 1 show the calculated orbit trajectories on the poloidal cross-section between the $R$ positions from 0.35 m to 0.53 m and $Z$=0.00 m of the fundamental resonant electrons. The black and green contours show the closed and opened magnetic surfaces, respectively. The numbers of actually calculated orbit trajectories were much more than the orbits shown in the figure. The left and right figures show the orbit trajectories of the electrons with initial + $v_\parallel$ and - $v_\parallel$, respectively. Figure 2 shows the banana orbits of the trapped electrons on the poloidal cross-section of the 2nd harmonic resonant electrons. The number of actually calculated orbits was much larger than the orbits shown in the figure. The left figure shows the banana orbits of the trapped electrons with initial + $v_\parallel$. The right figure shows the banana orbits of the trapped electrons with initial - $v_\parallel$.
Depending on the ECR for various conditions, the resonant electrons were travelled both in the parallel and the antiparallel directions to the magnetic field of the tokamak. The resonant electrons with initial positive $v_\parallel$ were travelled in the parallel direction to the magnetic field, while the resonant electrons with initial negative $v_\parallel$ were travelled in the antiparallel direction to the magnetic field. These two types of resonant electrons with initial positive and negative $v_\parallel$ contribute current in opposite directions, respectively. A significant number of confined electrons with initial positive $v_\parallel$ were maintained portions of the orbits in the open magnetic surfaces that may cause to shift the plasma outward along the equatorial plane. In the summary, several codes required to calculate and evaluate the ECR as well as orbits have been developed with the EFIT code. Various criteria of the relativistic ECR conditions were properly taken into consideration for the orbit analysis. Additionally, the developed code could be applicable to analyze the $\alpha$-particle (alpha) orbits of the burning plasmas.
- Hanada K., et al 2017 Nucl. Fusion 57 126061
- Havlicek, J., et al 2007 WDS’07 Proceedings of Contributed Papers Part II— Physics of Plasmas and Ionized Media WDS’07 234-239
Speaker: Dr Md Mahbub Alam (ICS, Bangladesh Atomic Energy Commission, Dhaka, Bangladesh) -
Plasma waves naturally occur in various forms in magnetically confined plasmas. With a broad categorization into cold waves, hot (kinetic) waves, and coupled cold-hot waves, the plasma waves play integral functions in fusion plasma physics, such as particle confinement, fluid instabilities, and radiative processes. Some plasma waves, in particular low-frequency electrostatic (ES) waves, affect the diffusive transport of particles 1. Alfven waves and ion cyclotron harmonic waves interact with energetic particles, altering the energy distribution of the particles [2, 3]. The emission of electron cyclotron waves is an important channel of radiative loss of energy.
The amplitude and dispersion relation of each plasma wave depend on the local plasma parameters of the region where the wave is excited. The best example is the electron cyclotron emission (ECE); its frequency is proportional to the local magnetic field strength, and the intensity is proportional to the local electron temperature. The ECE radiometry is one of the most reliable diagnostics for the measurement of the electron temperature profile. On the KSTAR tokamak, an advanced imaging radiometry system has extended the capability of the conventional 1-D profile radiometry to 2D and quasi-3D imaging of magnetohydrodynamic (MHD) fluid modes and turbulent eddies [4].
Ion cyclotron emission (ICE) is another example that carries information about the plasma. On the KSTAR, we have developed a versatile radio frequency (RF) spectrometer system consisting of broadband antennas (about 0.1~2 GHz), 8-channel filter-banks, and high-speed digitizers [5]. Using the RF spectrometer, we recently showed that harmonic ICEs occur in the edge of an H-mode plasma [5], and the ICE spectrum depends on the distribution of energetic ions and the local electron density [6]. By fully integrating the RF spectrometer with the ECE imaging system, we found that the spectral change of ICEs has a high correlation with the state of the macroscopic fluid modes in the edge region [7], implying a mutual interaction between ICEs and fluid modes.
The ECE and ICE examples suggest that other plasma waves also have high potential as diagnostic signals. The existing RF diagnostics implemented in several tokamaks and linear machines use various forms of antennas, depending on the wave frequencies of interest. The use of antennas is a relatively easy and low-cost method of resolving the frequency spectrum of the RF emissions. However, the frequency spectrum measured by an antenna provides no information on the wavenumber, which makes it difficult (although not impossible) to determine the type of wave. Furthermore, the ES waves or the waves that attenuate too much before escaping the plasma boundary are beyond the reach of antennas.
We propose a new RF diagnostic system to enable local measurement of plasma waves and determination of wavenumber as well as frequency spectrum. The new concept relies on the idea that a plasma wave in the RF range can modulate the intensity of ECE. The modulation of ECE can be detected outside the vacuum vessel using the conventional mm-wave heterodyne technique. We have implemented the first such RF diagnostic system utilizing multiple mm-wave mixer antennas in the ECE imaging system on the KSTAR. The mm-wave-based RF (mwRF) diagnostic system provides high-resolution (~1 MHz) measurement with a broad frequency range (0.1~8 GHz).
We have applied the mwRF diagnostic to the plasma waves associated with a variety of MHD and kinetic phenomena such as sawtooth crashes, runaway electrons, and plasma disruptions. For the case of the edge transport barrier (pedestal) collapse, we observe intense narrow-band emissions in the whistler frequency range (~ 1 GHz). The measurements at two different toroidal positions (about 18.5° apart) show that intense waves of similar spectral characters are observed only along a specific direction (approximately the magnetic field line). This observation supports that the onset of pedestal collapse occurs at a localized region in the edge [8, 9] and generates RF waves propagating mostly parallel to the magnetic field line.
We now use the mwRF diagnostic system to determine the wavenumber of the plasma waves embedded in the modulation of ECE emissions. We envision that passive diagnostics utilizing plasma waves will be one of the most practical solutions as reactor-relevant diagnostics beyond ITER. This work was supported by the NRF of Korea under grant No. NRF-2019M1A7A1A03088456 and BK21+ program.
References:
(1) R. E. Waltz, W. Pfeiffer, and R. R. Dominguez, Nucl. Fusion 20, 43 (1980)
(2) W. W. Heidbrink, Phys. Plasmas 15, 055501 (2008)
(3) R. O. Dendy and K. G. McClements, Plasma Phys. Control. Fusion 57, 044002 (2015)
(4) G. S. Yun et al., Rev. Sci. Instrum. 85, 11D820 (2014)
(5) S. G. Thatipamula et al., Plasma Phys. Control. Fusion 58, 065003 (2016).
(6) B. Chapman et al., Nucl. Fusion 57, 124004 (2017)
(7) M. H. Kim et al., Nucl. Fusion 58, 096034 (2018)
(8) G. S. Yun et al., Phys. Rev. Lett. 107, 045004 (2011)
(9) J. E. Lee et al., Sci. Reports 7, 45075 (2017)Speaker: Gunsu YUN (Pohang University of Science and Technology) -
Chinese Fusion Engineering Testing Reactor (CFETR) aims to bridge the technical gaps between ITER and the first commercial fusion power plant [1]. The physics design of the operation scenarios should provide an integrated solution for the plasma to meet the key mission goals for CFETR subject to engineering constraints. For this purpose the optimization of scenarios and the exploration of operating space have been performed by integrated modeling simulations and system code studies in order to address the outstanding challenges that emerged from previous work presented in the last IAEA FEC.
Previous sensitive study of operating space by system code studies showed there were a reasonable number of feasible operating points for CFETR in steady state with high fusion power [2]. However, there were unresolved gaps between the 0-D design developed by the system code study and the preliminary scenarios developed by the integrated modeling simulations as shown in the last FEC [3]. Compared to the 0-D system study the simulations showed more pessimistic performance and/or more challenging requirements on external actuators such as high external current drive power and deep fueling. Another challenge is that the hybrid scenario and the steady-state scenario required quite different compositions of current drive systems to meet the long-pulse and high-fusion target for CFETR. Recently this problem has been solved by optimizing the pedestal parameters, the current drive techniques and the safety factor profiles in the integrated modeling simulations with self-consistent core-pedestal coupling.
Several new operating points are found in the simulations including a long pulse hybrid scenario and a steady-state scenario both with 1 GW fusion power (see Figure 1 for kinetic profiles). The designed hybrid scenario has a high density plasma (<$n_e$>$_{vol}/n_{Greenwald}\approx1.0$, $n_{e,ped}/n_{Greenwald}\approx0.9$) with good confinement ($H_{98y2}\approx1.2$) and a flattened q profile in the deep core ($r/a$<~0.4). The global ideal MHD modes ($n$=1~8) are stable since the value of $\beta_N$ is moderate, while the size of ELMs at pedestal is small ($ΔW/W_{ped}$<~1%). The steady state scenario relies on higher confinement enhancement ($H_{98y2}$~1.33) to achieve the same fusion power with a lower plasma current and a lower pedestal compared to the hybrid scenario. The confinement enhancement is ascribed to a local magnetic shear reversal at mid-radius ($r/a$~0.5) sustained by local RF current drive and the bootstrap current. It is not necessary to have deep fueling, which was required in the preliminary simulation shown in the last FEC. The position of magnetic shear reversal is optimized to make $\beta_N<4l_i$ for the stabilization of the global MHD modes ($n$=1,2). The pedestal of steady state scenario is expected to be compatible with the grassy ELMy condition [4] since $\beta_P$ is sufficiently high (~2.2) to suppress the more volatile balloonig modes.
A single composition of top-launched high frequency EC waves, helicon waves launched from high field sides and high energy neutral beam with large tangential radius can sustain both the scenarios according to the following consideration. Both helicon waves and high energy neutral beams have demonstrably higher current drive efficiency than EC waves and are useful to reduce the total power requirement for current drive. However, their maximum power will be limited by the available ports through the breeding blankets. Both EC waves and helicon waves can be optimized to drive localized current profiles to sustain the reversed shear in the steady state scenario while the neutral beam current drive profile is quite broad.
Ongoing work in scenario development includes verification of these solutions with models of varying physics fidelity and sensitivity study of the operating points due to the uncertainties in physics and technology. The latter is being performed with iteration between integrated modeling simulations and system code studies.
Based on the above integrated modeling results, new system studies are started using a flexible System Code Framework (SCoF). The framework is being developed to build comprehensible, adaptable and extensible system codes which can incorporate the evolving output from the ongoing design research for CFETR to make comprehensive system studies. The first application of SCoF produces a system code to benchmark the 0-D system study by the GASC code [5] with the output of the integrated modeling simulations. The benchmark shows that the fusion power can be overpredicted in a system code by about 30% if the Shafranov shift is neglected, and by about 20% if simple global polynomial profiles are used to approximate the kinetic profiles for transport evolved H-mode plasma with self-consistent pedestals. The consistency between the 0-D system code study and the integrated modeling simulation can be improved significantly by this benchmark (See the table in Figure 2).
Based on SCoF the system code is then adapted to explore the sensitivity of the operating point of the hybrid scenario. Operating points are calculated with a statistical sampling about a mean value over the ion temperature, the electron temperature, the plasma current and the radius of plasma boundary at the outer midplane (i.e., $R_{out}=R_0+a$) while <$n_e$>$_{vol}/n_{Greenwald}$=1.0 and the composition of current drive power is kept fixed. Increasing $R_{out}$ is an attempt in the recent research to enhance the control of vertical displacement events. The sensitivity study by the system code shows that increasing both the plasma size a and the major radius $R_0$ together (by increasing $R_{out}$) will decrease either the fusion power or the pulse duration $\tau_{flattop}$ (if keeping the fusion power by increasing plasma current $I_p$) (see Figure 3). This behavior is consistent with the recent observation in the integrated modeling simulations.
Further development is ongoing to incorporate more technology modules relevant to the CFETR design and scaling laws of physics components, obtained from detailed modeling, such as the parametric dependences of the pedestal, the efficiency and location for current drives, the characterization of the kinetic profiles with transport barrier in the core and at the edge, and so on. Then a more complete sensitivity study with high fidelity can be performed to estimate the robustness of the scenarios. The implication of these studies for future fusion reactor design is discussed.
References
[1] Y. Wan et al., Nuclear Fusion 57, 102009 (2017).
[2] J. Morris et al., Fusion Engineering and Design 146, 574 (2019).
[3] G. Zhuang et al., Nuclear Fusion 59, 112010 (2019).
[4] Y.-R. Zhu et al., Nuclear Fusion 60, 046014 (2020).
[5] R. D. Stambaugh et al., Fusion Science and Technology 59, 279 (2011).Speaker: Dr Jiale Chen (Institute of Plasma Physics, Chinese Academy of Sciences) -
Studies are carried out examining the dependence and sensitivity of fusion power production, temperature and density pedestals on edge density fueling strength, current density profile, alpha heating, and magnetic field strength. The goal of these integrated ITER simulations is to identify dependencies that can impact ITER fusion performance.
The self-consistent predictive core-pedestal ITER scenario modeling is carried out. The Weiland anomalous transport model {1} is combined with the NCLASS module in order to compute the evolution of electron and ion temperature, density, toroidal and poloidal rotation, and flow shear profiles from separatrix to the magnetic axis. The nonlinear anomalous transport multifluid model includes ion temperature gradient, trapped electron, kinetic ballooning, peeling, collisionless, and collision dominated MHD modes. The model has been derived from kinetic theory using a small parameter $\epsilon ̴\sim 10^{-2}$. The model can simulate internal and external transport barriers, the L-H transition, the nonlinear Dimits upshift, particle and heat pinches, and poloidal spin up. The other remarkable consequence of model is that it reproduces the experimental power scaling of the energy confinement time, $τ_E \sim P^{-2/3}$.
The simulations are started with prescribed sources and a guessed L-mode profile and evolve to L-H transition and temperature and density pedestal self-consistently as shown in Fig. 1. There are no assumptions in the simulations that suggest that there will be an L-H transition or where should be the temperature and density barriers. In the simulation shown in Fig. 1, the magnetic-q is just above 2 at the separatrix and density at the edge is below the Greenwald limit. Although, the particle pinch is present in the simulations, a peaked density profile is not obtained. However, there is no problem in reaching the design performance of ITER with fusion Q = 11.5 and with temperature pedestal heights around 3.5 keV.
The auxiliary heating and current drive profiles are suggested by the central ITER team {2} and the Alpha heating is computed from an analytic formula depending on the local density and the temperature. The neoclassical transport is added to the ion channel.
In general, the ion modes dominate in the interior while electron modes dominate on the barrier. In some ITER simulations both electron and ion modes are found to be stabilized by flow shear in the barrier although the electron mode is considerably stronger without flow shear. There are also cases where the electron channel is not completely stabilized.
The density and temperature pedestals are found to be quite sensitive to the particle source at the edge as shown in Fig.2. The edge particle source is assumed to be Gaussian in space and its amplitude is increased around 20%. The increase in the edge particle source has two effects. First, it raises the edge density and second it raises the reaction rate close to the edge, thus increasing the temperature fluxes that results in weak temperature pedestal barriers. Clearly, there must be a particle pinch somewhere to get in the particles from the edge source that raises the edge particle density as can be seen in Fig. 2.
The increase in magnetic-q (reducing current density) tend to trigger more kinetic ballooning modes at the edge and decrease in magnetic-q means that peeling gets worse but ballooning gets better. Moreover, it appears that large edge magnetic-q tends to give hollow density (not shown here) while small edge magnetic-q gives normal density profile. This behavior is consistent with the result that n∼1/q {3}.
Simulations are carried out with and without Alpha heating. The effects of Alpha heating are found to reduce the slope of the H-mode pedestal. This is expected to significantly ease the problem of ELMs causing damage to the first wall.
Simulations are also performed with B-field increased. A pronounced edge barriers are found when the strength of B-field is increased. The increased B-field have a distinct effect on the MHD modes at the edge. The interesting feature is found that the barrier collapses around q = 2 as MHD stability has reduced to having the barrier replaced by a continuous slope but with about the same fusion Q value. This would actually be good for a reactor.*One of the authors, T. Rafiq, would like to acknowledge support by US DOE Grants DE-SC0013977 and DE-FG02-92ER54141.
{1} J. Weiland, Stability and Transport in Magnetic Confinement Systems (Springer, New York, Heidelberg, 2012); T. Rafiq et al., Phys. Plasmas 20, 032506 (2013).
{2} TER Physics Basis Editors, “Progress in ITER physics basis II confinement” Nucl. Fusion 47, S1–S413 (2007).
{3} Garnier, et. al., Phys. Plasmas 24, 012506 (2017).Speaker: Dr Tariq Rafiq (Lehigh University) -
The Advanced FRC is a Field Reversed Configuration maintained by neutral beam (NB) injection and electrode biasing (EB), with scrape-off-layer (SOL) pumping and electron heat confinement provided by expander divertors. This alternate magnetic confinement system has been developed at TAE Technologies, Inc in the C-2 [1,2], C-2U [3,4], and C-2W [5] series of devices. In this paper we summarize the recent developments in the simulation of equilibrium, stability, and transport of this configuration. For illustration, the C-2W Advanced FRC configuration [5] is sketched in Figure 1.
As indicated in the figure, the plasma is confined in a region which includes a true magnetic null at the O-point and at the two X-points. The magnetic field reversal and closed magnetic geometry is provided largely by diamagnetic current, which peaks around the separatrix. The ion orbits are comparable in size to the FRC and many are betatron orbits rather than drift orbits. About half of the plasma lies inside the separatrix, and about half is in the SOL. The SOL is collisionless and exploits a combination of magnetic and electrostatic confinement.
Equilibrium of the FRC is studied with several models including one which allows multiple ion species with strong toroidal ion rotation [6]. This has recently been coupled to a fast ion Monte Carlo code which models the NB injected fast ions, to create a hybrid fluid/kinetic equilibrium model which can accommodate a significant (> 50%) anisotropic fast ion pressure component. It is found that the separatrix shape is influenced by a combination of external magnetic field shaping, the ratio of edge pressure to O-point pressure, and the degree of anisotropy of the energetic ion population.
Global stability of the FRC is studied using the FPIC and HYM [7] hybrid PIC codes. These codes agree on the well-known kinetic stabilization of the internal MHD tilt mode of the FRC. It has previously been shown with HYM that NB injection into FRC plasmas can have a stabilizing or destabilizing effect on global MHD modes depending on the mode and beam parameters [8]. On C-2W it has been shown experimentally that EB can be stabilizing or destabilizing to radial modes depending on the sign of applied bias. A remarkable discovery of the C-2W experiment is a new operating regime where these external actuators, NB injection and EB, have a synergistic effect when applied simultaneously, leading to much more stable and long-lasting plasmas than when either actuator is used alone. The FPIC and HYM codes are being used to understand this important synergistic effect.
Global confinement in macroscopically stable FRC plasmas is due to a combination of perpendicular confinement by magnetic field inside the separatrix, and parallel confinement by magnetic field gradients and electric fields in the SOL. The perpendicular and parallel transport are coupled [9]. Reduced modeling of global transport has been performed using a hybrid fast ion + MHD code to characterize the global interaction of perpendicular transport in the FRC, parallel outflow in the SOL, and field line expansion and electrostatic potential formation in the expander divertor [10]. The same code has been used to study equilibrium relaxation, spin-up, and development of equilibrium rotation velocity under the influence of EB and NB external actuators.
Parallel electron heat transport modeling in the SOL has been performed using a custom developed 1d2v continuum code [11]. As expected, it has been shown that a pre-sheath potential is formed in the expanding magnetic field of the divertor, which helps to confine plasma electrons and reduces the amount of cold electron emission from the walls. Simulations predict that parallel electron heat loss is close to the minimal theoretical limit, a result which has been validated by experimental diagnostics. This helps to explain another remarkable discovery of the C-2W experiment, which is a high Te operating regime.
Perpendicular turbulent transport is being modeled using the 3D particle-in-cell codes ANC [12] and GTC [13]. In previous simulations [14–16] a gyrokinetic particle push was used meaning that the magnetic null regions had to be removed from the simulation domain. Nevertheless, non-linear simulations found qualitative agreement in fluctuation spectrum with experimental Doppler Back Scattering measurements [15]. To include the magnetic null regions in global simulations, which entail a significant fraction of figure-8 and betatron orbits, a “blended” drift-Lorentz particle pusher [17] has recently been implemented, and a new 𝛿𝑓 model has been developed and applied. Using this new particle model, linear and non-linear results are consistent with the previous gyrokinetic simulations. The new turbulence simulations also include kinetic electron effects which were previously removed in global simulations [15,16] (but kept in local simulations [12,14]). The resulting model has been used to compute the perpendicular ion and electron heat conductivity due to turbulent transport.
The combination of insights gleaned from simulations of plasma self-organization and equilibrium, the influence of external actuators on global stability, and global, coupled perpendicular and parallel transport, are being used to help design a next step Advanced FRC device.
References
1 M.W. Binderbauer, et al, Phys. Rev. Lett. 105, 045003 (2010).
[2] M. Tuszewski, et al, Phys. Plasmas 19, 056108 (2012).
[3] M. W. Binderbauer, et al, Phys. Plasmas 22, 056110 (2015).
[4] H. Y. Guo, et al, Nat. Commun. 6, 6897 (2015).
[5] H. Gota, et al, Nucl. Fusion 59, 112009 (2019).
[6] L. Galeotti, D. C. Barnes, F. Ceccherini, and F. Pegoraro, Phys. Plasmas 18, 082509 (2011).
[7] E. V. Belova, R. C. Davidson, H. Ji, and M. Yamada, Phys. Plasmas 11, 2523 (2004).
[8] E. V. Belova, R. C. Davidson, H. Ji, and M. Yamada, Phys. Plasmas 13, 056115 (2006).
[9] L.C. Steinhauer and H.L.Berk, Phys. Plasmas 25, 022503 (2018).
[10] M. Onofri, et al, Phys. Plasmas 24, 092518 (2017).
[11] S. Gupta, et al, in Bull. Am. Phys. Soc. (American Physical Society, 2017).
[12] D. P. Fulton, et al, Phys. Plasmas 23, 056111 (2016).
[13] Z. Lin, T. S. Hahm, W. W. Lee, W. M. Tang, and R. B. White, Science 281, 1835 (1998).
[14] C. K. Lau, et al, Phys. Plasmas 24, 082512 (2017).
[15] C. K. Lau, et al, Nucl. Fusion 59, 066018 (2019).
[16] J. Bao, et al, Phys. Plasmas 26, 042506 (2019).
[17] R. H. Cohen, et al, Nucl. Instrum. Methods Phys. Res. Sect. A. 577, 52 (2007).Speaker: Sean Dettrick (TAE Technologies, Inc) -
Recently, improved high-performance plasma operation has been significantly extended towards more ITER and CFETR related high beta steady-state regime with optimization of current profile (βP ~ 2.5 & βN ~ 1.9 with ITB +ETB of using RF & NB and βP ~ 1.9 & βN ~ 1.5 with eITB + ETB of using pure RF) on EAST 1. The ITB formation and sustainment company with optimization of the current profiles, which seem to be due to different MHD-modes, similar with other devices 2. The mechanisms of linking the observed changes in MHD-behaviour and current profile seem to play an important role for ITB formation and sustainment in high beta steady state plasma on EAST.
First demonstration of a >100 s time scale long-pulse steady-state scenario with a good plasma performance (H98(y2) ~ 1.1, with eITB) has been achieved on EAST using the pure radio frequency (RF) power heating and current drive. A typical discharge, with NB injection based on the >100 s long pulse discharge parameter, is shown in Figure 1. Better confinement was obtained with flat central safety factor q~1 profile, with a flat boundary is about 0.2 m, and sawtooth-free in plasma discharges, accompanied by a long live mode (LLM) with a few times of current relaxation time. The LLM is localized in the core with the amplitude (10-20 Gauss) directly measured by Faraday-effect based polarimetry, and mode number (1,1) is determined by toroidally and poloidally separate diagnostics. The helical displacement of LLM is estimated to be as large as 2-3 cm, implying helical structure in the core may play an important role in current relaxation to sustain flat q profile. The measured local peaking radial magnetic fluctuation of LLM is ~90 Gauss, which is at least 10% equilibrium poloidal magnetic field inside q=1 surface, suggesting a 3D equilibrium topology.
The ITB and ETB are both observed with an optimizied flat central q profile in the long-pulse high βN operation, as shown in Figure 2. The ITB of ion temperature, electron temperature and electron density has been obtained simultaneously during the fishbone event. And the ITB+ETB discharges sustain a few seconds with the long live fishbone in the H-mode discharge 80496, the central q profile is flat and q ~ 1 inside ρ<0.3. The sustainment of internal transport barrier of electron density is accompanied with m/n =1/1 fishbone while it has no fishbone in the phase without ITB at t = 3.7 s. Further analysis shows that the m/n =1/1 fishbone signals located at the region of ρ=0.3 which is consistent with central flat q profile and ITB region of electron density. The formation and sustainment of the central flat q profile could be associated with the presence of the fishbone mode. Also, relationship between the formation of ITB and fishbone in EAST high βN ELMy H-mode discharge is confirmed in discharge 56933 in ref. 1. This current clamping effect is similar with other hybrid scenarios like in other devices. It suggests that some no-linear coupling exist between the MHD behaviors and the current profiles. The peak of the stored energy during the full shot revealed that the plasma was well constrained.
The 1/1 mode almost exists during the entire discharge period for improved H mode on EAST, and is a common phenomenon for many similar shots, which is similar to the LLM found on ASDEX-U and MAST 2. The over driven off axis current and fast transport of current from the core area give rise to larger difference between qmin and q=1 rational surfaces, which makes the internal mode with low m,n number more hard to be driven unstable. The combination of ECRH and LHCD plays an important role in sustaining flat q profile to avoid sawtooth crash. The 1/1 mode might play important role in shaping current density profile and sustaining the ITB in the high-performance plasma on EAST. The detailed process or mechanism based on internal measurements about how those modes can affect current profile is still missing. It is speculated that 1/1 modes may have interactions with background turbulence and play a role in current relaxation to sustain flat q profile and high-performance plasma with ITB..
1 X. Gao et al., Nucl. Fusion 57 056021(2017); B.N.Wan, et al., Nucl. Fusion 59 112003(2019); X. Z.Gong, et al., Nucl. Fusion 59 086030(2019).
2 J.Stober et al., Nucl. Fusion 47 728(2007); E. Joffrin et al., Nucl. Fusion 43 1167(2003); I Chapman et al., Nucl. Fusion 50 045007(2010).
Speaker: Prof. HAIQING LIU (Institute of Plasma Physics, Chinese Academy of Sciences) -
The KSTAR uses the NBI (neutral beam injection) as a majority of heating and current drive and has been exploring the inboard-limited ITB (Internal Transport Barrier) as an alternative candidate to achieve a high performance regime since 2016. The approach with the inboard limited configuration to avoid the H-mode transition prior to the formation of the ITB was effective at a given L-H transition characteristics and heating resources in KSTAR. The NBI power more than 4 – 5 MW under a limited L-mode was a key of the ITB access during the 2016 and 2017 campaign. The ITB formed in both ion and electron thermal channels, and performances are comparable to the usual H-mode in KSTAR. A stable ITB discharge, which was sustained for about 7 s, was generated in a weakly reversed $q$-profile with the maximum available NBI power of 5.0 MW $^{1}$.
However the maximum available NBI power was limited to 3 MW due to technical difficulties during the last two campaigns. Meanwhile the doubled capacity of the in-vessel cryopump (IVCP) allowed the recycling control more practical. Under these conditions, we attempted to extend the operating window by controlling the plasma shape and position, and the experiment successfully demonstrated that the ITB is accessible and sustainable with a marginal NBI heating power of about 3 MW.
The key control parameters of the experiment were the triangularity ($\delta$) and vertical position ($Z_{p}$) of the plasma $^{2}$. The shape control attempted to divert the plasma to a vertically shifted Upper Single Null (USN), with a marginal touch of the inboard limiter, so that the plasma can remain in L-mode at the boundary. To do this, the $dsep$, which is the distance between the closest separatrix just outside the discharge boundary and the limiter touch point, goes to 0 cm at 3.0 s. We have also slowly increased the $\delta$ about 0.3 until 5.0 s and moved up the plasma about ±5 cm from the midplane. Here, the NBI off-axis heating provides current density profile modification and it flattens the $q$-profile. This was intended in the vertically shifted USN configuration, and it was found that the duration of ITB performance was found to be related to the striking point of the upper divertor.
Figure 1 shows a result of the shaped ITB in 2018. We have applied 2.8 MW of NBI power. The plasma control reduced the attached area to the inboard limiter and moved up the plasma slowly to feel the vertical off-axis current drive. In this particular shot, we have reduced the electron density with a less 2nd gas puffing (reduced by ~ 75%), and the formed ITB with the $T_{i}$ ~ 9 keV lasted for about 1.5 s until the 8 inboard pellet injections at $t$ = 5.0 s terminated it. Without the density control, the ITB lasted shorter or experienced thermal sawtooth oscillations. The stored energy and the $\beta_{N}$ are 350 kJ and 1.6, respectively, in the discharge. The ITB foot is located roughly at R = 2.0 m which corresponds to $\rho$ = 0.3. The ion toroidal velocity is even faster during this high $T_{i}$ discharge.
We have carried out analysis of shaped ITB discharges. No significant instability was observed in the Mirnov spectrum. Even at a particular ITB discharge with sawtooth oscillations on both ion and electron temperatures we were only able to see high $n$ (= 5) dominant weak fluctuations. This MHD-resistant characteristic can be thought of as making the ITB discharge robust and reproducible, and it should be related with the shape of $q$-profile. During the first observation of the ITB in 2016, a flat $q$-profile has been observed in the central region of $ρ$, and the flatness tends to be monotonous to be distinguished by the difference of $q_{0}$ $^{3}$. The $q_{0}$ is near $q$ = 1 during the 2018 experiment with < 3.0 MW of NBI power, while we could observe a weakly reversed $q$-profile with $q_{0}$ ~ 2 at a higher power (~ 5.0 MW, 2016 campaign). We have got clear pressure profiles during the period of stable ITB, and this helped clear analysis of the discharge. Here we can see the formation of the ITB reduces both ion and electron thermal diffusion in the region of $\rho$ < 0.3.
In 2019 campaign, the density was measured at a level of 60 - 70 % lower than in the previous year. This is a partial failure of the other diagnostic device we use, but experienced a temperature overshoot that occurred at lower density at around 3 s during the ITB onset. The high temperature gradient in the core during the overshoot caused the barrier unsustainable and this can be prevented with the appropriate additional gas puffing before the ITB onset timing. The ITB is terminated by beam blips for 10 ms for measuring the ion temperature or impurity accumulation in the vacuum vessel. At higher NBI power above 3 MW, this termination is greatly alleviated, but more technical solution was to adjust striking point of the upper divertor. KSTAR demonstrated that this plasma shape control enables ITB to be sustained for more than 10 seconds at relatively low NBI power.REFERENCES
$^{1}$ J. Chung et al, Nucl. Fusion 58, 016019 (2018)
$^{2}$ J. Chung et al, 46th EPS conference, P4.1092 (2019)
$^{3}$ J. Chung et al, Rev. Sci. Instrum. 89, D10112 (2018)Speaker: Dr Jinil Chung (National Fusion Research Institute) -
Lower hybrid waves (LHW) are absorbed in the scrape-off layer (SOL), and then the heated plasma follows the magnetic field lines in the co-current and counter current directions, which intercepts the LHW antenna limiter and divertor plate [1,2]. Hot spots are observed on the guard limiter, and heat flux striations on the divertor plate are observed on both the ion and electron drift sides. In EAST, it was found that a strong electron-ion side asymmetry existed on the striated heat flux. Striated heat flux at the ion side distributed widely in the toroidal direction, while the striated heat flux on the electron side was much weaker and hard to observe at low plasma density. This paper presents the electron-ion side asymmetry on LHW induced striated heat flux, and the divertor geometry affect on striated heat flux in the EAST.
The striated heat flux was found to move to the outer strike point (OSP) during plasma current ramp-up3. Fig. 1 shows the outer divertor heat flux under lower single null (LSN) discharge, with 0.5MA plasma current and ~3.7×1019/m3 plasma current. The OSP is located at R~1.775m, and the striated heat flux peaks at R~1.797m. The peak striated heat flux on the ion side is ~0.65MW/m2, compared to 0.13MW/m2 on the electron side. The heat flux at the lower OSP is larger than at the upper OSP since this was a LSN discharge and most of heat from the main plasma flows to the lower divertor. Note that the heat source for the striated heat flux is from SOL heating generated by LHW absorption in the SOL, not outflux from the main plasma 1. The striated heat flux on the ion side is larger than the heat flux at the OSP, which indicates most of LHW power is deposited in the SOL. Recent experimental results in C-MOD indicated that roughly 80% of the LHW is observed to be promptly deposited in the edge plasma with a parasitic mechanism [4].
Previous results indicated that LHW absorption in the SOL increased with the plasma density and decreased with the outer gap[5]. Here we show that the divertor geometry also significantly affects the LHW absorption in the SOL with stable plasma density and outer gap. The maximum striated q// in Fig. 2(a) represents the qpeak mapped to the outer midplane. The maximum striated q// on the ion side decreased from 11MW/m2 to 6MW/m2 when the drsep changed from -2cm to 2cm (Fig. 2(a)). The plasma density was stable ~2.5×1019/m3 and the outer gap was maintained at ~5cm, while the q// on the hot spots at the ion side decreased by 2.5MW/m2 when drsep changed from -2cm to 0cm. The striated heat flux and hot spots on the electron side were not observed at low plasma density (2.5×1019/m3 ). This indicates the LHW absorption in the SOL was reduced when divertor geometry was changed from LSN to upper single null (USN). The striated q// is ~2cm away from the guard limiter in the midplane, but ~3-5 times larger than the maximum q// at the hot spots. This indicates the LHW absorption in the SOL is stronger when the absorption is closer to the separatrix.
At higher plasma density ~3.5×1019/m3, the striated heat flux appeared at the electron side. Fig. 3 shows an EAST L-mode with Ip=0.4MA and 1.4MW LHW. The striated qpeak at the ion side and qpeak on hot spots at the ion side decreased from double null (DN) to USN, which could be induced by both the decreasing plasma density and changes in divertor geometry. However, the striated qpeak at the electron side increased from 0.2MW/m2 to ~0.3MW/m2 when drsep increased from 0 to 1.5cm, despite the decreasing plasma density from 4.7×1019/m3 to 3.7×1019/m3. Meanwhile, the qpeak on the hot spots at the electron side increase from 0.9 to 1.2MW/m2 when the divertor geometry was changed from DN to USN.
However, the mechanism for the observed ion-electron side asymmetry on the LHW induced striated heat flux is still unknown, e.g. not explicable via Landau damping. Further research is needed for LHW edge loss control and first wall protection.
This work was supported by the US DOE contract DE-SC0016915 and National Natural Science Foundation of China (Grant No. 11775259).Reference
1 K.F. Gan, et al., Physics of Plasmas 26, 072506 (2019)
2 P. Jacquet, et al., Nucl. Fusion 51 (2011) 103018
3 K.F. Gan, et al., J. Nucl. Mater. 438, S364–S367 (2013)
[4] I. C. Faust, et al., Physics of Plasmas 23, 056115 (2016)
[5] K. M. Rantamaki., et al 2005 Plasma Phys. Control. Fusion 47 1101-8
Speaker: kaifu gan (university of tennessee,knoxville) -
Disruptions can have root causes in density limits, radiation cooling of the edge or MHD instabilities. The latter include external kink modes, tearing modes and Neoclassical Tearing Modes (NTM) that slow down and can eventually lock to the wall {1}. Operating a tokamak with lowest disruption rate is not only a matter of understanding the stability limits of a plasma, but also of developing a robust scenario to prevent/mitigate operator errors and reduce the failure rate of the control and heating systems. In order to preserve the machine integrity of large tokamaks, the number of disruptions must be reduced to a minimum or – better - eliminated. For example, on ITER no disruption shall go unmitigated at current above 8.4MA because of the severity of mechanical loads {2}.
In order to assess robust scenarios on ITER that approach the H-L back transition with minimum disruptivity during the termination phase, we have implemented in TRANSP {3} a scheme to control the burn phase and the current ramp-down phase, which mimics the plasma response to external actuators. Free-boundary transport simulations have been integrated with feedback control on the line averaged density, active control of NTMs {4} and a parametrization for the edge boundary conditions from SOLPS {5}. The latter has been modified from the original to allow dynamic control of seeding impurities and pellet fueling based on estimated values of edge localized modes (ELM) heat load and heat load to the divertor. The pressure pedestal width and height are evolved based on peeling-ballooning stability calculations. These simulations indicate that even for an H-mode exit at current between 12.5MA and 14MA there is a high probability of success to control the plasma until shut-down (see Fig.1). The plasma maintains vertical stability and radial control against the large inward shift that follows the large drop of poloidal beta during the H-L transition. This trajectory is safe within the limits of the transport and stability models used herein and for variations of the input parameters within 20% of the operating point.
The plasma cross-section needs to be reduced and the plasma centroid guided downward in the ramp-down to avoid vertical instability {6}. For the current rate of 0.21MA/s used here, which is the fastest rate during normal operation, the slowest reduction rate of the plasma cross-section is 500cm2/s. This results from free-boundary calculations that evolve the magnetic equilibrium and the transport in a tightly coupled scheme, with electron density during the H-mode ramp-down phase between 65% and 80% of the Greenwald fraction.
Auxiliary heating is used until the current is reduced to 7.5MA, since ITER cannot ramp-down the current with Ohmic heating only. Since the outer gap increases when the plasma shrinks, the electron cyclotron (EC) system is the only option to provide core heating in L-mode, because ion cyclotron (IC) coupling would be lost.
Sustaining H-mode during the ramp-down has implications for NTM control. The EC Upper Steering Mirror is designed to provide a highly focused beam. During the ramp-down phase the EC deposition broadens and the current drive needed for NTM control increases up to the maximum available. The requirements for NTM control must be compatible with the needs for heating the plasma core, which sets a lower limit on the current of about 12MA. Including all these constraints in the simulations leads to the conclusion that it does exist a safe trajectory requires an H-L transition at 12.5-14MA.
Theoretical models are not mature enough to model self-consistently the density evolution and the transition to/from H-mode in time-dependent simulations for scenario development. This is why designing experiments that mimic the conditions expected on ITER and validating the models against these data is critical for extrapolation.
It is important to notice that – while the control of the external input power is at a mature stage – experience on controlling the self-heating is lacking. Thus, while present-day experiments give us confidence on how to control the plasma current and shape to avoid disruptions due to low-q and vertical stability, the interplay between D-T fueling mix and heating power to access safely the H-L transition is a completely new area, affected by large uncertainties in the modeling.
Experiments on present devices are most needed to inform ITER on how to control the H-L transition in the presence of alpha heating, for example by leveraging the synergy between Ion Cyclotron waves and Neutral Beam fast ions. This is where simulations like the ones described here can guide designing dedicated experiments to inform ITER on how to design robust discharges, from startup to termination.This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, under contract number DE-AC02-09CH11466. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization.
{1} T. Hender et al, Nuclear Fusion 47, S128 (2007).
{2} M. Lehnen et al, Journal of Nuclear Materials 463, 39 (2015).
{3} R.J. Hawryluk, "An Empirical Approach to Tokamak Transport", in Physics of Plasmas Close to Thermonuclear Conditions, ed. by B. Coppi, et al., (CEC, Brussels, 1980), Vol. 1, pp. 19-46. Computer software, USDOE Office of Science (SC) DOI:10.11578/dc.20180627.4
{4} F.M. Poli et al, 2017 Nucl. Fusion 58 016007
{5} A.R. Polevoi et al 2017 Nucl. Fusion 57 022014
{6} Y. Gribov et al, 2015 Nucl. Fusion 55 073021Speaker: Francesca Poli (PPPL) -
To predict energy and particle transport in future tokamaks we cannot use experimental measurements as boundary condition. Therefore, we need integrated modelling from the SOL to the plasma center.
On the other hand, transport in the various plasma regions is known to different degree. In order to increase our confidence on the transport predictions, we need to validate the available transport models or assumptions against existing data, or at least to quantify the impact of the respective uncertainties on the overall confinement performance.
Core transport has been studied with theory-based quasi-linear models for more than two decades, with significant progress both in terms of verification in comparison to nonlinear gyro-kinetic simulations and validation in the description of the main channels of heat and particle transport.
However, detailed aspects of the quasi-linear models can still contain inaccuracies or uncertainties, particularly regarding the precise level of predicted stiffness, or the transport levels approaching the plasma periphery and close to the pedestal. Moreover, effects that can have a predominant non-linear character, such as the stabilization due to fast ions and to beta, are difficult to consistently capture within a quasi-linear description.
In this paper we show recent progress both in integrated modelling [1] and in the detailled validation of quasi-linear theory-based transport models in the core, for both electron-heated and ion-heated ASDEX Upgrade H-mode plasmas. The transport characteristics, like the dominant turbulence, are addressed. Two quasi-linear transport model are considered: TGLF [2] and QuaLiKiZ [3]. The effect of fast ions is retained only in terms of ion dilution.
The ASTRA code [4] provides the simulation frame, including theory-based core transport and a pedestal model, which allows us to determine the pedestal pressure for a given pedestal width $\Delta_{ped}$ by means of a transport constraint.
The boundary conditions of the density and temperature profiles at the separatrix are derived with a regression-based formula, Eq. (2) in [1], depending on the particle fluxes of D and seeding impurity, as well as on the pumping speed. The sensitivity of the predicted plasma thermal energy on the boundary condition is discussed. Multiple parallel ASTRA simulations with different $\Delta_{ped}$ are run, then the MISHKA code [5] selects the case with highest stable pedestal pressure. The experimental stored thermal energy is predicted significantly better than with the established IPB98(y, 2) scaling, as shown in Fig. 1. for a selected set of stationary plasma phases featuring a heating power scan, a gas puff scan, and a current scan.Experiments have been performed featuring a scan of the ion heat flux at constant total heating, by progressively moving from full on-axis to full off-axis NBI heating [6]. Two discharges at different levels of central ECRH are selected to assess the profile stiffness with the theory-based models TGLF and QuaLiKiZ.
TGLF-sat1 predicts particle transport very accurately in all cases. $T_e$ and $T_i$ are also close to the experimental profiles in the discharge with low $P_{ECRH}$ ($T_e/T_i \approx 1.3$), whereas they are overpredicted for high $P_{ECRH}$ ($T_e/T_i \approx 1.9$), as summarised in Fig. 2.GK simulations show a significant de-stiffening by fast ions for $T_e/T_i \sim 1$, so we can conclude that overall TGLF-sat1 is less stiff than the experiment. For electron transport in H and D plasmas, gyro-kinetic simulations indicate a significant role by the ETG at higher collisionality (typically $n_e \sim 5-9 ~ 10^{19} m^{-3}$), whereas at lower collisionality the TEM is dominant, leading to the prediction that ITG and TEM will prevail over the ETG for electron heat transport at the low collisionalities of a burning plasma. Modelling with TGLF matches the kinetic profiles with good accuracy, but for a too high sensitivity on $T_e/T_i$. QuaLiKiZ is more accurate at high collisionality.
[1] T. Luda di Cortemiglia et al., accepted for Nuclear Fusion (2020)
[2] G. Staebler et al., Phys. of Plasmas 23 (2016) 062518
[3] C. Bourdelle et al., Phys. of Plasmas 14.11 (2007) 112501
[4] G. V. Pereverzev, P. N. Yushmanov, IPP report 5/42 (1991)
[5] A. B. Mihailovskii, Plasma Phys. Rep. 23 (1997) 844
[6] F. Ryter et al., Nuclear Fusion 59 (2019) 096052Speaker: Dr Giovanni Tardini (Max-Planck-Institut fuer Plasmaphysik) -
Tungsten is foreseen as plasma facing material in next generation tokamaks (ITER/DEMO). It is thus crucial to understand and predict tungsten transport to prevent detrimental behaviour such as central tungsten accumulation leading, in worst scenarios, to disruptions.
In the framework of integrated modelling, ASDEX Upgrade and WEST discharges (both machines operating in full tungsten environments) are studied. Temperature and density profiles are evolved according to heat/particle sources and transport. Neoclassical transport is computed from NEO 1 and turbulent transport from the kinetic transport model QuaLiKiz 2. A faster, 10 dimensional Neural Network version of QuaLiKiz [3,4] is also used to tackle the real-time transport modelling challenges.
In modelled AUG NBI heated H-mode plasmas, within the transport solver ASTRA [5], the W accumulation and its avoidance are found to be determined by a competition between two mechanisms, i.e., the central particle source produced by NBI heating and the electron heating from ECRH. The former causes central peaking of the plasma density and the subsequent inward tungsten neoclassical convection. This accumulation can be compensated by the increase of $T_e/T_i$ with ECRH heating, resulting in increase central turbulent diffusion, thus reducing the effect of the central particle source (Fig.1) [6]. The competition of these two mechanisms is reproduced in a control oriented integrated modelling framework with the transport code RAPTOR [7], including fast turbulent transport modelling via Neural Networks of QuaLiKiz. In Fig. 2, the central and midradius electron normalised density gradients ($R/L_{n}$) are shown for increasing ECRH power. It is observed that the central $R/L_n$ decreases with $P_{ECRH}$ due to the increased central turbulent transport. On the other hand $R/L_n$ at midradius increases due to reduced collisionality [8] and reduced impact of the particle source compared to the turbulent pinch. Such fast and robust predictive capabilities with respect to central density peaking with the ECRH power, paves the way towards real-time core tungsten density mitigation.
Long pulse WEST L-mode plasmas with dominant radio-frequency electron heating and low torque are also modelled within ASTRA. These plasmas are particularly relevant considering scenarios of the Pre-Fusion Power Operation phase 1 of ITER where cyclotron resonance heating will be used (20 to 30 MW of ECRH), without neutral beams. It is found that predictions of the electron density and temperature profiles, modelled with QuaLiKiz, are inconsistent with experimental observations. Comparisons with first principle gyrokinetic simulations (GKW [9]) identified the reduced model for the collision operator as the main culprit [10]. Predictions of the midradius turbulent particle peaking (no central particle source) between GKW and QuaLiKiz for such collisional plasmas feature discrepancies in both high and low ion temperature gradient (Fig. 3) due to overstabilisation of Trapped Electron Modes driven by the electron temperature gradient. These discrepancies contributed to spurring development of the QuaLiKiz collisionality model to alleviate this issue.
Reduced kinetic-theory-based models (QuaLiKiz neural network) for transport modelling in view of plasma scenario optimisation and tungsten accumulation avoidance, are proven to be powerful and accurate tools. In specific regimes relevant for ITER operations, additional development are still required with constant validation against first principle modelling.
Acknowledgments: This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement
No 633053. The views and opinions expressed herein do
not necessarily reflect those of the European Commission.References:
1 E. Belli et al, PPCF 50, 095010 (2008)
2 C. Bourdelle et al, PPCF 58, 014036 (2016)
3 K. L. van de Plassche et al, PoP 27, 022310 (2020)
[4] K. L. van de Plassche et al, EPS (2019)
[5] G. V. Pereverzev et al, IPP-Report, IPP 5/98 (2002)
[6] P. Manas et al, 22$^{nd}$ Meeting of the ITPA Transport and Confinement Topical Group
[7] F. Felici et al, Nuclear Fusion 51, 083052 (2011)
[8] C. Angioni et al, Physical Review Letters 90, 205003 (2003)
[9] A. G. Peeters et al, Computer Physics Commnunications 180, 2650 (2009)
[10] X. Yang et al, submitted to Nuclear FusionSpeaker: Pierre Manas -
The development of tokamak start-up operation scenario often relies on operator’s experiences, rather than more robust approach based on numerical modelling. Such a trial-and-error approach has fortunately worked to find start-up recipes in small or medium size devices, but increases the risk of delays to experiments. Moreover, it would not be appropriate anymore for a large superconducting tokamak like ITER. First plasma operation in ITER is planned in 2025, and it is of crucial importance to ensure robust successful start-up. However, ITER has a risk of failure in full ionization of the prefilled gas (i.e. plasma burn-through), which is an essential condition for the start-up operation, because the toroidal electric field available is limited to 0.33V/m, much lower than the value typically used in now-a-days tokamaks i.e. 1 V/m. In order to assist the heating power in the plasma burn-through phase, a few MW of ECH power is planned for ITER First Plasma, but there still is a large uncertainty as the ECH absorption efficiency is very low (~ a few %) due to the low Te during the plasma burn-through phase. Therefore, it is now important and timely to develop a reliable plasma burn-through modelling tool including ECH to further optimize the operation scenario for ITER First Plasma.
This paper presents the summary of the code benchmark of plasma burn-through modelling codes, which was a joint modelling activity in ITPA-IOS for 2018 ~ 2020. For the first time, there was an extensive comparison between the plasma burn-through modelling codes presently available - DYON 1 [2] [3] [4], SCENPLINT [5] [6] [7], and BKD0 [8] [9] [10]. The benchmark activities were done in three steps, adding more complexities to the modelling in each step.
The first step was to verify the energy balance, particle balance, and circuit equations used in the three codes. With simple settings in the modelling e.g. constant plasma volume, no impurity, no ECH, etc, and with the parameter set foreseen in ITER First Plasma e.g. Hydrogen, Vloop =12V, Bt=2.65T, Bp=2mT, Vv=1000m3, R=5.65m, a=1.6m, ITER First Plasma was simulated, and the calculated threshold value of prefill gas pressure required for plasma burn-through was compared. Without any adjustment in the source codes to provide similar simulation results one another, interestingly, the three codes already predicted the same value of the threshold prefill gas pressure i.e. 0.8mPascal (see Figure 1), implying the overall consistency between the codes for the simplified modelling case. Each term in energy balance, particle balance, and circuit equation has been compared in detail, and the differences in the detailed simulation results were identified. The three codes were adjusted to match the detailed simulation results each other, and could successfully provide almost identical values for all the simulation results, such as time evolution of the plasma current, temperature, and density. This confirms the identified terms are the main differences for the simplified modelling case.
The second step was to validate the modelling against experimental data. For this, experimental data at JET such as loop voltage, plasma volume, and prefill gas pressure were used as input data, and impurity content evolution was also modelled in the codes. In this exercise, it was found that the plasma volume evolution and electromagnetic modelling of the eddy current in passive structure are important. Some differences were also found between codes which turned out to result from different atomic data. DYON and BKD0 use ADAS data, while SCENPLINT uses V. E. Zhogolev’s data [11]. Adjusting the term related to plasma volume evolution and eddy current, DYON and BKD0 were able to provide the same results, but SCENPLINT had different results, in particular, the impurity content evolution and radiation due to the different atomic data. The difference of the atomic data was investigated, and it was found that the radiation due to transitions without changing of main quantum number is taken into account in V. E. Zhogolev’s data, while ADAS does not include it.
The third step was to validate the modelling codes against ECH-assisted plasma burn-through discharges. Stand-alone simulations of ECH modules in each code were compared, and only small difference in the ECH absorption efficiency was found. Together with the ECH modules, the three codes simulated KSTAR discharges, and provided similar results until the carbon impurity become important for energy balance. This confirms that the three codes are consistent for ECH-assisted plasma burn-through, except impurity evolution because of the different atomic data.
After summarising the findings in the above joint benchmark activities, this paper will present the plans to develop reliable plasma burn-through modelling tool, in preparation of operation scenarios for ITER First Plasma.
1Hyun-Tae Kim et al, "Enhancement of plasma burn-through simulation and validation in JET," Nuclear Fusion, p. 103116, 2012.
[2]Hyun-Tae Kim et al, "Physics of plasma burn-through and DYON simulations for the JET ITER-Like Wall," Nuclear Fusion, p. 083024, 2013.
[3]Hyun-Tae Kim et al, "PSI effects on plasma burn-through," Journal of Nuclear Materials, p. S1271, 2013.
[4]Hyun-Tae Kim et al, "Plasma burn-through simulations using the DYON code and predictions for ITER," Plasma Physics and Controlled Fusion, p. 124032, 2013.
[5]Y. Gribov et al, "Chapter 8: Plasma operation and control," Nuclear Fusion, p. S385, 2007.
[6]V. A. Belyakov et al, "Analysis of initial stage of plasma discharge in tokamaks: Mathematical model formulation, simulation results, comparison with experiments," in International Conference on Physics and Control, Saint Petersburg, 2003.
[7]V. A. Belyakov et al, "Plasma initiation stage analysis in tokamaks with TRANSMAK code," Plasma devices and operations, pp. 193-202, 2003.
[8]G. Granucci et al, "Experiments and modeling on FTU tokamak for EC assisted plasma start-up studies in ITER-like configuration," Nuclear Fusion, p. 093025, 2015.
[9]D. Ricci et al, "Operational parameters for EC assisted start-up in ITER," in 43rd EPS, Leuven, 2016.
[10]D. Ricci, "Discharge recovery by means of EC assisted start-up," in 45th EPS, Prague, 2018.
[11]V. E. Zhogolev, "Impurity radiation from the peripheral plasma," Kurchatov Institute of Atomic Energy, Moscow, 1992.Speaker: Dr Hyun-Tae Kim (Culham Centre for Fusion Energy)
-
-
-
M. Nakamoto, H. Kajitani, T. Suwa, Y. Takahashi, M. Yamane, T. Baba, K. Sakamoto, K. Yoshizawa, Y. Uno, A. Ishikawa, M. Nakahira, N. Koizumi, M. Inoue 1, E. Fujiwara 1, T. Shichijyo 1, K. Kuno 2, T. Minato 2, T. Hemmi 3 and C. Luongo 3
National Institutes for Quantum and Radiological Science and Technology, 801-1 Mukouyama, Naka-shi, Ibaraki 311-0193, Japan
1 Mitsubishi Heavy Industries, LTD, 1 Minamifutami, Futami-cho, Akashi 674-0093, Japan
2 Mitsubishi Electric Corporation, 1-1-2 Wadasaki-cho, Hyogo-ku, Kobe, Hyogo 652-8555, Japan
3 ITER Organization, Route de Vinon-sur-Verdon, CS 90 046, 13067, Francee-mail: nakamoto.mio@qst.go.jp
In the last report, we announced the completion of the first coil case structure for ITER Toroidal Field (TF) coil [1]. Since then, the assembly of winding pack (WP) with the coil case has been completed. In this paper, we report the completion of the first ITER TF coil in Japan. The successful assembly of TF coil was achieved by developing manufacturing techniques to overcome some challenges; such as: (i) positioning of WP inside coil case to optimize deviation in current center line (CCL), a magnetic property, of a coil under a few millimeters, (ii) satisfying the tolerances on interfaces with other coils and supports, sometimes sub-millimeter, in spite of welding deformation during coil case welding, (iii) impregnation of narrow gaps between WP and coil case with high viscosity charged resin, and (iv) tracing CCL through the assembly process and transferring the information to the coil case.
ITER TF magnet is a D-shaped coil consisting of superconducting WP enclosed in a massive austenitic stainless-steel case with a height of 16m and a width of 9m. A set of 18 TF coils will be installed in the tokamak to create a magnetic field to confine the plasma within the donut-shape vacuum vessel. Among the set of 18 coils, 8 TF coils along with a spare are manufactured by Japanese domestic agency (DA) and remaining 10 TF coils are manufactured by EU DA. In Japan, TF coil is assembled in vertical posture with straight portion of D-shape facing the bottom to realize the bilateral symmetricity of the coil. The assembly is performed in the following steps; 1) insertion of WP into straight portion of coil case, 2) positioning of curved portion of coil case and welding with straight portion, 3) closure of coil case with inner cover plates, 4) impregnation of gap between WP and coil case, 5) final machining of interface surfaces and holes, and 6) dimensional inspections and electrical tests. It is important that generated magnetic field be toroidally uniform; therefore, those 18 coils shall have the same shape and the magnetic property to be assembled evenly around the vacuum vessel regardless of differences in manufacturing strategies, tooling and postures between two DAs. In order to achieve common requirements, unprecedented and severe tolerances are defined on both the CCL and the interfaces.
The CCL of a coil is the center line of an equivalent single turn coil, which has the same magnetic field properties at far away locations. The CCL shall exist within 2.6mm diameter circle over the extent of straight portion of more than 7m while keeping sufficient gaps between WP and coil case for impregnation [2]. In order to position WP to accommodate both requirements, the target WP position within coil case was assessed in advance using the individual dimensional inspection data of WP and coil case [3]. The WP position was monitored by laser tracker during insertion so that the fine adjustment of 0.1mm order was possible. After positioning of WP, measurement results of gaps between WP and coil case showed the validity of the prior assessment.
To install 18 TF coils around ITER vacuum vessel to create uniform magnetic field, very severe structural tolerances are defined at interfaces. Since the welding deformation has large impact on the positions and shapes of interfaces, extra materials were left on critical interfaces to be machined to nominal after welding. On the other hand, leaving too much of extra materials will lead to longer machining duration. To avoid unnecessary schedule delay, amounts of extra materials were minimized. Where extra materials were found not sufficient, two DAs collaborated to compensate all 18 coils in a common deviation under guidance of ITER Organization (IO) to harmonize the shape of final coils. This harmonization strategy was a great way to manage two conflicted issues, manufacturing schedule and accuracy, and a big achievement as international collaboration project.
For impregnation of the gaps between WP and coil case, there was a concern in breakage of the seal between WP and coil case at the terminal region, where conductor ends and instrumentation wires are extracted from WP, during evacuation before impregnation and pressurization after resin injection due to the relative movement of WP with respect to coil case. To prevent the relative movement, partial vacuum chamber is installed over the terminal region to balance the pressures on both sides of the seal. The pressure balance was monitored by pressure gauge and feedbacked to pressure control inside the partial vacuum chamber. However, filler added to impregnation resin as crack prevention increased viscosity of resin and therefore, the resin’s internal pressure during resin injection. For the first coil, adjustment of pressure balance required full attention of operator since resin’s internal pressure directly affected the pressure inside coil case. For the second coil, addition of relay tank between coil case and resin injection system prevented sudden change of resin pressure and improved manageability of pressure balance.
Since WP of TF coil is enclosed in coil case, CCL positional information needs to be transferred to reference points on coil case. As mentioned earlier, TF coil is assembled standing on straight portion of D-shape, resulting in gravitational deformation in WP. The amounts of deformation are about 1mm at the straight portion and 4mm at the curved portion. After coil case welding, deformation of WP is corrected by pushing curved region of WP against coil case by bolts [2] in accordance with the prior assessment using structural analysis. However, the resulted WP shape differed from the prior assessment. Since the WP is pushed against coil case, residual stress in coil case from welding affected the final WP shape. Even though we incorporated a step to release this residual stress by adjusting supports of coil case before WP shape correction, some stress remained in the coil case. This is unavoidable situation which must be treated carefully by measurement and feed-back during the fabrication process.
The first TF coil was completed in January 2020 successfully. For the following coils, the gap-filling impregnation process was improved due to experience of the first TF coil and implemented on the second coil effectively. WP of the third coil was inserted in coil case and is now in welding phase. TF coil assembly process is now routinized and its adequacy is proven. The TF coil will arrive ITER site in April 2020 for acceptance tests and installation in Tokamak, showing a progress in ITER project.[1] M. Nakahira et al., Nuclear Fusion, Vol59, No8, 086039 (2019)
[2] N. Koizumi et al., IEEE Trans. App. Superconductivity, Vol30, No4, 2971673, (2020)
[3] N. Koizumi et al., IEEE Trans. App. Superconductivity, Vol29, No5, 4200505, (2019)Disclaimer: The views and opinions expressed herein do not necessarily reflect those of the ITER Organization
Speaker: Mio Nakamoto -
In a recently conducted test for assessing compatibility of accelerator grid of Neutral beam 1 for their performance at 150 C, failure has been evidenced across an electrodeposited (ED) bond layer, which forms a vacuum boundary with cooling medium. This happens to be the first instance where an electrodeposited bond has been subjected to Hot Helium Leak test under operational conditions of temperature (150 C) and pressure (>2MPa), according to the ITER specifications. Presently there exists no recommended procedure for carrying out an assessment of the bond integrity for electrodeposited surfaces and also no codes are presently available to qualify the process compatibility for the operational requirements including application at around 150 C. Therefore, to arrive at the root cause, it was necessary to undertake a systematic evaluation study to identify the root cause and establish the adhesion characteristics at room as well as operational temperatures. This also serves the objective to assess and to arrive at an understanding of the possibility of recommending a metric for the strength parameters for such bonds. To carry out this assessment, multiple electrodeposited samples ~20, were drawn under different bath conditions. These samples have been subjected to a specially designed test procedure in order to arrive at an assessment of the strength parameters for the bond, the failure of which is considered to be the root cause. Hence, the present study is purported to present itself as a procedure for evaluation of this important manufacturing process. This is possibly first such initiative of its kind.
Samples were prepared under three different bath conditions with the parameter variation for electrolyte composition, additives, electrical parameters, time, temperature and pH. Adhesion, being one of the major area of consideration for the pressure application, a specific test, called ‘Push’ test has been configured (Fig. 1) in such a way that, it simulates the operational scenario and thus evaluate the stress on the bond area in the operational condition. The most important element of this test, is in the identification of procedure that, would test the performance of the bond for the thinnest cross section of surface contact of the deposit with the base material.
Fig. 1 shows the configuration of the cross sections where the top layer is ED and corresponds to cross section for the apertures of the grid segments. In addition, tests related to metallurgical assessments had to be developed to understand the microstructural characteristics at the transition region between the electrodeposited layer and the base material.
Studies performed across the samples show that interface of the base material and the electrodeposited layer is significantly impacted with the change in bath condition in terms of its adhesion strength, typically in the range of 15-20 %. The most significant observation comes out from the high temperature (150 C) assessment, where degradation in adhesion strength occurs to the tune of by ~30-50 % as compared to base material. This range in degradation is closely co-related to the change in the bath condition. Further, the Load Vs. Extension behaviour differs when the specimen is subjected to tests at elevated temperature. The graph (Fig. 2) reveals the absence of the elastic zone in case of high temperature zone for all the samples from different baths.
Further, the present case being a pure metal, it could also be possible to cross verify the outcome of the strength test through a hardness test (Fig. 3) across the base material, interface zone and the electro-deposited layer. These correlations of strength and temperature along with their dependency on the bath conditions establish a metric to develop the application based design for such components.
Metallurgical assessments to understand the microstructural characteristics across the base material, transition zone and electrodeposited layer (Fig. 4) shows the difference in the material growth characteristics for base material and the ED layer under different bath conditions thereby resulting in the variation of the adhesion properties at the interlayer.
In conclusion, a process has been established in form of experiments where ~20 samples, from different baths, have been subjected to tests to evaluate and obtain a statistical variation in the quality of the bond at room as well as at elevated temperature of 150 C. Test results shows the variation in the bond strengths is highly dependent on the bath quality, thereby establishing the root cause of failure in the hot leak tests of the grid segments. The results also establish the need for qualification of the bond integrity and improvisation of the bath characteristics, if required, to ensure a reliable application of ED process for the actual components. Recommendation of the qualification process is as follows; (1) carrying out and interpreting the specially designed push test for samples (2) Co-relating the strength with the hardness parameters (3) study of microstructural characteristics and (4) application of process on production pieces.
The details of the study performed above and the qualification procedure developed shall be presented and discussed. This forms an important database for similar components in fusion devices like ITER using ED process as one of the important manufacturing sequence.
1] Jaydeep Joshi et. al : Technologies for realization of Large size RF sources for –ve neutral beam systems for ITER - Challenges, experience and path ahead, Nuclear Fusion- 102937.R1Acknowledgement:
Support from M/s PVA Industrial Vacuum Systems, M/s Germany, Research Instruments, Germany and their sub-suppliers, in terms of providing the samples is sincerely acknowledged.Speaker: Jaydeepkumar Joshi (ITER-India (Institute for Plasma Research)) -
The present contribution is devoted to the neutral beam injectors (NBIs) for ITER heating and current-drive. First, updated information is provided about the development status of the entire NBI prototype (MITICA); starting in 2021, the first experiments will be dedicated to high-voltage holding tests in vacuum. Then the contribution describes the full-scale prototype of the NBI ion source (SPIDER) and the activities performed in the first two years of operation, devoted to investigating the operational regimes of the ion source and verifying the performances of the various plants in view of extensive beam operation. Some improvements have already been implemented; others are being prepared for the next shutdown. The characterisation of the SPIDER plasma is presented and the results of the first beam operations are reported.
The ITER experiment represents the next step in realising nuclear fusion as an environment friendly energy source for the future. To reach fusion conditions and to control plasma configurations, two heating and current-drive neutral beam injectors (NBIs) will provide the ITER plasma with 17MW each, by accelerating negative hydrogen/deuterium ions. The requirements of ITER neutral beam sources (40A D- ions at 1MeV for up to 1 hour, 46A H- ions at 870keV for up to 1000s) are so challenging that current-voltage-duration have never been simultaneously attained yet. So, in the dedicated Neutral Beam Test Facility (NBTF) at Consorzio RFX (Italy), an extended R&D activity is aimed at reaching full performances and optimizing reliable operation, in time for ITER experiments. To speed up the beam development, imposed by the tight ITER schedule, the NBTF includes two experiments: MITICA, full-scale ITER NBI prototype, and SPIDER, full-scale prototype of the ITER NBI source with 100keV particle energy, whose simpler construction allows anticipating several R&D activities with respect to MITICA. SPIDER operation started in spring 2018 aiming at investigating source and beam uniformity (over a 1m×2m area), negative ion current density, beam optics in conditions relevant to ITER requirements. MITICA will focus on beam acceleration, in terms of optics (divergence <7mrad, aiming <2mrad) and high-voltage holding in vacuum, and on beam propagation, governed by neutralisation and by electrostatic removal of residual ions.
Concerning MITICA, all power supplies and auxiliary systems were tested and installed; the vacuum vessel was completed, whereas the in-vacuum mechanical components are under procurement by F4E (see fig. 1).
Installation, commissioning and test of the 1MV power supplies started in autumn 2018 and are nearing their end. In particular, insulating tests for high-voltage components were successfully completed, after solving various technical issues and by adopting a step-by-step procedure during the integration of the sub-systems procured by different Domestic Agencies (European and Japanese). Power supply integrated tests with a dummy load (1MV, 50MW, 2s) are ongoing in 2020 and include the simulation of accelerator grid breakdowns using a short circuit device installed inside the vacuum vessel in the place of the beam source. These tests will allow verifying the full performances of the power supply systems and their reliability during grid breakdowns in normal operation. In the present contribution, the activities for completing and commissioning the power supply system, together with the results obtained during the integrated tests, are presented.
In the first two years, SPIDER operations generated a wealth of experimental information, which provided insight into the source performance and raised operational issues that must be resolved in view of extensive beam operation and particularly of MITICA. It is worth mentioning that, unlike any other existing NBIs, the entire beam source assembly of MITICA (and SPIDER) lies inside the vacuum vessel, so that it is surrounded by the background gas at low pressure in which the beam propagates. RF-induced overvoltages around RF drivers are found to result in electrical discharges when the gas pressure inside the vessel exceeds a threshold. In order to carry out experiments with all 8 RF drivers, the gas conductance between source and vessel was reduced by installing a mask limiting the number of beamlets to 80 out of 1280, thus lowering the gas pressure near the RF drivers. The mask will be removed when improved pumping system will be available (ongoing design). As for the SPIDER RF system, mismatch between self-oscillating frequency of the generators and resonance frequency of drivers and plasma can result in frequency jumps with consequent plasma stop; this condition was demonstratedly avoided by implementing feedforward control of the generator frequency and tailoring the RF frequency to plasma conditions. Several circuits and diagnostics are plagued with RF noise; the cause was shown to be due to currents flowing in the capacitive voltage dividers measuring the RF power; an alternative solution will soon be tested. Experiments allowed also clarifying the cause (layout of RF circuits) of interference between neighbouring drivers along with identifying some limitations to maximum RF power; solving these issues however requires major modifications of the RF circuits. Experiments showed that the magnetic filter field currents (aimed at extending the negative ion lifetime) affect the plasma performances; the cause lies in the SPIDER magnetic filter topology, which exhibits a null point and a large intensity inside the RF drivers; thus distorting electron trajectories. This issue was addressed by modifying the magnetic field topology installing additional busbars. Other issues regard: extension of plasma grid and bias plate voltages to comply with the higher plasma potential found in operations without caesium; characterisation of caesium evaporators; limitation to 30kV instead of 100kV of the acceleration power supply voltage.
In the meantime, the plasma was characterised by means of spectroscopy (including optical diodes). Additionally, at the beginning of 2020 an experimental campaign was devoted to performing spatially resolved measurements of plasma parameters by means of an array of electrostatic sensors installed on a temporary, remotely controlled structure entering the plasma through accelerator grid apertures. Electron temperature and density are shown in fig. 2: as expected with low magnetic filter field, plasma expands out of the drivers, whereas temperature decays gently towards the plasma grid (PG).
During the first extraction of negative particles from the source, features of negative ion beam and of co-extracted electrons were studied and correlated with plasma parameters. Particularly, the magnetic filter field effectiveness in reducing the co-extracted electron current was verified (down to ~30 times the H- current), along with its influence on negative ion current. The first characterisation of the SPIDER beam, in terms of beamlet divergence and deflection, was compared with numerical models while varying source parameters. The calorimetrical estimate of the current, which is not affected by secondary particles, is lower than the corresponding electrical measurements (see fig. 3).
The negative ion beam is confirmed to exhibit values of current density (up to ~25A/m2) and optics parameters (down to ~25mrad) similar to those usually obtained in volume operation. Finally, in 2020 for the first time, caesium will be injected into the SPIDER source to increase the negative ion density; the results of the first campaign will be described.
After these experimental activities, SPIDER will enter a long shutdown, to carry out a set of modifications identified either during the procurement phase or during the first year of operation, like the aforementioned rearrangement of the RF circuit configuration and the enhancement of the vacuum system to keep the vessel pressure low with no plasma grid mask.Speaker: GIANLUIGI SERIANNI (ItRFX) -
Introduction
Electron cyclotron systems of fusion installations are based on powerful millimetre wave sources – gyrotrons, which are capable to produce now megawatt microwave power in very long pulses. Gyrotrons for plasma fusion installations usually operate at frequencies 40-170 GHz. Requested output power of the tubes is about 1 MW and pulse duration is between seconds and thousands seconds (depending on plasma machine parameters). To provide operation with indicated parameters the gyrotrons have very large transverse cavity sizes, output barrier windows made of CVD diamond discs, electron beam collectors apply effective energy recovery. The main partners in the development of gyrotrons in Russia are the Institute of Applied Physics and industrial company GYCOM Ltd. The most developed gyrotrons are now the tubes for ITER which demonstrate parameters corresponding to ITER requirements. High-level parameters were also achieved with long-pulse 140 GHz gyrotrons developed for EAST and KSTAR installations. Some steps were done in development of higher frequency (230-700 GHz) gyrotrons for future plasma installations and for plasma diagnostics. Novel ideas were proposed to enhance gyrotron operation.
170 GHz gyrotron for ITER
In ITER installation there will be 24 gyrotron systems with 1 MW power each. Russian contribution consists of 8 gyrotron systems. ITER requirements include also high efficiency of the gyrotrons over 50%, possibility of power modulation with frequency up to 5 kHz, compatibility of the gyrotron complex with ITER control system. Development of the gyrotron system for ITER is based on solution of many very difficult scientific and engineering problems. At present the main purpose of the system modification is an enhancement of the system reliability and implementation of all gyrotron systems into ITER machine and its control and safety system.
In May, 2015 a Russian Prototype of ITER Gyrotron System was completed and its operation was demonstrated. The system includes gyrotron oscillator, liquid-free superconducting magnet, supplementary magnets, several electric power supplies, cooling systems control and protection systems, and other auxiliary units. The gyrotron system shows reliable operation with required parameters. In 2016-2019 three serial gyrotron systems were fabricated. All these three ITER gyrotron systems showed reliable operation in 1000 second pulses at megawatt power and efficiency higher than 50% (fig.1). The gyrotron microwave beam fed with low losses the corrugated HE11 waveguide of 50 mm diameter. The measured X-ray radiation and stray microwave radiation do not exceed safety levels. One more 170 GHz ITER gyrotron was delivered for EU team for testing microwave components.
Higher power and higher frequency gyrotrons
Development of a higher power gyrotron in Russia is going on along two directions: power enhancement in well tested gyrotron operating at TE25.10 mode and development of a new gyrotron with a new operating mode – TE28.12. Detail analysis of the test results showed that a slightly modified ITER gyrotron prototype is capable to operate at power 1.2 MW. First tests of the modified tube are rather encouraging: microwave power 1.2 MW at MOU output was demonstrated in 100 second pulses with efficiency of 53%. The gyrotron prototypes with TE28.12 operating mode were tested at 1.2 MW power with pulses up to 500 seconds.
Calculations of higher frequency gyrotrons for future plasma installations (230-300 GHz) were done. The calculations show that 0.2-0.3 MW gyrotrons can be made on the base of easily commercially available 10T magnets with 100 mm bore diameter. 1 MW tubes require at least 150 mm diameter magnets. 0.2 MW/ 260GHz/ CW tube is now in fabrication. Microwave power of 200 kW was demonstrated with a 660 GHz pulse gyrotron. CW megawatt power gyrotron with 230 GHz frequency is under development. 10 T magnet for the gyrotron with 150 bore is in fabrication by JASTEC. Simulations also show that using the gyrotron frequency stabilization and oscillator phase locking helps to provide stable gyrotron operation at very high modes (as TE56.24), at very high frequencies (as 350 GHz) with megawatt power. are encouraging. The latter parameters are required for future plasma machines with high magnetic fields.
References
- ITER NEWSLINE. 26Feb, 2018. https://www.iter.org/newsline/-/2931
- http://www.iterrf.ru/
- M.K.A. Thumm, G.G. Denisov, K. Sakamoto and M.Q. Tran. High-power gyrotrons for electron cyclotron heating and current drive. Nucl. Fusion 59 (2019) 073001 (37pp).
Speaker: Prof. Grigory Denisov (Federal Research Center Institute of Applied Physics RAS) -
This paper presents a progress of the achievement of performance tests of ITER-gyrotrons developed in QST and design of dual-frequency (170 GHz and 104 GHz) gyrotron to enhance various operation scenarios in ITER such as characteristics studies of H-mode/ELM at low magnetic field. Major achievements of the ITER gyrotron developments are as follows: (i) Manufacturing of 6 out of 8 sets of ITER gyrotrons was completed. Factory acceptance test (FAT) in QST has been progressed and 2 of 6 gyrotrons achieved required specifications such as 1 MW / 300 s / 50 %, 5 kHz modulation with ≥ 0.8 MW etc. The 50 mm diameter waveguide transmission-line and a matching optics unit (MOU) for ITER were newly introduced to perform the operation test at the same environment as ITER-site and excitation of HE11 mode purity of ≥ 95 % at the waveguide inlet was also successfully demonstrated, satisfying the requirement. (ii) Design of dual-frequency gyrotron, which is able to operate continuous wave of 1 MW power, was successfully completed.
Introduction
An ITER electron cyclotron heating and current drive system will be installed for initial breakdown, assist of electron heating during current ramp-up, on/off-axis current drive for the steady-state operation, plasma instability control using 24 sets of 170 GHz gyrotrons. The required specifications of the ITER gyrotron are 1 MW/continuous wave (CW) with efficiency of 50% and 5 kHz modulation capability with ≥ 0.8 MW. The QST procures eight sets of the gyrotron and four of them are planned to be used for first plasma. Six gyrotrons have been manufactured in the period of 2016 ~ 2019. In FEC2018, FAT results of a first ITER gyrotron were reported [A]. After the first plasma phase, low field operation at 1.8 T is being considered to evaluate threshold power for a H-mode. If the H-mode plasmas are generated, some issues such as a validity of L-H scaling, an ELM control and divertor heat loads will be demonstrated in early stage. The 1.8 T operation desires ~100 GHz RF for plasma breakdown, ramp up, heating and current drive in second harmonics and 170 GHz in third harmonic. In the QST, the multi-frequencies 1MW oscillation at 104 GHz / 137 GHz / 170 GHz / 203 GHz were demonstrated using a proto-type ITER gyrotron [B] However, the pulse length was only a few seconds for 104GHz and 137GHz. The design study of a dual frequency (104 GHz and 170 GHz) gyrotron to achieve long pulse operation of 1 MW power has been carried out in 2019.(i) Operation performance tests of ITER gyrotrons
The FAT of the 2nd ITER gyrotron was completed in 2019. The output power of 1.05 MW and total efficiency of 50.6 % in the long pulse operation was achieved. The oscillation frequency was 169.85 GHz. 20 similar operations were repeatedly performed 20 times with 25 % duty cycle and the operation reliability of 95 % was demonstrated. In 5 kHz full-power modulation, the maximum output power of 0.90 MW was achieved. As summarized in Fig.1, the 2nd gyrotron demonstrated the same performance as the 1st gyrotron.
The FAT of two gyrotrons were carried out using a prototype MOU and 63.5 mm diameter waveguides, which were assumed in the previous specifications. The integrated test introducing ITER relevant waveguides (50 mm diameter) and the MOU was carried out for the third gyrotron to simulate the operation at the ITER-site. As shown in Fig.2, a parabolic mirror and a phase-correction mirror in the MOU were newly designed to focus the beam size and flatten the phase of electric field at the waveguide inlet. It was shown that the mode purity of HE11 exciting at the waveguide inlet was 95.5%, which was comparable to the design value of 96.1 %. The power transmission efficiency in the MOU was 95.5 %, which was equivalent to the design value. These results satisfied the ITER requirement. At present, the 3rd ITER gyrotron is under commissioning and the output power of 1 MW for 100 s has been already achieved.
(ii) Design optimization of dual-frequency gyrotron of 170 GHz and 104 GHz
Issues for improving operation performance at 104 GHz were RF loss scattered in the gyrotron, which was about three times larger than at 170 GHz, and the spread of RF-sidelobe. To achieve the lower loss-power at 170 GHz and 104 GHz, Gauss mode content and directivity of radiated beam at an aperture of internal mode converter must be improved by modifying the inner wall-surface of the mode converter. In addition, a beam power distribution larger than -20 dB, which has non negligible power-level, should also be within the window aperture by modifying the curvature and the tilt of internal four mirrors located above the mode converter. As the results, the Gauss mode content achieved 95.4 % (170 GHz) / 97.5 % (104 GHz), much better than the ITER gyrotron, which are 94.5 % (170 GHz) / 90.7 % (104 GHz). These power transmission efficiency in the gyrotron achieved 98.8 % comparable to the design of ITER gyrotron. Both beams pass through the window center and are within -20 dB as shown in Fig.2. Moreover, two mirrors in the MOU were designed to improve the effective coupling of the beams to a 50 mm-diameter waveguide. Since the discrepancy of radiation angles at the window becomes only 0.1 º by design optimization of the mode converter, the same MOU mirrors for both beams are applicable. The coupling efficiency from Gaussian to HE11 mode of 95.6 % (170 GHz) and 96.1 % (104 GHz) has been obtained. As shown in Fig.2, the total transmission efficiency between the mode converter and the waveguide inlet is 95 % (170 GHz) / 93 % (104 GHz), higher than the ITER gyrotron design of 94 % (170 GHz) / 81 % (104 GHz). The blank circles in the Fig.3 represent the experimental power transmission efficiencies at the mode converter, at the output window, and at the waveguide. The experiment agrees with the design within error of 1-2%, it shows that the new design promises the long pulse, 1 MW operations at both 170 GHz and 104 GHz.
Conclusion
Manufacturing of 6 sets of ITER gyrotrons and FAT of 2 sets were completed. The output power of 1 MW with the efficiency of 50 % for 300 s operation and fast power modulation of 0.9 MW for 60 s were achieved. New MOU demonstrated HE11 mode purity of 95.6 %. The design of a dual-frequency (104 GHz and 170 GHz) gyrotron comparable to the same performance as the current ITER gyrotron was successfully completed.References
[A] Y. Oda, et al., Nuclear Fusion 59 086014 (2019).
[B] R. Ikeda, et al., J Infrared Milli Terahz Waves 38, 531 (2017).Speaker: Dr Ryosuke Ikeda (National Institutes for Quantum and Radiological Science and Technology) -
Shattered pellet injection (SPI) systems that form cryogenic pellets of low and high-Z impurities in a pipe-gun [1] for injection to mitigate disruptions have been fabricated and installed for use in thermal mitigation and runaway electron dissipation experiments on JET and KSTAR. These systems are to support disruption mitigation research for ITER and are based on an ORNL 3-barrel design for flexible pellet size and variable pellet composition studies [1]. The services for gas supply, vacuum, cryogenic cooling, and control are provided by the host institution with collaborative operation and experimental investigations organized by the host and through the ITER disruption task force [2].
The SPI systems for JET and KSTAR have a common feature of 3 different size pellets that are formed in-situ and collimated into a single injection line that enters the vacuum vessel. The pellets are fired by high pressure gas or mechanical punch and are shattered in stainless steel tubes with a 20-degree bend that are mounted inside the vacuum vessel of the tokamak, vertically on JET and horizontally on KSTAR. The JET installation shown in Fig. 1 has the unique feature of vertical SPI mounting and injection with the shatter plume aimed toward the inner wall to intercept known runaway electron (RE) beam locations generated from argon gas injection induced disruptions. Observations of the shattered pellet plume in plasmas on JET shown in Fig. 1 verify that the trajectory is as designed.
The JET SPI pellet sizes are 4.5, 8, and 12.5 mm diameter with lengths that are 30-50% longer. The two large sizes can optionally be operated with a mechanical punch for release of pure neon and argon pellets, that otherwise cannot be fired with gas alone. The control of the SPI is through a programmable logic controller programmed to automate the formation of the pellets and control vacuum components. Verification of the D2 propellant gas removal by the vacuum system was achieved by firing gas without a pellet and measuring that < 0.25% of the total gas fired ended up in the torus, showing excellent removal to prevent misinterpretation of the SPI performance by gas reaching the plasma before the pellet.
The KSTAR SPI installation described in Ref. [3] and shown in Fig. 2a has two identical 3-barrel SPIs that are mounted on the midplane with identical shatter tubes inside the vessel aimed to the plasma center after traversing a 20-degree shatter tube. Unlike the JET SPI that uses cold helium gas, these SPIs are cooled with a cryocooler that provides enough cooling to achieve 8 K pellet formation temperatures. Cooldown takes under 2 hours to be at pellet formation conditions, and it takes 5-15 minutes for the formation depending on the pellet size. The pellet sizes are 4.5, 7, and 8.5 mm with lengths that are 40% larger than the diameter. The pumping system and infrastructure were all provided by NFRI [3] and became operational in Nov. 2019. The performance of the propellant gas removal system was determined by firing the SPI with gas only into an empty torus where less than 0.1% of the gas was detected, but while the torus was being pumped. One of the key reasons for the KSTAR dual SPI installation is to investigate the performance of simultaneous injection of SPI pellets from ports on opposite sides of the machine. The identical SPI systems make this research possible as the systems have shown good synchronization between SPIs as shown in Fig. 2b where 7 mm D2 pellets fired from both SPIs arrive at their respective microwave cavities only 0.13 ms apart. Initial thermal mitigation experiments with neon-D2 mixtures have been performed with single and dual SPIs. These results show good assimilation of the pellet material into the plasma at the time of injection [4].
Installation and operation of these SPI systems has provided useful lessons learned in the implementation of this SPI technology and valuable experience in optimizing the performance of the formation and firing of the pellets. Experience on JET has shown a high reliability of pellet firing with only 1 failure to fire in over 150 attempts into plasmas since Sept. 2019. Punch operation with pure impurity pellets has also been especially reliable at firing pellets, but frequently results in broken pellets out of the injector .
[1] L. R. Baylor, et al., Nucl. Fusion 68 (2019) 211.
[2] S. Jachmich et al., to be submitted to 28th IAEA Fusion Energy Conference (2020).
[3] S.H. Park, et al., ISFNT 2019, Fus. Eng. Des. 154 (2020) 111535.
[4] J. Kim et al., to be submitted to 28th IAEA Fusion Energy Conference (2020).This work was supported by the US DOE under contracts DE-AC05-00OR22725 and DE-FC02-04ER54698 and by the ITER Organization (TA C18TD38FU) and carried out within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission or the ITER Organization. Also supported by Korean Ministry of Science and ICT under the KSTAR project.
*See the author list of E. Joffrin et al. accepted for publication in Nuclear Fusion Special issue 2019, https://doi.org/10.1088/1741-4326/ab2276Speaker: Larry R. Baylor (Oak Ridge National Laboratory) -
The mitigation of thermomechanical and runaway loads during disruptions and Vertical Displacement Events (VDEs) in ITER is essential for the project to execute the ITER Research Plan culminating (1) in the demonstration of the fusion power production goals (Q = 10 inductive operation for 300-500 s and Q = 5 for 1000 s and in steady-state up to 3000 s). To mitigate these loads ITER is equipped with a Disruption Mitigation System (DMS). The original baseline concept for was found insufficient to provide the required degree of mitigation following a review by experts from the ITER Members in 2017 and this lead to major change in the development of the DMS for ITER both in the concept, now based on a Shattered Pellet Injection (SPI), and the design approach, now led by the ITER Organization for integration and design and with support for ITER Members institutes for R&D organized under a Task Force to address open detailed design issues. This presentation reports on progress on the ITER DMS technical design and integration while other presentations at this conference address specific tokamak experiments and modelling activities to address open R&D design issues (2 , 3, 4).
The DMS is presently at the conceptual design level with its interfaces with other systems fixed at the preliminary design level not to affect the development of interfacing systems towards their final design or manufacturing level. The DMS utilizes toroidally distributed Shattered Pellet Injectors integrated in three equatorial and three upper port plugs (12 in EP02, 6 in EP08, 6 in EP17 and 1 each in UP02, UP08, UP14). While the injectors on the equatorial level are dedicated to thermal load mitigation, current quench control and RE prevention and RE energy dissipation, the purpose of the injectors on the upper ports are for late current quench mitigation. All DMS locations will be equipped with a dedicated gas supply providing the pellet material gases and high-pressure propellant gas, which is now integrated into the ITER design. An existing cryogenic supply system, as well as a gas venting system, are also incorporated in the ITER design to support the DMS needs in all required ports. In order to provide a DMS that performs with high reliability and availability, this first of a kind system has to fulfil key crucial requirements; these are the current design drivers of this system: a) defined and reproducible pellet integrity and pellet acceleration process; b) monitoring of pellet integrity and minimised interaction between the pellet and the flight tube wall; c) optimised shattering unit.
One of the main challenges for the DMS plant is the integration of the required components in a small and extremely harsh environment considering nuclear qualification and maintenance. The change of DMS concept following the 2017 review has lead, in addition, to integration challenges with diagnostic systems located in the same ports. To address these integration issues specific actions have been implemented: a) in the port plugs where the DMS are installed, pellet flight tubes’ openings have been optimised to ensure pellet survivability and to reduce neutron streaming and activation of equipment; b) in the port interspace (small area inside the bioshield), the DMS injectors have been designed to minimise the impact on the neighbouring diagnostics, while providing sufficient human access; c) in the port cell behind the bioshield extensive gas handling and cryogenic distribution units enable the control of all relevant process variables, in order to reliably operate the DMS; d) in equatorial port two (EP02) integration issues are being assessed with respect to the Core Imaging X-Ray Spectroscopy diagnostic. This diagnostic provides measurements of ion temperature and plasma rotation and its design has been substantially modified for compatibility with the two DMS systems sharing the port. The analysis of this design change shows impacts on its measurement capabilities regarding spatial/time resolution and for the measurable magnitude of the toroidal rotation, which are being minimized by design iterations. Other diagnostics systems sharing ports with DMs, such as the visible infrared viewing systems in the upper ports are not significantly affected.
A DMS Task Force has been created to support the DMS design and has two main activities, the design validation through experiments and modelling and the optimisation of the SPI technology (5). The technology programme addresses the main DMS design drivers through development and optimisation of key components to fulfil the ITER mitigation requirements. R&D under the technology programme covers issues such as : a) fundamental studies, including systematic tests and optimisation of the pellet formation and release process; b) the creation of a support laboratory, providing a test bed to assess the performance of key components such as the shattering bend; c) the development of pellet launch to optimize release mechanisms (fast valve and punch); d) the development of optical pellet diagnostics to diagnose pellet alignment, pellet integrity and pellet parameters.
The DMS TF work program on design validation consists of dedicated experiments accompanied by theory and modelling activities. These efforts are supported by significant contributions through the domestic activities of the ITER partners. The ITER DMS strategy relies on densification of the plasma to avoid runaway electron formation and on radiating most of the thermal and magnetic energy spatially as uniform as possible to avoid first wall melting. Densification requires the injection of multiple pellets, a scheme that is tested in JET (2), DIII D (6) and KSTAR (3), for the latter with an SPI system that can inject a total of four identical pellets from two toroidally opposite locations. These experiments, together with the tokamak J-TEXT, are providing information on size and energy scaling of SPI mitigation performance for model validation. While the technology programme assesses the parameter space for pellet fragmentation, the experiments planned for ASDEX-Upgrade will focus on finding the optimum fragment size for maximum material assimilation. In support of the experimental activities, the theory and modelling activities programme will provide physics-based extrapolation from the experimental results to ITER (4). This will allow narrowing down design parameters for the ITER DMS such as the pellet velocity, the fragment size distribution, and the pellet composition and it will increase the confidence level that the ITER DMS can fulfil its purpose.
The paper will described status of the DMS design, its integration into the ITER baseline and the status of the technology R&D and will provide a summary of progress of the design validation experimental and modelling activities.
References
(1) ITER Research Plan, ITR-18-03, ITER Organization, 2018.
(2) Jachmich, S., et al., this conference.
(3) Kim, J.-H., et al., this conference.
(4) Nardon, E., et al., this conference.
(5) Lehnen, M., et al., 2018 IAEA Fusion Energy Conf., Ahmedabad, India, pap. EX/P7-12.Speaker: Dr Timothy C. Luce (ITER Organization) -
The Final Design Review (FDR) of the ITER Plasma Control System (PCS) for First Plasma will be held in July 2020 following the conceptual and preliminary designs [1,2] to prepare for First Plasma operation scheduled for the end of 2025. ITER operation follows the Staged Approach of the ITER Research Plan (IRP) [3]. The main goals of the First Plasma campaign include achieving a plasma current Ip > 100 kA for a duration > 100 ms and possibly up to 1 MA for up to 3 seconds duration at half toroidal field of 2.65 T, followed by engineering operation to test the main superconducting coils up to full current. The First Plasma PCS design also includes the full architecture of the PCS that allows implementing high performance control algorithms in the future without changing the PCS architecture.
This phase includes control of the plasma current, radial plasma position, and plasma shape for a nearly circular plasma limited on either the inboard or outboard temporary stainless steel poloidal limiters (Figure 1). Disruption force limits of the attachments of these temporary limiters to the vacuum vessel require control algorithms to ensure that the plasma current remains < 1 MA at the nominal toroidal field of 2.65 T at R = 6.2 m. Following the first plasma attempts at 2.65 T, there will be an engineering operation phase to commission the superconducting central solenoid (CS), poloidal field (PF), and toroidal field (TF) coils to full current, up to 5.3 T nominal TF. If there are difficulties achieving first plasma at 2.65 T, the IRP and the PCS design foresees an option possibly to achieve first plasma at 5.3 T nominal TF. This would be easier than at 2.65 T because of the increased connection length to achieve plasma breakdown and the improved ionization and absorption efficiency of ECH power at the fundamental gyrotron frequency [4].
The plasma initiation phase also includes control of the gas injection of hydrogen to achieve and maintain the required prefill pressure in the range of 0.5 – 1 mPa and the initial density rise as well as the control of electron cyclotron heating (ECH) at 170 GHz. The low startup electric field ~0.3 V/m, large stray fields due to eddy currents in the vacuum vessel up to 1.5 MA, and the large volume and neutral fueling source limit the allowed prefill range and increase the required power for breakdown. The injected ECH power will be up to 5.8 MW for up to 300 ms from 7 of 8 available gyrotrons from one upper launcher with four beams crossing the breakdown region about 40 cm below the midplane and three beams crossing about 40 cm above the midplane (Figure 1). The beams will be reflected off mirrors mounted to the inner wall into an absorbing beam dump in an equatorial port to avoid scattering stray ECH power around the vessel that could damage in-vessel components. The absorbed power of a single pass is only expected to be of order a few percent of the injected power at half field. Feedback of the prefill pressure timed with the ECH pulses, the PF null, and vertical field swing will be used for robust plasma initiation control. This phase also includes initial exception handling algorithms for possible plant system and diagnostic faults, excess stray ECH power, plasma initiation failures, and start-up runaway electrons.
All control algorithms and synthetic diagnostic models for this phase, including magnetics, H, a single radial chord interferometer, and hard x ray monitor for runaway electrons have been developed in the Plasma Control System Simulation Platform (PCSSP) [5] in Matlab/Simulink for thorough testing in simulation prior to operation.
The PCS design is documented in the PCS database (PCSDB) using Enterprise Architect that records all requirements and tracks their compliance in the design [6]. The controllers are all documented in the PCSDB with explicit links between them and the PCS architecture including supervision, events, and exception handling, which allows checking the impact of any design changes on all aspects of the PCS design. The database includes performance requirements and the test results carried out to assess the design as well as the use cases and commissioning procedures.
Extensive simulations of plasma breakdown and First Plasma scenarios have been carried out with the TRANSMAK [7] and DINA [8] codes including vacuum vessel eddy currents and simplified plasma transport modeling of sputtered impurities from the stainless steel first plasma protection components. The simulations show that plasma initiation is sensitive to the impurity content, but, if breakdown is achieved, the plasma current will rise to ~ 0.5 MA and by switching the CS and PF voltages to zero, ensure Ip remains < 1 MA to protect vacuum vessel limiter housings.1 J. A. Snipes, et al., Fusion Eng. and Design 89 (2014) 507.
[2] J. A. Snipes et al., Nucl. Fus., 57 (2017) 125001.
[3] ITER Research Plan, ITER Technical Report ITR-18-03 (2018), https://www.iter.org/technical-reports?id=9 .
[4] P. C. de Vries and Y. Gribov, Nucl. Fus. 59 (2019) 096043.
[5] M. Walker, et al, Fusion Eng. and Design 96-97 (2015) 716.
[6] M. Cinque, at al., IEEE Trans. Plasma Sci., doi: 10.1109/TPS.2019.2945715 (2019).
[7] Mineev A.B. et al, 25th IAEA Fusion Energy Conf. (St. Petersburg, Russia, 2014) PPC/P3-20 (www-naweb.iaea.org/napc/physics/FEC/FEC2014/fec2014-preprints/255_PPCP320.pdf).
[8] Khayrutdinov R.R. and Lukash V.E. 1993 J. Comput. Phys. 109 193.
Speaker: Joseph Snipes (ITER Organization) -
The Chinese Helium Coolant Ceramic Breeder (HCCB) Test Blanket Module (TBM) and its ancillary systems (together called Test Blanket System or TBS) is one of important steps for the China magnetic confinement fusion development, which will contribute to validate the key tritium breeding blanket technologies under the burning plasma environment, including tritium extraction, heat removal, integration safety, etc. In the past two years, the design of HCCB TBS was continuously optimized based on the analysis and R&D results. Also the safety of HCCB TBS has been assessed to support the preliminary design.
As the core component, the HCCB TBM-set design was optimized, considering thermo-mechanical performance and manufacturability. The major change is the optimization of coolant flow diagram to avoid the local high temperature, though it is in the limit range. Also the purge gas flow path is adjusted to mitigate the tritium concentration at the corners of breeding zones. After the design optimization, the tritium performance of TBM still keeps at the same level as before, about 58mg/full power day. The thermo-hydraulic and thermo-mechanical performance has been checked, and the results show that the temperature and stresses are below the allowable value. Following the assembly scheme, the semi prototype of TBM has been fabricated and tested, which verified the manufacturability of the TBM design. The further optimization is ongoing taking into account the compressed pebble bed behaviors of functional materials and its neutronics impact. The integration scheme of the TBM shield is also updated considering the difficulty of the fabrication process, which will not change the structure design of TBM shield and the mockup fabrication is under plan. Considering the tritium release during the operation, one enclosure with the local detritiation system has been added for the helium cooling system (HCS) to reduce the tritium concentration in the room and the seismic analysis has been performed to verify this update. The design of the neutron activation system (NAS) has been also optimized according to the experiment and the investigation of key components. Besides, the I&C architecture has been developed to realize the conventional control, interlock and safety functions of ITER and HCCB TBS. The control by state machine requested by ITER for conventional control has been practiced and achieved in the helium testing loop (HeCEL-1).
ITER had been recognized as a basic nuclear installation (INB) by French nuclear safety authority ASN. The HCCB TBS system design should follow various nuclear safety requirements to control the radioactive risks from tritium and other neutron activation products produced during operation.The detailed nuclear analysis has been performed to identify source term inventories. The dose rate has been evaluated to demonstrate the radiation levels in different locations during ITER operation and maintenance phases which meet the ITER radiation zoning requirements. Multiple confinement barriers are designed to control tritium and mobile radioactive dusts, allowing the contamination level in each room to be compliance with ITER ventilation zoning. ALARA principle is being implemented in the design of HCCB TBS. Actions, including choices of advanced materials, improvement of shielding, reduction of activated materials and release of tritium, optimization of operation and maintenance activities, etc., have been considered to minimize the radioactive release and radioactive exposures to personals and public.
The tritium transport analysis is important to understand the tritium behavior in the HCCB TBS. After the code benchmark with other ITER members, a steady-state system level tritium transport for the updated HCCB TBS design has been performed. The amount of tritium transported to different parts of the system and released to rooms has been assessed. The results show that the tritium inventory is 0.39mg in the structure materials of tritium extraction system (TES) and 26.5mg in HCS structure materials. Also the integrated dynamic tritium transport analysis is ongoing, which will provide the dynamic changing of tritium concentrations in fluids, tritium inventory in solid materials, and tritium release towards outside based on the ITER operation scenario.
Several potential envelop accidents have been identified and analyzed, including loss of flow of cooling system; loss of coolant (break of TBM cooling channel or HCS pipes) in different locations, break of tritium systems, break of heat exchanger between HCS and ITER water cooling systems etc. During all accidents, because of the intrinsic safety characteristics of fusion device, after fusion plasma shutdown passively or actively, the decay heat in HCCB TBS will be removed through thermal radiation to large surface of ITER machines, without further over-temperature issues. The release of radioactive contents will be controlled by safety measures. No major safety consequences have been identified through accident assessments.According to the agreement between China and ITER, after operation in ITER, the irradiated HCCB TBM is expected to be transported back to China for further post irradiation tests (PIE) using special transport casks. Remaining rad-wastes from dismantling are assumed to be processed and disposed in French rad-waste facilities. Characteristics of these rad-waste and safety issues for handling them have been assessed.
Currently, the manufacture feasibility and safety of the HCCB TBS have been preliminarily verified by the R&D and safety assessment. Before the finalization of the design of HCCB TBS, the more detailed R&D and experiments will be performed to check the reliability and safety, also the compliance with the French regulations will be consulted with the agreed notified body (ANB).
Acknowledgments
This work was supported by the National Key R&D Program of China with grant numbers 2017YFE0300601, 2017YFE0300503 and 2017YFE0300502.Speaker: Xiaoyu WANG (Southwestern Institute of Physics)
-
-
-
-
-
Here we report recent progresses of laser fusion energy research in Japan, especially on the fast ignition scheme. For the fast-track to the laser fusion energy, we are investigating the fast-ignition plasma physics to realize optimal compression of a fusion fuel as well as efficient heating of the compressed fuel. In this scheme, we have demonstrated the efficient heating of high density plasmas with compression of the solid sphere by using a multi kJ lasers. The results imply the FI could have burning of a laser fusion plasma with a few 100kJ including 50-100kJ heating laser. Furthermore, we are developing a high repetition high power laser system: 10kJ/10-100Hz laser system to realize data driven analysis of fusion plasmas for the optimal target design as well as test of laser fusion reactor engineering with a fusion subcritical reactor including fusion power generation. These extensive studies are being made with under the fast ignition consortium and laser fusion strategy committee of IFE forum in Japan.
laser fusion plasma physics for the fast ignition
In the fast ignition (FI) scheme, an ultra-intense laser light generates high energy particles such as relativistic electron beam (REB) to heats a high density compressed fuel. The first demonstration of fast heating of laser compressed fuel1 has been made with a cone geometry for the efficient propagation of the heating laser pulse to generate the REB close to the high density core plasmas.
Details of the heating mechanisms has been investigated through experiments and simulations. In general, REB can heat the core plasma with 3 different mechanisms such as drag heating, return current heating and diffusive heating [2]. The drag and diffusive heating processes are important especially for the efficient heating by REB in the FI scheme. To improve the coupling of REB to the core plasma, the magnetized fast-isochoric heating (MFI) scheme had been designed and demonstrated. Delivering REB with a guidance of external magnetic fields at kT-level had been realized with the GEKKO XII (GXII) + LFEX, a ps/ kJ laser [3]. The REB, which has inherently a large divergence, was successfully guided along the magnetic field lines to the core, resulting twice higher laser-to-core coupling for the drag heating than that in a case without the guidance [4]. Another important heating mechanism in the FI is the diffusive heat driven by a strong radiation pressure of the heating pulse toward the core plasmas [5]. This scheme has been based on the physic of a relativistic laser -plasma interaction and becomes significant with a multi-picosecond kJ laser. The diffusive heating is the main heating process in the current experiment with relatively low density core plasmas, e.g. 0.1g/cm2 [6], and the drag heating will be comparably important in the ignition-scale core (0.3 -0.5 g/cm2).
Improving the coupling of the heating laser to the appropriate energy of REB is also crucial for the efficient heating. One is focusing of multiple coherent laser beamlets spatially and temporally overlapped, thus producing an interference pattern in the laser focus, has significantly improved the laser energy conversion efficiency into the REB, compared to one laser beam with the same energy as the four beamlets combined [7]. Another idea to the higher coupling efficiency is utilization of shorter wavelength laser light to generate relatively lower energy REB, which then has a shorter stopping power. A 527nm laser light will make expectation for future realization of fusion gain of more than 10 with a heating laser energy of 70kJ for the DT fuel with ρR=2 g/cm2 and 500 g/cm3. Furthermore, higher efficiency could be expected with fundamental (1053 nm) and second harmonics (527nm) mixed wavelength laser lights.
Optimal fuel compression for the FI is critical to realize the efficient ignition of the laser fusion. The FI scheme can optimize separately the processes of the fuel compression and the heating of the compressed fuel. We find that a solid spherical fuel target will be more feasible, to stably obtain high-density fuel compression without a risk of the hydrodynamic instability. Moreover, the solid sphere target has an advantage in more simple fabrication process, as compared with the shell target, especially for the cryogenic fuel fabrication. High density compression of the solid sphere attached a Au cone target, which could be hydro dynamically stable, had been experimentally demonstrated to be ρR~0.1 g/cm2[8]. Based on 2D simulation results, about 130kJ would be required to obtained ρRmax =1.0 g cm−2 using a solid sphere and multi-step laser pulse [9]. More efficient coupling > 20% would be expected through the optimal heating processes to initiate ignition and burning of the compressed fuel with 50-100kJ heating laser energies [10].
High repetition laser system development and fusion energy reactor engineering
We are developing a 100J/100Hz high power lasers with an active mirror amplification scheme using 10 cm Yb:YAG ceramics pumped by laser diodes. Based on this system, we are now proposing a new type of high power laser system as shown Fig. 1 to explore a variety of new fields of sciences, which is called “J-EPoCH”. This facility integrates all the state-of-arts high power laser technologies, based on the 160 beams of 100Hz /100J laser module, providing high repetition 10kJ long pulse lasers, 5-20PW short pulse lasers and different kinds of laser plasma accelerators, and laser-driven radiation sources such as x-rays and neutrons.
This multi-purpose high repetition laser system will also open a new frontier of laser fusion energy development. The laser fusion strategy committee of IFE forum in Japan organized around 40 members from 20 institutes is considering the roadmap utilizing this high repetition laser system toward laser fusion reactor. The high repetition system will realize data-driven analysis of fusion plasma for the optimal target design as well as testbed of laser fusion reactor engineering with a fusion subcritical reactor [11]. This fusion subcritical reactor enables early implementation of all laser fusion engineering assessments such as wall materials, optical elements, pellet injection and power generation, e.g. fusion electric power of a few Watt using a large aspect shell target under the 1 Hz laser operation.1 R. Kodama et al., Nature 412, 798 (2001); Nature 418, (2002) 933
[2] Y. Sentoku at al., Physics of Plasmas 14, (2007) 122701
[3] S. Fujioka et al., Scientific Reports 3, (2013) 1170
[4] S. Sakata et al., Nature Commun. 9, (2018) 3937
[5] N. Iwata et al., Natture Commun. 9, (2018) 623
[6] K. Matsuo et al., Phys. Rev. Lett. 124, (2020) 035001
[7] A. Morace et al, Nature Commun. 10, (2019) 2995
[8] H. Sawada et al., Appl. Phys. Lett. 108, (2016) 254101
[9] H. Nagatomo et al., Nuclear Fusion 59, (2019) 106055
[10] T. Johzaki et al., Physics of Plasmas 15, (2008) 062702
[11] A. Iwamoto et al., 11th Int. Conf. on Inertial Fusion Sci. and Appl. 2P01-04 (Sep.23 2019)Speaker: Prof. Ryosuke Kodama (Institute of Laser Engineering, Osaka University) -
In the laser-driven indirect drive scheme for inertial confinement fusion (ICF), the capsule diameter is typically limited to ~2 mm in order to attain quasi-round implosions with currently available laser energy in cylindrical hohlraums. This geometrical factor restricts the energy coupling efficiency from the hohlraum to the capsule to be ~10% [Ref.1]. We report the first series of experiments on NIF demonstrating ~ 30% energy coupling to an Al capsule in a rugby-shaped Au hohlraum [Ref.2]. 3.0mm- and 3.4mm-diameter Al capsules are driven in Au rugby hohlraums at 1MJ and 1.5MJ laser energy, respectively. Measurements of in-flight capsule size, mass remaining, velocity, bang time and neutron yield show good agreement with simulations, consistent with ~300kJ coupling at 1MJ drive and ~500 kJ coupling at 1.5MJ drive. It is found that the shell shape during the implosion is sensitive to the rugby dimensions so that the rugby shape is an effective knob for symmetry tuning. A round imploded shell has been achieved near the bang time for the 3.4mm capsule.
In comparison to a conventional cylindrical hohlraum that has a straight wall, the curved wall of a rugby hohlraum [Ref.3,4] affects the laser irradiation mainly in two aspects: enlarging the laser spot size and enhancing the specular reflection of the laser beams. This results in ~1.4-1.8x larger laser spot and lower intensity of the beams, which is helpful to reduce beam blocking due to the wall bubble expansion. The larger incident angle also leads to higher reflection of the beams, which points toward inside of the rugby hohlraum and is helpful to increase the drive at the waist. These beneficial effects and the simulation setup have been discussed in detail in [Ref.5].
The experiments were preformed using a standard 1D and 2D x-ray radiography platform on NIF. The setup is illustrated in Fig. 1 (a). A Zr foil located at 12 mm from the target center was irradiated by 8 NIF beams to generate a 16 keV backlighter. Measurements of the velocity and mass density profile by the streaked radiography provided a shell kinetic energy of 34 kJ with ~1 MJ laser drive, which is 2-4x that which was achieved in the recent high-foot shots with a ~2 MJ laser drive. Given that the typical rocket efficiency for imploding Al ablator is ~ 10%, 34 kJ shell kinetic energy corresponds to > 300 kJ coupled to the capsule. Fig. 1 (b) shows the simulated energy absorption by the capsule at three scales, 0.7x, 0.9x and 1.0x. The energy coupled to the capsule reaches 350, 500, and 650 kJ with 1.0, 1.5 and 2.0 MJ drive, respectively. Good agreement between the simulated and measured quantities including peak radiation temperature, in-flight radius, velocity, shell kinetic energy and shell width supports the high energy coupling at 0.7x scale [Ref.2].
![(a) Experimental setup (from [Ref.2]). (b) Simulated absorbed energy by the capsule vs time at 3 scales.](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/1190/fig1.jpg)
The sensitivity of the laser energy distribution to the incident angle makes it possible to tune the implosion symmetry by adjusting the rugby wall shape. We have carried 3 NIF shots with different rugby dimensions using a 4-5 ns long reverse-ramp pulse shape with peak power 300-400 TW. The x-ray radiographs are shown in Fig. 2. The wide rugby in Fig. 2 (a) produced a quite prolate shell shape, indicating more drive at the waist than at the laser entrance hole. A narrower rugby with smaller waist diameter in Fig. 2 (b) reduced asphericity. Finally a scale-up of the narrow rugby produced a very round implosion as shown in Fig. 3 (c). This symmetry tunability by rugby wall shape will be very useful for the design of future campaigns.
Nuclear diagnostics were enabled in another shot at 0.9x scale with 7 mg/cc DT gas fill in the Al capsule, providing additional constraints on the energy coupling to the capsule. The measured nuclear burn history is in good agreement with simulated value as shown in Fig. 2(d). The yield reaches 78% of pre-shot prediction. The good agreement between multiple measurements and simulations indicates a coupling energy of ~ 500 kJ in the 1.5MJ drive experiments [Ref.6].
To summarize, we have performed measurements of symmetry, nuclear bang time, neutron yield, in-flight capsule size, and velocity of Al capsules with diameter 3.0-3.4mm in Au rugby hohlraum. The good agreement between data and simulations indicates 500 kJ coupled to the capsule with 1.5 MJ drive. It is demonstrated that the implosion symmetry can be tuned effectively by adjusting the rugby shape. These results open new opportunities for both the mainline single-shell scheme and the double-shell designs toward ignition in ICF, which is a critical step for the development of IFE.
[Ref.1] R. Betti and O. A. Hurricane, Nature Phys. 12, 435 (2016).
[Ref.2] Y. Ping, V. A. Smalyuk, P. Amendt, R. Tommasini, J. E. Field, S. Khan, D. Bennett, E. Dewald, F. Graziani, S. Johnson, O. L. Landen, A. G. MacPhee, A. Nikroo, J. Pino, S. Prisbrey, J. Ralph, R. Seugling, D. Strozzi, R. E. Tipton, Y. M. Wang, E. Loomis, E. Merritt, and D. Montgomery, Nature Phys. 15 (2019) 138.
[Ref.3] P. A. Amendt, C. Cerjan, A. Hamza, D. E. Hinkel, J. L. Milovich and H. F. Robey, Phys. Plasmas 14 (2007) 056312 ; P. A. Amendt, C. Cerjan, D. E. Hinkel, J. L. Milovich, H. S. Park and H. F. Robey, Phys. Plasmas 15 (2008) 012702.
[Ref.4] M. Vandenboomgaerde, J. Bastian, A. Casner, D. Galmiche, J.-P. Jadaud, S. Laffite, S. Liberatore, G. Malinie, and F. Philippe, Phys. Rev. Lett. 99 (2007) 065004.
[Ref.5] P. Amendt, D. Ho, Y. Ping, V. Smalyuk, S. Khan, J. Lindl, D.Strozzi, R. Tommasini, M. Belyaev, C. Cerjan, O. Jones, W. Kruer, N. Meezan, H. Robey, F. Tsung, C. Weber, and C. Young, Phys. Plasmas 26 (2019) 082707.
[Ref.6] Y. Ping, V. A. Smalyuk, P. Amendt, et al. submitted to High Energy Density Physics for IFSA conference proceedings.Speaker: Y. Ping (LLNL) -
A promising repeatable laser system producing multi-kilojoule of pulse energy has been basically designed for realization of the fast-ignition-based inertial fusion energy (IFE) reactor. Two cultivated core key technologies ensure high reliability of the proposed design. First, using our novel bonding technology, a cryogenic active-mirror amplifier has been developed to enable 100 Hz repeatable operation at 100 J of pulse energy. Kilojoule energy can be easily obtained for the fuel compression by combining a number of the amplifiers into a single beam. The other is the high-contrast, petawatt laser technology for heating laser: LFEX-laser demonstrated 2 PW, 2 kJ, 1 ps with a high intensity contrast more than 1 x 10^10.
Understanding of the central ignition scheme of IFE reveals significant difficulties of the ignition achievement through many experiments and theoretical calculations by using a number of kilo-joule class beam lines at the huge facilities in National Ignition Facility (LLNL) and OMEGA facility (LLE, Univ. Rochester) and so on. Recently, at Osaka university, the highly efficient production of an ultra-high-energy-density plasma of 2.2 Peta-Pascal has been successfully demonstrated with the fast ignition scheme by using the 2 PW LFEX-laser.[A] This remarkable progress of the fast ignition research accelerates realization of a practical IFE reactor based on fast ignition scheme. However, a repeatable high power laser system, which is an absolutely essential technology of the IFE reactor, is still missing. In a latest fusion reactor design, as multiple moderate-in-size reactors operate together, the required repetition rate for the laser system comes up to 100 Hz. So far, all high-energy laser systems used for the IFE studies, including the above lasers, are a single shot laser, and there is no repeatable laser system in kilo-joule class over the world. Serious bottlenecks for repeatable laser operation are thermal effects caused by large heat loaded in the laser material of the energy amplifier, wave front distortion, induced birefringence, and material fracture. The first application of ceramic active-mirror to high energy fusion laser as a laser amplifier reduces these effects dramatically to enable repeatable operation.
The active-mirror laser amplifier shows, in principal, two excellent capabilities of a high heat removal due to conductive cooling with a metal heat sink and a low wave front distortion due to the heat flow direction parallel to the laser propagation, shown in Fig. 1. In addition, a ytterbium-doped YAG (Yb:YAG) ceramic is used as a laser material instead of the conventional neodymium-doped laser glass (Nd:glass). The high thermal conductivity of the YAG ceramic is about one-order of magnitude higher than that of Nd:glass, this high thermal conductivity reduces the material temperature rise. Moreover, the Yb:YAG ceramic is cryogenically cooled to increase the thermal conductivity further. A cryogenic active-mirror has been already demonstrated at high repetition rate of a few hundred hertz, the aperture size was, however, as small as 1 cm class and the pulse energy was around 1 J.[B,C] A thermally strong bonding of the ceramic to the metal heat sink is considerably difficult. And to make matters worse, as the cryogenic operation temperature is quite different from the bonding temperature, an intense internal stress of the ceramic is not avoidable in several-centimeter aperture size for more than 10 J due to a different thermal expansion between the ceramic and the metal. By developing a novel bonding technology, which is under patent pending, to relax the stress, 7 cm x 7 cm active-mirror has been successfully bonded and 10 J high pulse energy has been obtained at 10 Hz with the four diode-pumped cryogenic active-mirrors, for the first time, in Fig. 2. A higher repetition rate of 100 Hz will be demonstrated and a higher pulse energy of 100 J will be upgraded by enlarging the beam aperture size in the near future.
Another core key technology is the petawatt laser technology. The kilojoules chirped-pulse amplification (CPA) has been already developed in our LFEX-laser at 2 PW, 2 kJ, 1 ps to supply pulses for the plasma experiments. An excellent intensity contrast of more than 1 x 10^10 is obtained with the original pulse cleaner in Fig. 3. The large aperture essential optics used in CPA such as a dielectric grating (92 cm x 42 cm, 1740 grooves/mm) has been already developed, see Fig. 4.
A promising repeatable laser system at multi-kilojoule, named “J-EPoCH”,[D] has been basically designed for various high-energy-density plasma applications, and roughly consists of two laser systems of a nano-seconds power laser and a petawatt laser. The power laser is a 16 kJ diode-pumped solid-state laser system with 160 beam lines, each of which has a 100 J, 100 Hz active-mirror laser module. The petawatt laser is a Titan-doped sapphire (Ti:sapphire) laser system. Using about half beamlines of the power laser system as a pump source after second harmonic generation, kilo-joule energy is obtained in the femtosecond to picosecond time domain. Extremely higher thermal conductivity of cryogenic Ti:sapphire, compared with cryogenic Yb:YAG, is plenty of thermal strength for 100 Hz operation. The appropriate combination of beam lines from the both systems will help the first power generation demonstration in the fast-ignition-based IFE reactor.[E][A] K. Matsuo et al., Phys. Rev. Lett. 124, pp. 035001-1 - 035001-5(January 2020).
[B] M. Divoky et al., Opt. Lett. 40, pp. 855- 858 (March 2015).
[C] C. Baumgarten et al., Opt. Lett. 41, pp. 3339-3342 (July 2016).
[D] R. Kodama, AAPPS-DPP2018, P9 in Kanazawa, Japan (12-17, Nov. 2018).
[E] A. Iwamoto and R. Kodama, IFSA2019, 2P01 in Osaka, Japan (23 September 2019).Speaker: Prof. Junji Kawanaka (Institute of Laser Engineering, Osaka University) -
Inertial Confinement Fusion (ICF) schemes are designed to heat and compress DT fuel to conditions exceeding the Lawson criterion ($p \tau$) using implosion, which greatly amplifies the pressure of a driver (~100 MBar) to the conditions necessary for laboratory-scale ICF (~100s GBar). The National Ignition Facility (NIF) focuses on the laser indirect drive approach to ICF, in which laser energy is converted to x-ray radiation in a `hohlraum’, which drives the fuel-containing capsule$^1$. This process is inefficient, with ~10% coupling efficiency from the laser energy to energy absorbed by the capsule typical. Of the energy absorbed the capsule material, only ~10% is converted into kinetic energy of the imploding fuel and internal energy of the fuel at stagnation. One focus of the program is to improve this coupling efficiency and enable larger implosions within the current capabilities of the NIF laser; this is a route towards increasing $p \tau$ on NIF and is relevant for future ignition experiments or approaches towards inertial fusion energy since the coupling efficiency feeds into requirements for driver size and target gain. In parallel, several degradation mechanisms have been identified that impact implosions on NIF, notably low-mode drive asymmetry and mix induced by target defects or engineering features. Detailed studies of these mechanisms have been conducted to identify routes towards improved performance.
In 2017-2018, record fusion yields on NIF were produced with designs that utilized high-density-carbon (HDC) capsules and low-gas-fill hohlraums$^2$. These previous campaigns explored studies over several parameters, with the combination of data and theoretical scaling arguments suggesting that increasing the capsule size could be a favorable tactic for further improving performance$^3$, in part from the increased coupling efficiency from the driver to the fuel. With a constant available laser energy and power, a recent campaign pursued an increased capsule size in comparable hohlraums, substantially reducing the case-to-capsule ratio (CCR), from ~3 to ~2.7. A schematic of the target, compared to a previous campaign$^4$, is shown at the left of Fig. 1. Empirical metrics for the implosion symmetry$^5$ and initial data demonstrate that control over the symmetry is needed to prevent highly distorted implosions at this CCR. We use wavelength detuning ($\Delta \lambda$) between inner beams, which drive the hohlaum waist, and outer beams, which drive near the poles, to semi-empirically adjust the amount of cross-beam energy transfer (CBET) between these beams, which provides control over the implosion symmetry$^6$. This wavelength detuning can be adjusted on every shot to control the shape. Example data are shown on the right of Fig. 1, demonstrating that the application of 1Å of $\Delta \lambda$ changes the implosion shape from oblate to prolate. Unlike previous campaigns that employed wavelength detuning in high-gas-fill hohlraums$^7$, our data demonstrate that the shape is symmetric throughout all stages of the implosion.
With control over the shape we have conducted an initial series of three cryogenically layered DT fueled shots to assess the integrated implosion performance. The first two implosions revealed a higher than expected level of high-Z ablator material mixing into the fuel. This was successfully mitigated with two techniques: first, by increasing the fuel thickness to provide an additional buffer against mix, and second by using capsules with improved quality, or fewer seeds for deleterious hydrodynamic instabilities. These changes result in an implosion with record values for NIF for the capsule absorbed energy, fuel kinetic energy, and hot-spot internal energy, shown in Fig. 2. The highest performing shots in this campaign (N191007 and N191110) are denoted and have record values for coupled energy. Notably, these implosions have achieved record fuel energy at very modest values of other design parameters, especially the velocity. Simple scaling relations$^3$ expect the fusion performance to increase strongly as a function of both scale factor ($S$) and velocity ($v$) as approximately $Y \propto v^{7.7} S^{4.4}$. Increasing the velocity of these implosions is therefore a clear direction for future exploration.
In parallel with the effort to increase energy coupling to the capsule and fuel, understanding known degradation mechanisms has been a dedicated effort of the program. Several sources of unintentional directional mode-1 drive asymmetry have been identified, including diagnostic windows in the target, random variation in the laser delivery, and anisotropic capsule thicknesses. Reducing these sources of asymmetry is expected to negate the impact of deleterious mode-1 drive asymmetry on current implosions. Similarly, several sources of high-Z material mixed into the fuel have been identified. First, engineering features such as the membrane that holds the capsule within the hohlraum and the tube used to introduce fuel into the capsule are sources of material mix into the hot spot. Second, defects introduced during the capsule manufacturing process, including ‘pits’ on the surface and ‘voids’ within the material, can cause high levels of mix into the fuel. Mitigation mechanisms for these varied sources of deleterious mix are being pursued. The degradations, when mitigated, are expected to both improve performance of current implosions and enable experiments in more aggressive parameter space.
In summary, we have conducted a campaign to improve the energy coupling efficiency for NIF implosions, by fielding a larger capsule to absorb more energy from the x-ray producing hohlraum, and have achieved record fuel kinetic energy and hot-spot internal energy with this approach. Increasing the coupling efficiency from the laser drive to the fuel is advantageous for improving performance on NIF and for projections to inertial approaches to fusion energy. In parallel, several sources of degradation mechanisms have been studied, with the causes identified and mitigation techniques in development to enable higher performing implosions.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
1: J. Lindl, Phys. Plasmas 2, 3933 (1995)
2: S. Le Pape et al., Phys. Rev. Lett. 120, 245003 (2018); K. Baker et al., Phys. Rev. Lett. 121, 135001 (2018)
3: O.A. Hurricane et al., Phys. Plasmas 26, 052704 (2019)
4: A.L. Kritcher et al., “Symmetric fielding of the largest diamond capsule implosions on the NIF”, submitted to Phys. Plasmas (2020)
5: D.A. Callahan et al., Phys. Plasmas 25, 056306 (2018)
6: A.L. Kritcher et al., Phys. Rev. E 98, 053206 (2018)
7: O.A. Hurricane, Nature Physics 12, 800 (2016)Speaker: Alex Zylstra (Lawrence Livermore National Laboratory) -
TAE Technologies, Inc. (TAE) is a privately funded company pursuing an alternative approach to magnetically confined fusion, which relies on field-reversed configuration (FRC) plasmas composed of mostly energetic and well-confined particles by means of a state-of-the-art tunable energy neutral-beam (NB) injector system. TAE’s current experimental device, C-2W (also called “Norman”) shown in Fig. 1, is the world’s largest compact toroid (CT) device [1] and has recently made significant progress in FRC performance, producing record breaking, high temperature advanced beam-driven FRC plasmas, dominated by injected fast particles and sustained in steady state for up to 30 ms, which is limited by NB pulse duration as can be seen in Fig. 2. C-2W has been producing significantly better FRC performance than the preceding C-2U experiment [2], in part due to Google’s machine learning framework for experiment optimization, which has contributed to the discovery of a new operational regime where novel settings for the formation sections yield consistently reproducible, hot, stable plasmas.
In order to produce such high performance FRC plasmas, C-2W operates with the following key features [1]: reliable dynamic FRC formation scheme via colliding and merging two oppositely-directed CT plasmas (relative collision speed up to ~1000 km/s); tangential, co-current NB injection (NBI) into the FRC with high input power (total up to ~21 MW) and intra-discharge variable energy (15–40 keV) functionality; flexible edge-biasing electrode systems for stability control in both inner and outer divertors; neutral gas density control via ~2000 ${m^3/s}$ pumping capability in the divertors; external magnetic field fast control capabilities, such as field ramp, and active feedback control of the FRC plasma using Trim and Saddle coils; and 50+ dedicated plasma diagnostics in the main confinement region and in divertors to characterize FRC and open-field-line plasma performance.
In the recent C-2W experiments, adequately controlled external magnetic-field profile throughout the machine and proper gas injection/fueling have led to more effective edge biasing from electrodes to globally stabilize plasma; thus, improving the efficiency of the NB-to-FRC coupling so that more plasma heating and current drive are obtained. Due to this synergistic effect of combining effective edge biasing and NBI on C-2W, beam-driven FRCs have achieved a high temperature regime as shown in Fig. 2; total temperature ${T_{tot}}$ >2 keV (${T_{tot}}$ = ${T_e+T_i}$, based on a pressure balance), electron temperature ${T_e}$ >250 eV. Other typical plasma parameters are: averaged electron density <${n_e}$> ~${1–3×10^{19} m^{-3}}$, trapped magnetic flux (based on rigid-rotor model) ${\phi_p}$ ~5–10 mWb, and external axial magnetic field ${B_e}$ ~1 kG. To date, the optimum C-2W discharges have reached ${T_{tot}}$ up to ~3 keV and ${T_e}$ ~300 eV at the peak inside the FRC.
Based on a careful global power balance analysis detailing input/loss channel characteristics and plasma timescales [2], there appears to be a strong positive correlation between ${T_e}$ and energy confinement time. The previously reported C-2/2U scaling of the electron energy confinement time ${\tau}$${_{E,e}}$ still persists at the higher ${T_e}$ (i.e. collisionless plasma regime) in C-2W, as shown in Fig. 3. Given uncertainties in the measurements and assuming the power-law model, regression analysis shows that ${\tau}$${_{E,e}}$ is approximately proportional to ${T_e^2}$ when fitting for the entire ensemble of the C-2W data set.
Dedicated equilibrium and transport simulations have been performed to better understand FRC global stability and confinement, NBI and edge biasing effects, and turbulence in the FRC core and open-field-line plasmas. Simulations predict that parallel electron heat loss is close to the minimal theoretical limit, which has been experimentally validated by end-loss energy analyzers in the outer divertor. Nonlinear kinetic simulations also qualitatively agree with experimental fluctuation measurements, where turbulent transport is greatly reduced by sheared flows due to edge biasing. Google also contributes to advanced data analysis where their Bayesian inversion algorithm reconstructs plasma density profiles and high-frequency fluctuations. Disruptions of fast-ion orbits can be studied from plasma displacement inferred from reconstructions, where detailed correlations with magnetic probes provide further insights into energy loss mechanisms. This paper will review the highlights of C-2W program, including recent experimental results of significantly advanced FRC performance as well as simulations. Future plans will be reported as well.
[1] H. Gota et al., Nucl. Fusion 59, 112009 (2019).
[2] H. Gota et al., Nucl. Fusion 57, 116021 (2017).
![FIG. 2. Time evolutions of excluded-flux radius, electron temperature, total temperature, and thermal energy of FRCs for shots 113301 and 114534 in C-2W, compared with shot 43833 in C-2U [2].](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/749/2.jpg)
![FIG. 3. Confinement scaling: global energy confinement time as a function of electron temperature in C-2/2U [2] and C-2W.](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/749/3.jpg)
Speaker: Dr Hiroshi Gota (TAE Technologies, Inc.)
-
-
-
The commissioning of the power supplies for superconducting coils in JT-60SA has been done with dummy load (Inductance: 7.64 mH, Resistance: 6.995 m$\Omega$) since June 2019. The most important results are that (i) Integrated operation of the different PS components was completed successfully, (ii) High voltage generation of the rated voltage of 5 kV by Switching Network Unit (SNU) was performed properly and (iii) DC current interruption with rated current of 20 kA by Quench Protection Circuit (QPC) was achieved. These results are a prerequisite to achieve the integrated power supply operation with superconducting coils in JT-60SA and, in a next stage, contribute to the power supply commissioning of ITER and DEMO.
The detailed design of power supplies to provide the needed DC power to Toroidal Field (TF) and Poloidal Field (PF) superconducting coils for JT-60SA was started from 2011 and the installation and individual test for each power supply component has been completed by 2018, with a strong collaboration between Japan and EU. Each PF circuit consists mainly of three power supply components: the “Base PS” ac/dc thyristor converters for steady-state operation with low voltage (~ 1 kV) and high current (20 kA) ratings, the SNU and the QPC. In addition, the PS system includes the Motor Generator and a set of Vacuum Circuit Breakers providing the ac high voltage to the Base PS, the low voltage auxiliaries and the water cooling system. It is essential to verify the performance of the overall system using dummy load before real plasma experiment with superconducting coils in JT-60SA. To this aim, the commissioning for each PF circuit was performed by a dedicated supervisory control system which was newly developed for JT-60SA to achieve high-speed (4kHz) and real-time control of power supplies [A].
The main goal of these combined tests was to verify that the different power supply systems, which were individually tested, could work together in a coordinate way, following the time sequence, the commands and the references generated by the supervisory control system, both in normal operation and in case of fault, activating the correct and coordinate protection sequence when required. During the performance of the combined tests, it has been found that the implementation of the time sequence was not homogeneously programmed in all devices, causing unexpected operations, and it has been necessary to modify the firmware of some components. Thanks to these improvements, it has been possible to perform a complete 20 kA pulse on a PF circuit connected to the dummy load, with the coil current shape provided by the Base PS according to the assumed plasma scenario, and with the correct operation of the SNU and the QPC.
Figure 1 show current and voltage shapes indicating SNU operation at t=0s and also QPC operation by a simulated quench signal.
The SNU [B] is necessary to generate the high voltage (5 kV) required for plasma ignition. This is obtained by pre-charging the coil up to the nominal current of 20 kA, then inserting in the circuit a resistor with pre-settable value, obtaining a fast current variation, and finally by-passing the resistor when the plasma ignition is obtained. This function is managed by the operation of a hybrid switch for DC current interruption and requires a close coordination with the operation of the Base PS.
The QPC [C] has the function to protect the superconducting coil by dissipating the large energy stored in the coil inserting in the circuit a dump resistor in the case of coil quench or other faults requiring a fast discharge, with the requirement to complete the 20 kA DC current interruption in 350ms. The QPC main components are a hybrid switch for DC current interruption, a pyrobreaker (explosive fuse) for backup, and a dump resistor for coil energy consumption.
It was observed that the SNU can interrupt circuit current of 20 kA to commutate its current to the SNU resistor and can generate 5 kV against the coil in plasma ignition phase. Moreover, it was also observed that the QPC can interrupt the circuit current properly by the simulated quench signal. This result is that the system combined with the Base PS, the SNU and the QPC can operate properly.In addition, during the combined tests it has been possible to verify the correct operation of the SNU and the QPC with nominal current provided by the Base PS, reaching their maximum rated voltage. These tests were already performed in factory, but not yet repeated after the on-site installation, due to the unavailability of a PS rated for 20 kA at the time of the individual commissioning. Indeed, some problems due to a wrong connection of a voltage transducer in the QPC have been detected when operating at current higher than 10 kA and solved by modifying the connection layout.
[A] S. Hatakeyama, et. al., Fusion Eng. Des. 146 (2019) 1652
[B] A. Lampasi, et.al., Energies 11 (2018) 996
[C] E. Gaio, et. al., Nucl. Fusion 58 (2018) 075001Speaker: Katsuhiro Shimada (National Institutes for Quantum and Radiological Science and Technology) -
The JT-60SA tokamak is a large-scale fusion installation located in Naka (Japan) to be operated as a satellite for the ITER tokamak and also to explore DEMO plasma regimes 1. JT 60SA first plasma will be run end 2020 and will represent the energization of the worldwide largest superconducting cryomagnetic system in operation for fusion. The magnet system is composed of 18 Toroidal Field (TF) coils, 6 Equilibrium Field coils and the Central Solenoid. In total 20 TF coils were procured by Alstom-GE (France) and ASG (Italy) and were extensively tested in Cold Tests Facility (CTF) in CEA Saclay (France). All TF coils were tested along a given acceptance procedure, and furthermore, among the TF coils, two of them, bound to be used as spare coils (TFC02 and TFC19) were tested within a specific test program, called Advanced Tests Activity (ATA) aiming at reinforcing the knowledge on TF coils behavior by expanding the experimental database. For these specific tests, more instrumentation (temperature sensors) was installed on TFC02 to get a better accuracy on the behavior of each hydraulic unit length of the coil windings. ATA tests were conducted with a wide range of various parameters (current, temperature rate, etc…) in order to quantify the influence of each of them on the coil performances, the overall aim of the ATA program being to provide experts with an extended database to allow consolidated analyses for predictions for future tokamak operation.
On the other hand, in the aim of consolidating its modelling capacities, CEA team developed several codes to address the prediction of JT-60SA magnet performances in operation. A quasi-3D simulation code was built [2], coupling THEA (thermohydraulics) and Cast3M (thermal) and besides a multiphysic platform OLYMPE [3] was also developed to allow, among other goals, parametric studies to be led in an efficient way.
In the paper we present in a first part, an extensive study, where the ATA tests carried out on the TFC02 coil are tentatively represented with the coupled tool THEA-Cast3M inserted in OLYMPE structure. In this goal, several operational parameters are varied to match the whole experimental data. For innovative purpose, we chose to consider in our studies, three parameters which are usually not explored in the modelling approaches:
- The quality of contact between the casing and the winding pack (WP). Indeed the casing is heated by external thermal loads and will transfer its heat to the conductor in the WP, this transfer being driven by the quality of contact between the two pieces. An example of a configuration is visible in Fig.1 showing the location of the contact faces considered in the study. For instance those contacts can intrinsically vary along WP perimeter and also in time between the operating phases: steady state (no WP deformation) and current loads (WP asymmetric deformation).
- The cooling channels (CC) efficiency, the CCs being located in two zones of the casing and expected to divert some of the casing heat, mitigating the heat flux to WP. They are also shown in Fig.1. The CC properties can vary along their thermal contact quality with the casing.
- The conductor winding hydraulic properties (see Fig. 1), that will drive the coolant (helium) mass flow and therefore its heat exhaust capacities. The hydraulic properties have a statistical scattering and therefore should be included in the sensitivity approach regarding impacting factors for the helium temperature.
We will show the different optimal configurations, matching with the experimental data (that can be seen with Fig. 1, with the pancakes outlet temperatures variations in TFC02).
The most appropriate model setting will be exposed and justified.
In a second part, the simulation tools optimally set is run with the TF model in tokamak configuration in a coupled configuration with the cryoplant model represented with Simcryogenics code [4][5].
Several operation cases are analyzed and commented with the nominal operation parameters and some boundary cases. A statistical approach is also conducted using the strands performance database and results in a virtual coils performance database in operation conditions that shows a probability curve of the TF coils performances in normal operation. This database strands as a ground for a future comparison with the experimental data expected to be issued from the JT-60SA tokamak commissioning.1 P. Barabaschi, Y. Kamada, H. Shirai and the JT-60SA Integrated Project Team, “Progress of the JT-60SA project”, Nucl. Fus. Vol. 59 n°11 (2019).
[2] Q. Le Coz et al., “Quench simulation of a DEMO TF coil using a quasi-3D coupling tools”, IEEE Transactions on Applied Superconductivity Vol. 28 n°3, Art. n° 4203105 (2018)
[3] L. Zani et al., “OLYMPE, a multi-physic platform for fusion magnet design: development status and first applications”, presented at CHATS conference, submitted to Cryogenic (2020).
[4] S. Varin et al., “An update of dynamic thermal-hydraulic simulations of the JT-60SA cryogenic system for preparing plasma operation”, presented at CHATS conference, submitted to Cryogenic (2020).
[5] Bonne et al., “Simcryogenics: a Library to Simulate and Optimize Cryoplant and Cryodistribution Dynamics”, presented at CEC / ICMC 2019, submitted to Mat. Science and Engineering, Advances in Cryogenic Engineering (2020).Speaker: Louis ZANI (CEA-IRFM) -
The SST-1 machine is the first medium size superconducting tokamak operational at Institute for Plasma Research India. It comprises of set of toroidal (TF) as well as poloidal field (PF) superconducting coils system. In order to cool and maintain these magnets under superconducting state, a dedicated specially designed helium cryogenic system of cold capacity 1350 W at 4.5 K along with its auxiliary systems are operational since 2003. The helium cryo system was designed with customized requirements looking at the tokamak applications in mixed mode of operation i.e. Refrigeration (650 W at 4.5 K) and liquefaction rate of 7 g-s-1 simultaneously. Similar state of the art process for large cryogenics systems have been reported in EAST (China) 1, KSTAR (South Korea) 2 and ITER (France) [3] tokamaks systems in the world. The SST-1 cryo system comprises of helium and liquid nitrogen distribution systems to facilitate the cooling requirements of SST-1. Recent experiments of SST-1 have shown that cryogenic heat loads are more than installed cold capacity. Due to this fact, the system cool down goes into the “status-quo” in temperatures at ~12 K and further cool down is not possible. This critical situation of “status-quo” was understood and resolved by adopting the following procedures: (a) By grouping and distribution of the PF coils, optimization of the cryogenic plant process. First, we replaced common PF coils distribution to three groups having equal path lengths. (b) Providing the best possible pressure heads to each PF groups during cool down and using turbine–C to get cold capacity with active cooling of all paths [4]. (c) Identification and minimization efforts of the cryogenic heat loads from undue sources within some parts of SST-1 and outside of auxiliary system of SST-1.
It has been shown that the simultaneous cool down of the TF and PF coils are possible while achieving superconducting transition in all the coils except PF-5 (lower). Cryo heat loads minimization and mitigation within the SST-1 and current feeders system (CFS) to the best possible level, about 100 W at 4.5 K was reduced on to the helium cryogenic system. Efforts on uniform hydraulic distribution achieved while realizing the upgradation of integrated flow distribution and control system. To enhance the reliability and availability of the PLC based automation system, the old Eurotherm PC 3000 PLC system is retrofitted with new Eurotherm T2750 PLC. While adopting the above procedures, we obtained more than 2 weeks of cryogenic stability to charge the TF magnets at 4.7 kA and several hundreds of plasma discharges have been facilitated. After demonstration of longer cryo stable operation along magnet cooling and simultaneously production of ~ 50 l/h of liquid helium (refer Figure 1), we envisaged the possibility to connect the PF-3 current leads to obtain shaped plasma experiments in SST-1. Before installing the PF-3 current leads, electrical safety procedures are required to avoid any kind of Paschen discharge conditions within the PF-3 coils due to reflected voltages arise from the Ohmic coil discharge. For charging PF-3 (Upper/Lower) coils, four current leads are fastened on current feeders system (CFS) providing adequate electrical isolation using G-10 CR intermediate flange and fasteners. Welding of different vacuum breaks and isolators are very critical due to being exposed to high temperature. The electrical insulation is also major challenges on the surface exposed to high voltage (about 3 kV) due to induction. Some surfaces of CLs and feeders are insulated by Kapton tape with GFRP tape while rest of the surfaces will be covered in-situ conditions. The vacuum of cryostat between CFS and SST-1 is isolated to protect the coils and SST-1 from the Paschen-discharge as well as modularity point of view. On helium hydraulics side, pressurized supercritical helium cryogenic circuit will be re-routed and qualified for helium leak tightness. Similarly, liquid helium (LHe) supply lines from main header is re-routed in CFS chamber. All cryogenic circuits will be provided by adequate flexibility by suitable loops, bellows and bends to protect from the undue thermal stresses. Most of the low temperature surfaces (except current leads and feeders) will be covered by superinsulation MLI (Multi-Layer Insulation). Various instrumentation Viz. temperature sensors, voltage taps, pressure drop transducers and flow meters deployed for operational and diagnostics aspects. The PLC and SCADA have been programmed for the real time operation and control of PF-3 current leads. The photograph of PF-3 current leads integration with CFS is shown in Figure 2. The paper, a brief review of the installed cryo sub-systems as well as the plans of simultaneous cool down of the TF and PF coils of SST-1, performance of newly installed PF3 (Top/Bottom) current lead pairs will be discussed.
References
1 Hongyu Bai et al 2002 Construction of a 2 kW/ 4 K helium refrigerator for HT-7U Physics Sciences & Technology Vol 4 No. 3 1305-1310.
2 H.S. Chang, et al 2008 The on-site status of the KSTAR helium refrigeration system AIP Conference Proceedings 985 doi: 10.1063/1.2908582.
[3] L. Serio, ITER Organization, 2007 The ITER Cryogenic System IBF conference 2007, Nice (France).
[4] P.N. Panchal et al 2019 Cryogenic process optimization for simultaneous cool down of the TF and PF superconducting coils of SST-1 Tokamak, ICEC 2019, Oxford, UK.
Speaker: Vipulkumar Tanna (Institute for Plasma Research) -
Assembly of 1 MV power supply (PS) components to produce 1 MeV negative ion beams have been completed for the ITER neutral beam test facility (NBTF). To realize 1 MV insulation after the assembly of long and complicated components, (i) dust and particle during the assembly were controlled, (ii) the transmission line (TL) with a total length of 100 m composed of 22 series connected vessels was assembled with an accuracy of less than ± 2.5 mm, and (iii) contact resistance at more than 500 connecting points for inner electrical conductors were controlled. Finally, high-voltage has been successfully insulated in the final configuration with all components. The present achievement contributes to an insulation design verification of the ITER NB PS system.
(Introduction)
In the ITER neutral beam (NB) system, deuterium negative ion beam with 1 MeV, 40 A for 3600 s is required for plasma heating and current drive. The NB Test Facility (NBTF) is now being constructed in Padova, Italy, to achieve such high requirement prior to the ITER operation. Japan procures DC 1 MV power supply (PS) components for the NBTF and the ITER NB. Since this PS consists of various kinds of component with 100 m long, manufacturing and factory test have been performed over five years, and shipping and installation have been proceeded in parallel step by step as presented in the last FEC [A]. The on-site assembling technique for such complicated 1MV PS components is one of challenges not only for the ITER NB but also typical electrical components. As recent progress in the last two years, completion of assembly and also high-voltage insulation test of DC 1 MV power supply components are reported.(Assembly of DC 1 MV power supply components)
The DC 1 MV PS system as shown in Fig.1 consists of five DC generators (DCG), DC filter (DCF), transmission line (TL), 1 MV insulating transformer, high-voltage deck 2 (HVD2) to provide cooling water and gas from the grounded potential to 1 MV and HV bushing acting as a bulkhead between a vacuum-insulated beam source and the gas-insulated TL and also an insulating feedthrough for conductors and pipes. The TL with a total length of 100 m is composed of 22 series-connected pressure vessels with a typical diameter of 2 m and length of 5 m. Conductors with five-different voltages from 0.2 to 1 MV are allocated inside the vessel filled with SF6 gas. As for the installation of such the TL extending from indoor to outdoor, (i) the particle content in the environment was confirmed to be less than 0.14 mg/m$^3$. To mount sensitive internal parts inside the vessel, a temporal dustproof room was utilized to avoid dust. As the geometrical requirement on the installation, (ii) a fine positioning is quite important to connect vessels in series over 100 m. The routing of the TL identical to that in ITER is complicated, and hence vessels shall be installed with an accuracy of less than $\pm$ 2.5 mm taking into account the specification of the bellows equipped at several points to absorb the displacement. Among the vessels of the TL for instance, the heaviest one with 18 ton, 3 m in diameter, 7 m in long, encloses the surge block core for a surge protection in the breakdown, whose location was designed through a circuit analysis. The designed location was inevitable to have sufficient ability to absorb the surge energy, however, that was just at the building wall penetration, where the overhead crane cannot access. To realize the fine positioning even in such location, the tank was lifted down indoor and rolled to outdoor with special rollers on the slide rail. At the final connection in the TL3, the position of the vessel was confirmed to be with the accuracy of $\pm$2 mm which can be absorbed with the bellows.Inside the vessels, inner conductors at 0.2~1 MV potential are allocated and each conductor has a connection point every 2~3 m. In order to connect in series over 100 m in the TL, a flexible connector was utilized to absorb axial and angular displacement by sliding and swinging the head of the connector. Meanwhile, the number of connecting point is more than 500 in the TL and high-current conductor up to 5000 ampere is included. Therefore, the contact resistance at a connection point should be suppressed as small as possible to avoid excess heat generation at the connecting point. Thus, according to the current, the criteria of the contact resistance were set. As a result, the typical contact resistance was ~4$\mu\Omega$ for a high-current conductor and ~40 $\mu\Omega$ on average for the low-current (less than 750 ampere) conductor, which were less than allowable values of 9.6 $\mu\Omega$ and 50 $\mu\Omega$, respectively.
(High-voltage insulation test of DC 1 MV PS components)
The 1 MV PS system is required to operate stably for 1 hour. Characteristics of DC long-pulse insulation is perfectly different from that in AC insulation (or DC short-pulse insulation). During DC long-pulse insulation, the voltage distribution changes from capacitive distribution to resistive one. Taking into account of the time constant, test conditions were selected and agreed by ITER as (1) 1.2 MV for 1 hour at 120 $\%$ of rated voltage, (2) 1.06 MV for 5 hour for long-pulse insulation and (3) repetitive ramp-up from 1.06 to 1.265 MV (five times) to simulate over-voltage. High-voltage insulation test (HV test) were performed as the first site acceptance test with the testing power supply under the above test condition. The HV test was performed with testing power supply (TPS) in the upstream (DCG and DCF, HV test-1) and downstream (TL~HV bushing, HV test-2~5) area, respectively as shown in Fig.1. During the HV test 2~5, the HVD1 procured by EU, 1 MV insulating transformer and HV bushing were connected to TL sequentially to mitigate a risk of failure propagation in the PS components. Figure 2 shows the measured TPS current at 1 MV for each HV test and its content was analyzed. At the HV test 1, the measured current showed good agreement with the estimated one and the insulation capability has been confirmed. Taking into account a contribution from each component obtained in the HV tests 2~4 and the factory test, the TPS current in the final configuration was also analyzed. In the HV test 5, miscellaneous current was found. It could be the leak current attributing to exposing the TL to the atmosphere after the HV test 4, the measured current was comparable to the estimated one, which also showed electrical soundness in addition to the insulation capability. And then, finally, all HV tests have been successfully completed.The present achievement contributes to push forward with a start of the second site acceptance test integrating with the invertor system and DCGs to be done in n the 2nd quarter of 2020, which leads to an insulation design verification of the ITER NB PS system.
(Reference)
[A] H. Tobari et al., Proc. IAEA FEC, FIP/P1-10 (2018).
Speaker: Dr Hiroyuki Tobari (National Institute of Quantum and Radiological Science and Technology) -
Jing Wu1*, Jiming Chen1, Liman Bao2, Pinghuai Wang1, Kun Wang3, Stefan Gicquel2, Xiaobo Zhu1, Qian Li1, Hui Gao1, Weishan Kang1, Rene Raffray2
1Southwestern Institute of Physics, Chengdu 610041, China
2ITER Organization, 13115 St Paul Lez Durance, France
3China Int’l Nuclear Fusion Energy Program Execution Center, Beijing 100862, China*E-mail: wj@swip.ac.cn, Tel: +86-28-82820873
The design of the 12% first wall(FW) panels of enhanced heat flux(EHF) type which China undertakes has been largely improved for better maintainability and manufacturability. Contrasting to the old one, new design realized accessible pipe connections for non-destructive testing(NDT) between fingers and the fingers mechanically assembling to the central beam(CB). It largely reduces the number of cover plates welding to various water-coolant headers and consequently reduce the water leak risk in ITER operation. It was simplified the plasma-facing surface profile with optimized Be tile shapes. As a result, the number of welds, surface facets and Be tiles reduced by 49.5%, 10% and 14.3%, respectively. Thermal mechanical analysis has been performed, consistent results were obtained with ITER requirements.
As the in-vessel and plasma-facing components, and ITER EHF FW panels need to withstand surface heat fluxes up to 4.7MW/m2 with a thermal fatigue life of 15,000 cycles1. The old design adopted 15 mm heavy section welding joint between fingers and CB, and 178 small welding covers were also adopted for its cooling channels[2]. There were more than 507 welds excluding the large number of CuCrZr/316L(N) explosion bonding and the Be/CuCrZr diffusion bonding interfaces. Water connection among them were inside the panel, welds for the pipe connection were not accessible for NDT and repairing. It gave rise to high risk of cooling water leakage and made it difficult to replace any defective fingers after assembly. Welding deformation could hardly be controlled to the requiring FW assembly tolerances. Learning from experience of semi-prototype program, such kinds of drawbacks shall be solved.
Detailed new design of an ITER EHF FW panel is shown in figure 1. It contains 32 standard fingers, 8 edge fingers and 1 CB. To conform with the magnetic field and to avoid any leading edge of unacceptable heat load, both type of the fingers have toroidal shaping but additional poloidal shaping for the edge fingers. Edge fingers have inclined Be surface in poloidal direction leading to irregular inner hypervapotron cooling channel and sharp corner profile of Be tiles. This specific design is unavoidable and may cause higher Be surface temperature than the standard fingers due to longer heat transfer path. Comparing with old design, main characteristics of the new one is mechanical assembly with dovetail slot and bolting connection between its 40 fingers and the 316L(N) CB, which allows demounting fingers from CB without damaging them whenever finger repairing is required. The external pipe welds for cooling channels connection have also been employed, they allow 3 times cutting and weld repairing. The number of cover plates and corresponding welds has now been reduced from 178 to 36, and covers area was enlarged better for welding. The number of facets located on fingers upper surface have been optimized from 20 to 18. Shape and layout of Be tiles are both improved, total type of Be tiles reduces from 168 to 77 and total number of Be tiles reduces from 2400 to 2056.
Figure 1. A 3D model of the pipe connection assembly type ITER EHF FW panel
Due to huge model of new type EHF FW panel for thermo-mechanical analysis, it is divided into standard finger pair, edge finger pair and CB to simplify hydraulic analysis. As boundary condition for analysis of CB, finger pairs are replaced by "U pipes" with relevant pressure drops and velocities that are gotten from the analysis of the finger pairs. Heat transfer coefficient(HTC) values computed from hydraulic analysis results are employed for thermal analysis. Heat flux, and nuclear heat are loaded as boundary conditions. The heat flux is loaded on the whole Be surface in a value of 0.35MW/m2 and on an local area at 4.7 MW/m2. Total computing period for thermal analysis is 8 cycles and each cycle includes 400s burn and 1400s dwell.
Thermo-hydraulicanalysis results show that the flow is reasonably distributed and coolant velocity in most region is in the range of 1~10m/s. Pressure drop between general inlet and outlet is 0.35MPa which is less than target value of 0.4MPa for each FW panel. Maximum temperature of the CB and standard finger pairs in the transient thermal analysis are about 409.21℃and 650.56℃, respectively. Both meet the design criteria[3]. For edge finger pairs the irregular hypervapotron cooling channel makes it increase to846.54℃. As ITER FW is designed by test, further high heat flux test shall be implemented to validate whether it satisfy ITER's requirements.
Mechanical analysis utilizes temperature results obtained from the thermo-hydraulic analysis, and loads coolant pressure and other boundary conditions. Analysis results show that maximum deformation of edge finger pair, standard finger pair and CB are 4.18mm, 3.53mmand 2.8mm, respectively. The maximum stress is 1038MPa located on CB. For high stress region, linear analysis shows that they all satisfy ITER SDC-IC rules. Thermal fatigue assessment was performed, and the results will be presented.References
1 Bao L.. 1.6.P1A.CN.01 Blanket First Wall Annex B[R]. ITER Document Management(IDM). ITER_D_GEA8UD, v1.3, Nov. 2016.
[2] M.N. Sviridenko, et al. Analyses results of the EHF FW Panel with welded fingers [J]. Fusion Engineering and Design, 2014, 89:937–948.
[3] A.R. Raffray,et al. The ITER blanket system design challenge[J]. Nuclear Fusion, 2014, 54:033004 (18pp).Speaker: Mr Jing Wu (Southwestern Institute of Physics) -
In this presentation, we report recent advances in the development of the CFQS quasi-axisymmetric stellarator as a joint project of National Institute for Fusion Science, Japan and Southwest Jiaotong University, China. The quasi-axisymmetric stellarator (QAS) offers good plasma confinement properties with low aspect ratio, giving a prospect to become a compact fusion reactor. MHD equilibrium of the CFQS was designed based on that of the CHS-qa [1,2]. The modular coil (MC) system, toroidal field coils (TFCs), poloidal field coils (PFCs), and support system have been designed. Analysis based on finite element method showed sufficient robustness of support system for 1 T operation. A mock-up of MC is now being constructed by a manufacturing company in China to check the manufacturability and achieved accuracy of the MC. The project is steadily proceeding on schedule to achieve a first plasma in 2021.
A QAS has an axisymmetric magnetic field configuration in the special magnetic flux coordinates, Boozer’s coordinates. The guiding center orbit of particle in plasma is determined by the structure of magnetic field in Boozer’s coordinates, therefore the neoclassical transport of the QAS can be reduced to a level of similar to tokamak. Moreover, QAS does not essentially require plasma current with retaining the advantage of the steady-state operation capability. Therefore, QAS offers a great opportunity to overcome disadvantages in both tokamaks and helical systems, e.g., major disruption in tokamaks, enhanced neoclassical transport and/or suppressed plasma rotation due to non-axisymmetric ripple in helical systems simultaneously. Furthermore, in the QAS, the low aspect ratio can be achievable unlike other present stellarators while keeping good confinement property. Therefore, QAS has a great potential to realize the compact fusion reactor in the future.
The CFQS will be the first QAS in the world. The device will be constructed in China, under the international joint project of National Institute for Fusion Science, Japan and Southwest Jiaotong University, China [3]. By using this experimental device, we will perform proof-of-principle experiment to verify the effectiveness of the QAS concept. The physics design for the CFQS is almost completed [4, 5]. Good confinement property in the context of neoclassical transport is realized by introducing the QAS configuration. Due to QA configuration, substantial bootstrap current will be spontaneously generated in finite condition. However, analysis tells us that no significant effect due to bootstrap current will be seen in the QAS property. MHD instability is stabilized by the magnetic well property, and by using Mercier criteria we confirmed that interchange mode can be stabilized. Also, turbulent transport is investigated by the GKV code. The result shows that magnitude of turbulent transport is similar to that of axisymmetric-limit configuration.
The engineering design is now being performed intensively. The main parameters of the CFQS are shown in Table 1. The total number of MCs was chosen to be 16 with 4 different types of coils. Normal conductor of copper is used for coils. For flexibility of magnetic field configuration, 12 TFCs and two pairs of PFCs are designed. The magnetic coil system and vacuum chamber of the CFQS are depicted in Fig. 1. In order to resist large electro-magnetic forces while keeping large space necessary for diagnostics and heating systems, we have designed the support structures, which consist of coil cases with cage-like support structure and center pillars. The cage-like support structure is formed by the top frame, bottom frame, and eight outboard side pillars as shown in Fig. 2. Analysis based on the finite element method was performed by ANSYS/Mechanical to check the von-Mises stress, deformation, and elastic strain. These results show that this support system is sufficiently robust for 1 T operation.
**Table 1 ** Main designed parameters of the CFQS
Magnetic field strength 1.0 T Major Radius 1.0 m Minor radius 0.25 m Aspect ratio 4.0 Toroidal periodic number 2 MCs 16 (4 types) TFCs 12 (3 types) PFCs 4 (2 pairs)
At present, the mock-up MC, which is the most complicated in shape, is now being constructed by a manufacturing company in Hefei, China to check the feasibility of manufacturing and the accuracy of the MC. The computerized numerical control machine was employed to finalize the surface of winding mould. The copper conductor wrapped with layer insulator material was wound onto this winding mould, and the vacuum pressure impregnation (VPI) process finished as shown in Fig.3. Now, various tests, e.g. impulse test to check isolation and running test, are being carefully performed. The first plasma will be obtained in 2021. The CFQS will show the effectiveness of the QAS concept experimentally, and hence, a new possibility of advanced magnetic confinement devices.
[1] S. Okamura et al., Nuclear Fusion 41 (2001) 1865.
[2] K. Matsuoka et al., Fusion Science and Technology 46 (2004) 378.
[3] M. Isobe et al., Plasma Fusion Res. 14 (2019) 3402074.
[4] A. Shimizu et al., Plasma Fusion Res. 13 (2018) 3403123.
[5] H. Liu et al., Plasma Fusion Res. 13 (2018) 3405067.Speaker: Dr Akihiro Shimizu (National Institute for Fusion Science) -
This report focuses on the development of the thermal insulation devices including thermal shield (TS) and cryostat for the superconducting tokamak JT-60SA.
- Design, manufacturing and acceptance test were successfully completed
by 2019 and installation will be done by March 2020.
- The technique and knowledge to realize high accuracy manufacturing
and short time installation of these devices will contribute to the
ITER construction and DEMO design.JT-60SA (Super Advanced) as a superconducting tokamak is under construction in the Naka Fusion Institute of QST. The tokamak assembly had started since Jan. 2013 and the superconducting coils, the toroidal magnetic field (TF) coils and equilibrium magnetic field (EF) coils, were installed into the tokamak in early 2018. The size of the JT-60SA tokamak is over 13.5m diameter and 15.5m high although this is almost a half of ITER, and the tokamak components were manufactured with some millimeter precision. The design on each component is required to endure the electromagnetic force and seismic load, moreover, to absorb the thermal displacement with the clearance to the other surrounding components. These are installed into the tokamak along with the assembly coordinate built in the torus hall, and consequently, positioned in millimeters order as designed to reduce the magnetic field error below 10^-4 B_tor 1.
The thermal shield is designed to insulate superconducting magnets from the radiation heat intrusion of the vacuum vessel and cryostat in the cryostat vacuum environment. This structure is designed as the double wall with the 80K pipe set inside between the walls on the condition of the high-pressure gas law, and the double wall is composed of thinner stainless plates to reduce the thermal stress due to the operational temperature gradients and electrically insulated in every 20-degree toroidal sector to reduce the eddy current during the operation. The gravity support of suspend type is designed to absorb the thermal displacement and to keep the clearance each other at operational temperature so that heat intrusion to the superconducting coil was reduced by heat conduction distance. The entire TS is supported at the horizontal port thermal shield (HPTS) from the base on each TF coil. The TS surface facing to the cryostat vessel is entirely covered with low emissivity sheets layered as multilayer-thermal insulation (MLI) as shown in FIG.1, and the TS is maintained at the cryogenic temperature with 80K helium gas flow during the operations.
In the TS assembly, the 20-degree TS sectors manufactured in the geometrical tolerance of ±5mm was fixed onto the assembly jigs and positioned by using the laser tracker measurement in the torus hall and kept the over 30mm clearance between TS and each other component by the onsite customization to reduce the misalignments when sectors were installed to the tokamak and joined together with the mechanical couplers. In 2019, Port thermal shield (PTS) and the gravity support for the entire thermal shield were installed as shown in FIG.2. and middle cryostat thermal shield (MCTS), which was close to the cryostat lateral vessel body region, was installed. The precision of CTS position has been successfully kept customizing the suspending support arm flanges jointed to the HPTS.
The cryostat is designed to provide vacuum environment of 10^-3 Pa to insulate superconducting magnets thermally from the ambient and the major dimensions are described as the size of this tokamak. The manufacturing size is designed to be divided mainly in three parts, such as base, cylindrical section and lid, and each part is split in pieces because of the Japanese domestic transportation limit. The base (cryostat base, CB) and cylindrical section (cryostat vessel body cylindrical section, CVBCS) were manufactured in the EU and sent to the QST Naka-site. The CB was completely installed in early 2013 as the first component of JT-60SA and the CVBCS has been installed recently in later 2019.
The cryostat is to be integrated by bolt-joint with the onsite vacuum seal-welding, which requires high precision in each large part split, so that both of CB and CVBCS were pre-assembled each other when those were manufactured at the factory. The orientation at each port in the CVBCS was kept in ±8mm and these results of pre-assembly2 has been successfully reflected to the reamer pins for shortening the onsite assembly only in a few months. The closure lid at the cryostat top was manufactured as the two half-pieces in Japan, and the 60-degree sector mock-up trial to predict the welding distortion kept the precisions of each piece as a spherical shape 8m radius of within ±8mm and outer 11.5m diameter of ±5mm. These pieces will be weld-jointed as one onsite, and this lid will be set on the top of the CVBCS and fixed with mechanical clamps at the groove in each flange and seal welded by March 2020.Rest top region of the CTS and Cryostat will be assembled while feeders for the superconducting coils and cryogenic pipes are installed. This paper reports the steady progress on-schedule of the device developments on the JT-60SA and its manufacturing toward the start of the plasma operation in 2020.
1 Y.Shibama, et al., Fusion Eng. Des. 125 (2017) 1–8.
2 J.Botija, et al., Fusion Eng. Des. 146 (2019) 822–826.
Speaker: Yusuke Shibama (NATIONAL INSTITUTES FOR QUANTUM AND RADIOLOGICAL SCIENCE AND TECHNOLOGY) -
Abstract In the frame of the EUROfusion breeding blanket research activities, two reference blanket concepts are developed, the helium cooled pebble bed and the water cooled lead lithium (WCLL) blankets, which represent the most attractive designs for a DEMO reactor $\left[ 1\right] $. Test Blanket Modules (TBMs) derived from these concepts will be tested in ITER.
The design of breeding blankets represents a major challenge for fusion reactor engineering due to the performance requirements and the severe operating conditions in terms of heat load and neutron flux. Liquid metal alloys such as lead-lithium, PbLi, are considered as breeder material due to their lithium content and as coolants because of their large thermal conductivity and the possibility to be operated at high temperature. On the other hand, the motion of the electrically conducting breeder in the plasma-confining magnetic field induces electric currents and generates strong electromagnetic forces that modify significantly the velocity distribution in the blanket compared to hydrodynamic conditions and increase pressure losses $\left[ 2\right] $. Magnetohydrodynamic (MHD) pressure drops have to be carefully quantified, since excessive values can jeopardize the feasibility of the considered blanket concept. The present work investigates numerically liquid metal MHD flows in manifolds of a WCLL TBM. Velocity and pressure distributions are analyzed.Problem description A view of the WCLL TBM manifold design is shown in Figure 1a. It consists of two long poloidal ducts, which are electrically connected across a common wall. The liquid metal has to flow along a series of expansions and contractions due to the presence of horizontal stiffening plates that separate the breeder chambers arranged in a column. This type of MHD flow is known to cause additional pressure drop compared to flow in straight ducts $\left[ 3\right] $. Liquid metal flows are investigated in the model geometry depicted in Figure 1b whose dimensions are taken from the most recent WCLL TBM design. A uniform magnetic field is imposed in toroidal direction.
Numerical simulations are performed for periodic fractions of the manifolds by assuming a stepwise decrease/increase of the flow in the feeding/draining manifolds. Velocity and pressure distributions in the entire manifolds that feed/drain 8 breeder zones are then reconstructed. The pressure drop depends linearly on the mean velocity and quadratically on the magnetic field strength. Therefore, results can be extrapolated and scaled, e.g. to other flow rates, depending on design specifications.
Calculations have been performed by using a finite volume code. Equations describing the MHD flow are implemented in the open source software OpenFoam $\left[ 4\right] $. Accurate simulations of MHD flows are pretty demanding since they require a proper resolution of thin boundary layers that form along all walls. Their thickness reduces by increasing the intensity of the magnetic field $\left[ 5\right] $. Moreover, electrically conducting walls provide closing paths for electric currents and their resolution is also crucial to determine the total current density in the fluid. About 4·$10^{6}$ nodes are needed in the fluid and in the wall to resolve an eighth fraction of the manifold.Objectives of the study According to the ITER schedule the conceptual design for the WCLL TBM should be completed by 2020. Therefore, there is urgent need to reduce technical uncertainties of the present design. With this purpose, a task has been initiated by EUROfusion to investigate numerically liquid metal MHD flows in manifolds that distribute and collect PbLi in the WCLL TBM. Previous experimental and theoretical analyses for comparable manifolds of a helium cooled PbLi blanket showed the decisive role of the manifolds in determining pressure drop and flow distribution in the blanket module $\left[ 6\right] $. Results presented here serve four main purposes:
- to provide an estimate of pressure drop in the WCLL TBM manifold, which has to be compatible with the pumping capability in the PbLi loop
- to be compared with data from experiments that are currently prepared at the Karlsruhe Institute of Technology to study MHD flows in a scaled mock-up of a WCLL TBM. This permits a further validation of the implemented physical models and the employed predictive tool
- to complement the experimental data in order to obtain an overview of the investigated MHD phenomena and to derive scaling laws to extrapolate results to realistic ITER operating conditions
- to support correct interpretation of surface electric potential measurements in terms of velocity distribution inside the test-section. Especially when walls are thick and induced currents have to be taken into account a comparison with numerical simulations is compulsory.
Brief overview of numerical results Figure 2a shows as an example a 3D view of the velocity distribution in the considered geometry for the position at the bottom of the TBM where the fluid is supplied to duct 2, while the weak flow in duct 1 is only driven by electromagnetic coupling caused by leakage currents. In Figure 2b the vertical component of the velocity is plotted along the radial dashed lines in the two channels. Due to electromagnetic coupling across the common dividing wall the flow in duct 2 pulls that in the core of duct 1 in the same direction. This leads to a buildup of pressure along the flow direction that drives the backward oriented jets in duct 1 (red curve in Figure 2b).
A principle sketch of the pressure distribution in the manifolds along the poloidal direction is depicted in Figure 3. Since the cross-section of the two channels remains constant, while the flow rate changes along the poloidal path of the liquid metal flow, the pressure exhibits a non-linear profile. This results in different pressure drops in the breeder units and a corresponding non-homogeneous partitioning of the fluid among them. The latter is further increased by the fact that the draining manifold has a cross-section larger than the one of the feeding duct and hence a smaller pressure gradient. From the calculated pressure distribution flow rates in the breeding zones can be estimated in order to get a picture of the way in which the liquid breeder distributes.
References
$\left[ 1\right] $ G. Federici et al, “An overview of the EU breeding blanket design strategy as an integral part of the DEMO design effort”, Fusion Engineering and Design, 141, pp. 30-42, 2019.
$\left[ 2\right] $ G. Aiello, C. Mistrangelo et al, "MHD issues related to the use of lithium lead eutectic as breeder material for blankets of fusion power plants", Magnetohydrodynamics, 51, pp. 185-193, 2015.
$\left[ 3\right] $ L. Bühler, “A parametric study of 3D MHD flows in expansions of rectangular ducts”, Fusion Science and Technology, 52, pp. 595-602, 2007.
$\left[ 4\right] $ C. Mistrangelo, L. Bühler, "Development of a numerical tool to simulate magnetohydrodynamic interactions of liquid metals with strong applied magnetic fields", Fusion Science and Technology, vol. 60, p. 798, 2011.
$\left[ 5\right] $ S. Smolentsev et al., “An approach to verification and validation of MHD codes for fusion applications”, Fusion Engineering and Design, 100, p. 65-72, 2015.
$\left[ 6\right] $ L. Bühler, C. Mistrangelo, H.-J. Brinkmann, C. Koehly, “Pressure distribution in MHD flows in an experimental test-section for a HCLL blanket,” Fusion Engineering and Design, 127, pp. 168-172, 2018.Speakers: Chiara Mistrangelo (Karlsruhe Institute of Technology (KIT)), Prof. Leo Bühler (Karlsruhe Institute of Technology) -
The 3D nonlinear equilibrium and its associated magnetic topology are investigated on EAST for the first time for future understanding on the mechanism of how the Resonant Magnetic Perturbation (RMP) mitigates or suppresses the Edge-Localized Mode (ELM). Recently, a nonlinear transition from mitigation to suppression of the ELM by using RMP on EAST is observed $^{1}$. To understand the RMP mechanism in different ELM phases, the nonlinear 3D resistive equilibrium is studied by HINT code $^{2}$, and impacts of the plasma rotation on the magnetic topology are investigated. To suppress or mitigate ELM strongly, the deep penetration of RMP field or the relevant edge topological change resulting from nonlinear plasma response is supposed to be the key factor.
The ELMs control in the H-mode operation is a key issue in the future tokamak fusion reactor research like ITER, and one of the most effective proposals of type-I ELM control is RMP $^{3}$. To investigate the RMP mechanism, the plasma parameter profile in the pedestal region and the edge magnetic topological change are the key points. The HINT code, directly solving the dynamic equations of the magnetic field and pressure based on the relaxation method, has the unique advantages of 3D MHD equilibrium with magnetic islands and stochastic edge field induced by RMP, which is maintained by NIFS group. Here we apply the HINT code to study 3D equilibrium with RMP on EAST and develop the plasma flow effect part to study the nonlinear interaction between RMP and plasma rotation effect on magnetic topology.
The nonlinear transition between ELM mitigation and suppression by scanning of the phase difference between upper and lower RMP coils $\delta \Phi _{UL}$ with n = 1 RMP on EAST is obtained $^{1}$. In the EAST shot #55272, the ELM frequency behaves in a significantly different way at different phase. For ELM suppression or strong mitigation phase, the RMP field could penetrate deeply. While for the weak mitigation phase, the RMP field is shielded by plasma response. The 3D equilibrium is calculated for EAST shot #55272 in different ELM phase. Figure 1 (a) shows the pressure profile calculated by HINT for without RMP (t = 3.9 s), strong ELM mitigation phase (t = 4.2 s) and suppression phase (t = 4.5 s) when the RMP penetrates. The pressure profiles in the plasma core region for the different phases are almost identical while the profiles in the pedestal region are significantly different with RMP penetration, as figure 1 (b) shows. Or equivalently, as shown in figure 1 (c), the pressure gradient decreases in the pedestal region with significant 3D edge magnetic field topology change when the RMP penetrates. It can be inferred that the bootstrap current density will decrease with the pressure gradient degradation, resulting in ELM suppression or strong mitigation concerning peeling-ballooning mode. Here, the evidence of plasma pressure gradient degradation with RMP penetration is presented on EAST for the first time from the point of view of 3D nonlinear equilibrium.
The RMP penetration is a nonlinear process essentially. The nonlinear interaction between the RMP and plasma rotation is also the key point for the final 3D equilibrium. The RMP field could be shielded by plasma response, resulting in a weak ELM mitigation case. Based on the HINT model, the plasma response effect is simulated by introducing plasma flow in the evolution of magnetic field calculation. Since HINT employs a resistive model, the initial surface plasma flow cutting the RMP field can generate the electric field when RMP superposed, leading to the screening current by Ohm’s law to shield the RMP field. In the meanwhile, the flow velocity is also changed by the momentum equation to keep its self-consistency. Or the nonlinear interaction between the RMP field and plasma rotation is treated as a nonlinear process of equilibrium with plasma flow in the modeling. As HINT is based on the single-fluid model, here we technically divide the plasma flow into two parts, the parallel flow and perpendicular flow concerning the initial 2D magnetic field line given by EFIT. Recently the module of parallel flow part is developed. Figure 2 shows the magnetic topology in the weak ELM mitigation phase for EAST shot #55272 calculated by HINT with or without parallel flow. The 3/1 magnetic island width becomes small when the parallel flow is taken into account. Since the main component of RMP field is the radial component $\vec{B}_r$, the parallel flow can generate the electric potential by $\vec{v}_{\parallel0}\times\vec{B}_r$ term and change the current density profile according to Ohm’s law in the nonlinear equilibrium calculation. The additional current component induced by the initial parallel flow could give rise to magnetic island healing topology, resulting in the reduction of the RMP field penetration depth. To model the plasma response more precisely, the perpendicular flow is important and will be included in the future work. Here, the plasma response effect with RMP is modeled by introducing the plasma flow in the 3D nonlinear equilibrium calculation and the magnetic island healing is presented with the parallel flow, which could give the interpretation of the RMP mechanism in weak ELM mitigation case.
Reference:
1 Y. Sun et al 2016 Phys. Rev. Lett. 117 115001
2 Y. Suzuki 2017 Plasma Phys. Controlled Fusion 59 054008
3 T. E. Evans et al 2006 Nat. Phys. 2 419Speaker: JIE HUANG (National Institute for Fusion Science) -
For tokamaks to be attractive as the core of future fusion based power plants, it must operate in steady state or at least quasi-steady state without plasma current disruptions. As is evident from the predictions for ITER based on present day tokamak research, a major challenge would be to avoid disruptions in majority of plasma discharges with full plasma parameters achieving >99% good shots. Or, in discharges where disruptions become unavoidable, one should be able to diagnose it early enough to take effective actions for mitigation of thermal and electromagnetic loads. Presently, major research and development efforts are underway in most fusion machines across the world to develop a robust mitigation technique that would be effective in ITER. This R&D would be very useful and may even decide parameters and operational space of future DEMO reactors. A critical aspect of disruption avoidance or mitigation is the early detection of an impending disruptive event, long enough to take corrective measures. Generally disruptions can occur in tokamaks due to a variety of reasons, e.g., plasma hitting density or q limits resulting in various MHD instabilities, core impurity accumulation, H-L mode transitions, loss of plasma position control and so on. Due to this multi-dimensional nature of the problem, it is often difficult to simultaneously track all plasma and machine parameters for effective disruption prediction. Perhaps a better way for effective predictions technique would be based on machine learning technique, which has received a lot of interest in recent times across all tokamaks [1-3].
Aditya is a small sized tokamak with modest plasma parameters (R/a=0.75m/0.25m, Ip(max)=0.25 MA, Bt=1.2-1.5T) and a circular, limited configuration. Still it is very effective for disruption studies. This paper describes a novel tool for quick data visualization and parameter selection for disruption prediction based on a machine learning technique applied on Aditya data. Recently Sharma et al. [4] have proposed convolution neural network (CNN) based binary classification scheme to classify disruptive and non-disruptive shots, building on a database of time series data of about 100 shots with 10 parameters of Aditya tokamak. This time series data is quantized, converted into 1000 images, and classification is performed using CNN. Their scheme achieved 91.22% mini-batch training accuracy in 500 epochs. We have selected a data set of 2000 Aditya discharges, which include both disrupted and non-disrupting ones for this study. The objective of our research is to develop a numerical tool for prediction of plasma disruptions based on state of the art machine learning algorithms applied on diagnostic data which will be compatible with the real time hardware based solution for avoidance or mitigation actions. To enhance the intuitive understanding and association of input data of disruptions, a 2D visualization tool has been developed. This uses color-coding of normalized parameters and decision surfaces projecting probable disruption space for offline and real time dataset. The system will be validated for time series prediction of plasma disruption. We have first used a sub-set of 1000 labeled shots, each having 156 recorded parameters (thus a total dataset of 156000), each shot with a sampling rate of 5000 samples per second. These shots are manually labeled in five shot-types of normal, disruption, no-discharge, small discharge and others. Preliminary investigation shows the database has a large redundancy of information for desired prediction of disruption and optimization of the number of parameters is required to reduce computation time. For this, we have designed an elaborated offline artificial neural network (ANN)-based correlation algorithm to compute the score of each parameter with respect to plasma current. Six top score parameters were found consistent with different combinations of above dataset. However, in second set of 1000 shots, shot-type information was not included. A shot-type classification model is developed using 25 sets of feedforward ANNs working concurrently using all samples of each parameter as input. The combined result of the ANNs predicts the shot-type with overall 96.5% accuracy, whereas share of disruption classification accuracy is 99.0%. This model is further applied to the second set of 1000 shots with unknown shot-type, where the results were validated by human expert with 99% accuracy. This is possibly due to the human error of sub-set one taken care by neural network.
Each dataset of Aditya has its own characteristic like intensity, bandwidth, offset, life-cycle etc. Choice of ideal parameters for prediction needs multi-variable optimization over large number of shots. Visualizing the combined information at a glance is found useful for decision making. Information gain of each parameter is color-intensity coded to view in 2D plane. The viewing parameters are sorted along the X-Y axis such that the information density is highest to lowest across the diagonal as shown in Fig. 1, in which 20736 parameters are displayed with intensity coding cells representing information density. Each cell represents a parameter of a shot which can be viewed in a pop-up plot. The darker zone towards top-right corner represents low information area and the bottom-left corner, which is the area of interest, has high information gain intensity. The map is associated with the back-end database to view the parameter plot of each cell out of the 156000 samples. The display is further enhanced by color coding of the shot-type to increase the contrast between disruptive and normal shots. In these maps, the low information zones are dark and corresponding plots are flat. One of the important advantages of these maps is to locate missing data as a dark spot in high information density zone. These maps are very useful in identifying the right parameters for training using the ANN. Details of the above maps and preliminary results using this ANN tool will be presented in this paper.
References:
- William Tang et al, 2016, 26th IAEA Fusion Energy conference, 17-22 October, Kyoto, Japan, EX/P6-47.
- K.J. Montes et al 2019 Nucl. Fusion 59 096015
- Julian Kates-Harbeck et al, 2019, Nature, 568 (7753)
- Sharma, P. et al, 2018. “A proposed method for disruption classification in tokamak using convolutional neural network” in Towards Extensible and Adaptable Methods in Computing (pp. 179-193). Springer, Singapore.
Speaker: Indranil Bandyopadhyay (ITER-India, Institute for Plasma Research) -
M. Nakamoto, H. Kajitani, T. Suwa, Y. Takahashi, M. Yamane, T. Baba, K. Sakamoto, K. Yoshizawa, Y. Uno, A. Ishikawa, M. Nakahira, N. Koizumi, M. Inoue 1, E. Fujiwara 1, T. Shichijyo 1, K. Kuno 2, T. Minato 2, T. Hemmi 3 and C. Luongo 3
National Institutes for Quantum and Radiological Science and Technology, 801-1 Mukouyama, Naka-shi, Ibaraki 311-0193, Japan
1 Mitsubishi Heavy Industries, LTD, 1 Minamifutami, Futami-cho, Akashi 674-0093, Japan
2 Mitsubishi Electric Corporation, 1-1-2 Wadasaki-cho, Hyogo-ku, Kobe, Hyogo 652-8555, Japan
3 ITER Organization, Route de Vinon-sur-Verdon, CS 90 046, 13067, Francee-mail: nakamoto.mio@qst.go.jp
In the last report, we announced the completion of the first coil case structure for ITER Toroidal Field (TF) coil [1]. Since then, the assembly of winding pack (WP) with the coil case has been completed. In this paper, we report the completion of the first ITER TF coil in Japan. The successful assembly of TF coil was achieved by developing manufacturing techniques to overcome some challenges; such as: (i) positioning of WP inside coil case to optimize deviation in current center line (CCL), a magnetic property, of a coil under a few millimeters, (ii) satisfying the tolerances on interfaces with other coils and supports, sometimes sub-millimeter, in spite of welding deformation during coil case welding, (iii) impregnation of narrow gaps between WP and coil case with high viscosity charged resin, and (iv) tracing CCL through the assembly process and transferring the information to the coil case.
ITER TF magnet is a D-shaped coil consisting of superconducting WP enclosed in a massive austenitic stainless-steel case with a height of 16m and a width of 9m. A set of 18 TF coils will be installed in the tokamak to create a magnetic field to confine the plasma within the donut-shape vacuum vessel. Among the set of 18 coils, 8 TF coils along with a spare are manufactured by Japanese domestic agency (DA) and remaining 10 TF coils are manufactured by EU DA. In Japan, TF coil is assembled in vertical posture with straight portion of D-shape facing the bottom to realize the bilateral symmetricity of the coil. The assembly is performed in the following steps; 1) insertion of WP into straight portion of coil case, 2) positioning of curved portion of coil case and welding with straight portion, 3) closure of coil case with inner cover plates, 4) impregnation of gap between WP and coil case, 5) final machining of interface surfaces and holes, and 6) dimensional inspections and electrical tests. It is important that generated magnetic field be toroidally uniform; therefore, those 18 coils shall have the same shape and the magnetic property to be assembled evenly around the vacuum vessel regardless of differences in manufacturing strategies, tooling and postures between two DAs. In order to achieve common requirements, unprecedented and severe tolerances are defined on both the CCL and the interfaces.
The CCL of a coil is the center line of an equivalent single turn coil, which has the same magnetic field properties at far away locations. The CCL shall exist within 2.6mm diameter circle over the extent of straight portion of more than 7m while keeping sufficient gaps between WP and coil case for impregnation [2]. In order to position WP to accommodate both requirements, the target WP position within coil case was assessed in advance using the individual dimensional inspection data of WP and coil case [3]. The WP position was monitored by laser tracker during insertion so that the fine adjustment of 0.1mm order was possible. After positioning of WP, measurement results of gaps between WP and coil case showed the validity of the prior assessment.
To install 18 TF coils around ITER vacuum vessel to create uniform magnetic field, very severe structural tolerances are defined at interfaces. Since the welding deformation has large impact on the positions and shapes of interfaces, extra materials were left on critical interfaces to be machined to nominal after welding. On the other hand, leaving too much of extra materials will lead to longer machining duration. To avoid unnecessary schedule delay, amounts of extra materials were minimized. Where extra materials were found not sufficient, two DAs collaborated to compensate all 18 coils in a common deviation under guidance of ITER Organization (IO) to harmonize the shape of final coils. This harmonization strategy was a great way to manage two conflicted issues, manufacturing schedule and accuracy, and a big achievement as international collaboration project.
For impregnation of the gaps between WP and coil case, there was a concern in breakage of the seal between WP and coil case at the terminal region, where conductor ends and instrumentation wires are extracted from WP, during evacuation before impregnation and pressurization after resin injection due to the relative movement of WP with respect to coil case. To prevent the relative movement, partial vacuum chamber is installed over the terminal region to balance the pressures on both sides of the seal. The pressure balance was monitored by pressure gauge and feedbacked to pressure control inside the partial vacuum chamber. However, filler added to impregnation resin as crack prevention increased viscosity of resin and therefore, the resin’s internal pressure during resin injection. For the first coil, adjustment of pressure balance required full attention of operator since resin’s internal pressure directly affected the pressure inside coil case. For the second coil, addition of relay tank between coil case and resin injection system prevented sudden change of resin pressure and improved manageability of pressure balance.
Since WP of TF coil is enclosed in coil case, CCL positional information needs to be transferred to reference points on coil case. As mentioned earlier, TF coil is assembled standing on straight portion of D-shape, resulting in gravitational deformation in WP. The amounts of deformation are about 1mm at the straight portion and 4mm at the curved portion. After coil case welding, deformation of WP is corrected by pushing curved region of WP against coil case by bolts [2] in accordance with the prior assessment using structural analysis. However, the resulted WP shape differed from the prior assessment. Since the WP is pushed against coil case, residual stress in coil case from welding affected the final WP shape. Even though we incorporated a step to release this residual stress by adjusting supports of coil case before WP shape correction, some stress remained in the coil case. This is unavoidable situation which must be treated carefully by measurement and feed-back during the fabrication process.
The first TF coil was completed in January 2020 successfully. For the following coils, the gap-filling impregnation process was improved due to experience of the first TF coil and implemented on the second coil effectively. WP of the third coil was inserted in coil case and is now in welding phase. TF coil assembly process is now routinized and its adequacy is proven. The TF coil will arrive ITER site in April 2020 for acceptance tests and installation in Tokamak, showing a progress in ITER project.[1] M. Nakahira et al., Nuclear Fusion, Vol59, No8, 086039 (2019)
[2] N. Koizumi et al., IEEE Trans. App. Superconductivity, Vol30, No4, 2971673, (2020)
[3] N. Koizumi et al., IEEE Trans. App. Superconductivity, Vol29, No5, 4200505, (2019)Disclaimer: The views and opinions expressed herein do not necessarily reflect those of the ITER Organization
Speaker: Mio Nakamoto -
Shattered pellet injection (SPI) systems that form cryogenic pellets of low and high-Z impurities in a pipe-gun [1] for injection to mitigate disruptions have been fabricated and installed for use in thermal mitigation and runaway electron dissipation experiments on JET and KSTAR. These systems are to support disruption mitigation research for ITER and are based on an ORNL 3-barrel design for flexible pellet size and variable pellet composition studies [1]. The services for gas supply, vacuum, cryogenic cooling, and control are provided by the host institution with collaborative operation and experimental investigations organized by the host and through the ITER disruption task force [2].
The SPI systems for JET and KSTAR have a common feature of 3 different size pellets that are formed in-situ and collimated into a single injection line that enters the vacuum vessel. The pellets are fired by high pressure gas or mechanical punch and are shattered in stainless steel tubes with a 20-degree bend that are mounted inside the vacuum vessel of the tokamak, vertically on JET and horizontally on KSTAR. The JET installation shown in Fig. 1 has the unique feature of vertical SPI mounting and injection with the shatter plume aimed toward the inner wall to intercept known runaway electron (RE) beam locations generated from argon gas injection induced disruptions. Observations of the shattered pellet plume in plasmas on JET shown in Fig. 1 verify that the trajectory is as designed.
The JET SPI pellet sizes are 4.5, 8, and 12.5 mm diameter with lengths that are 30-50% longer. The two large sizes can optionally be operated with a mechanical punch for release of pure neon and argon pellets, that otherwise cannot be fired with gas alone. The control of the SPI is through a programmable logic controller programmed to automate the formation of the pellets and control vacuum components. Verification of the D2 propellant gas removal by the vacuum system was achieved by firing gas without a pellet and measuring that < 0.25% of the total gas fired ended up in the torus, showing excellent removal to prevent misinterpretation of the SPI performance by gas reaching the plasma before the pellet.
The KSTAR SPI installation described in Ref. [3] and shown in Fig. 2a has two identical 3-barrel SPIs that are mounted on the midplane with identical shatter tubes inside the vessel aimed to the plasma center after traversing a 20-degree shatter tube. Unlike the JET SPI that uses cold helium gas, these SPIs are cooled with a cryocooler that provides enough cooling to achieve 8 K pellet formation temperatures. Cooldown takes under 2 hours to be at pellet formation conditions, and it takes 5-15 minutes for the formation depending on the pellet size. The pellet sizes are 4.5, 7, and 8.5 mm with lengths that are 40% larger than the diameter. The pumping system and infrastructure were all provided by NFRI [3] and became operational in Nov. 2019. The performance of the propellant gas removal system was determined by firing the SPI with gas only into an empty torus where less than 0.1% of the gas was detected, but while the torus was being pumped. One of the key reasons for the KSTAR dual SPI installation is to investigate the performance of simultaneous injection of SPI pellets from ports on opposite sides of the machine. The identical SPI systems make this research possible as the systems have shown good synchronization between SPIs as shown in Fig. 2b where 7 mm D2 pellets fired from both SPIs arrive at their respective microwave cavities only 0.13 ms apart. Initial thermal mitigation experiments with neon-D2 mixtures have been performed with single and dual SPIs. These results show good assimilation of the pellet material into the plasma at the time of injection [4].
Installation and operation of these SPI systems has provided useful lessons learned in the implementation of this SPI technology and valuable experience in optimizing the performance of the formation and firing of the pellets. Experience on JET has shown a high reliability of pellet firing with only 1 failure to fire in over 150 attempts into plasmas since Sept. 2019. Punch operation with pure impurity pellets has also been especially reliable at firing pellets, but frequently results in broken pellets out of the injector .
[1] L. R. Baylor, et al., Nucl. Fusion 68 (2019) 211.
[2] S. Jachmich et al., to be submitted to 28th IAEA Fusion Energy Conference (2020).
[3] S.H. Park, et al., ISFNT 2019, Fus. Eng. Des. 154 (2020) 111535.
[4] J. Kim et al., to be submitted to 28th IAEA Fusion Energy Conference (2020).This work was supported by the US DOE under contracts DE-AC05-00OR22725 and DE-FC02-04ER54698 and by the ITER Organization (TA C18TD38FU) and carried out within the framework of the EUROfusion Consortium, receiving funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission or the ITER Organization. Also supported by Korean Ministry of Science and ICT under the KSTAR project.
*See the author list of E. Joffrin et al. accepted for publication in Nuclear Fusion Special issue 2019, https://doi.org/10.1088/1741-4326/ab2276Speaker: Larry R. Baylor (Oak Ridge National Laboratory) -
The Chinese Helium Coolant Ceramic Breeder (HCCB) Test Blanket Module (TBM) and its ancillary systems (together called Test Blanket System or TBS) is one of important steps for the China magnetic confinement fusion development, which will contribute to validate the key tritium breeding blanket technologies under the burning plasma environment, including tritium extraction, heat removal, integration safety, etc. In the past two years, the design of HCCB TBS was continuously optimized based on the analysis and R&D results. Also the safety of HCCB TBS has been assessed to support the preliminary design.
As the core component, the HCCB TBM-set design was optimized, considering thermo-mechanical performance and manufacturability. The major change is the optimization of coolant flow diagram to avoid the local high temperature, though it is in the limit range. Also the purge gas flow path is adjusted to mitigate the tritium concentration at the corners of breeding zones. After the design optimization, the tritium performance of TBM still keeps at the same level as before, about 58mg/full power day. The thermo-hydraulic and thermo-mechanical performance has been checked, and the results show that the temperature and stresses are below the allowable value. Following the assembly scheme, the semi prototype of TBM has been fabricated and tested, which verified the manufacturability of the TBM design. The further optimization is ongoing taking into account the compressed pebble bed behaviors of functional materials and its neutronics impact. The integration scheme of the TBM shield is also updated considering the difficulty of the fabrication process, which will not change the structure design of TBM shield and the mockup fabrication is under plan. Considering the tritium release during the operation, one enclosure with the local detritiation system has been added for the helium cooling system (HCS) to reduce the tritium concentration in the room and the seismic analysis has been performed to verify this update. The design of the neutron activation system (NAS) has been also optimized according to the experiment and the investigation of key components. Besides, the I&C architecture has been developed to realize the conventional control, interlock and safety functions of ITER and HCCB TBS. The control by state machine requested by ITER for conventional control has been practiced and achieved in the helium testing loop (HeCEL-1).
ITER had been recognized as a basic nuclear installation (INB) by French nuclear safety authority ASN. The HCCB TBS system design should follow various nuclear safety requirements to control the radioactive risks from tritium and other neutron activation products produced during operation.The detailed nuclear analysis has been performed to identify source term inventories. The dose rate has been evaluated to demonstrate the radiation levels in different locations during ITER operation and maintenance phases which meet the ITER radiation zoning requirements. Multiple confinement barriers are designed to control tritium and mobile radioactive dusts, allowing the contamination level in each room to be compliance with ITER ventilation zoning. ALARA principle is being implemented in the design of HCCB TBS. Actions, including choices of advanced materials, improvement of shielding, reduction of activated materials and release of tritium, optimization of operation and maintenance activities, etc., have been considered to minimize the radioactive release and radioactive exposures to personals and public.
The tritium transport analysis is important to understand the tritium behavior in the HCCB TBS. After the code benchmark with other ITER members, a steady-state system level tritium transport for the updated HCCB TBS design has been performed. The amount of tritium transported to different parts of the system and released to rooms has been assessed. The results show that the tritium inventory is 0.39mg in the structure materials of tritium extraction system (TES) and 26.5mg in HCS structure materials. Also the integrated dynamic tritium transport analysis is ongoing, which will provide the dynamic changing of tritium concentrations in fluids, tritium inventory in solid materials, and tritium release towards outside based on the ITER operation scenario.
Several potential envelop accidents have been identified and analyzed, including loss of flow of cooling system; loss of coolant (break of TBM cooling channel or HCS pipes) in different locations, break of tritium systems, break of heat exchanger between HCS and ITER water cooling systems etc. During all accidents, because of the intrinsic safety characteristics of fusion device, after fusion plasma shutdown passively or actively, the decay heat in HCCB TBS will be removed through thermal radiation to large surface of ITER machines, without further over-temperature issues. The release of radioactive contents will be controlled by safety measures. No major safety consequences have been identified through accident assessments.According to the agreement between China and ITER, after operation in ITER, the irradiated HCCB TBM is expected to be transported back to China for further post irradiation tests (PIE) using special transport casks. Remaining rad-wastes from dismantling are assumed to be processed and disposed in French rad-waste facilities. Characteristics of these rad-waste and safety issues for handling them have been assessed.
Currently, the manufacture feasibility and safety of the HCCB TBS have been preliminarily verified by the R&D and safety assessment. Before the finalization of the design of HCCB TBS, the more detailed R&D and experiments will be performed to check the reliability and safety, also the compliance with the French regulations will be consulted with the agreed notified body (ANB).
Acknowledgments
This work was supported by the National Key R&D Program of China with grant numbers 2017YFE0300601, 2017YFE0300503 and 2017YFE0300502.Speaker: Xiaoyu WANG (Southwestern Institute of Physics) -
Core transport in present tokamaks is mostly ascribed to micro-turbulence driven by the non-linear saturation of ion-scale ITG-TEM [1] instabilities ($k_\theta\rho_i\le1$, where $k_\theta$ is the poloidal wave number and $\rho_i$ the ion Larmor radius). It has been shown that electron-scale ETGs [2] ($k_\theta\rho_e\le1$) can also impact the heat transport, also exchanging energy with ITG-TEM turbulence by multi-scale coupling [3-9]. This topic of investigation gains a particular relevance due to its potential impact on devices like ITER, dominated by electron heating. ETG modes have been shown to play a role in plasmas with mixed ion and electron heating, since a proper balance of ion heating, decreasing the ETG threshold in $T_e$ gradient (which increases with increasing $T_e/T_i$ [10]), and electron heating (pushing $T_e$ gradient towards threshold while increasing the threshold due to $T_e/T_i$ increase), could destabilize them. Also all mechanisms that stabilize ITGs, such $E\times B$ or fast ions from neutral beams (NBI) and/or ion cyclotron resonance heating (ICRH), due to multiscale interactions open a window favourable for ETG destabilization.
The response of the $T_e$ profiles to the applied heating can be experimentally investigated by performing normalized electron heat flux scans and/or RF power modulation analysis. The two methods can be used in conjunction to extract information on the dependence of the gyro-Bohm normalized electron heat flux $q_{egB}$ on the normalised $T_e$ logarithmic gradient $R/L_{Te}$, yielding experimental values for the threshold $R/L_{Te,crit}$ for the onset of turbulent transport and for the ‘electron stiffness’ $\partial q_{egB} /\partial R/L_{Te}$. The experimental results can be compared with the output of gyrokinetic (GK) simulations, which infer both $R/L_{Te,crit}$ (fast linear runs), and the dependence of the saturated heat flux on $R/L_{Te}$ (more costly nonlinear runs). Resolving both ion and electron scales (i.e. performing nonlinear multi-scale simulations) is computationally very demanding and just became possible in the last years.
In order to access a broad range of parameters, a great effort is actually devoted to analyse different machines, comparing experimental and numerical results, within the framework of EUROfusion and of the ITPA Transport & Confinement group. In this paper, the analysis of plasmas of three different tokamaks, i.e. the Joint European Torus (JET, at Culham, UK), ASDEX Upgrade (AUG, at Garching, DE) and the Tokamak à Configuration Variable (TCV, at Lausanne, CH), is presented. Dedicated plasma discharges have been analysed experimentally and modelled numerically, by means of GK codes (GENE [11] and GKW [12]) and reduced quasi-linear models (TGLF [13] and QuaLiKiz [14]). The results of the different tokamaks concur to make a general picture indicating that ETGs could also be important for electron heat transport in fusion relevant conditions, in particular when $T_e\sim T_i$ with consistent fast ion density.
TCV is equipped with an NBI system, that allows the plasma to achieve $T_e\sim T_i$ in conjunction with high $R/L_{Te}$ (due to ECRH), allowing to access parameters compatible with ETGs. Two dedicated L-mode discharges, with $B_0=1.41$ T, $I_p=170$ kA have been performed with a different proportion of deposited ECRH power on- vs off-axis to perform a heat flux scan. Each pulse presented different phases corresponding to a different proportion of NBI/ECRH power to vary $T_e/T_i$, with ECRH both steady and modulated to allow a perturbative analysis. Both the experimental analysis and GK modelling (linear multi-scale and nonlinear ion-scale simulations) tend to indicate a possible role of ETGs at mid-radius when both ECRH and NBI are injected simultaneously, and at a larger toroidal radius $ρ_{tor}=0.7$ also when only ECRH is injected. In the former case, the main mechanism which explains the failure of ion scales alone to explain the experimental fluxes, is the stabilisation of ion-scales by the fast ions that are produced by the NBI. These results, published in [6], provide hints of a contribution of ETGs to electron heat transport in TCV plasmas.
Experiments on the AUG tokamak to study electron heat transport [5] have produced H-mode discharges with $B_0=2.5$ T, $I_p=0.8$ MA, injecting 2.5 MW of ECRH (steady and modulated to perform the perturbative analysis) and 5 MW of NBI in order to have $T_e\sim T_i$. Different discharges had different proportions of ECRH power deposition on- vs off-axis, in order to obtain the heat flux scan. At mid-radius, both the electron heat pulse diffusivity $χ^{HP}$ (from perturbative analysis) and $q_{egB}$ (from steady state scan), indicate strong turbulence levels above $R/L_{Te}\sim$ 6-7, leading to a moderate/high electron stiffness, consistent with the possible presence of ETGs. Both GENE and GKW linear-gyrokinetic simulations predict a role for ETGs for $R/L_{Te}>6$, based on an effective model for nonlinear turbulence saturation [15]. Both ion-scale and multi-scale simulations are being performed in order to analyse the most representative AUG pulse, to test the existence of an ETG ‘wall’ limiting the achievable $R/L_{Te}$. The preliminary multi-scale results are in agreement with TGLF in indicating a >30% contribution of ETGs to the electron heat flux for the cases close to threshold, setting high electron stiffness above it (see figure 1-2).
Following early results pointing to an important role of ETGs in JET [4], very recently dedicated sessions on ETGs have been performed at JET. Both L- and H-mode plasmas have been obtained, with $B_T=3.3$ T, $I_p=2$ MA, injecting 0-20 MW of NBI and up to 6 MW of ICRH (H minority, mainly heating electrons), achieving heat flux scans for a range of $T_e/T_i$ values. The preliminary analysis of the experimental data indicates that JET results are very similar to the AUG ones, with a strong increase of the electron stiffness for $R/L_{Te}>6$ (see figure 3). L-mode cases, in particular, allow to obtain sufficiently large values of $q_{egB}$ at large $R/L_{Te}>6$, giving the hint of a possible ETG ‘wall’. In parallel, high performance hybrid discharges are analyzed in order to study the ETG impact on these scenarios. Both sets of data are being modelled by means of single scale GK simulations and reduced models, in order to set the basis for heavier multi-scale GK simulations.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014– 2018 and 2019– 2020 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This work was also conducted under the auspices of the ITPA Topical Group on Transport and Confinement. We acknowledge the CINECA award under the ISCRA initiative, for the availability of high performance computing resources and support.
[1] W. Horton, Rev. Mod. Phys. 71, 735 (1999)
[2] W. Dorland et al., Phys. Rev. Lett. 85, 5579 (2000)
[3] N.T. Howard et al., Nucl. Fusion 56, 014004 (2016)
[4] N. Bonanomi et al. Nucl. Fusion 58, 124003 (2018)
[5] F. Ryter et al., Nucl. Fusion 59, 096052 (2019)
[6] A. Mariani et al., Nucl. Fusion 59, 126017 (2019)
[7] S. Maeyama et al., Phys. Rev. Lett. 114, 255002 (2015)
[8] A. Marinoni et al., Nucl. Fusion 57, 126014 (2017)[9] C. Holland et al., Nucl. Fusion 57, 066043 (2017)
[10] F. Jenko et al., Phys. Plasmas 8, 4096 (2001)
[11] F. Jenko et al., Phys. Plasmas 7, 1904 (2000)
[12] A. Peeters et al., Comput. Phys. Commun. 180, 2650 (2009)
[13] G.M. Staebler et al., Phys. Plasmas 23, 062518 (2016)
[14] J. Citrin et al., Plasma Phys. Control. Fusion 59, 124005 (2017)
[15] G.M. Staebler et al., Nucl. Fusion 57, 066046 (2017)Speaker: Alberto Mariani (Istituto per la Scienza e la Tecnologia dei Plasmi, CNR, Milano (Italy)) -
Recent experiments in JET-ILW have been successfully exploring a high-performance H-mode scenario with no gas dosing at low $q_{95}$ ($I_p=3$ MA, $B_t$=2.8 T, $q_{95}=$ 3.2) and low triangularity, with peak neutron rates reaching values of 3.6$\times 10^{16}$ s$^{-1}$. This was enabled by operation at very low gas fueling, which is challenging in JET with the metal wall due the need to control the W influx into the core region. By starting H-mode operation at high density, applying a high level of gas injection early during the NBI heating phase to avoid ELM-free phases, it was possible to reduce the gas puffing to very low levels ($\approx 10^{21}$e/s), achieving a high performance, low density regime (called no-gas regime in the rest of the text) with averaged $n_e\approx3\times 10^{19}$ m$^{-3}$ and Greenwald fraction of ≈0.35, amongst the lowest ever achieved in JET-ILW. Operation at such low densities allows decoupling ions from electrons, resulting in higher $T_i/T_e$ than what obtained in conventional ELMy H-modes at higher densities and similar heating power.
One of the best examples of this no-gas scenario (#94900) is shown in Fig. 1, compared to the reference ELMy H-mode discharge (#94777) at similar heating power. Both discharges are heated by 20 MW of neutral beam injection (NBI) and up to 4 MW of ion cyclotron resonance heating (ICRH). In discharge #94900 the gas puffing is switched off at 8.5 s, resulting in a strong decrease in edge density (from 7$\times 10^{19}$ m$^{-3}$ to 2$\times 10^{19}$ m$^{-3}$) and a significant increase in the density profile peaking. This is accompanied by an increase in pedestal temperatures ($T_{e,ped}\approx$1.5 keV, $T_{i,ped}\approx$2 keV), enhanced toroidal rotation ($v_{tor,0}\approx$450 km/s), improved core ion confinement ($T_{i,0}\approx$15 keV, $T_{i,0}\approx 2\times T_{e,0}$) and ELMs substantially smaller and faster than those of the reference with gas fuelling ($f_{ELM}\approx$60 and 25 Hz respectively). In the no-gas phase, the ion temperature, stored energy and neutron rate continuously increase until the appearance of a core MHD mode ($n$=4) triggered by a sawtooth crash (see sharp drop in the neutron rate at 10.5 s, Fig. 1(e)), suggesting the performance is limited by MHD rather than transport. It must be noted that this MHD event does not lead to a disruption, the discharge survives and is landed safely. Density and radiation remain essentially constant after the gas puff is switched off, indicating particle transport is fast enough to provide adequate density and impurity control. This behavior differs from what observed in the hot-ion H-mode developed in JET-C(1) where density and radiated power increased constantly during long ELM free periods, eventually leading to a radiation collapse and back transition to L-mode.
Despite the absence of gas puffing and strong electron density peaking of the no-gas scenario, which typically lead respectively to an increase in W source and to strong inward impurity convection, central impurity accumulation does not take place, the temperature of the outer target and the total radiated power are comparable to the reference discharge. Both discharges also show very similar 2D radiation patterns, with strong localization on the LFS midplane at $\psi_N$>0.8 and very low central values. Due to the much lower electron density at the pedestal top, the no-gas discharge reaches similar radiated power to the reference with a factor 4 increase in mid-Z (Ni/Fe/Cr/Cu) and high-Z (W) impurity concentrations, which in turn are the cause for the increased LOS-integrated $Z_{eff}$ measurement (Fig. 1(c)). Due to the strong localization of these impurities on the LFS midplane, their increased concentration does not affect the plasma center where $Z_{eff}$ remains < 1.4 as in the reference, so core dilution is also kept under control.
The no-gas scenario exhibits remarkably good absolute and normalized performance, albeit transient, reaching peak values of $H_{98}\approx$1.4, $\beta_N\approx$2.2, $W_{MHD}\approx$9 MJ, and this is achieved at much lower collisonality ($\nu_{e,ped}^*$< 0.1, close to ITER values) than the reference ELMy H-mode at higher density. The electron pedestal pressure is also slightly smaller than the reference, albeit at lower density and higher temperature, but there is a significant increase in core electron and ion pressures, resulting in a 35% increase in global energy confinement at the maximum stored energy. Improved transport driven by the increase in sheared $E\times B$ flow [2] is thought to contribute to both the strong peaking of the density profile and the improved performance. Additional effects associated with the large population of fast ions present in these plasmas, as shown in [3], might also play a role in the overall improved thermal transport. The added impurity control is provided by the increased ion temperature screening enhanced by the extreme toroidal rotation [4].
An especially interesting feature of the no-gas regime is the marked reduction in ELM size compared to conventional ELMy H-mode plasmas. With the decrease in edge density the type I ELMs are replaced by very small ELMs at a much higher frequency. ELM size increases as the pedestal pressure increases, but remains significantly smaller than those obtained in the reference pulse at higher density and similar heating power. We note that in those conditions the link between ELM size and pedestal collisionality and/or edge density typically found for Type I ELMs is lost [5]. Both the electron density and temperature pedestals become wider, and there is a substantial reduction in the maximum $\nabla n_e$ as the edge density decreases, resulting in a lower maximum $\nabla P_e$. Pedestal stability analysis indicates that the edge operating point is below the peeling-ballooning boundary, which might explain the absence of large type I ELMs. The underlying physics mechanisms responsible of the onset of these small ELMs are still a matter of ongoing investigation.
The new no-gas H-mode regime recently demonstrated in JET-ILW provides a valuable opportunity to study the confinement properties and ELM dynamics of high temperature plasmas with temperature and density profiles substantially different from those obtained in the conventional scenarios. With the aim of improving our understanding and increasing the accuracy of extrapolations for ITER, this scenario allows validating existing transport models and investigating the role of different physics mechanism involved in the observed improved energy confinement and impurity control.
ACKNOWLEDGEMENTS
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the EURATOM research program 2014-2018 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This research was supported in part by grant FIS2017-85252-R of the Spanish Research Agency, including ERDF-European Union funding.
References
1. The JET team, Plasma Phys. Control. Fusion 37 (1995), p. A35
2. J. Garcia et al., Plasma Phys. and Control. Fusion 61 (2019) 104002
3. J. Garcia et al., Nucl. Fusion 55 (2015) 053007
4. C. Angioni et al., Physics of Plasmas 22 (2015) 055902
5. A. Loarte et al., Plasma Phys. and Control. Fusion 45 (2003) 1549-1569Speaker: Elena De La Luna (Laboratorio Nacional de Fusión, CIEMAT, 28040, Madrid, Spain) -
State-of-the-art deep-learning disruption prediction models based on the Fusion Recurrent Neural Network (FRNN) (1) have been further improved. Here we report the new capability of the software to output not only the “disruption score,” as an indicator of the probability of an imminent disruption, but also a “sensitivity score” in real-time to indicate the underlying reasons for the imminent disruption. As an indicator of possible causes for future disruptions, the “sensitivity score” can provide valuable physics-based interpretability for the deep-learning model results, and more importantly, provide targeted guidance for the control actuators when implemented into any modern Plasma Control System (PCS). This achievement represents a significant step forward in moving from modern deep-learning disruption prediction to real-time control— a major advance since the last IAEA meeting— that brings novel AI-enabled capabilities needed for the future burning plasma ITER system.
The main findings in this paper all help address the basic issue/perception that advanced Machine Learning/Deep learning methods are generally hard to interpret. Results presented used large amounts of data from JET and DIII-D congruent with publications such as prominent deep learning NATURE paper (1). Moving beyond this work on tokamak disruptions, the current IAEA paper addresses and answers in addition to “when” a shot is going to disrupt, some compelling reasons to explain “why” it disrupts by carrying out sensitivity studies.
A new scheme as shown in Fig. 1 is introduced in which control of actuators can be engaged by AI-enabled disruption predictors. The FRNN code used 16 total – including 14 zero-D scalar signals and two 1D profile signals. In this total of 16 channels, 5 are post-processed data with the remaining 11 channels being real-time data. We have obtained the real-time “point names” for all 16 signals and have subsequently focused on training the prediction model based on the real-time data for control purposes.
We highlight here the fact that the FRNN deep learning code can be readily extended to using many more channels of information. For example, some DIII-D signals that are known to be highly relevant physics-based channels are: (i) “n1rms” – a signal correlated with the activities of the n =1 modes with finite frequency, including the neoclassical tearing modes (NTMs) and sawtooth; (ii) the bolometer data that reflects the impurity content of the plasma; and (iii) “q- min” – the minimum value of the safety factor which directly relates to important physics such as the kink modes. These considerations have strongly motivated us to include the associated channels directly into the improved performance development of the deep learning FRNN software with the goal of clearer identification of the physics most responsible for the dangerous disruption events with associated guidance toward targeted guidance for the control actuators. The potential for significant improvement over existing traditional algorithms targeting these signals for plasma condition and disruption control comes from the fact that deep-learning/AI models have the distinct advantage of being able to carry out supercomputing-enabled robust predictions for complex physical systems with a huge number of potentially correlated features, without the necessity of manual “feature engineering”. This enables the capability to deliver predictions for unseen conditions, such as new plasma parameters including those associated with future larger devices. Moreover, another key advantage of deep-learning enabled predictive capabilities is the ability to carry out forecasts significantly earlier in the evolution of the plasma state under consideration, especially when more physics-based channels are included in the model input. For example, in preliminary recent studies, the average alarm time for FRNN disruption predictions from models trained with the n=1 mode amplitude channel can be around 100 ms earlier than that from models trained without this information. The basic enabling capabilities with deep learning FRNN performance on large databases were established and explained previously (1).
Overall, the key point made in this paper is that when more physics related channels are included, valuable new insights can be gained on the mechanisms contributing to disruptions. For example,
in the right upper panel of Fig. 2, the evolution of the FRNN disruption score for DIII-D shot #164582 is displayed. While this score rises above the disruption alarm threshold at around 3000ms, such information alone would only enable the PCS to respond in a single and preprogrammed way -- such as injecting gas to terminate the shot. In contrast, our new real-time sensitivity score (shown on the right lower panel) clearly indicates that FRNN can categorize this potentially dangerous plasma condition as being impurity related, which could in turn lead to tearing modes and locked modes that did in fact show up later in the shot at around 3400 ms (as indicated by the cyan-colored line in the left top panel). Therefore, FRNN is demonstrably able to construct much more efficient real-time strategies for the PCS; i.e., in this instance, it could initiate actuator control to reduce impurities or weaken tearing modes impact. In addition, it is found that different shots could have completely different sensitivity score distributions at alarm time. Consequently, in future work, with more data from the control actuators together with the aid of proper plasma simulators, FRNN can be developed to: (i) construct a tree-shape “policy network” for the different actuators in the PCS; and (ii) update this tree at each time step according to the feedback from real-time plasma dynamics. The overall impact here would be the capability to output the best real-time strategy for keeping the plasma in a disruption-free regime for a significant duration.Summarily, in addition to providing a robust “disruption score,” the present deep learning studies compute a “sensitivity score” for each physics-connected channel (as shown in Fig. 2). In addition to studying the physics in subcategories (2) of disruptions, these “sensitivity scores” for each channel can also indicate the detailed reason for the coming disruption, and in turn provide guidance to the PCS with more precise and direct information for the control actuators.
(1) J. Kates-Harbeck, A. Svyatkovskiy, W.Tang, Nature 568 (7753), 526; (2) P.C. de Vries, et al, Nucl. Fusion, 51, 053018.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-FC02-04ER54698; DE-AC02-09CH11466
Speaker: Prof. William Tang (Princeton Plasma Physics Laboratory) -
A disruption predictor based on deep learning method is developed in HL-2A. Its structure is specially designed to deal with data from fusion devices, and shows to have a better performance on disruption prediction problem than ordinary models commonly used in computer science[1.]. Based on this deep learning algorithm, in-depth research are carried out in three aspects. Firstly, different structures of the 1.5-D layer are tried in order to find the best way to merge information from different input channels. Secondly, different preprocessing methods are tried to find a better way to deal with the discrete data in the input, e.g., power of NBI heating. These data seems to be harmful to the performance of disruption predictor without a properly preprocessing. Finally, an interpretation method is applied on the disruption predictor, which can be used to give the causes of disruptions along with the disruption alarms.
The model in this research is mainly based on a 1.5-Dimensional Convolutional Neural Network (1.5-D CNN), which is good at dealing with signals from multi-channels with great divergence. The Disruption Predictor uses shots 20000-29999 in HL-2A to train the machine learning model, and uses shots 30000-31999 to optimize hyper parameters. When tested on shots 32000-36000 in HL-2A, it reaches a True Positive Rate (TPR) of 92.2% and a True Negative Rate (TNR) of 97.5% with 30ms before the disruption. A trade-off between TPR and TNR can be realized according to the cost of false alarms and missed alarms during application by changing the alarm threshold, as showed in Fig.1. There is also a trade-off between prediction advance times and accuracy, which is showed in Fig.2.
A comparison is made between the accuracy of four different models to predict the disruptions in HL-2A. And 1.5-D CNN+LSTM shows to perform much better than other traditional CNN-based models. LSTM also shows to perform an important role to achieve better accuracy.
To solve the remaining problems in the algorithm mentioned above, and reach a better performance, in-depth research are carried out in three aspects.
The first problem is how to deal with multi-channels data with great divergence from fusion device. Although the 1.5-D CNN has solved the problem to a certain extent, since information from all channels are merged simply in one layer, this structure may not be able to find some complex multi-channels information. Since that, several different versions of 1.5-D layer are compared to find a better way to merge the information from different channels.
The second problem is about the preprocessing of data. Most of the input channels of disruption predictor are continuous variables, i.e., the distributions of their value locate in a continuous range. But there are some exceptions. For example, the power of NBI heating will always be selected from several individual values, then the sharp rising edge and falling edge will bring troubles to the training of machine learning models. Several different preprocessing method are tried to solve this problem, and the result shows that the model will perform better during the transition stage with properly preprocessing.
Finally, a model interpretation method is applied on the prediction model to decide which of the input channels are more closely related to a certain disruption. And the result shows a good coherence with the disruption causes. For example, the model shows to rely on signals from mirnov probes during MHD instability induced disruptions, and to rely on bolometer signals during radiation induced disruptions. This model interpretation method can be used to automatically give the disruption causes, which will be quite helpful for the active avoidance of disruption. To check the feasibility of this method, several main types of disruption are defined in HL-2A and a simple Bayesian classifier is developed to classify these types of disruptions.References
[1.] Z. Yang, et al, Nucl. Fusion 60 (2020) 016017
Speaker: Mr Zongyu Yang -
In recent years the multi-scale interaction between large-scale tearing modes and micro-scale turbulence has been found to be of paramount importance for thoroughly understanding the tearing mode physics and the island-induced transport, which will ultimately lead to developing a more effective method of the tearing mode control and optimizing the plasma performance in fusion devices, such as ITER [1-2]. In this work, the impact on micro-turbulence, meso-scale quasi-coherent modes (QCMs) and large-scale zonal flows as well as plasma rotations by naturally rotating $m/n$=1/1 and 2/1 islands and by externally magnetic perturbation-induced static $m/n$=2/1 island has been investigated using 2D ECE imaging and reflectometers in HL-2A and J-TEXT tokamaks [3-5]. The results indicate that for sufficient large islands there exist strong interactions between the tearing mode (TM) island and turbulence, including QCMs and zonal flows. The critical island width is in general consistent with the theoretical prediction [6].
In the case of rotating $m/n$=1/1 island generated in NBI heated plasmas at HL-2A, it is observed that for large islands ($W_c\geq$ 10$\rho_i\approx$ 4 cm) the electron temperature ($\tilde{T}_e$) and density fluctuations ($\tilde{n}_e$) are modulated periodically by the island rotation, with minimum amplitude at the O- and maximum at the X-point, respectively. The turbulence modulation is localized merely in the inner area of the island due to significant alteration of local profiles and gradient-driven turbulence. Evidence also reveals that for large islands turbulence spreading can occur across the island O-point from the region outside the island. The experimental observations of the turbulence modulation and spreading effects with large island are in general agreement with simulations for the trapped electron mode instability [3].
In the case of rotating $m/n$=2/1 island in HL-2A ohmic plasmas, it is found that for sufficient wide islands ($W_c\geq$ 4.5 cm) a quasi-coherent mode (QCM, peaked at $\sim$100-180 kHz) in density fluctuations is excited at the island boundary as the island O-point passes by, where the local $T_e$ profile is steepened, as shown in figure 1. Statistical analysis indicates that for the QCM excitation, a threshold value of $T_e$ gradient is needed and the QCM is solely observed in low density discharges, consistent with the linear ohmic confinement regime. These experimental evidences and also the linear stability calculations both suggest that the observed QCMs are driven by the trapped electron mode instability. Bi-spectral analysis further shows that there exists nonlinear coupling between the tearing mode and QCMs, whereas no nonlinear interaction is observed between the QCMs and ambient turbulence. The results verify that the observed QCMs are linearly driven by locally enhanced $T_e$ gradient with large islands, but not driven by ambient turbulence via nonlinear energy coupling [4].
In J-TEXT ohmic plasmas, the plasma rotations, geodesic acoustic mode (GAM) zonal flows and turbulence are found to be significantly modified by a static $m/n$=2/1 island induced by externally applied resonant magnetic perturbations (RMP). Whereas after the island formation the edge toroidal rotations shift from the counter-$I_p$ to co-$I_p$ direction and the perpendicular rotations from the electron to ion diamagnetic drift direction, both of rotation changes do not show much difference between the island O- and X-point. However, the turbulence level at the O-point is substantially lower than that at the X-point. The amplitude of the GAM zonal flows measured outside the island is damped and the peak frequency slightly increase after the RMP. In addition, at certain island size (e.g., $W\approx$ 3.8 and 4.5 cm in figure 2) the nonlinear coupling among ambient turbulence inside the island is considerably enhanced through the inverse energy cascading. However, this situation does not occur for a narrower or wider island, as depicted in figure 2. These results suggest a profound influence of the island size on the nonlinear interplay of turbulence and turbulent transport [5].
References:
[$1$] L. Bardóczi et al 2016 Phys. Rev. Lett. 116, 215001
[$2$] K. Ida et al 2018 Phys. Rev. Lett. 120, 245001
[3] M. Jiang et al 2019 Nucl. Fusion 59, 066019
[4] M. Jiang et al 2020 submitted to Nucl. Fusion
[5] M. Jiang et al 2019 Nucl. Fusion 59, 046003
[6] R. Fitzpatrick 1995 Phys. Plasmas 2, 825Speaker: Min Jiang (Southwestern Institute of Physics) -
Future DT operation in ITER and DEMO will face a significant number of challenges. From the physics point of view, the change from DD to DT plasmas is poorly understood. There are indications that the core confinement, the ELM behavior, the pedestal confinement and the Scrape of Layer behavior can significantly change from DD to DT. From the operational point of view, scenarios with high enough alpha power production are essential to demonstrate that efficient electrical power generation is possible but this also requires techniques for power exhaust, ELM and impurity sources and accumulation control in the presence of enough additional input heating power and a metallic wall environment.
In order to address the relatively scarce knowledge of DT plasmas and in support of safe future DT operations, JET has developed a scientific program in DD, TT and DT with the aim of supporting and minimizing the risks of the transition from DD to DT plasma operation in ITER 1. Such program helps to understand and document the physics characteristics of DT plasmas but also to provide scenario solutions integrating several key aspects such as core transport, impurity accumulation avoidance, power exhaust, enough fusion power generation to reveal alpha particle effects and potential difficulties for operation.
To this end, JET has gone through an upgrade of diagnostics and heating systems, in particular, Neutral Beam Injection (NBI) input power, which has delivered the record of 32MW. With such power, high performance scenarios have been developed following two main routes, i.e., the baseline scenario with $q_{95}$~3, β$_N$~2, H$_{98}$(y,2) ~1 at high current and magnetic field and the hybrid scenario with $q_{95}$~4, β$_N$~2.4, H$_{98}$(y,2) ~1.3 at reduced current and low central magnetic shear.
Compared to previous campaigns [2], a new baseline JET DD neutron rate record in the ITER Like Wall (ILW), R$_{NT}$=4.1x10$^{16}s^{-1}$, has been attained as shown in figure 1. This high performance was possible at current Ip=3.5MA and magnetic field B$_T$=3.35T through a complex non-linear interaction between edge and core plasma regions. From the pedestal side, high frequency ELM’s help to flush impurities. This is obtained by combining 50%/50% particle source with pacing pellets and neutral gas puff as it allows simultaneously good confinement and impurity flushing. The pedestal density is relatively low for such a high current, ~4x10$^{19}m^{-3}$, but the strong peaking, which is significantly high in these conditions of low core NBI fueling, leads to a core density ~1x10$^{20}m^{-3}$. Rotation and its shear are high both for the core and the pedestal with a Mach number reaching 0.6 at mid-radius. The ExB shearing of such plasmas, which usually are in the ITG regime, have been shown to be important for the decoupling of ions and electrons [3], with Ti/Te~1.7 in the best performing discharge at H$_{98}$(y,2)=1.0, and for increasing the density peaking through an increase of the inward pinch [4]. The radiation power is mostly localized in the low field side pedestal for the whole discharge and never penetrates in the inner core, which allows long and stable flat-top pulses up to 5s.
This new baseline scenario at high Ip and with relatively low density and high temperature at the pedestal offers an attractive alternative to hybrid scenarios as no specific q profile tailoring or core improved confinement is required in order to obtain high neutron rate. Although the Greenwald fraction is moderate, G$_{fr}$=0.65, a key point is the density peaking, which provides core densities compatible with high fusion power. Furthermore, the ion pedestal temperature, 1.3keV, provides a route for avoiding W accumulation through neoclassical screening. The relevance of this scenario for ITER is strong for the long pulse operation program [5] as it can soften some of the constraints for the hybrid scenario in terms of off-axis current.
The hybrid route has also stablished a new neutron record generation compared to previous campaigns, R$_{NT}$=4.7x10$^{16}s^{-1}$. This has been obtained in conditions of reduced β$_N$=2.4 by working at higher magnetic field, i.e. 3.45T and $q_{95}$=4.5 rather than 2.8T used in previous campaigns with $q_{95}$=4.0. Actually, temperature and density profiles are essentially identical for both magnetic field showing that, at least in this regime, high confinement appears to be largely independent of the magnetic field. However, it has been also obtained that β$_p$≥1 seems to be necessary to get good confinement. Such hybrid plasmas are type I ELMy H-modes with Ti/Te ~1.5. They are prone to suffer from impurity accumulation when the low gas used in order to obtain high confinement leads short phases of too low ELM frequency. Further techniques, such as pacing pellets, will be used to maintain an adequate level of ELM frequency, and hence impurity flushing, but at reduced gas injection.
A key element that has significantly improved the robustness of both scenarios is the use of real time control techniques which have been applied for ELM, fueling and impurity accumulation control. With such techniques, the disruptivity has been significantly decreased to ~9% for the hybrid and to ~25% for the baseline when previously reached 60%.
An extensive exercise of ‘predict first’ modelling has been carried out with the aim of predicting the DT fusion performance expected in the future JET-DT campaign. As a first step, DT equivalent fusion power from recent plasmas at high power has been compared to past predictions from lower power discharges showing a reasonable agreement when using validated models, such as TGLF or QuaLiKiz, for core transport [6,7].
DT extrapolations to the maximum available power at JET show that 11-16MW of fusion power are possible both for the baseline and hybrid routes. In particular, favorable core isotope effects are found for the hybrid scenario in conditions of low turbulent transport, such as strong rotating and high core thermal and fast ion beta plasmas.
Finally, in baseline plasmas, electron heating by alpha power can be dominant when ICRH is temporary removed. Such characteristic can be used to demonstrate alpha heating in ITER relevant conditions for the first time. These predictions will be compared to TT and DT plasmas when they will be obtained in the upcoming campaigns at JET.References
- E. Joffrin et al 2019 Nucl. Fusion 59 112021
- L. Garzotti et al 2019 Nucl. Fusion 59 076037
- Hyun-Tae Kim et al 2018 Nucl. Fusion 58 036020
- J Garcia et al 2019 Plasma Phys. Control. Fusion 61 104002
- A.R. Polevoi et al 2015 Nucl. Fusion 55 063019
- J. Garcia et al 2019 Nucl. Fusion 59 086047
- F. Casson et al., submitted to Nucl. Fus.
Speaker: Jeronimo Garcia (CEA IRFM) -
Dimensionless experiments test the invariance of plasma physics to changes in the dimensional plasma parameters, when the canonical dimensionless parameters, such as rho_star, nu_star, β, q, ... are conserved [1], [2]. In particular, isotope identity experiments exploit the change in isotope ion mass A = mi/mp to obtain plasmas with identical dimensionless profiles in the same tokamak. In order to keep rho_star, nu_star, β and q fixed when also varying A, the plasma current, toroidal magnetic field and the density and temperature must scale, respectively, as I_P, B_T ~ A^(3/4), n ~ A and T ~ √A [3]. However, conditions at the plasma boundary, such as influx of neutral particles, may introduce additional physics, potentially invalidating this approach. Moreover, although the isotope mass appears explicitly only in the parameter rho_star_i, changing A in experiment will affect all plasma kinetic profiles, both in the core and edge plasma, therefore achieving an isotope identity is not trivially expected a priori.
An isotope identity pair was first obtained in JET with C wall (JET-C), with Type I ELMy H-modes in H and D [3]. Remarkably, the scaled thermal energy confinement time Ωi τE,th = B τE,th/A (with Ωi the ion cyclotron frequency), scaled ELM frequency, A f_ELM/B, and scaled sawtooth frequency, A f_saw/B, were all matched, confirming the invariance principle throughout the entire plasma radius despite the different physics processes governing the edge, core confinement and plasma centre regions [3]. We have revisited the isotope identity technique in JET with Be/W ITER-like wall (JET-ILW) with H and D plasmas, both in L-mode and type I ELMy H-mode regimes. Additional aspects are addressed, compared to [3]: i) improved edge kinetic profiles enable careful assessment of profile similarity in the H-mode pedestal region, whose confinement and linear MHD stability is strongly affected by neutrals in JET-ILW [4], [5]; ii) we also seek similarity of the Mach number profile, M ~ ω_tor R √mi/√ kTe (toroidal rotation velocity normalized to ion sound speed) and investigate the role of rotation and ExB shear on the heat and particle transport channels, as ExB shear is the relevant parameter for stabilization of Ion Temperature Gradient (ITG) turbulence, which dominates in the core of NBI-heated JET-ILW plasmas.
An L-mode isotope identity pair was obtained in H (1.44MA/1.74T) and D (2.5MA/3.0T) NBI-heated plasmas [6]: the rho_star, nu_star, beta, q and Ti/Te (= 1) profiles, as well as the line averaged Zeff (=1.4) were all matched in the core confinement region, where the dominant instabilities are ITG modes for both isotopes. The dimensionless thermal energy confinement time, Ωi τE,th, and the scaled core plasma heat diffusivity, A χeff/B_T, are identical in H and D within error bars, indicating lack of isotope mass dependence of the dimensionless L-mode τE,th. This result is broadly consistent with the weak, positive isotope scaling of the dimensional τE,th in NBI heated JET-ILW L-modes [7], which originates at the plasma edge, where both particle and heat transport are larger in H than in D [7], [8]. The scaled input power required to obtain the identity, P_abs/B_T^(5/3), is indeed consistent with Bohm-like transport [9]. Predictive flux driven simulations of the H and D identity pair with JETTO-TGLF are in very good agreement with experiment for both isotopes: the stiff core heat transport, typical of JET-ILW NBI heated L-modes, overcomes the local gyro-Bohm scaling of gradient-driven TGLF, explaining the lack of isotope mass dependence in the core confinement region of these plasmas [6]. Although the M-profiles are not matched in H and D (but within 50% from each other), the effect of E×B shearing on the predicted heat and particle transport channels is found to be negligible for these low beta and low momentum input plasmas [6].
In type I ELMy H-modes, an isotope identity pair was obtained in H (H-NBI, 1.0MA/1.0T) and D (D-NBI, 1.7MA/1.7T), with low triangularity (delta) plasma shape. The companion tritium plasma (T-NBI, 2.3MA/2.3T) is planned for the upcoming JET-ILW campaign in tritium. The scaled ne and Te profiles were matched in H and D over ~ 25 x τE,th’s - both in the core and pedestal regions (see Figure 1) - as well as the q, Ti/Te (=1) profiles and line averaged Zeff (= 1.4). Matching the scaled ELM frequency (A f_ELM /B_T = 54 Hz/T) - using f_ELM control via feedback on gas injection in the D discharge - was key to achieve similarity of the scaled pedestal density profiles. This is because ne,PED decreases with increasing f_ELM in low delta JET-ILW type I ELMy H-modes [4], where the ELM losses are primarily convective. While the input power required to obtain the isotope identity, P_abs ~ B_T, is consistent with gyro-Bohm transport [9], Ωi τE,th increases strongly with A, in line with the strong isotope mass scaling derived from the regression of the dimensional energy confinement time of H and D type I ELMy H-modes, τE,th ~ A^0.4 [7], [10]). The Mach-number profiles are not identical in H and D, but the ExB shear is similar for both isotopes, suggesting the latter as the relevant parameter (and not M) for achieving the identity in ITG dominated H-modes.
In the isotope identity experiments presented here we have sought a match in the ion rho_star, as implicitly assumed in the theory of [1], therefore the electron rho_star profile, rho_star_e = (me Te)^(1/2)/(a B_T) is not matched. As verified experimentally, this is justified in our case because the plasmas are ITG dominated in the confinement region. On the other hand, in plasmas where electron scale turbulence dominates, matching rho_star_e and not rho_star_i would be required. Thus, care must be taken when using wind tunnel experiments to extrapolate to future experiments [11]. It is doubtful whether this approach would be valid in plasmas where ion – and electron-scale turbulence co-exist with similar strength.
References: [1] Connor J W and Taylor J B 1977 Nucl. Fusion 17 1047; [2] Luce TC et al., 2008 Plasma Phys. Control. Fusion 50 043001; [3] Cordey J G et al. 2000 Plasma Phys. Control. Fusion 42 A127; [4] Maggi CF et al., 2015 Nucl. Fusion 55 113031; [5] Frassinetti L et al. 2019 Nucl. Fusion 59 076038; [6] Maggi C F et al. 2019 Nucl. Fusion 59 076028; [7] Maggi CF et al., 2018 Plasma Phys. Control. Fusion 60 014045; [8] Bonanomi N et al., 2019 Nucl. Fusion 59 126025; [9] Cordey JG et al. 1996 Plasma Phys. Control. Fusion 38 A67; [10] Weisen H et al., 2018 IAEA FEC; [11] Garcia J et al., 2019 3rd Asia-Pacific Conference on Plasma Physics
Acknowledgements. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission
Speaker: Costanza Maggi (CCFE) -
The Final Design Review (FDR) of the ITER Plasma Control System (PCS) for First Plasma will be held in July 2020 following the conceptual and preliminary designs [1,2] to prepare for First Plasma operation scheduled for the end of 2025. ITER operation follows the Staged Approach of the ITER Research Plan (IRP) [3]. The main goals of the First Plasma campaign include achieving a plasma current Ip > 100 kA for a duration > 100 ms and possibly up to 1 MA for up to 3 seconds duration at half toroidal field of 2.65 T, followed by engineering operation to test the main superconducting coils up to full current. The First Plasma PCS design also includes the full architecture of the PCS that allows implementing high performance control algorithms in the future without changing the PCS architecture.
This phase includes control of the plasma current, radial plasma position, and plasma shape for a nearly circular plasma limited on either the inboard or outboard temporary stainless steel poloidal limiters (Figure 1). Disruption force limits of the attachments of these temporary limiters to the vacuum vessel require control algorithms to ensure that the plasma current remains < 1 MA at the nominal toroidal field of 2.65 T at R = 6.2 m. Following the first plasma attempts at 2.65 T, there will be an engineering operation phase to commission the superconducting central solenoid (CS), poloidal field (PF), and toroidal field (TF) coils to full current, up to 5.3 T nominal TF. If there are difficulties achieving first plasma at 2.65 T, the IRP and the PCS design foresees an option possibly to achieve first plasma at 5.3 T nominal TF. This would be easier than at 2.65 T because of the increased connection length to achieve plasma breakdown and the improved ionization and absorption efficiency of ECH power at the fundamental gyrotron frequency [4].
The plasma initiation phase also includes control of the gas injection of hydrogen to achieve and maintain the required prefill pressure in the range of 0.5 – 1 mPa and the initial density rise as well as the control of electron cyclotron heating (ECH) at 170 GHz. The low startup electric field ~0.3 V/m, large stray fields due to eddy currents in the vacuum vessel up to 1.5 MA, and the large volume and neutral fueling source limit the allowed prefill range and increase the required power for breakdown. The injected ECH power will be up to 5.8 MW for up to 300 ms from 7 of 8 available gyrotrons from one upper launcher with four beams crossing the breakdown region about 40 cm below the midplane and three beams crossing about 40 cm above the midplane (Figure 1). The beams will be reflected off mirrors mounted to the inner wall into an absorbing beam dump in an equatorial port to avoid scattering stray ECH power around the vessel that could damage in-vessel components. The absorbed power of a single pass is only expected to be of order a few percent of the injected power at half field. Feedback of the prefill pressure timed with the ECH pulses, the PF null, and vertical field swing will be used for robust plasma initiation control. This phase also includes initial exception handling algorithms for possible plant system and diagnostic faults, excess stray ECH power, plasma initiation failures, and start-up runaway electrons.
All control algorithms and synthetic diagnostic models for this phase, including magnetics, H, a single radial chord interferometer, and hard x ray monitor for runaway electrons have been developed in the Plasma Control System Simulation Platform (PCSSP) [5] in Matlab/Simulink for thorough testing in simulation prior to operation.
The PCS design is documented in the PCS database (PCSDB) using Enterprise Architect that records all requirements and tracks their compliance in the design [6]. The controllers are all documented in the PCSDB with explicit links between them and the PCS architecture including supervision, events, and exception handling, which allows checking the impact of any design changes on all aspects of the PCS design. The database includes performance requirements and the test results carried out to assess the design as well as the use cases and commissioning procedures.
Extensive simulations of plasma breakdown and First Plasma scenarios have been carried out with the TRANSMAK [7] and DINA [8] codes including vacuum vessel eddy currents and simplified plasma transport modeling of sputtered impurities from the stainless steel first plasma protection components. The simulations show that plasma initiation is sensitive to the impurity content, but, if breakdown is achieved, the plasma current will rise to ~ 0.5 MA and by switching the CS and PF voltages to zero, ensure Ip remains < 1 MA to protect vacuum vessel limiter housings.1 J. A. Snipes, et al., Fusion Eng. and Design 89 (2014) 507.
[2] J. A. Snipes et al., Nucl. Fus., 57 (2017) 125001.
[3] ITER Research Plan, ITER Technical Report ITR-18-03 (2018), https://www.iter.org/technical-reports?id=9 .
[4] P. C. de Vries and Y. Gribov, Nucl. Fus. 59 (2019) 096043.
[5] M. Walker, et al, Fusion Eng. and Design 96-97 (2015) 716.
[6] M. Cinque, at al., IEEE Trans. Plasma Sci., doi: 10.1109/TPS.2019.2945715 (2019).
[7] Mineev A.B. et al, 25th IAEA Fusion Energy Conf. (St. Petersburg, Russia, 2014) PPC/P3-20 (www-naweb.iaea.org/napc/physics/FEC/FEC2014/fec2014-preprints/255_PPCP320.pdf).
[8] Khayrutdinov R.R. and Lukash V.E. 1993 J. Comput. Phys. 109 193.Speaker: Joseph Snipes (ITER Organization) -
$\qquad$ Characterizing and understanding the power threshold conditions for ITER to achieve H-modes ($P_{LH}$) is a major goal of a series of L-H transition experiments undertaken at JET since the installation of the ITER-like-wall (JET-ILW), with Beryllium wall tiles and Tungsten divertor [1,2,3,4]. In this contribution we report on results from L-H transitions studies in H, D and new almost pure $^4Helium$ plasmas, and compare the results with ITER predictions. The most notable result is that the density at which $P_{LH}$ is minimum, $n_{e,min}$, is considerably higher for $^4He$ than for D, and strongly influenced by shape.
$\qquad$ A detailed analysis of the pre-transition $E_r$ profiles across the ne scan in D and $^4He$ find matching qualitative changes in the $E_r$ profile. In high field NBI heated D plasmas, we report on power balance analysis and its impact on $n_{e,min}$. Modelling of the plasma SOL does show differences in the heat flux required to drive a transition between H and D (in the high $n_e$ branch), and $^4He$ plasmas are also being studied.Characterizing the L-H transition power threshold for H, D, $^4He$: $n_{e,min}$, ion heat flux, $E_r$
$\qquad$ The interest on $^4He$ plasmas is not purely academic, and our data brings surprises. The ITER Research Plan includes a low toroidal field Pre-Fusion Operating Power phase with either Hydrogen or Helium plasmas in order to study H-modes as early as possible, before the nuclear phase that starts with D plasmas. A prediction of $n_{e,min}$, was made inspired on the studies of Ryter [5], who observed in AUG that a sufficient edge ion heat flux is necessary to achieve a sufficient radial electric field (shear). Assuming pure electron heating in ITER, $n_{e,min}$ has been evaluated on the basis of 1.5-D transport modelling as the density at which the ratio of edge ion power flux to total edge power flux starts to saturate with increasing density. The result of this modelling is that $n_{e,min}$~$0.4 n_{GW}$, independent of the ion species [6]. The transition condition in that model is itself based on the assumption that the He power threshold, $P_{LH}(He)$, is 1.4 x $P_{LH}(D)$, while $P_{LH}(H)$=2x$P_{LH}(D)$ [7].
$\qquad$ We find that for Horizontal Target plasmas the estimate of $n_{e,min} \sim 0.4 \times n_{GW}$ is in agreement with the D data, but $n_{e,min}$ is closer to $0.5 \times n_{GW}$ for H, and to 0.6 for Helium. NBI heated H plasmas have higher $P_{LH}$ than RF heated ones, the reasons are still being investigated. The data points with no black outline correspond to transitions during an NBI blip, so $P_{LH}$ is probably higher. In fact radiation is considerably higher for the dominantly RF-heated Horizontal Target Helium plasmas at low density, so the auxiliary power required for the L-H transition to take place is lower for H than for He below 3.3×$10^{19} e/m^3$. Above $n_{e,min}$(He), D and He have similar $P_{LH}$, below the Martin scaling, while H has a much higher $P_{LH}$.
In the 1.8 T Corner dataset $n_{e,min}$ is not so easily identified. Above 0.4$n_{GW}$, $P_{LH}$ in Corner is generally higher than in Horizontal Target for all species, approaching the Martin scaling for D and He, much higher for H.
$\qquad$ The strong shape effect shown in all L-H transition datasets at JET is in apparent contradiction with the ion heat channel determining $n_{e,min}$. A detailed study of the relation between ion heat flux and $n_{e,min}$ in 3T, 2.5 MA D plasmas, now with $T_i$ measurements, is underway [9]. We find the e-i exchange term is subdominant and unlikely to determine $n_{e,min}$.
$\qquad$ In a dataset with Horizontal Target, 2.4 T, 2 MA, NBI-heated plasmas (not shown), we find that $n_{e,min}(He) \sim 0.7 \times n_{GW}$, while $n_{e,min}(D) \sim 0.4 \times n_{GW}$. Above $n_{e,min}(He)$, D and He have similar $P_{LH}$. In this case we are attempting to reproduce the ITER transport models and $P_{LH}$ predictions and contrast them with the data. For these plasmas Doppler reflectometry shows that the $E_r$ profile of the low ne branch for both D and He plasmas has a modest $E_r$ well inboard of the separatrix and a sharp peak further out, while the high density branch has a clear $E_r$ well, but no peak near the separatrix.
$\qquad$ DIII-D results show a ~30% increase in $n_{e,min}$ of He plasmas relative to D [10], lower than our 50% shift. AUG studies show no difference in $n_{e,min}$ between H, D and He, and the same $P_{LH}$ for D and He plasmas[11]. In AUG $H+^4He$ mixtures [14], more than 20% $n_{He}/(n_{He}+n_D)$ is needed to see a change in $P_{LH}$(H), while <10% suffices in JET NBI heated plasmas [4]. C-Mod results [12] show He data in the low $n_e$ branch for $n_e<0.3 \times n_{GW}$, while in D $P_{LH}$ increases with density, indicating a shift in $n_{e,min}$. Above $n_{e,min}(He)$, $P_{LH}$ in JET-ILW is similar for D and He, therefore the increase in $P_{LH}$ due to higher $n_{e,min}$ is compensated by the lower power required to access it, since ITER had assumed $P_{LH}(He)$=$1.4\times P_{LH}$(D).
Simulations of L-H transitions for hydrogen isotopes with the HESEL[15] model find that $P_{LH}$ decreases with increasing mass number like $A^{-1.2}$. Results in $^4{He}$ plasmas are expected soon.
Summary and Outlook:
Our results question the logic that supports He for access to H-mode in the early operating phase of ITER, but not necessarily the final power estimate. Detailed analysis is on-going, to provide better understanding of the mechanisms involved and produce an improved prediction. Novel $E_r$ measurements will enable a more detailed understanding of L-H transitions in D and $^4He$ plasmas.
A Tritium campaign is planned at JET for summer 2020. We expect to obtain L-H transition data for pure Tritium, H+T and H+$^4He$ mixtures. This should inform future experiments in JET and ITER.
References:
[1] CF Maggi et al 2014 Nucl. Fusion 54 023007
[2] E Delabie, Proc. of the 25th IAEA FEC, Saint Petersburg, Russia, EX/P5 (2014)
[3] J Hillesheim et al Proc. of the 26th IAEA FEC, Kyoto, Japan, (2016)
[4] J Hillesheim et al Proc. of the 27th IAEA FEC, Gandhinagar, India (2018)
[5] F Ryter et al 2013 Nucl. Fusion 53 113003
[6] ITER Research Plan within Staged Approach, ITR-Report 18-003 (2018) p 351
[7] D McDonald et al, Plasma Phys. Control. Fusion 46 519 (2004)
[8] U. Kruezi et al 2020 JINST 15 C01032
[9] P. Vincenzi et al., 46th EPS conf., P2.1081 , Milano, Italy (2019);
[10] Gohil et al, Nucl. Fusion 51 (2011) 10;
[11] F Ryter et al, Nucl. Fusion 54 083003 (2014);
[12] C.E. Kessel et al 2018 Nucl. Fusion 58 056007;
[13] C. Silva, to be presented at EPS 2020;
[14] U Planck, submitted to NF
[15] Nielsen AH et al, Physics Letters A 379 (2015) 3097–3101.Speaker: Dr Emilia R. Solano (Laboratorio Nacional de Fusion, CIEMAT, SPAIN) -
Introduction
Electron cyclotron systems of fusion installations are based on powerful millimetre wave sources – gyrotrons, which are capable to produce now megawatt microwave power in very long pulses. Gyrotrons for plasma fusion installations usually operate at frequencies 40-170 GHz. Requested output power of the tubes is about 1 MW and pulse duration is between seconds and thousands seconds (depending on plasma machine parameters). To provide operation with indicated parameters the gyrotrons have very large transverse cavity sizes, output barrier windows made of CVD diamond discs, electron beam collectors apply effective energy recovery. The main partners in the development of gyrotrons in Russia are the Institute of Applied Physics and industrial company GYCOM Ltd. The most developed gyrotrons are now the tubes for ITER which demonstrate parameters corresponding to ITER requirements. High-level parameters were also achieved with long-pulse 140 GHz gyrotrons developed for EAST and KSTAR installations. Some steps were done in development of higher frequency (230-700 GHz) gyrotrons for future plasma installations and for plasma diagnostics. Novel ideas were proposed to enhance gyrotron operation.
170 GHz gyrotron for ITER
In ITER installation there will be 24 gyrotron systems with 1 MW power each. Russian contribution consists of 8 gyrotron systems. ITER requirements include also high efficiency of the gyrotrons over 50%, possibility of power modulation with frequency up to 5 kHz, compatibility of the gyrotron complex with ITER control system. Development of the gyrotron system for ITER is based on solution of many very difficult scientific and engineering problems. At present the main purpose of the system modification is an enhancement of the system reliability and implementation of all gyrotron systems into ITER machine and its control and safety system.
In May, 2015 a Russian Prototype of ITER Gyrotron System was completed and its operation was demonstrated. The system includes gyrotron oscillator, liquid-free superconducting magnet, supplementary magnets, several electric power supplies, cooling systems control and protection systems, and other auxiliary units. The gyrotron system shows reliable operation with required parameters. In 2016-2019 three serial gyrotron systems were fabricated. All these three ITER gyrotron systems showed reliable operation in 1000 second pulses at megawatt power and efficiency higher than 50% (fig.1). The gyrotron microwave beam fed with low losses the corrugated HE11 waveguide of 50 mm diameter. The measured X-ray radiation and stray microwave radiation do not exceed safety levels. One more 170 GHz ITER gyrotron was delivered for EU team for testing microwave components.
Higher power and higher frequency gyrotrons
Development of a higher power gyrotron in Russia is going on along two directions: power enhancement in well tested gyrotron operating at TE25.10 mode and development of a new gyrotron with a new operating mode – TE28.12. Detail analysis of the test results showed that a slightly modified ITER gyrotron prototype is capable to operate at power 1.2 MW. First tests of the modified tube are rather encouraging: microwave power 1.2 MW at MOU output was demonstrated in 100 second pulses with efficiency of 53%. The gyrotron prototypes with TE28.12 operating mode were tested at 1.2 MW power with pulses up to 500 seconds.
Calculations of higher frequency gyrotrons for future plasma installations (230-300 GHz) were done. The calculations show that 0.2-0.3 MW gyrotrons can be made on the base of easily commercially available 10T magnets with 100 mm bore diameter. 1 MW tubes require at least 150 mm diameter magnets. 0.2 MW/ 260GHz/ CW tube is now in fabrication. Microwave power of 200 kW was demonstrated with a 660 GHz pulse gyrotron. CW megawatt power gyrotron with 230 GHz frequency is under development. 10 T magnet for the gyrotron with 150 bore is in fabrication by JASTEC. Simulations also show that using the gyrotron frequency stabilization and oscillator phase locking helps to provide stable gyrotron operation at very high modes (as TE56.24), at very high frequencies (as 350 GHz) with megawatt power. are encouraging. The latter parameters are required for future plasma machines with high magnetic fields.
References
- ITER NEWSLINE. 26Feb, 2018. https://www.iter.org/newsline/-/2931
- http://www.iterrf.ru/
- M.K.A. Thumm, G.G. Denisov, K. Sakamoto and M.Q. Tran. High-power gyrotrons for electron cyclotron heating and current drive. Nucl. Fusion 59 (2019) 073001 (37pp).
Speaker: Prof. Grigory Denisov (Federal Research Center Institute of Applied Physics RAS) -
This paper presents a progress of the achievement of performance tests of ITER-gyrotrons developed in QST and design of dual-frequency (170 GHz and 104 GHz) gyrotron to enhance various operation scenarios in ITER such as characteristics studies of H-mode/ELM at low magnetic field. Major achievements of the ITER gyrotron developments are as follows: (i) Manufacturing of 6 out of 8 sets of ITER gyrotrons was completed. Factory acceptance test (FAT) in QST has been progressed and 2 of 6 gyrotrons achieved required specifications such as 1 MW / 300 s / 50 %, 5 kHz modulation with ≥ 0.8 MW etc. The 50 mm diameter waveguide transmission-line and a matching optics unit (MOU) for ITER were newly introduced to perform the operation test at the same environment as ITER-site and excitation of HE11 mode purity of ≥ 95 % at the waveguide inlet was also successfully demonstrated, satisfying the requirement. (ii) Design of dual-frequency gyrotron, which is able to operate continuous wave of 1 MW power, was successfully completed.
Introduction
An ITER electron cyclotron heating and current drive system will be installed for initial breakdown, assist of electron heating during current ramp-up, on/off-axis current drive for the steady-state operation, plasma instability control using 24 sets of 170 GHz gyrotrons. The required specifications of the ITER gyrotron are 1 MW/continuous wave (CW) with efficiency of 50% and 5 kHz modulation capability with ≥ 0.8 MW. The QST procures eight sets of the gyrotron and four of them are planned to be used for first plasma. Six gyrotrons have been manufactured in the period of 2016 ~ 2019. In FEC2018, FAT results of a first ITER gyrotron were reported [A]. After the first plasma phase, low field operation at 1.8 T is being considered to evaluate threshold power for a H-mode. If the H-mode plasmas are generated, some issues such as a validity of L-H scaling, an ELM control and divertor heat loads will be demonstrated in early stage. The 1.8 T operation desires ~100 GHz RF for plasma breakdown, ramp up, heating and current drive in second harmonics and 170 GHz in third harmonic. In the QST, the multi-frequencies 1MW oscillation at 104 GHz / 137 GHz / 170 GHz / 203 GHz were demonstrated using a proto-type ITER gyrotron [B] However, the pulse length was only a few seconds for 104GHz and 137GHz. The design study of a dual frequency (104 GHz and 170 GHz) gyrotron to achieve long pulse operation of 1 MW power has been carried out in 2019.(i) Operation performance tests of ITER gyrotrons
The FAT of the 2nd ITER gyrotron was completed in 2019. The output power of 1.05 MW and total efficiency of 50.6 % in the long pulse operation was achieved. The oscillation frequency was 169.85 GHz. 20 similar operations were repeatedly performed 20 times with 25 % duty cycle and the operation reliability of 95 % was demonstrated. In 5 kHz full-power modulation, the maximum output power of 0.90 MW was achieved. As summarized in Fig.1, the 2nd gyrotron demonstrated the same performance as the 1st gyrotron.The FAT of two gyrotrons were carried out using a prototype MOU and 63.5 mm diameter waveguides, which were assumed in the previous specifications. The integrated test introducing ITER relevant waveguides (50 mm diameter) and the MOU was carried out for the third gyrotron to simulate the operation at the ITER-site. As shown in Fig.2, a parabolic mirror and a phase-correction mirror in the MOU were newly designed to focus the beam size and flatten the phase of electric field at the waveguide inlet. It was shown that the mode purity of HE11 exciting at the waveguide inlet was 95.5%, which was comparable to the design value of 96.1 %. The power transmission efficiency in the MOU was 95.5 %, which was equivalent to the design value. These results satisfied the ITER requirement. At present, the 3rd ITER gyrotron is under commissioning and the output power of 1 MW for 100 s has been already achieved.
(ii) Design optimization of dual-frequency gyrotron of 170 GHz and 104 GHz
Issues for improving operation performance at 104 GHz were RF loss scattered in the gyrotron, which was about three times larger than at 170 GHz, and the spread of RF-sidelobe. To achieve the lower loss-power at 170 GHz and 104 GHz, Gauss mode content and directivity of radiated beam at an aperture of internal mode converter must be improved by modifying the inner wall-surface of the mode converter. In addition, a beam power distribution larger than -20 dB, which has non negligible power-level, should also be within the window aperture by modifying the curvature and the tilt of internal four mirrors located above the mode converter. As the results, the Gauss mode content achieved 95.4 % (170 GHz) / 97.5 % (104 GHz), much better than the ITER gyrotron, which are 94.5 % (170 GHz) / 90.7 % (104 GHz). These power transmission efficiency in the gyrotron achieved 98.8 % comparable to the design of ITER gyrotron. Both beams pass through the window center and are within -20 dB as shown in Fig.2. Moreover, two mirrors in the MOU were designed to improve the effective coupling of the beams to a 50 mm-diameter waveguide. Since the discrepancy of radiation angles at the window becomes only 0.1 º by design optimization of the mode converter, the same MOU mirrors for both beams are applicable. The coupling efficiency from Gaussian to HE11 mode of 95.6 % (170 GHz) and 96.1 % (104 GHz) has been obtained. As shown in Fig.2, the total transmission efficiency between the mode converter and the waveguide inlet is 95 % (170 GHz) / 93 % (104 GHz), higher than the ITER gyrotron design of 94 % (170 GHz) / 81 % (104 GHz). The blank circles in the Fig.3 represent the experimental power transmission efficiencies at the mode converter, at the output window, and at the waveguide. The experiment agrees with the design within error of 1-2%, it shows that the new design promises the long pulse, 1 MW operations at both 170 GHz and 104 GHz.Conclusion
Manufacturing of 6 sets of ITER gyrotrons and FAT of 2 sets were completed. The output power of 1 MW with the efficiency of 50 % for 300 s operation and fast power modulation of 0.9 MW for 60 s were achieved. New MOU demonstrated HE11 mode purity of 95.6 %. The design of a dual-frequency (104 GHz and 170 GHz) gyrotron comparable to the same performance as the current ITER gyrotron was successfully completed.References
[A] Y. Oda, et al., Nuclear Fusion 59 086014 (2019).
[B] R. Ikeda, et al., J Infrared Milli Terahz Waves 38, 531 (2017).Speaker: Dr Ryosuke Ikeda (National Institutes for Quantum and Radiological Science and Technology) -
The mitigation of thermomechanical and runaway loads during disruptions and Vertical Displacement Events (VDEs) in ITER is essential for the project to execute the ITER Research Plan culminating (1) in the demonstration of the fusion power production goals (Q = 10 inductive operation for 300-500 s and Q = 5 for 1000 s and in steady-state up to 3000 s). To mitigate these loads ITER is equipped with a Disruption Mitigation System (DMS). The original baseline concept for was found insufficient to provide the required degree of mitigation following a review by experts from the ITER Members in 2017 and this lead to major change in the development of the DMS for ITER both in the concept, now based on a Shattered Pellet Injection (SPI), and the design approach, now led by the ITER Organization for integration and design and with support for ITER Members institutes for R&D organized under a Task Force to address open detailed design issues. This presentation reports on progress on the ITER DMS technical design and integration while other presentations at this conference address specific tokamak experiments and modelling activities to address open R&D design issues (2 , 3, 4).
The DMS is presently at the conceptual design level with its interfaces with other systems fixed at the preliminary design level not to affect the development of interfacing systems towards their final design or manufacturing level. The DMS utilizes toroidally distributed Shattered Pellet Injectors integrated in three equatorial and three upper port plugs (12 in EP02, 6 in EP08, 6 in EP17 and 1 each in UP02, UP08, UP14). While the injectors on the equatorial level are dedicated to thermal load mitigation, current quench control and RE prevention and RE energy dissipation, the purpose of the injectors on the upper ports are for late current quench mitigation. All DMS locations will be equipped with a dedicated gas supply providing the pellet material gases and high-pressure propellant gas, which is now integrated into the ITER design. An existing cryogenic supply system, as well as a gas venting system, are also incorporated in the ITER design to support the DMS needs in all required ports. In order to provide a DMS that performs with high reliability and availability, this first of a kind system has to fulfil key crucial requirements; these are the current design drivers of this system: a) defined and reproducible pellet integrity and pellet acceleration process; b) monitoring of pellet integrity and minimised interaction between the pellet and the flight tube wall; c) optimised shattering unit.
One of the main challenges for the DMS plant is the integration of the required components in a small and extremely harsh environment considering nuclear qualification and maintenance. The change of DMS concept following the 2017 review has lead, in addition, to integration challenges with diagnostic systems located in the same ports. To address these integration issues specific actions have been implemented: a) in the port plugs where the DMS are installed, pellet flight tubes’ openings have been optimised to ensure pellet survivability and to reduce neutron streaming and activation of equipment; b) in the port interspace (small area inside the bioshield), the DMS injectors have been designed to minimise the impact on the neighbouring diagnostics, while providing sufficient human access; c) in the port cell behind the bioshield extensive gas handling and cryogenic distribution units enable the control of all relevant process variables, in order to reliably operate the DMS; d) in equatorial port two (EP02) integration issues are being assessed with respect to the Core Imaging X-Ray Spectroscopy diagnostic. This diagnostic provides measurements of ion temperature and plasma rotation and its design has been substantially modified for compatibility with the two DMS systems sharing the port. The analysis of this design change shows impacts on its measurement capabilities regarding spatial/time resolution and for the measurable magnitude of the toroidal rotation, which are being minimized by design iterations. Other diagnostics systems sharing ports with DMs, such as the visible infrared viewing systems in the upper ports are not significantly affected.
A DMS Task Force has been created to support the DMS design and has two main activities, the design validation through experiments and modelling and the optimisation of the SPI technology (5). The technology programme addresses the main DMS design drivers through development and optimisation of key components to fulfil the ITER mitigation requirements. R&D under the technology programme covers issues such as : a) fundamental studies, including systematic tests and optimisation of the pellet formation and release process; b) the creation of a support laboratory, providing a test bed to assess the performance of key components such as the shattering bend; c) the development of pellet launch to optimize release mechanisms (fast valve and punch); d) the development of optical pellet diagnostics to diagnose pellet alignment, pellet integrity and pellet parameters.
The DMS TF work program on design validation consists of dedicated experiments accompanied by theory and modelling activities. These efforts are supported by significant contributions through the domestic activities of the ITER partners. The ITER DMS strategy relies on densification of the plasma to avoid runaway electron formation and on radiating most of the thermal and magnetic energy spatially as uniform as possible to avoid first wall melting. Densification requires the injection of multiple pellets, a scheme that is tested in JET (2), DIII D (6) and KSTAR (3), for the latter with an SPI system that can inject a total of four identical pellets from two toroidally opposite locations. These experiments, together with the tokamak J-TEXT, are providing information on size and energy scaling of SPI mitigation performance for model validation. While the technology programme assesses the parameter space for pellet fragmentation, the experiments planned for ASDEX-Upgrade will focus on finding the optimum fragment size for maximum material assimilation. In support of the experimental activities, the theory and modelling activities programme will provide physics-based extrapolation from the experimental results to ITER (4). This will allow narrowing down design parameters for the ITER DMS such as the pellet velocity, the fragment size distribution, and the pellet composition and it will increase the confidence level that the ITER DMS can fulfil its purpose.
The paper will described status of the DMS design, its integration into the ITER baseline and the status of the technology R&D and will provide a summary of progress of the design validation experimental and modelling activities.
References
(1) ITER Research Plan, ITR-18-03, ITER Organization, 2018.
(2) Jachmich, S., et al., this conference.
(3) Kim, J.-H., et al., this conference.
(4) Nardon, E., et al., this conference.
(5) Lehnen, M., et al., 2018 IAEA Fusion Energy Conf., Ahmedabad, India, pap. EX/P7-12.Speaker: Dr Timothy C. Luce (ITER Organization) -
In a recently conducted test for assessing compatibility of accelerator grid of Neutral beam 1 for their performance at 150 C, failure has been evidenced across an electrodeposited (ED) bond layer, which forms a vacuum boundary with cooling medium. This happens to be the first instance where an electrodeposited bond has been subjected to Hot Helium Leak test under operational conditions of temperature (150 C) and pressure (>2MPa), according to the ITER specifications. Presently there exists no recommended procedure for carrying out an assessment of the bond integrity for electrodeposited surfaces and also no codes are presently available to qualify the process compatibility for the operational requirements including application at around 150 C. Therefore, to arrive at the root cause, it was necessary to undertake a systematic evaluation study to identify the root cause and establish the adhesion characteristics at room as well as operational temperatures. This also serves the objective to assess and to arrive at an understanding of the possibility of recommending a metric for the strength parameters for such bonds. To carry out this assessment, multiple electrodeposited samples ~20, were drawn under different bath conditions. These samples have been subjected to a specially designed test procedure in order to arrive at an assessment of the strength parameters for the bond, the failure of which is considered to be the root cause. Hence, the present study is purported to present itself as a procedure for evaluation of this important manufacturing process. This is possibly first such initiative of its kind.
Samples were prepared under three different bath conditions with the parameter variation for electrolyte composition, additives, electrical parameters, time, temperature and pH. Adhesion, being one of the major area of consideration for the pressure application, a specific test, called ‘Push’ test has been configured (Fig. 1) in such a way that, it simulates the operational scenario and thus evaluate the stress on the bond area in the operational condition. The most important element of this test, is in the identification of procedure that, would test the performance of the bond for the thinnest cross section of surface contact of the deposit with the base material.
Fig. 1 shows the configuration of the cross sections where the top layer is ED and corresponds to cross section for the apertures of the grid segments. In addition, tests related to metallurgical assessments had to be developed to understand the microstructural characteristics at the transition region between the electrodeposited layer and the base material.
Studies performed across the samples show that interface of the base material and the electrodeposited layer is significantly impacted with the change in bath condition in terms of its adhesion strength, typically in the range of 15-20 %. The most significant observation comes out from the high temperature (150 C) assessment, where degradation in adhesion strength occurs to the tune of by ~30-50 % as compared to base material. This range in degradation is closely co-related to the change in the bath condition. Further, the Load Vs. Extension behaviour differs when the specimen is subjected to tests at elevated temperature. The graph (Fig. 2) reveals the absence of the elastic zone in case of high temperature zone for all the samples from different baths.
Further, the present case being a pure metal, it could also be possible to cross verify the outcome of the strength test through a hardness test (Fig. 3) across the base material, interface zone and the electro-deposited layer. These correlations of strength and temperature along with their dependency on the bath conditions establish a metric to develop the application based design for such components.
Metallurgical assessments to understand the microstructural characteristics across the base material, transition zone and electrodeposited layer (Fig. 4) shows the difference in the material growth characteristics for base material and the ED layer under different bath conditions thereby resulting in the variation of the adhesion properties at the interlayer.
In conclusion, a process has been established in form of experiments where ~20 samples, from different baths, have been subjected to tests to evaluate and obtain a statistical variation in the quality of the bond at room as well as at elevated temperature of 150 C. Test results shows the variation in the bond strengths is highly dependent on the bath quality, thereby establishing the root cause of failure in the hot leak tests of the grid segments. The results also establish the need for qualification of the bond integrity and improvisation of the bath characteristics, if required, to ensure a reliable application of ED process for the actual components. Recommendation of the qualification process is as follows; (1) carrying out and interpreting the specially designed push test for samples (2) Co-relating the strength with the hardness parameters (3) study of microstructural characteristics and (4) application of process on production pieces.
The details of the study performed above and the qualification procedure developed shall be presented and discussed. This forms an important database for similar components in fusion devices like ITER using ED process as one of the important manufacturing sequence.
1] Jaydeep Joshi et. al : Technologies for realization of Large size RF sources for –ve neutral beam systems for ITER - Challenges, experience and path ahead, Nuclear Fusion- 102937.R1Acknowledgement:
Support from M/s PVA Industrial Vacuum Systems, M/s Germany, Research Instruments, Germany and their sub-suppliers, in terms of providing the samples is sincerely acknowledged.Speaker: Jaydeepkumar Joshi (ITER-India (Institute for Plasma Research)) -
Nonlinear two-fluid MHD simulations reveal the role of resonant field penetration in ELM suppression and density pump-out in low-collisionality ITER-Similar-Shape (ISS) plasmas in the DIII-D tokamak$^1$. The operational window for ELM suppression in DIII-D ISS plasmas coincides with calculations for magnetic island formation at the pedestal top ($n_e<3\times10^{19}m^{-3}, B_r/B_t >10^{-4}, \nu_e^*<0.3$) based on nonlinear MHD simulations using the TM1 code$^2$. Key phenomenology in experiment are reproduced in the simulations including density pump-out for field penetration at the foot of the pedestal and pedestal pressure and width reduction due to magnetic island formation at the top of the pedestal. TM1 simulations also reproduce the observed $q_{95}$ width of ELM suppression windows in DIII-D. Analysis indicates that wide $q_{95}$ windows of ELM suppression may be accessible in DIII-D and ITER by operating at higher toroidal mode number.
For our analysis we use the cylindrical initial value nonlinear two-fluid MHD code TM1$^2$ with the helical magnetic field boundary condition provided by the ideal MHD code IPEC$^3$. The fully toroidal ideal MHD code IPEC calculates the perturbed 3D magnetic equilibrium (vacuum field plus ideal MHD kink/peeling response to the I-coils) in the actual magnetic geometry. TM1 takes the measured kinetic profiles from DIII-D before the RMP is applied, including initial transport coefficients and neoclassical resistivity from TRANSP. Multiple helical field harmonics from IPEC are applied at the simulation boundary for a given toroidal mode number (e.g. $m/n=6/2, 7/2, 8/2, 9/2, 10/2, 11/2, 12/2$ harmonics for $n=2$ RMP). The TM1 simulation then predicts the penetration or screening of these resonant fields in the plasma interior and the effect of these fields on the electron density and temperature profile. The TM1 model solves the torque balance that governs the bifurcation from screening to penetration of resonant fields including diamagnetic drifts that are important in the pedestal. TM1 also solves the electron continuity and energy transport equations, taking into account enhanced collisional (parallel) transport across magnetic islands.
TM1 simulations show that RMPs readily penetrate into the collisional foot of the DIII-D pedestal near the separatrix ($\psi_N>0.98$), generating a narrow region of edge stochasticity and enhanced collisional transport. The enhanced transport leads to density pump-out, reproducing the magnitude of the observed density reduction observed in experiment$^{1,4}$. While pump-out is ubiquitous at low collisionality, special conditions are required for ELM suppression in DIII-D ISS plasmas, including low plasma density ($<3\times10^{19}m^{-3}$), high co-$I_p$ toroidal rotation and high RMP amplitude ($B_r/B_t >10^{-4}$)$^5$. These conditions correspond to the requirements for resonant field penetration at the top of the pedestal from TM1 simulations. Figure 1 (a) shows the similarity in the TM1 predicted (white) and measured electron pressure profile (yellow) during pumpout and ELM suppression using $n=2$ RMPs in a DIII-D ISS plasma with $q_{95}\approx4.1$, together with the Poincaré plot showing penetration of resonant fields at the top and foot of the pedestal. ELMs are suppressed when the pedestal height and width fall $\approx15\%$ below the EPED model prediction$^6$. Several hundred non-linear TM1 simulations were performed to derive the following scaling relation for the field penetration threshold at the pedestal top in DIII-D ISS plasmas
$\ \ \ \ \ \ \ \ \ \ \ \ \ \ B_{r,th}^{scale}/B_t=3.5\times10^{-2}n_e^{0.7}|\omega_E+\omega_{*e}|^{0.9}B_t^{-1}, \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ $(1)
where $\omega_E$ and $\omega_{*e}$ are the $E\times B$ and electron diamagnetic drift frequency. This scaling closely resembles analytic estimates for error field penetration in the plasma core$^7$ and is quantitatively consistent with the threshold for ELM suppression in DIII-D. A 2D contour plot of the penetration threshold $B_r/B_t$ at the pedestal top versus density and $\omega_E$ is shown in Figure 1 (b). The RMP amplitude for penetration decreases (yellow to blue) as the density and the magnitude of $\omega_E$ decrease, consistent with experiment for ELM suppression$^5$. A database of $n=2$ RMP discharges are overlaid onto Figure 1 (b), with ELM suppression (yellow stars) and ELMing (black stars). The boundary contour between suppression and ELMing corresponds to the maximum available RMP amplitude $B_r/B_t\sim3.5\times10^{-4}$ in the experiment (white dashed line), representing a remarkable level of agreement between experiment and nonlinear MHD theory for the ELM suppression threshold. Analysis of a model ITER $Q=10$ discharge$^8$ predicts that the threshold for $n=3$ penetration at the top of the pedestal will be significantly lower ($B_r/B_t\sim2\times10^{-5}$) than in DIII-D due to the lower rotation and diamagnetic frequency expected in ITER.
TM1 simulations also reveal the conditions required for wide $q_{95}$ windows of ELM suppression. Wide $q_{95}$ windows are essential for operational flexibility in ITER. However, near the threshold of resonant field penetration at the top of the pedestal, only narrow $q_{95}$ windows of ELM suppression are observed in DIII-D. These narrow windows are reproduced using TM1 simulations and EPED model predictions. Figure 2 shows the predicted pedestal pressure reduction (color contour) versus $q_{95}$ from TM1 simulation and its intersection with the $n=2$ and $n=3$ RMP amplitude on the plasma boundary obtained from IPEC (horizontal dashed lines). For similar RMP amplitudes on the plasma boundary, $n=3$ produces wider $q_{95}$ windows of ELM suppression than $n=2$, consistent with DIII-D experiment$^5$. Figure 2 (c) shows a TM1 simulation with $n=4$ RMPs in DIII-D, indicating window overlap for the same ISS plasma condition. The TM1 simulations predict access to wide operational window of ELM suppression with relatively weak pressure reduction at higher-$n$. Analysis of model ITER equilibria also reveals a similar trend with toroidal mode number, suggesting that operating with dominant $n=4$ RMP, accessible to the ITER ELM coil design, may be advantageous for ITER operation.
To conclude, TM1 simulations quantitatively reproduce ELM suppression windows and density pumpout in the DIII-D tokamak. ITER simulations suggest that the threshold for penetration will decrease relative to DIII-D due to the expected lower $E\times B$ and diamagnetic frequency expected in ITER. Our simulations account for the observed width of $q_{95}$ windows of ELM suppression in DIII-D and indicate that $n=4$ RMPs may be effective to produce wide windows of ELM suppression in ITER.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-AC02-09CH11466 and DE-FC02-04ER54698.
Reference:
1 Q.M. Hu et al., Phys. Plasmas 26, 120702 (2019).
2 Q. Yu and S. Günter, Nucl. Fusion 51, 073030 (2011).
3 J.K. Park and N.C. Logan, Phys. Plasmas 24, 032505 (2017).
4 Q.M. Hu et al., Nucl. Fusion submitted (2019).
5 C. Paz-Soldan et al., Nucl. Fusion 59, 056012 (2019).
6 P.B. Snyder et al., Phys. Plasmas 19, 056115 (2012).
7 R. Fitzpatrick, Phys. Plasmas 5, 3325 (1998).
8 F.M. Poli et al., Nucl. Fusion 58, 016007 (2018).Speaker: Dr Raffi Nazikian (Princeton Plasma Physics Laboratory) -
Since the initial JET operations with the metal wall (JET-ILW), the experimental results have shown a pedestal pressure in baseline plasmas that tends to be 10-20% lower than in the corresponding earlier carbon wall operations (JET-C) [1]. While this degradation seems mainly correlated with the high fueling rates typical of JET-ILW [2,3] and/or the lack of carbon impurity [4,5], an exhaustive and comprehensive explanation for the lower pedestal performance has not been achieved yet. This work will address the role of fueling and its goals are:
- to prove that the lower pedestal performance in D fuelled JET-ILW plasmas are due to the higher separatrix density ($n_e^{sep}$) produced by the higher neutral pressure,
- to describe the corresponding physics mechanisms that lead to the pedestal degradation.
In the baseline scenario of JET-ILW, operations with no gas-fueling rate have been extremely challenging due to the problems related to tungsten influx and divertor heat loads. Since most of the JET-C plasmas have no gas fueling, a direct comparison of JET-ILW and JET-C pedestals obtained with identical engineering parameters is not possible. A further complication is related to the fact that the peeling-ballooning (PB) stability model (implemented with ideal MHD equations) does not describes correctly the experimental JET-ILW results (the experimental pedestal with high fueling rates does not seem to reach the stability boundary when the ELMs are triggered [1,2,3]). Therefore, the work is based on two levels. First, the work focuses on the empirical understanding of the pedestal behavior in JET. Then, based on these results, an investigation of the pedestal transport and an extension of the PB stability analysis is done with the GENE [6] and JOREK [7] codes.
Figure 1 shows the height of the electron pedestal pressure ($p_e^{ped}$) versus $n_e^{sep}$ for a set of JET plasmas with the same engineering parameters apart from fueling rate and divertor configuration. The JET-C dataset has higher $p_e^{ped}$ than the JET-ILW dataset. However, the two datasets align very well in the $p_e^{ped}$-$n_e^{sep}$ diagram. Moreover, the JET-ILW pulses with lowest $n_e^{sep}$ reach a pedestal pressure comparable to JET-C. This suggests that the separatrix density is one of the key parameters to understand the difference between carbon and metal wall. The higher $n_e^{sep}$ in JET-ILW is likely due to the higher neutral pressure, as recently discussed for JET-ILW [8] and AUG [9], produced by the higher gas fueling and/or different recycling. In figure 1, note that the subsets with different divertor configurations show no systematic difference, strengthening the hypothesis that the neutral pressure plays the key role.
The standard PB stability analysis performed with ideal MHD can only partially explain the empirical trend. This is shown in figure 1 by the red line, which represents the pressure predictions obtained with the Europed code [10]. The increase in $n_e^{sep}$ initially leads to a sharp reduction in the predicted $p_e^{ped}$. This is due to the fact that the increasing $n_e^{sep}$ is intrinsically linked to the outward shift of the $n_e$ position ($n_e^{pos}$), shifting the $p_e$ profile and destabilizing the PB modes [3,11]. While this explains rather well the JET-C trend, the effect saturates at high $n_e^{sep}$. The prediction significantly overestimates the experimental $p_e^{ped}$ for the JET-ILW pedestal with high $n_e^{sep}$.
Therefore, the next steps are to understand the mechanisms that (1) set the pedestal gradient and (2) trigger the ELMs at high $n_e^{sep}$. First of all, we note from figure 1 and figure 2(a) that the reduction of pedestal gradient is correlated with the increase in $n_e^{sep}$ and in $n_e^{pos}$-$T_e^{pos}$. The increase in these parameters leads to the increase of $\eta_e$ (ratio between $n_e$ and $T_e$ gradient length) [3], which in turn can destabilize microinstabilities, increase turbulent transport [12, 13] and hence reduce the pressure gradient. This hypothesis is under investigation with GENE and is supported by preliminary results shown in figure 2(b), where the growth rates (mainly of ETG modes) are higher in pedestals with higher $n_e^{sep}$ and higher $n_e^{pos}$-$T_e^{pos}$ [14,15].
Then, it is necessary to understand the ELM triggering mechanisms. The discrepancy between the experimental results and the ideal MHD results is quantified with the ratio $\alpha_{crit}/\alpha_{exp}$ (where $\alpha_{crit}$ is the normalized pressure gradient predicted by ELITE and $\alpha_{exp}$ is the experimental one). Figure 3 shows that $\alpha_{crit}/\alpha_{exp}$ increases with increasing resistivity. This suggests that resistivity might have a destabilizing effect on the PB modes, as theoretically discussed in [16]. This hypothesis is currently under investigation with the non-ideal MHD non-linear code JOREK.
The picture that is emerging is the following. Due to higher gas fueling rate / different re-cycling, JET-ILW has higher neutral pressure than JET-C. This leads to higher $n_e^{sep}$ and higher $n_e^{pos}$, producing higher $\eta_e$, increasing the turbulent transport and reducing the pedestal gradient. In turn, the lower pedestal gradient leads to a lower temperature inside the separatrix, increasing the resistivity and making resistive effects on the MHD stability non-negligible.
References
[1] Beurskens M. et al., NF 54 043001 (2014)
[2] Maggi C. et al., NF 57, 116012 (2017)
[3] Frassinetti L. et al., NF 59 076038 (2019)
[4] Giroud C et al., IAEA FEC, EX/3-4 (2018)
[5] Beurskens M. et al., NF 56 056014 (2016)
[6] T.Görler et al., JCP 230, 7053-71 (2011)
[7] Huysmans G. et al. NF 47, 659 (2007)
[8] Frassinetti L. et al, to be submitted to NF
[9] Kallenbach A. et al., NME 18, 166 (2019)
[10] Saarelma et al., PPCF 60, 014042 (2018)
[11] Dunne M. et al., PPCF 59, 014017 (2017)
[12] Hatch D.R et al., NF 57, 036020 (2017)
[13] Kotschenreuther M. et al., NF 57 064001 (2017)
[14] Chapman B. et al., to be presented at the 47th EPS conference
[15] Hatch D.R et al., NF 59, 086056 (2019)
[16] Wu et al., PoP 25, 092305 (2018)
Speaker: Lorenzo Frassinetti (KTH, Royal Institute of Technology) -
Good confinement of the fusion-born alpha particles is essential to ensure adequate burning plasma performance in next-step fusion devices. Among the processes determining this confinement, instabilities triggered by energetic particles (EPs) may play a major role, and are currently being studied in various tokamaks using auxiliary power sources to sustain EP populations. Instabilities resulting from fusion-born alphas, on the other hand, can only be observed in deuterium-tritium (D-T) plasmas. Since DTE1, the D-T campaign conducted in the Joint European Torus (JET) in 1997, the device has undergone significant changes, among which the installation of a Be/W ITER-like wall (ILW) and the development of new diagnostics directly relevant to the physics of energetic ions, in particular alphas. The preparation of a new D-T campaign (DTE2) in JET [Joffrin2019] thus includes various developments relevant to burning plasmas [Sharapov2008]. As JET is currently the only tokamak in which D-T plasmas can be produced, DTE2 constitutes the only opportunity to experimentally document the physics of alphas, and validate the numerical tools used to simulate their effects before ITER comes into operation.
Among the instabilities related to the presence of EPs, alpha-driven Toroidal Alfvén Eigenmodes (TAEs) have received some attention in the past. The rationale is that the features of the alpha population differ significantly from those of energetic ions created by external sources. As a result, the instability itself differs and its impact on the plasma performance remains to be evaluated. Because of the relatively low values of normalized alpha pressure ($\beta_\alpha$) attained in the only two magnetic confinement fusion devices capable of D-T operation to this day, TFTR [Nazikian1997] and JET [Sharapov1999], core-localized alpha-driven TAEs have been difficult to observe unambiguously. From these experiments and from results obtained during the present effort in JET [Dumont2018], it has been established that their observation requires i) a sufficient alpha pressure, ii) an elevated safety factor (q), iii) an “afterglow phase” consisting of abruptly switching off all external EP sources and rely on the longer slowing-down of alphas compared to other ions present in the pulse to isolate their impact, including the destabilization of TAEs. The afterglow has been key to the success of the experiments performed in TFTR [Nazikian1997]. In terms of scenario, these conditions translate into i) low density to favour large electron and ion temperatures, ii) large NBI power to maximise the fusion yield, iii) no ICRH power before the afterglow phase to exclude any contribution from ICRH-driven ions to the TAE drive, iv) an elevated q-profile. In preparation for DTE2, advanced scenarios fulfilling these requirements have been under development in deuterium plasmas during the last experimental campaigns. In pulses at 3.4T/2.5MA, NBI waveforms have been fine-tuned to inject the power early in the pulse and thus obtain elevated q-profiles, while fulfilling the requirements of the ILW in terms of beam shine-through. Operating at line-integrated densities in the range $5-9\times 10^{19}\text{m}^{-2}$ has allowed clear Internal Transport Barriers (ITBs) to be observed in JET-ILW.
The resulting ion temperatures in the range $10-15\text{keV}$ at peak performance yield large neutron rates from D-D reactions in NBI+ICRH discharges ($2.8\times 10^{16}\text{s}^{-1}$) and in NBI-only discharges ($2.5\times 10^{16}\text{s}^{-1}$). On the other hand, in some instances, phases during which ELM-free/type-I ELM alternate set in before the period of interest, and result in impurity influxes deleterious to the performance and possibly inducing early pulse terminations. A large effort has thus been devoted to ELM and impurity control without resorting to ICRH power. Pellet pacing has been found to be the most efficient method for these plasmas, and has allowed pulses with no ELM-free/type-I ELM periods to be obtained. Despite the use of these methods, predicting the time of peak performance remains difficult because it results from a trade-off between ITB build-up and impurity accumulation. A real-time control algorithm has therefore been developed and successfully tested to start the afterglow at the best possible time during the pulse, i.e. when the neutron rate reaches its peak value. Finally, discharges entirely fuelled by the Tritium Injection Modules (TIMs) relevant to the upcoming TT and D-T campaigns have been successfully demonstrated.
In order to produce EP populations and probe the TAE stability in these D plasmas, ICRH power has been employed. Hydrogen (H) minority heating at 51MHz results in the destabilization of core-localized TAEs with properties approaching those expected for alpha-driven TAEs in DTE2.
Modelling the stability in these discharges allows predictive simulations for DTE2 to be refined. Helium 3 ($^3$He) minority heating at 33MHz has also been tested. Although it requires more fine-tuning compared to H minority heating, the advantage of this scheme is that it results in the creation of energetic $^3$He populations which are particularly well diagnosed by the EP and neutron diagnostics installed in JET. As a result, this type of pulse provides essential information regarding EP transport in the presence of elevated q-profiles, a topic fully relevant to the preparation of robust and performant advanced scenarios for ITER [Sips2005].Overall, the extrapolation of the best-performing NBI-only pulses obtained so far to D-T predicts that $\beta_\alpha(0)>0.1\%$ should be attained, which is compatible with the observation of core-localized alpha-driven instabilities and measurement of resulting induced EP transport, thus encouraging the completion of the present developments in view of DTE2.
[Joffrin2019] E. Joffrin et al, Nucl. Fusion 59 112021 (2019)
[Sharapov2008] S.E. Sharapov et al, Fus. Sci. Tech. 53 989 (2008)
[Nazikian1997] R. Nazikian et al, Phys. Rev. Lett. 78 2976 (1997)
[Sharapov1999] S.E. Sharapov et al, Nucl. Fusion 39 373 (1999)
[Dumont2018] R.J. Dumont et al, Nucl. Fusion 58 082005 (2018)
[Sips2005] A.C.C. Sips et al, Plasma Phys. Control. Fusion 47 A19 (2005).This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
Speaker: Dr Remi Dumont (CEA, France ) -
A series of experiments have been executed at JET to assess the efficacy of the newly installed Shattered Pellet Injection (SPI) system in mitigating the effects of disruptions. In this contribution, the results from these JET SPI experiments are presented and their implications for the ITER disruption mitigation scheme discussed.
An effective Disruption Mitigation System (DMS) that minimizes thermal excursions, mechanical forces and effects of Runaway Electrons (RE) is mandatory for successful operation of ITER. The chosen ITER-DMS concept [1] is based on disruption mitigation through injection of shattered pellets, so-called SPI, which promises a deeper and faster penetration of radiating material compared to the commonly used Massive Gas Injection (MGI) valves as demonstrated on DIII-D [2].
Gaps in the understanding of the effectiveness of disruption mitigation with SPI still remain. To assist in addressing them, a new SPI system has been brought into operation during the most recent JET campaign and the efficiency in mitigating thermal and electromagnetic loads and in dissipating the energy of an existing RE-beam has been studied. The JET SPI experiments are a key element to allow extrapolation to ITER in terms of size, plasma current and plasma energy. Moreover, the ITER-like wall is essential to address injections with low impurity content without being impacted by high carbon levels. These experiments serve as basis for the benchmarking of models that are required to predict the performance of the ITER-DMS. The JET SPI is a three-barrel injector system capable of delivering neon, deuterium, argon and mixed pellets. By using mechanical punches, which are required to dislodge argon pellets, the velocity of the fired pellet and the resulting shard size and speed can be varied significantly [3].
In ITER, thermal load mitigation must ensure a high radiation fraction during the thermal (TQ) and current (CQ) quenches, while keeping the radiated power asymmetries to acceptable levels. To determine the optimum amount of neon needed to maximise the radiation, the fraction of neon injected together with deuterium into typical JET low and high energy H-mode plasmas was scanned for magnetic and thermal energies ranging respectively from 3.0 to 16 MJ and from 1.0 to 7.0 MJ. High radiated energies are reached for neon atom quantities >3-4x1021. However, due to potential variations in pellet velocities and shard size distributions, the achieved radiation level varies. Analysis of fast camera observations of the pellet shard ablation indicates differences in the material delivery that can be expected to affect the radiation efficiency. The achievable radiation levels depend on the fraction of thermal energy in the pre-disruptive plasma, though it should be noted that radiation asymmetries have an important impact on the assessment of the thermal load mitigation efficiency [4]. In addition, such asymmetries can lead to a local radiation flash sufficient to melt some areas of ITER beryllium first wall (FW). In order to determine the toroidal peaking factor of the radiation, either a good toroidal coverage with the bolometry diagnostic is required, or the toroidal distribution must be inferred from the change of the radiated power at a given toroidal location as a function of the location of the locked mode phase. At JET, the total radiated power can be estimated at two different toroidal locations [5]. By superimposing a non-axisymmetric n=1 perturbation field [6], the radiation asymmetry was measured as a function of the locked mode phase. The experiments indicate that asymmetries decrease with increasing plasma energy. A dedicated modelling programme accompanies these experiments to assist the data interpretation and to provide the extrapolation to ITER [7].
Low radiation levels during the unmitigated CQ are expected to lead readily to melting of the ITER FW due to high plasma thermal fluxes. This has in fact already been observed in JET after unmitigated disruptions with plasma currents > 2.0 MA. It must therefore be ensured that the magnetic energy will be fully radiated and hence that a late material injection, i.e. after the TQ, is still effective. In order to test the mitigation scheme, where the DMS is triggered after a TQ, a density limit disruption has been provoked on JET by deuterium MGI into an ohmic plasma, with the SPI triggered such that the shards arrive ~9 ms after the start of the CQ (see Figure 1). As a result, the CQ has been accelerated and the radiated energy has been more than doubled compared to the unmitigated reference.
To keep the electromagnetic forces on the ITER blanket modules within the design limits, the CQ times must remain within a range 50-150 ms [8]. The JET experiments have clearly demonstrated the controllability of the CQ rate by injecting different quantities of neon. Moreover, the effects of asymmetric Vertical Displacement Events are found to be fully mitigated with SPI, when the vertical displacement is detected with sufficient warning [9].
For ITER, initial modelling predicts that 6x1024 hydrogen atoms must be assimilated by the plasma to avoid the generation of REs [10]. This scheme was successfully tested at JET by injecting deuterium with the SPI into RE-prone disruptions initiated by a pure argon MGI. In ITER, injection of multiple pellets is required to achieve sufficiently high density for avoidance of REs. The material assimilation during dual injection in JET using the largest two barrels was assessed. It was found, similarly to DIII-D, that the simultaneous arrival of the shards from the two pellets is essential to maximise the assimilation. The available time to inject high-Z material for heat load mitigation after an attempt has been made to raise the density for RE avoidance was determined by injecting a large, pure deuterium SPI into a high thermal energy plasma. Cooling durations up to the TQ of several tens of ms were achieved. On the contrary, by doping the deuterium pellet with just 2% neon, the cooling duration is reduced to a few ms, which would impose severe constraints on the synchronisation of the ITER-DMS injectors. These results are in agreement with modelling for ITER using the INDEX code, where the dilution cooling by pure deuterium SPI is found to avoid an immediate radiative collapse and delays the TQ trigger [11].
In the case of accidental RE generation in ITER, a scheme to dissipate their energy must be in place. Injections of both pure argon and neon shards into a fully developed RE-beam in JET can reduce the RE current. However, in contrast to high-Z injections, it was observed that pure deuterium injection could avoid high energy deposition during the final loss phase of the RE-beam.
On ITER hydrogen is planned for use both as propellant and main pellet material for RE avoidance throughout all operational phases, raising the issue of a potential detrimental effect on the subsequent pulses due to dilution of the plasma main species with residual hydrogen. Tests in H-mode plasmas at JET are helping ITER to determine whether or not this will be an issue.[1] M.Lehnen et al., Proceedings 27th IAEA-FEC (2018).
[2] N.Commaux et al., Nucl.Fus. 56, 2016, 046007.
[3] L.Baylor et al., subm. 28th IAEA-FEC (2020).
[4] U.Sheikh et al., subm. 28th IAEA-FEC (2020).
[5] J.Lovell et al., subm. Rev.Sci.Instr.
[6] S.Jachmich et al., 22nd PSI-conference (2016).
[7] E.Nardon et al., subm. 28th IAEA-FEC (2020).
[8] M.Sugihara et al., Proceedings 24th IAEA-FEC (2012).
[9] S.Gerasimov et al., subm. 47th EPS-conference Plasm.Phys. (2020).
[10] J.R.Martin-Solis et al., Nucl.Fus. 57, 2017, 066025.
[11] A.Matsuyama et al., subm. 28th IAEA-FEC (2020).Speaker: Stefan Jachmich (ITER Organization) -
The present contribution is devoted to the neutral beam injectors (NBIs) for ITER heating and current-drive. First, updated information is provided about the development status of the entire NBI prototype (MITICA); starting in 2021, the first experiments will be dedicated to high-voltage holding tests in vacuum. Then the contribution describes the full-scale prototype of the NBI ion source (SPIDER) and the activities performed in the first two years of operation, devoted to investigating the operational regimes of the ion source and verifying the performances of the various plants in view of extensive beam operation. Some improvements have already been implemented; others are being prepared for the next shutdown. The characterisation of the SPIDER plasma is presented and the results of the first beam operations are reported.
The ITER experiment represents the next step in realising nuclear fusion as an environment friendly energy source for the future. To reach fusion conditions and to control plasma configurations, two heating and current-drive neutral beam injectors (NBIs) will provide the ITER plasma with 17MW each, by accelerating negative hydrogen/deuterium ions. The requirements of ITER neutral beam sources (40A D- ions at 1MeV for up to 1 hour, 46A H- ions at 870keV for up to 1000s) are so challenging that current-voltage-duration have never been simultaneously attained yet. So, in the dedicated Neutral Beam Test Facility (NBTF) at Consorzio RFX (Italy), an extended R&D activity is aimed at reaching full performances and optimizing reliable operation, in time for ITER experiments. To speed up the beam development, imposed by the tight ITER schedule, the NBTF includes two experiments: MITICA, full-scale ITER NBI prototype, and SPIDER, full-scale prototype of the ITER NBI source with 100keV particle energy, whose simpler construction allows anticipating several R&D activities with respect to MITICA. SPIDER operation started in spring 2018 aiming at investigating source and beam uniformity (over a 1m×2m area), negative ion current density, beam optics in conditions relevant to ITER requirements. MITICA will focus on beam acceleration, in terms of optics (divergence <7mrad, aiming <2mrad) and high-voltage holding in vacuum, and on beam propagation, governed by neutralisation and by electrostatic removal of residual ions.
Concerning MITICA, all power supplies and auxiliary systems were tested and installed; the vacuum vessel was completed, whereas the in-vacuum mechanical components are under procurement by F4E (see fig. 1).Installation, commissioning and test of the 1MV power supplies started in autumn 2018 and are nearing their end. In particular, insulating tests for high-voltage components were successfully completed, after solving various technical issues and by adopting a step-by-step procedure during the integration of the sub-systems procured by different Domestic Agencies (European and Japanese). Power supply integrated tests with a dummy load (1MV, 50MW, 2s) are ongoing in 2020 and include the simulation of accelerator grid breakdowns using a short circuit device installed inside the vacuum vessel in the place of the beam source. These tests will allow verifying the full performances of the power supply systems and their reliability during grid breakdowns in normal operation. In the present contribution, the activities for completing and commissioning the power supply system, together with the results obtained during the integrated tests, are presented.
In the first two years, SPIDER operations generated a wealth of experimental information, which provided insight into the source performance and raised operational issues that must be resolved in view of extensive beam operation and particularly of MITICA. It is worth mentioning that, unlike any other existing NBIs, the entire beam source assembly of MITICA (and SPIDER) lies inside the vacuum vessel, so that it is surrounded by the background gas at low pressure in which the beam propagates. RF-induced overvoltages around RF drivers are found to result in electrical discharges when the gas pressure inside the vessel exceeds a threshold. In order to carry out experiments with all 8 RF drivers, the gas conductance between source and vessel was reduced by installing a mask limiting the number of beamlets to 80 out of 1280, thus lowering the gas pressure near the RF drivers. The mask will be removed when improved pumping system will be available (ongoing design). As for the SPIDER RF system, mismatch between self-oscillating frequency of the generators and resonance frequency of drivers and plasma can result in frequency jumps with consequent plasma stop; this condition was demonstratedly avoided by implementing feedforward control of the generator frequency and tailoring the RF frequency to plasma conditions. Several circuits and diagnostics are plagued with RF noise; the cause was shown to be due to currents flowing in the capacitive voltage dividers measuring the RF power; an alternative solution will soon be tested. Experiments allowed also clarifying the cause (layout of RF circuits) of interference between neighbouring drivers along with identifying some limitations to maximum RF power; solving these issues however requires major modifications of the RF circuits. Experiments showed that the magnetic filter field currents (aimed at extending the negative ion lifetime) affect the plasma performances; the cause lies in the SPIDER magnetic filter topology, which exhibits a null point and a large intensity inside the RF drivers; thus distorting electron trajectories. This issue was addressed by modifying the magnetic field topology installing additional busbars. Other issues regard: extension of plasma grid and bias plate voltages to comply with the higher plasma potential found in operations without caesium; characterisation of caesium evaporators; limitation to 30kV instead of 100kV of the acceleration power supply voltage.
In the meantime, the plasma was characterised by means of spectroscopy (including optical diodes). Additionally, at the beginning of 2020 an experimental campaign was devoted to performing spatially resolved measurements of plasma parameters by means of an array of electrostatic sensors installed on a temporary, remotely controlled structure entering the plasma through accelerator grid apertures. Electron temperature and density are shown in fig. 2: as expected with low magnetic filter field, plasma expands out of the drivers, whereas temperature decays gently towards the plasma grid (PG).During the first extraction of negative particles from the source, features of negative ion beam and of co-extracted electrons were studied and correlated with plasma parameters. Particularly, the magnetic filter field effectiveness in reducing the co-extracted electron current was verified (down to ~30 times the H- current), along with its influence on negative ion current. The first characterisation of the SPIDER beam, in terms of beamlet divergence and deflection, was compared with numerical models while varying source parameters. The calorimetrical estimate of the current, which is not affected by secondary particles, is lower than the corresponding electrical measurements (see fig. 3).
The negative ion beam is confirmed to exhibit values of current density (up to ~25A/m2) and optics parameters (down to ~25mrad) similar to those usually obtained in volume operation. Finally, in 2020 for the first time, caesium will be injected into the SPIDER source to increase the negative ion density; the results of the first campaign will be described.
After these experimental activities, SPIDER will enter a long shutdown, to carry out a set of modifications identified either during the procurement phase or during the first year of operation, like the aforementioned rearrangement of the RF circuit configuration and the enhancement of the vacuum system to keep the vessel pressure low with no plasma grid mask.Speaker: GIANLUIGI SERIANNI (ItRFX) -
Significant stabilizing effect of kinetic thermal ions is found for the LHD plasmas. The kinetic MHD simulations for the LHD plasmas at high magnetic Reynolds number show that the high beta plasmas can be maintained since the saturation level of the pressure driven MHD instabilities is significantly reduced by the kinetic thermal ions. This results from the fact that the response of the trapped ions to the instabilities is weakened by the precession drift motion in the three-dimensional magnetic field. In the LHD experimental results, the instabilities do not cause significant degradation of the plasma confinement for the Mercier parameter $D_I<0.2\sim0.3$[A]. The linear MHD stability analysis for the plasmas with $D_I=0.2\sim0.3$ in [A] showed that the linear growth rates of the ideal interchange modes with low mode number are $0.01\sim 0.015 / \tau_a$ where $\tau_a$ is the Alfvén time, which are close to the linear growth rate of the interchange mode analyzed in this paper. The supercritical stability of the LHD plasmas well above the Mercier criterion can be attributed to the precession drift motion of the trapped ions in the three-dimensional magnetic field.
In the LHD experiments, about 5% of the volume averaged beta value is achieved without large MHD activities in the inward shifted LHD configurations where the magnetic axis is shifted inward to the center of the helical coils. However, previous theoretical studies based on the MHD model predicted significantly more unstable MHD modes which were not observed in the experiments [B]. The kinetic MHD analysis for low magnetic Reynolds number showed that the kinetic thermal ions suppress the resistive ballooning modes[C]. Although the saturation level is also suppressed, the decrease of the central pressure occurs in the same way as the MHD model. In this study, nonlinear evolution for the high S number has been carried out by the kinetic MHD simulation in order to investigate whether high beta plasmas can be maintained.
The numerical calculations are done by MEGA code where thermal ions are treated by the drift kinetic model and the electrons are treated by the fluid model [C-E]. The MHD equilibrium is constructed by the HINT code where the central pressure is assumed to be 7.5%[B]. In this equilibrium, the peripheral region is Mercier unstable. Using the pressure obtained from the HINT code, PHINT, the initial pressure profiles are set to be $P_{e,eq}=P_{i\perp,eq}=P_{i\parallel,eq}=P_{HINT}/2$ where $P_{e,eq}$ and $P_{i\perp,eq}$ ($P_{i\parallel,eq}$) are the electron pressure and the ion pressure perpendicular (parallel) to the magnetic field. The initial density profile is assumed to be uniform and the initial ion’s distribution is assumed to be a Maxwellian distribution.
Figure 1 shows the orbit of a trapped ion and the profile of the perturbed Pe of the ballooning mode. In the LHD, the precession drift motion of the trapped ions is not only toroidal direction but also poloidal direction. Thus the precession drift frequency of the thermal trapped ions with respect to the mode phase ($\omega_d$) can be larger than the linear growth rate of the instability ($\gamma$). For such case, the trapped ions can move in both positive and negative perturbed regions as shown in Fig.1. Then the response of the trapped ions to the instability is weakened since the influence of the instability on the trapped ions is smoothed. Figure 2 shows the profiles of perturbed $P_{i\perp}$ along the black curve in Fig.1 where the profiles in Fig.2(b) are obtained by neglecting the curvature and gradient B drifts of the ions for suppressing the precession drift motion. The response of the trapped ions to the instability is significantly weakened by the precession drift motion. This results in the suppression of the perturbed Pi⊥ leading to the decrease of the linear growth rate[D]. This suppression effect becomes more effective when $\omega_d \gg \gamma$.
Figure 3 compares the dependence of the linear growth rate for $n\le 7$ on the $S$ number between the MHD model and the kinetic MHD model where $n$ ($m$) is the toroidal (poloidal) mode number. For the low S region, the most unstable modes are resistive ballooning modes of $n\ge4$. As the S number increases, the linear growth rate of the resistive ballooning modes for the kinetic MHD model are significantly reduced comparing with the MHD model. For the high $S$ region, the most unstable mode for the MHD model becomes the ideal interchange mode with (m,n)=(3,2). Since the linear growth rate is comparable to ωd, the stabilizing effect of the kinetic thermal ions also appears in the interchange mode. It is noted that the interchange mode for the kinetic MHD model is the resistive mode as opposed to the ideal mode for the MHD model.
Figures 4 and 5 show the nonlinear simulation results for $S=10^7$ where the most unstable mode is the interchange mode with $(m,n)=(3,2)$. Figure 4 shows time evolution of profile of the perturbed pressure for the MHD model and the perturbed $P_e$ for the kinetic MHD model in a poloidal cross section. For the MHD model, the mode expands to the core region in the initial nonlinear phase ($t =2079\tau_a$) and then the central pressure decreases as shown in Fig.5(a). On the other hand, for the kinetic MHD model, the high beta plasma is maintained at the saturated state where the pressure is defined as $P=P_e+(2P_{i\perp}+P_{i\parallel})/3$. The mode does not expand to the core region as shown in Fig.4. Moreover, as shown in Fig.5(b), the perturbation of $P_{i\perp}$ of the $(m,n)=(0,0)$ mode is significantly suppressed. This results from the fact that the orbit of the trapped ions in the nonlinear phase do not significantly deviate from the equilibrium orbit due to their weak response to the instability.
[A] K.Y. Watanabe et al., Nucl. Fusion 45 (2005) 1247.
[B] M. Sato et al., Nucl. Fusion 57 (2017) 126023.
[C] M. Sato and Y. Todo, Proc. 27th IAEA-FEC (Ahmedabad, INDIA, 22–27 October 2018) [TH/P5-25]
[D] M. Sato and Y. Todo, Nucl. Fusion 59 (2019) 094003.
[E] M. Sato and Y. Todo, submitted to J. Plasma Phys.
Speaker: Dr Masahiko Sato (National Institute for Fusion Science) -
The ITER Disruption Mitigation System (DMS) should ensure that heat loads,
ElectroMagnetic (EM) loads, and Runaway Electron (RE) impacts remain tolerable during
ITER disruptions. The design of the Baseline ITER DMS, which shall be available from the
beginning of ITER operation, relies on Shattered Pellet Injection (SPI). Up to 24 pellets may
be injected from 3 equatorial ports, plus 3 pellets from upper ports. Several key parameters
however remain to be defined, such as the injected species, the size of the pellets or the
characteristics of the flight tube front end (which determine the shattering). An international
DMS Task Force (TF) has been launched in 2018 in order to urgently inform the Baseline
DMS design (which has to be fixed by 2022), as well as to consider options for a possible
later DMS upgrade [M. Lehnen et al., 27th IAEA FEC, Gandhinagar, India, 2018]. The DMS
TF comprises 3 divisions: Technology, Experiments and Theory & Modelling (T&M). The
present contribution summarizes the T&M activities.The most critical issue is the risk of large (multi-MA) RE beam generation [B. Breizman et
al., Nucl. Fusion 59 (2019) 083001]. Open questions are the amount of hot tail generation and
the amplification of RE seeds by the avalanche mechanism during the Current Quench (CQ).
Concerning the hot tail issue, several actions are underway: 1) the modelling of hot tail
generation in present experiments, in particular in DIII-D, using available numerical tools; 2)
test particle studies in 3D non-linear MHD simulations to study electron transport (and in
particular losses of hot tail electrons due to field line stochasticity) and electron parallel
momentum dynamics during the Thermal Quench (TQ); 3) the development of more
sophisticated numerical tools coupling 3D non-linear MHD and electron kinetics; and 4) a
study on the possibility to reduce hot tail generation in ITER by diluting the plasma with a
pre-TQ pure D2 or H2 SPI [A. Matsuyama et al., this conference].
Even if hot tail generation is negligible, large RE beams may still be produced during the
nuclear phase of ITER operation, due to small but unavoidable RE seeds from tritium beta
decay or Compton scattering of gamma rays emitted by the activated wall, combined with the
very large avalanche gain expected in ITER, which may reach ~1016 [Hender et al., Nucl.
Fusion 47 (2007) S128] or even more [L. Hesslow et al., Nucl. Fusion 59 (2019) 084004].
According to [J.R. Martín-Solís et al., Nucl. Fusion 57 (2017) 066025], this risk would be
mitigated if the plasma could assimilate, in a uniform fashion, a large enough quantity (~20-
40 times the plasma content) of H2 or D2 (in addition to a small quantity of Ne, which is
necessary to radiate the thermal energy and mitigate EM loads by controlling the CQ
duration). However, this work is currently being revisited [T. Fülöp et al., this conference]
using more accurate models (e.g. in what concerns the effect of the partial screening of the
nuclear charge for non-fully stripped ions [L. Hesslow et al., Nucl. Fusion 59 (2019) 084004]
or finite aspect ratio effects [C. McDevitt et al., Plasma Phys. Control. Fusion 61 (2019)
054008]). Furthermore, the critical question of whether the plasma can assimilate the required
amounts of material with sufficient spatial uniformity is being investigated by 1.5D transport
[A. Matsuyama et al., this conference] as well as 3D MHD simulations.
An alternative scheme for RE beam avoidance, based on repeated SPI during the CQ in
order to deplete small RE seed populations before they get amplified by the avalanche, is also
considered. Simple estimates suggest that this scheme may work, motivating a more detailed
investigation.
On the other hand, if a large RE beam forms, its mitigation appears difficult. This is due in
particular to the fact that the beam will move up as its current decreases. T&M efforts
regarding RE beams have thus shifted to understanding the beam termination and how impact
damage may be minimized. In this respect, seemingly benign impacts observed recently at
DIII-D and JET when performing a D2 SPI into a RE beam are the subject of particular
attention [C. Paz-Soldan et al., this conference].
Thermal loads during the TQ are another important issue. Their mitigation requires
radiating most of the thermal energy content of the plasma with minimal toroidal and poloidal
peaking factors. Non-linear 3D MHD simulations with JOREK, M3D-C1 and NIMROD show
that simultaneous dual SPI from toroidally opposite ports can substantially reduce radiation
asymmetries, as illustrated in Figure 1. Work is underway to assess the effect of imperfect
synchronization between the different pellets.
In parallel to the above-described efforts dedicated to providing urgently-needed input for
the DMS design, actions are underway to 1) improve theories and models of disruptions and
SPI; 2) benchmark and validate modelling tools; and 3) explore alternative solutions in case
the Baseline SPI-based DMS turns out not to be fully effective.
Work on model improvements focuses in particular on pellet physics, involving a better
description of electron kinetics and radiation in the ablation cloud, and the coupling of
dedicated pellet codes to 3D non-linear MHD codes.
A detailed benchmark between non-linear MHD codes has been performed for
axisymmetric impurity injection simulations [B. Lyons et al., Plasma Phys. Control. Fusion 61
(2019) 064001] and is being pursued in 3D. Non-linear MHD simulations of SPI are broadly
consistent with experimental observations in DIII-D [C. Kim et al., Phys. Plasmas 26 (2019)
042510] and JET [D. Hu et al., APS-DPP 2018] and detailed comparison is in progress. RE
generation models are also being compared to experimental data, showing promising
agreement for JET massive gas injection cases [L. Hesslow et al., J. Plasma Phys. 85 (2019)
475850601].
In the present contribution, we will overview the progress in the above-mentioned topics and
outline directions for future work.Speaker: Eric Nardon (CEA) -
Disruption in the TOKAMAK device is generally known as one of the most harmful events. The subsequent event of the thermal quench and the current quench cause collateral heat-damage and structural damages. These two potential sources of danger are relatively well known because it is easy to conceive that the confined thermal energy and the magnetic field energy associated with the plasma current in the tokamak system are pouring out at the disruption event. The magnetic energy associated with the plasma current induces the eddy and the direct contact of the plasma current to the surrounding wall causes halo. Both currents flowing through the tokamak structure unleash huge JxB forces which can deteriorate the structural solidity of the machine. In addition to the direct JxB forces by eddy and halo current with the toroidal magnetic field, there is a relatively unknown harm source. Especially during the disruption followed by the vertical movement (VDE), several devices reported the toroidally rotating halo current and were projected toward ITER by multi-machine scaling [Myers 2018]. The biggest concern of the rotating halo current is that one of the vibration modes of machine structure might resonate with the rotating frequency of halo current. Likewise, other tokamaks referred to in C.E.Myer's paper, KSTAR also showed the same phenomena. Figure 1 shows the typical observation of the rotating halo current.
The dynamics of the halo current observed in KSTAR is complicated. The direction of the rotation changes shot to shot and also during a shot implying that it is not caused by the MHD mode riding on the rotating plasma. KSTAR plasma rotates fast and it is not comparable with the slow halo rotation and its direction is always co-current especially for the H-mode plasma. How can it rotates opposite to the plasma rotation and why its speed is far low than plasma rotation should be explained. Even though the projection to ITER is urgent it seems there is no relevant explanation of the relatively low rotation frequency and direction changes so far. Through data analysis of the KSTAR disruption event from 2015 to 2018, we found ample examples of the rotating halo current changing its direction during a shot. We propose a new physics model that elucidating the dynamics of frequency and the change of the rotation direction of the halo current. The model is assuming that the tokamak plasma can be regarded as a spinning rigid body and its angular momentum tilted during an abrupt VDE event for some reason. Then the external torque exerted on the plasma results in a precession motion of the toroidal plasma likewise a tilted spinning top. Depending on the external torque whether it tends to stand upright the spinning top or tilt it more, the precession direction could be changed. The main sources of the external torque are the magnetic field by PF coil currents and the induced current on the passive plate. Two torque compete and the stronger one determines the precession direction. As the plasma approaches the passive plate during a VDE event, the induced current on the passive plate dominates the PF coil effects. Figure 2 shows the external torques by the PF coils and the induced current as the plasma moves downward.
The precession of the plasma torus makes the rotation of the contact point of the plasma with the wall and it results in a rotation of the halo current. The complicated dynamics of the rotation speed and direction variance during the VDE would be discussed in this paper with a detailed explanation of the model. Also, the statistical analysis of the KSTAR data based on this model for the recent experiments will be presented.
Speaker: ByoungHo Park (National Fusion Research Institute) -
The JT-60 Super Advanced (JT-60SA) tokamak construction have been achieved respecting the requirements of very tight tolerance for the assembly and by handling very heavy components in a very close space environment. The construction of this large superconducting tokamak represents a big step forward in the world nuclear fusion history, opening the road for ITER and DEMO. Precise assembly is required, not only to avoid mechanical interference, but also to obtain good plasma performance by less magnetic error field. To complete this work, unique and well-considered procedures were introduced. In this paper, the developed technologies and their results are reported, focusing on the assembly of the final sector of vacuum vessel, central solenoid and in-vessel components.
The JT-60SA project has been designed and constructed under the frameworks of the Satellite Tokamak Programme of the Broader Approach Agreement between EU and Japan, and the Japanese national programme in order to contribute directly to the ITER project and to DEMO by addressing key engineering and physics issues. The JT-60SA is under construction in the Naka Fusion Institute of QST. The tokamak assembly has started in Jan. 2013 and the toroidal magnetic field coils (TFCs), the equilibrium magnetic field coils and 340 degree of the vacuum vessel were assembled and positioned into the Torus Hall in 2018 [A], aiming to complete the construction by the end of Mar. 2020.-
Vacuum vessel final sector assembly
After the last FEC [A], the final sector of the vacuum vessel (VV) was welded on-site. The VV is a double shell stainless steel structure with 18mm thickness for each shell, designed to provide a high toroidal one-turn resistance with large torus vessel volume. The VV shells and cooling water provide neutron shielding. The VV is baked at 200℃ by nitrogen gas. In 2019, the VV final 20 degree sector was completely weld-assembled, and its gravity supports were installed into the tokamak. The 18 sectors of the 10m-diameter and 7m-high VV were assembled onsite by welding and the welding contraction was predicted and compensated to achieve the required high precision (typically $\pm$10mm and $\pm$20mm at the inboard and outboard walls respectively). Absence of defects in the welds was confirmed by the non-destructive examinations. The VV is designed to be supported by 9 gravity supports to resist the seismic and operational loads, and to absorb the thermal expansion by radial flexible plates. The required tight tolerance on the positioning of the plates of the VV gravity supports with respect to the tokamak centre axis of $\pm$1 mm was successfully achieved. -
Central solenoid assembly
One of the challenging assemblies was the insertion of the central solenoid (CS), because it is a huge component with 11m height, 2.1m diameter and about 100 tons weight [B] with a minimum clearance between TFC in-bore and CS outer surfaces of 10mm (see Fig. 1). Assembly tolerance of the CS magnetic axis is $\pm$2mm to minimize magnetic error field. In order to achieve this tight requirement, about 200 points on the CS outer surface were measured by the laser tracker with a small measurement error of less than 0.1mm beforehand. The vertical alignment between the top and bottom of magnetic axis was achieved within 4mm. During the insertion, the laser tracker measurement was used in real time, tracking one point on CS inside. The CS insertion was carefully controlled by continuous adjustment and visual inspections, tracking data during the vertical insertion to precisely achieve the CS horizontal and circumferential positions. About 80 pressure sensors on the points which were more exposed to the risk of contact were monitored to identify collision between CS and TFC. Finally, a precise centering of the magnetic axis within $\pm$1.4mm with a vertical tilt of 1.6mm was achieved.
-
In-vessel components assembly
For the first plasma operation, in-vessel components such as various sensors, first walls and the upper divertor [C] have been installed. Precise assembly for these plasma facing components is required to avoid unacceptable local heat load on the inboard first wall for the limiter configuration at the plasma initiation and on the upper divertor for the divertor configuration. A final alignment of $\pm$1mm of the graphite tiles is required. The waving of the VV surface generates a strong challenge to this accuracy. The C-type steels (see fig.2) representing the interface between the VV surface and the graphite tiles have been precisely machined based on the VV surface measurements obtained by the laser tracker with T-probe. There were 15 points for this measurement along the interface of the C-type steels to the VV surface, and the interfaces were machined along optimum curve by these measured points. By the precise measurement and machining an accuracy of the graphite tile surface alignment within $\pm$1mm has been achieved.
[A] Y. Shibama, et al., in Proceedings of the 27th IAEA Fusion Energy Conference, 22-27 October 2018, Gandhinagar, India, IAEA-CN-258 FIP/P7-37 (2018).
[B] H. Murakami, et al., IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 28, NO. 3, APRIL 2018.
[C] D. Tsuru, et al. Phys. Scr. T171 ( 2020 ) 014023.Speaker: Dr Go Matsunaga (National Institutes for Quantum and Radiological Science and Technology) -
-
In magnetically confined fusion devices, nonlinear wave-wave interaction has been noticed to play important roles in the production of new modes. On NSTX, nonlinear interactions among low-frequency energetic particle modes (EPMs) and high-frequency toroidal Alfvén modes (TAEs) have been reported $[1]$. On JET, a 3/2 neoclassical tearing mode (NTM) is stabilized through the nonlinear coupling among 3/2, 4/3 and 7/5 modes $[2]$. On HL-2A, high frequency coherent modes can be driven by nonlinear wave-wave coupling $[3]$. Routinely, bispectral analysis is applied to detect the nonlinear interaction $[4]$. However, a number of statistical ensembles are necessary for the bispectral analysis. Here, we propose to use Hilbert transform $[5]$ for the analysis of nonlinear wave-wave interaction without many ensembles.
Two spontaneously excited waves could generate new waves through nonlinear interaction. Supposing there are two waves b and c, wave b is cos$\it{\theta}_b$ = cos(2$\pi\it{f}_bt+\it{\phi}_b$), and wave c is cos$\it{\theta}_c$ = cos(2$\pi\it{f}_ct+\it{\phi}_c$). Here, $\it{\theta}_b$ and $\it{\theta}_c$ are the time dependent instantaneous phases, and $\it{\phi}_b$ and $\it{\phi}_c$ are the initial phases. If nonlinear interaction exists between b and c, the generated new wave d should be the product of the two waves, cos $\it{\theta}_d$ = cos $\it{\theta}_b$ cos$\it{\theta}_c$ = cos(2$\pi(\it{f}_b+\it{f}_c$))t+$\it{\phi}_b$+$\it{\phi}_c$)) + cos(2$\pi(\it{f}_b-\it{f}_c$)t+$\it{\phi}_b$-$\it{\phi}_c$))/2. Wave d will have the frequency components of $\it{f}_b+\it{f}_c$, and the initial phases have to obey $\it{\phi}_d$ = $\it{\phi}_b$+$\it{\phi}_c$. This condition is actually $\it{\theta}_d$ = $\it{\theta}_b$+$\it{\theta}_c$. For simplification, we focus on the case of $\it{f}_d$ = $\it{f}_b+\it{f}_c$, the analysis of $\it{f}_d$ = $\it{f}_b-\it{f}_c$ is actually in a similar way. Let’s generate a synthetic signal of 10 ms with a sampling rate of 1 MHz, including three waves b, c, d and some random noise. The frequencies of the three waves are $\it{f}_b$ = 30 kHz, $\it{f}_c$ = 95 kHz and $\it{f}_d$ = $\it{f}_b+\it{f}_c$ = 125 kHz. If wave d is the result of mode coupling between wave b and wave c, the phase relation $\it{\theta}_d$ = $\it{\theta}_b$+$\it{\theta}_c$, i.e. $\it{\theta}_d-\it{\theta}_c$ = $\it{\theta}_b$ has to be satisfied. In Fig. 1(a), wave b (blue curve), wave c (red curve), wave d (orange curve) and the sum of b, c, d and a random noise (purple curve) are plotted. In Fig. 1(b), the power spectral density (PSD) of the purple signal in Fig. 1(a) shows the peaks of three waves at 30 kHz, 95 kHz, and 125 kHz. Supposing $\it{\theta}_b$ jumps at 0.5 ms, 1 ms, (in the intervals of 0.5 ms) to 9.5 ms (while $\it{\theta}_c$ does not jump), to satisfy the phase relation $\it{\theta}_d-\it{\theta}_c$ = $\it{\theta}_b$, $\it{\theta}_d$ has to jump at the same timings. In Fig. 1(a), the phase jumps of wave b and wave d at 0.5ms are labeled. The phase difference between wave d and wave c $\Delta\it{\theta}_{dc}$ = $\it{\theta}_d-\it{\theta}_c$, and the phase of wave b $\it{\theta}_b$ for Fig. 1(c) are shown in Fig. 1(d). The jump of both $\it{\theta}_{dc}$ and $\it{\theta}_b$ at 0.5ms is clearly observed. Due to the phase relation, the two phases are locked together. The fast fourier t ransform (FFT) bicoherence spectrogram is shown in Fig. 1(d). A bright spot is observed at ($\it{\theta}_c$, $\it{\theta}_b$) = (95, 30) kHz, indicating that the nonlinear coupling exists among waves b, c and d.
Fig. 2 (a) and (b) are the power spectrogram and the bicoherence spectrogram of the mirnov coil signal, respectively. The nonlinear coupling among AM1, AM2 and TM has been identified by observing the bright spot at (f1, f2) = (129 kHz, 10 kHz) and (f1, f2) = (139 kHz, 10 kHz). $\it{\theta}_{AM1}$, $\it{\theta}_{AM2}$ and $\it{\theta}_{TM}$ stand for the instantaneous phases of AM1, AM2 and TM, respectively. In Fig. 2 (b), the blue curve is the phase delay between AM1 and AM2, i.e. $\Delta\it{\theta}_{12}$ = $\it{\theta}_{AM1}-\it{\theta}_{AM2}$, and the red curve is $\it{\theta}_{TM}$. We could observe that $\Delta\it{\theta}_{12}$ and $\Delta\it{\theta}_{TM}$ are roughly synchronized with each other, and the maximum value of cross-correlation coefficient between them r($\Delta\it{\theta}_{12}$, $\it{\theta}_{TM}$) reaches 0.84, as shown in Fig. 2 (c). As a counter-example, we check the phases of AM1, AM2 during 696-698 ms and 8-12 kHz band-passed waveform during 676-678 ms. The first two waves and the last wave are in different timings, and the TM during 676-678 ms is not in 8-12 kHz, as shown in the power spectrogram in Fig. 2 (b). Therefore there are basically no nonlinear coupling among these three modes. The phase difference between AM1 and AM2 during 696-698ms $\Delta\it{\theta}_{12}$, and 8-12 kHz band-passed wave during 676-678ms $\it{\theta}_{8-12kHz}$ have been checked in Fig. 2 (e). In this figure, the phases are not synchronized at all, which means that nonlinear interaction does not exist. The maximum value of cross-correlation coefficient between $\Delta\it{\theta}_{12}$ and $\it{\theta}_{8-12kHz}$, r($\Delta\it{\theta}_{12}$, $\it{\theta}_{8-12kHz}$) is only 0.1, as shown in Fig. 2 (f).
Figure 2: (a) The power spectrogram; (b) The FFT bicoherence spectrogram; (c) The blue curve is the time evolution of $\Delta\it{\theta}_{12}$, and the red curve is the time evolution of $\it{\theta}_{TM}$. (d) The time evolution of r($\Delta\it{\theta}_{12}$, $\it{\theta}_{TM}$) during [-50, 50]$\mu$s. (e) The blue curve and the red curve are the time evolutions of $\Delta\it{\theta}_{12}$, and $\it{\theta}_{8-12kHz}$, respectively. (f) The time evolution of r($\Delta\it{\theta}_{12}$, $\it{\theta}_{TM}$) during [-50, 50]$\mu$s.As a summary, the Hilbert transform allows us to track the phase of a mode without many ensembles. In our work, the nonlinear interactions among the 139kHz and 129kHz AMs and the 10kHz TM on HL-2A were checked using phase tracking with Hilbert Transform. Results show that the phase delay between two AMs $\Delta\it{\theta}_{12}$ is approximately synchronized with the phase of TM $\it{\theta}_{TM}$ and the maximum of the normalized cross-correlation is 0.84. In a counter example, the maximum of normalized cross-correlation is only 0.18.
This work was supported by National Key R&D Program of China under Grant Nos. 2018YFE0310200 and 2017YFE0301203, National Natural Science Foundation of China under Grant No. 11775069, Sichuan Science and Technology Program Grant Nos. 19GJHZ0154.References:
$[1]$ Crocker N A, Peebles W A, Kubota S, Fredrickson E D, Kaye S M, LeBlanc B P and Menard J E 2006 Phys. Rev. Lett. 97 045002
$[2]$ Raju D, Sauter O and Lister J B 2003 Plasma Phys. Control. Fusion 45 369
$[3]$ Shi P W et al 2019 Nucl. Fusion 59 086001
$[4]$ Kim Y C and Powers E J 1979 IEEE Trans. Plasma Sci. 7 120
$[5]$ Ohshima S et al 2014 Rev. Sci. Instrum. 85 11E814Speaker: Linge Zang (Southwestern Institute of Physics) -
As the planned JET DT campaign [1] draws nearer, the necessity of high quality data, especially kinetic data, is more apparent than ever. This is especially the case of ion temperature data, which have required substantial work to overcome the diagnostic difficulties encountered in the early years of the JET ITER Like Wall (ILW) project, due to reduced light impurity levels and the presence of unwelcome tungsten lines in the light impurity charge exchange spectra (CXS) used for ion temperature and toroidal rotation measurements. Ion temperatures are usually determined from one of more impurities, rather than from the hydrogenic species, although measurements using the hydrogenic species have also been developed [2]. This paper shows how an analysis of the ion (fuel and impurities) power balance (PB) has helped to overcome these issues [3]. Such an analysis is in fact necessary for high power density discharges in which the temperature difference between different ion species can approach 10%, thereby impacting fusion yield calculations.
Temperature differences between species depend on conditions such as heating power to the different species and collisionality, with temperature differences between species limited by inter-species thermal equipartition and transport. In its simplest application, a PB calculation allows the determination of the the maximum sustainable temperature difference between ions and electrons, |Ti-Te|, thereby allowing to reject grossly erroneous measurements. A PB analysis is also required for estimating the errors of the ion and electron heat fluxes prior to any species-resolved transport analysis. The ion-ion temperature differences arise from the fact that collisional heating scales as Z2/A, but is heavily mitigated by strong inter-ion equipartition. Differences are usually found to be small (a few %), however, they can approach 10% in low to moderate density, high power discharges, such the ones under development for the JET DT campaign [1]. In fig.1 we show for several impurity species the local ratio Tz/Ti as a function of Qi/Q110, where Qi is the net ion heat flux in the main (deuterium) species and the normalisation is to Q110 which is a fictituous collisional equipartition power (equipartitionality) defined as , where ceq=3.2610-32 W eV1/2 m3, L11 is the Coulomb logarithm for hydrogen-hydrogen collisions, ni the density of species i and the integral extends from the magnetic axis to the flux surface under consideration. The most important assumption underlying this analysis is that the heat transport coefficients ci and cz be equal. In the JET core plasma of the highest power discharges Qi/Q100 >0.1, i.e. Tz/TD>1.05 for non-hydrogenic impurity species. In such cases the underlying temperature needs to be corrected using the PB analysis, rather than be assumed to be equal to the measured impurity temperature. We also note that for A>7, Tz/Ti is essentially species-independent. This is convenient, because measurements from several impurity species can be combined into a single composite profile without any PB corrections. Fig.1 is for trace impurity concentrations, however iterative calculations show that Tz/Ti decrease only weakly for Znz/nD up to 0.3, which is well above typical values.
In fig.2 we show Tz/TD for Qi/Q100=0.1 (at the high end of JET conditions) as a function of the deuterium fraction in a D-T plasma. Here Qi=QD+QT refers to both main species. While differences between impurities with A>7 remain small, we note large differences between hydrogenic temperatures. The differences between deuterium and tritium, about 5.5% at nD=nT, are however not likely to significantly affect fusion performance.
Finally, this analysis lead to a method for the reconstruction ion temperature profiles from ion temperature data available at only one or a small number of spatial locations [3]. We noticed that the equipartition power Qie is a nearly constant fraction, radially, of the deposited power Qis as calculated by heating codes. The actual values vary between discharge types and and are typically in the region 0.07-0.3 for discharges with dominant ion heating. A single measurement, such as the one provided for the plasma core by a 59Ni26+ K-shell at X-ray line at 1.6Å, measured with a crystal spectrometer, is sufficient to determine the value of Qie/Qis, providing, together with the PB analysis, acceptable impurity and main ion temperature profile when no CXS measurements are available.
[1] E. Joffrin et al., Nuclear Fusion 2019, https://doi.org/10.1088/1741-4326/ab2276
[2] B. A. Grierson et al, 2012, Review of Scientific Instruments 83, 10D529
[3] H. Weisen et al, 2020 Nucl. Fusion 60 036004 https://iopscience.iop.org/article/10.1088/1741-4326/ab6307/pdfSpeaker: Mr Henri Weisen -
Experiments in L-mode plasmas on HL-2A tokamak show that the electron and impurity transport is related to the normalized electron temperature gradient with opposite trends. In discharges with inner-deposited ECRH, the increase of the normalized electron temperature gradient in the confinement zone (0.25≤ρ≤0.5, ρ is the normalized minor radius) tends to pump out electrons and accumulate the trace Al impurity injected by laser blow-off (LBO), leading to slightly hollow electron density profiles meanwhile with an overall increase of impurity density. It is also found that the combined effects of the impurity expulsion by core magneto-hydrodynamic (MHD) activity in the plasma center and the strong impurity influx driven by turbulence in the outer confinement region is responsible for the strong hollowness of impurity density profiles. By contrast, the decrease of the normalized electron temperature gradient in discharges with outer-deposited ECRH leads to centrally peaked electron density profiles and a much smaller impurity accumulation. Gyrokinetic simulation has been performed with the gKPSP code
[1], confirming that the local change of the normalized electron temperature gradient affects the relative strength of the the ion temperature gradient (ITG) and trapped electron mode (TEM), eventually determining the electron particle transport in a way consistent with the experimental observations. Comparison of simulations with and without collisionallity shows that the plasma collisionality can significantly reduce the growth rate of the TEM and the associated outward particle flux.
In order to study the effect of turbulence on particle transport on HL-2A, we performed a set of experiments in which the power deposition of ECRH was modulated to locally alter electron temperature gradient as the turbulence drive[2]. According to the micro-instability analysis, the TEM will dominate over the ITG if a critical value of the normalized electron temperature gradient is reached. FIG.1 illustrates the dependence of experimental profiles of the electron temperature and its normalized gradient on the ECRH power deposition position, along with an additional profile measured in the phase without ECRH as reference. The inner-deposited ECRH increases electron temperature (from 0.7 keV to 1.8 keV) and its gradient (from 5 keV/m to 11 keV/m) around the power deposition radius of ρ=0.28, whilst the electron heating effect of the outer-deposited ECRH is less localized and lifts the electron temperature profile across the whole radius within the deposition position of ρ=0.7. In FIG.2, after the injection of Al impurity at 600 ms, the soft X-ray (SXR) reconstructed by a Bayesian tomogaphy method[3]appears to be deeply hollow in the inner-deposited ECRH case, whereas it is centrally peaked in the outer-deposited ECRH case. In particular, in discharges with inner-deposited ECRH, core MHD instabilities and sawtooth crashes are observed to have a dramatic influence on the impurity transport in the region inside the q=1 surface[4], as evidenced by the time evolution of emission profiles as shown in FIG.3. In three consecutive time points ($t_3$=602 ms→$t_5$=605.07 ms), the impurity density keeps increasing in the region outside the q=1 where an extraordinarily large density gradient has been built up, which is probably due to the turbulence characteristics of that region. Later on, this deeply hollow profile is flattened by a sawtooth crash (occurring at $t_5$=605.07 ms) on a time scale of approximately 0.15 ms, leading to a sharp increase of the impurity density inside the q=1 surface.
References:
[1]Qi L. et al 2016 Phys. Plasmas 23 062513.
[2]DeBoo J.C. et al 2010 Phys. Plasmas 17 056105.
[3]Li D. et al 2013 Rev. Sci. Instrum. 84 083506.
[4]Sertoli M. et al 2015 Nucl. Fusion 55 113029.Speaker: Dong Li -
In this paper, the JT-60SA cryogenic system and the results of the commissioning and annual operations are summarized. During the commissioning of the cryogenic system, performances for each component and the automatically controlled operation sequence have been confirmed. Notably, mitigation of cryogenic heat load fluctuations in large superconducting tokamak machines is essential. The automatic mitigation of variable loads becomes a challenging control to perform on large scale refrigeration system for tokamaks. Hence, for JT-60SA cryogenic system, an active control method for cryogenic heat load fluctuation has been successfully tested. We have established an active control method for cryogenic heat load fluctuation in the JT-60SA cryogenic system. The commissioning was completed toward an integrated commissioning test of the whole JT-60SA system for the first plasma discharge.
The National Institutes for Quantum and Radiological Science and Technology (QST) is constructing a superconducting tokamak JT-60SA at Naka Fusion Institute in Japan under an international collaboration between Japan and the European Union. To cool the superconducting magnets (total 650 tonnes), the thermal shields (100 tonnes), and the cryopumps (1.3 tonnes) and to supply cold gas for the High-Temperature Superconducting Current Leads (HTS-CL), a cryogenic system with an equivalent refrigeration capacity of 9 kW at 4.5 K has been provided as a French contribution to the project. To induce plasma currents and control the plasma shape and position, equilibrium field (EF) coils and Central Solenoid (CS) provide variable magnetic field flux to the plasma. The cryogenic heat load generated in the magnets varies significantly during the plasma experiment. The variation of the heat load leads to rapid changes in the cold return flow to the Refrigeration Cold Box (RCB), which has to be balanced by the compressors and turbines. The stationary 4 K heat load of 2.3 kW quickly reaches 7.4 kW during dynamic magnet operation. One of the design challenges for the cryogenic system was to mitigate the impact of the heat load fluctuation using a thermal damper with a dedicated active control system.
The cryogenic system mainly consists of the RCB and the auxillary cold box (ACB), warm compressors (WCS), and gaseous He storage tanks. Figure 1 shows a simplified flow diagram of the JT-60SA cryogenic system. The RCB supplies cold helium gas to the magnets, thermal shields, HTS-CLs, and the cryopumps through the ACB during cool down. During magnet and cryopump operation, liquid He (LHe) is collected in a 7-m3 thermal damper at about 4.4 K and a smaller LHe tank at 3.6 K in the ACB to precool the supercritical helium (SHe). The required flow rate to the cryogenic users is the largest among the existing fusion experimental devices. During nominal operation, the cryogenic flow rates are as follows:- Two cold helium circulators supply a total 1.8 kg/s of SHe at 4.4 K
for the 18 toroidal field coils, six EF coils, four CS modules, and
the coil structures. - Besides, the RCB supplies 25 g/s of cold He gas at 50 K for the HTS-CLs, 404 g/s at 80 K for the thermal shields,and 270 g/s at 3.7 K for the cryopumps.
Commissioning of the refrigerator and annual operations for the cryogenic system have been performed [a, b]. The main test items were as follows;
- Performance measurement for each component (WCS, cold circulators,
turbines, cold compressors, oil separators). - Operation sequence tests: purification, cool down, warm up, short standby mode, the baking mode of the plasma vacuum vessel, and nominal operation modes.
- Performance test for a stationary equivalent 4.4-K refrigeration
capacity of 6.5 kW. - Heat load fluctuation tests simulating different plasma scenarios
- Recovery after a loss of electrical power
The active reaction to heat load fluctuation is essential for the cryogenic operations of the magnet system, which requires variation of current operation in the tokamak, including ITER. Such a control system was challenging so far in a large scale cryogenic system.
An advanced active mitigation method of heat load fluctuation has been demonstrated in the JT-60SA cryogenic system, using the following methods;- Reduction of the flow rate of circulated SHe after plasma discharge to avoid the sudden hot He return from the magnet system to the damper.
- Temporary absorption of the heat by helium in the thermal damper and accumulation and pressurization of the evaporated He gas to stay within the suction capacity of the WCS.
In the test, the heat load was stepwise increased to 7.4 kW (heat load: 100%) using heaters. In the annual operation, control parameters to adapt various heat load conditions have been investigated. Figure 2 shows typical results of mitigated heat load fluctuations for various heat load conditions. The control system regulated the return flow to the RCB within the capacity of the WCS and kept the turbine inlet temperature above 27 K slightly above the turbine´s trip temperature of 26 K.
We have established the active control method to mitigate the impact of cryogenic heat load fluctuation on the JT-60SA cryogenic system. These results contribute to the preparation of the stable operation of the magnet system in JT-60SA plasma experiments during integrated commissioning of JT-60SA in 2020.
[a] K. Kamiya et al., IOP Journal of Physics: Conf. Series 897 (2017)012015
[b] C. Hoa et al., IOP Conf. Series: Materials Science and Engineering 278 (2017) 012104Speaker: Mr Kazuya Hamada (National Institutes for Quantum and Radiological Science and Technology) - Two cold helium circulators supply a total 1.8 kg/s of SHe at 4.4 K
-
In this paper, we present the recent experimental results of cross phase influence on turbulent momentum and particle transport in the edge of HL-2A tokamak. The mathematical expressions for cross phases are derived in Fourier domain. The fluctuations and turbulent flux are measured by Langmuir probes. For Reynolds stress, prominent phase scattering in the strong shear layer is found, which indicates a phase slipping state as predicted by theory. With ECRH, the cross power is more concentrated in the high $ \gamma cos(\varphi) $, which shows an agreement with the dynamics of spectral symmetry breaking in the development of poloidal torque. For particle flux, phase scattering increases with RMP, which could be induced by field line stochastisation due to magnetic flutter. Comparative statistical study shows that, fluctuations in particle flux exhibit a big deviation from Gaussian distribution, while fluctuations in Reynolds stress don’t. This suggests that edge transport models based on quasi-linear theory need to be reconsidered. This may be related to L-H transition physics. The cross phase dynamics may be also relevant to I-mode, which shows L-mode like density profiles, but H-mode like temperature profiles, with phase slipping in heat flux but locked in particle flux.
High confinement regimes are expected to be the operation scenarios of future fusion reactors such as ITER. A fundamental feature of these regimes in magnetic confinement devices is the formation of transport barriers, and the dramatically decreased turbulence transport. The primary physics involved in transport reduction is the effect of $ E\times B $ velocity shear owing to a decrease in the intensity of fluctuations, as well as the change in the cross phase between them $ {^1} $. The transient increase in $ E\times B $ sheared flows plays an important role in the L-H transition. The turbulent flux of momentum—Reynolds stress—is a mechanism thought to be responsible for the generation of sheared flow driven by turbulence $ {^2} $. Reynolds stress and fluxes are sensitive to cross phase. However, usual quasilinear theory does not treat phase as dynamic. Study of cross phase influence on turbulent transport and the interaction between sheared flow and cross phase, is significant for understanding the high confinement regimes.
Ensemble-averaged Reynolds stress or particle flux is written as the product of fluctuation intensities and a cross phase factor: $ 〈 \tilde{v} _r \tilde{v }_\theta 〉=\sigma_{ \tilde{v }_r }\cdot \sigma_{ \tilde{v} _\theta}\cdot X_{RS} $, $ 〈 \tilde{n} \tilde{v}_r 〉=\sigma _{\tilde{n}} \cdot \sigma_{ \tilde{v }_r} \cdot X_{PF} $. By using Fourier transform, the mathematical expressions for $ X_{RS} $ and $ X_{PF} $ are derived:
$ X_{RS}=\frac{\sum_\omega P_{ \tilde{v }_r \tilde{v }_r}(\omega)^{1/2} P_{ \tilde{v }_\theta \tilde{v }_\theta}(\omega)^{1/2} \gamma_{ \tilde{v }_r \tilde{v }_\theta} (\omega)\ cos\varphi_{ \tilde{v }_r \tilde{v }_\theta}(\omega)}{(\sum_\omega P_{ \tilde{v }_r \tilde{v }_r}(\omega))^{1/2}\ (\sum_\omega P_{ \tilde{v }_\theta \tilde{v }_\theta}(\omega))^{1/2}}, $ (1)
$ X_{PF}=\frac{\sum_\omega P_{ \tilde{n } \tilde{n } } (\omega)^{1/2} P_{ \tilde{v }_r \tilde{v }_r}(\omega)^{1/2} \gamma_{ \tilde{n } \tilde{v }_r}(\omega)\ cos\varphi_{ \tilde{n } \tilde{v }_r} (\omega)}{(\sum_\omega P_{ \tilde{n } \tilde{n }}(\omega) )^{1/2}\ (\sum_\omega P_{ \tilde{v }_r \tilde{v }_r} (\omega) )^{1/2}}. $ (2)
Figure 1 shows the cross phase evolution of Reynold stress in Ohmic and ECRH heated L-mode discharges. Figure 1(a1-a3) show the normalized cross power profile between $ \tilde{v }_r $ and $ \tilde{v }_\theta $, and its distribution on coherence $ \gamma $. $ \gamma $ drops in the strong shear layer near [-1,1] cm. This shows the turbulence decorrelation by the sheared flow. Figure 1(b1-b3) show the normalized cross power distribution on cross phase angle $ \varphi $. The phase scattering in the strong shear layer is larger than other regions. The phase angle tends to distribute in [-$ \pi $/2,$ \pi $/2] at [-2,-1] cm, and [$ \pi $/2,$ \pi $] at [1,2] cm. This is consistent with a novel theory, which indicates two different states of the cross phase: slipping in strong shear regimes and locked in weak shear regimes $ {^3} $. Figure 1(c1-c3) show the cross power distribution on X factor $ \gamma cos(\varphi) $. X factor changes from negative to positive from SOL to edge. With ECRH, the cross power is more concentrated in the high X factor region. This exhibits good agreement with the dynamics of spectral symmetry breaking in the development of poloidal torque $ {^4} $.
Figure 2 shows the cross phase evolution of particle flux in Ohmic discharges with and without RMP. Figure 2(a1) shows that the cross power is more concentrated in the positive $ \varphi $~0.2 at [0.5,1.5] cm, and concentrated in the negative $ \varphi $~-0.2 at [-1.5,-0.5] cm. The phase scattering in the strong shear layer near [-0.5,0.5] cm is larger than other regions. Figure 2(a2) and (b2) show that , with RMP, the phase scattering increases and the cross power is more concentrated in the higher coherence. This could be induced by field line stochastisation due to magnetic flutter.
Figure 3 shows the joint PDF of $ \tilde{v }_r $-$ \tilde{v }_\theta $ and $ \tilde{n } $-$ \tilde{v }_r $ at ±1.25cm in Ohmic discharges. Unlike the joint PDF of $ \tilde{v }_r $ and $ \tilde{v }_\theta $, the center of $ \tilde{n} $-$ \tilde{v }_r $ PDF shows a large deviation from (0,0) point. This indicates the distributions of fluctuations have a big deviation from Gaussian whose skewness should be 0. Particle flux is necessarily nonlinear in $ \nabla n $, leading to an asymmetric PDF tail due to intermittent transport events. In contrast, the Reynolds stress is secondary to the particle flux. It has weaker $ \nabla n $ dependence, and so is closer to Gaussian. More statistical studies (skewness, kurtosis, Hurst parameter) of cross phase are planned for future work.References
1 Boedo, J.A., et al., Physical Review Letters, 2000. 84(12): p. 2630-2633.
2 Diamond, P.H. and Y.B. Kim, Physics of Fluids B, 1991. 3(7): p. 1626-1633.
3 Guo, Z.B. and P.H. Diamond, Physical Review Letters, 2015. 114(14): p. 145002.
4 Long, T., et al., Nuclear Fusion, 2019. 59(10): p. 106010.Speaker: Dr Ting Long (Southwestern Institute of Physics) -
Core density profile peaking and particle transport have been recently extensively studied on several tokamaks [1,2,3]. In JET, the earlier research of the significance of the NBI fueling on density peaking 4 was recently confirmed when thanks to the development of the gas puff modulation technique it was found that in ITG dominated plasmas, NBI is responsible for typically half of the density peaking 1. Here we aim to quantify the role of NBI fueling in contributing to density peaking in JET by executing identity discharges between the ICRH and NBI heated plasmas. By performing dimensionally matched identity plasmas, we can also learn how the different heating systems affect plasma confinement, MHD, impurities, radiation, pedestal, ELMs, the effect of rotation on transport and gas puff modulation properties.
Up to 15s H-mode plasmas with 8MW of ICRH power were executed on JET, with the JET record high injected ICRH energy of 108MJ. The ICRH discharges, using the H minority heating scheme at 3% minority concentration, were stationary without any major MHD activities or impurity accumulation. NBI heated discharges were executed consecutively to match the dimensionless plasma profiles of q, ρ, υ, βn and Ti/Te using the same shape. Both pulses had gas puff modulation at 3Hz throughout the discharges to extract perturbative particle transport coefficients. The ELM frequencies are 75Hz and 40Hz for the ICRH and NBI pulses, respectively, with mixed ELMs and affected by the gas puff modulation.
The time traces of the key parameters for the two discharges are shown in figure 1 and the main dimensionless and dimensional profiles are compared in figure 2. The dimensionless profiles of q, ρ, υ, βn and Ti/Te≈1 were matched within 5% difference except in the central part of the plasma (ρtor<0.3). The most significant difference, in addition to rotation or Mach number (see table 1), is the density profile which is ~50% more peaked for the NBI discharge than for the ICRH peaked discharge. Although the dimensionless profiles are well matched, there are other differences between the discharges that could modify the density profile and particle transport, in addition to the obvious influence of the NBI fueling to explain the increase in density peaking 1.
The main engineering and global dimensionless parameters are compared in table 1. The general conclusion is that 8MW of NBI and ICRH (roughly 4MW goes to ions and 4MW to electrons for each heating case) create similar plasma profiles and performance in JET despite various differences, such as larger steady-state gas puff needed in the NBI pulse to get the same pedestal height in ne. The main differences are the toroidal rotation (10km/s counter-Ip in ICRH pulse and 110km/s co-Ip for the NBI pulse), power deposition profiles, fast ion content (14% versus 8%), ELM characteristics, radiation and heavy impurity concentration. Although Zeff is very similar between the two pulses, the concentration of heavy impurities, such as W and Ni, is a factor 2-4 higher in the case of the ICRH discharge. Therefore, the higher W concentration and higher W peaking in ICRH plasmas does not explain the higher ne peaking in NBI plasmas. Concerning the differences in the MHD activity, the sawtooth frequency goes from 3Hz to 6Hz from the ICRH to NBI pulses and similar trend is seen for fish-bone activity. NTMs are minor in both cases. The effect of MHD activity is limited to the central region (ρtor<0.3), and does not affect the core (0.3<ρtor<0.8) transport analysis. The pedestal is similar between the shots and thus does not explain the higher ne peaking in NBI plasma. The possible candidates left are the rotation and fast ion content modifying particle transport or the NBI fueling.
Gas puff modulation technique has been developed with high quality time-dependent density profile measurements to determine particle transport coefficients on JET 1. The experimental density modulation amplitude and phase together with the determined perturbative D ̃ and V ̃ profiles are shown in figure 3. The phase profile is flatter in the case of ICRH, giving rise to higher diffusion, which is clearly seen in figure 3. Although these profiles represent perturbative transport coefficients, this suggests that the steady-state particle transport could also be different between the two pulses and contribute to the difference in ne peaking.
In order to quantify how large fraction from the difference in ne peaking between the ICRH and NBI heated plasmas is due to the NBI particle source or whether there are also transport changes due to rotation or fast ion content in play, integrated transport modelling with ETS/TGLF will be performed. For the first time integrated modelling by including the actual gas puff modulation with time varying density modulation is performed with JINTRAC/QualiKiz. The perturbative transport coefficients from the transport modelling can be directly compared to the experimental ones. Finally, gyro-kinetic GENE simulations are performed to quantify the role of NBI fueling versus changes in particle transport.
The ICRH versus NBI identity plasmas in JET show that the NBI fuelled discharge has 50% higher density peaking. Otherwise the ICRH and NBI heated identity plasmas yield similar plasma parameters and performance despite the obvious differences in rotation and fast ion content. This result is valid at 8MW of heating power level, however, it remains to be seen how this promising result will scale to larger power levels and to larger devices.
1 T. Tala et al., Nucl. Fusion 59, 126030 (2019).
2 S. Mordijck et al., “Collionality driven turbulent particle transport changes in DIII-D H-mode plasmas”, submitted to Nucl. Fusion (2019).
3 E. Fable et al., Nucl. Fusion 59, 076042 (2019).
4 L. Garzotti et al., Nucl. Fusion 46, 994 (2006).Speaker: Tuomas Tala (VTT, Association Euratom-Tekes) -
The Lithium Tokamak eXperiment-$\beta$ (LTX-$\beta$), the upgrade to LTX, is designed to utilize low recycling walls and the resultant gradient-free temperature profiles [D. P. Boyle et al., Phys. Rev. Lett. 119, 015001 (2017)] to robustly stabilize ion and electron temperature gradient-driven modes. Low recycling and the resultant low collisionality in the scrape-off layer (SOL) plasma [R. Majeski et al., Phys. Plasmas 24, 056110 (2017)] are also expected to broaden the divertor (or limiter, in LTX-$\beta$) power footprint to reduce peak power handling requirements. In LTX-$\beta$, confinement and the SOL plasma collisionality both depend strongly on global recycling. Global recycling has a strong effect on confinement in many tokamak experiments, yet no systematic study of the variation of confinement with recycling has ever been undertaken. Here we report on the preliminary results of the first such study, in LTX- $\beta$ (Fig. 1).
The upgrade has approximately doubled the toroidal field (to >3 kG) and will double the plasma current (to 150 – 175 kA), compared to LTX. Plasma pulse length is limited to ~50 msec. All plasma facing surfaces are coated with lithium, using the new lithium evaporators shown in Fig. 1. LTX-$\beta$, like LTX, employs a heated high-Z liner which can liquefy the lithium coatings, to provide liquid lithium plasma facing surfaces. Short-pulse (up to 6 msec) neutral beam injection has been added - up to 600 kW (35A at 18.5 kV) of neutral beam power has been injected. As of this submission, modifications to increase the plasma current are still underway, and the current is limited to <100 kA, so that neutral beam injected fast ion first orbit losses are significant. Modeling results for the ion orbit losses, along with results at increased plasma current, will be presented. Flat electron temperature profiles were observed during operation of LTX with low recycling lithium coated walls prior to the outage for the upgrade to LTX-$\beta$, and there are now indications that the low field side edge temperature in LTX-$\beta$ increases by 2-3× to ~60 eV within 8-10 msec after gas puffing is terminated (Fig. 2).
Electron temperatures of 200 – 250 eV have been observed for discharges with ~ 90 kA plasma current; an electron temperature profile is shown in Fig. 3.
At present the Thomson scattering diagnostic cannot view the plasma axis, at R~33 cm in this case, due to background light from the reflective lithium coated PFCs. Additional Thomson scattering upgrades are being installed, and will be operational by summer 2020, including supplemental views on the high field side of the plasma axis, as well as sightlines in the plasma SOL.Measurements of the pressure evolution in the torus before, during and after a discharge indicate that the lithium wall retains > 2/3 of the fueled hydrogen. An array of Lyman-α detectors viewing the high field side plasma contact area is being installed, and will be used with DEGAS 2 modeling to determine the recycling coefficient.
LTX-$\beta$ is a collaborative effort, with major participation from Oak Ridge and Lawrence Livermore National Laboratories (ORNL and LLNL), as well as the University of Wisconsin, UCLA, the University of Washington, Princeton University, and LiWall Fusion (a private company). The ORNL Charge Exchange Recombinant Spectroscopy (CHERS) system views lithium impurity light. Initial results indicate that the discharge ion temperature is 40 – 80 eV, for discharges in which the peak electron temperature is 100 – 125 eV.
High field side swept Langmuir probes have been installed, and are providing edge electron temperatures and densities. A “vacuum suitcase” based surface exposure probe is operational. Samples are transferred to an x-ray photoelectron spectroscopy (XPS) facility which shares a building with LTX-$\beta$; the time interval between exposure on LTX-$\beta$ and completion of the surface analysis is ~90 minutes. Continuous vacuum pumping ensures that minimal additional contamination occurs during transfer. UCLA is upgrading the LTX reflectometer and interferometer capabilities for both coherent and turbulent density fluctuation measurements. An upgrade to the frequency-modulated continuous-wave reflectometer will double the previous tracking rate of the density profile, and the same hardware will be used as a high-$k_{r}$ backscattering diagnostic. A new dual-channel tunable fixed-frequency reflectometer will simultaneously monitor density fluctuations at two locations in the plasma core. A new detector will increase the sensitivity of the 1 mm interferometer. Far-forward scattering measurements with this diagnostic will probe density fluctuations with $k_{perp}<2 cm^{-1}$.
The most recent results from LTX-$\beta$ will be presented.
This work was supported by USDoE contracts DE-AC02-09CH11466, DE-AC05-00OR22725, and DE-AC52-07NA27344.
Speaker: Richard Majeski (Princeton Plasma Physics Lab) -
Introduction
In ITER the density of deuterium and tritium will be controlled by injection of cryogenic pellets, with D and T isotopes separately to allow active isotope ratio control. Due to technical limitations and high pedestal temperatures the pellets will be ablated at the plasma periphery, about at the pedestal top. For deeper penetration one have to invoke curvature drift and inward particle transport. Another parameter governing the ITER pellet fuelling is rather small relative size of the pellets which is difficult to match in medium size machines but just possible by small pellets in JET. This paper present unique pellet fuelling dataset from JET addressing aforementioned issues. As a result for the first time pellets of ITER like relative size are injected into high current high power JET plasmas.Deuterium pellets into hydrogen plasma
The first experiment [1.] was performed with plasma fuelling provided by hydrogen gas and hydrogen beams. In addition, RF heating at second harmonic hydrogen resonance is used. Plasma current was 1.4MA. After stationary hydrogen H-mode is established, fuelling by ~20mm3 deuterium pellets from the high field side is applied (see figure 1). The reduction from nominal pellet size of 40mm3 was possible by “double cut” method. During injection of deuterium pellets into hydrogen plasma the isotope mix ratio was measured by four independent methods: Balmer-alpha spectroscopy, Penning gauges, charge exchange spectroscopy in the plasma core, and from neutron rate. The last uses the modelling of particle transport JETTO code combined with HPI2 pellet deposition code. All methods show that isotopic mix is close to $n_D/\left(n_H+n_D\right) = 0.45$ during the flattop.
The pellet fuelling flux required for this mix is $\Phi_{pel}=0.045 P_{aux}/T_{e,ped}$ or $\Phi_{pel,D\rightarrow D}=0.10P_{aux}/T_{e,ped}$ to extrapolate to the case of deuterium pellets and deuterium plasma. The front coefficient 0.1 is within a factor 1.4 - 2 the same than found in our previous pellet fuelling experiments in pure deuterium plasmas in AUG [2].
When pellet is injected into plasma it creates transiently a zone of reversed gradient of deuterium density. As a consequence, the deuterium ion particle flux reverses from outwards to inwards and consequently deuterium concentration in the core increases as manifested by increase of neutron rate signal in figure 1. This situation is transient. When evaluated during post pellet phase at $r/a=0.5$ the deuterium particle duffusivity is $D_D=1.38m^2/s$ or $D_D/\chi_{eff}=0.41$.
The neutron rate after the first pellet in figure 1 increases faster than predicted by JETTO Bohm-gyro Bohm model indicating that penetration of deuterium to the plasma core is faster at the beginning of pellet train compared to the later pellet fuelling phase. The gyro-kinetic simulations show that indeed in ion temperature gradient regime the diffusivities for both ion species are higher than the electron diffusivity $D_H \sim D_D>D_e$ [3]. The detailed gyro-kinetic modelling of core particle transport under our condition of isotope mixing by pellets is now ongoing and is subject of separate paper [3].
Hydrogen pellets into deuterium plasma
In second experiment hydrogen pellets are injected into deuterium plasma while the gas fuelling was minimised. The plasma current was set to 1.4MA and 3MA where the low current plasma was aimed to connect to the experiment described above. At 3MA the heating power was ~26 MW. Both pellet sizes, $40mm^3$ and $6mm^3$, are used with 10Hz and 40Hz resp. Under these conditions the mix ratio $n_H/(n_H+n_{D)})\sim 0.4$ is obtained. For each current and pellet size a reference plasma is produced with deuterium pellets into deuterium plasma to complete the dataset, including 3.5MA, high power plasma.
In the above dataset, in particular data with smaller $6mm^3$ pellets are unique as they have the pellet/plasma particle ratio and the ablation depth the same as fuelling pellets in ITER. High temporal resolution interferometry is used to calculate amount of deposited particles during ablation and compared with pellet size from microwave cavity measurement in order to determine injection efficiency. The data indicate high injection efficiency for ablation length $r/a>0.15$ and falling sharply for shallower pellets. The low efficiency for shallower pellets is correlated with prompt ELMs. No difference between hydrogen and deuterium pellets is observed so far. When confirmed these results are important for ITER, in particular that the pellet deposition efficiency is a result of a delicate balance between losses while pellet ablates inside the pedestal and transport of pellet particles beyond the pedestal by curvature drift. Finally note that these observations refer to the time interval during pellet ablation which lasts about 1 ms. After pellet is ablated the behaviour of deposited material is controlled by ELMs, which could be quite complex and will be analysed separately.Acknowledgements
This work was carried out within the framework of the EUROfusion Consortium and received funding from the Euratom research and training programme 2014–2018 and 2018-2020 under grant agreement No. 633053 and from the RCUK Energy Programme grant No. EP/P012450/1. The views and opinions expressed herein do not necessarily reflect those of the European Commission.[1.] Valovič M et al Nucl Fusion 59 (2019) 106047
[2] Valovič M et al Plasma Phys. Control. Fusion 60 (2018) 085013
[3] Marin M.et al this conferenceSpeaker: M Valovic (CCFE, Culham Science Centre, Abingdon, OX14 3DB, UK,) -
The 2018 ITER Research Plan states that “Operation of ITER will have to strongly focus on avoiding disruptions with a high success rate and on mitigating those in which avoidance techniques fail” (1). The development of a disruption mitigation system for ITER will not suffice. We discuss a nonlinear effect that can allow the RF current drive stabilization of larger islands than would otherwise be possible (2). As we discuss, disruption studies on JET suggest that this could have a substantial impact on the rate of disruptions in ITER.
It has been estimated that ITER will need to maintain a mitigated disruption rate of less than 1% to keep cumulative damage to the first wall at an acceptable level. JET has had a 16% rate of unintended disruptions with the ITER-like wall (3). Furthermore, clean-up after each mitigated disruption will lead to a loss of valuable machine time, and each mitigated disruption will also be associated with some unavoidable risk.
The nonlinear effect that we have identified, “RF current condensation”, concentrates the RF driven current near the island O-point and thereby increases the efficiency of RF current drive stabilization of islands. Fig. 1 shows the predictions of a simple model to estimate the effect of RF condensation on the efficiency. The efficiency metric, the ratio of the resonant component of the RF driven current to the total RF driven current (4), is widely used. The deposition model takes the linear contribution to the power deposition to be a constant inside the island and to vanish outside the island (4). The figure shows the stabilization efficiency as a function of normalized power, $P\equiv W_i^2w^2P_0/(4n\kappa_{\perp}T_0)$, where $W_i$ is the island width, $w$ is the ratio of the wave phase velocity to the thermal velocity, $n$ is the density, $\kappa_{\perp}$ is the cross-field thermal diffusivity, $T=T_0+\tilde{T}$ is the perturbed temperature at the O-point of the RF heated island, and $P_{RF}=P_0\exp(w^2\tilde{T}/T_0)$ is the RF power deposition. This expression for the power deposition is a consequence of the fact that, for electron cyclotron waves and lower hybrid waves, the power is deposited on the electron tail. The expression for $P_{RF}$ leads to a nonlinear thermal diffusion equation in the island whose solution determines the profile of the RF driven current. The efficiency increases rapidly as $P$ approaches a bifurcation point of the nonlinear thermal diffusion equation in the island, above which we must include additional physics to determine the saturation of the temperature. Additional pieces of physics that can saturate the temperature increase are depletion of the wave energy (5) and temperature profile stiffness above the microinstability threshold (6). The efficiency increases further as the power is increased above the bifurcation threshold. Hysteresis behavior arises that can cause stabilized islands to shrink to smaller widths than would otherwise be the case.In ITER, a relatively small fraction of the available EC power will be used for routine ECCD control of neoclassical tearing modes, stabilizing islands when they are still small to minimize the impact on the power balance. We address the situation where an off-normal event leads to the appearance of a large island that threatens to cause a disruption. In that circumstance, it may be desirable to use all available power to avoid a disruption.
On JET with the ITER-like wall, 95% of the disruptions are preceded by the appearance of locked islands (3). A statistical analysis of JET disruption data has found that there is a distinct island width at which the plasma disrupts, corresponding to a width of about 30% of the minor radius (7). These observations suggest that, regardless of the initial trigger, the islands are playing a key role in initiating the thermal quench. Islands grow on a global resistive time scale. However, ramp-down of a plasma discharge also occurs on a resistive time scale, and it can itself trigger a disruption if it is too fast. One advantage of ECCD stabilization is that it is intrinsically fast. The imposition of RF current drive in a magnetic island produces a stabilizing electric field on an electron-ion collision time (8). There is a longer actuator time scale, but that time scale is still fast compared to the resistive time scale on which the islands grow (9).
The use of ECCD for off-normal event response will involve larger islands and higher ECCD powers than routine stability control. The conventional model used for studies of ECCD active stability control breaks down in this context (2). The conventional model assumes that the local acceleration of electrons is not affected by the presence of the island. This model neglects the sensitivity of the power deposition and electron acceleration to the perturbation of the temperature in the island. There is a nonlinear feedback on the temperature in the island, with the increased power deposition giving rise to a further increase in the temperature. The combination of this with the sensitivity of the RF-driven current to the temperature can produce the RF current condensation effect. The nonlinear effects must be taken into account for aiming of the ray trajectories when large islands are present to avoid a “shadowing” effect that can deplete the EC power before the center of the island is reached.
The RF condensation effect motivates a reevaluation of the potential utility of lower hybrid current drive (LHCD) for island suppression. Although the LHCD deposition profile is intrinsically broader than that of ECCD, LHCD is more sensitive to temperature perturbations, and the condensation effect localizes it. Fig. 2 shows a calculation done with the GENRAY ray tracing code and the CQL3D Fokker-Planck code of the effect of imposed temperature perturbations on the LH deposition at the $q=2$ surface in an ITER equilibrium (10). There is the prospect that a broad LHCD profile used to maintain a steady state tokamak can provide automatic stabilization of islands by RF current condensation.
A high fidelity code for simulation of RF condensation has been constructed, with the power deposition along EC ray trajectories calculated by GENRAY, then mapped into a magnetic island. A self-consistent solution of the thermal diffusion equation in magnetic island geometry feeds back to determine the perturbed temperature profiles along the ray trajectories, and the resulting modification of the power deposition. Fig. 3 shows the result of 500 calculations of the bifurcation threshold for a 20% island in an approximate ITER L-mode equilibrium as a function of poloidal and toroidal launch angles and launch position. The perturbed temperature at the bifurcation point has been plotted as a function of a parameter that measures the sensitivity of the power deposition to the temperature perturbation. For each point plotted here, the EC power was incrementally increased until the bifurcation threshold was encountered, with the EC power constrained to lie below 20 MW.This work was supported by U.S. DOE contracts DE-AC02-09CH11466, DE-SC0016072, DE-FC02-04ER54698, and DE-FG02-91ER54109.
References
(1) ITER Technical Report ITR-18-003 (2018).
(2) A. H. Reiman and N. J. Fisch, Phys. Rev. Lett. 121, 225001 (2018).
(3) S. N. Gerasimov et al, IAEA FEC, IAEA-CN-258/151 (2018).
(4) C. Hegna and J. Callen, Phys. Plasmas 4, 2940 (1997).
(5) E. Rodriguez, A. H. Reiman and N. J. Fisch, Phys. Plasmas 26, 092511 (2019).
(6) E. Rodriguez, A. Reiman, N. Fisch, arXiv:2001.09044.
(7) P.C. de Vries et al, Nucl. Fusion 56, 026007 (2016).
(8) A. H. Reiman, Phys. Fluids 26, 1338 (1983).
(9) F.M. Poli et al, Nucl. Fusion 58, 016007 (2018).
(10) S. Frank, A. Reiman, N. Fisch, P. Bonoli, to be submitted.Speaker: Allan Reiman (Princeton Plasma Physics Lab) -
The disruption mitigation technology remains the key issue of safe and reliable device operation in future large tokamaks including ITER [1,2]. Several approaches have been proposed and experimentally tested in contemporary devices, which demonstrate opportunities of massive gas, pellets, dust and liquid gets injection in preventing the avalanche as the most dangerous mechanism of the runaway electrons. Physics of the avalanche [3] is determined by a very high electric field generated in tokamak at the final stage of the thermal quench that provides conversion of the plasma current from thermal electrons to runaways. It was shown that effective tool for the runaway avalanche mitigation is a fast growth of the plasma density above so called Rosenbluth density via techniques mentioned [4]. This density value is 100~1000 times higher than the plasma operation density. The mass of injected matter being in a kilogram range negatively affects technology systems sited the in-vessel and requires long-term recovery of the tokamak device in the created conditions. In this report, we analyze a novel approach aiming at an essential reduction of seeds causing the avalanche runaway electron generation after the thermal quench but does not use injection into the device vacuum vessel a large mass of gas, liquid or solid/dust matter.
The essence of the approach is to inject a projectile into the plasma from the material that is from the list of PFC materials (W, C, Be). The Fig.1 shows a schematic of the approach using the tungsten rod ~8 mm in diameter and 80 mm in length, that crosses the plasma volume with the velocity of 0.8 km/s perpendicular to the toroidal magnetic field. The projectile is injected just after the thermal quench (TQ) and crosses the plasma dimension in equatorial plane (4 m) at a time of ~5 ms. It collects all runaway electrons existing in plasma sequentially cleaning magnetic surfaces from runaways remained in the plasma after TQ (Fig.1). Such cleaning allows us to escape the runaway avalanche (or delay the time of the avalanche development) if the amount of primary runaways born during the plasma operation would be significantly reduced. The optimal scenario for this technology uses the following steps: control of the plasma stability and switching on the rail gun at the finish of the thermal quench; accelerating the projectile during 1.5 ms in the equatorial zone of device being aimed at collection of the seed electrons crossing the plasma during 5 ms; additional reconnection events stimulating seed runaway losses; capturing the injected projectile inside the collector sited inside the inner blanket zone of the tokamak-reactor.
Estimations of the tungsten projectile approach are performed for Basic Plasma Performance regime of ITER as in Ref. [5]. The stopping power due to interaction of runaways with the W projectile is evaluated by the relation L(mm) = 0.62[E(MeV)-0.106] from Ref. [6]. Runaway electrons within the 1~25 MeV energy range are terminated by the W projectile with the 8 mm dimension along the magnetic field. The 80 mm length and the 0.8 km/s speed of the projectile were chosen to provide existence of the projectile shadow at the magnetic surface needed to collect runaways during their ~800 toroidal transits within one-third part of the minor radius where the main source of seeds is expected.
Fig. 2 demonstrates temperature and density profiles prior and after the TQ together with the seed hot tail density nseed ~ 5x1012 m-3 that is evaluated assuming the 1 ms temperature decay time as in Ref. [7] and uniform profiles during TQ. The estimated nseed value corresponds to the hot-tail seed current Iseed ~ 5 kA. Fig. 3 demonstrates evolution of the total and runaway currents during the current quench (CQ) stage of the disruption calculated for the strong electric field approach [2] with the Ohmic decay time of ~0.3 s corresponding to TeCQ = 40 eV, Zeff = 1.7 and the e-fold avalanche time of ~19 ms. One can see that for Iseed ~ 5 kA the large runaway current of Irun ~ 6 MA will replace the total current at ~0.5 s. Two orders the Iseed reduction results in delay of the runaway current rise so that the replacement will take place at ~1.0 s with Irun ~ 1.5 MA. This seems acceptable since the poloidal magnetic energy content is reduced by two orders of magnitude in comparison with that for the initial plasma current 15 MA.
Simulations of the projectile-plasma interaction show that the projectile surface temperature of ~ 1500 K will not exceed the tungsten sublimation threshold 5828 K, so that evaporation of the projectile can be neglected.
Estimates of the characteristics of a 0.6 m railgun made in accordance with [8] demonstrate the possibility of accelerating a projectile weighing 80 g to a speed of 800 m/s for 1.6 ms at a 1 MA railgun current in the magnetic field of a tokamak reactor ~5 T.
In conclusion, it should be noticed that the proposed approach to disruption mitigation via reduction of the seed runaway electron population immediately after TQ seems prospective. Collecting the seed runaways provided by the fast speed injected projectile from refractory materials (W or C) is capable to reduce the runaway current below of about MA even for 100-fold reduction of the seed current in conditions of a tokamak reactor. A stronger reduction of the seed current seems possible by optimization of the approach proposed. Additional reconnection events initiated by this projectile [9] during penetration through the plasma volume are favorable for suppression of seed runaways.References.
[1] Lehnen M., Maruyama S. 23 March 2018 ITER Technical Report 18-002.
[2] Breizman B.N et. al. 2019 Nucl. Fusion 59 08300.
[3] Sokolov Yu.A. 1979 JETP Lett. 29 244.
[4] Rosenbluth M.N., Putvinski S.V. 1997 Nuclear Fusion 37 1355.
[5] Lehnen M. et. al., 2015 J. Nucl. Mater. 463 39.
[6] Kiselev V.A. et. al., 2012 Instruments and Experimental Techniques 55 347.
[7] Smith H.M., Verwichte E. 2008 Phys. Plasmas 15 072502.
[8] Bobashev S.V. et. al., 2016 Technical Physics Letters 42 309.
[9] Kuteev et. al., 2006 JETP Letters 84 239.
Speaker: Mr Vladimir Sergeev (Peter the Great St. Petersburg Polytechnic University) -
Disruption mitigation remains a unresolved critical issue for ITER. Over 90% of the massive stored energies in ITER must be radiated for a mitigated disruption [1]. The most promising present strategy is to inject a large amount of material into the plasma using a shattered pellet injector (SPI). This study presents an exploration of SPI parameters on mitigation performance and the mitigation efficacy measured using the recently installed SPI on JET.
The JET SPI system incorporates three barrels of different size that can be used in parallel. Scans were conducted varying neon/deuterium mixtures, pellet size and velocity, arrival interval between multiple pellets and target plasma conditions[2]. Two scenarios with an ITER relevant $q_{95}$ were selected; 2.5MA/2.5T and ~1.0MA/1.15T. Both had nominal auxiliary heating of 15MW which was scanned in dedicated experiments. The 2.5MA scenario was selected due to its high total thermal energy and the 1.0MA scenario provided a relatively high thermal energy fraction ($f_{th}$) due to its lower magnetic energy content. The chosen figures of merit for SPI performance were radiated energy fraction ($f_{rad}$) and the current quench duration ($\tau_{CQ}$). Although $f_{rad}$ produces a non-dimensional, multi-machine value, it is subject to large uncertainties that stem from the inference of radiated energy and the calculation of coupled magnetic energy. $\tau_{CQ}$ provides a clearer physical value that is simpler to model but is insensitive to the target plasma or the energy balance.
Large uncertainties in the radiated energy emerge due to strong asymmetries associated with SPI injection and a lack of specialized diagnostics with toroidal resolution. JET’s bolometery cannot yield tomographic inversions in mitigation experiments as its cameras are toroidally separated. Simulations employing phantoms suggest that the radiation location in a given poloidal plane can influence the inferred radiated energy obtained from one bolometer camera by up to 50% [3]. This has led to the development of an SPI specific bolometer analysis algorithm that infers a correction factor based on the location of radiation emission distribution from other diagnostics reducing the estimated uncertainty to 25% for a single toroidal plane. The recently revived KB1 bolometery system, comprising of four toroidally separated, poloidally identical lines of sight, is being incorporated into the new radiated energy analysis technique by providing a toroidal scaling function. Preliminary results indicate higher asymmetry for higher thermal energies that are particularly relevant to ITER as it may significantly alter inferences from previous $f_{rad}$ measurements with high $f_{th}$ [4].
Scans of auxiliary heating power with fixed plasma parameters, shown in Fig. 1, indicate a decreasing $f_{rad}$ with increasing $f_{th}$ as previously observed by the massive gas injection system on JET. Higher toroidal asymmetries may, however, have led to lower radiated powers being measured at the bolometer camera location, and thus a reduced $f_{rad}$ would be measured at high $f_{th}$. With the simpler $\tau_{CQ}$ metric, a shorter $\tau_{CQ}$ was observed with increasing $f_{th}$ implying a more resistive plasma after the thermal quench, most likely due to increased neon content assimilation. This is contrary to the trend observed with $f_{rad}$ and unexpected based on previous studies [5]. To reduce the influence of thermal energy on ablation rate, a range of neon/deuterium mixtures were injected into a fixed plasma configuration. $\tau_{CQ}$ was found to be proportional to the injected neon quantity, irrespective of pellet size and thus the additional deuterium in the pellet. Pellets broken at launch performed as well as intact pellets unless a significant portion of the pellet arrived after the thermal quench and could not be ablated. This complete dataset provides a range of velocities, ablations rate and penetration depths for comparison with modelling.
Scans using multiple barrels simultaneously investigated the superposition of pellets with variations in arrival times. Pellets from the 8mm and 4mm barrels resulted in small $\tau_{CQ}$ changes with variations in arrival time and no overall apparent trend. A timing scan for the 12.5mm and 8mm barrels gave an increase in $\tau_{CQ}$ of ~15-20% for simultaneous arrival and a doubling of density during the current quench. This shows promise for tailoring the neon and deuterium content for parallel RE avoidance and will be further investigated experimentally.
SPI operational space has been explored and optimised using $f_{rad}$ and $\tau_{CQ}$ as metrics. Trends based on $\tau_{CQ}$ indicate improved mitigation at higher $f_{th}$ and for simultaneous arrival of multiple pellets, both positive indicators for the ITER SPI system. The current uncertainties in radiated energy do not allow for quantification of $f_{rad}$, strongly motivating improved diagnostic hardware for future mitigation experiments. Ongoing three-dimensional modelling efforts with synthetic diagnostics will be applied in the analysis of these discharges with a goal of generating a reliable mitigation efficacy metric[6].
References
[1] M. Lehnen, Nuc. fusion 51.12 (2011): 123010
[2] L. Baylor, this conference
[3] J. Lovell, in preparation
[4] E. Joffrin, Proc. 26th IAEA FEA, 2016
[5] P. de Vries, PPCF, 54 (2012) : 124032
[6] E. Nardon, this conferenceSpeaker: Umar Sheikh -
The control of edge localized mode is crucial for protecting the plasma facing components in the magnetic fusion reactor. In the investigation of ELM control, it has been commonly observed that the pedestal turbulence enhances during edge localized mode (ELM) mitigation with supersonic molecular beam injection (SMBI) [1], lower hybrid current drive (LHCD) [2] and impurity seeding [3][4]. The enhancement of the turbulence with LHCD is interpreted by the radial wavenumber spectral shift of turbulence on HL-2A [2]. It seems that a common model of the turbulence enhancement could be used for the ELM control with these external source inputs [5]. However, the reduction of the turbulence intensity observed during ELM suppression with LBO impurity seeding indicates that the turbulence variation during these two phases should result from different mechanism.
Previous result roughly shows the effect of the quantity of the seeded impurities on ELMs, where the quantity is estimated by the diameter of the laser pot of the LBO system [4]. However, this estimation could be inaccurate respecting that the energy of the laser source is not always the same. In further investigation, the maximum value of the radial profile of radiation power density $P_{max}$ is used to indicate the quantity of the seeded impurities. Its time averaged value $\overline{P_{max}}^{t_2}_{t_1}$ is used for statistical research concerning the crucial role of the impurity transport. Here,$t_1$ is the time when the radiation starts to increase and $t_2$ is the time when the mitigation or suppression ends. It has been found that two thresholds of the impurity density exist for ELM control in the normalized statistic result as shown in Figure I. There is no effect on the ELMs when $\overline{P_{max}}^{t_2}_{t_1} \leq P_{mitig}\simeq 20$, whereas ELMs would be mitigated when $ P_{mitig} \leq\overline{P_{max}}^{t_2}_{t_1} \leq P_{Suppr}\simeq 40$, and the ELMs could be suppressed when $\overline{P_{max}}^{t_2}_{t_1} \geq P_{suppr}\simeq 40$.
In addition, contrary effects of the LBO impurity seeding have been found on the pedestal turbulence, which are due to different mechanisms. During ELM mitigation by impurities, it has been found that the turbulence intensity is increased by reducing the pedestal shear rate, which is similar to the results of the ELM mitigation experiments with LHCD [2]. It has been shown that the modification of the pedestal shear rate by LHCD is due to its ion diamagnetic term $\nabla(\nabla P_i/n_iq_i)$. However, it has been demonstrated that the reduction of pedestal $v_{E \times B}$ shear rate by impurity seeding is attributed to its toroidal velocity term $\nabla(v_{\Phi}B_{\theta})$ in which the ion toroidal velocity $v_{\Phi}$ is increased. Unexpectedly, the ions are accelerated in the toroidal direction, although the flow damping should be enhanced due to the increasing of the electron-ion collision rate by impurity seeding. Good agreement has been found between the experimental observation and simulation result on the turbulence behavior as shown in Figure II. During ELM suppression, the pedestal turbulence is significantly reduced. This reduction of the turbulence could result from the decreasing of the linear growth rate of the ITG mode as predicted by the theoretical simulation [6]. The dual effects of impurity on the pedestal turbulence dominate separately during the ELM mitigation phase and ELM suppression phase. Hitherto, results from HL-2A show a potential approach to control the ELMs by regulation of the pedestal turbulence when appropriately controlling the spatial distribution and quantity of the impurity at the pedestal. Furthermore, the understanding of the unexpected acceleration of the ions in toroidal direction when the electron-ion collision rate increases is a crucial issue, which would be the priority of our research on the next step.References
[1] W.W. Xiao et al., Nucl. Fusion 52, 114027 (2012)
[2] G.L. Xiao et al., Nucl. Fusion 59, 126033(2019)
[3] Y.P. Zhang et al., Nucl. Fusion 58, 046018 (2018)
[4] W.L. Zhong et al., Nucl. Fusion 59, 076033 (2019)
[5] G.L. Xiao et al., Phys. Plasmas 26, 072303 (2019)
[6] J.Q. Dong et al., Phys. Plasmas 2, 3412(1995)Speaker: Guoliang Xiao (Southwestern Institute of Physics，China) -
Due to concerns of erosion and damage to divertors caused by Type-I edge-localized modes (ELMs) in ITER tokamak$[1]$, resonant magnetic perturbation (RMP) has been proposed as an efficient technique for ELM control and has been extensively studied in the past few decades. Complete suppressed of type-I ELMs by RMPs was first demonstrated in DIII-D $[2]$. Since then, suppression or mitigation of ELMs has been achieved in several present day devices. It has been speculated that application of RMP may modify the turbulence behavior of the plasma, though there has so far been no conclusive evidence that enhanced turbulence by RMP is responsible for ELM control. Here, we report the first observation of an edge coherent mode (ECM) in the HL-2A H-mode plasmas with ELM mitigation by RMP, with a frequency range of 2–25 kHz. It was found that the mode is located at the edge pedestal region with steep pressure gradient, being excited by the three-wave interaction of turbulence enhanced by the application of RMP. The ECM drives a significant outflow of particles and heat flux, thus providing a channel for continuous extra particle and heat transport across the pedestal during the ELM mitigation.
Mitigation of Type-I ELMs, utilizing the n = 1 RMP field, is observed in HL-2A as shown in Fig.1.The D$_{\alpha}$ signal shows a reduction in amplitude and an about two-third increase in frequency of the ELM bursts [Fig. 1(a)]. A strong reduction in the amplitude of the peak heat flux deposited onto the divertor targets, due to ELM bursts, was also observed [Fig. 1(b)]. Figure 1(c) shows reduced heat flux and a broader flux distribution during the ELM mitigation phase. The MARS-F toroidal modeling for this type of HL-2A discharges has suggested a strong edge-peeling response as the main reason leading to the ELM mitigation $[3]$. A key finding during this experiment is the occurrence of small 2kHz coherent oscillations (ECM), appearing between the reduced Type-I ELM events or riding on them as shown in Fig.2.
The impact of RMP on the plasma turbulence is examined by a new BES system covering the density fluctuation in the pedestal region. A pronounced increase in the density fluctuation level is found in the power spectrum after RMP is switched on. To gain a deeper insight into the RMP impact on the plasma turbulence and the ECM, we have investigated the possible nonlinear interaction or energy cascading among different turbulent components. We found that the nonlinear coupling between different frequency components was substantially enhanced during RMP as shown in Fig.3. The bi-spectrum exhibits strong coupling during RMP as compared with that before RMP, suggesting that the ECM is excited by the three-wave interaction of turbulences enhanced by RMP.
Furthermore, clear evidence of particle exhaust by the ECM has been obtained, via a close examination of the measurement of particle flux in the plasma edge region as shown in Fig.4.The envelope of the saturate particle flow exhibits an oscillation at about 2 kHz, suggesting that the ECM contributes to the outward particle transport. This significant correlation indicates that the ECM can indeed induce a continuous outward particle flux, leading to extra particle transport during the application of the n=1 RMP in HL-2A.
This work was supported by NSFC under Contracts No. 11875018.References
$[1]$ A. Loarte et al, Nucl. Fusion 54, 033007 (2014)
$[2]$ Evans T.E. et al, Phys. Rev. Lett. 92, 235003 (2004)
$[3]$ Liu et al, Phys. Plasmas 24, 056111 (2017)Speaker: yi liu -
Pedestal formed in the plasma edge of high confinement mode (H-mode) strongly affects edge localized modes (ELMs) burst. Thus, understanding the physics of the pedestal instabilities is key to reducing the uncertainties associated with the realization of burning plasma conditions and the appropriate control of ELMs. Fusion power depends strongly on the pedestal pressure. RF waves, such as electron cyclotron wave (ECW) and lower hybrid wave (LHW), have obvious effects on the increase of plasma pressure. It was observed that the change of ELM frequency by ECRH is owing to its role in modifying the edge pressure profile rather than driving the edge plasma current $[1]$. However, there is no clear evidence to explain the decrease of the pressure gradient and the increasing population of the higher ELM frequency band $[2]$. In this work, comparative studies of the impact of ECRH and LHW on pedestal instabilities in the type-I ELMy H-mode discharges are carried out in HL-2A.
For the H-mode plasmas with ECRH, the ELM behaviors change accordingly. Fig. 1. illustrates the typical characteristics of type-I ELM, where the frequency increases with the heating power. In this shot, the power of ECRH is deposited around r = 25 cm. The ECW was perpendicularly injected into the plasma from low field side (LFS). Plasma radiation power increases lightly in the front part of ECW injection. The particle loss and energy loss keep almost constant for the periods with and without ECRH. To study the impact of ECRH on ELMs, the power deposition location was scanned by changing the toroidal filed (B$_{t}$). The ECRH heating powers are almost identical (0.5~0.6 MW) and the ELM periods are about 6 ms without ECRH. However, it decreases from 4.5 ms to 2.7 ms with ECRH power deposition location moving from plasma core region (r/a = 0.35) to edge region (r/a = 0.8). The result suggests that edge ECRH makes it easier to ELM burst earlier.
For the H-mode plasmas with LHW, a pedestal instability quasi-coherent mode (QCM) has been observed after the L-H transition and before the first ELM burst. It was located in pedestal region of the type-I ELMy H-mode. The characteristic frequency of QCM gradually decreases from 50 kHz to nearly 20 kHz before the first ELM, as shown in Fig. 2. At 850 ms when the LHW heating was switched on, the core MHD in Fig. 2 (c) disappeared and the QCM was excited at the same time. Signals of divertor probes illustrate that the particle transport was gradually enhanced during the occurrence of the QCM. The radial wave number of QCM is k$_r$ ~ 0.8 cm$^{-1}$ and it is radially propagating outward. Besides, the poloidal wave number is k$_{\theta}$ ~ 1.4 cm$^{-1}$ and propagates in electron diamagnetic direction. The occurrence and evolution of QCM is strongly associated with the pedestal dynamics and ELM activities, which could be externally controlled by LHW and ECRH.
In order to study the physical mechanism of QCM occurrence and ELM behavior changes, edge plasma parameters are measured, in which the toroidal rotation plays a key role. For QCM, it has been observed that the characteristic frequency is linear proportional to the toroidal angular rotation frequency. The rotation can be affected by LHW and ECRH. For instance, ECRH decreases the plasma rotation not only in the central region but also in the edge region as shown in Fig.3, in which two time slices before and after ECRH switch-on are compared. The theoretical work predicted that the sheared toroidal rotation can reduce the growth rate of the high n modes $[3]$. The results in HL-2A indicate that the reduction of co-current toroidal rotation increases the ELM frequency. As the shear of the plasma rotation has a stabilizing effect on MHD modes, the reduced V$_t$ and softened rotation shear might have destabilized effect on ELMs. More details on the relations among QCM, ELMs and pedestal structure will be illustrated, and its simulation results will also be presented.References
$[1]$ Horton L.D. et al 2004 Plasma Phys. Control. Fusion 46 B511
$[2]$ Burckhart A. et al 2016 Nucl. Fusion 56 056011
$[3]$ Connor J.W. et al 2004 Plasma Phys. Control. Fusion 46 B1Speaker: Dr Jie Wen (Southwestern Institute of Physics) -
Runaway electrons (REs) are a crucial issue for future large tokamaks, especially during disruptions, due to the local impact of RE beam and large thermal loads they can place on the plasma facing components [1]. Therefore, a very active field of research has been opened up in the past decades on RE dynamics during disruptions [2]. Utilizing the newly developed key systems in the HL-2A tokamak for the study of RE dynamics during disruptions, such as: hard X-ray (HXR) camera, laser blow-off (LBO) system, and massive gas injection (MGI) system, the effects of lower hybrid current drive (LHCD) and laser blow-off (LBO) on RE dynamics during disruptions have been systematically investigated. RE generation during disruptions has been successfully avoided for the first time by the LBO-seeded impurity. However, the enhancement of RE generation during disruption with LHCD has been found. With the aid of a hard X-ray (HXR) camera, the physical mechanism of the RE dynamics during disruptions has been observed and these allow a detailed analysis of the generation and evolution of the REs. In this paper, the effects of LHCD and LBO on RE dynamics during disruptions in HL-2A will be demonstrated and discussed.
The experiments presented in this paper were carried out in the HL-2A tokamak with the following parameters: plasma current $I_P= (120-200)~kA$, toroidal magnetic field $B_t = (1.2-1.8) ~T$ and electron density $n_e = (0.5-2)\times 10^{19} ~m^{-3}$. A typical RE generation enhancement during disruption induced by LHCD is shown in Fig. 1, where the temporal evolution of the main parameters is plotted for shot 27908. LH power with 0.18 MW was launched into the Ohmic plasma during the current flat-top phase at 600 ms. A significant rise in the HXR signal with the LHW injection, as shown in Fig.1 (b), which indicates that the LHWs were absorbed by the plasma and a large number energetic electrons were produced in the plasma [3]. The plasma was triggered to disrupt by massive gas injection (MGI) at 700 ms. A runaway current plateau was formed during the current quench phase. Compared with the plasma disruptions without LHCD, the disruptions with LHCD generate very easily a runaway current plateau. To understand the physical mechanism, the HXR camera is used to determine the evolution of energetic electrons. The HXR measurement results show that energetic electrons are created by LHWs and their energy reaches 200 keV during LHCD, then the electrons are converted directly into runaway electrons during disruptions due to the rapid dropping in the plasma temperature and rapid increasing in the toroidal electric field.
Using the newly developed LBO system, RE generation has been successfully avoided during disruptions. Fig.2 is the time evolution of the main parameters of a typical RE generation avoidance during disruption. Before disruption, Iron impurities were injected into the plasma by LBO at 980 ms. With the impurity injection, strong magnetic fluctuation is produced in the plasma, as shown in Fig.2(d). At 990 ms, plasma disruption was triggered by MGI. It can be observed that no runaway current plateau was formed during disruption. Fig.3 is a comparison of the images of RE beams, which were obtain by an infrared camera, during disruptions with LHCD and LBO discharges. It can be clearly seen that there is no RE beam during disruption with LBO. The measurement results from the HXR camera show that almost all energetic electrons are lost under strong magnetic fluctuation before disruption. That is, the “seed” electrons that form the RE beam are “killed” [4], which prevents or suppresses the generation of REs during disruption. Moreover, the dependence of the RE generation avoidance on the LBO parameters has been found, indicating the impurity quantity and injection time is likely the key parameter for RE avoidance during disruptions.
References
[1] Breizman B. et al. 2019 Nucl. Fusion 59 083001
[2] Martin-Solis J. R. et al. 2004 Nucl. Fusion 44 974
[3] Y.P.Zhang, D.Mazon, X.L.Zou, et al., 2019 AIP Advances 9 085019
[4] Lehnen M. et al. 2008 Phys. Rev. Lett. 100 255003Speaker: Dr Yipo Zhang (Southwestern Institute of Physics) -
N. Zhang$^1$, Z. C. Yang$^1$, Y. Liu$^1$, Y. Q. Liu$^2$, T. F. Sun$^1$,X. Q. Ji$^1$, P. Piovesan$^3$, V. Igochine$^4$, D. L. Yu$^1$, S. Wang$^1$, G. Q. Dong$^1$, R. Ke$^1$ , J. M. Gao$^1$, W. Deng$^1$, N. Wu$^1$, Q. W. Yang$^1$, M. Xu$^1$ and X. R. Duan$^1$, The HL-2A Team, The ASDEX Upgrade Team and The EUROfusion MST1 Team
$^1$ Southwestern Institute of Physics, P. O. Box 432, Chengdu 610041, China
$^2$ General Atomics, PO Box 85608, San Diego, CA 92186-5608, USA
$^3$ Consorzio RFX, Corso Stati Uniti 4, I-35127 Padova, Italy
$^4$ Max-Planck-Institut fur Plasmaphysik, EURATOM Association, Garching, GermanyControlling large edge localized modes (ELMs) is critical for tokamaks operating in H-mode, due to potentially severe consequences on material damages caused by ELM bursts in future large scale devices such as ITER $[1]$. Resonant magnetic perturbation (RMP) has been extensively applied to mitigate or suppress ELMs $[2]$. In this work, we report two new recent results on the effect of the $n=1$ ($n$ is the toroidal mode number) RMP fields on ELMs and the associated plasma transport. One is the experimental result on the HL-2A tokamak, where large type-I ELMs were for the first time on this device suppressed by the applied $n=1$ RMP. The other is the toroidal modeling study on the plasma core flow damping by the applied $n=1$ RMP, with computational results quantitatively agreeing with experiments in ASDEX Upgrade $[3]$.
In HL-2A experiments, we find that an increase in turbulence around the pedestal region plays an essential role in the ELM suppression dynamics. The plasma flow damping due to RMP may be a direct cause of the observed increase of turbulence. Figure 1 shows a typical ELM suppression discharge with $4.9 kAt$ current in the coils. The pulse has the toroidal magnetic field of $B_T=1.34$ T and the plasma current of $I_p=120 kA$.The plasma edge safety factor is $q_{95}=3.8$. Within one ELM cycle after switching on the RMP, the large type-I ELM disappears. After the perturbation field is turned off later on in the discharge, large type-I ELMs return.
We mention that suppression of type-I ELMs in HL-2A is globally confirmed across all diagnostics utilized to monitor the ELM properties, including the particle flux to the divertor target plate and heat flux measured by an infrared camera.To gain deeper insight into the impact of RMP on the plasma edge behavior during ELM suppression, we have investigated the changes in the pedestal height, the plasma rotation, and the turbulence in the pedestal region. The key finding is an observed increase of the turbulence level [Fig.1(e)] at the pedestal foot, which in turn appears to effect a mild decrease in the pedestal height. Further MHD calculations show that the peeling-ballooning modes become stable during the RMP-on phase. Another, and perhaps the most interesting experimental observation, is the significant decrease in the edge poloidal flow measured by a Doppler reflectometry, after application of the RMP. Reduction of edge poloidal flow may be in direct correlation with the enhancement of turbulence in the pedestal foot. We also find that the impulsive transport associated with ELMs is replaced by an increase in the turbulent transport. More detailed study on the flow damping dynamics due to RMP and its role in enhancing the turbulence at the pedestal foot is on-going.
Flow damping has also been observed in the ASDEX Upgrade ELM control experiments. In one set of experiments, it was found that the core plasma toroidal rotation was significantly reduced by the RMP field induced $m/n = 1/1$ plasma response. We carry out both linear and quasi-linear plasma response modelling, assuming the $n=1$ RMP field as in experiments. The computational tools that we utilize are the MARS-F $[4]$ and MARS-Q $[5]$ codes. For the linear plasma response, several aspects are primarily investigated: (i) comparison of the poloidal spectra between the vacuum field and the plasma response fields, (ii) the plasma boundary corrugation due to 3D fields, (iii) comparison of toroidal torques, including the neoclassical toroidal viscous (NTV) torque $[6]$ and torques associated with the Maxwell and Reynolds stresses. For the quasi-linear plasma response, we investigate the (physically) non-linear interaction between the core plasma toroidal flow and the plasma response to the RMP field. The toroidal torques usually act as the sink term in the momentum balance equation, leading to flow damping $[7, 8]$.
The simulation results are summarized in Figure 2. Linear response computations show a large internal kink response [Fig. 2(a)] when the plasma central safety factor $q_0$ is just above 1. This internal kink response induces neoclassical toroidal viscous (NTV) torque [Fig. 2(b)] in the plasma core, which is significantly enhanced by the precessional drift resonance of thermal particles in the super-banana regime. Quasi-linear simulation results reveal a core plasma flow damping by about $25\%$, agreeing well with experimental observations, with the NTV torque playing the dominant role [Fig. 2(c) and (d)]. Sensitivity studies indicate that the internal kink response and the resulting core flow damping critically depend on the plasma equilibrium pressure, the initial flow speed, the coil phasing and the proximity of $q_0$ to 1. No appreciable flow damping is found for a low $\beta_N$ plasma. A relatively slower initial toroidal flow results in a stronger core flow damping, due to the enhanced NTV torque. Weaker flow damping is achieved as $q_0$ is assumed to be farther away from 1. Finally, a systematic coil phasing scan finds the strongest (weakest) flow damping occurring at the coil phasing of approximately 20 (200) degrees, again quantitatively agreeing with experiments. This study points to the important role played by the internal kink response in plasma core flow damping in high-beta hybrid scenario plasmas such as that foreseen for ITER.
References
$[1]$ A. Loarte et al. Nucl. Fusion 54 033007 (2014)
$[2]$ Evans T.E. et al 2004 Phys. Rev. Lett. 92 235003
$[3]$ Piovesan P. et al 2017 Plasma Phys. Control. Fusion 59 014027
$[4]$ Liu Y.Q. et al 2000 Phys. Plasmas 7 3681
$[5]$ Liu Y.Q. et al 2013 Phys. Plasmas 20 042503
$[6]$ Sun Y. et al 2010 Plasma Phys. Control. Fusion 52 105007
$[7]$ Zhang N. et al 2018 Phys. Plasmas 25 092502
$[8]$ Zhang N. et al 2017 Phys. Plasmas 24 082507Speaker: Neng Zhang (Southwestern Institute of Physics) -
Pellet injection is used in tokamaks and stellarators for fueling, ELM pacing and disruption mitigation. Injection of shattered pellets is a critical part of the envisaged ITER disruption mitigation system. Rapid deposition of a large amount of material is expected to result in a quick cooling of the entire plasma. However, it has recently been demonstrated that a considerable transfer of thermal energy from the electrons of the background plasma to the ions accompanies a localized material injection [1]. This is the result of the ambipolar expansion along the magnetic field line of the cold and dense plasma cloud left behind the ablated pellet. If the cloud is heated at a constant rate, the ions accelerated by the ambipolar electric field acquire half the total energy transferred to the cloud if radiation losses are negligible. If the heating source is depleted and the heating rate drops as the cloud expands, the majority of the energy is transferred to the ions.
In the present work, we investigate the role of the ambipolar energy transfer mechanism in the global plasma energy balance, in particular its role in disruption mitigation cases. In conventional mitigation scenarios, it is assumed that impurity radiation and radial heat transport in stochastic magnetic fields are the two mechanisms responsible for the dissipation of the pre-quench electron thermal energy. We show that ambipolar energy transfer is a competitive mechanism and may even dominate the electron energy balance. As a result, the thermal quench timescale is shortened compared to a uniform injection of impurities, since a part of the pre-quench electron thermal energy is transferred to the plasma ions rather than being dissipated to the wall on the thermal quench electron timescale.
The total radiation losses from a partially ionized gas are the combination of radiation in spectral lines due to electron impact excitation and dielectronic recombination, and radiation in the continuous spectrum due to radiative recombination and bremsstrahlung. Line radiation plays the dominant role in radiative losses of optically thin plasma. However, the mean free path of a line photon is typically much shorter than that of the continuum radiation. Therefore, for relatively high densities the line radiation losses can be significantly reduced due to re-absorption [2]. Continuum radiation will be the dominant energy loss mechanism is such a case.
The initially very dense pellet cloud is not transparent to line radiation and radiates predominantly in the continuous spectrum. The cloud transitions to volumetric line radiation as it expands and becomes optically thin for the lines.
Figure 1 shows the breakout of the electron, ion and the total coupled energy for an expanding dense Argon cloud heated by ambient electrons that cool down during the cloud expansion in accordance with total energy conservation. The plasma and the pellet parameters are relevant to tokamak disruption mitigation scenarios, and heat conduction along stochastic field lines is assumed to be negligible. The details of the pellet ablation process are not considered as we assume fast localized material deposition on a field line as the initial condition for the subsequent ambipolar expansion. These results are obtained by solving hydrodynamic equations for the expansion of the dense plasma cloud (see Ref. [1]) taking into account continuum radiation losses according to the atomic radiation data from Ref. [3]. The cloud drifts due to magnetic field inhomogeneity are also neglected.
For the case shown in Fig. 1 approximately 40% of the total pre-quench electron energy stored on a field line is transferred to the pellet ions and the rest is eventually radiated. The effective size (length) of the cloud along the field line is also shown in Fig. 1.
The relative amount of radiated energy and energy deposited in the ions is found to be sensitive to the initial cloud density and the heating rate. For instance, the ions get 60% of the total energy if the initial cloud density is half as high as in Fig. 1 (assuming the same per-particle heating rate).
The ambipolar electron-ion energy transfer mechanism should be taken into account in the design of the ITER disruption mitigation system [4]. The pre-quench electron energy transferred to ions will eventually be dissipated but on a longer timescale. At the same time, the thermal quench timescale is shortened by this mechanism, and the radiated energy is toroidally localized.
References
[1] Pavel Aleynikov, Boris N. Breizman, Per Helander and Yuriy Turkin, J. Plasma Phys. 85, 905850105 (2019)
[2] Ya. B. Zel’dovich, Yu. P. Raizer, Physics of Shock Waves and High-Temperature Hydrodynamic Phenomena (Academic Press, New York and London, 1966)
[3] H. P. Summers, and M. G. O’Mullane., Atomic data and modelling for fusion: the adas project., AIP Conference Proceedings. 1344, 1 (2011)
[4] M. Lehnen et al, R&D for reliable disruption mitigation in ITER Preprint: 2018 IAEA Fusion Energy Conf. (Gandhinagar, India, 22–27 October 2018) EX/P7-12
Speaker: Pavel Aleynikov (Max-Planck Institut für Plasmaphysik, Greifswald, Germany) -
In magnetically controlled fusion devices, improving plasma performance is crucial for enhancing the confinement efficiency. An improved regime, high confinement mode (H-mode), has been chosen as the standard operating scenario for the international thermonuclear experimental reactor (ITER). Recently, the energy confinement improvement by externally seeded low or medium Z impurities has been observed both in low confinement mode (L-mode) and H-mode plasmas. In metallic wall tokamaks, the energy confinement is lower than that in carbon wall case with identical configuration and input power. Fortunately, the confinement loss with respect to the carbon devices can be recovered by seeding impurity gas, such as nitrogen seeding in JET and ASDEX Upgrade [1, 2]. More recently, in the carbon wall tokamak of the HL-2A, the core energy confinement of the H-mode plasma is improved by edge-deposited neon or argon impurities seeded by the supersonic molecular bean injection (SMBI), which is a useful fuelling method with higher fuelling efficiency than that of gas puffing. During the edge-core coupling process after the impurity seeding, distinct responses of electron and ion thermal transport to impurity seeding is found for the first time. The study suggests that the impurity ions have diverse effect on the different scale turbulence and the resulting thermal transport. The distinct responses are accompanied with the increase of ion temperature. Further improvement of core ion temperature in the H-mode plasmas is beneficial for fusion reaction in the future devices.
In the HL-2A H-mode plasmas, the impact of spontaneously accumulated and externally seeded impurities on plasma turbulence and edge-localized modes (ELMs) has been intensively investigated [3-5]. The response of the resulting transport to impurity ions is expected to affect the plasma energy confinement. As Fig.1 shows, the confinement is improved by neon impurity seeding in the ELMy H-mode. The neon SMBI pulse length is 1ms. The plasma density moderately increases to $0.7$ times of the Greenwald density limit after the SMBI. The impurity increases the plasma radiation power. For the features of the ELMs, the amplitude is almost identical as evaluated from the divertor $D_\alpha$ signal intensity. However, the ELM frequency decreases to $0.3-0.5$ times lower than that before the SMBI. BOUT++ simulation predicted that the shift of the pedestal profile could change the peeling-ballooning stability. The prolonger inter-ELM periods allow the plasma to build a higher pedestal density. Regarding the ion temperature, it increases both at the plasma edge and core after the impurity seeding as shown in Fig.2(a). Similarly, this observation is also found in the HL-2A experiment with argon impurity seeding by SMBI. Figure 2(b) shows the time traces of electron and ion temperature increment ratios ($r_e$ and $r_i$) before and after the neon seeding by SMBI. It shows that the electron temperature increment ratio is almost unchanged at the edge and core plasma region after the impurity seeding. On the contrary, the increment ratio of ion increases to $0.2-0.4$ after the SMBI. The results indicate that the ion and electron heat flux exhibits distinct responses to the impurity seeding. It suggests that the electron and ion thermal transports are decoupled by the impurity seeding. The decoupled ion thermal transport contributes to an improved energy confinement. Theoretical studies predicted that the effect of impurity on the electron and ion scale turbulence is quite different. Impurity ions can induce the scale separation between electron and ion thermal transport. The seeded impurity could change the core heat flux via the decoupling transport, resulting in a higher ion temperature profile and an enhanced energy confinement in the H-mode plasmas.
Reference
[1] M.N.A. Beurskens et al 2013 Plasma Phys. Control. Fusion 55 24043
[2] G.P. Maddison, et al 2014 Nucl. Fusion 54 073016
[3] W.L. Zhong et al 2016 Phys. Rev. Lett. 117 045001
[4] Y.P. Zhang et al 2018 Nucl. Fusion 58 046018
[5] W.L. Zhong et al 2019 Nucl. Fusion 59 076033Speaker: Dr Wulyu Zhong (Southwestern Institute of Physics) -
INTRODUCTION
Understanding the interaction of Alfvén Eigenmodes (AEs) and energetic particles, namely fusion alphas, is of utmost importance to the operation and performance of future tokamaks, such as ITER and SPARC. In the JET tokamak, a toroidal array of eight in-vessel antennas was installed 1 to actively excite stable AEs with frequencies ranging 25 – 250 kHz, typical of BAEs, RSAEs, and TAEs in JET. Independent phasing of each antenna allows AEs with low to intermediate toroidal mode numbers, |n| < 10, to be probed; this is a significant improvement over previous studies conducted with saddle coils 2 limited to |n| < 2 (and frequencies > 60 kHz). During the 2019 JET deuterium campaign, over 5000 resonances were actively excited – and their frequencies ω_0, damping rates γ, and toroidal mode numbers measured – in over 500 JET plasmas. A large database of these parameters has been populated, from which new physics can be extracted and compared with MHD and gyrokinetic simulations. Similar measurements are expected in the upcoming JET hydrogen and tritium campaigns, and these data will help prepare for operation in the DT campaign.
DAMPING RATE MEASUREMENTS FOR LARGE FAST ION POPULATIONS
Recent JET experiments have seen record NBI heating powers over 32 MW, and stable AEs have been resonantly excited at total heating powers (NBI + ICRH) up to 35 MW. Damping rate measurements γ/ω_0 at such high power allow the investigation of AE driving and damping mechanisms in the presence of large fast ion populations. Latest observations are shown in Fig. 1 as probability density functions, constructed assuming each measurement is Gaussian with mean equal to γ/ω_0 and standard deviation equal to the associated uncertainty. While the effects of fast ion drive and Landau damping must be untangled from the dataset, lower damping rates are observed more often during NBI and ICRH. Though not shown here, it is found that γ/ω_0 decreases with increasing plasma current and density and depends on the magnetic configuration (e.g. limiter vs x-point, elongation, triangularity, etc.).
INTERMEDIATE TOROIDAL MODE NUMBERS AND THE FAST ION PRESSURE GRADIENT
This work extends the previous JET AE database [3] to intermediate toroidal mode numbers, |n| ~ 5 – 7, which are those predicted to interact most strongly with fusion alphas in JET [4]. Measurements of unstable AEs indicate the existence of modes with |n| ≤ 7, although higher |n| ~ 20 can be resolved by the magnetics. Comparisons of toroidal mode numbers with opposite signs can assess the fast ion drive since its destabilization term is proportional to n while no other damping effects depend on the sign
of n [5]. In this dataset, there are over 25 observed modes with 5 ≤|n|≤ 7 in plasmas with total heating power (NBI + ICRH) ranging 5 – 30 MW.COMPUTATIONAL STUDIES OF TOROIDAL AND BETA-INDUCED ALFVÉN EIGENMODES
Computational studies of AE stability have been performed using both MHD and gyrokinetic codes. Recent work [6] reported the damping of TAEs during high NBI power in JET, and the MHD code MISHKA [7] confirmed the core-localization of TAEs during ICRH heating. A complementary study [8] simulated both stable and destabilized TAEs in JET with the gyrokinetic code GTC [9], quantifying the dominance of ion Landau and radiative damping over continuum and electron Landau damping. Comparisons between experimentally-measured and GTC-predicted damping rates are in good agreement, as seen in Fig. 2. Additional GTC simulations are being performed to quantify (i) the drive from the ICRH minority ion population and (ii) the damping due to NBI, for TAEs with n = 4, 5, and 6. Moreover, GTC is being used for the first time to characterize experimentally-observed low n, low frequency modes – postulated to be BAEs – and determine their radial mode structure. Preliminary results indicate that modes with n = 2, 3, and 4 can be driven by an experimentally-relevant fast ion population.
![Damping rates measured by the Alfvén Eigenmode Active Diagnostic (AEAD) compared to GTC predictions for three times in two JET pulses. Adapted from Fig. 8 of Ref. [8].](https://www1.psfc.mit.edu/iaea-fec20/porkolabm/fig2_Porkolab_IAEA2020_reduced.jpg)
ACKNOWLEDGEMENTS
This work was supported by US DOE through DE-FG02-99ER54563, DE-AC05-00OR22725, and DE-AC02-05CH11231. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training program 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
REFERENCES
1 P Puglia et al 2016 Nucl. Fusion 56 112020
2 A Fasoli et al 1996 Phys. Rev. Lett. 76 7
[3] D Testa et al 2011 Nucl. Fusion 51 043009
[4] WW Heidbrink 2002 Phys. Plasmas 9 2113
[5] A Fasoli et al 2002 Plasma Phys. Control. Fusion 44 B159
[6] RJ Dumont et al 2018 Nucl. Fusion 58 082005
[7] GTA Huysmans et al 1998 Nucl. Fusion 38 179
[8] V Aslanyan et al 2019 Nucl. Fusion 59 026008
[9] Z Lin et al 1998 Science 281 1835Speaker: Dr Roy Tinguely (MIT) -
The nonlinear interactions between energetic particle (EP) and Alfvén waves are very important for astrophysics and high temperature plasma physics, especially for magnetically confined fusion plasma, because they will affect the redistribution and transport of EPs significantly. When the EPs have sufficiently strong pressure gradient, they can excite a non-normal mode, named energetic particle mode (EPM) [1-4], which emerges as discrete fluctuation out of the shear Alfvén continuum at the frequency for maximal wave-EP power exchange above the threshold condition due to continuum damping. Therefore, the critical level of tolerable EP losses in a fusion device can be more severe in the presence of EPM.
The early nonlinear theories [5-7] have expected the role of the EPs radial displacements on the time evolution of a strongly driven mode as the EPM. If the EPs move outward, they can locally alter the EP gradient and destabilize a new wave that transports them further, where a new wave is destabilized, much like different runners do in a relay race. The model for the nonlinear evolution is named relay runner model (RRM) [6]. The expected self-consistent nonlinear evolution of the EPM and the EPs have not been unambiguously observed in experiments up to now, because the evolution of the modes are very fast. In this paper, we will give a new experiment evidence about the EPM avalanche in the HL-2A tokamak.
The strong EPM instabilities, with frequency chirping-down rapidly and in the range of 35-70 kHz, driven by 42 keV beam ions are observed, immediately, after the turning on of tangential injection NBI. The typical chirping-down process of EPM is shown in the spectrogram of Mirnov signal in Fig.1 (a) and (b). It is obvious that the amplitude of the mode ($A$) increases to the maximum rapidly during t=618.62-618.68 ms, while the mode frequency ($f$) drops rapidly from 54 to 45 kHz. Then, A decreases slowly, and the f decreases to 37 kHz during t=618.70-619.50 ms. The peak fluctuation amplitude $\delta B/B$ is about $(4-5)\times10^{-4}$. The ratio of typical growth time of EPM to Alfvén time is $\tau/\tau_{A}\sim 200$, which is in the expected range where convective EP redistributions can take place [7-9].
Evolution of mode numbers near the amplitude rising process is confirmed by the phase shift of the filtered poloidal and toroidal Mirnov waves, as shown in Fig.2. It is obviously that the mode propagates in ion diamagnetic drift direction poloidally. The poloidal mode number changes rapidly from m=2 to 3, and then becomes 4 step by step, which are labeled by the black, brown and pink lines in Fig.2 (a), respectively. The change of m can be completed within 0.06 ms (between t=618.64-618.70ms), and the corresponding frequency changes from 54 to 45 kHz during that time interval. The toroidal mode number is confirmed always as n=1 during the whole chirping down process, as shown in Fig.2 (b). It is indicated that the EPMs move from the core (q=m/n=2, where q is the safety factor) to the edge (q=3 and 4) of plasma, gradually. The EPM mode structure and their evolution in poloidal cross section are obtained by tomography of SXR arrays. The m/n=2/1 mode mainly locates on the core of plasma at first. Then, the modes propagate to the edge gradually. At last, the edge (m=3 and m=4) elements become dominant.
According to the RRM model, for dominant circulating particles driven EPM at transit resonance, as in the HL-2A experiments, the relationship between the EMP chirping rate $\dot{\omega}$ and $A$ satisfies the scaling $A \propto q\dot{\omega}$. The relationship has been obtained by experiment data, as shown in Fig.1 (c). It can be seen that the amplitude of Mirnov signal increases rapidly during t=618.60-618.68 ms, which is labeled by the red line in Fig.1 (a). The center frequency of the mode is labeled by the black circles in Fig.1 (b). The q value timing change rate of the frequency ($\dot{f}$) is roughly proportional to mode amplitude, e.g., $A \propto q\dot{f}$, as shown in Fig.1 (c).
For a single n coherent fluctuation, one obtains, from the conservation of the extended phase space Hamiltonian, $\dot{E}/\dot{r}=-m_i\Omega_i \omega r/m$. On the other hand, from wave-particle phase-locking for maximized power exchange, the derivative of the EPM frequency with time ($\dot{\omega}$) can be expressed as $\dot{\omega}=\dot{\omega_{tr}}\approx -\Omega_{i}r\dot{r}/(nR_0^2q^3)$. Therefore, the radial velocity of outwards energetic ions can be expressed as $\dot{r}=-nR_0^2q^3\dot{\omega}/(\Omega_ir)$. Experimentally, the average radial velocity of the waves ($V_p$) can be estimated by the ratio of the distance from q=2 to 4 rational surfaces ($\Delta r$) to the time interval ($\Delta t$). Compare to the above result, we obtained $|\dot{r}/V_p|\approx O(1)$, i.e., the observed mode radial propagation velocity is comparable to that of Eps, as predicted by RRM model.This work is supported in part by the National Key R$\&$D Program of Cnina under Grant Nos. 2017YFE0301202 and 2018YFE0304102, and by NNSF of China under Grant Nos. 11875024, 11875021 and 11835010.
Reference:
[1] L. Chen, Phys. Plasmas 1, 1519 (1994).
[2] F. Zonca et al., Plasma Phys. Control. Fusion 48, 537 (2006).
[3] L. Chen and F. Zonca, Review of Mordern Physics 88, 015008 (2016).
[4] W.W. Heidbrink, Phys. Plasma 15, 055501 (2008).
[5] R.B. White et al., Phys. Plasmas 26, 2958 (1983).
[6] F. Zonca and L. Chen, 6th IAEA TCM on Energetic Particles in Magnetic Confinement Systems, Oct. 12-14, 1999, JAERI, Naka, JAPAN.
[7] L. Chen and F. Zonca, Nucl. Fusion 47, S272 (2007).
[8] F. Zonca, et al., Nucl. Fusion 45, 477 (2005).
[9] F. Zonca et al., Plasma Phys. Controlled Fusion 57, 014024 (2015).
[10] L. Chen et al., Nucl. Fusion 28, 389 (1988).Speaker: Dr Liming Yu (Southwestern Institute of Physics) -
Alpha particles are the key players of a burning plasma as they provide the self-heating required for the sustainment of the fusion burn. At the same time, however, there is only little experimental knowledge on their properties, mostly because of the limited availability of deuterium-tritium (DT) plasmas. Among the challenges that the scientific program of the Joint European Torus (JET) is preparing to face is thus the unambiguous observation of alpha particle physics effects. This encompasses the excitation of alpha-driven Toroidal Alfvén Eigenmodes (TAE) [Ref. 1], as well as the documentation of the general impact of the alpha particles on a number of plasma properties such as, for instance, their expected positive impact on transport through the stabilization of turbulence.
This ambitious scientific goal requires the development of a dedicated scenario at JET, where the production of alpha particles is maximised at limited input power. Of special benefit would here be the capability to produce a steady state supra-thermal deuterium and/or tritium distribution function, as this can in principle maximize the DT reactivity, which is not achieved in a solely thermal plasma. Among the tools that can realize this scenario in DT is an application of the novel ‘3 ion’ scheme [Ref. 2], which is based on ion-cyclotron resonance heating (ICRH) in a mixed species plasma and has recently been applied at JET in D-$^3$He. In this contribution we present the findings of this experiment, which led to the generation of a significant amount of alpha particle through the $^3$He(d, p)$\alpha$ reaction. We also demonstrate that these plasmas anticipate many of the peculiarities that are commonly associated to alpha particles, without the technical complications of a DT environment, and are thus worth studying per se.
In the set of D-$^3$He JET experiments we have conducted, deuterium ions from the neutral beam injection (NBI) system were accelerated by ICRH. The concentration of $^3$He ions ~20-25% was chosen to locate the ion-ion hybrid (IIH) layer in the plasma core, where the wave polarization is particularly favourable for ICRH absorption by D-NBI ions through their Doppler shifted fundamental resonance. This leads to the acceleration of D ions up to the MeV range, which is unambiguously demonstrated by a large variety of diagnostics data. Among these are a factor ≈ 10 enhancement of the DD neutron rate, the observation of high energy tails in the neutron spectrometers and neutral particle analyzers, the production of gamma-rays from nuclear reactions driven by the energetic deuterons, and many others. A peculiarity of the scenario is the capability to change the plasma reactivity by modifying the NBI/ICRH heating mix at fixed input powers up to about 15 MW. In all these plasmas, $^3$He acts as an element for ICRH acceleration of D ions and as the target of the $^3$He(d, p)$\alpha$ fusion reaction, and we can produce alphas at the level of 10$^{16}$ particle/s and with a mean energy around 4 MeV, with a spectral width that depends on the average energy of the fast deuterium. Of particular relevance is here the unique capability that JET has to determine the image of the fusion born alpha particle source, which is made possible by a tomographic inversion of the 16.4 MeV gamma-ray emission from d+$^3$He fusion reactions using data from the recently enhanced gamma-ray cameras [Ref. 3] (figure 1).
As MeV range ions are produced, the plasma responds in a peculiar way. Despite a dominant electron heating predominantly constrained in the very core region where the IIH layer occurs, we observe T$_i$≈T$_e$ throughout the plasma and we achieve core temperatures of about 8 keV at moderately high electron densities of ≈ 6‧10$^{19}$ m$^{-3}$. This suggests an important contribution of MeV range fast ions in the mitigation of turbulence in a scenario with dominant electron heating that, in many respects, mocks up some of the heating conditions expected by alpha particles in DT [Ref. 4].
Another common observation is the presence of a large variety of fast ion driven MHD and, in particular, of Reversed Shear Alfvén Eigenmodes (RSAE), that persist also in the main heating phase, suggesting that a non-monotonic q-profile is unexpectedly achieved.
Furthermore, we have developed a “slowing down” scenario, whereby the NBI source is switched off while ICRH persists (figure 2), as a way to study the decay of the energetic deuteron and fusion born alpha populations. In the “after glow” phase of this discharge there is a spatial change of the alpha particle source (figure 1), which is accompanied by long lived elliptic Alfvén eigenmodes (EAE). Numerical simulations of the drive and damping of these modes are being carried out to establish the contribution of the fusion born alphas to the drive of the observed EAEs, with application to the possibility of developing a new scenario for the destabilization of $\alpha$-driven AEs studies in DT in NBI/ICRH plasmas at moderate input power levels.
We finally discuss the implications of these results for JET DT and ITER, in particular with respect to the facets of alpha particle physics that these plasmas anticipate, what can be learnt from them and their readiness level in view of DT.Acknowledgements This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
[Ref. 1] R. Dumont et al. “Scenario preparation for the observation of alpha-driven instabilities and transport of alpha particles in JET DT plasmas”, this conference
[Ref. 2] Y. Kazakov et al., Nature Physics 13, 973–978 (2017)
[Ref. 3] D. Rigamonti et al. Rev. Sci. Instrum. 89, 10I116 (2018)
[Ref. 4] J. Garcia et al. Phys. Plasmas 25, 055902 (2018)
Speaker: Dr Massimo Nocente (Dipartimento di Fisica, Università di Milano-Bicocca) -
Gamma-ray spectrometry of the plasma [1] is one of the tools giving information on the heating efficiency. The source of gamma-ray is nuclear reactions between energetic confined ions and plasma impurities, i.e. Be and bulk plasma ions. Gamma-ray diagnostics allow monitoring the energy distributions of the fusion products, ions accelerated during ICRF heating and plasma fuel ions and provide an effective optimization of high-performance discharge scenarios with additional plasma heating (NBI and ICRF).
In the recent JET experiments at 3.7T/2.5MA, D-NBI ions were accelerated to MeV energies using the 3-ion ICRH scheme D–(D NBI)–3He [2] in D3He mixed plasmas. There are three essential components in the plasma: thermal deuterium, 20-25% of thermal 3He and D-NBI ions that effectively adsorb ICRF power at the mode conversion layer in the plasma core.
In the experiments, two large volume LaBr3(Ce) Ø3”x6” spectrometers [3] with vertical and tangential lines-of-sight (LoS) were used. They allow measuring high-resolution and time-resolve spectra up to ~30 MeV during plasma discharge. In some high performance discharges the vertical LaBr3(Ce) detector was replaced with a high-resolution HpGe spectrometer [4] for measurements of the Doppler broadening of gamma-lines in recorded spectra. In addition to these highly efficient spectrometers, the gamma-ray camera, consisting of 19 compact LaBr3(Ce) detectors [5] with 10 horizontal and 9 vertical LoS, was used for obtaining 2D gamma-ray emission profiles measurements. In the recorded during plasma discharges spectra of both vertical and tangential spectrometers, gamma-ray lines corresponding to transitions in the nuclei 10Be and 10B excited in the 9Be(D,pγ)10Be and 9Be(D,nγ)10B nuclear reactions were identified. The line-integrated energy distribution function of the fast D-ions was reconstructed with a specially developed gamma-ray spectrum analysis code DeGaSum [6]. This code allows reconstructing the fast D-ion energy distribution using the measured gamma-ray line intensities together with the known excitation functions of the reactions. To obtain intensities of the gamma-rays generated in the plasma discharge, we used the spectrometer response functions calculated for monoenergetic gamma-rays in the energy range 0.5 - 30 MeV. For these calculations both the vertical and tangential spectrometer LoS models were used. An example of the data processing results is presented in figure 1a, where one can see both measured gamma-ray spectrum (black line) and the restored energy distribution of the gamma-rays emitted from plasma (red line). Figure 1b shows the reconstructed energy distribution of the confined D-ions that was obtained using 3.37-MeV gamma-ray line from 9Be(D,pγ)10Be reaction and 2.86- and 3.59-MeV lines from 9Be(D,nγ)10B reaction. In the Maxwellian approximation, the effective temperature of fast D-ions is ~600 keV in the plasma discharge #94701.
In these experiments, due to a high concentration of 3He in the plasma and strong population of the energetic D-ions, the 3.6-MeV alpha-particles were generated in the fusion reaction 3He(D,p)4He. The source of the fusion-born alpha-particles (rate and spatial profile) could be measured with 3He(D,γ)5Li reaction, which is a weak branch of the main fusion reaction; the 3He(D,γ0)5Li/3He(D,p)4He branch ratio is ~3·10^(-5). This reaction gives rise gamma-rays with energy ~16.7 MeV. It was identified in the recorded spectra. Measuring intensity of the 16.7- MeV gammas, the fusion born alpha-particle rate production was estimated in discharges, i.e. ~4.2·10^13 m-3·s-1 for the #94701 discharge. The spatial distribution of the confined alpha-particles was obtained by measurements of the 4.44-MeV gamma-rays from the 9Be(α,nγ)12C reaction [2]. This line is strong and clearly seen in the recorded spectra (figure 1a). These gamma-ray measurements have provided a comprehensive test of the diagnostics and analysis methods that are required for alpha-particle studies in forthcoming DT-experiments.
In fusion reactors, the source of DT alpha-particles and their behaviour in the plasma should be under control to provide the high fusion performance. The presented paper demonstrates the capability of the gamma-ray spectrometry for such control. The inferred deuterium energy distributions, as well as the assessment of the D3He fusion rate, have allowed optimizing the ICRF heating scenario.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. The work was also partially funded under contract No. 17706413348180002230/19-19/01 of 18/03/2019 between the State Atomic Energy Corporation ROSATOM, Institution “Project Center ITER” and Ioffe Institute.[1] Kiptily, Cecil F.E. and Medley S.S., Plasma Phys. Control. Fusion 48 (2006) R59–R82
[2] Kazakov Ye.O. et al, Nature Physics 13 (2017) 973–978
[3] Nocente M. et al Rev. Sci. Instrum. 81, 10D321 (2010);
[4] Tardocchi M. et al PRL 107 (2011) 205002
[5] Nocente M. et al Review of Scientific Instruments 87, 11E714 (2016);
[6] Shevelev A.E. et al Nucl. Fusion 53 (2013) 123004Speaker: Margarita Iliasova (Ioffe Institute) -
To verify the CarMa0NL modelling for COMPASS-U, the numerical results are cross-validated with general analytical predictions [Pustovitov V. D., Nuclear Fusion 55 113032 (2015)]: the computed vertical force on the tokamak wall is found to be almost zero during fast transients, as it should be. This test proves the credibility of the model and computational method. The role of poloidal eddy current (which is absent in some approaches) in the dynamics of radial force on vacuum chamber is found to be essential. This justifies the choice of CarMa0NL, which additionally is able to describe real 3D geometry of the conducting structures, as a primary tool for disruption studies for COMPASS-U.
Strong electromagnetic loads on tokamak vacuum vessel and in-vessel components ('wall') caused by an abrupt termination of plasma discharge must be modelled thoroughly to guarantee structural integrity of future fusion reactors. In the absence of a predictive theory for plasma behavior during disruptions, the design of tokamak chambers still remains a challenging task. Recently some general rules [1] have been established in theory that allow to test the credibility of modelling. The present study aims to verify these rules numerically for the wall configuration with passive stabilizing plates (PSPs) to complement the similar analysis performed for the wall without PSPs [2].
Many tokamaks, specifically AUG, COMPASS-U, DTT, EAST, JT-60SA, KSTAR and WEST employ PSPs to improve vertical stability of elongated plasmas. Hereinafter, we scrutinize the COMPASS-U, which is a high-field tokamak presently in the final design phase. It will operate at the toroidal magnetic field, plasma current and elongation up to $B=5\ T$, $I=2\ MA$ and $\kappa = 1.8$, respectively. The upper (UP) and lower plate (LP) are placed inside the vacuum vessel (VV) as shown in Fig. 1 (a). The resistive decay time for toroidal current induced in these copper PSPs is $\tau_{psp}=66~ms$, which is 22 times larger than the same value for VV made of Inconel ($\tau_{vv}^{tor}=3~ms$). This complicates electromagnetic and mechanical analysis of the wall response. We study a number of different disruption scenarios with CarMa0NL [3] to understand the dynamics of electromagnetic loads on the wall. Preliminary results presented in Fig. 1 confirm analytical predictions [1] for the COMPASS-U setup with PSPs: the total vertical force on the wall is zero for fast transients.
We analyse the fastest event one can expect on COMPASS-U: a linear 0.1 ms long TQ, as shown in Fig. 1 (k), is followed by a 3.3 MA/ms CQ (f). The safety factor $q_{95}$ first rises because of the current decay, but then, upward motion shrinks plasma cross-section and drops $q_{95}$ down to 1 (n). At this point we trigger a 0.01 ms CQ that terminates plasma. The above 2 MA CQ induces strong vertical forces on the vessel, $F_{vv}^z$, upper, $F_{up}^z$, and lower plate, $F_{lp}^z$, but the total vertical force on the wall (VV+PSPs), $F_\Sigma^z$, remains almost zero (h). The maximum of $\mid F_\Sigma^z \mid$ is reached in $\Delta t \approx \tau_{vv}^{tor}$ after the end of CQ. The same constant $\tau_{vv}^{tor}$ characterizes the grows of $F_{up}^z$ and $F_{lp}^z$ in Fig. 1 (d). While not yet predicted by analytics, also the minimum of radial force on vessel is delayed by $\Delta t \approx\tau_{vv}^{tor}$. We conclude that for fast ($<<\tau_{vv}^{tor}$) transients, $F_\Sigma^z\approx0$ at $t<<\tau_{vv}^{tor}$.
Until now, the largest forces on the COMPASS-U wall have been found for slow vertical displacement events (VDEs), such as shown in Fig. 2. However, to increase the credibility of numerical predictions we cross-validate our results with analytics [1] for the limit case of fast transients.
References:
[1] Pustovitov, V. D., Nucl. Fusion55, 113032 (2015).
[2] Pustovitov, V. D., Rubinacci, G. & Villone, F.Nucl. Fusion57, 126038 (2017).
[3] Villone, F., Barbato, L., Mastrostefano, S. & Ventre, S.Plasma Phys. Control. Fusion55, 095008 (2013).
Speaker: Vadim Yanovskiy (Institute of Plasma Physics of the Czech Academy of Sciences) -
Optimized stellarators promise comparable energy confinement to existing tokamaks. Recent results from the Wendelstein 7-X device are consistent with this possibility. At the same time, stellarators offer considerable advantages, such as disruption-free operation without a Greenwald density limit. On the other hand, transport in optimized stellarators will be dominated by turbulence, unlike earlier stellarators and as in large tokamaks. There is also a large parameter space for configurations which is difficult to explore with experiments. Whole-volume gyrokinetic simulation is a powerful and flexible tool which can be applied to understand existing stellarator experiments and design improved types.
The whole-volume gyrokinetic code XGC has been extended for stellarators, and linear benchmarks for ion temperature gradient-driven modes (ITGs), neoclassical physics, and particle confinement have been published[1-3]. These give confidence in the code implementation, as well as shedding light on timely topics such as the properties of microinstabilities in Wendelstein 7-X. This topic is beginning to be explored experimentally, but observational investigations are incomplete and can be supported and enhanced by numerical modeling. This capability of XGC can be used to guide scenario development and optimization as well as theoretical understanding of linear microinstabilities.
In this work, the previous linear capability for microinstability simulations is first extended to the nonlinear regime. As a benchmarking exercise, the stellarator and tokamak versions are compared in a nonlinear tokamak simulation. The equations solved should be equivalent, but the numerical implementation is substantially different, making this a good test of the new nonlinear stellarator modelling capability. In the figure above, the nonlinear evolution of the electrostatic potential by ITG is plotted comparing the two codes. The agreement is almost exact, except for slight differences in the nonlinear phase which are expected due to the stochastic nature of the nonlinear dynamics.
Previous linear simulations of the quasi-isodynamic stellarator Wendelstein 7-X are then used as a point of comparison for linear simulations in the quasi-axisymmetric design QUASAR. The quasi-axisymmetric design is of particular interest because no experimental implementation exists, meaning that numerical simulations provide the best current data on their performance. The figure below shows that the normalised linear growth rate for ITG is roughly equivalent in the two devices for the same normalised temperature gradient strength. This linear result shows similiarity in the linear stability of the two devices despite considerable differences in the three dimensional structure of the confining magnetic field.
Having established the validity of the XGC nonlinear model for stellarators, and established confidence in linear simulations of the QUASAR configuration with XGC, this analysis is then extended to nonlinear simulation. The figure below shows the ITG mode structure in terms of electrostatic potential perturbation in the nonlinear phase. Radially extended streamers, as the radically localised temperature gradient profile relaxes, are observed, with comparably strong saturated perturbation amplitude over a much larger proportion of the device poloidally than in the linear regime. The turbulent heat flux has also been calculated.
Simulations such as these offer novel insight into the behaviour of turbulence in stellarator devices and the possibility to examine the differences between turbulent behaviour in unbuilt device types. This capability will be used to compare the resistance of different proposed stellarator designs to turbulent transport.
- M. D. J. Cole et al., Phys. Plasmas 26, 032506 (2019)
- T. Moritaka et al., Plasma 2(2), 179-200 (2019)
- M. D. J. Cole et al., Phys. Plasmas 26, 082501 (2019)
Speaker: Michael Cole (PPPL) -
JET has addressed one of the key issues for the baseline scenario in ITER by demonstrating for the first time a high-performance H-mode with high confinement using a neon (Ne)-seeded radiative divertor. Although the ITER Tritium plant is being designed to deal with both nitrogen (N) and Ne, the use of N leads to the formation of tritium-containing ammonia 1. Its accumulation in the ITER divertor cryopumps will require a more time-costly high temperature regeneration and will reduce the plant duty cycle 1. On the other hand, on current devices with all-metal plasma-facing components, N generally provides the best performance with respect to a carbon dominated environment. It would be beneficial for ITER to use Ne for divertor radiation and plasma boundary simulations for ITER show that this should be equally possible with N and Ne 1. Experimental results on the full-tungsten ASDEX Upgrade, and up to now on JET, had always found insufficient compression of Ne to allow simultaneous divertor power load control and sufficiently low pedestal plasma degradation. It had been expected that the larger physical size of JET in comparison to ASDEX Upgrade would improve matters regarding this compression, allowing JET to approach more closely the situation on ITER. Now, new experiments at JET with a higher input power (30MW) have finally yielded a Ne-seeded scenario (Fig. 1) which behaves similarly to the high-performance seeded H-modes previously only achievable with N impurity.
Stationary Type I ELMy H-modes have been obtained with Ip=2.5 MA, $B_T$=2.7 T, $P_{NBI}$=25 MW,$P_{RF}$=5 MW, at high triangularity ($\delta$ =0.4) with an ITER-like divertor configuration with both inner and outer strike points on the vertical divertor target plates 2. These plasmas are a significant step towards the ITER requirements, achieving a confinement factor $H_{98(y,2)}$~0.9, normalised pressure of $\beta_N$~2.0, Greenwald fraction of $\langle n \rangle$/ $n_{GW}$~0.94, an effective charge of $Z_{eff}$~2.0, radiation fraction of $f_{rad}$~0.5, energy confinement time of $\tau_E$=200ms with low power deposition on the W divertor. The ELM size is ~4-7% of the stored energy. Compared with their N-seeded counterpart discharges 2, these Ne-seeded plasmas require 10 MW more input power to obtain the same improved pedestal pressure. In addition, whilst the N-seeded discharges have natural ELMs with $f_{ELM}$ ~ 45 Hz (with ELM size of 4% of the stored energy), in the case of Ne, the ELMs are irregular at approximately 7Hz frequency, with smaller fluctuations in the inter-ELM period (Fig. 1).
At high enough input power and seeding rate (the latter being required for sufficient divertor power load reduction), the key to obtaining significant confinement is the improvement of the pedestal pressure (Fig.2), as was observed with N-seeding 2. The mechanism(s) leading to an increase in pedestal pressure and temperature at JET are still being actively investigated for Ne, N and even C. In the new high-performance Ne-seeded discharges, the radiation distribution remains more centred on the X-point region than for the N-seeded cases 2. However, some of the features found with improved N-seeded plasmas 2 are also observed with respect to the unseeded case: unchanged position of the pedestal density and width, whilst the pedestal temperature width increases by ~40% in poloidal flux; the separatrix density remains constant; the relative distance between the pedestal density and temperature increases from 1.6% to 2.5% of the poloidal flux.Apart from their importance in terms of ITER baseline demonstration with relevant divertor geometry and plasma-facing materials, a key contribution of these seeded plasmas is to provide a well-diagnosed set of target discharges for plasma boundary simulations. Such simulations, performed with the SOLPS code suite, are the basis for the ITER divertor design and fuel cycle 1. If the JET results for N and Ne can be matched, this significantly increases the confidence in the ITER divertor design basis and will confirm that Ne can be deployed as a seed gas in ITER. The experimental programme leading to these high-performance seeding results has been accompanied from planning to execution by an extensive modelling effort 3 using the SOLPS-ITER code. Scans have been made for these JET discharges with both N and Ne, with fluid drifts activated, and including upstream cross-field transport profiles adjusted to reproduce the H-mode pedestal profiles and the measured near-SOL heat flux channel width. Starting with lower input power cases, for which data where initially available, fair agreement has been obtained between code and experiment regarding divertor target profiles and radiation distributions. At higher input power, corresponding to the new discharges reported here, code runs are underway, and results will be reported at the conference.
JET has thus now demonstrated for the first time that Ne seeding is compatible with good confinement and power load control in an all metal environment. This is an important step towards improving confidence in the scientific basis for high power seeded operation in ITER, as well as providing a route to demonstrate an integrated scenario in JET-DT.
Acknowledgements. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission or of the ITER Organization
References: 1 R.A. Pitts et al, Nuclear Materials and Energy 20 (2019) 100696 2 C. Giroud et al., Plasma Phys. Control. Fusion 57 (2015) 035004 3 E. Kaveeva et al, PSI 2020.Speaker: Dr Carine Giroud (UKAEA) -
Achieving high-$\beta_N$ for current and future tokamaks is a challenging and important issue, where \beta_N is the normalized toroidal beta. High-$\beta_N$ is beneficial for the ignition and fusion reaction, as well as the ratio of bootstrap current is proportional to $\beta_N$. Recently, on HL-2A a high-performance region, combining edge and internal transport barriers (double transport barriers, dubbed as DTBs), had been obtained by high-power NBI and LHW heating, and shown in the Fig.1. Usually, the internal transport barrier (ITB) forms just after the NBI injection, and the ITB foot locates near q=1 surface where the magnetic shear is weak and the flow shear is significant to suppress effectively turbulence fluctuations. Subsequently, the center/edge ion temperature and toroidal rotation both decrease/increase, meanwhile, the plasma density increases, as a result that the pedestal creates and L-H transition occurs. The ELM-free H-mode sustains around 40 ms with DTBs. Accompany with the density ascending, the type-I ELM emerges and \beta_N reaches maximum at the moment. In this scenario, $\beta_N$ >3.0 is realized, and corresponding $H_{98}$~1.3, $f_{bs}$~30% and G~0.4. Meanwhile, the high-$\beta_N$ scenario has also been successfully modeled using integrated simulation codes, i.e. OMFIT and METIS. A steady-state high-$\beta_N$ (t~0.8s and $\beta_N$ >2.0) scenario is also obtained by the pure NBI.
In such high-$\beta_N$ plasmas, there are abundant MHD instabilities, including low-frequency MHD oscillation and high-frequency coherent mode in the edge, and neoclassical tearing mode (NTM) and Alfvén modes in the core (shown in the Fig.2), as well as complex MHD dynamics, e.g. nonlinear wave-wave and wave-particle interactions. Some new physics problems need yet to be resolved, namely how sustain and control MHD and transport barrier to achieve higher and steady-state high-$\beta_N$.
References
[1] C. M. Greenfield, et al, Phys. Rev. Lett. 86 (2001) 4544.
[2] K. H. Burrell, et al, Phys. Rev. Lett. 102 (2009) 155003.
[3] E. R. Solano, et al, Phys. Rev. Lett. 104 (2010) 185003.
[4] W. Suttrop, et al, Plasma Phys. Control. Fusion 45 (2003) A151.Speaker: Wei Chen (Southwestern Institute of Physics) -
The report presents the results of the design development of the equatorial diagnostic port #11 of the final project level and the upper diagnostic ports ## 02,07,08 of the preliminary design level. Ports are being developed at the Budker Institute of Nuclear Physics of the Siberian Branch of the Russian Academy of Sciences (BINP SB RAS or shot BINP) in close cooperation with the ITER central team in the framework of port standardization (DP-IPT). The second part of the report is devoted to the preparation of technological processes for the manufacture and assembly of diagnostic ports at the port integrator site.
The diagnostic ports integrated by BINP, both upper and equatorial, are made according to a single unified principle - modular design. In-vessel part of the diagnostic ports consists of the following main components: diagnostic first wall (DFW), diagnostic shielding module (DSM) port plug structure (PPS) (see Figure 1).
These components are manufactured by different manufacturers and subsequently assembled together at the port integrator site at BINP. The diagnostic shielding modules manufacturing is under responsibility of the port integrator, the other two components are developed and supplied by a central team.
The basis of the DSM design is a metal c-shaped frame, which provides the transfer of forces from the DFW to the PPS. This design allows to provide the 1necessary rigidity and structural integrity independent of the internal location of the diagnostics (see figure 2,3). The space around the diagnostics is filled with a shielding trays consisting of a stainless-steel tray and a boron carbide ceramic bricks. Since the shielding trays are standardized and have a discrete step, there are significant gaps between the diagnostics and shielding trays. To eliminate these gaps, a stainless-steel blocks (backfilling) are used that are installed on the DSM and fill this empty space as much as possible (see Figures 2,3).
Such a constructive solution allows to have a very flexible system that allows, on the one hand, the ability to provide the necessary cutouts for lines of sight and diagnostic equipment inside the DSM, on the other hand, to ensure the strength of the DSM design independent of these cuts.
The big advantage of using a standardized approach to the design of a port plugs allows to use the same technological and assembly solutions for all ports, which significantly reduces the cost and time of production preparation for manufacturers with more than one port.
As part of the preparation for the production of diagnostic port components, a series of technological studies (R&D) was carried out (both at the BINP and the ITER central team) on the possibility of using the proposed technologies for the manufacture of components. A technology of a neutron protection from boron carbide has been developed. It was shown, as a result of a set of tests (including gas separation measurement), that boron carbide is a suitable material for ITER and can be placed in a vacuum vessel in large volumes (tons). A typical cooling system has been developed for DSM of the equatorial and upper ports. A prototyping of complicated places was carried out, including the manufacture of a gun-drilling channels, welding of blind holes and long welds on water channels.The results of these technological studies were presented in the form of reports about the study and are stored in the ITER database. In the future, all these technological developments can be used by other ITER members for the technological substantiation of the design and preparation of the production of DSMs.
The next key point in preparing the integration of ITER diagnostic ports is the preparation of an assembly hall on the territory of the BINP. The complexity of creating an assembly hall for diagnostic ports lies in the aggregate of key requirements for technological processes, in dimensions and masses of assembled products, as well as in the place of production of various components.
The main technological requirements are compliance with the cleanliness of the premises according to the requirements of the first class of cleanliness according to RCC-MR during all assembly operations and testing, plus a carry outing of an acceptance tests in a clean room. The critical restrictions on the creation of the assembly hall are the mass of the equatorial port plug 48 tons and the length of the upper port plug (about 6 meters). To assemble the upper port plug, the necessary space under the crane hook should be at least 13 meters. Due to the presence of a large number of different components from different manufacturers (DFW, PPS, diagnostic components) that are delivered at different times, it is necessary to provide for the possibility of transportation and storage of all delivered components in a clean room of the assembly site, regardless of the stage of work on it.
As part of the preparation for assembly operations of large-sized products (diagnostic port plugs), a unique assembly hall was developed and created a unique complex of an integration hall.
The report gives a detailed data of the progress of work in a preparation for the integration of diagnostic ports on the territory of the BINP SB RAS.Speaker: Aleksandr Burdakov (Budker Institute of Nuclear Physics) -
Recent experiments at the JET tokamak with ITER-Like-Wall studied intrinsic rotation in a large tokamak, addressing questions related to the effects of collisionality and hydrogen isotope type on the amplitude of the measured toroidal rotation and rotation reversals of Ohmic plasmas. The isotope effect on the intrinsic rotation was investigated by comparing the rotation of the main ion in Hydrogen (H) and Deuterium (D) discharges. In order to assess the influence of machine size the JET studies were complemented by identity experiments performed at the DIII-D tokamak, providing a direct comparison of JET intrinsic rotation data with that of a medium size tokamak.
Increasing rotation shear in the plasma core is valuable for increasing thermal confinement, and yet what determines the shape of the rotation profile remains unclear. JET experiments studied the effect of density on the shape of the core rotation profiles of Ohmic divertor plasmas. Density scans in both H and D were performed at JET for the study of rotation reversals, in low triangularity configurations with $\mathrm{B}_{\mathrm{T}}=2.7\,\mathrm{T}$ in two different plasma currents ($\mathrm{I}_{\mathrm{p}}=1.7\, \mathrm{MA}$ and $2.3\,\mathrm{MA}$) and, in plasmas with a lower toroidal field, $\mathrm{B}_{\mathrm{T}}=1\,\mathrm{T}$, $\mathrm{I}_{\mathrm{p}}=0.9\,\mathrm{MA}$. The latter, were JET pulses for a JET-DIII-D identity experiment. The toroidal rotation of the main ion was measured from $\mathrm{H}_{\alpha}$ or $\mathrm{D}_{\alpha}$ charge exchange spectrum obtained during short bursts of Neutral Beam Injection (NBI).
At JET, as the density increased, two consecutive core rotation reversals (where the gradient of rotation changes sign in the middle of the plasma) were observed (Figures 1 (a) and (b)). At low densities co-current rotation decreased with a reversal on the core from peaked to hollow profiles. Further increasing the density leads to restoration of monotonic profiles, co-rotation now increasing as a function of density. Although rotation reversals in Ohmic plasmas have been a common observation in small and medium size tokamaks {1}, the JET experiments described here, were the first clear observation of rotation reversals in a large tokamak.
The phenomenology is similar in H and D, however the magnitude of the core rotation was found to depend on isotope type, stronger co-current rotation and larger co-current rotation gradients observed in H (Fig. 1 (b)). In all plasma configurations deeper counter-current rotation was observed in D. In addition, the critical densities for reversal, were found to depend on isotope type. Comparison of H and D rotation profiles at the same density, shows that the rotation difference is at the plasma core and not at the edge (Fig. 2).
An important question, for reliable rotation prediction for ITER, is whether changes in the core rotation shear, thought to be a result of turbulent transport, depend on machine size and if it will scale into a significant effect in ITER. In order to investigate how machine size can affect the magnitudes of intrinsic rotation and rotation shear, a JET-DIII-D identity experiment was performed at JET and DIII-D. Rotation was measured in DIII-D plasmas {2} that matched the JET plasma shape, $\mathrm{q}_{95}$ and other dimensionless parameters thought to be key for intrinsic rotation and intrinsic rotation reversals, such as the normalized gyro-radius, collision frequency and plasma pressure. Rotation reversals as a function of density were also observed in DIII-D. Both rotation reversals occur at lower densities in JET. Extrapolation to a larger machine such as ITER implies that intrinsic rotation profiles are likely to be in the peaked co-current rotating regime at the densities anticipated.
In JET and DIII-D the low density rotation reversal was observed near the transition from the linear to the saturated Ohmic confinement regime (Fig. 3). Linear gyro-kinetic calculations, with codes TGLF and GS2, show that in both JET and DIII-D the low density rotation reversal occurs close to the density of transition from dominant TEM to ITG instabilities. However, from JET data where a clear high density branch with peaked rotation profiles was observed, it is concluded that the instability type would not explain the rotation peaking at higher densities, since both hollow and peaked profiles are observed with dominant ITG.
Comparisons between intrinsic rotation data and ongoing turbulent momentum transport calculations have been performed. Non-linear calculations with the low-flow model in GS2 {3} show changes of sign in rotation gradient, consistent with the observation of peaked to hollow to peaked profiles as the density increased, though the modelled rotation shear is lower than that measured.
Bibliography
{1} J. Rice et al. Nucl. Fusion 51, 083005 (2011)
{2} B. Grierson et al. Phys. Plasmas 26, 042304 (2019)
{3} F. I. Parra and M. Barnes. Plasma Phys. Control Fusion 54, 045002 (2015)![DIII-D (a) energy confinement time and (b) central toroidal rotation in $[rad/s]$ (averaged over $r/a=0.0-0.3$) versus average line density; for a D pulse with $I_{\mathrm{p}}=1.0\, \mathrm{MA}$, $B_{\mathrm{T}}=2.0\,\mathrm{T}$ (plasma shape matched to JET). Near confinement transition from linear to saturate, the core rotation changes from peaked to hollow.](https://i.ibb.co/Y8Jb7HQ/Nave-FEC20-3-small-v4.png)
Acknowledgments
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 under grant agreement No 633053. In addition, it has received financial support from Fundacao para a Ciencia e Tecnologia through contracts UID/FIS/50010/2013, IF/00483/2014/CP1214/CT0008 and grant PD/BD/105877/2014. The views and opinions expressed herein do not necessarily reflect those of the European Commission. The work was supported by US DOE under contracts DE-FC02-04ER54698 and DE-AC02-09CH11466. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.Speaker: Dr Maria Filomena F. Nave (Instituto de Plasmas e Fusão Nuclear, Associação do Instituto Superior Técnico para a Investigação e o Desenvolvimento) -
Abstract
The nuclear fusion reactor ITER (International Thermonuclear Experimental Reactor) foresees a Pressure Suppression System (PSS) in order to manage a Loss of Coolant Accident (LOCA) or other over pressurization accidents in the Vacuum Vessel (VV) which has a pressure limit fixed at 150 kPa (abs).
This system (VVPSS) has a key safety function because a large internal pressure in the VV could lead to a breach of the primary confinement barrier. The pressure suppression is ensured by discharging the steam, produced by the LOCA, in 4 tanks, 100 m3 of volume partially filled by water, where it condenses (Figure 1).
Differently by the applications in the nuclear fission reactors, the steam condensation occurs in ITER at sub-atmospheric pressure conditions.
No previous applications of direct steam condensation at sub-atmospheric conditions have been done. Therefore, experimental assessment of VVPSS has been necessary considering the direct condensation of the steam in the ITER thermal-hydraulic conditions.
Under financial support of ITER Organization an extensive experimental program has been performed at the Department of Civil and Industrial Engineering (DICI) of the University of Pisa using a small-scale experimental rig (1/22 scale factor) [ref1]-[ref3]. Figure 2 shows the reduced scale experimental rig.
The main component is the Reduced Condensation Tank (RCT). It is deeply instrumented with 28 temperature sensors and 8 pressure transducers. An internal vertical sparger with several holes discharges the steam inside the water. The steam mass flow rate, produced by an electric steam generator, is controlled by means of Vortex or Coriolis measurement devices.
Different temperature, pressure and steam mass (flow rate per hole) conditions and sparger patterns have been investigated in about 400 tests. The experimental results demonstrated that there are strong differences between the steam direct condensation at atmospheric and sub-atmospheric pressure.
First of all the condensation regimes (stable, oscillation, chugging and bubbling condensation) occur for different values of unit steam mass flow rate per hole (Kg/sm2). Moreover these condensation regimes depend on the pressure inside of the tanks. Figure 3 illustrates the map of condensation regimes determined by means of the experimental program.
The experimental results permitted to elaborate a suitable similitude analysis which describes the steam condensation at sub-atmospheric pressure.
ITER Organization launched an international tender in order to perform an experimental activity on a Large Scale facility which the aim was the assessment the scale laws determined in the small scale facility. University of Pisa won this international tender and has built the large scale experimental rig, shown in Figure 4.
This facility, presently in commissioning phase, is about a full scale from the geometric point the view (1/1.09 scale factor) and it is 1/10 scale factor from the steam mass flow rate point of view. In fact the condensation tank has a volume of 92 m3 and the electric steam generator produces 0.5 Kg/s mass flow rate of steam. The actual tanks of VVPSS have 100 m3 of volume and the maximum steam mass flow rate for each tank in the envisaged accidental scenario (Large LOCA accident) is 5 Kg/s. The built facility has the possibility also to simulate the full scale accidental scenario. By means two 15 m3 high pressure vessels it is possible to produce 5 Kg/s of steam for an interval of time equal to the foreseen accidental scenario.
This paper describes the main results obtained by means of the tests performed on the small scale rig. Moreover the Large Scale experimental rig built at the University of Pisa and the test matrix, which will be performed, are illustrated. These tests will permit to verify the scale laws determined in the previously performed research program on the small scale rig.
The results of this research program are of fundamental importance in order to assess the reliability of VVPSS.References
[ref1] Rosa Lo Frano-Dahane Mazed--Marco Olcese-Donato Aquaro Daniele Del Serra Igor Sekachev Guglielmo Giambartolomei Investigation of vibrations caused by the steam condensation at sub-atmospheric condition in VVPSS Fusion Engineering and Design Volume 136, Part B, November 2018, Pages 1433-1437
[ref2] D. Mazed, R. Lo Frano D. Aquaro D. Del Serra, I. Sekachev, M. Olcese - Experimental investigation of steam condensation in water tank at subatmospheric pressure Nuclear Engineering and Design, ISSN: 0029-5493, Vol: 335, Page: 241-254 Publication Year2018
[ref3] R.Lo Frano,et al. Experimental investigation of functional performance of a vacuum vessel pressure suppression system of ITER Fusion Eng. Des.,122(2017), pp.42-46Speaker: Prof. Donato Aquaro (University of Pisa (Italy)) -
A transition from an interchange mode with high mode numbers $(m,n)$ for $m$ (poloidal) and $n$ (toroidal) to a non-resonant $(m,n)=(1,1)$ mode is found in the nonlinear magnetohydrodynamic (MHD) simulation for a Large Helical Device (LHD) plasma with net toroidal current. This transition occurs when the rotational transform is closed to unity in the core region. Because partial collapses caused by (1,1) modes are observed in the LHD experiments with net toroidal current, this transition to the non-resonant mode is a candidate for explaining the partial collapses.
In the magnetic confinement of plasmas, it is crucial that the plasmas are stable against MHD instabilities. Therefore, the stability property is extensively studied in the LHD experiments. In such experiments, partial collapse phenomena are observed when net toroidal current is driven by the neutral beam injection so that the rotational transform $\iota / 2 \pi$ is increased [1]. These collapses are always caused by (1,1) modes. These modes are considered to be pressure driven modes because the equilibria are strongly Mercier unstable. However, according to the theory of the pressure driven modes, the linear growth rate is larger for higher mode numbers. Therefore, it has been required to explain why the partial collapses are caused by the (1,1) modes in LHD.
This problem is investigated by means of three-dimensional nonlinear MHD simulations for LHD plasmas with net toroidal current. In the simulations, the HINT[2] and the MIPS[3] codes are utilized for the equilibrium and the nonlinear dynamics calculations, respectively. In the equilibrium calculation, the vacuum magnetic configuration corresponding to the experiment in Ref.[1] is used. The pressure profile $P_{eq}$ and the axis beta $\beta_0$ are employed as $ P_{eq} = P_0 ( 1 - 0. 68 \rho^2 - 0. 32 \rho^4 ) $ and $\beta_0 =1.6\%$, respectively. Here $\rho $ denotes the square root of the normalized toroidal magnetic flux. The current density of the net toroidal current is assumed as $ J_{eq} = J_0 ( 1 - \rho^2 )^4 $ and the total current $I=80$kA for $B_0 = 3$T are chosen. The values of $\beta_0$ and $I$ are close to those of the experiment in Ref.[1]. In this equilibrium, $\iota / 2\pi $ is close to unity and a no-shear point appears in the core region as shown in Fig.1 (a) and (b). Similar rotational transform profiles are often observed in the experiments.
In the nonlinear dynamics calculation, the values of $S=5.6\times 10^6$, $\nu = \chi_\perp = 2.5 $m$^2$/s are employed for the magnetic Reynolds number, the viscosity and the perpendicular heat conductivity, respectively. These values are relevant to the experiments. Also, $\chi_\parallel = 10^{3} \chi_\perp $ is assumed for the parallel heat conductivity. At $t/\tau_A = 500$ in the early nonlinear phase, the (3,3) component is dominant as shown in Fig.1 (a), where $\tau_A$ denotes the Alfv$\acute{\mbox{e}}$n time. This mode is a typical interchange mode resonant at the $\iota /2 \pi = 1$ surface. These mode numbers are higher than those in the observations. However, at $t/\tau_A =800$ in the further nonlinear phase, the (1,1) component appears and becomes dominant as shown in Fig.1 (b), which corresponds to the observations. This mode is not resonant at the $\iota / 2 \pi = 1 $ surface but is localized around the no-shear region. Therefore, a transition from the resonant (3,3) mode to the non-resonant (1,1) mode occurs. Figure 2 (a) shows that the magnetic field becomes locally stochastic in the unstable region due to the $m=1$ global convection. This convection also causes the partial collapse of the total pressure from one poloidal direction as shown in Fig.2 (b). This collapse tendency is similar to the experimental observation as shown in Fig.2 (c).
The case with no net toroidal current ($I=0$) is also examined as a reference. Here, $\beta_0 =3.0\%$ is employed. As shown in Fig.3 (a), the rotational transform has a monotonically increasing profile. In early nonlinear phase, the (3,2) component is dominant and the (1,1) component is negligibly small. This perturbation is a typical interchange mode resonant at the $\iota / 2 \pi = 2/3$ surface. The total pressure profile decays with a triangular deformation as shown in Fig.3 (b). In the further time evolution, other sideband modes are enhanced and the total pressure decays in multiple poloidal directions. These mode numbers and decay property do not match the observed partial collapse shown in Fig. 2(c). Therefore, it is necessary for the transition that the rotational transform is changed by the net toroidal current so as to have no-shear region where the value is close to unity. Such a profile in the safety factor can appear also in tokamak sawtooth collapses. Therefore, extension of the analysis for flat rotational transform profiles is expected to provide common knowledge for avoiding collapses in current carrying plasmas.
![(a) Puncture plots of the magnetic field lines and patterns of positive (red) and negative (blue) pressure perturbations, (b) bird's eye view of the total pressure and (c) the pressure profile along $z=0$ line at $t/\tau_A=800$ for $I=80$kA. A typical profile obtained in [1] is also plotted in (c).](https://workshop.nifs.ac.jp/fec2020/image/60-ICHIGUCHI-image-fig_2.png)
[1] S. Sakakibara, et al., 2015 Nucl. Fusion, 55 083020.
[2] Y. Suzuki, et al, Nucl. Fusion 46 (2006) L19.
[3] Y. Todo, et al, Plasma and Fusion Res. 5 (2010) S2062.Speaker: Katsuji ICHIGUCHI (National Institute for Fusion Science) -
The NSTX-U Recovery Project is completing the transition from the design phase to the fabrication and installation phase in the early part of 2020. The design that has been established is based on the findings of the 2017 Extent of Condition review, and includes significant technical improvements to the machine core and auxiliary systems, in order to provide a high-reliability use facility. The Project had a successful DOE baseline review in August 2019; this review and subsequent DOE authorizations included significant elements of long lead procurement, and many components are presently in fabrication. This paper will describe key technical aspects of the Recovery Project design, as well as progress towards Project completion.
A key aspect Recovery Project scope are the new inner-PF coils. Six new coils are being fabricated, based on specific design improvements relative to the previous coils. This includes the elimination of in-line braze joints and simplification of the winding pattern to ease manufacturing. An image of the first production PF-1a coil, with two of four winding layers near complete, is shown in Fig. 1
. The Recovery Project is also fabricating a new Center Stack (CS) assembly to provide the inner vacuum boundary and to support the coils (see Fig. 2)
. The 4.2 m tall CS casing is under fabrication, with all forgings having been completed. These will be welded and post-machined to provide the appropriate tolerances. Compared to the original NSTX-U CS casing, this component has improved weld designs, enhanced heating and cooling features to support bakeout and operations, and tolerances that support Project alignment requirement.The new CS assembly also includes numerous other new features shown in Fig. 2. The PF-1a and PF-1b coils are preloaded in Inconel 718 slings; this preload provides mitigation of thermal stresses due to water cooling after the pulse. Double O-ring seals are used at each of the primary vacuum interfaces, reducing both the likelihood of major leaks and the level of permeation.
Other key fabrication elements of the machine core scope have started, including:
• New plasma facing components are being fabricated for the CS casing and all divertor regions. In high heat flux regions, these utilize fine-grain isotropic graphite, are castellated to reduce the thermal stresses, and have “fishscaling” to eliminate leading edge heating.
• NSTX-U utilizes in-vessel passive stabilizing plates to reduce the growth rates of instabilities such as VDEs and RWMs. These plates have large loads during disruptions, and the Recovery Project is implementing improved welds and brackets to counteract these forces.
• The outer PF coils are being realigned to meet stringent error field requirements.A major component of Recovery scope involves improving the neutron shielding of the NSTX-U test cell. A new concrete labyrinth wall was poured at the primary test cell entrance, and roof blocks were fabricated (See Fig. 3)
. Additionally, many penetrations in the test cell walls are being filled with grout, neutron putty, or polyethylene sheets.The Recovery Project is implementing a new Personnel Safety System (PSS). This includes a set of configuration managed safeguards against contact hazards, a trapped key system for managing the facility configuration, and a new IEC 61511 compliant access control system using a safety-class PLC. These systems will significantly enhance the operational safety of the facility.
The Recovery Project is making rapid progress, with first plasma scheduled for the summer of 2021.
This work was supported by U.S. DOE Contract D-AC02-09CH11466 and DE-AC05-00OR22725. The project appreciates analysis support provided by Oak Ridge National Labs.
Speaker: Mr Stefan Gerhardt -
A helical coil designed to passively generate non-axisymmetric fields during a plasma disruption has been shown (via electromagnetic circuit and linear MHD modeling) to be effective at deconfining runaway electrons (REs) before the RE beam current grows to dangerous levels. Magnetic equilibria from DIII-D RE experiments were used to calculate the toroidal electric field generated during the current quench phase of a disruption, which in turn drives current in a proposed n=1, m=1 in-vessel helical coil (Fig.1), without the need for any external power supplies or disruption detection diagnostics. Modeling using both the JFIT and TokSys codes predicts coil currents up to 6% of pre-disruption plasma current (Fig.2). The coil geometry was systematically varied to maximize the magnetic perturbation at the plasma rational surfaces, resulting in an optimized coil where $\delta$B/B~$10^{-2}$. The new REORBIT module of the MARS-F code was used to model the full non-axisymmetric magnetic field and trace RE orbits to determine the coil’s effect on RE deconfinement, with up to 45% of the RE population lost after 200$\mu$s. Electromagnetic and thermal stresses on the coil are calculated to be well within operational limits for installation in DIII-D, and scale favorably to a reactor-size device.
An optimized coil geometry was constructed with the goal of maximizing the non-axisymmetric magnetic perturbation in the disrupting plasma. Equilibrium fits from runaway electron experiments in DIII-D were used to approximate the magnetic geometry for field harmonic calculations. The coil geometry was parametrized in several ways, and the optimal shape was found to have shallow pitch angle and few discrete horizontal and vertical segments, localized at the midplane center-post (Fig.1a). This position minimizes the distance between coil and the disrupting plasma, maximizing the loop voltage along the coil as well as the magnetic perturbation amplitude in the plasma. This design also avoids the large number of outboard ports present on DIII-D and other tokamaks.
In order to estimate the toroidal electric field generated during a disruption, the JFIT and TokSys codes were used to model the inductively coupled coil-plasma-vessel system. Both models used full 3D inductive coupling terms to calculate the evolution of the axisymmetric plasma current, predicting peak coil currents of up to 6% of the initial plasma current which are well in excess of the theoretical 2% threshold for meaningful RE deconfinement {1}. Most importantly, the coil inductance and resistance are suitable to match the disruption current quench timescale (Fig.2a) (14ms in DIII-D, 190ms in ITER). This allows the coil to passively respond to the disruption and generate a 3D magnetic perturbation with the necessary amplitude and timing for runaway electron deconfinement. The resulting vacuum field amplitude is greater than the theoretical threshold of $\delta$B/B~$10^{-3}$ {2}, and is well-matched to the edge plasma helicity (Fig. 2b).
The amount of RE deconfinement caused by this optimized coil was modeled using the new REORBIT module of the linear MHD code MARS-F {3}. After calculating the full non-axisymmetric magnetic field for toroidal mode numbers up to n=6, the REORBIT code traces test particles along field lines to determine their confinement time as a function of initial energy and pitch angle. Simulations extending out to 240$\mu$s post-disruption found that 18% of the total RE population was deconfined by expected coil currents (6% I$_\text{P}$). Doubling the coil current (12% I$_\text{P}$) further increased the loss fraction to 45% of the total RE population (Fig.1b). Unlike previous studies {4} these results do not require significant resonant plasma response or island formation; in fact, the modeled plasma response in this study was found to be negligible even though the vacuum magnetic perturbation is well-aligned with the edge helicity. Thus it is likely that these results are pessimistic, and that equilibria with lower q$_{\text{edge}}$ would show increased RE deconfinement due to non-trivial resonant response.
Scaling estimates are favorable for the operation of a passive RE deconfinement coil on a larger reactor-scale device, with only marginal decreases in relative coil current and magnetic perturbation amplitude (Fig.2). The DIII-D coil-plasma-vessel system was scaled up by a factor of 3.75 to approximate an ITER-sized device while maintaining the same edge safety factor and aspect ratio. Interestingly, these constraints dictate that I$_\text{coil}$/I$_\text{P}$ and $\delta$B/B are independent of size; the slight differences in Fig.2 are due solely to the change in vacuum vessel conductivity (Inconel to stainless steel). The mechanical and thermal stresses of such a coil on an ITER-sized device are only slightly larger than the DIII-D case. Even considering the worst-case high-current scenarios in scaling from the DIII-D (I$_\text{P}$ = 2MA) to ITER (I$_\text{P}$ = 15MA) sizes, J$\times$B stress on the coil increases from 30% to 40% of stainless steel tensile yield strength and total coil temperature rise due to joule heating increases from $\Delta$T = 20$^{\circ}$C to 50$^{\circ}$C.
In summary, a helical in-vessel coil was designed with the goal of passively deconfining REs prior to energetic beam formation via large non-axisymmetric magnetic perturbations. The linear MHD code MARS-F was used to test this hypothesis, and the coil was found to be effective at substantially increasing RE orbit losses in DIII-D equilibria. Such a coil is robust to mechanical and thermal stresses, reacts passively and quickly to a disruption event, and scales favorably to larger reactor-size devices. The experimental confirmation of these findings on an existing tokamak should be a high-priority goal of the international fusion community, and serious thought should be given to its applicability to future fusion devices.
Work supported by General Atomics Internal funds.
{1} Boozer, Plasma. Phys. Control Fusion 53 (2011) 084022
{2} Boozer, Phys. Plasmas 19 (2012) 058101
{3} Liu, Parks, Paz-Soldan, Kim, and Lao, Nucl. Fusion 59 (2019) 126021
{4} Smith, Boozer, and Helander, Phys. Plasmas 20 (2013) 072505Speaker: David Weisberg (General Atomics) -
The extreme energy content in ITER makes the disruptions a matter of grave concern. The current strategy of disruption mitigation in ITER relies on the injection of cryogenic pellets into the disruptive plasma [1]. Pellet ablation is an essential factor in disruption mitigation, which calls for dedicated theoretical support in modeling this process in the plasma and establishing related constraints on the disruption mitigation system.
This work addresses the following aspects of pellet-plasma interaction physics:
• high-Z pellet ablation;
• thermal electron response to the pellet during ablation;
• pellet interaction with runaway electrons.
We present a first principle kinetic calculation of the power deposition from energetic electrons into the cold halo of an ablating high-Z pellet. For high Z, the velocity distribution of the hot electrons is nearly isotropic, and we use this feature to simplify and solve the electron kinetic equation, including the effect of electron gyro-motion in a magnetic field [2]. The resulting ablation rate for the high-Z pellets can be estimated as
$\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ G\sim \frac{4\pi M}{{{\left( \pi {{e}^{4}} \right)}^{2/3}}}{{\left( \frac{1}{{{m}^{1/2}}M} \right)}^{1/3}}R_{p}^{4/3}\frac{n_{\infty }^{1/3}T_{\infty }^{11/6}}{{{Z}^{7/6}}{{\left( \ln {{\Lambda }_{ei}} \right)}^{1/2}}{{\left( \ln {{\Lambda }_{ee}} \right)}^{1/6}}},$
where $e$ and $m$ are the electron charge and mass, $M$ is the ion mass, ${{R}_{p}}$ is the pellet radius, ${{n}_{\infty }}$ and ${{T}_{\infty }}$are the density and temperature of the ambient electrons, and $\ln {{\Lambda }_{ei}}$ and $\ln {{\Lambda }_{ee}}$are Coulomb logarithms for elastic and inelastic collisions.
![Reduction of ablation rate due to elastic scattering. Predictions by Sergeev et al. $[3]$ (red line) are too high. The difference between our results (black lines) and those by Parks $[4]$ (blue diamond) shows significant sensitivity of the heat deposition to elastic scattering.](http://i.ibb.co/wr5x5sf/fig1.png)
This rate is much lower than the pre-existing estimates, as shown in Fig. 1. We find that strong elastic scattering of the incident electrons reduces the role of electrostatic shielding significantly. The new expression for the heat deposition provides an updated input for fluid simulations of the pellet ablation process.
We also consider the impact of pellets on plasma electrons [5]. The cold, dense pellet absorbs some of the incident hot electrons and emits secondary electrons to maintain quasi-neutrality. The required balance is provided by a sheath potential. The resulting electron distribution function in the ambient plasma may, therefore, deviate significantly from the Maxwellian (see Fig. 2), which can affect the heat flux into a low-Z pellet. The pellet also tends to modify the distribution of the incident electrons and the heat flux via energy-dependent absorption. This heat depletion is of particular interest for resonant magnetic surfaces.
In a cold post thermal quench plasma, runaway electrons can carry a significant fraction of the initial plasma current. Our first principle estimates show that any pellet injected to dissipate a 10 MA runaway electron current in ITER will evaporate virtually instantly to form a gas cloud once exposed to the runaway electrons. This is in line with the recent observations in DIIID, where there was no significant difference between the runaway electron dissipation by pellets or by massive gas injection [6].
Work supported by the U.S. Department of Energy Contracts DEFG02-04ER54742 and DESC0016283.
References
[1] L.R. Baylor, et al., Fusion Science and Technology 68 (2), 211-215 (2015).
[2] A.K. Fontanilla and B.N. Breizman, Nucl. Fusion 59, 096033 (2019).
[3] V.Yu. Sergeev, et al., Plasma Physics Reports 32, 5 (2006).
[4] P.B. Parks, The ablation rate of light-element pellets with a kinetic treatment for penetration of plasma electrons through the ablation cloud, Submitted to Phys. Plasmas (2019).
[5] D.I. Kiramov and B.N. Breizman, Phys. Plasmas 27, 022302 (2020).
[6] D. Shiraki, et al., Nucl. Fusion 58, 056006 (2018).Speaker: Boris Breizman (Institute for Fusion Studies, The University of Texas, Austin, Texas, 78712 USA) -
Plasma rotation in thermonuclear fusion plasma plays an important role for particle and energy confinement and it can stabilize magnetohydrodynamic instabilities if some rotation level is achieved. In early studies, the relation between plasma confinement and radial electric field E_r was studied via measurements of the electric potential distribution and also via spectroscopic measurements of plasma rotation. Since no significant effect of E_r on plasma confinement was observed, little attention was given to it. However, after the discovery of H-mode, the need for accurate measurements of E_r, and consequently, plasma rotation, became necessary. Nowadays, interest in the electric field, whether by its origin or by influence in plasma confinement, is one of the most current topics in nuclear fusion research.
In TCABR, plasma rotation has been measured using Doppler shift of both the C^{+5} spectral line emission (529.05 nm) [1-2]. In Figure 1, (b) poloidal and (c) toroidal rotation measurements of C^{+5}. In panels (a-b), C^{+5} poloidal rotation measurements are compared with the rotation profile expected from neoclassical theory. As can be seen, the observed C^{+5} poloidal rotation in TCABR is well described by neoclassical theory.
Plasma rotation in the presence of anomalous processes and toroidal momentum damping are introduced phenomenologically into neoclassic equations in [3]. So, the next phase of this work will focus on the estimation of radial electric field and the comparison of the toroidal velocity with the model proposed in [3-4].
ACKNOWLEDGMENTS
This work was supported by the São Paulo Research Foundation FAPESP
(proc: 2018/09734-0).[1] J. H. F. Severo, et al Nuclear Fusion, vol. 43, no. 10, p. 1047, 2003.
[2] J. H. F. Severo, et al Nuclear Fusion, vol. 49, no. 11, p. 115026, 2009.
[3] V. I. Bugarya, et al Nuclear Fusion, vol. 25, no. 12, p. 1707, 1985.
[4] Rozhansky V and Tendler M 1996 " Plasma Rotation in Tokamak" in Rev. Plasma Physics v.19, ed. B.B.Kadomtsev (N.Y.:Consultans Bureau).Speaker: Jose Helder Severo (University of Sao Paulo) -
Ion Cyclotron Resonance Frequency (ICRF) heating plays an important role in many present day experiments and it is one of the auxiliary heating methods that will be used in ITER. In this contribution, we will review the recent key ICRF results from the JET and ASDEX Upgrade (AUG) tokamaks in preparation of ITER.
In the recent JET campaigns, the focus has been in the preparation of integrated scenarios for high fusion performance with long duration (Pfus=15MW for 5s) (1) and alpha physics (2) in the forthcoming campaign with deuterium-tritium (D-T) fuel mixture. Following the successful earlier characterization and optimization of hydrogen minority heating for the use in JET scenario plasmas (3), the new developments include the integration of He-3 minority heating in high-performance D plasmas for improved bulk ion heating compatible with the control of central high-Z impurity accumulation in the presence of the ITER-like metallic wall. With He-3 minority heating, the best plasma performance in terms of the neutron rate and plasma energy content was obtained at a low He-3 concentration of ~2%. The resulting modest He-3 consumption is advantageous in light of lower operational cost when using He-3 minority heating in ITER. It is also well in line with earlier computational multi-code work (4) for ITER where good absorption performance with a He-3 concentration of ~3% was found. In the coming JET campaigns, which will include a campaign with tritium and D-T plasmas, these experiments will be extended to the studies of He-3 minority heating and second harmonic heating of tritium, which are the two main ICRF heating schemes planned for ITER full-field operation in 50%-50% D-T plasmas. Further ICRF options for JET D-T campaign are discussed following a recent review (5).
On AUG, novel applications of ICRF waves for plasma heating have become possible through the improved operating space of ICRF system and, in particular, its extended frequency range (6). It has allowed the application of second harmonic heating of hydrogen on AUG for improved core electron heating in the ITER-baseline-like plasmas with pure wave heating (i.e. without NBI-induced torque to simulate ITER burning plasma conditions) (7). The extended frequency range has also been instrumental for the experiments using third harmonic ICRF heating of NBI-injected deuterons for fast ion studies and for further development of fast ion and neutron diagnostics. Figure 1 shows a typical discharge with a more than two-fold increase of the D-D fusion rate due to ICRF-accelerated deuterons achieved with this scheme in AUG. As a continuation of the very successful earlier experiments with this scheme on JET (8), this ICRF development on AUG has provided for the first time a means for simultaneous controlled variations and measurements of both the confined and the non-confined parts of ICRF-driven fast deuterium distribution. Furthermore, analysis and modeling of JET and AUG experiments in D, H-D, H-He-4 and D-He-4 plasmas heated with He-3 minority heating and the so-called three-ion ICRF schemes (9) have provided improved insights on core ICRF physics as well as nonlinear electromagnetic stabilization of ion temperature gradient (ITG) modes by fast ions (10,11).
The rich variety of new ICRF scenarios in various plasma scenarios (H-mode and improved confinement regimes) in the two devices of different sizes has formed a challenging test bed for the validation of numerous modelling tools. We will discuss some representative examples from the comparisons of experimental results with the ICRF modelling code PION (12). PION computes the ICRF power absorption and the distribution functions of the resonant ions in a self-consistent way. Thanks to its speed, it forms a part of the automated data processing chain at JET, and has recently been installed in the ITER Integrated Modelling and Analysis Suite (IMAS) for integrated predictive modelling of ITER.
Despite its relatively simple physics model, we find that PION reproduces successfully many features observed in the recent ICRF experiments on JET and AUG. For example, in the case of modelling the novel three-ion-schemes, it reproduces the strong ion cyclotron damping by third ion species despite its low concentration, strong ICRF acceleration of resonant ions into the MeV range, and the dependence of confined and lost resonant ions distribution functions on experimental parameters (13). Our results increase our confidence in the applications of PION such as those reported in (14, 15) for predictive simulations of future experiments planned in the JET D-T campaign and ITER.
Acknowledgements This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
1. L. Garzotti et al. 2019, Nucl. Fusion 59, 076037
2. R. Dumont et al. 2018, Nucl. Fusion 58, 082005
3. M.J. Mantsinen et al. 2017, EPJ Web of Conferences 157, 03032.
4. R. Bilato et al., 2014, AIP Conference Proceedings 1580, 291.
5. E. Lerche et al., RFPPC 2019.
6. Vl. Bobkov et al., RFPPC 2019.
7. T. Pütterich et al., 27th IAEA Fusion Energy Conference (FEC2018), EX/P8-4; T. Pütterich et al, EPS2020.
8. S.E. Sharapov et al. 2016, Nucl. Fusion 56 112021 and references therein.
9. Ye.O. Kazakov et al., this conference and references therein.
10. N. deOliveira et al., 2020, submitted to Nuclear Fusion.
11. A. Di Siena et al 2019 Nucl. Fusion 59, 124001.
10. L.-G. Eriksson, T. Hellsten and U. Willén 1993, Nucl. Fusion 33, 1037.
11. M.J. Mantsinen et al., 2019 Europhysics Conference Abstracts, vol. 43C, O5.102.
12. D. Gallart et al., RFPPC 2019.
13. I.L. Arbina et al., 2019 Europhysics Conference Abstracts, vol. 43C, P4.1079.Speaker: Prof. Mervi J. Mantsinen (ICREA and Barcelona Supercomputing Center) -
Shattered pellet injection (SPI) has been selected as the baseline disruption mitigation (DM) system for ITER. SPI utilizes cryogenic cooling to desublimate low pressure (<100 mbar) gases onto a cold zone within a pipe gun barrel, forming a cylindrical pellet. Pellets are dislodged from the barrel and accelerated using either a gas driven mechanical punch or high-pressure light-gas delivered by a fast opening valve. SPI technology is currently deployed and operational on DIII-D, JET, and KSTAR. These SPI systems are used in experiments for physics scaling of ITER thermal mitigation and runaway electron dissipation/avoidance. The pellet sizes used for these machines are in the range of 4 to 12.5 mm in diameter with length to diameter ratios (L/D) of ~1.5. The current plan for ITER SPI is to utilize pellets that are 28 mm in diameter with an L/D of ~2. The large pellet sizes, high steady-state magnetic fields, and limitations of operating in a radiation environment render much of the current technology unusable. In addition to technology improvements, a deeper understanding of pellet material properties, formation, and release is being developed for implementation in future SPI designs, specifically ITER.
The present solenoid driven propellant valve used on the SPIs, cannot operate in the high steady-state magnetic field due to the magnetic circuits utilized to open the valve. A new valve is being designed and tested for operation in the ITER magnetic field and tritium environment. This valve operates by inducing eddy currents in an aluminum “flyer plate” [1]. The force from the eddy currents lifts the plate, which unseats the valve to allow high pressure light-gas to flow into the breech of the pipe gun to dislodge the pellet. The valve and corresponding power supply have undergone various design upgrades and multiple testing campaigns. Lifecycle testing with and without an external magnetic field will be conducted to ensure safe and reliable operation through the life of the valve. An image of the valve test stand can be seen in Fig. 1. Measurements have been made of the amount of gas delivered by the valve. Star CCM+ CFD simulations of the valve gas flow were conducted and validated against experimental data. The CFD model was then updated with an ITER relevant SPI barrel geometry. The breech volume was optimized for the maximum force on the pellet. The limitation on having the smallest possible breech volume is the thermal conduction to the cold-zone from close proximity to the room temperature surface.
The desire to use pure neon and argon in the SPI for runaway electron dissipation has introduced new challenges for the SPI technology. Pure neon and argon pellets are much more difficult to dislodge from the cold zone of the barrel due to high material strength compared to deuterium. The force needed to dislodge a pellet increases with size making release of the 28 mm ITER pellets a challenge. To better understand how to most effectively dislodge and accelerate pellets, the process of how they are formed, and the important influencing factors, such as temperature and pressure, must be understood and documented.
An experiment has been designed to measure the shear strength of pellets as a function of temperature. The experiment uses a load cell to measure the force applied by a motor driven shaft, which slowly compresses the rear of the pellet until it dislodges. The peak force is used to calculate the shear strength of the material in a pipe gun barrel. Four different barrel sizes were used to measure the shear strength of pellets made from pure deuterium, pure neon, deuterium-neon mixtures, and pure argon. These planned measurements, combined with ANSYS simulations of stresses produced by high-pressure gas flow and a mechanical punch striking the pellet, will allow for a more complete understanding of the pellet release process. This experiment will also be used to determine whether formation pressures and temperatures impact pellet breakaway strengths.Understanding the pellet formation process and how pellet strength and release are related is an important aspect of SPI design. The propellant valve and mechanical punch must be designed to achieve the release parameters for the ITER SPI system to be reliable and successful. Pellets must also be able to survive the sometimes-treacherous flight path between the barrel and the plasma. Experiments were conducted to determine the pellet fracture threshold for pure deuterium, pure neon, various deuterium-neon mixtures, and pure argon pellets by measuring the perpendicular impact velocity [2]. It was found that pure deuterium, neon, and argon pellets have normal velocity thresholds for fracturing of 20, 8, and 6 m/s and there was no size dependence on these values. Deuterium-neon mixture pellets have a fracture threshold velocity that is dependent on the mass percentage of neon in the mixture. After pellets traverse their flight paths they are shattered before entering the plasma by impacting a bent tube or angled plate. A theoretical model of resulting fragment size in the plume, based on brittle fracture mechanics, has been developed [3]. This model is intended to provide a statistical description of the shattered material for computational simulations of disruption mitigation events using SPI. This model has successfully reproduced the fragment size distributions from the shatter tube geometries used on JET, DIII-D, and KSTAR. Figure 2 shows the estimated fragment size distribution for ITER sized pure deuterium, neon, and argon pellets. Estimated pellet speeds are 500, 350, and 150 m/s for deuterium, neon, and argon and the angle of the shatter surface is assumed to be 20-degrees. The deuterium and neon pellets are assumed to be accelerated using high pressure gas and the argon pellet is assumed to be accelerated using a mechanical punch. For a pure neon pellet to be fired using high pressure gas a thin shell of deuterium can be used to ease breakaway.
All these advances in technology and physical understanding of the SPI formation and shattering process will be incorporated in the ITER DMS design. An ITER prototype SPI test stand is under construction at Oak Ridge National Laboratory that incorporates our enhanced physical understanding of the SPI process. A new flyer plate valve will be used to dislodge and accelerate pellets that will be formed in conditions for rapid pellet formation, successful breakaway, guide tube survivability, and optimal shattering.
[1] S. A. Bozhenkov et al., Rev. of Sci. Inst. 78, 033503 (2007)
[2] L. R. Baylor et al., Nuc. Fus. 59 (2019)
[3] T. E. Gebhart et al., IEEE Trans. Plas. Sci. (2020)Speaker: Dr Trey Gebhart (Oak Ridge National Laboratory) -
This paper presents numerical modelling to assess the fast shutdown scenario by shattered pellet injection (SPI) in a baseline strategy of the ITER Disruption Mitigation System (DMS). A new versatile 1.5D disruption simulator INDEX is applied for this purpose and a SPI ablation/assimilation model has been implemented. Pre-thermal quench (pre-TQ) H$_2$/D$_2$ SPI is proposed as a promising scheme to avoid hot-tail runaways and the simulation shows that the scheme is consistent with the material injection capabilities in the present DMS design.
Simulation Method:
The INDEX code used in this paper is a 1.5D tokamak modelling code coupled to an external circuit model, which has recently been developed at QST for disruption load assessment. The axisymmetric conductor model of ITER was implemented and benchmarked successfully against the DINA code [V. Lukash et al 2005 Plasma Phys. Control. Fusion 47 2077]. To evaluate the particle source due to SPI, the motion of thousands of pellet shards is traced numerically as marker particles. The ablation scaling laws, e.g., Ne/D$_2$ composite pellet ablation scaling [P. B. Parks, TSD Workshop, Princeton 2017], are used to calculate the ablation of shards and underlying ablation physics is discussed. Though the present 1.5D modelling comes with simplifications and does not capture 3D aspects of SPI, its easy-to-use and fast computation capabilities allow assessing sensitivity of the mitigation process on key plasma and SPI engineering parameters.Pure H$_2$/D$_2$ SPI Scheme to Raise the Electron Density for RE avoidance:
Runaway electrons (REs) are an important issue for ITER operation because a very large avalanche gain was predicted by the theory and also because the high temperature of fusion core plasma can produce significant runaway seeds through the hot-tail mechanism. It has been proposed to use SPI with a mixture of a small quantity of Ne and a large quantity of D$_2$ to avoid REs [J. R. Martín-Solís et al 2017 Nucl. Fusion 57 066025]. However, it is uncertain whether the plasma can assimilate the required amounts of material, because a TQ will likely be triggered when the cold front reaches the q = 2 surface, at which point the energy is radiated away and not available for pellet ablation anymore.This paper proposes a different scheme based on a two-step SPI: first, pure H$_2$/D$_2$ SPI to raise the electron density; second, Ne or Ne mixed SPI to radiate the stored energy. At the first step, we revisit massive deuterium injection [M. N. Rosenbluth et al 1996 Proc. 16th Int. Conf. on Fusion Energy, Montreal Vol. 2 p. 979] with the assumption that dilution cooling does not trigger an immediate TQ when the pellet shards reach the q = 2 surface. To support this, we present INDEX simulations of plasma and current profiles after SPI in figure 1. It is seen that comparing with Ne mixed SPI case (top), pure H$_2$/D$_2$ SPI (bottom) does not cause a radiative collapse of the electron temperature profile to 10 eV range, and does not perturb the current profile. After dilution cooling, the slowing down time of hot electrons (whose velocity is much faster than the thermal velocity) can be shorter than TQ times (0.1-1ms), while the voltage is maintained below the critical threshold for runaways. Therefore, raising the density in such a way has the benefit of avoiding hot-tail runaways.
Open questions remain regarding the role of intrinsic impurities, whether densities significantly higher than the Greenwald value can transiently be achieved, and the available time window for the subsequent Ne injection to radiate the stored thermal and magnetic energies. Detailed INDEX simulations will be presented on these aspects. Experimentally, a super-dense plasma with the electron density up to $7\times 10^{21}$/m$^3$, far exceeding the Greenwald limit, is observed transiently after pre-TQ D$_2$ SPI at DIII-D [N. Commaux et al 2010 Nucl. Fusion 50 112001], which may support the present scheme. Careful investigations are also needed for the post-TQ phase in ITER because even if the hot-tail mechanism is avoided, the tritium decay and Compton scattering act as steady sources of REs in nuclear operation. As a possible option to mitigate REs, we consider performing repetitive injections of H$_2$/D$_2$ pellets to deplete the RE population by collisions of REs with the shards.
Assimilation Efficiency of Multiple Injection in ITER Reference Scenarios:
Another highlight of this paper is an extensive study of the assimilation efficiency of SPI. Different injection quantities, shard sizes, injection velocities, and pellet compositions are compared. In the present DMS design, more than twenty flight tubes are prepared in three equatorial ports. Each pellet of diameter D = 28.5 mm and length over diameter ratio L/D = 2 can deliver up to $2\times 10^{24}$ hydrogen atoms. Because any pellets bigger than this size are not compatible with the planned pulse repetition rates, quantities > $2\times 10^{24}$ atoms must be delivered by ‘multiple injection.’Example simulation data are illustrated in figure 2(a) for 15 MA DT H-mode operation with thermal energy of 367 MJ (red) and for selected scenarios during the non-nuclear operation (blue). The SPI assimilation is shown in terms of the density increase averaged over the volume inside r/a < 0.8. The red curve shows close to 100 % assimilation for injection of up to 10 barrels as can be seen from the dashed line that represents the injected quantity divided by the plasma volume. We found that the pellet penetration depth increases with the injected quantities, and at least three pellets are needed to raise the density at the plasma center. Conversely, the blue curve, the efficiency for the non-nuclear scenario plasmas, saturates at smaller quantities because of lower thermal stored energy available for ablation and ionization of the injected material.
Figure 2(b) compares the assimilation with different pellet compositions, showing that a mixed amount of Ne < 5% of a single pellet can degrade the efficiency. No gain at all, or even a loss of the efficiency was observed for mixed pellets. Also regarding the aspect of the assimilation efficiency, our separate injection scheme of pure H$_2$/D$_2$ SPI is consistent with the injection capabilities of the present design.
Speaker: Dr Akinobu Matsuyama (National Institutes for Quantum and Radiological Science and Technology) -
The efficacy of ITER's DMS must be evaluated within the next several years and, since in-situ evaluation is impossible, verified and validated simulations are critical. High fidelity 3D initial value simulations of Shattered Pellet Injection and Dispersive Shell Pellet (SPI and DSP) simulations show favorable verification and validation in cross code comparisons and comparisons to experiment. These successes develop the necessary confidence in their predictive capabilities of the Disruption Mitigation System (DMS) in ITER. Furthermore, simulations allow a flexible test bed to optimize and refine existing strategies and to develop new ones. To perform these critical 3D nonlinear initial value simulations we have implemented a particle based SPI and DSP model in the NIMROD code coupled to the single fluid resistive MHD equations modified with impurities and radiation[VIzzo2006].
The table summarizes the favorable comparison of NIMROD SPI simulations of the impurity scan to the DIII-D experiments[CCKim2019]. Both thermal quench times and radiated fractions show good agreement and the same trends. Comparisons of benchmark cases using a stationary, axisymmetric, on-axis source to the M3D-C1 results[BLyons2019] show very good agreement. Complementary efforts by the 3D MHD code MARS and Fokker-Planck code CQL3D demonstrate progress on modeling runaway electrons (RE) and the current quench. Dispersive Shell Pellets have demonstrated the advantage of an ``inside-out'' thermal quench where the central core temperature drops first and heat flux is subsequently inward. DSP modeling also shows potential advantages for RE loss and radiated energy fraction.SPI simulations show that the ablating fragment drives strong parallel flows that transport the impurities along flux tubes and govern the thermal quench evolution. This parallel flow, caused by excess pressure (see figure) from ablation, is halted when the ``head bites the tail'', limiting the overall spreading of impurities, and accounting for the observed radiation asymmetry peaking near the injector. SPI simulations also show that as the thermal quench proceeds, the peak radiation lags behind the ablating fragment and peaks in the accumulated cold impurities that builds up in the wake of the fragment trajectory. Impurity scans of mixed deuterium/neon SPI pellets show a more benign thermal quench due to the enhanced transport and dilution cooling caused by the addition of deuterium suggesting optimal pellet mixtures exist. Multi-injector SPI simulations demonstrate significant suppression of the n=1 mode, the usual culprit for late thermal quench instabilities depicted in the figure at t=1.6ms. Symmetric dual-injector and tri-injector simulations show that suppressing the n=1 instability results in a significantly more benign thermal quench with reduced radiation asymmetry and greater radiation fraction. The observations from these simulations, greatly assisted by 3D visualization tools and animations of the nonlinear evolving fields, are exploited and synthesized and tested in additional scenarios combining multiple injectors, different mixture pellets, and the addition of Massive Gas Injection to reduce the radiation asymmetry and overall heat loads. The most promising will be submitted for experiments on DIII-D and other tokamaks.
DSP simulations of the pre-thermal quench (TQ) phase indicate that a non-perturbative shell is acheivable at quantities similar to the DIII-D experimental shell mass, which is the key to the high radiated energy fraction associated with DSP. Simulations carried through the end of the TQ show large amplitude MHD fluctuations ($\delta$B/B>$10^{-2}$) at the time of the plasma current spike--associated with current profile redistribution--which destroys all confinement by stochastizing fields and reducing field line connection lengths by at least three orders of magnitude. After the plasma current spike, which is of comparable amplitude to that measured in DIII-D experiments, no runaway electron tracer-particles remain confined. This result suggest that inside-out cooling may be particularly favorable to runaway electron de-confinement during the TQ, which is an unanticipated benefit of the concept. A large amplitude current spike is also observed in the stationary, axisymmetric, on-axis source benchmark cases with M3D-C1.
The generation and acceleration of runaway electrons poses an even greater uncertainty and threat than mitigation of the thermal load. Initial one-way coupling of time evolved thermal quench NIMROD fields to CQL3D demonstrates the sensitivity of runaway electron evolution to details of the thermal quench[RWHarvery2019]. These CQL3D-NIMROD simulations demonstrate that knock-on sources become more important in cases of slower, more realistic thermal quench. MARS-F modeling[YLiu2019] shows mitigation of a post-disruption, high-current runaway beam in DIII-D, due to the 1 kG level of magnetic field perturbation produced by a fast growing $n=1$ resistive kink instability. The RE loss is shown to be independent of the particle energy or the initial location of particles in the configuration space. Distributions of the lost REs to the DIII-D limiter surface show poloidally peaked profile near the high-field-side of the torus. Higher perturbation field level and/or higher particle energy also result in REs being lost to the low-field-side of the limiter surface. These complementary simulations will be used to compare and test a particle-in-cell runaway electron model implemented in NIMROD based on the hybrid kinetic-MHD energetic ion model. Ultimately, this hybrid kinetic-MHD RE model will provide comprehensive predictive 3D initial value simulations of the entire disruption mitigation scenario: from disruption trigger to thermal quench mitigation to current quench and runaway electron generation and mitigation, involving multiple injections and different strategies.
NIMROD DMS simulations of the Q=10 ITER baseline scenario show that many of the same characteristics are seen in ITER thermal quenches as those observed in DIII-D, particularly the dominance of an n=1 instability in the final thermal collapse. These simulations will be compared along side DIII-D and other (e.g. JET and KSTAR) tokamak DMS simulations and an initial assessment of the viability of the proposed DMS in ITER will be made.
This material is based upon work supported by US DOE-OFES, under Awards DE SC0018109 - Center for Tokamak Transient Simulations, DE-SC0016452 - SCREAM, DE-FG02-95ER54309 - Fusion Theory, DE-FC02-04ER54698 - DIII-D, and GA ITER Contract ITER/CT/14/4300001108.
Speaker: Charlson C. Kim (SLS2 Consulting/General Atomics) -
Plasma Exhaust and Plasma Wall Interaction are subjects of intense studies in fusion energy research for the understanding of the amount of heat loads and the lifetime of Plasma Facing Components. In order to ensure reliable predictive edge modeling in this context, it is mandatory to determine the transport properties of the Scrape Off Layer (SOL), a region largely influenced by the presence of turbulent filaments which contribute to particle and energy losses in both L and H modes. From the ITER divertor perspective, to keep the power fluxes acceptable for target material, high neutral pressure and partial detachment are needed to ensure maximum tolerable loads 1. Thus experimental investigation of SOL transport needs to be extended to these regimes.
Presently the regimes matching the ITER divertor conditions are obtained with high gas throughput and high density. In L-Mode these conditions are associated with a density shoulder i.e. progressive flattening of the density scrape off layer profile at high density [2-4]. It has been shown that density shoulder appear starting from high-recycling regimes and become broader after target density rollover [5], even though differences have been observed depending on divertor geometry [6], or if high recycling conditions are achieved through impurity seeding rather than high fueling [6,7]. The density shoulder is actually accompanied by an increase of the filamentary activity [4,5], with an increase of their associated heat and particle transport [4]. Preliminary investigations suggested that similar inter-ELM SOL density profile broadening is observed also in H-mode [4,5,8], with a stronger dependence on the neutral pressure 5. The possible increase of convective heat and particle fluxes to the wall poses serious issues in terms of acceptable sputtering yield of the first wall. In H-mode, with highly dissipative divertor, the plasma changes its stability moving towards a small-ELM regime [9] where a clear increase of the SOL density decay length is observed. Despite the large effort, a comprehensive understanding of the mechanism leading to an H-mode shoulder formation is presently lacking and this motivated a joint experimental program within the Eurofusion framework. The present contribution will show an unique comparison of H-Mode SOL density shoulder properties across 3 different devices, JET, ASDEX-Upgrade (AUG) and TCV focusing on the SOL profile evolution in different divertor recycling states, correlating the profile modification with different turbulent SOL plasma transport.On JET, 2MA/2.3T low $\delta$ plasma with 16 MW of applied NBI power were analyzed, with different levels of fueling exploring different divertor shapes in order to tackle the dependence of neutral compression as well [10]. On AUG 0.8MA/2.5T scenarios at different power levels (from 3 to 17 MW) and different fuelling schemes were analyzed in order to explore a wide range of divertor parameters and recycling states. Finally on TCV high-$\delta$ low current (0.18 MA) discharges were investigated with 1 MW of NBI heating with different fueling levels and locations. In all the devices we have identified conditions where inter-ELM density profiles exhibit a clear profile broadening as shown in figure 1.
To access the contribution of SOL turbulence in modifying the SOL profile, fluctuations in the main SOL and at the wall have been investigated as shown in
figure 2. On AUG, filaments velocities of inter-ELM filaments have been determined using the Thermal Helium Beam diagnostic and compared with the fluctuations observed in high-recycling state during the small-ELM regime. The comparison of the Probability Distribution Function (PDF) of these velocities is shown in figure 2 (b) and a clear increase of the filament velocity during high recycling state is observed. For TCV and JET we show the PDF of the ion saturation current density J$_s$ as measured at the wall respectively in panels (a) and (c) of figure 2.
In high density/high recycling state more skewed PDFs are observed for both the machines suggesting an increase of the fluctuation induced convective transport towards the first wall.
In high density/high recycling state more skewed PDFs are observed for both the machines suggesting an increase of the fluctuation induced convective transport towards the first wall. These experimental results form an excellent basis to benchmark SOL modelling under various conditions in differently sized machines,
providing a more complete characterization of the explored conditions in terms of divertor properties, upstream profiles, SOL fluctuation and induced transport and pedestal evolution. These observations consequently will strongly contribute to improve the understanding of SOL transport in conditions relevant fo the ITER divertor operation.
Acknowledgment
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014 - 2018 and 2019 - 2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.This work was supported in part by the US Department of Energy under Award Number DE-SC0010529.This work was supported in part by the Swiss National Science Foundation
References
1. Pitts, R et al., Nucl. Mat. Energy, 20, 100696 (2019)
2. Asakura, N et al, J. Nucl. Mater., 241-243, 559 (1997)
3. LaBombard, B et al, Phys. Plasmas 8, 2107 (2001)
4. Carralero,D et al, Nucl. Fusion, 57, 056044 (2017)
5. Vianello, N et al, Nucl. Fusion, 57, 056044 (2017)
6. Wynn, A et al, Nucl. Fusion,58, 056001 (2018)
7. Kuang, A et al, Nucl. Mater. Energy 19, 295 (2019)
8. Muller, H W et al, Journ. Nucl. Mater., 463, 739 (2015)
9. Harrer, G et al, Nucl. Fusion, 58, 112001 (2018)
10. Tamain, P et al, J. Nucl. Mater., 463, 450 (2015)†See the author lists of B. Labit et al 2019 Nucl. Fusion 59 086020, †† H. Meyer et al 2019 Nucl. Fusion 59 112014 and †††E. Joffrin et al. 2019 Nucl. Fusion 59 112021
Speaker: Nicola Vianello (Consorzio RFX) -
MHD stability at edge pedestal in QH-mode plasmas in DIII-D and JT-60U was analyzed by considering plasma rotation and ion diamagnetic drift effects. It was found for the first time the coupled rotation and ion diamagnetic drift effects can stabilize a kink/peeling mode in both experiments in case the rotation direction is counter to the plasma current direction although it has been recognized the rotation alone destabilizes the mode regardless of its direction. The stabilizing mechanism was identified as reduction of the destabilizing energy flow from rotation to the mode due to the coupling between rotation and ion diamagnetic drift. The coupling can be harnessed to both stabilize and destabilize the kink/peeling mode by switching the rotation direction. This trend could be the reason QH-mode plasmas in DIII-D and JT-60U favor toroidal rotation counter to the plasma current direction.
In H-mode regime in tokamaks, edge localized modes (ELMs) often appear and induce large heat load to divertors. Since the heat load is unacceptable for future large reactors like ITER and DEMO, it is necessary to suppress or mitigate the ELMs. Quiescent H-mode (QH-mode) is one of the promising candidates realizing ELM suppression and high confinement performance with reactor-relevant plasma parameters $^1$. One of the characteristics of QH-mode is the appearance of edge harmonics oscillations (EHOs) although ELMs disappear. Since QH-mode can be obtained experimentally when plasma current $I_p$ and sheared rotation counter to the $I_p$ direction (counter-$I_p$ direction) are large near plasma surface, current-driven MHD (kink/peeling) mode destabilized by rotation has been recognized as the trigger of the EHOs. In the plasmas, the ion diamagnetic drift frequency $\omega_{*i}$ can be comparable to the plasma rotation frequency $\Omega_{v \times B}$, hence, the E×B rotation frequency $\Omega_{E \times B} =\Omega_{v \times B} + \omega_{*i}$ has been considered as a strong candidate of the rotation responsible for QH-mode. However, it has not been identified numerically the importance of $\Omega_{E \times B}$ for realizing QH-mode even in nonlinear MHD simulations $^2$.
In this study, to identify the key mechanism explaining the role of rotation for triggering the EHOs, we analyze the impact of plasma rotation on kink/peeling mode stability in QH-mode plasmas in DIII-D and JT-60U by considering $\Omega_{v \times B}$ and $\omega_{*i}$ simultaneously with the linear extended MHD stability code MINERVA-DI $^3$.
The analyzed DIII-D plasma is #153440/1.725sec., and the JT-60U plasma is E42868/6.5sec.; in both plasmas,
the EHOs were observed with the dominant $n=2$ harmonic in DIII-D and $n=2$ and $4$ harmonics in JT-60U, where $n$ is the toroidal mode number. The $\Omega_{v \times B}$ profile is determined with the measured rotation of carbon $\Omega_{v \times B, C}$, where the toroidal component of this rotation is in the counter-$I_p$ direction. In both plasmas, when analyzing the MHD stability without the $\omega_{*i}$ effect, the stability boundary on the ($j_{ped,max}, \alpha_{max}$) stability diagram indicates the kink/peeling mode is slightly destabilized by $\Omega_{v \times B, C}$ rotation because the mode can be unstable with a small amount of $j_{ped,max}$ as shown in Fig. 1 (a) (DIII-D), where $j_{ped,max}$ is the maximum flux averaged current density at pedestal, and $\alpha_{max}$ is the maximum normalized pressure gradient. The trend is consistent with the previous results discussing the impact of sheared rotation on the ideal kink/peeling mode stability $^4$. However, when considering the $\omega_{*i}$ effect, the kink/peeling mode in rotating plasmas requires a larger amount of $j_{ped,max}$ for destabilization as shown in Figs. 1 (b) (DIII-D) and (c) (JT-60U); namely, rotation stabilizes the MHD mode in both plasmas. Note that the operation points exist near the stability boundary of the kink/peeling modes whose $n$ number is $2$ in DIII-D and $4$ in JT-60U when analyzing the stability with the rotation and $\omega_{*i}$ effects, and the $n$ numbers correspond to the dominant harmonics of the EHOs in both plasmas; the error bars on the point are determined to be $\pm20\%$ of $j_{ped,max}$ and $\alpha_{max}$.The physics mechanism stabilizing the kink/peeling mode by the coupled $\Omega_{v \times B, C}$ and $\omega_{*i}$ effects was investigated with the energetics established in $^3$. In ideal MHD case, an MHD mode can be destabilized due to energy flow from a fast-rotating equilibrium to a slow-rotating mode; this is called as dynamic pressure. When including the $\omega_{*i}$ effect, it was identified a part of the energy flow from an equilibrium to a mode is proportional to the product of $\Omega_{v \times B, C}$ and $\omega_{*i}$, and the modulation of dynamic pressure through the coupling between $\Omega_{v \times B, C}$ and $\omega_{*i}$ can stabilize the kink/peeling mode. The result also clarifies the coupling can be harnessed to both stabilize and destabilize the mode by switching the rotation direction, because the sign of the energy flow due to the coupling depends on the direction. In fact, when inverting the direction from $\Omega_{v \times B, C}$ to $-\Omega_{v \times B, C}$ in the DIII-D plasma, the stability boundary on the stability diagram moves downward as shown in Fig. 2.
Based on the result, it is possible to make a hypothesis explaining the reason why QH-mode discharges in DIII-D and JT-60U favor toroidal rotation in the counter-$I_p$ direction. A plasma can be assumed as axisymmetric before the transition to QH-mode, but when a kink/peeling mode becomes unstable, the plasma has non-axisymmetric distortion. Such distortion can accelerate plasma rotation in the counter-$I_p$ direction near plasma surface, the rotation which is induced by neoclassical toroidal viscosity as observed in DIII-D $^5$. The kink/peeling mode will stop growing due to the stabilization by the rotation, and as the result, the plasma with non-axisymmetric distortion will be sustained to keep the rotation, the distortion which could be observed as EHOs. In this case, the operation point will be between the stability boundaries with and without rotation effect. The results in Figs. 1 (b) and (c) show the condition can be satisfied, because the error bar of the point is in the area.
$^1$ K.H. Burrell et al., Phys. Rev. Lett. 102,155003 (2009).
$^2$ F. Liu et al., Nucl. Fusion 55, 113002 (2015).
$^3$ N. Aiba, Plasma Phys. Control. Fusion 58, 045020 (2016).
$^4$ X. Chen et al., Nucl. Fusion 56, 076011 (2016).
$^5$ A.M. Garofalo et al., Phys. Rev. Lett. 101,195005 (2008).Speaker: Dr NOBUYUKI AIBA (National Institutes for Quantum and Radiological Science and Technology) -
ITER Council has decided to reduce the number of port from three to two for testing TBM due to the introduction of a Disruption Mitigation System during November 2018. Since then, India is proposing Helium Cooled Solid Breeder (HCSB) concept to be tested in one half of the two ports available for the same in ITER. Indian HCSB concept is having Reduced Activation Ferritic Martensitic Steel (RAFMS) as the structural material and Lithium Meta Titanate (Li2TiO3) and beryllium as the tritium breeder material and neutron multiplier respectively. Based on the neutronic analysis which ensures the adequate tritium-breeding ratio (TBR), the design of the edge-on configuration of HCSB blanket has been optimized by doing extensive thermal–hydraulic, CFD analyses thermo-mechanical analyses including high pressure coolant loads, electromagnetic, seismic load conditions in compliance with ITER safety requirements and guidelines 1. Some the analysis results of the optimized HCSB blanket are shown in fig. 1. Transient thermal analysis has also been simulated for the ITER transients. Results obtained from steady state and transients show that the maximum temperature of all materials are well within their limiting values. Various option of breeder modules like flaring in radial direction, comparisons of radial-toroidal and poloidal-toroidal modules and their advantages have been studied in detail in our earlier works 2
The Indian blanket related R&D program is focused on the development and characterization of different functional (Li2TiO3, Be, PbLi etc.) and structural materials, tritium technologies and fabrication technologies for the blanket module. The results from these R&D activities are being used for the design and development of various subsystems of HCSB blanket module. The spherical form (pebble) of ceramic breeder material (Li2TiO3) and beryllium multiplier are kept as a packed bed in the breeder module. The fabrication process of these materials will have the challenge to meet the demands of having suitable compositions, properties and their functional requirements. At every stage of the preparation, various characterizations on pebble and pebble bed are performed to meet the desired properties required for fusion blanket. Density and temperature dependence thermal conductivity shows the influence of porosity on the thermo-mechanical properties (crush strength and thermal conductivity) of Li2TiO3 pebbles. A test facilities for the measurement of effective thermal conductivity (Keff) of the pebble bed using steady-state Axial and transient hot-wire method have been designed and built indigenously at IPR. Fig. 2 shows the measured Keff value of Li2TiO3 pebble bed under Helium gas and compared with available literature results [4].
The Indian blanket related R&D program is focused on the development and characterization of different functional (Li2TiO3, Be, PbLi etc.) and structural materials, tritium technologies and fabrication technologies for the blanket module. The results from these R&D activities are being used for the design and development of various subsystems of HCSB blanket module. The spherical form (pebble) of ceramic breeder material (Li2TiO3) and beryllium multiplier are kept as a packed bed in the breeder module. The fabrication process of these materials will have the challenge to meet the demands of having suitable compositions, properties and their functional requirements. At every stage of the preparation, various characterizations on pebble and pebble bed are performed to meet the desired properties required for fusion blanket. Density and temperature dependence thermal conductivity shows the influence of porosity on the thermo-mechanical properties (crush strength and thermal conductivity) of Li2TiO3 pebbles. Test facilities for the measurement of effective thermal conductivity (Keff) of the pebble bed using steady-state Axial and transient hot-wire method have been designed and built indigenously at IPR. Fig. 2 shows the measured Keff value of Li2TiO3 pebble bed under Helium gas and compared with available literature results 3. India specific RAFM steels (IN-RAFMS) has been produced for the structural material of HCSB blanket. The final composition of IN-RAFM steel has been decided with varying tungsten and tantalum content. The 9Cr-1.4W-0.06Ta wt% RAFMS is found to show better impact properties than other RAFM steel heats [4]. Various thermo-physical and thermo-mechanical properties of IN- RAFMS material have been studied for the generation of material qualification database as per ITER requirement. Further R&D needs to be done for the material properties database up to the radiation damage expected in ITER (> 1 dpa).
A FW mock-up having 10 cooling channels as shown in fig. 3has been successfully fabricated using modified 9Cr-1Mo steel which has similar properties of RAFM and successfully tested in high heat flux (HHF) facility using high temperature, high pressure helium loop (HELOKA) at KIT Germany. The mock-up has been tested for both normal and accidental operating conditions of ITER surface heat fluxes. The experimental results of maximum and average front surface temperature of the mock-up shows a good agreement with the FE simulation within the variation of ±2%, which is within the uncertainty margin except for some local hotspots [5]. The FW temperature for LOFA condition for 25% flow and for NO flow conditions are also within the allowable margin of deviation. Similar to HELOKA loop, the installation of an Experimental Helium Cooling Loop (EHCL) is in progress at IPR.
In summary, over the last few years, the joint team comprising of IPR and other Indian institutes has made significant progress in development of materials, fabrication technology, conceptual design, engineering design and various R&D areas relevant to TBM system development. This paper will presents the design optimization and analyses of HCSB blanket and current status of different R&D developments which provides a roadmap for the future activities planned in realizing Indian HCSB TBM for testing in ITER.
Key words: HCSB, Blanket, thermal-hydraulics, Li2TiO3, pebble bed
References:
1 D. Sharma et. al., “Design and analysis of manifolds for Indian HCCB blanket module”, Fusion Engineering and Design 129, (2018), 40–57.
2 Paritosh Chaudhuri et. al., “Comparative studies for two different orientations of pebble bed in an HCCB blanket”, Plasma Sci. Technol. 19 (2017) 125604.
3 M. Panchal et. al., “Measurement of effective thermal conductivity of Li2TiO3 pebble bed by transient hot-wire technique”, Communicated in Fusion Engineering and Design.
[4] A N Mistry et. al., "Status of India-specific Reduced Activation Ferritic-Martensitic steel and fabrication technologies development for LLCB TBM", Fusion Engineering and Design 125, (2017) 263-268
[5] S. Ranjithkumar et. al., “Performance assessment of the Helium cooled First Wall mock-up in HELOKA facility”, Fusion Engineering and Design 150 (2020), 111319.
Speaker: Dr Paritosh Chaudhuri (Institute for Plasma Research) -
Perhaps the most attractive fusion reactor concept is the stellarator since it has minimal recycling power, minimal auxiliary systems and no time dependent electro-magnet systems. However, progress has been delayed by two formidable challenges: obtaining sufficient
confinement in three dimensional fields and engineering the magnetic configuration with sufficient precision at low cost. Stellarators have experienced a rebirth in recent years due to fundamental breakthroughs in the calculation and optimization of the confinement properties of three-dimensional magnetic systems. Optimized stellarator experiments have shown conclusively that the neoclassical ion-confinement has been raised to values similar to tokamaks (1) - thus the first challenge is beginning to be met and further optimization is in development (2). However along with these improved physics-driven design criteria came substantial magnet complexity and precision requirements that are extremely demanding ($\sim10^{-3}$) (3). Thus, the second challenge, magnet simplification has been identified as a critical research need for stellarators in a recent report (4).
Recently, a new concept using permanent magnets and simple planar coils for making the complex fields required by stellarators was proposed [5,6]. Calculations show that permanent magnets can broaden the space of achievable stellarator configurations, reduce construction costs and increase availability. Thus ultimately, they can reduce the cost of fusion electricity.
The technical goal of the current activity is to develop the technology required to achieve stellarator fields using permanent magnets and to verify that the fields meet accuracy requirements. The assembled magnetic field structures will be verified with measurements made using Hall probes. An important feature of the design will be a system of adjustments that enable tuning of the magnet positions to minimize error fields within tolerance. The permanent magnet assembly will be designed so that it can be a part of a planned but as yet unfunded future stellarator that will re-use some components of the NCSX device.
In addition to the reduced assembly cost, the use of demountable permanent magnets will also provide a viable sector maintenance scheme that will increase the availability of the reactor. This reduces the cost of electricity by reducing down time. Additionally, one of the barriers to stellarator innovation and research is the cost of high complexity which has been prohibitive for smaller research programs. Reduction in complexity will lower the cost barrier enabling the pursuit of novel physics optimization strategies. Furthermore, we have already shown that designs that were unattainable with coils can be easily achieved with permanent magnets.
A method of designing practical magnets that can be mounted to a relatively simple cylindrical mount with radial ribs was devised for the preconceptual design. The computational methods described above are then used to design an array of trapezoidal pyramid shaped magnets. It is envisioned that these magnets will be contained within stainless steel boxes the back face of which will be the interface to a radial rib that acts as the mounting point for the magnet. The magnets will be held to the rib with screws. The radial ribs will in turn be mounted to a central support cylinder.
The design activities on this concept have just begun and we will continue to investigate addition mounting concepts and magnet distributions. The details of this design effort as well as the additional design options under consideration will be presented.(1) “Overview of first Wendelstein 7-X high-performance operation” T. Klinger, et al., Nucl. Fusion 59 (2019) 112004
(2) See the Hidden Symmetries activity website at: https://hiddensymmetries.princeton.edu
(3) “Engineering cost & schedule lessons learned on NCSX” R. Strykowsky, et al., 2009 IEEE/NPSS Symposium on Fusion Engineering, DOI: 10.1109/FUSION.2009.5226449
(4) “Stellarator Research Opportunities: A Report of the National Stellarator Coordinating Committee”, D. A. Gates, et al., Journal of Fusion Energy 37 (2018) 51 https://www.osti.gov/biblio/1414416-stellarator-research-opportunities-report-national-stellarator-coordinating-committee
(5) M. Zarnstorff, S. Cowley, and C. Forest, “Simple method to create 3D and poloidal magnetic fields by permanent magnets for efficient steady-state plasma confinement”, Provisional Patent, 2019.
(6) “Stellarators with permanent magnets” P. Helander, et al., Accepted for publication in Physical Review Letters January 2020, https://arxiv.org/abs/1907.01363 , https://journals.aps.org/prl/accepted/4d070YdbT431f67083cc6383685fd810890bedb52Speaker: Dr David Gates (PPPL) -
In Wendelstein 7-X, the vacuum rotational transform, $\bar\iota$, has a rather small shear and does not cross any major rational surfaces. Nevertheless, during plasma operation it can be modified by electron cyclotron current drive (ECCD) in such a way that the resulting iota profile passes through low-order rational values, potentially triggering magnetohydrodynamic (MHD) events.
Indeed, W7-X plasmas are sometimes subject to repetitive collapses of core confinement, which were observed during co- and counter central ECCD [1]. These phenomena are periodic, rapid “crashes” of the electron temperature reminiscent of tokamak “sawtooth” instabilities. In some discharges these events lead to a complete termination of the entire plasma on the time scale of a few ms. These large crashes are usually accompanied by a noticeable (~0.6kA) current jump in the total plasma current. Even though the origin of these MHD instabilities is not yet clear, the fast crash is likely to involve the formation of magnetic islands and magnetic reconnection.
Since the deposition profile of ECCD is usually very localized, it results in a strong distortion of the rotational transform $\bar\iota$. For co-ECCD (defined as locally increasing $\bar\iota$), it can cause $\bar\iota$ to pass through unity, which is detrimental for MHD stability. Experimentally the $\bar\iota$-profile is rather uncertain, therefore information about the profile form has to be inferred from simulations.
To calculate the pre-crash $\bar\iota$-profile (Fig.1, left plot, green line), the plasma current evolution has been simulated from the poloidal flux diffusion equation using NTSS code [2]. The temperatures, the bootstrap current density, and the parallel electric conductivity have been modeled self-consistently
with the electron cyclotron resonance heating (ECRH). The ECCD current density has been calculated with the Travis ray-tracing code [3] taking into account the real launching geometry, beam size and plasma density and temperature profiles. The total plasma current evolves slowly due to the compensation of the ECCD current by an inductive plasma current. But the current density profile
variation produces a change of the $\bar\iota$ profile, which is a superposition of the ”vacuum” $\bar\iota$, generated by the external stellarator coils, and the $\iota$ generated by the internal toroidal current. An $\bar\iota$ peak develops
in the core and crosses the $\bar\iota$ = 1 resonance (see Fig.1, left plot, green line). Note that the rotational transform passes through unity twice. Under such conditions, linear calculations suggest that the plasma is ideally stable but unstable to resistive MHD modes.The main question addressed in this work is what happens to the plasma configuration as a result of this instability. This question could be answered by nonlinear resistive MHD simulations. However, there is no code capable of such a simulation in the geometry of W7-X. In this work, we use a model based on Taylor relaxation [4, 5] to predict the nonlinear redistribution of the plasma current caused by these events.
The key theoretical assumption is that, in the crash, the plasma will seek to minimize its magnetic energy while keeping magnetic helicity and toroidal flux fixed. This assumption leads to the prediction of a unique plasma state, where the current flows in the direction of the magnetic field and the current density is proportional to the field strength:
$$ \nabla \times {\bf B} = \mu {\bf B} \ \ \ \ (1) $$ where $\mu$ is a constant. We suggest that the nonlinear result of resistive instability in W7-X may be such a Taylor-relaxed state. It is not easy to numerically construct equilibria with prescribed helicity. We therefore scan the values of $\mu$ and solve Eq. (1) numerically to obtain a range of possible post-relaxation states. Given those, we select the one whose helicity equals the pre-quench value. The pre-quench toroidal magnetic flux and the plasma boundary shape are also assumed to be conserved during the relaxation.
To calculate the pre-crash equilibrium we use VMEC equilibrium code [6]. For the calculations of possible relaxed states, the SPEC (Stepped Pressure Equilibrium Code) code [7] is used. SPEC can find minimal-plasma-energy states, subject to the different constraints, in nested sub-volumes, by extremizing the energy functional. Relaxation and magnetic reconnection is allowed in each volume.
Our results are in agreement with the experimentally observed current «jumps» during large crashes in W7-X. Furthermore, we cross-check our results with edge diagnostic, e.g. infrared (IR) thermographic systems. We show that a sequence of several crashes (Fig.1, right plot) at the plasma edge, i.e. increase of the toroidal current, and a change in the island divertor geometry, resulting in a move of the divertor strike lines.
References:
[1] Wolf et al., Physics of Plasmas 26, 082504 (2019).
[2] Turkin Y. et al., Physics of Plasmas 18 (2), 022505 (2011).
[3] Marushchenko N. et al., Computer Physics Communications 185 (1), 165–176 (2014).
[4] Taylor J.B., Phys. Rev. Lett. 33, 1139 (1974).
[5] Taylor J.B., Rev. Mod. Phys. 58, 741 (1986).
[6] Hirshman, S., Merkel, P. et al., Computer Physics Communications 43 (1), 143–155 (1986).
[7] Hudson, S., et al., Physics of Plasmas 19 (11), 112502 (2012).Speaker: Ksenia Aleynikova (Max-Planck Institut für Plasmaphysik, Greifswald, Germany) -
The termination of high performance plasmas in tokamak devices with high Z metal plasma facing components presents challenges related to the influx of heavy impurities which, if not kept under control, cause an increase of the radiative losses, radiative cooling and high probability of disruption.
A number of key players in these dynamics have been identified by intensive research performed after the first years of operation in tungsten machines as AUG and JET in preparation of ITER operation. Inward neoclassical convection related to the peaking of the density profile, poloidal asymmetries, plasma rotation and centrifugal effects, temperature screening, pedestal temperature, pedestal density and ELMs control are among them [1-3].
The objective of D-T fuelled plasmas with high neutron yield in stationary conditions, foreseen in the near future at JET, focuses the operations towards high performance in terms of thermal energy content and plasma current and consequently with higher disruption risk. The reduction of such risks is being pursued for the specific features of the two plasma scenarios being developed, baseline (β¬N ~1.8, q95 ¬~ 3) and hybrid (β¬N ~2–3, q95~4) [4]. The high plasma current (Ip≥2.5MA) experiments based on the baseline scenario performed in the high power campaign C36b (2016) had 65% overall disruptivity (Ip,disr≥1.0MA) with 49% pulses ending with a disruption at Ip,disr≥2.0MA. The high plasma current (Ip≥2.MA) experiments based on hybrid scenario of the same campaign had 39% overall disruptivity with 21% of the pulses ending with a disruption at Ip,disr≥2.0MA. The inspection of the corresponding databases for the recent C38 campaign (2019) reveals a significant reduction of the disruption rate. The overall disruptivity for baseline has been 34% with 27% at Ip,disr≥2.0MA. For hybrid the overall disruptivity has been ~9% (16% considering only the Ip≥2.MA database) with 5% at Ip,disr≥2.0MA. Such numbers are better understood by taking into account the necessary explorative nature of scenario development, and, in particular, the progressive adaptation towards the target parameters of higher plasma current, 3.5-4 MA for baseline and 2.2-2.8 MA at flat top for hybrid. The key elements developed to obtain a smooth termination of the high performances discharges in JET and their effectiveness in reducing the disruptivity are discussed in this contribution.
The analysis of the previous experimental campaign and the data so far collected in the present campaign indicate that the combination of edge and core W control is needed to obtain a safe plasma termination, with the optimized use of the available actuators: gas and pellet for ELMs control, ramp-down waveform of the NBI heating power while maintaining a relevant ICRH additional power, sweeping of the separatrix hitting point on the divertor to reduce the heat load and to decrease the W source. These elements, common for baseline and hybrids although with a different tuning [5], have been tested in the C38 campaign and some of them are represented in the pair of discharges in Fig.(1) where an impurity influx has been deliberately generated by drastically reducing the gas fuelling during the main heating phase.
With respect to the typical termination scheme previously adopted the gas fuelling has been increased to burst the ELM frequency in order to favor impurities removal at the plasma edge and to reduce the divertor temperature. The slow NBI power ramp-down delays the H-L transition. This needs to be tuned to shorten the free-ELMs phase at the ramp-down starting. At high current and high density the ICRH central heating may be inefficient to counteract the core impurity effect being dominantly ion heating. However, a significant core electron heating is obtained by increasing the H minority fraction during the ramp-down as shown in the figure, where the central electron temperature is recovered and the disruption avoided.
A further adaptation of the termination scheme may be required for the application to the higher plasma current D-T scenarios in development. Moreover, since the termination time cannot be foreseen in case of off-normal events, real-time tools are being implemented in a dedicate event detector to identify the nature of the off normal condition, e.g. core impurity accumulation or edge temperature collapse, and to start the termination phase with a real-time controlled response. The new detectors include real-time tomography which can estimate the amount of radiated power from different region of interest [6] and a Generative Topographic Mapping algorithm [7] aiming to compute the probability of disruptive evolution for core and edge radiative collapse respectively. A termination algorithm aiming to optimize the input power waveform during the ramp-down in order to keep a safety margin to overcome the radiative losses both in H and L mode is also being planned.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014–2018 and 2019–2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
- C. Angioni, et al. et al 2014 Nucl. Fusion 54 083028
- F Köchl et al 2018 Plasma Phys. Control. Fusion 60 074008
- E. de la Luna et al. 2018 27th IAEA Fusion Energy Conference – IAEA CN-258, EX/2-1
- L. Garzotti et al 2019 Nucl. Fusion 59 076037
- J. Hobirk et al 2018 Nucl. Fusion 58 076027
- D.D. Carvalho et al 2019 JINST 14 C09011
- A. Pau et al 2019 Nucl. Fusion 59 106017
Speaker: Carlo Sozzi (Istituto per la Scienza e Tecnologia dei Plasmi ISTP-CNR Milano Italy) -
We present a new model of the sawtooth oscillation that can explain direct measurements which show $q_0$ well below 1 during the entire cycle, and observations that indicate the $q=1$ surface is not removed by the sawtooth crash.
In this model the crash is caused by an ideal instability that occurs when $q_0=2/3$.
This instability changes the topology of the nested magnetic surfaces around the axis into that of an alternating-hyperbolic point.
This stochastisizes a region around the axis that resets $q_0$ to a value above $2/3$ but below 1.
Understanding and controlling this internal disruption is important for future tokamaks such as ITER.The leading sawtooth models are by Kadomtsev (1) and Wesson (2).
Both these models predict that the crash occurs when $q_0\sim1$, through the action of the ideal $1/1$ internal kink mode.
This process removes the $q=1$ surface and resets $q_0$ to 1.
This is confirmed by direct measurements where the reset of the central safety factor is observed (3).
However, the internal kink can be stabilized in many ways (4), and the effects of this stabilization on the $1/1$ sawtooth is seen in experiments (5, 6).Furthermore, there are several observations that are irreconcilable with the Wesson and Kadomtsev models.
The safety factor was directly measured on TEXTOR using Faraday rotation imaging (7), and on TFTR using MSE (8, 9).
These two different experiments, performed using two different methods, on two different tokamaks gave the same result: The crash occured when $q_0=0.7\pm0.1$, and stayed well below 1 for the entire cycle.
Other measurements indicate that the $q=1$ surface is not always removed: The snakes (10,11) observed on JET, which are associated with islands on the $1/1$ surface, can survive multiple sawtooth crashes.
A close analysis of EAE activity additionally reveals shots where the EAEs associated with the $q=1$ surface are not removed by the sawtooth crash (12).We show that every configuration of the magnetic field around a closed field line (for example the magnetic axis) corresponds with an element $\mathsf{M}$ of the simple Lie group $\mathrm{SL}(2,\mathbb{R})$, the group of $2\times2$ real valued matrices with unit determinant.
This group contains distinct subsets, classified by their trace, in which the topology of the magnetic surfaces differ.
These subsets comprise the O-points ($|\mathrm{Tr}(\mathsf{M})|<2$), the X-points ($\mathrm{Tr}(\mathsf{M})>+2$), points on an intact $q=1$ surface ($\mathrm{Tr}(\mathsf{M})=+2$), and \emph{alternating-hyperbolic} points ($\mathrm{Tr}(\mathsf{M})<-2$).
An alternating-hyperbolic point is surrounded by hyperbolic surfaces just as an X-point, but the field line map includes reflection in the fixed point.During a discharge the magnetic field changes continuously, and so does the matrix $\mathsf{M}$ associated with a closed field line.
The matrix $\mathsf{M}$ can transition to a different subgroup of $\mathrm{SL}(2,\mathbb{R})$, where the field changes topology, when $\mathrm{Tr}(\mathsf{M})=\pm2$.
The traditional $1/1$ sawtooth involves such changes at the intact $q=1$ surface where $\mathrm{Tr}(\mathsf{M})=2$, which changes into an X-point and an O-point.
The O-point of the 1/1 island becomes the new axis with $q_0=1$.
A transition of a different nature occurs when $\mathrm{Tr}(\mathsf{M})=-2$, which is the case when $q_0=2/3$.
Here the magnetic axis can change into an alternating-hyperbolic point.When we investigate the stability of ideal MHD equilibria with $q_0=2/3$ using NOVA-K, we find an ideal $2/3$ mode localized on the axis.
This instability occurs slightly before $q_0$ reaches $2/3$, and has a higher growth rate than the $1/1$ internal kink for similar values of plasma $\beta$.
The growth rate and mode profile are shown in figure 1.The perturbed field associated with this mode is calculated from $\delta \mathbf{B} = \nabla\times\left(\mathbf{\xi}\times\mathbf{B_0}\right)$.
This perturbation causes an immediate change in topology of the magnetic surfaces around the axis into the alternating-hyperbolic configuration.
At high amplitudes this mode stochastisizes a region around the axis, and breaks the $q=1$ surface into a $3/3$ island chain.
A poincar\'e plot of the stochastisized field is shown in figure 1.This leads to the alternating-hyperbolic model for the sawtooth:
The internal kink is stabilized until $q_0=2/3$.
The unstable $2/3$ mode is triggered and the axis is changed into an alternating-hyperbolic fixed point, which stochastisizes the core region.
The magnetic connectivity in the stochastic region can then drive rapid reconnection, which would redistribute poloidal and toroidal fluxes and increase $q_0$.
The cycle repeats, until through slow current diffusion $q_0$ reaches $2/3$ again.This model predicts the crash well within the uncertainty of the measured value of $q=0.7\pm0.1$ (7, 8, 9), it predicts $q_0$ below 1 for the entire cycle, and it does not remove the $q=1$ surface, explaining the persistence of snakes and EAEs.
(1) B. B. Kadomtsev, Soviet Journal of Plasma Physics 1, 710 (1975).
(2) J. A. Wesson, Plasma physics and controlled fusion 28, 243 (1986).
(3) Y. B. Nam, J. S. Ko, G. H. Choe, Y. Bae, M. J. Choi, W. Lee, G. S. Yun, S. Jardin, and H. Park, Nuclear Fusion 58, 066009 (2018).
(4) I. T. Chapman, S. D. Pinches, J. P. Graves, R. J. Akers, L. C. Appel, R. V. Budny, S. Coda, N. J. Conway, M. De Bock, L. G. Eriksson, et al., Plasma Physics and Controlled Fusion 49, B385 (2007).
(5) C. Angioni, A. Pochelon, N. N. Gorelenkov, K. G. McClements, O. Sauter, R. V. Budny, P. C. De Vries, D. F. Howell, M. Mantsinen, M. F. F. Nave, et al., Plasma physics and controlled fusion 44, 205 (2002).
(6) D. J. Campbell, D. F. H. Start, J. A. Wesson, D. V. Bartlett, V. P. Bhatnagar, M. Bures, J. G. Cordey, G. A. Cottrell, P. A. Dupperex, A. W. Edwards, et al., Physical review letters 60, 2148 (1988).
(7) H. Soltwisch and H. Koslowski, Plasma physics and controlled fusion 37, 667 (1995).
(8) F. M. Levinton, S. H. Batha, M. Yamada, and M. Zarnstorff, Physics of Fluids B: Plasma Physics 5, 2554 (1993).
(9) M. Yamada, F. Levinton, N. Pomphrey, R. Budny, J. Manickam, and Y. Nagayama, Physics of plasmas 1, 3269 (1994).
(10) A. Weller, A. D. Cheetham, A. W. Edwards, R. D. Gill, A. Gondhalekar, R. S. Granetz, J. Snipes, and J. A. Wesson, Physical review letters 59, 2303 (1987).
(11) R. D. Gill, A. W. Edwards, D. Pasini, and A. Weller, Nuclear Fusion 32, 723 (1992).
(12) G. J. Kramer, C. Z. Cheng, Y. Kusama, R. Nazikian, S. Takeji, and K. Tobita, Nuclear fusion 41, 1135 (2001).Speaker: Christopher Smiet (Princeton Plasma Physics Laboratory) -
Jiming Chen1*, Pinghuai Wang1,Kun Wang2, Liman Bao3, Xiaobo Zhu1, Jing Wu1, Stefan Gicquel3, Qian Li1, Hui Gao1, Jialin Li1, Rene Raffray3, Ming Xu1, Xuru Duan1
1Southwestern Institute of Physics, Chengdu 610041, China
2China Int’l Nuclear Fusion Energy Program Execution Center, Beijing 100862, China
3ITER Organization, 13115 St Paul Lez Durance, France*E-mail: chenjm@swip.ac.cn, Tel: +86-28-82850415
The design of the ITER Enhanced Heat Flux (EHF) First Wall (FW) panels has been updated to improve repairability and manufacturabilityin their manufacturing phase. R&Ds to validate the design are conducted. The 40 plasma-facing fingers are now paired and assembled to the 316L(N) steel central beam with accessible pipe connections, which makes the welds of the pipes repairable and the fingers replaceable. The number of weldsis largely reduced for lower water leak risk and better overall deformation control, and all of them are no longer inaccessiblefor non-destructive tests. Analysis shows that the designis in good conformances with the SDC-IC criteria. Optimized welding and repairing technologies were obtained with thick pipe up to 2.5mm in a narrow welding space with stable welding shrinkage. While the assembly of fingers into pairs are well demonstrated with reliable welding quality, twice repairability and tolerable dimension control, that of the actual full-scale finger pairs assembling to the central beam is under investigated. A full-scale EHF FW protype will be manufactured for production process qualification. The hot helium leak test facility and a proposing ISO testing standard have been well established with leak rate sensitivity down to 10-11 Pa.m3/s at elevated operational temperature and pressure for its factory acceptance test. All of the technologies and examination results will be presented in this paper.
R&Ds on pipe connection welding are conducted between the fingers and fingers to the central beam using orbit welding. To accommodation the narrow space around the welds,one of the smallest welding heads was used. The R&Ds are aimed at obtaining the maximum pipe thickness at which both welding and repairing could reach full-penetration and meet criteria of ISO5817 level B. Various thickness pipes from 1.6 to 3mm were welded and examined.Results showed the most feasible one is 2mm from the engineering point of view.To confirming the performance, visual inspection, penetrant testing, radiographic testing, transverse tensile testing, transverse bend testing, metallographic examination and N-H-O content measurementwerecarried out. All satisfies the standard and the ITER requirements.Bothrepairing tests (by discard 3~6 mm across weld) and fitting tolerance tests were performed to provide guidelines for further finger assembly. Figure 1(a) shows welding shrinkage in normal welding and repairing for some test pipesin 2mm thickness. It is very stable on average for the 11 joints, showing this deformation can be well controlled in manufacturing the products. Fake 316L(N) fingers were paired by such kind of joint, showinggood geometry and position appearances as that in Figure 1 (b). The assembly of the finger pairs to the central beam is ongoing with the L-shape connection pipes in the same thickness. The actual FW fingers with Be/CuCrZr joints are under manufacturing. With such R&D results, a full-scale EHF FW prototype (FSP) will be manufactured.
An EHF FW panel consists of 32 standard fingers and 8 edge fingers. Each finger is an joint of Be tiles on the CuCrZr/316L(N) bimetallic material heat sink, in which a hypervapotron (HVT) cooling channel was utilized to enhance heat transfer of 2m/s flowing water with the solid wall. Based on previous R&Ds [1.], a 4mm-depth bottom grove is added to the HVT to make one order longer fatigue life. The bimetallic plate is made by explosion bonding [2.] with qualitycontrol by ultrasonic test to ensure no defect at the interface. Diffusion bonding Be/CuCrZr has been qualified [1.] for 12×12mm2 Be tiles in the previous semi-prototype program. To simplify the manufacturing, larger Be tiles up to 18×18mm2 is planned for the FSP. Small Mock-ups with such Be tiles were made by the same technologies, thermal fatigue test for validating its performances will be performed at 4.7 MW/m2 surface heat load for 15000 cycles. Unlike the toroidally profiled standard fingers, the edge fingers are double shaped in both toroidal and poloidal direction with irregular HVT cooling channels. This will make the Be surface having rather higher temperature at the maximum surface heat load for the edge fingers. To lessen its effects, small Be tiles as before is preferred to reduce thermal stress at the interface and in the heat sink. Further high heat flux test will be performed to evaluate its thermal performance. Elaborative design of the EHF FW panels will be done again based on the FSP experience.
The ITER FW panels are in-vessel components of vacuum class 1A. It is required that each panel shall be in good vacuum tightness with its global leak rate less than 1×10-10 Pa.m3/s in air equivalent at the 250℃operation temperature and 4 MPa water coolant pressure. To perform the test, a hot helium leak test facility was established in China, which has a 3 m3 vacuum chamber and a leak rate sensitivity around 1×10-11 Pa.m3/s. An EHF FW semi-prototype was tested in this facility, and results showed the leak rate could meet the requirement. The key is that the component shall be backed at 250~300℃ for more than 8 h for degassing. In addition, it is found the leak rate is vacuum level dependent (see Figure 2). To reach the sensitivity, the vacuum level shall be as low as 10-5 Pa. An ISO standard was proposed to conform the tests all over the world for fusion reactor components.
References
[1.] J.M. Chen, X. Liu, P.H. Wang, et al, Progress in developing ITER and DEMO first wall technologies at SWIP, Nucl. Fusion 60 (2020) 016005 (8pp).
[2.] Pinghuai Wang, Jiming Chen, Qian Li, et al, Study on microstructure and properties evolution of CuCrZr/316LN-IG explosion bonding for ITER first wall components, Fusion Eng. Des. 124 (2017) 1135-1139.Speaker: Jiming Chen (Southwestern Institute of Physics) -
It is an extreme challenge to reduce the transient peak of the heat load on the plasma facing components (PFC) in tokamak plasma$[$1$]$. One effective way is to increase the wetted area on the divertor target by splitting the strike point. Most of the experimental results show that the change of magnetic topology induced by RMP and LHCD is responsible for the strike point splitting$[$2-4$]$. However the JET experiment reported the observation of double peaks of particle flux on the divertor target without the variation of magnetic topology$[$5$]$, and the simulation work pointed out that the radial E×B drift could be a candidate for interpreting such change$[$6$]$. Recently, double strike points of the divertor particle flux has been observed with the upgraded divertor Langmuir probe arrays$[$7$]$ in the HL-2A ECRH plasmas, which does not change the magnetic topology as discussed in the previous reports$[$2-4$]$. Thus, it is necessary to study the strike point splitting mechanism with different plasma conditions. Here we firstly report the double strike points of divertor particle flux in HL-2A ECRH plasmas, and then present the simulation with SOLPS compared to the experimental observation.
HL-2A is a medium size tokamak with closed divertor, which is routinely operated under the lower single null (LSN) configuration with the clockwise direction of toroidal magnetic field B$_t$ and anticlockwise direction of the plasma current I$_p$. Figure 1 shows the typical evolutions of the parameters in L-mode with ECRH. As we can see, the particle flux peak increases clearly on both the outer (Figure 1 (c)) and inner (Figure 1 (d)) divertor target when the ECRH is turn on, and about two milliseconds later, the second peak appears gradually on the outer divertor target. In order to observe the evolution process clearly, the profiles of particle flux are displayed in Figure 1(c1) with the curves in different colors, and each curve represents the mean values of a period of 3ms indicated by the arrows. There is only a single peak at Z=-79.1cm before 309ms, and the secondary peak appears at Z=-76.7cm from 309ms to 312ms as shown by the black curve, which is followed an obvious increases after 312ms. The evolution shows there is physical process during the formation of the double strike points, which is relevant to the ECRH. However, the strike point splitting is not observed on the inner divertor target, and there is only a slight upward movement of the single peak. Moreover, the fueling also has an obvious effect on the distribution of the particle flux. The particle flux is increased due to the second fueling pulse at about 333ms on the outer divertor target, but the double-peak structure is weakened due to the fueling.
The preliminary simulation of SOLPS shown in Figure 2, consistent with the experimental observation, indicates the critical role of the poloidal E×B drift in exciting the double strike points. The simulation results show that there is only a single peak for the particle flux, while the double strike points are produced when the E×B drift is included, as shown by the black and red curves, respectively. The formation of the double strike points is possibly driven by E×B poloidal drift, as shown by the purple curve in Figure 2, and the green and blue arrows mean that the poloidal electric drift flux flows toward and away from the divertor target, respectively. Different directions of the poloidal drift results in the peaks and hollow of the particle flux. The E×B poloidal drift is created by the plasma potential V$_p$, which is non-monotonic distributed in the scrape off layer (SOL) because of ECRH. On the other hand, the simulation points out that strike point splitting may be dependent on the electron density as well. Further experiment analysis and simulation are ongoing to make a final conclusion on the formation mechanism for the double strike points observed in this scenario.
Acknowledgements
This work is supported by National Natural Science Foundation of China under Grant Nos. 11905052, 11875020, 11875017, 11805057, and National Key R&D Program of China under Grant Nos. 2017YFE0301203, 2017YFE0301106.
References
$[$1$]$ T. Eich et al., Nucl. Fus., 53, 093031 (2013).
$[$2$]$ M. Jia et al., Plasma Phys. Contr. F., 58, 056102 (2016).
$[$3$]$ K. Kim et al., Phys. Plasmas, 24, 052506 (2017).
$[$4$]$ Y. Liang et al., Phys. Rev. Lett., 110, 235002 (2013).
$[$5$]$ R. D. Monk et al., J. Nucl. Mater., 241-243, 396-401 (1997).
$[$6$]$ A. V. Chankin et al., Plasma Phys. Contr. F., 43, 299–304 (2001).
$[$7$]$ Z. H. Huang, et al., to submit RSI (2020).Speaker: Dr Na Wu (Southwestern Institute of Physics) -
This contribution presents a numerical assessment of the impact of density fluctuations on the electron cyclotron (EC) wave in view of neoclassical tearing mode control in European DEMO. We show that, using the current design for the EC system launching the EC wave from equatorial outboard plane, the quality of the EC current profile is severely affected by the density fluctuations located at the plasma edge. The EC wave is diffused and has a width roughly 4-12 times wider (depending on the turbulence parameters) than originally designed in the absence of fluctuations. Therefore, control of neoclassical tearing modes within given installed power of EC system might become problematic or even impossible with the allocated EC power. The main reason for the deterioration is shown to be the EC injection geometry - the propagation path from the equatorial outer midplane to the resonance q = 1.5, 2 surfaces is about 2 meters, therefore, the angle deflections imparted by the scattering process in the plasma periphery are amplified by this long propagation distance. A design iteration where the EC beam is launched from upper port is considered, this option is compromised in view of technology (uppor port access is very limited) but might be necessary to restore NTM control.
Neoclassical tearing modes (NTM) severely limit the performance of burning plasmas by providing fast perpendicular transport of heat and particles within the NTM. This will reduce the temperature and drop the fusion yield. Hence, automized real-time control of NTMs have been studied and proven in various experimental machines, e.g. [1, 2]. The control scheme is based on launching EC wave to replace part of the missing bootstrap currrent ultimately driving the NTM. Recently, this process was re-iterated in view of ITER, where the plasma regime is slightly different from the exprerimental machines where NTM control has successfully been demonstrated. It was realized [3] that transport due to density fluctuations might pose a real risk for NTM control. Later on, this was shown by numerical simulations [4]. Further interest in the field has resulted in experimental efforts to show that the transport can not only be calculated using numerical tools but also measured experimentally [5, 6].
This work for European DEMO is a natural extension to the work carried out for ITER [4]. We use the exact same numerical tool, the WKBeam code, and repeat the exercise only us-ing DEMO specific plasma and wave parameters, and geometry [7]. The current EC design of DEMO is based on last steerable mirrors located at the outer equatorial midplane ports, as indicated by Figure 1. The numerical model for the turbulent density fluctuations is the same as used in [4] (interestingly although plasma parameters are different, e.g. the scaling for perpendicular correlation length will result in exact same numbers for ITER and DEMO). Here we present scan over the fluctuation amplitude at the SOL and correlation length in the perpendic ular direction in Figure 2. In this Figure, the broadening of the beam is defined as relative increase of the FWHM of the beam deposition profiles. The values of broadening are of the order of 8, for the best educated guess for turbulence parameters (L_⊥ =2 cm, δ n_e /n_e =20 %). This large broadening is putting severe constraint to the power required to control the NTM. The sensitivity analysis shows that this level of broadening is expected even for different variations of density profiles and turbulence parameters, suggesting that the result is not much dependent on the model input for the fluctuations. In particular, even just 5% fluctuation level would cause broadening by a factor of
4. The fluctuation model indeed is the main uncertainty in the numerical WKBeam model so the obtained result can be considered reliable, although not exact. The reason for such high broadening of the beam is identified to be the large propagation length between the turbulence layer located at the density pedestal and the resonance surface where the narrow beam width is necessary to drive the current within the NTM. Indeed, in current DEMO design, this distance is roughly 2m while in ITER it is only 50cm.We are currently assessing quantitatively the advantages of launching the beam from upper port as done for ITER. This will consirerably reduce the propagation path. However, such option is technologically more challenging to realize and needs strong evidence of performance improvement before it will be considered. Moreover, we are carrying out a study to estimate the necessary amount of EC power necessary to control NTMs as was done in [3].
References
1 E. Kolemen et al., Nucl. Fusion 54, (2014)
2 F. Felici et al., Nucl. Fusion 52, (2012)
3 E. Poli et al., Nucl. Fusion 55, (2015)
4 A. Snicker et al., Nucl. Fusion 58, (2018)
5 O. Chellai el al., Phys. Rev. Lett. 120, (2018)
6 M. Brookman et al., EPJ Web of Conferences 147, (2017)
7 T. Franke et al., submitted to Fus. Eng. Des., ISFNT, (2019)Speaker: Dr Antti Snicker (Aalto University) -
At the present time, the preparing for physical start-up of tokamak T-15MD is completed in the National Research Center “Kurchatov Institute”. Tokamak T-15MD has the following parameters: R=1.48 m, a=0.67 m, B=2.0 T, Ipl= 2.0 MA 1. The electromagnetic system is capable of maintaining without overheating (more 60ºC) the plasma current of 1MA for 40s, 700 kA for 120 s, 500 kA for 250 s and 300 kA stationary. Plasma current drive can be provided by both injection of fast neutrals and EC- and LH - waves.
Experimental study program on Tokamak T-15MD 2 is goaled on the obtaining of physical and technological data for both ITER support and creation of the fusion neutron source for atomic energy needs. Tokamak T -15MD will be used as the test bed, on which the stationary injectors of neutral particles, ICR-, ECR- and LH- plasma heating systems will be mastered, the materials and technologies for the first wall and divertor, including graphite, tungsten and lithium, will be studied.
In the first stage of T-15MD tokamak assembly the 16 D- form toroidal field (TF) coils (Fig.1) have been installed with the required tolerance in the vertical plane (< 2 mm/5m height). In August, 2019 the electromagnetic system, consisting of TF and poloidal field (PF) coils, together with the four sectors of vacuum vessel have been assembled in experimental hall (Fig.2). Four sectors of the vacuum vessel have been welded and tested. The leaks were not found out in the shell welds. The accuracy of the TF coils installation and the efficiency of the PF coils will be verified by means of electron beam in argon after vacuum vessel evacuation.
For high vacuum pumping of the vessel the four turbo molecular pumps with total productivity ~ 10 m3/s (H2) and two cryogenic pumps with total productivity of 8 m3/s (H2) will be used.
All vacuum pumping equipment was tested both in manual and automatic mode using a control system.The magnet system, in-vessel components, the turbo molecular pumps, equipment of auxiliary plasma heating systems are all cooled by distilled water with pressure of 0.5 MPa and 1.0 MPa. The total delivery for distilled water cooling by pumps is 2000 m3/h and the total mass flow rate of river water in the heat exchangers is 1500 m3/h. The equipment for water cooling system was mounted and tested.
Power supply system for Tokamak T-15MD includes: two substations 110/10 kV, two substations 10/0.83 kV, thyristor convertors and different equipment. Total power consumption during the pulse with plasma current 2 MA and additional plasma heating of 20 MW will consist of 300 MVA. Substation No745 (NRC “Kurchatov Institute”) is connected with heat electro power station by means of oil-filled cable 110 kV. Two oil transformers 80 MVA, 110/10 kV each and one oil transformer 40 MVA, 110/10 kV, designed for pulse loads during experiments with plasma, were installed and tested. The 96 vacuum switchgears (10 kV, I=1000A, 2000A, 3150A), placed in building 95 were introduced into the operation. Substations Nos.1, 2 are intended to supply power to the magnet system of tokamak T-15MD and auxiliary plasma heating systems. Sixteen new three-phase pulsed transformers were installed in substation No1 and four transformers were installed in substation No2. Each of the transformer is connected with the thyristor convertor. Twenty new thyristor convertors made in Czech Republic to supply power to toroidal winding and three coils of inductor were installed.
Tokamak T-15MD will be operated using the information and control system. All the information and control system equipment, required for the implementation of physical start-up of tokamak T-15MD in 2020 is available. Final adjustment of the control and information system after connection of all technological subsystems to tokamak T-15MD will be done.
Tokamak T-15MD will be equipped with contemporary physical diagnostics. Inside the vacuum vessel more than 250 different electromagnetic probes are installed to measure the plasma current, loop voltage, magnetic fields, MHD activity, etc. Plasma density, electron and ion temperatures, radiation losses will be measured too. Tokamak T-15MD is surrounded by wooden entresols for experimental equipment allocation.
The gyrotron with frequency 82.6 GHz and power of 1 MW will be used for preionization. The physical start-up of T-15MD is scheduled for the end of December 2020.References
1 Khvostenko P.P. et al. Tokamak T-15MD - two years before the physical start-up // Fusion Engineering and Design, Volume 146, Part A, September 2019, pp. 1108–1112.
2 Melnikov A.V., Sushkov A.V., Belov A.M. et al. Physical program and diagnostics of the T-15 upgrade tokamak (brief overview) // Fusion Engineering and Design, 2015, vol. 96-97, pp. 306-310.Speaker: Mr Petr Khvostenko (NRC "Kurchatov Institute") -
We report recent advances in reducing the coil complexity for optimized stellarators. Three efforts have been dedicated. First, the FOCUS code which uses fully 3-D representations for coils and employs analytically calculated derivatives has been applied in designing coils for new stellarators. FOCUS allows searching for more design space and thus is able to find more possible solutions. Secondly, a Hessian matrix method has been introduced to quickly identify the sensitivity of important error fields to coil deviations. Instead of performing computationally expensive perturbation analysis, the new method can determine the worst scenario by ranking the eigenvalues of the Hessian matrix. Last but not least, the state-of-the-art code, FAMUS, can determine the optimal layout of the magnets in a given design space. By using permanent magnets, a design of half-Tesla NCSX configuration with only planar TF coils is obtained. These tools/methods can help the design of next-generation stellarators in the U.S. and the construction of a new experiment in China.
The stellarator is an attractive approach to fusion energy because it has low recirculating power and is free of disruptions. On the other side, the 3D nature of stellarators generally requires more complicated coils than axisymmetric configurations. This problem remains crucial, as evidenced by the cancellation of NCSX stellarator in the US and the construction delays of the W7-X experiment recently completed in Germany. Stellarator optimization is usually divided into two steps. First of all, a configuration with the desired physics properties is optimized. And then coils are designed to reproduce the target magnetic field.
Conventional coil optimization codes assume that the coils lie on a pre-scribed surface, namely the winding surface. It decreases the degrees of freedom in the geometrical representation of the coils, but a pre-supposed winding surface also provides strong limitations. We have developed the FOCUS code (1) which relaxes the need for the winding surface. By using fully 3-D representations, coils can freely move in the space. Thus, more design space can be explored. FOCUS also employs analytically calculated (first- and second-order) derivatives to apply fast, robust optimization algorithms. Figure 1 shows a candidate coil design for a proposed midscale quasi-helical configuration in Madison, US. The coils can precisely produce the required magnetic field to support the equilibrium and retain the exceptionally good confinement for alpha particles (less than 2% loss on the s=0.3 surface).
The requirement for high accuracy during coil fabrication and assembly is also challenging, as coil deviation might produce error fields that degrade plasma performance. An error field sensitivity analysis prior to machine construction is necessary to determine acceptable tolerance. Instead of performing computationally expensive perturbation analysis to scan possible deviations, we have introduced a Hessian matrix method (2) that can directly find the worst combination of coil perturbations. The quadratic approximation indicates that the departure in the figure of merit away from the optimum is the eigenvalue-weighted norm in eigenspace. The first principal eigenvector, which is associated with the largest eigenvalue, will have the most significant effect on error fields. The second-order derivatives of normal magnetic field error, resonant magnetic harmonics and quasi-symmetry calculated in FOCUS thus can be used to determine the importance of all possible coil perturbations. As shown in figure 2, the modular coils of CFQS can deviate in a certain direction to enlarge/eliminate the n/m=4/11 islands (3). Important guidance is then provided for the upcoming coil fabrication and assembly.
The idea of using permanent magnets to provide a secondary magnetic field was proposed recently (4-5). We have developed the FAMUS code (6) to apply the technique of topology optimization determining the optimal layout of permanent magnets. Figure 3 demonstrates the design of half-Tesla NCSX configuration using permanent magnets and only planar TF coils. For engineering simplicity, the magnets are all perpendicular to the supporting surface. The existing ports on the vacuum vessel are explicitly excluded. Minimum magnets are used and remarkably large access on the outboard side is achieved. This calculation is leading to a proposed optimized stellarator at Princeton, US (7).
The results strongly demonstrate that advances in numerical modeling and optimization can reduce the complexity of stellarator coils and unearth the required coil tolerance. These efforts will help boost the design and construction of next-generation stellarators for pursuing fusion energy.
*This research was supported by the U.S. Department of Energy under Contract No. DE-AC02-09CH11466.
(1) C. Zhu et al 2018 Nucl. Fusion 58 016008
(2) C. Zhu et al 2018 Plasma Phys. Control. Fusion 60 054016
(3) C. Zhu et al 2019 Nucl. Fusion 59 126007
(4) P. Helander et al 2020 Phys. Rev. Lett. 124 095001
(5) C. Zhu et al 2019 arXiv: 1912.05144
(6) C. Zhu et al 2020 (to be submitted)
(7) D. Gates et al 2020 this conferenceSpeaker: Dr Caoxiang Zhu (Princeton Plasma Physics Laboratory) -
In JET plasma with a carbon wall (JET-C), and most other existing tokamaks, exceeding a critical density results in a H-L back transition, even when well above the H-mode power threshold. In contrast, at high density, JET plasma with a Be/W ITER-like wall (JET-ILW) always enter a ‘dithering’ phase before the H-L back transition, which enables a ($\approx20\%$) higher H-mode density limit (HDL) than in JET-C [1]. Burning plasma devices, such as ITER, will operate at high density to enable partial or total divertor detachment. The observed JET-ILW results suggest ITER can operate in H-mode at density above that previously predicted. This paper studies the edge-SOL physics of the HDL and the dithering phase. A hypothesis for the dithering phase limit cycle is given and the implications for high density operation of burning plasma devices such as ITER is presented.
Across a dataset of JET-ILW plasma, the ballooning stability parameter, $ \alpha_{sep} $ increases with $n_{e,sep}$ until a critical value, $\alpha_{crit}$, is reached after which confinement degrades, similar to JET-C, AUG [2] or DIII-D [3]. However, for the JET-ILW, $n_{e,sep}$ can increase significantly higher, with the plasma being in the dithering phase. Thus, for JET-ILW $\alpha_{sep}\geq\alpha_{crit}$ results in confinement degradation but another mechanism must cause the H-L transition and so the HDL. A new, reliable estimator for JET $E_r$ profiles in the edge has been derived by combining high resolution Thomson scattering (HRTS) measures of edge-SOL decay lengths, with HRTS pedestal gradient measurements. As shown in figure 1a, JET-ILW radial ETB wells are observed with $E_{r,min}$ in the range $-15$ to $-60\ kV/m$ in high performance H-modes, consistent with previous CXRS results of AUG [4]. $E_{r,min}$ for the edge plasma well is observed to vary little between L-mode and dithering phases, figure 1a. The inset figure in figure 1 shows this also the case for a single discharge, #91676, which alternates from dithering phase to L-mode and back to dithering phase whilst input power and global plasma parameters remain relatively constant. In both #91676 and the full dataset, $n_e$ and $T_e$ measured at the well minima are significantly higher (both $>50\%$) for the dithering phase. This implies that a small edge barrier is maintained during the dithering phase and, hence, there is indeed a bifurcation in the plasma state between the two phases. Plasma potential can be estimated as $V_{SOL,u}\approx 3kT_{et}/e+0.71 k\left(T_{e,u}-T_{e,t}\right)/e$, and used to determine the peak SOL E-field, $E_{r,SOL}$ [5]. $E_{r,SOL}$ is generally higher in the dithering phase, figure 2. The inset figure shows that the difference is more outstanding for the phases in #91676. The observed behaviour of $E_{r,min}$ and $E_{r,SOL}$ implies that the positive $E_r$ shear gradient at the separatrix is higher in the dithering phase than in L-mode. The higher positive $E_r$ gradient at the separatrix sustains the marginal phase and enables access to higher density when compared to JET-C discharges which do not enter a dithering phase. The weakening of this positive $E_r$ shear gradient eventually triggers the final H-L back transition.
Across a dataset of JET-ILW L-mode, H-mode and dithering plasma, normalised SOL width is found to increase with increasing collisionality, figure 3, in agreement with previous observation on AUG [6] and the Goldston finite collisionality HD model for SOL broadening at high collisionality [7]. A hypothesis for the dithering H-mode phase close to HDL is proposed with H-L-H-L- oscillations following:- $n_e$ increases $\to$ $\nu^*_{SOL}$ increases $\to$ SOL broadens $\to$ $E_r$ shear decreases $\to$ H-L transition;
- $n_e$ decreases $\to$ $\nu^*_{SOL}$ decreases $\to$ SOL narrows $\to$ $E_r$ shear increases $\to$ L-H transition.
The study has shown that, at high density, JET-ILW plasma reach the edge ballooning limit, $α_{crit}$ nd their confinement reduces but without a transition to L-mode. Instead, the higher positive $Er$ gradient at the separatrix sustains a dithering phase, shown to be a bifurcation in plasma state. A hypothesis has been developed for the limit cycle during this phase. Whilst the dithering H-mode phase enables access to higher density ($\approx20\%$) when compared to JET-C discharges, global confinement is observed to be low, $H_{98(y,2)}\approx0.75-0.8$. The observed JET-ILW results suggest ITER can operate in H-mode at higher density, which is beneficial for the power handling issue, but more likely in dithering phase with lower confinement and broader SOL. However, if the plasma exhaust can be handled at sufficiently low densities, operating just below the dithering phase would be a promising regime for maximising core density, global confinement and fusion performance.
[1] A. Huber et al. Nucl. Fusion 57 (2017) 0860007; [2] T. Eich et al. Nucl. Fusion 58 (2018) 034001; [3] A.W.Leonard et al, IAEA-TM 2019; [4] E. Viezzer et al. Nucl. Fusion 53 (2013) 0530005; [5] P.C.Stangeby, ‘The Plasma Boundary of Magnetic fusion devices’ 2000; [6] HJ. Sun et al. Plasma Phys. Contr. Fusion 57 (2015) 125011; [7] RJ. Goldston, 2018, EPS, ‘Generalization of the Heuristic Drift Model of the SOL for Finite Collisionality’Speaker: Dr Hongjuan Sun (UKAEA/CCFE, Culham Science Centre) -
- Introduction : Disruptions are one of the major concerns in ITER and other future tokamaks [1]. A particularly troublesome type of disruption is a vertical displacement event (VDE) where control of the vertical position of the plasma column is lost. In addition to heat, particle flux, and energetic electrons impacting the first wall, significant electromagnetic loads will arise.. For realistic modelling of a VDE disruption, a detailed 3D model of the disrupting plasma and an accurate description of the conducting structures surrounding the plasma is required. The structure affects the plasma evolution itself and the plasma acts as a source of currents and fields which produce the electromagnetic loads. Most of the VDE modeling work to date has used the axisymmetric evolving equilibrium codes TSC [2], DINA [3], and CarMa0NL[4] to describe the disrupting plasma. This paper describes our efforts to extend this analysis by using the fully 3D MHD codes M3D-C1 [5], NIMROD [6], and JOREK [7]. We describe our efforts in benchmarking these 3 codes on VDE relevant calculations, validation with some experimental data, and projection to ITER vessel and plasma conditions. Attention is given to the role of the currents shared by the plasma and the structure (halo currents).
- 2D and 3D benchmarking of M3D-C1, NIMROD, JOREK in simplified geometry
We have undertaken two VDE code benchmarking exercises for the 3 MHD codes M3D-C1, NIMROD, and JOREK. The first was in 2D (axisymmetric) [8] and we obtained excellent agreement for the linear and early nonlinear growth rates and for the motion of the magnetic axis, induced wall currents, and distribution of the halo current. The second benchmark, currently in progress, is fully 3D, beginning at the time the plasma first makes contact with the vessel. - Simulation of the sideways force in JET: M3D-C1 simulations of asymmetric vertical dis-placement events (AVDEs) in JET are in progress. The simulations were compared with previous M3D simulations [9] and with the JET 2011-2016 ILW disruption database. The computed and experimentally measured Noll approximations to the sideways force are in reasonable agreement. An important feature of the simulations and the data is that the sideways wall force is small when the current quench time is less than the wall penetration time.
- 2D benchmark of M3D-C1 and CarMa0NL for ITER geometry with no halo currents
We performed a separate nonlinear but axisymmetric VDE benchmark between M3D-C1 and CarMa0NL using a uniform thickness resistive wall located at the position of the ITER first wall. A 6 cm thick wall with resistivity of 7.4E-7 ohm-m gives a time constant of 235 ms for the wall current to develop when a constant loop voltage is applied (no plasma). Modeling of both a “hot” and “cold” VDE with the 2 codes, both without halo currents present, gave excellent agreement. [10] - The effect of Halo Currents on the vertical force: We have performed extensive axisymmetric studies with M3D-C1 of the effect of different assumptions regarding the temperature distribution of the halo region during an axisymmetric VDE in ITER geometry with the wall model described in #4 above. These studies use spatially and time varying parallel and isotropic thermal conductivities to model the thermal quench which then caused the current quench with differing amounts of halo current. We found the result that when a larger force from the halo current is present, it is at least partially offset by a smaller force due to the induced toroidal currents [11]. Slower current quenches generally lead to larger vertical forces, but the presence of a halo region can modify this scaling. Also, sheath-type boundary conditions are found to slow the evolution, increasing impulse in NIMROD computations [12].
- 3D VDE simulations in ITER geometry: Building on these successful code benchmarks and axisymmetric parameter studies in ITER geometry, we have performed several 3D VDE simulations with M3D-C1 using the ITER vessel model described in 4.
Currents and fields from these simulations are input to the CARRIDI code for detailed force calculations. Our initial result is that the intrinsically 3D “sideways force” is much smaller than the vertical force, largely because the kinking motion of the plasma occurs on a time scale much faster than the vessel time constant. This is largely consistent with an earlier scaling obtained with the M3D code. [13] - Summary and Conclusions: For the first time, we have applied massively parallel implicit 3D magnetohydrodynamics codes to predict the wall forces produced during ITER disruptions using realistic values for the wall resistivity and time constant. Extensive code verification and validation has been performed. We are in the process of refining the wall model to make it more realistic, and in developing a tighter coupling of M3D-C1 with CARRIDI in order to refine our present results.
Acknowledgments: This work was supported by US DOE Contracts No. DE-AC02-09CH11466, and No. DE-SC0018001the SciDAC Center for Tokamak Transient Simulations, EUROfusion Researcher Fel-lowship programme under the task agreement WP19-335 20-ERG-DIFFER/Krebs, EUROfusion grant No. 633053, F4E contract F4E-OPE-0891),and by Italian MIUR under PRIN grant 20177BZMAH.
Disclaimer ITER is the Nuclear Facility INB no. 174. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization. This publication is provided for scientific purposes only. Its content should not be considered as commitments from the ITER Organization as a nuclear operator in the frame of the licensing process.
References: - T. Schioler, C. Bachmann, G. Mazzone, et al, Fusion Engin. Design, 86, 11, pp. 1963-1966, (2011).
- S. C. Jardin, N. Pomphrey, and J. Delucia, J. Comp. Phys., 66, pp. 481–507 (1986).
- R. Khayrutdinov and V. Lukash, J, Comp. Phys., 109, pp. 193–201 (1993)
- F. Villone et al., Plasma Phys. Control. Fusion 55 (2013) 09500
- S. Jardin, N. Ferraro, J. Breslau et al, Comput. Sci. & Disc., 5 p 014002 (2012)
- C. Sovinec, A. Glasser, T. Gianakon, et al, J. Comp. Phys, 195 pp. 355386 (2004)
- M. Hoelzl, P. Merkel, G. Huysmans, et al. JPCS, 401, 012010 (2012)
- I. Krebs, F. J. Artola, C. R. Sovinec, et al., “Axisymmetric simulations of vertical displacement events in tokamaks: A benchmark of M3D-C1, NIMROD and JOREK”, submitted to Phys. Plasmas (2020)
- H. Strauss, S. Jachmich, E. Joffrin, V. Riccardo, J. Breslau, R. Paccagnella, and JET contributors, Physics of Plasmas 24, 102512 (2017)
- S. Jardin, F. Villone, C. Clauser, N. Ferraro, N. Isernia, G. Rubinacci, S. Ventre, “ITER Disruption Simu-lations with Realistic Plasma and Conductors Modelling”, Proceedings of EPS, (2019)
- C. F. Clauser, S. C. Jardin, and N. M. Ferraro, Nuclear Fusion 59 126037 (2019)
- K. J. Bunkers, PhD Dissertation, University of Wisconsin-Madison (2019)
- H. Strauss, R. Paccagnella, and J. Breslau, Phys. Plasmas 17, 082505 (2010)
Speaker: Dr Stephen Jardin (Princeton Plasma Physics Laboratory)
-
-
-
The capability to suppress edge localized modes (ELMs) is crucial for the success of ITER because the transient heat loads on the divertor due to ELMs would reduce the lifetime of plasma facing components to unacceptable levels. ELMs can be suppressed with the application of resonant magnetic perturbations (RMPs). But a side effect of RMPs is enhanced particle flux, or density pump-out, that reduces the plasma density by up to 50% [Evans 2006] and can significantly degrade fusion efficiency. Kinetic-level understanding of these RMP-driven phenomena is essential for predicting ITER's performance but is incomplete as of yet. One particular puzzle from experimental observation [Evans 2006] and transport modeling [Hu 2019] is how RMPs interact with neoclassical and turbulent transport to produce density pump-out even before magnetic islands penetrate the pedestal top while at the same time keeping electron heat well confined (or even improving confinement). Advanced MHD-assisted gyrokinetic simulations with the code XGC in realistic divertor geometry based on a DIII-D H-mode discharge with $n=3$, even parity resonant magnetic perturbations (RMPs) now reproduce these two experimental observations.
This study utilizes the global total-f gyrokinetic particle-in-cell code XGC [Ku 2018] with an RMP field from a linear two-fluid M3D-C1 [Ferraro 2009] calculation for arguably the highest-fidelity study of RMP-driven plasma transport to date. The combination of neoclassical and turbulent particle flux, in the presence of neutral particle ionization and charge exchange, explains a 50% density pump-out in less than 100 ms in the pedestal up to normalized poloidal flux $\psi_N\approx 0.96$ as well as the suppression of electron heat flux in the steep pedestal slope ($\psi_N\approx 0.97$). The majority of the enhanced particle flux at $\psi_N \ge 0.985$, where the magnetic field is stochastic, is from neoclassical transport (cf. [Hager 2019]). Turbulent transport is enhanced at $\psi_N \le 0.985$, where nested flux-surfaces remain intact. On the $\psi_N\simeq 0.97$ surface, both the turbulent particle and electron energy fluxes increase significantly due to the RMPs (see Fig. 1). But the electron heat flux is largely suppressed as the result of a cancellation between outward heat flux at longer wavelength ($k_\theta \rho_i \le 0.3$) and inward heat flux at shorter wavelength ($k_\theta \rho_i \ge 0.3$).
The RMS turbulence amplitude is enhanced due to the RMPs throughout the pedestal with a strong $n=3$ component [Fig. 2 (a) and (b)]. Closer examination of the turbulent fluctuations shows that ion drift modes (ITG) are active from $\psi_N\approx 0.94$ inward and electron drift modes (TEM) are active from $\psi_N\approx 0.94$ outward with a median poloidal mode number $k_\theta \rho_i \approx 0.25$, where $\rho_i$ is the ion gyroradius. The most influential (to the transport fluxes) RMP-induced modification is the amplification of the TEMs from the pedestal shoulder $\psi_N=0.94$ outward. This is shown at $\psi_N=0.97$ in Fig. 3, where the amplification of the TEMs is consistent with a reduction of the ExB shearing rate. The insights gained from these high-fidelity simulations suggest that a reduction in the TEM mode growth rate by an external means, e.g. enhanced edge rotation, could be a method to reduce the density pump-out. Our gyrokinetic understanding will also improve longer time-scale/lower-fidelity (e.g. [Hu 2019]) modeling to predict the steady state plasma profiles under RMP field. These discussions will also be presented.
This work is supported by the US Department of Energy (DOE) under contract nos. DE-AC02-09CH11466, DE-FC02-04ER54698. Computing resources: INCITE program at ALCF (DE-AC02-06CH11357) and ERCAP award at NERSC (DE-AC02-05CH11231).
[Evans 2006] T. E. Evans, R. A. Moyer, K. H. Burrell et al., Nature Physics 2, 419-423 (2006)
[Hu 2019] Q. M. Hu, R. Nazikian, B. A. Grierson et al., Physics of Plasmas 26, 120702 (2019)
[Ku 2018] S. Ku, C. S. Chang, R. Hager et al., Physics of Plasmas 25, 056107 (2018)
[Ferraro 2009] N. M. Ferraro, S. C. Jardin, Journal of Computational Physics 228, 7742-7770 (2009)
[Hager 2019] R. Hager, C. S. Chang, N. M. Ferraro, R. Nazikian, Nuclear Fusion 59, 126009 (2019)Speaker: Robert Hager (Princeton Plasma Physics Laboratory) -
Introduction. Extrapolations to ITER and DEMO from existing smaller experiments alone are unreliable, especially for turbulent transport - requiring the aid of predictive simulations. The 3D fluid turbulence code GRILLIX [1–4] is used to study confinement improvement through turbulence suppression that is compatible with power exhaust. This contribution describes the validation against experimental edge and scrape-off layer (SOL) measurements from the ASDEX Upgrade tokamak (AUG). Particular focus is put on the role of sheared $E \times B$ flows in turbulence suppression, and the resulting profiles, i.e. fall-off lengths (‘SOL width’). Simulations are also checked against experimental scaling laws [5] over varying parameters, and against smaller machines like C-mod and TCV. Further, we investigate the effect of varying magnetic geometry, in shaped single null as well as in advanced divertor configurations (ADCs).
Verification and validation. We report on the first quantitative validation based on L-mode AUG experiments in standard single null divertor geometry, depicted in Fig. 01. We compare profiles and fluctuations of density and temperature, and the electric field. In agreement with experiment, the ballooning driven fluctuations in the confined region are moderate, < 4%, and rather high in the SOL, > 15%, with highly intermittent events. Obtaining these results was possible after several improvements in the physical model, numerical methods and computational performance. The code has been verified via the method of manufactured solutions [1]. Previously, it was shown to reproduce many features of limited and diverted tokamaks [1, 3], and validated against experimental results from the Large Plasma Device (LAPD) [2], explaining the mechanism of blob formation and propagation.
Shear flows and turbulence suppression. $E \times B$ flow shear [6] and magnetic shear [1] at the separatrix suppress turbulence in the pedestal, which is crucial for the desired confinement in ITER. The $E \times B$ flow shear is explained by the experimentally observed jump in the radial electric field $E_r$ across the separatrix. We show that this is due to different underlying mechanisms in the closed and open field line regions: $E_r \approx \partial_r p_\mathrm{i}/n$ in the confined region, transitioning to $E_r \approx −3 \partial_r T_\mathrm{e}$ in the SOL due to sheath boundary conditions [6]. Additional contributions from rotation and Reynolds stress are damped at higher temperature. The resulting pressure and electric field profiles are shown in Fig. 02.
Dedicated framework for diverted tokamaks. The plasma edge and SOL have to be studied together, because edge turbulence is affected by the plasma-wall interaction in the SOL, and SOL turbulence largely spreads from the plasma edge [7]. Their interface, the separatrix, not only allows to separate the plasma-wall interaction in the divertor from the confined plasma, but it also produces strong, turbulence suppressing magnetic shear. Accurately resolving this dynamics is vital for predicting both the heat flux on the divertor, and the confinement properties of the device. GRILLIX is able to do this thanks to the flux-coordinate independent (FCI) approach [8], which allows for a flexible and efficient treatment of arbitrary magnetic geometries.
Multi-scale electromagnetic model. The model in GRILLIX, by keeping all global dependencies [3], describes the evolution of equilibrium profiles together with turbulent fluxes, allowing fluctuations of arbitrary magnitude. This is important for the prediction of pedestal profiles, i.e. fall-off lengths like the SOL width. Electromagnetic effects are crucial in limiting the response of the parallel current via shear Alfvén waves, as opposed to resistivity [1]. However, electromagnetic cross-field transport (due to field-line flutter) is significant only at higher pressure ($\beta$) and primarily for heat.
Advanced divertor concepts. ADCs such as snowflake and Super-X are promising for decreasing the heat flux to reactor walls [9], but their experimental implementation is so far rare. GRILLIX can support advanced divertor design by simulations [4]. In the snowflake divertor, we find a more complex structure of the electric field then in single null: a potential maximum at the magnetic null, i.e. a convective cell, redirects part of the SOL flows to the secondary divertor legs. Together with the increased magnetic flux expansion, this is promising for improving heat exhaust in DEMO.
Parameter scalings. The comparison to experimental multi-machine scalings, such as the SOL width scaling by Eich et al. [5], is facilitated by direct simulations of different machines like AUG, C-mod and TCV, and by parameter scans. Both plasma pressure and collisionality increase turbulent transport through different mechanisms. Larger machine size and larger toroidal magnetic field decrease the vortex size compared to the machine scale, effectively increasing confinement. In addition to magnetic field strength, the shape of the poloidal magnetic field has a significant impact: turbulence can be suppressed both directly through magnetic shear, and indirectly through a different flow pattern. This effect is especially visible near the X-point.
Conclusion. GRILLIX was validated against AUG experiments and checked against multi-machine scaling laws, which is a crucial prerequisite for ITER and DEMO simulations. The code is able to predict turbulent fluxes, including their suppression in the pedestal, as well as the resulting equilibrium profiles, $E_r$ and fall-off lengths like the SOL width. This was shown both in standard single null as well as in advanced divertor concept geometry.
References
[1] A. Stegmeir et al., Phys. Plasmas 26, 052517 (2019)
[2] A. Ross et al., Phys. Plasmas 26, 102308 (2019)
[3] W Zholobenko et al., Contrib. Plasma Phys. e201900131 (2019),
https://doi.org/10.1002/ctpp.201900131
[4] T. Body et al., Contrib. Plasma Phys. e201900139 (2019),
https://doi.org/10.1002/ctpp.201900139
[5] T. Eich et al., Nucl. Fusion 53, 093031 (2013)
[6] B. Zhu et al., Phys. Plasmas 24, 055903 (2017)
[7] P. Manz et al., Phys. Plasmas 22, 022308 (2015)
[8] A. Stegmeir et al., Comput. Phys. Comm. 198, 139 (2016)
[9] H. Reimerdes et al., Plasma Phys. Control. Fusion 55, 124027 (2013)Speaker: Mr Wladimir Zholobenko (Max Planck Institute for Plasma Physics) -
In this paper we will present nonlinear full-$f$ electromagnetic gyrokinetic simulations of turbulence in the pedestal and scrape-off layer (SOL) region of a tokamak. The algorithms in the Gkeyll code solve the electromagnetic gyrokinetic equations using a continuum high-order discontinuous Galerkin scheme. The equations are written in a sympletic form in which the particle parallel momentum is used as a coordinate. The details of our algorithms are presented in[1,2], with previous work in simulating electrostatic turbulence in in helical SOL plasmas in[3,4]. As far as we are aware, these are the \emph{first} fully nonlinear full-$f$ electromagnetic simulations in the high-$\beta$ regime on open field-lines.
Our setup is a helical, open field-line model of a NSTX-like SOL. We use a non-orthogonal field-aligned coordinate system. The simulation domain is a flux-tube on the outboard side that wraps helically around the torus, with the field-lines intersecting metal divertor plates on either ends. The interaction of the plasma with the divertor plate is modeled using a sheath boundary condition which allows current fluctuations into and out of the divertor plates.(The results presented in this abstract focus on the SOL with a helical model, but we have extended the code to general geometry sufficient for a pedestal and SOL with a limiter, with X-point capabilities available soon.) The simulation parameters approximate SOL conditions in an H-mode deuterium plasma: $B_\text{axis}=0.5$ T, $R_0=0.85$ m, $a=0.5$ m, $T_{e0}=T_{i0}=40$ eV. (The plasma $\beta$ is enhanced by a factor of 10 relative to the steady-state experiment to show the robustness of the algorithm, even for ELMs or or other high-$\beta$ transients.) The plasma core supplies a particle source for the SOL.
The bad curvature drive causes interchange instabilities in which blobs are ejected intermittently radially outwards. This is seen in the density contours in Fig.1 in which mushroom like structures appear. Significant magnetic fluctuations of around $2.5\%$ are seen in $|\delta B_\perp|/B_0$, greatly modifying the transport as compared to when the electromagnetic terms are neglected. In Fig.2 we show the radial profiles of density, temperature and beta for both electrons and ions. These are compared to the case in which the electromagnetic terms have been turned off. It is seen that in the ES case the profiles are shallower, indicating that the radial transport is less in the EM case. In general, the inclusion of the EM terms makes the turbulence more intermittent, significantly changing the transport in the high-$\beta$ regime. Even though the magnetic field-lines, Fig.3, are tied right at the divertor plates, they slip due to sheath resistance, and in the interior, fluctuate and reconnect as the blobs drag the field-lines in their outward motion. The results summarized here are for turbulence in the SOL.The extension of these results to include the pedestal physics also will be presented.
-
Ammar H. Hakim, Noah R. Mandell, T. N. Bernard, M. Francisquez, G. W. Hammett, E. L. Shi, "Continuum Electromagnetic Gyrokinetic Simulations of Turbulence in the Tokamak Scrape-Off Layer and Laboratory Devices", Submitted to Phys. Plasmas
-
Tess N Bernard, Eric L Shi, KW Gentle, Ammar Hakim, Gregory W Hammett, Timothy Stoltzfus-Dueck,and Edward I Taylor. "Gyrokinetic continuum simulations of plasma turbulence in the Texas Helimak". Phys. Plasmas, 26(4):042301, 2019
-
N R Mandell, A Hakim, G W Hammett, and M Francisquez. "Electromagnetic full-f gyrokinetics in the tokamak edge with discontinuous Galerkin methods", J. Plasma Phys., 86(1):905860109, 2020
-
Eric L Shi, Gregory W Hammett, Timothy Stoltzfus-Dueck, and Ammar Hakim. "Full-f gyrokinetic simulation of turbulence in a helical open–field–line plasma". Phys. Plasmas, 26(1):012307, 2019
Speaker: Mr Ammar Hakim (Princeton Plasma Physics Labaroratory) -
-
During burning plasma operation on ITER, extrinsic impurity seeding will be mandatory for heat flux control at the tungsten (W) divertor vertical targets [1]. A very extensive database of SOLPS plasma boundary code simulations has been compiled for ITER [1], including the most recent advances, obtained with the SOLPS-ITER version, in which for the first time, fluid drifts have been included [2]. These simulations predict that partially detached divertor solutions at high divertor neutral pressure will be possible on ITER for baseline burning plasmas ($Q_{DT}=10$, power into the scrape-off layer $P_{SOL} = 100 MW$), with both neon (Ne) and nitrogen (N) low Z seeded impurity, and with impurity compression sufficient in both cases to maintain the majority of the radiated power in the target vicinity. Drifts are found to be relatively unimportant under such conditions. This is in contrast to observations on smaller devices with W divertors, such as ASDEX-Upgrade (AUG), in which Ne compression is reduced in comparison with N, core plasma performance is compromised and drift effects are stronger. However, Ne is preferred on ITER in DT plasmas to avoid impact on machine duty cycle due to the formation of tritiated ammonia [1]. It is thus critical that the fundamental controlling physics responsible for this behavior be understood, in particular the impact of scale size. This contribution identifies the key factors at play through a unique SOLPS-ITER simulation study in which Ne-seeded H-mode conditions in vertical target W divertor geometries are compared in the three devices, ASDEX-Upgrade (R=1.65 m), JET (R = 3.0 m) and ITER (R = 6.2 m), spanning a factor of more than 3 in linear dimension in almost equal size intervals.
For AUG and JET, the modelling parameters are inspired by existing experimental results, but do not attempt to match a particular discharge. High power, H-modes with $P_{SOL} = 12 MW$ , $q_{95}=5.5$ (AUG) and $P_{SOL}=20 MW$, $q_{95}=3.3$ (JET), and cross-field heat transport chosen to match the typical SOL widths, ${\lambda}_q$ observed on these devices under such conditions, give comparable power flows at the divertor entrance to those for the modelled ITER burning plasma with ${\lambda}_q = 3-4 mm$, $P_{SOL} = 100 MW$ , $q_{95}=3$. All simulations include fluid drifts and currents, with neutrals traced by the EIRENE code. Semi-detached divertor conditions are established in all cases using moderate Ne seeding and high deuterium throughput.
An important finding is that the impact of the poloidal and radial ExB drifts (redistributing plasma between the outer and inner divertors through the private flux region [3]) steadily decreases with increasing machine size. The divertor asymmetry associated with these drifts also thus decreases with size. As shown in Fig. 1, the high field side, high density front observed experimentally in AUG [4] and JET [4] is reproduced by the modelling, but is absent in the ITER simulation. On AUG this front reaches the X-point vicinity and can influence the pedestal plasma [5]. The same divertor asymmetry driven by the ExB drift drives a significant redistribution of Ne impurity, which tends to accumulate in the more detached inner divertor (Fig. 2). The resulting increased impurity radiation further exacerbates the divertor asymmetry, making it harder to achieve partially detached conditions in the outer divertor without impurity concentrations exceeding acceptable limits for core performance. This effect also decreases with increasing machine size. The relative importance of $\nabla B \times B$ and Pfirsch-Schlueter (P-S) driven flows in the SOL region also differs between ITER and smaller devices. Whilst in AUG, P-S flow provides the main contribution to flow reversal in near SOL, in ITER excess ionization in the strike point region will be the principal driver.
A second key size dependent effect concerns the electron temperature ($T_e$) distribution in the divertor, which depends on the X-point to target connection length. Due to the $T_e$ dependence of the parallel heat conductivity, the region of temperature change near the target or, in the case of detachment, the ionization front, is narrow and does not depend on the X-point position or machine size. The position of this layer thus determines the X-point $T_e$. For larger devices, with higher confinement, this temperature is higher for the same energy flow at the divertor entrance. Since low Z impurity radiation will be strongest in the region of comparatively low $T_e$, where many partially ionized states exist, impurity radiation in larger machines is more localized in the divertor. On ITER, this means that even though the strongly radiating region with Ne impurity is more extended than for N seeding at comparable radiated power, both are equally effective at divertor power dissipation. In addition, any Ne ions reaching the ITER pedestal region are fully stripped due to the high $T_e$ there under high performance conditions and cannot radiate, reducing the impact on pedestal power balance.
[1] R. A. Pitts et al 2019 Nucl. Mater. Energy 20 100696
[2] E. Kaveeva et al Nucl. Fusion in press: DOI 10.1088/1741-4326/ab73c1
[3]V Rozhansky et al 2018 Contrib. Plasma Phys. 58 540–546
[4] S Potzel et al 2015 Journal of Nucl. Mater. 463 541-545
[5] M G Dunne et al 2017 Plasma Phys. Control. Fusion 59 014017Speaker: Prof. Vladimir Rozhansky (Peter the Great St.Petersburg Polytechnic University)
-
-
-
Tungsten (W) and tungsten based alloys are candidate materials for plasma-facing components (PFCs) in future fusion reactors, largely due to their low erosion yield by physical sputtering and low retention of plasma fuel in them. The present work concentrates on erosion results obtained from two European tokamaks, ASDEX Upgrade (AUG) and WEST, which have operated with full-W first walls since 2007 and 2017, respectively. In particular, we elucidate the erosion behaviour of W PFCs during representative AUG plasma discharges, identify how various material and plasma parameters affect them, and determine the balance between their gross and net erosion, i.e., how material is migrating and re-depositing on the vessel. Results for W erosion during recent experimental campaigns on WEST are also presented for comparison.
Extensive sets of results are available on campaign-integrated erosion processes as well as on W sources in the different regions of the AUG torus [1, 2]. However, only recently the focus of the studies has shifted towards understanding erosion and re-deposition behavior during individual plasma discharges. This way, it has become possible to benchmark predictive simulations of W migration in the edge plasma. Our strategy has been exposing different sets of marker samples, mounted at the low-field side (outer) divertor region of AUG, to a pre-determined number of plasma discharges and modelling the obtained erosion and re-deposition profiles using the simulation codes ERO, SOLPS, and DIVIMP. The main research lines have been clarifying how the observed erosion patterns are influenced by the type of the plasma discharge (L-mode vs. H-mode) including transient phenomena like ELMs, by the plasma gas (deuterium (D) vs. helium (He)), and by the material and morphology of the PFCs. Most of the results are obtained by using gold (Au) or molybdenum (Mo) as proxies for W to eliminate the contribution of material influx from other regions of the torus; Indeed, even though Au, Mo, and W have different erosion yields, they exhibit comparable ionization and migration lengths. The physics picture is supplemented by the analyses of the W erosion profiles of WEST marker tiles, removed from the vessel after the recent completion of its phase I operations, and comparison to the existing W erosion data from JET [3].
To obtain measurable erosion rates, the AUG plasma scenarios have been designed to exhibit high electron temperatures up to Te=20-25 eV in the divertor plasma. In L-mode discharges, both net and gross erosion typically peak around the outer strike point, roughly following the shape of the Te profile. In the strike-point region, the gross erosion of the markers is 3-4 times larger than their net erosion while at a distance of a few cm from the most affected region, the net erosion rapidly decreases, even turning into net deposition [4]. In addition, the eroded particles are locally re-deposited in the immediate vicinity of their origin [5]. We estimate that the net/gross erosion ratio is always less than 50% in AUG L-mode plasmas. According to ERO simulations, this ratio systematically decreases as the Te drops or the plasma density increases [5].
In H-mode discharges with inter-ELM conditions similar to those in the L-mode, gross erosion during ELMs is 1-2 orders of magnitude higher than in between ELMs [6]. The overall contribution of ELMs on the W source is estimated to be >50% [1], while net erosion is enhanced by a factor of 2-3 [6], see Figure 1. Compared to L-mode data, re-deposits are observed at distances of several cm from the studied marker structures and, according to Figure 2, in the direction magnetically downstream of the markers, erosion can be highly reduced or completely eliminated.
In the absence of ELMs, ERO simulations predict impurities to have the largest effect on net erosion in regions where Te drops to values below 20 eV. However, without any impurities, erosion would be almost two orders of magnitude lower. The simulated erosion profile is typically more peaked than the experimental one and it exhibits a toroidal tail of re-deposited particles downstream of the markers. However, the particle density remains below the experimental detection threshold [5]. Flow of particles along the field lines as well as poloidal transport and diffusion across the field lines, driven by the E×B drift, are found to be important ingredients in explaining the measured erosion profiles [7].
In He plasmas, W erosion is intensified by the large mass and charge of the plasma species. Additionally, in the experiments discussed here the applied ion cyclotron resonance heating schemes can result in large influx of material from the main chamber to the divertor, leading to all the marker samples being covered with deposited layers [8]. When the extra W source is eliminated and the amount of impurities minimized, net erosion in He can be estimated to be 5-10 times larger than in comparable D plasmas.
Surface roughness reduces erosion by a factor of 5-10 and even turns net erosion into net deposition for roughness >1000 nm [9], see Figure 3. This happens mostly in areas where the E×B drift plays a significant role. Local net erosion of W appears to follow the surface topography: in shadowed areas W re-deposition dominates while for plasma-exposed regions erosion is uniform at length scales much larger than the different surface features. One therefore needs to carefully consider the surface features and their evolution during plasma operations to obtain a reliable estimate for gross/net erosion balance.
[1] R. Dux et al., J. Nucl. Mater. 390-391 (2009) 858
[2] A. Hakola et al., J. Nucl. Mater. 463 (2015) 162
[3] S. Brezinsek et al., Nucl. Fusion 59 (2019) 096035
[4] A. Hakola et al., Phys. Scr. T167 (2016) 162
[5] A. Hakola et al., ”ERO modelling of net and gross erosion of marker samples exposed to L-mode plasmas on ASDEX Upgrade”, 24th PSI Conference, Jeju, Korea, 2020.
[6] A. Lahtinen et al., ”Erosion and deposition studies during L- and H-mode plasma discharges in the outer strike point region of ASDEX Upgrade”, 24th PSI Conference, Jeju, Korea, 2020.
[7] A. Hakola et al., Nucl. Mater. Energy 12 (2017) 423.
[8] A. Hakola et al., Nucl. Fusion 57 (2017) 066015.
[9] A. Lahtinen et al., Europhysics Conference Abstracts 41F (2017), P2.119Speaker: Antti Hakola (VTT Technical Research Centre of Finland Ltd.) -
Present and future long-pulse tokamaks such as JT60-SA, ITER and DEMO will require increasingly advanced control methods to maximize the plasma performance and pulse duration while avoiding plasma disruptions. Given the cost and complexity of a single discharge, maximal use of automated approaches is strongly preferred above costly and error-prone trial-and-error discharge development approaches. In addition to continuous plasma control functions such as position, shape and kinetic (density, temperature) control and profile control (e.g. current density), also off-normal events, including (but not limited to) those that may lead to a disruption, need to be detected and handled in an automated way. At the same time, all these control tasks should be handled with a minimal set of actuators and diagnostics. For this purpose, a sophisticated set of plasma reconstruction, monitoring, supervision and control algorithms will be required. In this work we present the design, implementation, and integrated experimental application of multiple key elements of these advanced plasma control systems on TCV and ASDEX Upgrade.
On both ASDEX Upgrade and TCV, a model-based plasma state reconstruction system has been developed that combines real-time models for the profile evolution, including real-time source calculations for EC and NBI actuators [1], [2], with real-time diagnostic measurements. On TCV, real-time model-based plasma profile estimations have been coupled to equilibrium calculations, yielding for the first time a real-time ‘kinetic’ plasma equilibrium reconstruction with realistic pressure and current density profiles [3]. On ASDEX Upgrade, an improved real-time density profile estimation algorithm has been deployed, merging several diagnostics to provide reliable estimates also in situations where key diagnostic systems are compromised – for example in the case of loss of interferometer measurements during ICRH injection.
In addition, on TCV, a novel framework for real-time control including off-normal event handling has been designed, implemented and experimentally tested [4,5]. The system, as shown in the figure, consists of two layers: a tokamak-specific interface layer and a tokamak-independent task layer. The interface layer translates the tokamak-specific diagnostic and actuator signals into tokamak-independent, general descriptions of the state of the plasma and the tokamak. This allows control tasks to be handled by the ‘task-layer’, in a way that is general and portable across tokamaks. The task layer contains a plasma state/event monitor, which categorizes the state and events (if any) and sends this information to the supervisory controller. This supervisory controller generates, based on user-set rules, a prioritized list of control tasks that should be performed to respond to the present conditions of the plasma. An actuator manager then assesses the list of prioritized tasks and decides which tasks can be fulfilled with the presently available set of actuators, based on resource requests from a set of controllers handling each task. This system is capable of handling an arbitrary chain of events with an appropriate response depending on the severity of the event – with a response ranging from changing control references, re-allocation of actuators, or shutting down the discharge. The framework is generic and designed to be tokamak-independent, therefore it can readily be ported to other tokamaks.
This new control framework has been tested in numerous experiments on TCV. A first example demonstrated real-time re-assignment of ECRH actuators from q-profile and beta control to Neoclassical Tearing Mode (NTM) control in response to the appearance of the mode [5]. More recently, this control framework has been used for avoiding high-density limit disruptions: monitoring the distance from a threshold (experimentally determined in the 2D space of H98 and normalized edge density), the heating power and gas injection were controlled to remain at a safe distance from the disruptive limit. The disruption boundary and algorithm used is identical to one previously implemented on ASDEX Upgrade [6], proving the portability and generality of the approach. Work is ongoing to include real-time handling of other well-understood chains of events that can lead to disruptions, such as radiative peaking.
This generic control framework naturally supports more intelligent control, notably making tokamak control problems resource-aware. For example, intelligent NTM controllers that use real-time simulations of the Modified Rutherford Equation to predict the power required for stabilizing an NTM have been tested in simulation [7]. With knowledge of the resources required for the NTM stabilization (and other tasks), the supervisory controller has the ability to decide that a disruption can not be avoided (for example due to saturation of suitable actuators, or unexpected evolution of plasma quantities). The supervisor can then trigger further measures such as plasma ramp-down or triggering of a disruption mitigation system. By handling events that can be treated with conventional means, while flagging those that can not, this advanced plasma control system acts as a first line of defense to avoid unnecessary triggering of mitigating actions – which will be a crucial element of control for long-pulse tokamaks and fusion reactors.
While real-time control will be essential for future reactors, some quantities remain difficult to measure or control in real-time. For these quantities, it is important to be able to rapidly optimize the feedforward actuator references from shot to shot, to minimize trial-and-error and save experimental time. We report on new experiments on ASDEX Upgrade for automated shot-to-shot optimization of the plasma temperature, using the Iterative Learning Control (ILC) method, using a model of the response of the plasma temperature to the heating actuators to compute a correction to the commands of the heating systems. These experiments benefit from a newly developed system for actuator management for ASDEX Upgrade allowing several heating actuators to be used for the same control task [8]. This is particularly useful when replacing actuators that may fail during a discharge with other actuators that can fulfil a similar role.
Acknowledgements
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This work is supported in part by the Swiss National Science FoundationReferences
[1] E. Poli et al., Comput. Phys. Commun. 225, 36 (2018)
[2] M. Weiland et al. Nucl. Fusion, 58(8) 082032, (2018)
[3] F. Carpanese et al., submitted. to Nuclear Fusion (2020)
[4] T. Blanken et al, Nucl. Fusion 59(2) 026017 (2019)
[5] N. M. T. Vu et al., Fusion Eng. Des. 147 111260 (2019)
[6] M. Maraschek et al., Plasma Phys. Control. Fusion 60(1) 014047 (2018)
[7] M. Kong, et al. PhD Thesis, EPFL (2020)
[8] O. Kudlacek et al. Fusion Eng. Des., 146, 1145 (2019)
[9] S. Coda et al., Nucl. Fusion 59 112023 (2019)
[10] H. Meyer, et al., Nucl. Fusion 59 (2019) 112014
[11] B. Labit et al., Nucl. Fusion 59 086020 (2019)Speaker: Federico Felici (École Polytechnique Fédérale de Lausanne, Swiss Plasma Center) -
Disruption instability and formation of the accelerated electron beams represent one of the main problems in design of the cost-effective fusion reactor. To minimize consequences of the disruptions in tokamaks, several methods for predicting disruption, controlling plasma discharge at the initial stage of the instability, and fast quenching (stopping) a plasma discharge are considered [1].
In experiments on the T 10 tokamak with all-metal (tungsten, lithium) in-vessel elements, it was shown that transition from quasi-stationary discharge to a fast phase of disruption can be associated with development of arc plasma discharges at the limiters [2]. New technique based on special movable electric and magnetic probes installed in the T-10 tokamak allows monitoring of the arc discharges and provide reliable trigger parameters for disruption stabilization systems and/or for safe discharge termination systems.
Possibility of restoration of the stable plasma discharge after start of disruption (during series of thermal quenches) at high density is studied in experiments on the T 10 tokamak using programmed operation of the Ohmic current supply system and gas puffing. For the first time, complete restoration of the plasma discharge with recovering of the pre-disruption plasma parameters (electron density and temperature) is obtained by sequentially controlled decrease in the plasma current with regulation of gas flows in the peripheral regions of the plasma and subsequent increase of the plasma current. Discharge recovery is accompanied by stabilization of burst of the magnetohydrodynamic (MHD) modes and prevention of the accelerated electron beams.
Injection of intensive gas flows and solid-state macroparticles is actively used to safely stop plasma discharge in modern tokamak experiments and is considered as one of the main systems for preventing development of the accelerated electron beams in the tokamak reactor ITER [1]. One of the main limitations in using these systems in large-scale tokamaks is weak penetration of the gas flows into central zones of the plasma discharge and long time response of the injected particles techniques. This reduces reliability of the disruption mitigation and leads to needs to testing additional methods for safe discharge termination. To minimize damaging consequences of disruptions, several "novel" methods of fast gas flow injection are tested in the T 10 tokamak, including injection of neutral particles from targets with biasing, injection of impurities sprayed with high-power microwave waves, and ultra-fast gas injection by initiation of rapid chemical combustion reactions (CCR). Experiments with CCR in T 10 showed that plasma shutdown starts from abrupt growth of the m°=°2 MHD mode provided effective penetration of the fast gas flows into the central zones of the plasma columns and quick stop the plasma discharge with plasma current decay rate of up to 35°-°40°MA/sec and control system response time of up to 0.1°msec.
Generation of the directed fast gas flows by initiation of chemical combustion reactions is tested at the Laboratory bench. Preliminary analysis has shown possibility of using fast-flowing chemical reactions initiating by substances with increased radiation resistance to the fast neutron fluxes, temperature stability, and high stability in vacuum conditions in the ITER-like tokamak reactor.
The work is supported by ROSATOM Contract.[1] Lehnen M. et al, R&D for reliable disruption mitigation in ITER. Preprint: 2018 IAEA Fusion Energy Conf. (Gandhinagar, India, 22–27 October 2018) p.EX/P7-12.
[2] P. V. Savrukhin and E. A. Shestakov, Physics of Plasmas 26, 092505 (2019); https://doi.org/10.1063/1.5102112.
Speaker: Petr Savrukhin (NRC Kurchatov Institute) -
The paper reports unique observation of an ion cyclotron emission (ICE) from the ohmically heated plasma in the absence of energetic ions. The possibility to use the ohmic ICE for determination of the hydrogen isotopes ratio is discussed.
Described experiments were performed in the compact circular shaped limiter tokamak TUMAN-3M $(R(0)=0.53\:m$, $a=0.22\:m$, $B_T=1\:T$, $I_p=150\:kA$, $\overline{n}_e=(1\mbox{-}3)\cdot10^{19}\:m^{-3}$, $Ti(0)=0.2\:keV)$. No auxiliary heating was used in these experiments. The ICE appears as a set of equally spaced lines in the frequency spectra measured by in-vessel magnetic probes. The ICE spectra span from $5$ to $120\:MHz$ depending on a sort of working gas (hydrogen or deuterium) and probe position. In TUMAN-3M an ICE driven by Neutral Beam Injection was observed as well. Study of the NBI driven ICE is presented at this conference in a separate report by Askinazi L.G. et al.
Typical example of the ohmic ICE spectrogram obtained in deuterium plasma (shot #$17011907$) using a signal from the magnetic pick-up probe sited at the high field side of the torus, $R_{probe}=0.43\:m$, is shown on figure 1. The spectrogram contains up to nine harmonics of ICE with the fundamental one matched to the deuteron cyclotron frequency $f_{cD\mbox{-}I}$ at the probe location. Exact proportionality of a harmonic frequency to the local toroidal magnetic field is justified by coincidence of scaled $B_T$ trace with the third harmonic evolution as seen on the figure. The $\sim\;10$% variation of $B_T$, see figure 1, is caused by the features of the TF power supply operation. No variation of the ICE harmonic frequency with density was observed$^1$.
ICE spectrum obtained by FFT during $87\mbox{-}89\:ms$ of the shot #$17011907$ is shown on figure 2. The first peak of the spectrum coincides with the deuteron resonance frequency $f_{cD\mbox{-}I}=\omega_{cD\mbox{-}I}/2\pi=7.6\:MHz$ at $R_{probe}=0.43\:m$. Fundamental frequencies of the ICE detected by probes located at other major radii positions strictly match to the cyclotron resonances at the probes’ locations$^1$. The spectrum on figure 2 is characterized by a narrow width of the $f_{cD\mbox{-}I}$ peak. The FWHM of the peak $f_{cD\mbox{-}I}$ is $0.5\:MHz$. The narrowness of the peak indicates that source of the emission is strongly localized in both the major and minor radius directions. The size of the emitting area was estimated to be less than $2.5\:cm$. The peak width increases with harmonic number. FWHM of the fifth harmonic $f_{cD\mbox{-}V}$ is $3\:MHz$. It should be noticed that narrow $f_{cD\mbox{-}I}$ peak is the characteristic of the low density plasma, e.g. in shot #$17011907$ the average plasma density was $1.5\cdot10^{19}\:m^{-3}$. With the increasing density the width of the peak and amplitude of the higher harmonics rise. At average density of $2.7\cdot10^{19}\:m^{-3}$ FWHM of the $f_{cD\mbox{-}I}$ peak attains $2\:MHz$ and $f_{cD\mbox{-}IV}$ or even $f_{cD\mbox{-}V}$ has maximal amplitude in the spectrum.
Similar features of the ohmic ICE were found in the shots with hydrogen working gas although amount of observed harmonics was less than in deuterium – 7 instead of 9. The frequency of the fundamental harmonic $f_{cH\mbox{-}I}$ measured by the particular probe was two times higher than the one in deuterium plasma.
Above observations lead to idea to explore a possibility to measure an isotope ratio in the proton/deuteron plasma using the ohmic ICE. In order to do this a set of experiments was performed with additional puffing of deuterium into hydrogen discharge. In addition to the ICE measurements the Alfven oscillations$^2$, optical spectroscopy and neutral particle analysis were used to monitor isotope ratio. Figure 3 presents results of the measurements of relative concentration of deuterium in hydrogen/deuterium plasma – $n_D/(n_H+n_D)$ using optical spectroscopy and ratio of ICE peak intensities at $f_{cD\mbox{-}I}$ and at $f_{cH\mbox{-}I}$ coinciding with $f_{cD\mbox{-}II}$. Data on the figure 3 showed increase in the later from zero to $0.5$ during D-puffing phase. At the same time $n_D/(n_H+n_D)$ reveals similar tendency. These results prove a sensitivity of ICE to the hydrogen isotope ratio. Some disagreement between the ICE and spectroscopy techniques in the pre-puffing phase appears due to the insufficient contrast of the deuterium registration channel for hydrogen scattered light.
The measurement of the isotope ratio is based on the assumption of the linear dependence of intensity of the ICE at fundamental harmonic on the density of emitting ions. In order to be sure in the validity of the assumption the model describing a mechanism of the ohmic ICE generation is necessary. Until now in laboratory experiments ICE was found to be driven by resonant interaction with high energy ion population produced by fusion products or injected beam ions$^{3,4}$. Non-monotonicity or anisotropy of ion distribution function in velocity or physical space is considered as an energy source for the instability$^5$. Since there are no auxiliary heating or fusion reactions in the considered scenario a desired mechanism could not relay on the features of the energy distribution of fast ions and other explanation should exist for the observed ICE.
Two possibilities are considered as possible mechanisms of the ICE generation in the absence of energetic ions. First, the ICE at the multiple harmonics of IC resonance can be generated as a result of Ion Cyclotron Drift Instability (ICDI) excitation in presence of an plasma inhomogeneity$^6$. According to $^6$ the condition of ICDI excitation in single component plasma is: $\rho_i/\alpha>2(m_e/m_i)^{0.5}$, where $\rho_i$ – ion Larmor radius, $\alpha=n/(dn/dr)$ – radial scale of the density gradient, $m_e$ and $m_i$ – electron and ion masses, respectively. Above criterion could be fulfilled in the experimental conditions of TUMAN-3M. In this case relationship between amplitudes of isotope fundamental harmonic depends on ICDI saturation level which will be further analyzed. Another possible mechanism of the ohmic ICE is a thermal emission of gyrating ions. In this case amplitude of fundamental harmonic linearly depends on concentration of the thermal ions and proportional to $\omega^2$ in low to medium dense plasma.
Presented study demonstrates sensitivity of the ICE driven by thermal ions to hydrogen isotope ratio and opens a way to develop novel diagnostic tools for measurement of the isotope ratio in contemporary and future devices.
Acknowledgement. The study was supported by RSF grant 16-12-10285 – development of diagnostic tools for ohmic ICE measurements and by Ioffe Institute – support of the tokamak infrastructure.
References
1 Lebedev S.V. et al, EPJ Web Conf., v.149(2017), Art.No #03010
2 Abdullina G.I. et al, Tech. Phys. Lett., v.45(2019), 8, 790
3 Bhadra D.K. et al, Nucl. Fusion, V.26(1986), p.201
4 Cottrell G.A. et al, Nucl. Fusion, V.33(1993), p.1365
5 Coppi B. et al, Phys. Fluids, V.29(1986), p.4060
6 Mikhailovsky A.B., Nucl. Fusion, V.11(1971), p.323Speaker: Dr Sergei Lebedev (Ioffe Institute) -
ADITYA tokamak has been upgraded to ADITYA-U tokamak which is equipped to have shaped plasma operations. The main structural modification is the replacement of an old rectangular cross-section vacuum vessel by a new circular cross-section vacuum vessel. New poloidal coils (six) have been installed in between new vessel and toroidal field coils [1]. Additionally, as per the design requirement, the in-vessel structure of graphite first wall material has been added more in upgrade tokamak for toroidal and poloidal limiters. Accordingly, the surface area of graphite (IG-430) has been increased > 2 m2. Thus the surface area of graphite as plasma-facing components (PFCs) is more than 11 % of the total in-vessel SS304L surface in ADITYA-U, while it was only 2.5 % in ADITYA. Due to graphite porosity, the fuel gases like Hydrogen and Helium retention stays potentially high due to wall conditioning procedures. As result, high Hydrogen recycling has been observed in tokamak plasma discharges mainly from graphite PFC, which disturbs the H2 gas fueling control. The carbon contain impurities are highly observed in ADITYA-U during wall conditioning procedures and tokamak plasma discharge in residual gas analyzer and spectroscopy diagnostics due to bulk graphite. To overcome these issues, different types of wall coating are carried out in fusion devices using low-z materials like lithium, boron, silicon to enhance plasma performance. Lithium coating on PFCs and vessel wall is a proven technique for controlling fuel recycling and high-z impurity in various fusion devices. A well-established technique in ADITYA, the lithium coating had been carried out using insertion of two lithium rods into the vessel during glow discharge wall conditioning (GDC) [2].This technique was effective in ADITYA due to less graphite PFCs compare to ADITYA-U. The same technique of Lithium coating hasn’t been successful in ADITYA-U, as very low Lithium-I line radiation is detected in spectroscopy diagnostics during tokamak plasma discharge after the coating. Thus, this Lithium coating is not effected in impurity control and Hydrogen recycling during plasma discharge. As per Carbon-III line radiation, the carbon burn-through period during plasma breakdown is significant high in ADITYA-U compare to ADITYA. Thus carbon impurity must be controlled in ADITYA-U using uniform and qualitative Lithium coating on PFCs.
Thus, we introduced and developed various Lithium wall conditioning techniques for ADITYA-U as (I) Heated Lithium rods insertion in H2-GDC (II) Lithium fueling by developed evaporator system (III) Combination of Lithium fueling and H2-GDC (IV) Combination of Ar-H2 mixture GDC then Lithium fueling and H2-GDC. The technique-I, an upgraded version of ADITYA Lithium coating, as Li-rod is heated by resistive heaters up to 130°C during H2-GDC (10-4Torr, 0.5-1.5A, 300-600V). The high temperature of Li-rod increases the Li- evaporation rate during GDC and we observed maximum 22000 counts of Li-I line spectrum in plasma shot no.33103 with single Li-rod. In fig.1, the effect of lithium coating is observed in the comparison of two plasma shots for ECR pre-ionization experiments with and without Lithium. The parameters in both shots are same as 100 kW ECR (42 GHz) power, low loop voltage 11V, TF 1.28 T, pressure: ~ 2 x 10-4 Torr. The major effect of Lithium wall conditioning is observed in the reduction of Carbon-III line and Visible Continuum as effective reduction of high Z impurity. As result, the plasma performance has been improved in current rise, long flat top and duration. The H-α and Oxygen-II are reduced in Li effected shot in the initial stage.
The physical vapor deposition in solid-angle of evaporated Lithium technique is widely used in other fusion devices. Here we have developed the technique-II of the Lithium evaporation system to control Li-fueling to ADITYA-U. The system has been tested up to 550° C on test-bench and installed successfully at ADITYA-U machine. The uniform and hard bonding of evaporated Lithium on PFCs in forms of Li2O, LiH, LiC qualify the Li-coating strength. In general, the Li counts have been decreased after each plasma shots in most of the tokamak experiments, which is main drawback of lithium coating. The Li mono-layers are inert for H2 molecules with physically deposited Li atoms, but Li is more reactive with H2 during GDC [3]. Thus in technique-III, Li-fueling plus H2-GDC creates more Li-H composition (lithium hydride) on the PFCs. The thermal decomposition temperature of Li-H is very high 688.7° C compare to Li melting temperature 180.5° C. As result, the Li wall conditioning effect sustains for long period of plasma operation compare to physically deposited Li atoms.
In technique-IV, this concept of Li-coating is novel for fusion machine. In ADITYA, the Ar-H2 mixture GDC had been carried out successfully to control Oxygen and Carbon contain impurity compare to traditional H2-GDC [4]. As observed in Ar-H2 GDC, the Oxygen and Carbon contained impurities like H2O, CO, CO2 are reduced more than 5 times within ten minutes compare to traditional H2 GDC. After Ar-H2 GDC, the Oxygen-contained impurities have been disappeared in near surface regions of PFCs and vessel wall for few hours. This phenomenon of Oxygen disappearance gives advantage to Li-coating. Therefore more Li-C and Li-H compositions are created on PFCs during Lithium fueling and H2-GDC, after Ar-H2 mixture GDC. The Pulsed GDC of H2, He, Ar-H2 mixture has been tested successfully in ADITYA-U for limited gas fueling with better wall conditioning [5].Therefore combination of all above techniques and pulsed GDC is also useful for better lithium coating. The design and development of all the techniques of lithium wall conditioning for ADITYA-U have been presented in this paper. The effect of various Lithium wall conditioning techniques on ADITYA-U plasma discharges has been studied in details using various diagnostics like visible spectroscopy, VUV spectroscopy for high-Z concentration, plasma parameters such as current, density, temperature, SOL probes for temperature and density, bolometer for radiation power [6].
References:
[1] Ghosh J. et al 2016 “Upgrade of ADITYA tokamak with limiter configuration to ADITYA Upgrade tokamak with divertor configuration” Preprint: 2016 IAEA Fusion Energy Conf. (Kyoto, Japan, 17–22 October 2016) FIP/P4-46
[2] S.B. Bhatt, et al., “Different Types of Lithium Coating in Tokamak ADITYA” IEEE Transaction in Plasma Science, Vol.40, No.6, June 2012
[3] H. Sugai et al. “Lithium wall conditioning for fuel and impurity control”, Journal Vacuum, Vol 47, pages 981-984, 1996
[4] K.A. Jadeja, et al. “Experimental Study for Comparison of H2 and Ar–H2 Gas Mixture Glow Discharge Wall Conditioning in ADITYA Tokamak”, IEEE Transactions on Plasma Science, Vol. 44, No. 4, 722, April 2016
[5] K.A. Jadeja, et al. “Novel approach of pulsed-glow discharge wall conditioning in the ADITYA Upgrade tokamak” Nucl. Fusion 59 (2019) 086005 (8pp).
[6]M. B. Chowdhuri, et al. “Improvement of Plasma Performance with Lithium Wall Conditioning in Aditya Tokamak” Plasma Science and Technology, Volume 15, Number 2
Speaker: Mr Kumarpalsinh Jadeja (Institute for plasma Research) -
Tokamak is a toroidal device where plasma is confined by means of appropriate magnetic field configurations. Tokamak plasma is home to several magnetic Instabilities, which can lead to loss of confinement and even termination of plasma delivering very high heat load on the plasma facing components 1. As these magnetic instabilities are dependent on the magnetic field configurations inside the plasma column, a complete understanding of the plasma current density profile, for which these magnetic field configurations appear, is of great importance. Several procedures are already been followed in several tokamaks world-wide to figure out the pattern of plasma current density profile, and several authors have come out with many interesting outcomes, however having a dedicated diagnostics for the same is a necessity 2. Limited by the non-availability of the dedicated diagnostics in ADITYA-U, the main motivation of our work is to use the data from a set of magnetic probes for having the idea of the plasma current density profile scenario. In this work a novel approach has been adopted about such estimations using the data of magnetic pick-up probes in case of tokamak plasma. This work provides a simple method which will be very useful for studying the implications of plasma current density profile on plasma confinement and stability.
This work addresses a rigorous study of the impact of radial current density profile in a toroidal current on the magnetic field at different locations with respect to the tokamak geometry, by means of numerical methods. This study is then extended for the tokamak plasma, where the radial current density profile is given by
j(r)=j_0 (1-r^2/a^2)^γ , (1)
where j_0is the central current density (at plasma centroid), a is the minor radius, r is the radial distance from plasma centroid, and γ is the exponent that decide the exact pattern of radial current variation.
For the validation of our numerical study, we take the data from ADITYA-U tokamak and the corresponding magnetic pick-up probes. ADITYA-U is a medium size tokamak with major and minor radii to be 75 cm and 25 cm respectively. There are two set of Mirnov probes that are installed in ADITYA-U and each of the sets contains sixteen number of pick-up probes, equally spaced on a poloidal plane as shown in Fig 1. This study include a generation of plasma current density profile by means of numerical processes and hence its effect on the pick-ups at these magnetic probes. Generating current density profile according to Eq. 1, magnetic fields are calculated at all sixteen probe locations component-wise. Interestingly, the spatial variation pattern of magnetic field at probe locations remains same with the change of γ, though magnitudes differ from each other at the high magnetic field side (inboard) and low magnetic field side (outboard) of tokamak. Fig. 2 describes such a spatial profile where magnetic field at probe locations are calculated numerically for different profiles and plotted against the probe number. This mismatch of magnetic field values of different plasma current profiles at the inboard and outboard locations gives a clear dependency of magnetic profile pattern on the current density profile and this gives a subtle way to estimate the same in case of actual plasma current.
Similarly, magnetic field profile during plasma discharge is achieved from experimentally acquired data and the magnetic field is extracted in such a way that it only account for the plasma column and this is then analysed for the experimental estimation of the current density profile. Finally, we get several interesting results, followed by some important conclusions about plasma properties of ADITYA-U tokamak. The temporal variation of radial current density profile in an individual discharge and also in different discharges have been compared with other dependent events to justify the estimation of radial current density profile using this simple yet very useful technique.
References:
1 Wesson J Tokamaks 1997 Clarendon Oxford.
2 Soltwisch H 1992 Plasma Physics and Controlled Fusion 34 (12) 1669-1698.Speaker: Suman Aich (Institute for Plasma Research) -
Efficient RF current drive is essential for developing a steady-state operation scenario in a tokamak. This paper investigates the impact of lower hybrid wave interaction with the tokamak boundary plasma on wave power deposition on the Alcator C-Mod and EAST tokamaks. The results presented here suggest that the presence of edge density fluctuations in a tokamak may need to be considered in understanding wave propagation and absorption observed in a present-day lower hybrid current drive experiment operating in a multi-pass damping regime. Even in a reactor regime that will operate in a single-pass damping regime, providing a quiescent edge and scrape-off-layer plasma in front of the antenna may be critical to mitigate the first-pass parasitic wave interactions with the boundary plasma. This quiescent edge condition may be achieved by wall conditioning or an optimization of launch location such as high field side launch in a double null configuration.
This paper hypothesizes that the central power deposition widely observed in the present-day LHCD experiments is due to wave scattering by turbulence and/or parametric decay instabilities {1,2]. In a standard model without considering such interactions, the predicted power deposition profile is generally peaked at off-axis, which is not in agreement with the experimental observation. A heuristic approach is adopted by introducing spectral broadening mechanisms in both the perpendicular wave-vector {3} and the parallel refractive index (n$_{//}$) spaces {4}. The ray-tracing /Fokker-Planck (GENRAY {5}/CQL3D {6}) solver is utilized within the $\pi$-scope {7} framework.
A focus is given to identify the ray components that can be absorbed by the central plasma on the first pass in order to reduce the sensitivity of the power deposition profile to changes in plasma condition. Figure 1 shows that ray damping characteristics can be altered with a change in the initial orientation of the wave perpendicular vector with respect to the surface normal direction. In a typical modeling, the wave-vector is assumed to be fully directed to the surface normal direction. As shown in Figure 2, a rotation of the perpendicular wave-vector by ~-20 deg at the initial ray condition is found to be effective to reproduce the centrally peaked current profile observed in the C-Mod experiment {8}. C-Mod is a small-sized, high-density tokamak, which can induce a large variation of the poloidal mode number even with a moderate rotation of the perpendicular wave-vector.
On the other hand, EAST operates at a low magnetic field and density with a high aspect ratio, compared to C-Mod. This provides a unique plasma condition in determining lower hybrid wave propagation and absorption in terms of wave accessibility and geometrical n$_{//}$ up-shifts. While a kinematic analysis implies that such a power deposition is prohibited at a typical n$_{//}$ of 2.1 {9}, LHCD experiments on EAST also indicate a central deposition of the 4.6 GHz wave power. An introduction of either wave-vector rotation or high n$_{//}$ components may be necessary to interpret the experiment. For example, without a rotation in the perpendicular wave-vector, Figure 3 shows a window of n$_{//}$~4 that is necessary for central propagation and absorption for a typical EAST plasma with n$_{e,0}$ = 3x10$^{19}$ m$^{-3}$ and T$_{e,0}$ = 2 keV. In this case with a low plasma temperature, the initial launch point for the high n$_{//}$ ray is taken to be below the midplane in order to induce an adequate amount of n$_{//}$ up-shift from the poloidal component {3}, but within the poloidal extent of the 4.6 GHz grill antenna of EAST. The antenna spectrum may possess these high n$_{//}$ sideband components. A detailed investigation of two spectral broadening mechanisms will be presented with an aim to identify strategies for providing a quiescent boundary plasma for an optimized RF coupling.
This work has been supported by an International Collaboration Grant No. DE‐SC0010492 from US Department of Energy, the National Key R&D Program of China (No. 2016YFA0400600), and the National Natural Science Foundation of China (Nos. 11675214, 11775259, and11975266).
{1} E. H. Martin et al, Nucl. Fusion 59, 076006 (2019);
{2} S. G. Baek et al, Phys. Rev. Letters 121, 055001 (2018);
{3} P. T. Bonoli et al, Phys. of Fluids 25, 359 (1982);
{4} Y. Peysson, Plasma Phys. Control. Fusion 58, 044008 (2016);
{5} A. P. Smirnov and R. W. Harvey, Bull. Am. Phys. Soc. 40 1837 (1985);
{6} R. W. Harvey and M. McCoy, The CQL3D Fokker-Planck Code Proc. of the IAEA Technical Committee Meeting on Simulation and Modeling of Thermonuclear Plasmas, Montreal, Canada (IAEA, Vienna, 1992);
{7} S. Shiraiwa et al., Fusion Engineering and Design 112, 835 (2016);
{8} R. T. Mumgaard, MIT Phd. Thesis (2015);
{9} X. Zhai et al, Plasma Phys. Control. Fusion 61, 045002 (2019)Speaker: S. G. Baek (MIT PSFC) -
Implementation of suitable disruption mitigation technique remains the topmost priority for larger tokamaks including the ITER. The spontaneous disruption in ITER may probably be an unavoidable part, while operated with high performance D-T fuel [1]. Disruptions in ITER could produce very large heat loads on divertor targets and other Plasma Facing Components (PFC), and large electromagnetic forces on the Vacuum Vessel (VV) can lead to structural damages [2]. In order to avoid these detrimental consequences, disruption mitigation is an essential part of tokamak research. The disruption mitigation system (DMS) in ITER is based on massive gas injection (MGI) of impurities, in order to radiate the plasma stored energy and mitigate the potentially damaging effects of disruptions [3]. Although, MGI based technique is suitable for most disruption mitigation, there will be issues in case of having warning time less than 10 ms as may be the case for the onset of some disruptions in ITER [4]. This is due to the slow thermal velocity of the heavier impurity gas molecules, which limits the time needed to travel the distances before it reaches the plasma edge. At present, none of the currently planned disruption mitigations systems for ITER can respond on this time scale. The radiative dissipation of the plasma stored energy during major disruption in ITER by fast injection of massive pellets of low Z impurities, such as Li and Be pellet injection has been numerically modelled by Lukash et al. [5].
A novel concept of electromagnetically driven particle Injector (EPI) having fast time response has been developed and installed on the ADITYA-U tokamak. For the first time, ADITYA-U has experimentally demonstrated the use of electromagnetic pellet injection for firing the pellets into the tokamak plasmas for disruption mitigation studies. This is a significant development towards disruption control that can accomplish the critical need of disruption mitigation in future tokamaks including ITER. The experiments are conducted in ADITYA Upgrade (ADITYA-U), which is a medium sized (R0 = 75 cm, a= 25 cm) heated air core toroidal limiter tokamak, capable of producing circular as well as shaped plasma with single and double null open divertor configuration [6].
In these experiments, a projectile containing a radiative payload, weighing ~ 50 - 200 mg, is accelerated towards the core of the plasma with velocities of ~ 200-220 m/s using a linear coil gun accelerator, which can deposit micron-sized dust deep into the plasma core and radiate whole the plasma stored energy within few milliseconds. Lithium titanate (Li2TiO3) and Lithium carbonate (Li2CO3) particles of 50 – 80 micron are used in the experiments. Before performing the disruption mitigation experiment in ADITYA-U, the machine has been prepared for this experiment. The standard reference discharges of IP ~120 kA of 100 -120 ms duration have been established. Later, lithium titanate (Li2TiO3) impurity particles pellet has been injected into the ADITYA-U plasma during the plasma current flat-top at ~ 51.6 ms. The impurity particles reached the ADITYA-U core plasma within ~ 1.25 ms and causes fast termination of plasma current. The time evolution of two consecutive discharges of ADITYA-U, black curve (#33317) with particle injection and red curve (#33318) without pellet injection are shown in Figure 1, representing the radiative dissipation of the plasma stored energy in the discharge in which particles are injected. The spectroscopy measurements of Li line at 670.8 nm as shown in Fig. 2, confirms the signature of Li2TiO3 injection into the plasma. The chord averaged electron density and electron temperature, terminates very rapidly (Fig. 1(b) and 1(c) respectively), due to the increased of plasma radiation (Fig. 1(d)) after the impurity particles injections. The time evolution of relative intensity of visible radiation at 536 nm with and without particle injection for similar discharges during the disruption phase is shown in Figure 3. The Figure shows significant increase in the radiation level during disruption for the shot having particle injection.
During the experiment, fast visible imaging, high resolution video cameras on two different radial ports were installed on the ADITYA-U machine, to obtain the plasma images at high spatial and temporal resolution. One of the camera was kept for 2D tangential viewing, covers the entire cross-section of the limiter and the other camera covers the entire poloidal cross-section of the port, from where the impurity particles were injected into the plasma. The images clearly showed that the injected impurities radiate the energy from the core plasma and disrupted the whole plasma column within ~ < 2 ms. Furthermore, the disruptions triggered with particle injection using electromagnetically payload and disruptions triggered by massive gas-puff have been compared. Significantly different current quench characteristics have been observed in the disruptions triggered by particle injection, which occurred on faster time-scales as compared to those triggered by massive gas-injection. First results on disruptions in the ADITYA - U by electromagnetically driven fast injection of particles in to the plasma will be discussed in this paper.
References
[1] Strait E.J. et al 2019 Progress in disruption prevention for ITER Nucl. Fusion 59 112012.
[2] Lehnen M. et al 2015 Disruptions in ITER and strategies for their control and mitigation J. Nucl. Mater. 463 39-48.
[3] Hollmann E. et al 2015 Status of research toward the ITER disruption mitigation system Phys. Plasmas 22 021802.
[4] Raman R. et al 2015 Fast Time Response Electromagnetic Disruption Mitigation Concept Fusion Sci. and Tech. 68:4 797-805.
[5] Lukash V.E. et al 2014 Modeling of Major Disruption Mitigation by Fast Injection of Massive Li Pellets in ITER Like Tokamak-Reactor IAEA (2010) http://www-naweb.iaea.org/napc/physics/fec/fec2010/html/node1 91.htm#41805 (current as of Nov. 19, 2014).
[6] Tanna R. et al 2019 Overview of operation and experiments in the ADITYA-U tokamak Nucl. Fusion 59 112006.
Speaker: Joydeep Ghosh (Institute for Plasma Research) -
During non-local (“global”) L-H transitions found in various regimes of JET and JT-60U tokamaks earlier [1-3], the rise of Te,i and ne starts simultaneously in the spatial zone ≈0.3< r/a <1, while heat and density fluxes fall simultaneously in the same region. At the ITB-events in JT-60U and T-10 (circular tokamak with a limiter =30 cm, R=150 см, Вт= 2.5 Т), heat and density fluxes fall in a narrower internal spatial zone within ≈40-50% of the minor radius, see detail in [3-5].
The present report describes the new type of L-H transitions in plasmas with W-limiter and Li-coating called “semi-global”. Te starts to rise simultaneously at 0.2 r/a<0.6 similar to its behavior during an ITB-event in JT-60U and T-10. The density rises in a wider zone ≈0.3 < r/a <1 similar to a non-local L-H transition. The electron heat flux abruptly reduces in the spatial zone 0.2 < r/a < ≈ 1. The gradual formation of ITB occurs after transition. The rise of Te in the internal part of plasmas was mentioned at L-H transitions in circular tokamaks with the limiter TUMAN-3, JIPP T-IIU and TEXT-U [6-8] (light cooling of the periphery, like in our cases, was mentioned also). The formation of ITB measured with Thomson scattering was reported in [6]. Nevertheless, analysis of the non-locality of the electron heat flux jump was not done in [6-8].
The one spontaneous semi-global L-H transition was briefly described by us at EC-2018 workshop [9]. The spontaneous transitions (including dithering one) are observed at simultaneous co+contr ECCD by 2 gyrontrons only (total power 1.4-1.5 MW). The presence of dithering transitions (up to 5 H-mode phases with 5-10 ms length) means that the power is near critical value. The figure 1(a) shows the typical evolution of Te(r,t), line-averaged central density and energy content W at spontaneous and triggered semi-global transitions. In this particular case, neon puffing starts 5 ms before transition (PECRH=0.85 МW, see experiments [10]) and puffing starts 15 ms before transition in the regime with higher density and EC-power [11]. The plasma-wall interaction falls (D-beta level drops), Te starts to rise at 0.2 < r/a<0.6 together with the rise of density and W. The density at the edge increases 30% faster compare with that of in the bulk plasmas. The transition triggered by neon puffing is not a transition to RI mode since the level of radiation losses is small enough (below ~25%) and just starts to rise before the transition. Figure 1 (b) represents the typical profile of the electron heat flux delta(G Te ) at the transitions in all cases. In this case the transition at PEC=1.4 MW and co+co ECCD, was caused by spontaneous drop of Li-containing flake (jump of flux is constant in space up to the edge if one believe to ECE data at the periphery). In a contrast with [1-3], the rise of density is important at calculations of delta(G Te ) profile at transitions on T-10 by simple expression [1-5]: delta(G Te (r))S(r)= 3/2{volume integral of delta[d(nTe)/ dt]}, S(r) – enclosed surface and delta[d(nTe)/ dt] is the jump of derivative at the time of transition with the evolution of delta(n(r,t)) calculated from data derived by 16 channels interferometer. The transport gradually varies in space and time after transition and leads to the formation of ITB as one can see on the figure 1(b) (typical case). The similar case in the plasmas with co+contr ECCD has been described recently in [12].
The authors of the paper [13] show that the rise of central Te at the injection of small carbon pellets at ASDEX-U is explained by strong stabilization of the trapped electron mode turbulence due to the measured fast significant flattening of the density profile in the internal part of plasma column. This explanation is not valid in our cases since the density just starts to rise gradually after transition (even at the drop of flake the density varies abruptly at the edge only).
The paper describes the new type of transition called “semi-global” L-H transition. Te starts to rise simultaneously at 0.2 < r/a<0.6 similar to its behavior during an ITB-event in JT-60U and T-10. The density rises in a wider zone ≈0.3 < r/a<1 similar to a global L-H transitions at JET and JT-60U and the electron diffusivity falls in the same zone. The electron heat flux abruptly reduces in the spatial zone 0.2 < r/a< ≈ 1. The spontaneous transitions (including dithering one) are observed at simultaneous co+contr ECCD by 2 gyrontrons only. New triggers of the transitions are neon gas puffing and spontaneous drop of the Li-containing flakes in various regimes of ECCD and EC-power (observed at n line av (0) < 3 10 19/m3, Bt =2.3-2.5 T, Ip=200-250 kA). By our up to date knowledge, transitions with the non-local reduction of heat flux has never been reported in the tokamaks with limiter. An abrupt increase of energy confinement time tau-E at the moment of the transition is equal to 10-20% (small value of H-factor is typical at circular tokamaks). The H mode continues up to the double value of tau-E. The accumulation of impurities is absent. The part of Te rise can be explained by the abrupt reduction of the convective electron heat flux 2.5TeGn due to the decay of electron density flux Gn at the time of transition.
The authors thank Drs A.Ya Kislov, D.A. Kislov, N.A. Kirneva and many other members of the T-10 team for fruitful discussions and fine collaboration. The work was supported by ROSATOM Corporation.
References
1. Neudatchin S V, Cordey J G and Muir D J 20th EPS Conf. on Control. Fus. and Plasma Phys. (Lisboa) vol. I (Geneva : EPS), p 83 (1993)
2. Neudatchin S V, Takizuka T, Shirai H et al., Japan J. Appl. Phys. 35 3595 (1996)
3. Neudatchin S. V., Takizuka T., et al., Plasma Phys. Control. Fusion 44 A383-389 (2002)
4. Neudatchin S V., Inagaki S, Itoh K et al. 2004 J. Pl. and Fus. Res. Series 6 134
5. Neudatchin S.V, Shelukhin D.A., Mustafin N.A. 2017 J. Phys.: Conf. Ser. 907 012015
6. Andrejko M V, et al 1992 IAEA_PPCNFR_Wuerzburg_v1_p485 IAEA-CN-56/A-7-11
7. Watari T., Kumazawa R. et al 1990 Nucl. Fusion 30 1197
8. Roberts D. R. et al 1996 Plasma Phys. Control. Fusion 38 1117
9. Borshegovskiy A, Neudatchin S, Pimenov I et al 2019 EPJ Web of Conf. 203 02004
10. Kasyanova N.V., Rasumova K.A. et al, . 2018, in Procs. of 45th EPS Conf. on Pl. Ph, Prague, ECA, Vol 42A, P4. 1106
11. Kirneva N.A. et al, 2018, 45th EPS Conf. on Pl. Ph., Prague, ECA, Vol 42A, P4. 1081
12. Neudatchin S. V., Pimenov I. S, et al., 2019, J. of Ph.: Conf. Ser. 1383 012005
13. Angioni C. et al 2019 Nucl. Fusion 59 106007
Speaker: Dr Sergey Neudatchin (NRC Kurchatov Institute) -
Reaching good-quality H-mode and the development of ELM control techniques are among the main priorities for ITER during its non-active operations [1']. Recent encouraging experimental results at JET-ILW demonstrated a significant reduction in the H-mode power threshold for NBI-heated plasmas when a small amount of $^{4}{\rm He}$ ions, $n(^{4}{\rm He})/n_{e} \approx 10\%$, was added to hydrogen plasmas [2']. This finding motivated the ITER team to consider the use of H-$^{4}$He plasmas to widen the H-mode operational space in predominantly hydrogen plasmas [3]. Equally important, this mix also allows an application of the 3-ion ICRF scheme [4] with the off-axis heating of $^3$He minority ions, $n(^{3}{\rm He})/n_{e} < 1\%$, as proposed in [5], capable of additionally delivering up to 20MW of heating power in H + 10% $^4$He plasmas in ITER.
In this contribution, we report the results of physics studies on the ASDEX Upgrade (AUG) tokamak, in which the above mentioned ITER-relevant heating scheme was recently prototyped. The high efficiency of the 3-ion scheme for plasma heating was proven by applying an ICRF power ramp to increase the plasma stored energy and trigger L-H transitions in H-$^4$He plasmas (2.5T/0.8MA), see Fig. 1. Different combinations of the heating systems were successfully used to enter H-mode on AUG, including NBI+ECRH+ICRF (as in ITER), ECRH+ICRF, and ICRF-only. Similar to the earlier observations in H-$^4$He plasmas with hydrogen NBI and ECRH heating on AUG [6], the L-H transitions were reached at $P_{\rm LH} \approx 2-3\,{\rm MW}$ ($n_{e0} \approx 4 \times 10^{19}\,{\rm m}^{-3}$). Another important result of our studies on AUG is the demonstration that the 3-ion ICRF scheme with off axis $^3$He heating is compatible with avoiding tungsten accumulation. The 3-ion ICRF scheme with core $^3$He heating was also applied on AUG, resulting in new and interesting fast-ion physics effects, including the first observation of $^3$He-driven ion cyclotron emission on this tokamak.
![AUG has prototyped the heating scenario in non-active H-$^4$He plasmas with NBI, ECRF and ICRF systems that has a potential to widen H-mode operational space in ITER [3]. An ICRF power ramp using the 3-ion scheme $^4$He-($^3$He)-H was applied to trigger L-H transitions on AUG (2.5T/0.8MA).](https://i.ibb.co/x8BRXWh/Synopsis-Figure01-small.jpg)
In JET-ILW, we further advanced the 3-ion ICRF technique for heating mixed plasmas using injected fast NBI ions as a resonant ion component. Following the success of studies in H-D plasmas [7, 8], a controlled acceleration of D-NBI ions to higher energies with ICRF was recently demonstrated in mixed D-$^3$He plasmas. Figure 2 (a) shows an overview of JET pulse #94701 (3.7T/2.5MA), in which $6\,{\rm MW}$ of ICRF was applied in combination with $8\,{\rm MW}$ of NBI. The neutron rate was increased from $0.6 \times 10^{15}\,{\rm s}^{-1}$ in the NBI-only phase to $1.0 \times 10^{16}\,{\rm s}^{-1}$ in the combined ICRF+NBI phase of the pulse. For comparison, we also illustrate the plasma performance for the NBI-only pulse #94704 (characterized by dominant ion heating in the plasma core), which was performed at the same operational conditions as #94701. The stronger $T_{e}$ peaking and much higher $T_{e0}$ and neutron rate in pulse #94701, as compared to #94704, are easily understood by the presence of high-energy D ions due to efficient central deposition of ICRF power. However, the comparison of the measured plasma stored energy, $T_{e}$ and $T_{i}$ profiles at the same total auxiliary heating power ($P_{\rm aux.} = 14\,{\rm MW}$) and plasma density ($n_{e0} \approx 6 \times 10^{19}\,{\rm m}^{-3}$), as shown in Figs. 2 (b) and (c), hints at a significantly reduced transport in pulse #94701, characterized by a large fraction of fast ions in the plasma core. The analysis of pulse #94701 using the gyrokinetic code GENE is ongoing, aiming to understand better the transport in JET plasmas under the ITER-relevant conditions of dominant fast-ion electron heating in the plasma core and identify possible fast-ion effects on microturbulence. The observation at JET that in such plasmas $T_{i} \approx T_{e}$ is very promising for ITER, and is in line with recent theoretical studies showing that alpha particles can significantly stabilize ITG turbulence and reduce heat transport in ITER [9]. Furthermore, the developed 3-ion ICRF scheme in mixed D-$^3$He plasmas was used as a tool to generate alpha particles and validate successfully updated JET diagnostics for alpha measurements prior to future D-T operations [10].
The discussion of recent experimental studies on AUG and JET is complemented by illustrating the synergy between experimental and modeling developments in this field of plasma physics [8, 11]. An outlook of possible applications of the 3-ion schemes in JET and ITER is discussed, including promising schemes for the demonstration of alpha particle effects in the upcoming D-T campaign on JET [12], in particular, electron heating by alpha particles. While further more detailed analysis work remains to be done, e.g., better understanding the interaction of the fast ions with the plasma, MHD modes and turbulence, the results obtained recently on JET and AUG confirm the high efficiency of the novel 3-ion ICRF schemes for plasma heating and increase our confidence in extrapolating the application of these schemes to ITER.
Acknowledgements. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
[1'] B. Bigot et al., Nucl. Fusion 59, 112001 (2019)
[2'] J.C. Hillesheim et al., Proc. 44th EPS Conf. on Plasma Physics, P5.162 (2017)
[3] ITER Research Plan, report no. ITR-18-003 (2018)
[4] Ye.O. Kazakov et al., Nature Physics 13, 973-978 (2017)
[5] M. Schneider et al., EPJ Web. Conf. 157, 03046 (2017)
[6] U. Plank et al., Proc. 46th EPS Conf. on Plasma Physics, O2.111 (2019); also: in preparation for submission to Nucl. Fusion (2020)
[7] J. Ongena et al., EPJ Web. Conf. 157, 02006 (2017)
[8] M.J. Mantsinen et al., Proc. 46th EPS Conf. on Plasma Physics, O5.102 (2019); also: this conference
[9] J. Garcia et al., Phys. Plasmas 25, 055902 (2018)
[10] M. Nocente et al., this conference
[11] A. Kappatou et al., Proc. 45th EPS Conf. on Plasma Physics, O2.105 (2018)
[12] R. Dumont et al., this conferenceSpeaker: Yevgen Kazakov (Laboratory for Plasma Physics, LPP-ERM/KMS) -
Introduction -- Power exhaust solutions for a next step device must be compatible with good plasma performance. To reach sufficiently low divertor power loading impurity seeding is necessary. The amount of injected impurities required depends critically on both the maximum achievable separatrix density and the scrape-off layer width. Transient power loads due to type-I edge localised modes (ELMs) are expected to reduce the life time of the first wall in ITER and have to be avoided completely for reactor-sized devices.
Recent optimisation of ASDEX Upgrade and TCV scenarios [Ref.1,Ref.2] has led to a regime highly suitable for power exhaust. This promising regime combines a high plasma core performance ($H_{\mathrm{98,y2}} \simeq 0.9\,-\,1.0$, $n_{\mathrm{e,core}} \simeq 0.9\,n_{\mathrm{GW}}$) with a high separatrix density ($n_{\mathrm{e,sep}} \simeq\,0.4\,n_{\mathrm{GW}}$) at high triangularity, close-to-double-null. The most critical power exhaust parameter, the power fall-off length measured from divertor heat flux profiles in ASDEX Upgrade, is shown to widen up to a factor of four w.r.t. the ITPA-multi-machine (Eich) scaling of four whilst no type-I ELMs are present.
Edge characterisation -- The edge plasma reaches a ballooning parameter close to the ideal-MHD limit. HELENA calculations show that a small region just inside the separatrix is infinite-n ballooning unstable. Ballooning modes, which are generally driven by strong MHD-normalised pressure gradients and stabilised by magnetic shear, are destabilised in these discharges by the strong pressure gradient at the separatrix and the reduced local magnetic shear due to the close-to-double-null configuration. The onset of ballooning modes cause increased radial transport importantly at the pedestal foot and narrow the radial extent of the region with a steep pedestal pressure gradient. Peeling-ballooning stability calculations for these reduced pedestal widths show that type-I ELMs are avoided under these conditions.
Divertor characterisation -- Fuelling and seeding variations in dedicated experiments allow at ASDEX Upgrade to change the divertor state from high recycling to partial detachment, as foreseen for ITER {\em Q=10} operation. Fuelling from main chamber valves leaves the divertor attached, while detaching with divertor fuelling not changing the confinement properties.
In 2019, TCV has explored operation with a baffled divertor [Ref.3], separating the neutral pressure in the main chamber from the divertor neutral pressure. Under these conditions, the here mentioned regime can be recovered only if the fuelling from the divertor is increased by a factor ~3. The operational window for the regime has been successfully extended down to $q_{\mathrm{95}}$ values around 3.7.
For ASDEX Upgrade, we report in detail on the divertor target power load profiles. They are measured by high resolution infrared thermography. The flexibility to vary between an attached high recycling and detached divertor state allows to measure directly the scrape-off layer power fall-off length. Fig.1 shows outer divertor target heat flux profiles for phases with lowest fuelling (red, type-I ELMs present) and two levels of elevated fuelling (blue and green, no type-I ELMs present). Most importantly, a broadening of the power fall-off length is measured, up to a factor of four compared to the ITPA-multi-machine scaling [Ref.4]. These observations are consistent with the increased radial transport at the pedestal foot. The ballooning instabilities lead to a radial convection of filaments carrying the heat much further into the scrape-off layer and may give rise to the formation of a density shoulder [Ref.5]. Direct evidence stems from thermal helium beam analysis [Ref.6] for the scrape-off layer showing both an increased filament frequency and amplitude.
Integrated scenario -- A plasma start-up into a high confinement, partially detached H-mode without any type-I ELM is realised for the first time at ASDEX Upgrade. Fig.2 shows time traces for ASDEX Upgrade discharge # 37164. The H-mode is initiated at high $q_{\mathrm{95}}\,=\,7.8$ allowing easier access to a type-I ELM free regime. A ramp-up in plasma current after achieving the final close-to-double-null shape is performed eventually reaching $q_{\mathrm{95}}\,=\,4.6$. Additionally, for the first time, a double feedback controlled discharge is achieved using neutral beam injection to control $\beta_{\mathrm{pol}}\,=\,1.3$ and nitrogen seeding to control the divertor electron temperature of $T_{\mathrm{div}} = 10\,\mathrm{eV}$ (3-5s) and later of $T_{\mathrm{div}} = 6\,\mathrm{eV}$ (5-7s).
References
[Ref.1] HARRER, G. et al., Nuclear Fusion 58 (2018) 112001.
[Ref.2] LABIT, B. et al., Nuclear Fusion 59 (2019) 086020.
[Ref.3] THEILER, N. et al., this conference.
[Ref.4] EICH, T. et al., Nuclear Fusion 53 (2013) 093031.
[Ref.5] VIANELLO, N. et al., this conference.
[Ref.6] GRIENER, M. et al., Review of Scientific Instruments 89 (2018) 10D102.Speaker: Dr Michael Faitsch (Max-Planck-Institut für Plasmaphysik) -
Introduction – Achieving safe power and particle exhaust compatible with a high-performance core plasma is one of the main challenges towards commercial fusion power. Currently, the most promising solution is to operate a diverted tokamak in detached conditions with high divertor neutral pressure, high volumetric power dissipation, and a strong temperature and pressure gradient along the magnetic field towards the target plates [1]. Access and performance of detachment strongly depend on the divertor plate and wall geometry and the ability to confine neutrals and impurities to the divertor region. Important additional benefits such as reduced plasma and impurity density thresholds for detachment and/or passive stabilisation of the detachment front are expected from non-conventional divertor magnetic geometries.
Validation of current models and an improved understanding of the different mechanisms involved require experiments that separate magnetic topology from neutral trapping through wall geometry. To meet these needs, removable neutral baffling structures were installed on the Tokamak à Configuration Variable (TCV) [2] and a first experimental campaign was completed. Combined with the extreme magnetic shaping capabilities and substantial upgrades in heating power and divertor diagnostics, TCV now provides a unique testbed to disentangle the effects of magnetic topology from neutral effects and to validate state-of-the-art simulation tools.
Divertor closure and detachment threshold – The results of recent Ohmic L-mode density ramp experiments in a Single Null reference shape ($I_p$=250kA), comparing operation with and without the new baffles, confirm the main SOLPS-ITER predictions of how the baffles change divertor performance: a) an increase in divertor neutral pressure by a factor 2-5 for the same upstream conditions; and b) a $\sim$ 30% reduction of the upstream density threshold for detachment (Fig. 1). To further increase divertor closure and possibly detachment access, the position of the plasma with respect to the baffles was varied. This revealed that the reference baffle location, chosen based on modelling, is already close to the optimum; shifting the plasma closer to the baffles resulted in a substantial recycling flux from the baffles and a drop in divertor neutral pressure. In all these plasmas, the divertor neutral pressure is found to be a good parameter to describe the divertor state, with detachment starting at $\sim$ 0.15 Pa, irrespective of the upstream density, divertor closure, and fuelling location.
Detailed SOLPS-ITER drift simulations of these plasmas have been carried out and results are being compared with an extensive set of measurements, including two-dimensional profiles of density, temperature, parallel flow, and plasma potential measured across the entire outer divertor leg with a reciprocating divertor probe array housing 12 Mach probes. This comparison reveals a qualitative agreement between the simulated and measured target currents and their variation with density and drift direction. The experiments also confirm the model prediction of a negative potential well near the X-point in reversed field, and quantitatively match the predicted change in neutral pressure and detachment threshold with the baffles. The reasons for remaining differences with modelling are being investigated and benchmarked with simulations from the SolEdge2D-EIRENE code.Detachment in H-mode plasmas – Even stronger effects of the baffles are observed in H-mode experiments. The L-H power threshold is reduced substantially with the baffles and, for a given upstream density and level of neutral beam heating power, pedestal pressure and stored energy increase by $\sim$30%. The operational window for H-mode detachment also increased strongly with the added baffles: In Type-I ELMy H-mode plasmas at 170kA and with 1MW of neural beam heating, the outer target is close to detachment between ELMs, with target electron temperatures in the range of 5eV, rather than $\sim$20eV as in comparable unbaffled cases. Seeding nitrogen as impurity results in a larger than three-fold reduction of the outer target ion saturation current. This clear detached divertor state is corroborated by the inter-ELM CIII emission front receding smoothly up the divertor leg to the X-point, with no indications for a bifurcation-like transition as on other devices [3].
Divertor closure and alternative divertors – The compatibility of the new TCV baffles with a large range of alternative divertor geometries has been demonstrated experimentally, including Super-X, Snowflake, and X-Divertor geometries, and experimental and theoretical studies shed light on the separate effect of divertor closure and magnetic topology. In particular, recent SOLPS-ITER simulations [4] reproduce previous TCV Super-X results without baffles [5]. The modelling further reveals that the effect of a large strike-point major radius $R_t$ (the main Super-X property) can be counteracted by a simultaneous change in divertor neutral trapping. Both, strong baffling and a fixed poloidal incidence angle $\beta$ of the divertor leg, were required in the simulations to equalise neutral trapping between different geometries and to recover the expected detachment threshold reduction at large $R_t$ [4]. Motivated by this, $R_t$ and $\beta$ were varied in experiments with baffles from 0.62-1.05m and 50$^{\circ}$-130$^{\circ}$, respectively (examples in Fig. 1). This reveals a strong dependence of the detachment threshold on $\beta$ for $\beta$>90$^{\circ}$, yet a relatively weak dependence on $R_t$ and unexpected changes in target profile shapes with $R_t$, suggesting an importance of drift or turbulence effects.
Ongoing work is assessing the reliability of edge transport codes to reproduce these experimentally identified dependencies on magnetic topology and wall geometry and how they may extrapolate to higher power conditions. These investigations help to better understand the different effects which play a role for detachment and divertor-core compatibility and will allow for more reliable predictions for future devices.[1] Leonard et al., Plasma Phys. Control. Fusion 60, 044001 (2019)
[2] Fasoli et al., Nucl. Fusion 60, 016019 (2020);
[3] Jaervinen et al., Phys. Rev. Lett. 121, 075001 (2018)
[4] Fil et al., Plasma Phys. Control. Fusion 62, 035008 (2020)
[5] Theiler et al., Nucl. Fusion 57, 072008 (2017)Speaker: Christian Theiler (Ecole Polytechnique Fédérale de Lausanne (EPFL) – Swiss Plasma Center (SPC)) -
Power exhaust is one of the big challenges for future fusion reactors. In the EU programme, both conventional and alternative divertor approaches are studied. For a conventional divertor in EU-DEMO (1), more than 80% of the exhaust power needs to be dissipated before entering the SOL to keep the divertor in the detached regime, where the interaction of the plasma with the wall is significantly reduced (2). As the detached state needs to be sustained, it is essential to implement a real time control scheme which also allows for a sufficient margin to operate the device.
Radiation is the dominant energy dissipation process in a reactor relevant tokamak plasma. It can be increased by seeding of impurity gases. If the complete detachment is achieved by the injection of impurities into tokamaks with a full metal wall (such as ASDEX Upgrade (AUG) or JET), the divertor radiation concentrates in a small region at or above the X-point inside the confined region (3)(4), see Fig. 1. This so-called X-point radiator can be induced at AUG through nitrogen or argon seeding and presents within boundaries a stable regime of operation. This is contrary to carbon wall operation in which X-point radiation or a so-called MARFE is usually observed to lead to disruptions. Hence, in the operation with a tungsten wall, a reliable control of the X-point radiator is crucial for the success of the conventional divertor solution.
The vertical extent of the X-Point radiator at AUG is about 5cm and it can move further into the confined region; the vertical distance of the radiating region to the X-point depends, among other parameters, on the impurity influx and heating power. The radiator can be stably held up to 15cm above the X-point, corresponding to a normalized flux surface radius of ϱpol≈0.99.
In order to test the controllability of this regime and to gain a deeper understanding of the physics behind the X-point radiator, a real time control scheme has been implemented in AUG. The controller utilizes an array of poloidally-viewing AXUV diode bolometers to detect the location of the radiation peak in the vicinity of the X-point. The nitrogen seeding level is used as the actuator to move the radiation region either further into the core plasma or back towards the X-point, see Fig. 2. The details of the feedback control system and effects on the divertor and core plasma will be presented.
The new controller shows for the first time the active control of a fully detached state in H-mode. This will be compared with other existing detachment controllers, for example the control of the onset of detachment using the CIII emission front at TCV (5), and the possible application for future devices.
When the radiator penetrates strongly into the confined region (reaching 7cm above the X-point, which is at the foot of the pedestal at ϱpol<0.997), it is observed that the pedestal stability is modified, leading to reduced pedestal pressure gradients but similar performance further inside the confined plasma (4). In this regime, ELMs are suppressed (see bottom graph of Fig. 2). The ELM suppression avoids the intermittent heat fluxes of ELMs into the divertor, which would otherwise lead to a transient re-attachment of the divertor and significantly reduce the lifetime of a divertor in future reactors. This regime is currently being examined and will be discussed further.
The movement of the X-point radiator provides an operational margin between the beginning of complete detachment, where the radiator develops at the X-point, and a radiative collapse, where the radiator penetrates too far into the confined plasma. Thus, if it is possible to operate a future reactor in this regime without having detrimental effects, e.g. on the energy confinement, a real time controller could be implemented to regulate the position of the radiator and ensuring stable operation with a detached divertor at maximum radiated power fractions, possibly already incorporating ELM mitigation.
These experiments open up the possibility to reliably reduce the power flux across the separatrix such that a conventional divertor solution may be possible in EU-DEMO. At the same time, they also alleviate the challenge for alternative divertors.(1) G. Federici et al, Fusion Eng. Des. 136 (2018) 729–741
(2) M. Wischmeier et al, J. Nucl. Mat. 463 (2015), 22
(3) M. Bernert et al, Nucl.Mater.Energ. 12 (2017), 111-118
(4) F. Reimold et al, Nucl. Fus. 55 (2015), 033004
(5) T. Ravensbergen et al, EPS 2019Speaker: Matthias Bernert (Max-Planck-Institut für Plasmaphysik) -
A promising repeatable laser system producing multi-kilojoule of pulse energy has been basically designed for realization of the fast-ignition-based inertial fusion energy (IFE) reactor. Two cultivated core key technologies ensure high reliability of the proposed design. First, using our novel bonding technology, a cryogenic active-mirror amplifier has been developed to enable 100 Hz repeatable operation at 100 J of pulse energy. Kilojoule energy can be easily obtained for the fuel compression by combining a number of the amplifiers into a single beam. The other is the high-contrast, petawatt laser technology for heating laser: LFEX-laser demonstrated 2 PW, 2 kJ, 1 ps with a high intensity contrast more than 1 x 10^10.
Understanding of the central ignition scheme of IFE reveals significant difficulties of the ignition achievement through many experiments and theoretical calculations by using a number of kilo-joule class beam lines at the huge facilities in National Ignition Facility (LLNL) and OMEGA facility (LLE, Univ. Rochester) and so on. Recently, at Osaka university, the highly efficient production of an ultra-high-energy-density plasma of 2.2 Peta-Pascal has been successfully demonstrated with the fast ignition scheme by using the 2 PW LFEX-laser.[A] This remarkable progress of the fast ignition research accelerates realization of a practical IFE reactor based on fast ignition scheme. However, a repeatable high power laser system, which is an absolutely essential technology of the IFE reactor, is still missing. In a latest fusion reactor design, as multiple moderate-in-size reactors operate together, the required repetition rate for the laser system comes up to 100 Hz. So far, all high-energy laser systems used for the IFE studies, including the above lasers, are a single shot laser, and there is no repeatable laser system in kilo-joule class over the world. Serious bottlenecks for repeatable laser operation are thermal effects caused by large heat loaded in the laser material of the energy amplifier, wave front distortion, induced birefringence, and material fracture. The first application of ceramic active-mirror to high energy fusion laser as a laser amplifier reduces these effects dramatically to enable repeatable operation.
The active-mirror laser amplifier shows, in principal, two excellent capabilities of a high heat removal due to conductive cooling with a metal heat sink and a low wave front distortion due to the heat flow direction parallel to the laser propagation, shown in Fig. 1. In addition, a ytterbium-doped YAG (Yb:YAG) ceramic is used as a laser material instead of the conventional neodymium-doped laser glass (Nd:glass). The high thermal conductivity of the YAG ceramic is about one-order of magnitude higher than that of Nd:glass, this high thermal conductivity reduces the material temperature rise. Moreover, the Yb:YAG ceramic is cryogenically cooled to increase the thermal conductivity further. A cryogenic active-mirror has been already demonstrated at high repetition rate of a few hundred hertz, the aperture size was, however, as small as 1 cm class and the pulse energy was around 1 J.[B,C] A thermally strong bonding of the ceramic to the metal heat sink is considerably difficult. And to make matters worse, as the cryogenic operation temperature is quite different from the bonding temperature, an intense internal stress of the ceramic is not avoidable in several-centimeter aperture size for more than 10 J due to a different thermal expansion between the ceramic and the metal. By developing a novel bonding technology, which is under patent pending, to relax the stress, 7 cm x 7 cm active-mirror has been successfully bonded and 10 J high pulse energy has been obtained at 10 Hz with the four diode-pumped cryogenic active-mirrors, for the first time, in Fig. 2. A higher repetition rate of 100 Hz will be demonstrated and a higher pulse energy of 100 J will be upgraded by enlarging the beam aperture size in the near future.
Another core key technology is the petawatt laser technology. The kilojoules chirped-pulse amplification (CPA) has been already developed in our LFEX-laser at 2 PW, 2 kJ, 1 ps to supply pulses for the plasma experiments. An excellent intensity contrast of more than 1 x 10^10 is obtained with the original pulse cleaner in Fig. 3. The large aperture essential optics used in CPA such as a dielectric grating (92 cm x 42 cm, 1740 grooves/mm) has been already developed, see Fig. 4.
A promising repeatable laser system at multi-kilojoule, named “J-EPoCH”,[D] has been basically designed for various high-energy-density plasma applications, and roughly consists of two laser systems of a nano-seconds power laser and a petawatt laser. The power laser is a 16 kJ diode-pumped solid-state laser system with 160 beam lines, each of which has a 100 J, 100 Hz active-mirror laser module. The petawatt laser is a Titan-doped sapphire (Ti:sapphire) laser system. Using about half beamlines of the power laser system as a pump source after second harmonic generation, kilo-joule energy is obtained in the femtosecond to picosecond time domain. Extremely higher thermal conductivity of cryogenic Ti:sapphire, compared with cryogenic Yb:YAG, is plenty of thermal strength for 100 Hz operation. The appropriate combination of beam lines from the both systems will help the first power generation demonstration in the fast-ignition-based IFE reactor.[E][A] K. Matsuo et al., Phys. Rev. Lett. 124, pp. 035001-1 - 035001-5(January 2020).
[B] M. Divoky et al., Opt. Lett. 40, pp. 855- 858 (March 2015).
[C] C. Baumgarten et al., Opt. Lett. 41, pp. 3339-3342 (July 2016).
[D] R. Kodama, AAPPS-DPP2018, P9 in Kanazawa, Japan (12-17, Nov. 2018).
[E] A. Iwamoto and R. Kodama, IFSA2019, 2P01 in Osaka, Japan (23 September 2019).Speaker: Prof. Junji Kawanaka (Institute of Laser Engineering, Osaka University) -
Successful operation of ITER depends critically on disruption management for the Pre-Fusion Power Operation (PFPO) phase up through Fusion Power Operations (DT). The power-handling capabilities of the beryllium (Be) first-wall panels (FWP) and other plasma-facing components (PFC) must be preserved in the face of disruptions and vertical displacement events (VDE). This need should account for intentional, low-power events required for electromagnetic load validation in the early operation phases, along with unintentional events that will usually be mitigated by the ITER Disruption Mitigation System (DMS). Multiple factors of the disruptions and VDEs influence the time-dependent heat flux and energy deposition onto the PFCs, which then determine the increase in surface temperature, melt formation, and material loss from melt motion and vaporization. Even for relatively low plasma current scenarios during the early operational phases $(I_P = 5 MA)$, initial studies predicted Be melt damage from upward VDE current quenches (CQ) up to ~0.5 mm deep and lateral melt displacements up to ~10 cm for single events$^1$. The implication is significant given the 10mm Be armor thickness and castellated geometry of the ITER FWPs. Extensive damage to Be surfaces due to such events has already been clearly documented on JET$^2$.
This paper will detail the extensive studies being performed at ITER to estimate material damage to the Be FWPs during disruptions and VDEs, ensuring that the first-wall power-handling capabilities are maintained up through Fusion Power Operations. The simulation efforts described in (1) are significantly expanded to cover a range of VDE and disruption scenarios for PFPO phases and FPO. This scenario dataset allows for a broad series of parametric studies to be performed using a novel simulation workflow developed at ITER. The methodology is as follows: 2D magnetic flux profiles from the DINA code$^3$ provide input to the SMITER 3D field-line tracing software$^4$, producing 3D maps of perpendicular surface heat flux $q_\perp$ and magnetic field $\overrightarrow{B}$ on the FWPs. These maps are then used to compute the time-dependent melt formation and dynamics with the MEMOS-U code$^5$, accounting for heat flux reduction by plasma vapor shielding. The final step is to regenerate the FWP models in SMITER using the melt deformation data from MEMOS-U. This capability allows for an assessment of heat loads on the damaged panel from either steady-state scenarios or a subsequent VDE/disruption.
The DINA simulations provide parallel heat fluxes $q_\parallel$ and current densities in the halo region from a full power balance, including radial transport, ohmic heating, radiation and losses along field lines. Variations in the DINA scenarios include: $I_p$ (5 – 15 MA), disruption direction (up or down), Be impurity density (~0 – $3x10^{19} atoms/m^3$), and disruption type (VDE vs. major disruption). SMITER simulations use CAD-accurate, high-resolution FWP meshes to generate surface heat loads and wetted area patterns at different time intervals of the plasma scenario. Uncertainties in parameters calculated in DINA are studied in SMITER: power traversing the last closed flux surface, $P_{sol}$ (+/- 20%); scrape-off layer heat flux widths $\lambda_q$ (+/- 25%); and the limiting location of the disrupting plasma (+/- 10mm radial displacement). MEMOS-U simulations utilize the thermomechanical material properties of solid and molten Be, compiled from literature and the ITER Material Properties Handbook. A new library of vapor shielding efficiencies has been incorporated into the MEMOS-U simulations such that the importance of shielding can be studied by switching shielding effects on/off. The shielding is characterized using the 1-D PIXY code$^6$ taking into account plasma parameters from DINA, allowing for a time-dependent heat flux modified by the shielding effect as a function of $q_\perp$ and Be vaporization rate.
Preliminary results emphasize the importance of a multi-physics workflow in estimating a realistic lifetime for the ITER first wall. The introduction of Be vapor shielding, for example, significantly reduces the incoming $q_\perp$ impacting the FWPs, even for the less-severe, relatively long 5 MA VDE cases. The shield is shown to equilibrate rapidly, in ~1-2 ms, relative to the typical ~200 ms duration for VDE energy deposition. At peak surface temperatures of 1800 – 1900 K, and at VDE-relevant heat fluxes of 100’s of $MW/m^2$, the vapor shielding efficiency can approach 70 – 80 %. Such strong shielding substantially slows the surface temperature rise, leading to less-severe melt thickness, melt motion, and surface deformation. For the 5 MA upward VDE, the maximum Be thickness loss for upper FWP #9 is reduced by about 50%, from ~0.3 mm to ~0.14 mm. As the molten Be is displaced and re-solidifies, the height of melt ridges is reduced by about 60%, from ~0.9 mm to ~0.34 mm.
The geometries of these deformed zones, with excavated pits and melt ridges, influence the local power loading of the damaged panels when subjected to further heat loads, as will be the case for actual operation. Since the damage does not overlap with the peak energy deposition pattern at the secondary X-point, damage profiles from the majority of upward VDEs have negligible consequences for steady-state operations. However, as depicted in the heat flux maps of Figure 3, a successive 5 MA upward VDE onto damaged FWP surfaces is a concern. Local heat fluxes are modified due to the melt pits and ridges, leading to increases of $q_\perp$ by 15 - 35%.
Some sources of variation are less of a concern. A +/-10mm radial variation in VDE location, for instance, leads to a negligible difference (~10 um) in peak $q_\perp$ and melt excavation magnitudes. A comprehensive disruption budget must ultimately take into account the type of disruption, location of impact, and the order and combination in which such events occur. Some combinations may lead to ‘smoothing out’ of prior damage, while other combinations may enhance the damage profile. It is the evolution of these successive, 3D damage profiles that will degrade the FWP power handling capabilities and set a limit on lifetime. Work is ongoing to complete the parametric studies listed above, analyzing heat loads and melt damage for the worst-case grouping of disruption uncertainties. Those panel damage profiles will then be combined with sequential steady-state and disruption loads, as in Figure 3. In this way, the most limiting combination of events will be thoroughly explored and presented.
(1) J. Coburn et al., “First Wall Energy Deposition during Vertical Displacement Events on ITER” to be published in Physica Scripta (2020)
(2) I. Jepu et al., Nuclear Fusion 59 (2019) 086009
(3) V. E. Lukash and R. R. Khayrutdinov, 1996 Plasma Physics Reports, 22 (1996) 91
(4) L. Kos et al., Fusion Engin. & Des, 146 B (2019) 1796 – 1800
(5) E. Thorén et al, Nuclear Fusion 58 (2018) 106003
(6) K. Ibano et al, Nuclear Fusion 59 (2019) 076001Speaker: Dr Jonathan Coburn (ITER Organization) -
Positive plasma potential was observed for the first time in a core tokamak plasmas, conventionally characterized by negative potential. Direct measurement of the electric potential in the core plasma is of paramount importance for the understanding of the role of radial electric field E_r in the mechanisms regulating the toroidal plasma confinement. New experimental observations and theoretical description of the E_r formation based on neoclassical (NC) models in the core and turbulent dynamics in the edge of the T-10 tokamak are the goals of this paper. A Heavy Ion Beam Probe (HIBP) diagnostic was developed for T-10 (circular tokamak, B_0=1.7-2.42 T, R=1.5 m, a=0.3 m) to study the plasma potential with high spatial (<1 cm) and temporal (1 $\mu$s) resolution. Tl+ ions with energies E_beam up to 330 keV were used to probe the plasma from the edge to the core {1}. Low-density OH deuterium plasmas (n_е=1.0x10^19 m^–3,T_e<1.3 keV, T_i<0.6 keV) in T-10 are characterized by a negative potential up to $\varphi$= 1500 V near the centre ($\rho$=r/a=0.3), Fig. 1. The potential profile monotonically increases towards the periphery, forming the radially averaged electric field E_r = -75 V/cm. At the edge ($\rho$=1), HIBP data agrees with Langmuir Probe measurements (star in Fig. 1). The density rise due to gas puff is accompanied by an increase of negative potential. This is valid both for the steady-state phase of the discharge, and for the initial phase of the plasma current and density ramp-up. Powerful off-axis ($\rho$_EC=0.5) second harmonic X-mode ECRH with P_EC_off<1.7 MW (f_EC_off=144 GHz, gyrotrons A and C) leads to an increase of T_e up to 2 keV in the centre, as measured by ECE and shown in Fig. 2. It also causes a dramatic raise of the core plasma potential towards the positive values over the whole observation area, with the local potential increase $\Delta$$\varphi$=1000 V near the centre and E_r= - 20 V/cm. Potential profile has a sort of a plateau near the edge ($\rho$=0.8-1). An extra nearly on-axis ($\rho$=0.2) ECRH with P_EC_on<0.5 MW (f_ECon=129 GHz, gyrotron B) leads to a further increase of T_e up to 3.3 keV at the centre and a potential raise in the core plasma forming an extended area of positive E_r = 20 V/cm, from the core to the edge as presented in Fig. 1. Remarkably, the plasma potential in the TJ-II stellarator with similar size and plasma parameters shows positive electric potential for low-density plasmas with powerful second harmonic X-mode ECRH {2}.
Geodesic Acoustic Modes (GAMs) and broadband (f<400 kHz) turbulence of the plasma potential and density have been directly studied by HIBP from the plasma core to the edge and by Langmuir probe at the edge. GAMs altogether with higher frequency satellite are dominating in potential power spectra in OH plasma, GAM amplitude increases during ECRH. Both GAM and satellite have uniform structure with constant frequencies over a wide radial extension, exhibiting the features of global eigenmodes of plasma oscillations, as shown in Figs. 3 and 4, where f_GAM=18-22 kHz, and f_sat=23-25 kHz.
The main GAM peak has wider outer bound at the plasma edge than the satellite; f_GAM follows the theoretical scaling f_GAM~sqrt(T/m_i)/R (for T_e at $\rho$=0.7) for both OH and ECRH regimes in a wide temperature range, covering the whole operational limits of T-10. In addition to GAM, quasicoherent (QC) electrostatic mode with frequency 50-120 kHz takes place. In contrast to GAM, dominating in potential power spectra, QC-mode dominates in density power spectra. A Stochastic Low Frequency (SLF) mode with frequency <50 kHz is also seen in the plasma density and potential power spectra, density poloidal coherence and cross-phase. Bicoherence analysis has shown three-wave coupling between GAM and broadband turbulence, predominantly with QC. NC modeling was performed with various codes from the simple analytical approach {1} to the NC orbit code VENUS+δf {4}. Core radial profiles and the main tendencies like potential decrease with density raise and potential raise with T_e increase due to ECRH were explained by NC models within the experimental and modeling accuracy. Direct numerical calculation of the turbulent dynamics in the edge plasma by the 4-field {$\rho$, n, p_e, p_i} nonlinear two-fluid MHD Braginskii model based on {4} explains the E_r dynamics due to the influence of the Reynolds stress, in turn caused by potential phase relations, and the Winsor force. In summary, the results indicate the important features of $\varphi$ and E_r profiles and electrostatic turbulence: the negative potential growth with density (confinement time), change of potential sign to positive, and positive potential growth with ECRH (decrease of confinement time) may shed a light to the properties of plasma energy confinement.
The work was funded by Russian Science Foundation, project 19-12-00312.
References
1. A.V. Melnikov et al., Fusion Eng. Design, 146, 850-853 (2019)
2. A.V. Melnikov et al., Fusion Sci. Technol. 51, 31 (2007)
3. M.Yu. Isaev et al., Nucl. Fusion 49, 075013 (2009)
4. R.V. Shurygin and A.V. Melnikov, Plasma Phys. Reports, 44 263 (2018)
Speaker: Prof. Alexander Melnikov (NRC 'Kurchatov Institute') -
High performance advanced tokamak scenarios are very attractive for future burning plasmas. They can be achieved by elevating the central $q$-profile to values around unity to stabilize the sawtooth instability, which would otherwise reduce performance and could trigger deleterious instabilities. High-$\beta$ plasmas can develop such a flat elevated central $q$-profile in the presence of MHD modes that modify the current profile [1]. The self-regulating mechanisms leading to this anomalous evolution of magnetic flux can be referred to by the general term "magnetic flux pumping". At DIII-D, flux pumping was observed in the presence of a 3/2 tearing mode, as well as when inducing a helical core via external perturbation coils [2]. In the work presented here, experimental evidence of anomalous current redistribution due to the dynamo effect produced by a 1/1 quasi-interchange instability [3] is discussed. It is shown that the ability of the mode to redistribute the centrally driven current, and thereby to suppress sawteeth, scales with the plasma pressure. This is potentially important for future non-inductive tokamaks, as it could provide a way to redistribute the current driven by electron cyclotron current drive (ECCD), which drives current most efficiently in the plasma centre. The flux pumping mechanism would redistribute current outward, maintaining a flat central q-profile around unity and maximizing both current drive efficiency and plasma stability at high $\beta_N$.
![a) On-axis electric field from the 1/1 mode versus the electric field necessary to keep q$_0$ at 1 for different 3D nonlinear MHD simulations [4]. Blue and green are sawtooth-free cases, red and orange are sawtoothing, the labels identify the simulation parameters in the reference. b) Experimental parameter scan, $\beta_N$ versus on-axis ECCD.](https://home.mpcdf.mpg.de/~aburckha/IAEA2020/ECCDvsBetaN.png)
A theoretical model based on recent simulations suggests that flux pumping can occur in the presence of a saturated ($m=1$, $n=1$) interchange-like mode [4]. The flow in the convection cell combines with the perturbation of the magnetic field via the dynamo effect to generate an effective negative loop voltage in the plasma core. This prevents the central current density from peaking and thereby flattens the core $q$-profile. The mechanism is self-regulating such that the core q-profile is clamped to values close to unity. Figure 1a shows the central electric field resulting from a 1/1 mode predicted by simulations, plotted against the central loop voltage necessary to keep q$_0$ around unity. The latter depends on internal and external parameters that lead to central current peaking, like externally induced current drive. In the cases that lie above the line, the magnetic flux pumping mechanism is sufficiently strong to prevent sawtoothing, whereas in the cases below the line, $q_0$ is below unity and sawtoothing occurs. The simulation results suggest that the strength of the flux pumping mechanism depends on the core pressure. This dependency on $\beta_N$ stems from the pressure-driven nature of the 1/1 quasi-interchange mode. The simulations shown here use a generic tokamak geometry, but simulations based on ASDEX Upgrade (AUG) discharges are underway. The experimental results shown in figure 1b support the theoretical model and will be discussed below.
With the combination of the imaging motional Stark effect diagnostic (IMSE) [5] at AUG and the IDE equilibrium solver [6,7], changes of $q$ as small as 0.1 are measurable, even in the plasma center. Together with the current drive capabilities of the upgraded electron cyclotron resonance heating (ECRH) systems [8], AUG constitutes the ideal device to perform experiments that test these simulations. In discharges featuring a 1/1 mode, positive ECCD current was applied in several steps to decrease $q_0$ and trigger sawteeth. At the same time, an NBI power scan was performed to increase the $\beta_N$ value over the threshold necessary for the mode to suppress sawteeth at a given central current drive.
Figure 2 shows the heating power, $\beta_N$, ECCD, 1/1 mode activity and $q_0$ in an AUG discharge with 800kA plasma current, a $q_{95}$ of 5.2 and an $H_{98}$ factor of 1.1. Five phases are discussed in the following. The red shading indicates the presence of sawteeth, the blue shading their absence. The first phase starts with large sawteeth. After $t=1.55$s, the measured central $q$ profile remains flat and clamped around unity. The last large sawtooth is observed around 1.62s, but small sawteeth still remain. At 1.65s, 1/1 mode activity first appears. Once $\beta_N$ is increased above 2.4 the small sawteeth disappear as well, even when the central ECCD is increased. When the ECCD is increased further, above 130kA, sporadic small sawteeth reappear and become more frequent with more ECCD. In the last phase, the driven current is reduced and the sawteeth disappear again.
The bottom panel shows the modelled $q_0$ in red, resulting from an equilibrium reconstruction which takes external magnetic measurements, kinetic profiles, current diffusion and a sawtooth current redistribution model into account [7]. The blue curve shows the estimated $q_0$ when additionally taking into account the local measurements from the IMSE diagnostic, which will be referred to as "measured $q_0$". It can be seen that in phase III, without sawteeth, $q_0$ should drop well below unity if no other current redistribution mechanism were present besides neo-classical current diffusion. The modelled and measured $q$ profiles for this phase are shown in the right panel. The measurements show that the central safety factor stays stable around one, suggesting an anomalous modification of the current profile. At the beginning of phase IV, the modelled $q_0$ is sporadically increased to unity by sawteeth, but drops well below 1 between the sawteeth. Since such a low $q_0$ would immediately trigger a sawooth, this suggests that the flux pumping mechanism still plays a role, but is not strong enough to completely suppress the sawteeth. This can also be seen in the measured $q_0$, which remains closer to 1.
The measured $\beta_N$ and ECCD current from the different phases in this experiment are plotted in figure 1b (diamonds). The circles show the results from a similar discharge with more heating power, resulting in a higher $\beta_N$. For a comparison with the theoretical predictions (figure 1a), here $\beta_N$ is used as a proxy for the electric field that can be created by the 1/1 mode and the central ECCD current as a proxy for the electric field necessary to keep the central $q$ around unity. At a given ECCD current, $\beta_N$ needs to exceed a certain threshold to enter the sawtooth free regime. At a higher ECCD current, this threshold increases. This supports the simulation results from reference [4] where the flux pumping mechanism in the simulations is only able to prevent sawtoothing at sufficiently high $\beta_N$, and where the threshold is dependent on central current drive peaking.
In the proposed contribution, results from simulations based on the experimental data from AUG discharges will be presented. The qualitative and quantitative agreement with the electric field deficit in the experiment, calculated from the difference between the modelled and measured toroidal current, will be discussed.
[1] C. Petty et al, PRL, 102, 045005 (2009)
[2] P. Piovesan et al, NF, 57, 076014 (2017)
[3] S. C. Jardin et al, PRL, 115, 215001 (2015)
[4] I. Krebs et al, PoP, 24, 102511 (2017)
[5] O. Ford et al, RSI, 86, 093504 (2016)
[6] R. Fischer et al, FST, 69, 526-536 (2016)
[7] R. Fischer et al, NF, 59, 056010 (2019)
[8] J. Stober et al, PPCF, 62, 024012 (2020)Speaker: Dr Andreas Burckhart (Max Planck Institute for Plasma Physics) -
Substantial seeding of impurities into the divertor has been used for a long time in tokamaks to reduce the power and particle fluxes impacting on the divertor targets and is one of the main techniques to be utilised on ITER to facilitate stationary divertor operation [$1$]. There have been attempts to predict how the impurity concentration required for detachment should scale with different plasma parameters, such as the power crossing the separatrix, P$_{sep}$, and the upstream separatrix density, n$_{e,sep}$ [$2-4$]. In [$3$] the machine size scaling enters through the minor radius, a$_{min}$, due to the dependence on the poloidal magnetic field. Including the LH threshold scaling provides a dependence on B$_T^{0.88}$R$^{1.33}$[$4$]. Since there are no experimental results to test these scaling laws, this work uses spectroscopy to measure the divertor nitrogen concentration, c$_N$, in the outer divertor to examine the parameter dependencies both within and across the ASDEX Upgrade (AUG) and JET tokamaks.
From a database of N-seeded H-mode AUG discharges, spanning P$_{sep}$=3.5–12 MW and n$_{e,sep}$=1.8–4x10$^{19}$ m$^{-3}$, with line averaged core densities from 7–10x10$^{19}$ m$^{-3}$, and plasma currents from I$_p$=0.8–1.2 MA, the c$_N$ measurements at the onset of detachment will be presented. A similar database of JET pulses have been collected, with P$_{sep}$=12–16 MW, n$_{e,sep}$=1.6–2.7x10$^{19}$ m$^{-3}$, and I$_p$=2–2.5 MA. In these JET pulses, the outer target magnetic geometry is in a vertical configuration for a fairer comparison with both AUG and ITER. On both machines, the elongation is $\kappa$~1.7 while a$_{min}$~0.5 and a$_{min}$~0.9 on AUG and JET, respectively. The power to the outer divertor is calculated as ~P$_{sep}$/2.5 to account for power to the inner divertor and to the main chamber wall. Although the JET results still require validation, a first regression of the parameter dependencies on both devices gives c$_N$=15.9P$_{sep}^{1.32}$[MW] I$_p^{1.17}$[MA] n$_{e,sep}^{-3.44}$ [10$^{19}$ m$^{-3}$] (1+$\kappa^2$)$^{-1}$a$_{min}^{-2.85}$ [%] (see figure 1); a result in good agreement with [$3$] but with a stronger dependence on n$_{e,sep}$ (c$_{N,2P model}$~P$_{sep}$I$_p$n$_{e,sep}^{-2}$(1+$\kappa^2$)$^{-1}$a$_{min}^{-3}$). The stronger n$_{e,sep}$ dependence is also found from SOLPS-ITER simulations which show that c$_{Ne}$ ∝ n$_{e,sep}^{\alpha}$ where $\alpha$ varies between -2 and -4 depending on the power radiated below the X-point [$1$].
![Measurements of c$_N$ in the outer divertor in partially detached conditions on AUG (red) and JET (black) are plotted against the derived scaling law in (a). In (b) the normalised c$_N$ measurement is plotted against n$_{sep}$ with three best fit lines fitted to the function An$_{sep}^{\alpha}$ for $\alpha$=[-2,-3,-4].](https://ccfe.ukaea.uk/wp-content/uploads/2020/03/SHenderson_IAEA-FEC_Figure1.png)
Measurements of the N concentration in the outer divertor are taken when the divertor plasma reaches partial detachment. For AUG, the divertor target shunt measurements are used to provide a real-time estimate of the divertor temperature. Typically, the divertor is in a partially detached state when this real-time temperature is T$_{div}$~5 eV. Such a real-time estimate is not available on JET, and therefore spectroscopy is used to determine the detachment state. The ratio of the line-integrated N II intensity between sightlines viewing close to the divertor target and close to the X-point, which should vary according to the electron density front movement, typically follow the trends of J$_{sat}$ from the probe measurements and therefore provides a robust estimate of the detachment state. Using the ratio of sightline intensities is also possible on AUG, however T$_{div}$ is typically more reliable.
The N concentration is calculated using c$_N$=4$\pi$I$_{NII}$/(f$_N$+PEC$_{exc}$+f$_{N2}$+PEC$_{rec}$)/$\Delta$L/n$_{e,NII}^2$ [$5$] where $\Delta$L is the length of the N II emission region, I$_{NII}$ is the measured intensity, and PEC$_{exc/rec}$ and f$_{NZ}$ are photon emissivity coefficients and fractional abundances. The electron temperature and density, T$_{e,NII}$ and n$_{e,NII}$, are derived from spectroscopic N II line ratios. $\Delta$L is estimated based on inverted camera images on JET and SOLPS modelling on AUG. In partially detached conditions, this length is approximately 7cm on both devices and is localised in the private flux region. This is also consistent with the N II line ratios, which show electron temperatures mostly between 3-4 eV, close to the zero-transport prediction of maximum fractional ion abundance. This work will also examine the robustness of the measurement and provide comparisons to a simple approximation of c$_{N}$ derived from the ratio of the impurity and fuel gas valve fluxes. In stationary scenarios with high measured intrinsic N intensities, indicative of fully saturated vessel surfaces, the two measurements agree on AUG and JET as shown in figure 2; however, when the intrinsic N intensity is low, the two measurements can differ by an order of magnitude. While the gas valve ratios can provide a good proxy for the impurity concentration in the divertor neutral domain with fully saturated walls, spectroscopy provides a direct, instantaneous, line-integrated measurement in the outer divertor plasma.
Finally, this work will compare a database of JET discharges with horizontal outer target magnetic geometries to assess any differences due to divertor geometry. It is also clear that the scaling of the impurity concentration is dependent on the seeding gas species. This work has focused on N, due to the abundance of scenarios with N$_2$ seeding in current devices; still, it is important to assess results for Ne and Ar. The Ne II emission also radiates in the visible spectral range and scenarios on JET with Ne seeding will therefore be used to present a first assessment of the species scaling. It is not yet possible to measure the Ar concentration spectroscopically; however, a database of AUG discharges with different input powers and fractions of Ar and N$_2$ seeding will be presented to evaluate whether an optimum mixture of impurities exists with regards to the radiation distribution and plasma confinement.
[$1$] R. Pitts et al. 2019 NME 20 100696; [$2$] A. Kallenbach et al. 2016 PPCF 58 045013; [$3$] R. Goldston et al. 2017 PPCF 59 055015; [$4$] M. Reinke 2017 Nucl. Fusion 57 034004; [$5$] S. Henderson et al. 2018 Nucl. Fusion 58 016047 and †See the author lists of B. Labit et al 2019 Nucl. Fusion 59 086020, †† H. Meyer et al 2019 Nucl. Fusion 59 112014and †††E. Joffrin et al. 2019 Nucl. Fusion 59 112021
Speaker: Stuart Henderson (UKAEA) -
Here we report recent progresses of laser fusion energy research in Japan, especially on the fast ignition scheme. For the fast-track to the laser fusion energy, we are investigating the fast-ignition plasma physics to realize optimal compression of a fusion fuel as well as efficient heating of the compressed fuel. In this scheme, we have demonstrated the efficient heating of high density plasmas with compression of the solid sphere by using a multi kJ lasers. The results imply the FI could have burning of a laser fusion plasma with a few 100kJ including 50-100kJ heating laser. Furthermore, we are developing a high repetition high power laser system: 10kJ/10-100Hz laser system to realize data driven analysis of fusion plasmas for the optimal target design as well as test of laser fusion reactor engineering with a fusion subcritical reactor including fusion power generation. These extensive studies are being made with under the fast ignition consortium and laser fusion strategy committee of IFE forum in Japan.
laser fusion plasma physics for the fast ignition
In the fast ignition (FI) scheme, an ultra-intense laser light generates high energy particles such as relativistic electron beam (REB) to heats a high density compressed fuel. The first demonstration of fast heating of laser compressed fuel1 has been made with a cone geometry for the efficient propagation of the heating laser pulse to generate the REB close to the high density core plasmas.
Details of the heating mechanisms has been investigated through experiments and simulations. In general, REB can heat the core plasma with 3 different mechanisms such as drag heating, return current heating and diffusive heating [2]. The drag and diffusive heating processes are important especially for the efficient heating by REB in the FI scheme. To improve the coupling of REB to the core plasma, the magnetized fast-isochoric heating (MFI) scheme had been designed and demonstrated. Delivering REB with a guidance of external magnetic fields at kT-level had been realized with the GEKKO XII (GXII) + LFEX, a ps/ kJ laser [3]. The REB, which has inherently a large divergence, was successfully guided along the magnetic field lines to the core, resulting twice higher laser-to-core coupling for the drag heating than that in a case without the guidance [4]. Another important heating mechanism in the FI is the diffusive heat driven by a strong radiation pressure of the heating pulse toward the core plasmas [5]. This scheme has been based on the physic of a relativistic laser -plasma interaction and becomes significant with a multi-picosecond kJ laser. The diffusive heating is the main heating process in the current experiment with relatively low density core plasmas, e.g. 0.1g/cm2 [6], and the drag heating will be comparably important in the ignition-scale core (0.3 -0.5 g/cm2).
Improving the coupling of the heating laser to the appropriate energy of REB is also crucial for the efficient heating. One is focusing of multiple coherent laser beamlets spatially and temporally overlapped, thus producing an interference pattern in the laser focus, has significantly improved the laser energy conversion efficiency into the REB, compared to one laser beam with the same energy as the four beamlets combined [7]. Another idea to the higher coupling efficiency is utilization of shorter wavelength laser light to generate relatively lower energy REB, which then has a shorter stopping power. A 527nm laser light will make expectation for future realization of fusion gain of more than 10 with a heating laser energy of 70kJ for the DT fuel with ρR=2 g/cm2 and 500 g/cm3. Furthermore, higher efficiency could be expected with fundamental (1053 nm) and second harmonics (527nm) mixed wavelength laser lights.
Optimal fuel compression for the FI is critical to realize the efficient ignition of the laser fusion. The FI scheme can optimize separately the processes of the fuel compression and the heating of the compressed fuel. We find that a solid spherical fuel target will be more feasible, to stably obtain high-density fuel compression without a risk of the hydrodynamic instability. Moreover, the solid sphere target has an advantage in more simple fabrication process, as compared with the shell target, especially for the cryogenic fuel fabrication. High density compression of the solid sphere attached a Au cone target, which could be hydro dynamically stable, had been experimentally demonstrated to be ρR~0.1 g/cm2[8]. Based on 2D simulation results, about 130kJ would be required to obtained ρRmax =1.0 g cm−2 using a solid sphere and multi-step laser pulse [9]. More efficient coupling > 20% would be expected through the optimal heating processes to initiate ignition and burning of the compressed fuel with 50-100kJ heating laser energies [10].
High repetition laser system development and fusion energy reactor engineering
We are developing a 100J/100Hz high power lasers with an active mirror amplification scheme using 10 cm Yb:YAG ceramics pumped by laser diodes. Based on this system, we are now proposing a new type of high power laser system as shown Fig. 1 to explore a variety of new fields of sciences, which is called “J-EPoCH”. This facility integrates all the state-of-arts high power laser technologies, based on the 160 beams of 100Hz /100J laser module, providing high repetition 10kJ long pulse lasers, 5-20PW short pulse lasers and different kinds of laser plasma accelerators, and laser-driven radiation sources such as x-rays and neutrons.
This multi-purpose high repetition laser system will also open a new frontier of laser fusion energy development. The laser fusion strategy committee of IFE forum in Japan organized around 40 members from 20 institutes is considering the roadmap utilizing this high repetition laser system toward laser fusion reactor. The high repetition system will realize data-driven analysis of fusion plasma for the optimal target design as well as testbed of laser fusion reactor engineering with a fusion subcritical reactor [11]. This fusion subcritical reactor enables early implementation of all laser fusion engineering assessments such as wall materials, optical elements, pellet injection and power generation, e.g. fusion electric power of a few Watt using a large aspect shell target under the 1 Hz laser operation.1 R. Kodama et al., Nature 412, 798 (2001); Nature 418, (2002) 933
[2] Y. Sentoku at al., Physics of Plasmas 14, (2007) 122701
[3] S. Fujioka et al., Scientific Reports 3, (2013) 1170
[4] S. Sakata et al., Nature Commun. 9, (2018) 3937
[5] N. Iwata et al., Natture Commun. 9, (2018) 623
[6] K. Matsuo et al., Phys. Rev. Lett. 124, (2020) 035001
[7] A. Morace et al, Nature Commun. 10, (2019) 2995
[8] H. Sawada et al., Appl. Phys. Lett. 108, (2016) 254101
[9] H. Nagatomo et al., Nuclear Fusion 59, (2019) 106055
[10] T. Johzaki et al., Physics of Plasmas 15, (2008) 062702
[11] A. Iwamoto et al., 11th Int. Conf. on Inertial Fusion Sci. and Appl. 2P01-04 (Sep.23 2019)Speaker: Prof. Ryosuke Kodama (Institute of Laser Engineering, Osaka University) -
Inertial Confinement Fusion (ICF) schemes are designed to heat and compress DT fuel to conditions exceeding the Lawson criterion ($p \tau$) using implosion, which greatly amplifies the pressure of a driver (~100 MBar) to the conditions necessary for laboratory-scale ICF (~100s GBar). The National Ignition Facility (NIF) focuses on the laser indirect drive approach to ICF, in which laser energy is converted to x-ray radiation in a `hohlraum’, which drives the fuel-containing capsule$^1$. This process is inefficient, with ~10% coupling efficiency from the laser energy to energy absorbed by the capsule typical. Of the energy absorbed the capsule material, only ~10% is converted into kinetic energy of the imploding fuel and internal energy of the fuel at stagnation. One focus of the program is to improve this coupling efficiency and enable larger implosions within the current capabilities of the NIF laser; this is a route towards increasing $p \tau$ on NIF and is relevant for future ignition experiments or approaches towards inertial fusion energy since the coupling efficiency feeds into requirements for driver size and target gain. In parallel, several degradation mechanisms have been identified that impact implosions on NIF, notably low-mode drive asymmetry and mix induced by target defects or engineering features. Detailed studies of these mechanisms have been conducted to identify routes towards improved performance.
In 2017-2018, record fusion yields on NIF were produced with designs that utilized high-density-carbon (HDC) capsules and low-gas-fill hohlraums$^2$. These previous campaigns explored studies over several parameters, with the combination of data and theoretical scaling arguments suggesting that increasing the capsule size could be a favorable tactic for further improving performance$^3$, in part from the increased coupling efficiency from the driver to the fuel. With a constant available laser energy and power, a recent campaign pursued an increased capsule size in comparable hohlraums, substantially reducing the case-to-capsule ratio (CCR), from ~3 to ~2.7. A schematic of the target, compared to a previous campaign$^4$, is shown at the left of Fig. 1. Empirical metrics for the implosion symmetry$^5$ and initial data demonstrate that control over the symmetry is needed to prevent highly distorted implosions at this CCR. We use wavelength detuning ($\Delta \lambda$) between inner beams, which drive the hohlaum waist, and outer beams, which drive near the poles, to semi-empirically adjust the amount of cross-beam energy transfer (CBET) between these beams, which provides control over the implosion symmetry$^6$. This wavelength detuning can be adjusted on every shot to control the shape. Example data are shown on the right of Fig. 1, demonstrating that the application of 1Å of $\Delta \lambda$ changes the implosion shape from oblate to prolate. Unlike previous campaigns that employed wavelength detuning in high-gas-fill hohlraums$^7$, our data demonstrate that the shape is symmetric throughout all stages of the implosion.With control over the shape we have conducted an initial series of three cryogenically layered DT fueled shots to assess the integrated implosion performance. The first two implosions revealed a higher than expected level of high-Z ablator material mixing into the fuel. This was successfully mitigated with two techniques: first, by increasing the fuel thickness to provide an additional buffer against mix, and second by using capsules with improved quality, or fewer seeds for deleterious hydrodynamic instabilities. These changes result in an implosion with record values for NIF for the capsule absorbed energy, fuel kinetic energy, and hot-spot internal energy, shown in Fig. 2. The highest performing shots in this campaign (N191007 and N191110) are denoted and have record values for coupled energy. Notably, these implosions have achieved record fuel energy at very modest values of other design parameters, especially the velocity. Simple scaling relations$^3$ expect the fusion performance to increase strongly as a function of both scale factor ($S$) and velocity ($v$) as approximately $Y \propto v^{7.7} S^{4.4}$. Increasing the velocity of these implosions is therefore a clear direction for future exploration.
In parallel with the effort to increase energy coupling to the capsule and fuel, understanding known degradation mechanisms has been a dedicated effort of the program. Several sources of unintentional directional mode-1 drive asymmetry have been identified, including diagnostic windows in the target, random variation in the laser delivery, and anisotropic capsule thicknesses. Reducing these sources of asymmetry is expected to negate the impact of deleterious mode-1 drive asymmetry on current implosions. Similarly, several sources of high-Z material mixed into the fuel have been identified. First, engineering features such as the membrane that holds the capsule within the hohlraum and the tube used to introduce fuel into the capsule are sources of material mix into the hot spot. Second, defects introduced during the capsule manufacturing process, including ‘pits’ on the surface and ‘voids’ within the material, can cause high levels of mix into the fuel. Mitigation mechanisms for these varied sources of deleterious mix are being pursued. The degradations, when mitigated, are expected to both improve performance of current implosions and enable experiments in more aggressive parameter space.
In summary, we have conducted a campaign to improve the energy coupling efficiency for NIF implosions, by fielding a larger capsule to absorb more energy from the x-ray producing hohlraum, and have achieved record fuel kinetic energy and hot-spot internal energy with this approach. Increasing the coupling efficiency from the laser drive to the fuel is advantageous for improving performance on NIF and for projections to inertial approaches to fusion energy. In parallel, several sources of degradation mechanisms have been studied, with the causes identified and mitigation techniques in development to enable higher performing implosions.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
1: J. Lindl, Phys. Plasmas 2, 3933 (1995)
2: S. Le Pape et al., Phys. Rev. Lett. 120, 245003 (2018); K. Baker et al., Phys. Rev. Lett. 121, 135001 (2018)
3: O.A. Hurricane et al., Phys. Plasmas 26, 052704 (2019)
4: A.L. Kritcher et al., “Symmetric fielding of the largest diamond capsule implosions on the NIF”, submitted to Phys. Plasmas (2020)
5: D.A. Callahan et al., Phys. Plasmas 25, 056306 (2018)
6: A.L. Kritcher et al., Phys. Rev. E 98, 053206 (2018)
7: O.A. Hurricane, Nature Physics 12, 800 (2016)Speaker: Alex Zylstra (Lawrence Livermore National Laboratory) -
During burning plasma operation on ITER, extrinsic impurity seeding will be mandatory for heat flux control at the tungsten (W) divertor vertical targets [1]. A very extensive database of SOLPS plasma boundary code simulations has been compiled for ITER [1], including the most recent advances, obtained with the SOLPS-ITER version, in which for the first time, fluid drifts have been included [2]. These simulations predict that partially detached divertor solutions at high divertor neutral pressure will be possible on ITER for baseline burning plasmas ($Q_{DT}=10$, power into the scrape-off layer $P_{SOL} = 100 MW$), with both neon (Ne) and nitrogen (N) low Z seeded impurity, and with impurity compression sufficient in both cases to maintain the majority of the radiated power in the target vicinity. Drifts are found to be relatively unimportant under such conditions. This is in contrast to observations on smaller devices with W divertors, such as ASDEX-Upgrade (AUG), in which Ne compression is reduced in comparison with N, core plasma performance is compromised and drift effects are stronger. However, Ne is preferred on ITER in DT plasmas to avoid impact on machine duty cycle due to the formation of tritiated ammonia [1]. It is thus critical that the fundamental controlling physics responsible for this behavior be understood, in particular the impact of scale size. This contribution identifies the key factors at play through a unique SOLPS-ITER simulation study in which Ne-seeded H-mode conditions in vertical target W divertor geometries are compared in the three devices, ASDEX-Upgrade (R=1.65 m), JET (R = 3.0 m) and ITER (R = 6.2 m), spanning a factor of more than 3 in linear dimension in almost equal size intervals.
For AUG and JET, the modelling parameters are inspired by existing experimental results, but do not attempt to match a particular discharge. High power, H-modes with $P_{SOL} = 12 MW$ , $q_{95}=5.5$ (AUG) and $P_{SOL}=20 MW$, $q_{95}=3.3$ (JET), and cross-field heat transport chosen to match the typical SOL widths, ${\lambda}_q$ observed on these devices under such conditions, give comparable power flows at the divertor entrance to those for the modelled ITER burning plasma with ${\lambda}_q = 3-4 mm$, $P_{SOL} = 100 MW$ , $q_{95}=3$. All simulations include fluid drifts and currents, with neutrals traced by the EIRENE code. Semi-detached divertor conditions are established in all cases using moderate Ne seeding and high deuterium throughput.
An important finding is that the impact of the poloidal and radial ExB drifts (redistributing plasma between the outer and inner divertors through the private flux region [3]) steadily decreases with increasing machine size. The divertor asymmetry associated with these drifts also thus decreases with size. As shown in Fig. 1, the high field side, high density front observed experimentally in AUG [4] and JET [4] is reproduced by the modelling, but is absent in the ITER simulation. On AUG this front reaches the X-point vicinity and can influence the pedestal plasma [5]. The same divertor asymmetry driven by the ExB drift drives a significant redistribution of Ne impurity, which tends to accumulate in the more detached inner divertor (Fig. 2). The resulting increased impurity radiation further exacerbates the divertor asymmetry, making it harder to achieve partially detached conditions in the outer divertor without impurity concentrations exceeding acceptable limits for core performance. This effect also decreases with increasing machine size. The relative importance of $\nabla B \times B$ and Pfirsch-Schlueter (P-S) driven flows in the SOL region also differs between ITER and smaller devices. Whilst in AUG, P-S flow provides the main contribution to flow reversal in near SOL, in ITER excess ionization in the strike point region will be the principal driver.
A second key size dependent effect concerns the electron temperature ($T_e$) distribution in the divertor, which depends on the X-point to target connection length. Due to the $T_e$ dependence of the parallel heat conductivity, the region of temperature change near the target or, in the case of detachment, the ionization front, is narrow and does not depend on the X-point position or machine size. The position of this layer thus determines the X-point $T_e$. For larger devices, with higher confinement, this temperature is higher for the same energy flow at the divertor entrance. Since low Z impurity radiation will be strongest in the region of comparatively low $T_e$, where many partially ionized states exist, impurity radiation in larger machines is more localized in the divertor. On ITER, this means that even though the strongly radiating region with Ne impurity is more extended than for N seeding at comparable radiated power, both are equally effective at divertor power dissipation. In addition, any Ne ions reaching the ITER pedestal region are fully stripped due to the high $T_e$ there under high performance conditions and cannot radiate, reducing the impact on pedestal power balance.
[1] R. A. Pitts et al 2019 Nucl. Mater. Energy 20 100696
[2] E. Kaveeva et al Nucl. Fusion in press: DOI 10.1088/1741-4326/ab73c1
[3]V Rozhansky et al 2018 Contrib. Plasma Phys. 58 540–546
[4] S Potzel et al 2015 Journal of Nucl. Mater. 463 541-545
[5] M G Dunne et al 2017 Plasma Phys. Control. Fusion 59 014017Speaker: Prof. Vladimir Rozhansky (Peter the Great St.Petersburg Polytechnic University) -
TAE Technologies, Inc. (TAE) is a privately funded company pursuing an alternative approach to magnetically confined fusion, which relies on field-reversed configuration (FRC) plasmas composed of mostly energetic and well-confined particles by means of a state-of-the-art tunable energy neutral-beam (NB) injector system. TAE’s current experimental device, C-2W (also called “Norman”) shown in Fig. 1, is the world’s largest compact toroid (CT) device [1] and has recently made significant progress in FRC performance, producing record breaking, high temperature advanced beam-driven FRC plasmas, dominated by injected fast particles and sustained in steady state for up to 30 ms, which is limited by NB pulse duration as can be seen in Fig. 2. C-2W has been producing significantly better FRC performance than the preceding C-2U experiment [2], in part due to Google’s machine learning framework for experiment optimization, which has contributed to the discovery of a new operational regime where novel settings for the formation sections yield consistently reproducible, hot, stable plasmas.
In order to produce such high performance FRC plasmas, C-2W operates with the following key features [1]: reliable dynamic FRC formation scheme via colliding and merging two oppositely-directed CT plasmas (relative collision speed up to ~1000 km/s); tangential, co-current NB injection (NBI) into the FRC with high input power (total up to ~21 MW) and intra-discharge variable energy (15–40 keV) functionality; flexible edge-biasing electrode systems for stability control in both inner and outer divertors; neutral gas density control via ~2000 ${m^3/s}$ pumping capability in the divertors; external magnetic field fast control capabilities, such as field ramp, and active feedback control of the FRC plasma using Trim and Saddle coils; and 50+ dedicated plasma diagnostics in the main confinement region and in divertors to characterize FRC and open-field-line plasma performance.
In the recent C-2W experiments, adequately controlled external magnetic-field profile throughout the machine and proper gas injection/fueling have led to more effective edge biasing from electrodes to globally stabilize plasma; thus, improving the efficiency of the NB-to-FRC coupling so that more plasma heating and current drive are obtained. Due to this synergistic effect of combining effective edge biasing and NBI on C-2W, beam-driven FRCs have achieved a high temperature regime as shown in Fig. 2; total temperature ${T_{tot}}$ >2 keV (${T_{tot}}$ = ${T_e+T_i}$, based on a pressure balance), electron temperature ${T_e}$ >250 eV. Other typical plasma parameters are: averaged electron density <${n_e}$> ~${1–3×10^{19} m^{-3}}$, trapped magnetic flux (based on rigid-rotor model) ${\phi_p}$ ~5–10 mWb, and external axial magnetic field ${B_e}$ ~1 kG. To date, the optimum C-2W discharges have reached ${T_{tot}}$ up to ~3 keV and ${T_e}$ ~300 eV at the peak inside the FRC.
Based on a careful global power balance analysis detailing input/loss channel characteristics and plasma timescales [2], there appears to be a strong positive correlation between ${T_e}$ and energy confinement time. The previously reported C-2/2U scaling of the electron energy confinement time ${\tau}$${_{E,e}}$ still persists at the higher ${T_e}$ (i.e. collisionless plasma regime) in C-2W, as shown in Fig. 3. Given uncertainties in the measurements and assuming the power-law model, regression analysis shows that ${\tau}$${_{E,e}}$ is approximately proportional to ${T_e^2}$ when fitting for the entire ensemble of the C-2W data set.
Dedicated equilibrium and transport simulations have been performed to better understand FRC global stability and confinement, NBI and edge biasing effects, and turbulence in the FRC core and open-field-line plasmas. Simulations predict that parallel electron heat loss is close to the minimal theoretical limit, which has been experimentally validated by end-loss energy analyzers in the outer divertor. Nonlinear kinetic simulations also qualitatively agree with experimental fluctuation measurements, where turbulent transport is greatly reduced by sheared flows due to edge biasing. Google also contributes to advanced data analysis where their Bayesian inversion algorithm reconstructs plasma density profiles and high-frequency fluctuations. Disruptions of fast-ion orbits can be studied from plasma displacement inferred from reconstructions, where detailed correlations with magnetic probes provide further insights into energy loss mechanisms. This paper will review the highlights of C-2W program, including recent experimental results of significantly advanced FRC performance as well as simulations. Future plans will be reported as well.
[1] H. Gota et al., Nucl. Fusion 59, 112009 (2019).
[2] H. Gota et al., Nucl. Fusion 57, 116021 (2017).Speaker: Dr Hiroshi Gota (TAE Technologies, Inc.) -
Introduction. Extrapolations to ITER and DEMO from existing smaller experiments alone are unreliable, especially for turbulent transport - requiring the aid of predictive simulations. The 3D fluid turbulence code GRILLIX [1–4] is used to study confinement improvement through turbulence suppression that is compatible with power exhaust. This contribution describes the validation against experimental edge and scrape-off layer (SOL) measurements from the ASDEX Upgrade tokamak (AUG). Particular focus is put on the role of sheared $E \times B$ flows in turbulence suppression, and the resulting profiles, i.e. fall-off lengths (‘SOL width’). Simulations are also checked against experimental scaling laws [5] over varying parameters, and against smaller machines like C-mod and TCV. Further, we investigate the effect of varying magnetic geometry, in shaped single null as well as in advanced divertor configurations (ADCs).
Verification and validation. We report on the first quantitative validation based on L-mode AUG experiments in standard single null divertor geometry, depicted in Fig. 01. We compare profiles and fluctuations of density and temperature, and the electric field. In agreement with experiment, the ballooning driven fluctuations in the confined region are moderate, < 4%, and rather high in the SOL, > 15%, with highly intermittent events. Obtaining these results was possible after several improvements in the physical model, numerical methods and computational performance. The code has been verified via the method of manufactured solutions [1]. Previously, it was shown to reproduce many features of limited and diverted tokamaks [1, 3], and validated against experimental results from the Large Plasma Device (LAPD) [2], explaining the mechanism of blob formation and propagation.
Shear flows and turbulence suppression. $E \times B$ flow shear [6] and magnetic shear [1] at the separatrix suppress turbulence in the pedestal, which is crucial for the desired confinement in ITER. The $E \times B$ flow shear is explained by the experimentally observed jump in the radial electric field $E_r$ across the separatrix. We show that this is due to different underlying mechanisms in the closed and open field line regions: $E_r \approx \partial_r p_\mathrm{i}/n$ in the confined region, transitioning to $E_r \approx −3 \partial_r T_\mathrm{e}$ in the SOL due to sheath boundary conditions [6]. Additional contributions from rotation and Reynolds stress are damped at higher temperature. The resulting pressure and electric field profiles are shown in Fig. 02.
Dedicated framework for diverted tokamaks. The plasma edge and SOL have to be studied together, because edge turbulence is affected by the plasma-wall interaction in the SOL, and SOL turbulence largely spreads from the plasma edge [7]. Their interface, the separatrix, not only allows to separate the plasma-wall interaction in the divertor from the confined plasma, but it also produces strong, turbulence suppressing magnetic shear. Accurately resolving this dynamics is vital for predicting both the heat flux on the divertor, and the confinement properties of the device. GRILLIX is able to do this thanks to the flux-coordinate independent (FCI) approach [8], which allows for a flexible and efficient treatment of arbitrary magnetic geometries.
Multi-scale electromagnetic model. The model in GRILLIX, by keeping all global dependencies [3], describes the evolution of equilibrium profiles together with turbulent fluxes, allowing fluctuations of arbitrary magnitude. This is important for the prediction of pedestal profiles, i.e. fall-off lengths like the SOL width. Electromagnetic effects are crucial in limiting the response of the parallel current via shear Alfvén waves, as opposed to resistivity [1]. However, electromagnetic cross-field transport (due to field-line flutter) is significant only at higher pressure ($\beta$) and primarily for heat.
Advanced divertor concepts. ADCs such as snowflake and Super-X are promising for decreasing the heat flux to reactor walls [9], but their experimental implementation is so far rare. GRILLIX can support advanced divertor design by simulations [4]. In the snowflake divertor, we find a more complex structure of the electric field then in single null: a potential maximum at the magnetic null, i.e. a convective cell, redirects part of the SOL flows to the secondary divertor legs. Together with the increased magnetic flux expansion, this is promising for improving heat exhaust in DEMO.
Parameter scalings. The comparison to experimental multi-machine scalings, such as the SOL width scaling by Eich et al. [5], is facilitated by direct simulations of different machines like AUG, C-mod and TCV, and by parameter scans. Both plasma pressure and collisionality increase turbulent transport through different mechanisms. Larger machine size and larger toroidal magnetic field decrease the vortex size compared to the machine scale, effectively increasing confinement. In addition to magnetic field strength, the shape of the poloidal magnetic field has a significant impact: turbulence can be suppressed both directly through magnetic shear, and indirectly through a different flow pattern. This effect is especially visible near the X-point.
Conclusion. GRILLIX was validated against AUG experiments and checked against multi-machine scaling laws, which is a crucial prerequisite for ITER and DEMO simulations. The code is able to predict turbulent fluxes, including their suppression in the pedestal, as well as the resulting equilibrium profiles, $E_r$ and fall-off lengths like the SOL width. This was shown both in standard single null as well as in advanced divertor concept geometry.
References
[1] A. Stegmeir et al., Phys. Plasmas 26, 052517 (2019)
[2] A. Ross et al., Phys. Plasmas 26, 102308 (2019)
[3] W Zholobenko et al., Contrib. Plasma Phys. e201900131 (2019),
https://doi.org/10.1002/ctpp.201900131
[4] T. Body et al., Contrib. Plasma Phys. e201900139 (2019),
https://doi.org/10.1002/ctpp.201900139
[5] T. Eich et al., Nucl. Fusion 53, 093031 (2013)
[6] B. Zhu et al., Phys. Plasmas 24, 055903 (2017)
[7] P. Manz et al., Phys. Plasmas 22, 022308 (2015)
[8] A. Stegmeir et al., Comput. Phys. Comm. 198, 139 (2016)
[9] H. Reimerdes et al., Plasma Phys. Control. Fusion 55, 124027 (2013)Speaker: Mr Wladimir Zholobenko (Max Planck Institute for Plasma Physics) -
Introduction – Negative triangularity discharges were first studied on TCV to examine the effect of plasma shaping on energy confinement in ohmic, L-mode discharges 1. Subsequent experiments, using ECRH to stabilize MHD instabilities, showed an improvement of energy confinement in negative triangularity as compared to similar positive triangularity discharges 2. Modulated ECRH to allow measurement of $\chi$e, further revealed the improved energy confinement in negative triangularity over a large range of collisionality 3. The H-mode is foreseen as the operational scenario on ITER and fusion reactors [4]. It is plagued by edge instabilities (ELMs) that would seriously limit the lifetime of the plasma facing components rendering ELMy H-mode operation potentially impractical for a reactor. Negative triangularity offers the possibility of accessing H-mode grade energy confinement in L-mode with reduced edge pedestal pressure and without ELMs. It has the further benefit that, in H-mode operation, the pedestal height, which is strongly correlated to the energy loss per ELM, is reduced by a factor 4 [5]. A recent negative triangularity reactor study has shown the engineering feasibility of a negative triangularity reactor [6]. Recent experiments on DIII-D [7] have shown that $\beta$N $\approx$ 2.7 and HIPB98 > 1 are both possible with an L-mode edge. In addition, there is no degradation of energy confinement with heating power, confirming earlier studies [8]. These observations have encouraged a negative triangularity campaign on TCV with the goal of further understanding the root of the energy confinement improvement at high $\beta$ and validating the scenario as a realistic operating scenario for a future reactor.
![Relative density fluctuation amplitude n ̃⁄n as a function of radial position in a $\delta$ = 0.5 (red) and a $\delta$ = −0.25 (blue) TCV discharge; both discharges have 0.45MW EC heating at the plasma center. (Reproduced from [9].)](https://www.epfl.ch/research/domains/swiss-plasma-center/wp-content/uploads/2020/03/LPorte_IAEA2020_synopsis_Figure1-e1583853866636.jpeg)
Core Turbulence – To better understand the confinement improvement in negative triangularity a series of turbulence measurements were made on TCV. Over a wide range of plasma parameters the turbulence characteristics were measured using both correlation ECE [9] and tangential phase contrast imaging [10]. Figure 1 shows the relative density fluctuation amplitude as a function of minor radius for positive and negative triangularity discharges. There is a clear reduction in fluctuation amplitude across a significant fraction of the minor radius. By using NBH it was possible to vary both the Te/Ti ratio the effective collisionality. It was shown that the reduction of fluctuation amplitude in negative triangularity was maintained compared to positive triangularity as Te/Ti approached unity and when collisionality was reduced to ITER values.SOL Turbulence – A set of experiments has been performed to connect the core turbulence with turbulence in the scrape off layer. Both gas puff imaging (GPI) and reciprocating probe (RCP) measurements have been used. Figure 2 shows the evolution of the ion saturation current and the gas puff imaging brightness as discharge triangularity is stepped from -0.4 to -0.2 to 0.0. There is a clear reduction in SOL turbulence correlated with the improvement in energy confinement and there is a clear reduction in the plasma wall interaction with increasing negative triangularity.
The SOL measurements indicate that the consequences of negative triangularity are confined not solely to the plasma core but manifest themselves in reduced SOL fluctuations and, more significantly, on reduced plasma/wall interaction. It must be mentioned that the discharges described above were ohmic, limited discharges. It is planned to continue these experiments with diverted and additionally heated discharges.
High Performance – Understanding confinement improvement in negative triangularity is essential. It is also imperative to find the operational limits of negative triangularity. Consequently, a campaign was started to investigate the performance limits of negative triangularity. Both diverted and limited discharges were developed. In diverted discharges, where only partial matching of the shape was possible, $\beta$N doubled in negative triangularity compared to positive triangularity with H98(y,2) > 1. In limited discharges, it was possible to heat the plasma using NBH. With little discharge optimisation, $\beta$N $\approx$ 1.7 and H98(y,2) > 1 was achieved. Figure 3 gives an overview of one such discharge. As the beam power ramped up to 500kW there was a linear improvement of $\beta$N and of H98(y,2). There is no degradation of confinement with heating power. These discharges were operated with qsub>95 < 3, the ITER relevant range of qsub>95.
NTMs limited the high performance negative triangularity discharges. No attempt has been made to stabilize the NTMs neither using counter ECCD nor by increasing q95. This is planned during the next campaign. The discharges were operated with moderate negative triangularity and elongation which will be increased in future experiments.
The high-performance discharges were achieved in limited configurations. An effort is being made to produce stable, single null diverted discharges using TCVs independent real time control of shape and position.
Preliminary studies have achieved ITER (and DEMO) relevant plasma performance in negative triangularity, limited, NBI heated discharges. Future work will aim to explore the necessary conditions for negative triangularity to exhibit improved confinement.
Modelling – Efforts have been made to model negative triangularity discharges on TCV using the GENE code [11]. Experimental data has been used as input to the code and a study of the effect of experimental uncertainty on numerical results has been performed. Linear, flux tube simulations show that the discharges produced in TCV are all dominated by TEM and ETG instabilities. In the same linear, flux tube simulations and by varying experimental profiles within the experimental uncertainties it has been possible to reproduce experimental levels of heat flux.
- J-M. Moret et al, Physical Review Letters 79 (1997) 2057.
- A. Pochelon et al, Nuclear Fusion 39, No. 11Yc (1999) 1807.
- Y. Camenen et al, Nucl. Fusion 47 (2007) 510.
- ITER Physics Basis, Editors et al. Nucl. Fusion 39 (1999) 2137–2174
- A. Merle et al, Plasma Phys. Control. Fusion 59 (2017) 104001
- M. Kikuchi et al, 2019 Nucl. Fusion 59 056017
- M.E. Austin et al, 2019 Phys. Rev. Lett. 122, 115001
- O. Sauter et al, Phys. Plasmas 21, 055906 (2014)
- M. Fontana et al, Nucl. Fusion 58 (2018) 024002.
- Z. Huang et al, Plasma Phys. Control. Fusion 61 (2019) 014021.
- G. Merlo et al, Phys. Plasmas 26, 102302 (2019)
Speaker: Dr L. Porte (1Swiss Plasma Center (SPC), Ecole Polytechnique Fédérale de Lausanne (EPFL)) -
In the laser-driven indirect drive scheme for inertial confinement fusion (ICF), the capsule diameter is typically limited to ~2 mm in order to attain quasi-round implosions with currently available laser energy in cylindrical hohlraums. This geometrical factor restricts the energy coupling efficiency from the hohlraum to the capsule to be ~10% [Ref.1]. We report the first series of experiments on NIF demonstrating ~ 30% energy coupling to an Al capsule in a rugby-shaped Au hohlraum [Ref.2]. 3.0mm- and 3.4mm-diameter Al capsules are driven in Au rugby hohlraums at 1MJ and 1.5MJ laser energy, respectively. Measurements of in-flight capsule size, mass remaining, velocity, bang time and neutron yield show good agreement with simulations, consistent with ~300kJ coupling at 1MJ drive and ~500 kJ coupling at 1.5MJ drive. It is found that the shell shape during the implosion is sensitive to the rugby dimensions so that the rugby shape is an effective knob for symmetry tuning. A round imploded shell has been achieved near the bang time for the 3.4mm capsule.
In comparison to a conventional cylindrical hohlraum that has a straight wall, the curved wall of a rugby hohlraum [Ref.3,4] affects the laser irradiation mainly in two aspects: enlarging the laser spot size and enhancing the specular reflection of the laser beams. This results in ~1.4-1.8x larger laser spot and lower intensity of the beams, which is helpful to reduce beam blocking due to the wall bubble expansion. The larger incident angle also leads to higher reflection of the beams, which points toward inside of the rugby hohlraum and is helpful to increase the drive at the waist. These beneficial effects and the simulation setup have been discussed in detail in [Ref.5].
The experiments were preformed using a standard 1D and 2D x-ray radiography platform on NIF. The setup is illustrated in Fig. 1 (a). A Zr foil located at 12 mm from the target center was irradiated by 8 NIF beams to generate a 16 keV backlighter. Measurements of the velocity and mass density profile by the streaked radiography provided a shell kinetic energy of 34 kJ with ~1 MJ laser drive, which is 2-4x that which was achieved in the recent high-foot shots with a ~2 MJ laser drive. Given that the typical rocket efficiency for imploding Al ablator is ~ 10%, 34 kJ shell kinetic energy corresponds to > 300 kJ coupled to the capsule. Fig. 1 (b) shows the simulated energy absorption by the capsule at three scales, 0.7x, 0.9x and 1.0x. The energy coupled to the capsule reaches 350, 500, and 650 kJ with 1.0, 1.5 and 2.0 MJ drive, respectively. Good agreement between the simulated and measured quantities including peak radiation temperature, in-flight radius, velocity, shell kinetic energy and shell width supports the high energy coupling at 0.7x scale [Ref.2].
The sensitivity of the laser energy distribution to the incident angle makes it possible to tune the implosion symmetry by adjusting the rugby wall shape. We have carried 3 NIF shots with different rugby dimensions using a 4-5 ns long reverse-ramp pulse shape with peak power 300-400 TW. The x-ray radiographs are shown in Fig. 2. The wide rugby in Fig. 2 (a) produced a quite prolate shell shape, indicating more drive at the waist than at the laser entrance hole. A narrower rugby with smaller waist diameter in Fig. 2 (b) reduced asphericity. Finally a scale-up of the narrow rugby produced a very round implosion as shown in Fig. 3 (c). This symmetry tunability by rugby wall shape will be very useful for the design of future campaigns.
Nuclear diagnostics were enabled in another shot at 0.9x scale with 7 mg/cc DT gas fill in the Al capsule, providing additional constraints on the energy coupling to the capsule. The measured nuclear burn history is in good agreement with simulated value as shown in Fig. 2(d). The yield reaches 78% of pre-shot prediction. The good agreement between multiple measurements and simulations indicates a coupling energy of ~ 500 kJ in the 1.5MJ drive experiments [Ref.6].
To summarize, we have performed measurements of symmetry, nuclear bang time, neutron yield, in-flight capsule size, and velocity of Al capsules with diameter 3.0-3.4mm in Au rugby hohlraum. The good agreement between data and simulations indicates 500 kJ coupled to the capsule with 1.5 MJ drive. It is demonstrated that the implosion symmetry can be tuned effectively by adjusting the rugby shape. These results open new opportunities for both the mainline single-shell scheme and the double-shell designs toward ignition in ICF, which is a critical step for the development of IFE.
[Ref.1] R. Betti and O. A. Hurricane, Nature Phys. 12, 435 (2016).
[Ref.2] Y. Ping, V. A. Smalyuk, P. Amendt, R. Tommasini, J. E. Field, S. Khan, D. Bennett, E. Dewald, F. Graziani, S. Johnson, O. L. Landen, A. G. MacPhee, A. Nikroo, J. Pino, S. Prisbrey, J. Ralph, R. Seugling, D. Strozzi, R. E. Tipton, Y. M. Wang, E. Loomis, E. Merritt, and D. Montgomery, Nature Phys. 15 (2019) 138.
[Ref.3] P. A. Amendt, C. Cerjan, A. Hamza, D. E. Hinkel, J. L. Milovich and H. F. Robey, Phys. Plasmas 14 (2007) 056312 ; P. A. Amendt, C. Cerjan, D. E. Hinkel, J. L. Milovich, H. S. Park and H. F. Robey, Phys. Plasmas 15 (2008) 012702.
[Ref.4] M. Vandenboomgaerde, J. Bastian, A. Casner, D. Galmiche, J.-P. Jadaud, S. Laffite, S. Liberatore, G. Malinie, and F. Philippe, Phys. Rev. Lett. 99 (2007) 065004.
[Ref.5] P. Amendt, D. Ho, Y. Ping, V. Smalyuk, S. Khan, J. Lindl, D.Strozzi, R. Tommasini, M. Belyaev, C. Cerjan, O. Jones, W. Kruer, N. Meezan, H. Robey, F. Tsung, C. Weber, and C. Young, Phys. Plasmas 26 (2019) 082707.
[Ref.6] Y. Ping, V. A. Smalyuk, P. Amendt, et al. submitted to High Energy Density Physics for IFSA conference proceedings.Speaker: Y. Ping (LLNL) -
We propose a novel heating mechanism for ions in overdense plasmas by introducing two whistler waves along a strong magnetic field in the counter-beam configuration [A]. The essential process is the collapse of standing whistler waves within a short timescale comparable to the wave oscillation period. During the collapse, ions are accelerated by a static electric field and acquire a large amount of energy directly from the electromagnetic waves. This ion-heating mechanism could be applicable to various plasma phenomena in fusion energy sciences and astrophysics and become an alternative scheme of the fast ignition of the laser-driven inertial fusion energy (IFE). Ion temperature heated by standing whistler waves reaches up to the order of 10-100 keV [B], for example, 40 keV for imploded deuterium-tritium (DT) target and 500 keV for proton-boron (pB) plasmas.
The ultimate goal of fusion science is to heat ions to the high enough temperature to promote fusion reactions. In the laser-driven IFE scheme, the laser energy should be converted to the ion energy of fuels as much as possible through various laser-plasma interactions. However, there is a fundamental problem that electrons take away most of the laser energy at the first step of the interaction. Therefore, the development of direct energy transfer from electromagnetic waves to ions has a valuable meaning that overcomes the essential difficulty.
The motivation of our work is to explore an efficient mechanism of the energy transfer from the lasers to overdense ions directly. A possible mechanism of ion heating proposed recently is caused by collapsing standing whistler waves (see Fig. 1) [A]. The whistler wave is an electromagnetic wave traveling along a magnetic field line. The tangential electric field is right-hand circularly polarized to the direction of the external magnetic field. If the field strength exceeds a critical value $B_c = 10 (\lambda_0/1\mu{\rm m})^{-1}$ kT where $\lambda_0$ is the laser wavelength, then there is no cutoff density for the propagation of whistler waves. This crucial feature of the whistler wave enables the direct interaction of the electromagnetic waves and dense plasmas over the critical density $n_c$ with no magnetic field.
As for the initial setup to simulate the standing whistler wave heating, we investigate a simple configuration in which two circularly polarized (CP) lasers irradiate a thin overdense target from both sides. We perform particle-in-cell (PIC) simulations in the 1D Cartesian coordinates. The CP lasers with the frequency $\omega_0$ are injected from both sides of the boundaries of the computational domain. The polarization of the two lasers is right-hand circular toward the magnetic field direction. The normalized vector potential $a_0=e E_0/(m_e c \omega_0)$ characterizes the amplitude of the laser electric field $E_0$ in the vacuum, where $e$ is the elementary charge, $m_e$ is the electron mass, and $c$ is the speed of light. In our problem setup, the computational domain size is broader than the target thickness, so the laser lights coming from the boundaries propagate in the vacuum until they hit the target surface. The refractive index of the whistler wave, which is a function of the plasma density and magnetic field strength, determines the transmittance and reflectivity at the target surface. Since we consider the cases with a strong magnetic field, the transmittance to the whistler wave is typically more than 50%.
A series of PIC simulations reveal an effective use of the standing whistler waves in the heating part of the fast ignition scheme. A reasonable scenario for the enhancement of the energy gain is the standing whistler-wave heating of an imploded DT target by other long-pulse lasers. Here the energy gain is defined by the ratio of the generated energy by fusion reactions to the injected energy of the counter lasers. In our setup, the existence of a uniform magnetic field is a necessary ingredient. Then the cylindrical implosion perpendicular to the magnetic field line would be the best configuration.
We simulate the standing whistler-wave heating of such an imploded DT core plasma [B]. Figure 2 shows the energy spectrum of the DT core after the counter irradiation of heating lasers obtained by a 1D PIC simulation. In this run, the initial target density is assumed to be 30 times higher than the solid density, which corresponds to 1500$n_c$ for the laser wavelength of 1$\mu$m. The external magnetic field is set to be 2.1 MT (200$B_c$) and the injected laser intensity in the vacuum is 10$^{21}$ W/cm$^2$ ($a_0 = 20$). The DT core is found to be heated up to 40 keV successfully. The temperature of each species fitted by the Maxwellian distribution is 43, 45, and 2.0 keV for the deuterons, tritons, and electrons. The estimated energy gain for this case is just above unity. In the high-density limit, the energy gain increases in proportion to the density and laser pulse duration. Therefore, in order to enhance the gain, we need to make the imploded density much higher or the pulse duration of the whistler wave much longer.
In modern fusion researches, the DT reactions are commonly considered because of the lowest reaction temperature. However, the standing whistler-wave heating achieves potentially much higher temperature plasmas, which bring a possibility of ultraclean aneutronic fusion reaction such as the pB reaction. The thermal pB reaction works at the temperature over several hundreds of Kelvin. The ammonia borane (H$_6$BN) is found to be an idealistic target for the pB thermonuclear fusion. We have successfully demonstrated by PIC simulations that the protons and boron ions in the ammonia borane target are heated up to about 500 keV by the standing whistler-wave heating, where the aneutronic pB reaction is highly expected.
[A] T. Sano et al., “Ultrafast wave-particle energy transfer in the collapse of standing whistler waves,” Physical Review E, 100, 053205 (2019).
[B] T. Sano et al., “Thermonuclear fusion triggered by collapsing standing whistler waves in magnetized overdense plasmas,” Physical Review E, 101, 013206 (2020).
Speaker: Takayoshi Sano (Institute of Laser Engineering, Osaka University) -
The uncertainties surrounding the physics of plasma exhaust and its centrality in reactor design require a thorough evaluation of promising alternatives as a precautionary measure to avoid delays in DEMO, if the ITER solution for the divertor could not extrapolate to reactor relevant machines. In this contribution, we review the physics and engineering work carried out within EUROfusion’s work package DTT1/ADC on the subject (see Fig.1), showing with a quantitative assessment that alternative configurations provide a larger operating space than the single null according to multifluid simulations (in particular, lower Argon seeding levels and core concentration; lower separatrix density for comparable divertor protection; greater resilience to high power operations), but also highlighting the many engineering challenges that these configurations entail. The 3D engineering analysis of the alternative designs shows that the balance between port space for remote maintenance and the reinforcement of the supporting intercoil structures, stiffening the structure with respect to out of plane forces, is crucial to achieve acceptable solutions. In addition, active and passive magnetic control, pumping, neutronics and turbulence in alternative configurations will be discussed with quantitative analyses.
With alternative configuration we define here any divertor solution that cannot be qualified by ITER and it includes, but is not limited to, Double Null, Snowflake, Super-X and X divertors. As a risk mitigation strategy, EUROfusion has worked on understanding the physics and engineering of new exhaust configurations for reactor relevant devices and DEMO in particular. The objective is to provide a physics and engineering assessment of the usefulness and feasibility of alternative divertor configurations for DEMO by December 2023.
The primary activity of this project is to deliver integrated results through a “loop” where physics and the engineering concepts for sufficiently well-developed alternative designs are synergistically iterated and optimized, see Fig.1. In parallel, it develops more refined techniques and novel concepts that will become part of the “loop” when (or if) they reach maturity. The first iteration between physics and engineering is completed and its results will be presented. All alternative configurations investigated could be generated with external coils only and respecting the engineering force constraints on the poloidal field coils and the central solenoid.
As far as physics is concerned, detachment behaviour in the alternative configurations was assessed at an unprecedented level by comparing each design for an identical and large range of fuelling and Argon seeding influxes. The analysis is performed with state-of-the-art tools (SOLPS) so that future simulations will enable the correct understanding of complex effects such as the fluxes from neutral recycling and kinetic neutral effects. Also, the resilience of the operating point with respect to variations is systematically examined with this procedure for the first time, leading to a full map of the plasma response to changes in fuelling and seeding. Hence, detachment onset, depth and stability are obtained for these configurations, rather than just the operating point. The simulations show, for example, that the super-X and X-divertor configurations are effective in broadening the operating range by allowing acceptable heat fluxes and temperatures at the divertor at a significantly smaller Argon seeding level than the single null, a result reported in Fig.2. In the double null configuration, a lower fraction of the radiated power is localised in the core with respect to the single null, thus providing more margin.
The multifluid studies were complemented by 3D turbulent simulations, assessing for the very first time the behaviour of the anomalous perpendicular transport in alternative configurations. Four state of the art codes (GBS, GRILLIX, STORM and TOKAM3X) were used to simulate all the alternative configurations albeit on a reduce scale to make the computational cost manageable. The simulations show that turbulence is enhanced in the long leg of the Super-X divertor and that the Snowflake generates at the X-point a convective cell induced by drifts that helps redistributing the particles and heat to the four legs (this is similar to the churning mode, but the perturbation is electrostatic). This insight not only sheds light on the fundamental mechanisms behind alternative designs, but it will also help optimize the configurations in the next iteration by refining the SOLPS modelling.
In addition, a new detailed engineering analysis of the structural loads in the toroidal field coils of each configuration will be presented, including an assessment of the importance of the out of plane forces versus the hoop forces, which contribute to ~30% of the total. Our results show that former play an important role in taking the configurations beyond the acceptable stress threshold, thus suggesting that more rigid configurations could be beneficial. Full 3D builds (see Fig. 3), including port access for remote maintenance and intercoil structures have been produced and the technical complexities associated with them will be discussed, including the difficulties associated with the extraction of the divertor cassette in Super-X and Snowflake designs. Controllability of all the solutions has been investigated for variations of beta poloidal and Li of order of 10%. Large excursions of the strike point and significant changes of shape in the Snowflake configuration were observed, while the X-divertor configuration produces a visible but acceptable sweeping on the divertor plates and separatrix of the Super-X tends to get closer to the unprotected upper wall (24cm of vertical excursion). In addition, the paper will give a review of a number of issues associated with alternative configurations, such as helium pumping, which preliminary results predict to be acceptable in all configurations; force constraints on the structures during the whole discharge from ramp up to ramp down, which might be problematic; and detailed neutronic calculations of relative tritium breeding ratios, irradiation of the coils and nuclear heating performed with the Monte Carlo particle transport code MCNP based on geometry from the CAD design of the different configurations, with particular attention to the potential shielding that alternative configurations could provide.
The work carried out suggests that a milder version of the alternative configurations, combining some concepts together (for example a version of the Super-X with shorter leg and significant poloidal flux expansion) could be more suitable for reactor conditions.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
Speaker: Dr Fulvio Militello (UKAEA/CCFE, Culham Science Centre) -
The Japan Establishment for a Power-laser Community Harvest (J-EPoCH) is proposed as a next generation laser facility having multi-purpose high repetition laser beams at the maximum rate of 100 Hz. The omnidirectional 12 laser beams with 8 kJ would yield ~$10^{13}$ neutrons with a Large High Aspect Ratio Target (LHART) (1). As one of the applications of J-EPoCH, a laser fusion subcritical research reactor has been conceptually designed based on existing technologies. Moreover, we can conduct a variety of fusion engineering studies: energy conversion, tritium (T) breeding, neutron irradiation effects, etc. Feasibilities of T breeding and preliminary energy conversion studies are proven according to calculations by the Particle and Heavy Ion Transport code System (PHITS) (2).
Fusion researches have been focusing on plasma physics. ITER is under construction to prove the feasibility of fusion burning. Its operation will start with hydrogen plasma in 2025 and then will change to deuterium-tritium (DT) plasma in 2035 (3). During the phase of DT plasma experiments, T breeding and power generation will be partially studied with test blanket modules. After that, DEMO reactors are proposed without any integrated experiment as a fusion system. Especially tokamak DEMO reactors are designed to generate several GW of thermal energy and to produce ~100 kg of T per year (4, 5). On the other hand, the Laser Inertial Fusion Test (LIFT) reactor has been designed as an experimental reactor with three technical phases (6). The laser fusion system will be developed step by step. However no facility with a high repetition and high power laser system exists to demonstrate required technologies. Eventually, a lot of engineering issues still remain to realize the fusion reactors. We propose the laser fusion subcritical research reactor with J-EPoCH which makes it possible to conduct a variety of fusion engineering studies on energy conversion, T breeding, neutron irradiation effects, etc.
J-EPoCH would be capable to stably generate laser fusion pulse neutrons of $10^{13}$ n/shot with LHART as a laser fusion subcritical research reactor. Figure 1 shows the system which consists of a core, a target delivery system, a vacuum chamber and a vacuum pump. The core is replaceable to conduct various fusion engineering experiments. Each LHART is contained in a capsule, which is delivered at the center of the core shot by shot. A linear motor driven injector can deliver targets with high repetition of more than 1 Hz. The opportunities of neutron irradiation with fluence of ~6.6 x $10^{13}$$\, \mathrm{n/m^2}$ per shot would be provided at a radius of 10 cm. Figure 2 shows the cross section of the concentric core. Neutron – thermal (n-t) conversion by $\rm{B_4C}$ and T breeding by LiPb are preliminarily estimated by PHITS. The energy deposition to the $\rm{B_4C}$ core from a DT fusion reaction is calculated as shown in Figure 3. An alpha particle does not reach to the $\rm{B_4C}$ layer. The thermal energy of 8.72 MeV would be converted from a 14.1 MeV neutron in the $\rm{B_4C}$ layer. It corresponds to 14 J thermal energy per shot. Temperature raise of several mK should be observed per shot. The numerical estimation of the thermal energy conversion will be validated by experiments with the subcritical research reactor. Figure 4 shows T yield in the LiPb core by a fusion neutron. A T yield is expected to be 9.1 x $10^{11}$ /shot. The feasibility applying the subcritical research reactor to fusion engineering researches is proven above. Other candidate materials for n-t conversion and T breeding are able to be assessed by replaceable cores.
The laser fusion subcritical research reactor is attractive and worth constructing for systematical studies on neutron irradiation effects, n-t conversion, T breeding and so on as changing core materials and experimental conditions.
References
(1) T. Norimatsu, et al., J. Vac. Sci. Technol. A6(4), 2552 (1988).
(2) T. Sato, et al., J. Nucl. Sci. Technol. 55, 684-690 (2018).
(3) B. Bigot, et al., Fusion Eng. and Des. 146 (2019) 124-129.
(4) G. Federici, et al., Fusion Eng. and Des. 89 (2014) 882-889.
(5) K. Tobita, et al., Fusion Eng. and Des. 81(2006) 1151-1158.
(6) T. Norimatsu, et al., Nucl. Fusion 57 (2017) 116040.Speaker: Dr Akifumi Iwamoto (National Institute for Fusion Science) -
A world-class ultraintense laser LFEX at ILE, Osaka University directly heated a CD shell target, imploded by GEKKO XII(GXII) laser. Illuminating LFEX energy of 246 J increased the core internal energy by $23\pm 3$ J, leading to the conclusion that the heating efficiency is $9\pm 0.8$ %. The results encourage the fast ignition scheme fusion as a hopeful candidate of the fusion machine.
For the fast ignition fusion[ref1], the direct illumination of an ultraintense laser onto the core is the simplest technical and economical way. To demonstrate the feasibility of direct core heating, we have performed experiments with difficult illumination configurations, such that (a) LFEX is coaxial to the GXII bundled beams axis (Fig. 1(a)), and (b) LFEX is transverse to the axis (Fig.1(b)). (a) mode will be the simplest scheme, but the cutoff point is far from the core and much plasma clouds may block the hot electron and ion transports[ref2]. (b) mode may not be so simple to operate the power plant, but the cutoff point is close to the core and there are less cloud plasmas, which block the transport[ref3]. However, because the transverse mode results are not yet confirmed, we concentrate to the axial (a) mode. Counter illuminating 6 beams from the GXII has imploded a CD shell target. Ratio of shell radius $R$ to focal distance $d/R$ is -3. LFEX laser is directly focused to the core center.
The intensity on target is $0.5\sim 1\times10^{19}$ W/cm$^2$. We have estimated the LFEX fast heating efficiency from the following procedures.
![a) DD reacted-proton tracks without LFEX on 0.4 mm$\times$0.6 mm-square CR-39 films 5 cm far from the target[ref4]. DD reaction has two branches, one of which is D+D $\longrightarrow$ n(2.45 MeV)+ H${^3}$(0.82 MeV)(50 %). Here, we use the another branch, D+D $\longrightarrow$ p(3.02 MeV)+ T(1.01 MeV)(50 %). Red columns and arrow are spectral shifts and their peak along the transverse direction. Green to the axial direction, respectively. (b) X-ray streak images of shell flow. y axis is space. Shell center is at 0 $\mu$m. x axis is time. GXII peak is at 1.8 ns. The core emission is +200 ps later(upper arrow),and LFEX is +600 ps(lower arrow). (c) LFEX is at +200 ps, just at the core emission timing, showing the best compression. Timing jitter is $\pm100$ ps.](http://www.gpi.ac.jp/contents/candy/fig/FEC2020/kitagawa/Fig2FEC2020Kitagawa_ver1.jpg)
Without LFEX illumination, Fig. 2(a) shows a DD reacted proton energy peak (3.02 MeV) shift down to 750 keV (red arrow in Fig.2 (a)) due to the core plasma, yields an areal density of 0.016 g/cm$^2$ along the transverse direction. Peak shift to 1500 keV (green arrow in Fig. 2 (a)) yields 0.011 g/cm$^2$ to the axial direction[ref4]. Considering an ellipsoidal core figure, these results lead us to the core density $\rho=2.8\pm 0.3$ g/cm$^3$ and volume $V = 3.2\times10^{-7}$ cm$^3$.
An x-ray streak camera, in Fig. 2 (b), shows both the core emissions due to the implosion +200 ps after the GXII peak and the LFEX heating +600 ps after. The intensity of core emission here is same as that without LFEX, but once LFEX is just on the compression, as in Fig. 2 (c), the core emission becomes $\sim$20 times stronger than the emissions without LFEX. The stagnation period $\tau$, within which the maximum compression continues, $\leq 50$ ps. We assume that the neutron generation period is also close to $\tau$.
![(a) Neutron TOF signals from MANDALA (Multichannel single-hit neutron scintillation detector array[ref2]) without LFEX: $N_y= 3.5\times10^6/4\pi$, and (b) signals with LFEX of 246 J: Total Ny = $9.8\times10^6/4\pi$; Red curves are the best fitted spectra. (c) X-ray core emission from a pinhole camera corresponding to (a). (d) X-ray core emission corresponding to (b). Emission peak(d) is 4 times larger than (c).](http://www.gpi.ac.jp/contents/candy/fig/FEC2020/kitagawa/Fig3FEC2020Kitagawa_ver1.jpg)
$3.5\times10^6$ neutron yields(Ny) without LFEX illumination in Fig.3 (a) gives us the ion temperature $T_i$ to be 700 eV, if we assume $Ny = n_i^2/4<\sigma v >_{Ti} V \tau$. While, with LFEX at +200 ps, Ny of Fig.3 (b), $9.8\times10^6$ gives $T_i\sim$1.0 keV. The ion temperature increment $\Delta T_i$ is 0.3 $\sim$ 0.4 $T_i$, or $\sim$300 eV.Supposing $T_i \sim T_e$ in equilibrium, then $\Delta T_e\sim\Delta T_i\sim300$ eV with LFEX. The intensity of x-ray emission of the core (Fig.3(d)) is 4 times larger than the Fig.3 (c), leading to the electron temperature increment $\Delta T_e$ is ($4^{1/4}$-1)$T_e=0.4T_e$, if the core is yet in equilibrium. The increment of core energy is given by $\Delta E = \Delta E_e + \Delta E_i = 3/2(n_e\Delta T_e + n_i\Delta T_i) V$, where $\Delta E_e$ and $\Delta E_i$ are the electron and ion contributions, respectively. Assuming the core plasma is fully ionized, then $n_e=Z_{CD}n_i$, where $Z_{CD}$ is total charges of a CD ion and also using $\Delta T_e\sim\Delta T_i$, we could estimate $\Delta E$$ = 3/2V\Delta T_i n_i(Z_{CD}+1)$$ = 23\pm3$ J. Since the LFEX energy is 246 J, we expect the heating efficiency to be $9\pm 0.8$ % for the axial mode. Without LFEX, the implosion efficiency, defined as a core internal energy divided by an implosion laser energy (now GXII), is $53\pm7 {\rm J}/1.7 {\rm kJ} = 3.1\pm7$ %. With LFEX, we can estimate that the total implosion efficiency is improved to 85 J/1946 J = 4.4 %.
We will discuss the transverse mode later on.
[References]
[ref1]Y. Kitagawa et al., J. Physics: Conf. Series 688 012049 (2016).
[ref2]Y. Kitagawa et al., Phys. Rev. E 71 016403 (2005): J. Plasma Fusion Res. 81 384 (2005).
[ref3]Y. Kitagawa et al., Phys. Rev. Lett. 114 195002 (2015): Nucl. Fusion 57 076030 (2017).
[ref4]Y. Kitagawa et al., Phys. Rev. Lett. 75 3131 (1995).Speaker: Prof. Yoneyoshi Kitagawa (The Graduate School for the Creation of New Photonics Industries) -
Introduction
An interesting phenomenon known as density incrustation at the interface of high Z / low Z plasmas was recently reported in high energy density (HED) systems like inertial confinement fusion (ICF) 1. Radiation transport and hydrodynamic motion of materials are intricately coupled in HED systems and their interplay gives rise to several interesting phenomena. Density incrustation (DI) is one such phenomenon characterized by the appearance of a cooling layer at the interface. Formation of cooling layers also occur in radiative shocks where radiation transport from the shock front preheats the upstream matter, cools the matter downstream and leads to a peak in temperature at the shock front. However, the phenomenon of DI is physically distinct from that of radiative shocks. This may arise at the interface when high Z plasma maintained at very high temperature drives a shock wave as well as a Marshak wave into the adjacent low Z material. DI increases the Atwood number at the interface increasing the growth rate of RT instabilities which is detrimental to the implosion dynamics.
The phenomenon was first observed by Meng et al. 1 in their study on Au-CH plasmas. Au at normal density was held at a temperature of 1 keV adjacent to CH at 1 eV in their work. CH was at an initial density of 10 g/cc whereas the ambient density of CH is usually ~ 1 g/cc. A sharp rise in density at the interface accompanied by a drop in temperature was observed in their study. The phenomena may appear in direct drive fusion applications 2 and in gas filled hohlraums where an increase in density was observed in the laser ablated Au plasma adjacent to the filling gas [2, 3]. However temperatures ~ 1keV are usually not encountered in laser driven ICF experiments and the density of the low Z material is also not as high. The motivation behind the present work was to evaluate whether the physical phenomenon of DI appears at lower temperatures presently attainable in ICF experiments with low Z materials at their ambient density.Details of simulations
We performed 1D radiation hydrodynamics simulations using RADHYD1 developed in our group to study DI [4]. In our simulations, Au was maintained at 500 eV adjacent to CH at 1eV. The thicknesses of both the slabs were 1000 μm each. Au and CH were assumed to be at their ambient densities.
Material equation of state (EOS) and opacity data were generated from the wide range EOS library SBCRIS and OPIND package respectively. Solid and fluid phases were modeled separately and matched at normal density to generate a wide range EOS library SBCRIS for materials applicable to HED systems [5, 6]. The opacity of Au and CH plasma for the required range of densities and temperatures were generated using opacity package OPIND developed by us [7]. OPIND is based on Average atom model and is applicable for thermodynamic regime where local thermodynamic equilibrium is a valid approximation.Result and Discussion:
Our aim was to study whether the phenomenon of DI appears if Au plasma is maintained at lower temperatures. To this end, we varied the temperature of Au plasma from 300 eV to 500 eV. We observed that DI does occur and the magnitude of the incrustation increases steadily with rise in temperature as shown in Figure 1.
We observed an increase in the magnitude of incrustation with increase in initial density of the low Z plasma as well in our simulations. This indicates that at higher temperatures of Au plasma ~1keV and high density of CH plasma, higher magnitude of DI will be observed 1. This was verified in our simulations as well. The evolution of the DI at the interface and the shock wave inside CH is presented in Figure 2. We observed that the phenomenon is not transient and peak in density due to DI is maintained at later instants of time as well.
DI arises because of the difference in the spatial scales of hydrodynamic motion and radiation transport. The spatial scale of hydrodynamic motion (~ cm) is the maximum local sound speed multiplied by the time. The pressure in Au plasma changes due to hydrodynamic motion for this spatial scale. However, the spatial scale for radiation transport (~ μm) is mean free path of photons in Au plasma. For this spatial scale, temperature changes due to radiation transport at Au-CH interface.
Initially, Au plasma releases energy to the cold CH via radiation transport leading to formation of a cooling layer at the interface. Size of the cooling layer is large compared to Rosseland mean free path but small compared to the hydrodynamic spatial scale. So although temperature in the cooling region drops, pressure remains almost constant. This leads to a rise in density of Au plasma at the interface. These features are clearly visible in Figure 3.
Conclusion
We studied the phenomena of density incrustation arising at the interface of low-Z CH and high-Z Au. This combination of materials is commonly used in ICF experiments. We generated EOS and opacity data for the materials using SBCRIS library and program package OPIND respectively. In this work, Au is maintained at a high temperature (~ 500 eV) which drives shock wave and Marshak wave into CH. At the same time rarefaction occurs in Au at the interface. Simultaneously, there is a sharp increase in the density of Au at the interface. This arises because of the difference in the spatial scales of hydrodynamic motion and radiation transport and different opacities of materials at the interface.
Our results exhibit features similar to that observed by Meng et-al 1. However, they considered Au plasma at a temperature of 1 keV in their simulations leading to a much higher magnitude of DI. We observed the physical phenomena albeit the magnitude was smaller. We conclude that the appearance of DI is definitely a physical phenomenon and not a numerical artifact but its magnitude strongly depends on the temperature of Au plasma. DI may increase Atwood number leading to growth of RT instability in the HED systems as observed by Meng et al. Hence we have to look for conditions that can possibly reduce the magnitude of the incrustation to as low as possible. Maintaining high Z plasmas at lower temperature (upto 500 eV) and keeping the low Z material at ambient density do lead to reduced incrustation at the interface and hence reduced growth rate of RT instabilities.References
1. G. W. Meng et al., Matter and radiation at extremes, 1, 249 (2016).
2. M. Dunne et al., Physical Review Letters, 75, 3858 (1995).
3. G. Huser et al., Phys. Plasmas, 16, 032703 (2009).
4. M. K. Srivastava, C. Bhattacharya and M. Das, internal report, BARC (2017).
5. C. Bhattacharya and M. K. Srivastava, J. App. Phys., 102, 064915 (2007).
6. C. Bhattacharya, M. K. Srivastava and S. V. G. Menon, Physica B, 406, 4035 (2011).
7. M. Das, M. K. Srivastava, S. V. G. Menon, JQSRT 113, 286 (2012).Speaker: Dr Chandrani Bhattacharya (Bhabha Atomic Research Centre) -
The divertor target is the most intense plasma-surface interaction area in tokamaks. Currently, the control of power load on the targets becomes to one of the most important issue for high performance long-pulse discharges. EAST has achieved over 100 s high performance operation, however, its lower graphite divertor prevents its achievement of further high-power long-pulse discharges $[1]$. Tungsten (W) has been chosen as the divertor plasma-facing materials (PFMs) for ITER $[2]$ and the main candidate for CFETR. To demonstrate the performance of W as the divertor target material for reactor and solve the power exhaust problem, the upgrade of EAST lower divertor into W material is undertaken. In this work, the physical design of the new EAST lower W divertor is presented. By using the 2D edge plasma code SOLPS modeling $[3][4]$, the new optimized divertor shape for the configuration is proposed. The erosion and W transport during the external impurity seeding has also been studied using the new divertor shape.
The SOLPS5.0 code is applied for the modeling. At the beginning, to confirm the optimized shape of the EAST lower divertor, the carbon divertor is assumed with the physical and chemical sputtering included and the constant radial transport coefficients are used. The systematic examination of different target shapes, target angles and the pump locations has been carried out. By comparing the horizontal target, vertical target, the strike point position, the horizontal target with the SP locating near the closed corner is proposed. The simulation result reveals that the divertor closure plays significant role in divertor plasma, while the target inclining angle influences plasma slightly. It is confirmed that the pump slot location has a remarkable impact on the particle exhaust, which influences the divertor plasma significantly. Higher pumping speed reduces the neutral density and radiated power, thus increases Te as well as the heat flux to the target, as shown in Fig. 1. For pumping at the SOL side, the particle removal rate is the essential physical quantity to the divertor particle dynamics. For a given particle removal rate, however, the plasma conditions are insensitive to the pump location within the scrape-off layer (SOL) side.
After the confirmation of the new divertor shape, the performance of the W divertor is further investigated. The divertor target material uses W material by considering the particle reflections $[5]$. To study the effect of external impurity seeding and the control of W impurity in the newly designed divertor, the performance of argon (Ar) and neon (Ne) seeding into the plasma, and the resulting W target erosion and its impurity transport is simulated by the SOLPS $[4]$ coupled DIVIMP modeling $[6]$. Two external impurity gas seeding positions are compared, i.e. puffing at scrape off layer (SOL) and private flux region (PFR). The puffing rate scan with argon has been done. The power crossing the core-edge interface (CEI) $P_{SOL}$ = 4 MW, $n_{D+}$ at CEI is fixed to $4.5 \times 10^{19} m^{-3}$ ($n_{e,sep}$ ~ $1.5 \times 10^{19} m^{-3}$). The simulation results indicate that seeding at scrape-off layer (SOL) side is better than seeding at private-flux region (PFR) by considering the divertor power dissipation and impurity screening as shown in Fig. 2. By comparing the Ne and Ar seeding, it is found that to achieve the similar divertor plasma condition, Ar seeding is better for divertor impurity screening as shown in Fig.3 the W impurity density at the core region. However, Ar may cause more serious W erosion, resulting in severe core contamination by W impurity. Our study indicates the erosion of W is still a critical problem during the Ar seeding, even with Te at the target below 10 eV. To reduce the target erosion and enrichment of W impurity in the core region, the deuterium gas puffing at upstream SOL region combined with external impurity seeding at divertor region is proposed.
This work presents the divertor shape design of the EAST lower divertor to find out the optimized shape for the long-pulse discharge. Moreover, it illustrates application of the external impurity seeding to achieve the radiative divertor optimization for the W PFMs. Moreover, these studies will improve the understanding of W target sputtering and W impurity transport control during the radiative divertor discharges for CFETR/DEMO.
Acknowledgements
This work is supported by National Key R&D Program of China Nos. 2017YFA0402500, 2018YFE0301101, 2017YFE0301206, 2017YFE0300402 and 2017YFE0300501, National Natural Science Foundation of China under Grant No. 11775044.References
$[1]$ X.Z. Gong et al., Plasma Sci. Technol. 19 (2017) 032001.
$[2]$ R. A. Pitts et al., Nucl. Mater. Energy 20 (2019) 100696.
$[3]$ C. F. Sang et al., Nucl. Fusion 57 (2017) 056043.
$[4]$ C. F. Sang et al., Plasma Phys. Control. Fusion 59 (2017) 025009.
$[5]$ C. Zhang et al., Plasma Phys. Control. Fusion 61 (2019) 115013.
$[6]$ P. C. Stangeby and J. D. Elder, J. Nucl. Mater. 196-198 (1992) 258.Speaker: Prof. Chaofeng Sang (Dalian University of Technology) -
A new fluid plasma transport solver MAPS (MFEM 1 Anisotropic Plasma Solver) is being developed for the simulation of far scrape-off-layer (SOL) radio frequency antenna simulations. MAPS solves a coupled set of particle, parallel momentum, and energy equations for plasma and neutral species using a finite element approach based on the MFEM (Finite Element Discretization Library) framework 1. The code uses a generalized implementation supporting 2D or 3D geometries on an unstructured mesh, allowing for a grid that conforms to complicated far-SOL antenna and limiter structures. The highly anisotropic plasma transport is addressed via the use of high order finite elements, as used in extended MHD codes [2,3]. This choice of numerical methods can support cross-field drifts and kinetic corrections, which are typically neglected in 3D fluid plasma transport codes. The development of MAPS will lead to important verification tests with the leading 3D edge fluid plasma code EMC3-EIRENE [4], which uses a field-aligned mesh and a Monte-Carlo numerical method to handle the transport anisotropy.
MAPS solves continuity equations for a fluid neutral atom species and a single hydrogenic ion species, a total plasma parallel momentum equation, and energy equations for electrons and ions (the latter two in the form of static pressure evolution). A simplified set of these equations is written as
$\frac{\partial}{\partial t}n_n = \nabla\cdot(D_n \nabla n_n)-S_{iz}$
$\frac{\partial}{\partial t}n = -\nabla_{||}(n v_{||})+\nabla_{\perp}\cdot(D_{\perp} \nabla_{\perp} n)+S_{iz}$
$\frac{\partial}{\partial t}mnv_{||} = \nabla\cdot(\bar{\bar{\eta}}\cdot \nabla v_{||})-\nabla_{||}(p+mnv_{||}^2)+\nabla_{\perp}\cdot(mv_{||}D_{\perp} \nabla_{\perp} n)$
$\frac{3}{2}\frac{\partial}{\partial t}p = \nabla\cdot(n\bar{\bar{\chi}}\cdot \nabla T)+\nabla_{||}p-\frac{5}{2}\nabla_{||} pv_{||}$
$\;\;\;\;\;\;\;\;\;\;\;\;+\frac{5}{2}\nabla_{\perp}\cdot(n\chi_{\perp} \nabla_{\perp} T)-\frac{D_{\perp}}{n}\nabla_{\perp}n\cdot\nabla_{\perp}p + S_E$
where cross-field drifts, volume recombination, and electric fields are currently neglected. The static pressure equation is the same for electrons and ions, with an appropriate source term $S_E$ for each species (equipartition and terms due to plasma-neutral interactions). The diffusivity tensors $\bar{\bar{\eta}}$ and $\bar{\bar{\chi}}$ have the form $\bar{\bar{D}} = D_\perp(\bar{\bar{I}}-\bar{b}\bar{b})+D_{||}\bar{b}\bar{b}$ with the parallel components from classical theory and ad-hoc cross-field components. The particle source is $S_{iz}$, with neutral atoms taken to have a constant energy of ~3 eV. The transport equations are discretized using discontinuous Galerkin finite elements of arbitrary order. Time integration uses high-order singly-diagonally implicit Runge–Kutta (SDIRK) methods, with a novel time step selection using a PID controller [5]. Adaptive mesh refinement is based on a weighted error estimate of each field.
The anisotropy in the plasma transport is addressed by the use of high order finite elements. The method has been benchmarked against a general anisotropic diffusion equation test problem, as described in Ref. 4. Acceptable levels of numerical pollution of the cross-field transport are obtained for a diffusivity ratio of 106 - 109 using finite element orders > 2. The results also depend on the grid resolution [5]. Figure 1 shows results from a test problem where a constant temperature field on closed magnetic flux surfaces is perturbed by a Gaussian energy source. Acceptable levels of pollution are found for finite element orders > 2 with > 16 interpolation points across the feature of interest. The anisotropy benchmarks exhibit good behavior at extreme diffusivity ratios that can exist in the core of tokamak plasmas, however the target problem of far-SOL simulations will exhibit significantly lower anisotropy.
The full set of transport equations in MAPS is currently being benchmarked against a 2D fluid plasma transport code developed to study island transport in LHD [6]. Figure 2 shows a comparison of MAPS and analytic solutions for simple transport in an annular cylindrical geometry. Comparison for a more complicated case with an imposed island chain [6] is in progress. Further extensions of the code include the addition of an electric potential equation using the approach employed in BOUT++ [7]. Following this, MAPS will be coupled to a wave solver to explore issues of RF-plasma coupling, the effect of the ponderomotive force, and the impact RF-enhanced sheaths on divertor fluxes and sputtering. A natural coupling is to the MFEM Stix miniapp 1, also under development as part of the RF-SciDAC.
Work supported by the US DOE under DE-AC05-00OR22725
1 MFEM: Modular Finite Element Methods Library, mfem.org, doi.org/10.11578/dc.20171025.1248
2 S.C. Jardin, J. Comp. Phys 200 (2004) 133.
3 C.R. Sovinec, et al., J. Comp. Phys. 195 (2004) 355.
4 Y. Feng, et al., J. Nucl. Mater 266 (1999) 812.
5 M.L. Stowell, et al., APS-DPP, Portland OR, 2018.
6 M. Kobayashi, et al., PET conference, San Diego CA, 2019.
7 N.M. Li, et al., Comp. Phys. Comm. 288 (2018) 69.Speaker: Jeremy Lore (ORNL) -
We developed a hydrogen population code by combining the Neutral-Transport code with the rovibrationally resolved Collisional-Radiative model (NT-CR), EMC3-EIRENE code, and the Molecular Dynamics (MD) simulation of carbon (C) and tungsten (W) divertor plates. Using this code, (i) we successfully treated hydrogen molecule H$_2$ reactions relating to the molecular assisted recombination (MAR), which plays a significant role in the detached plasmas. Moreover, (ii) we calculated the H$_2$ population which is transported from the divertor plate in LHD plasma. (iii) We identified the H$_2$ MAR reaction which reduces the kinetic energy of plasma in the divertor region most dominantly. We also found that (iv) the distribution of the H$_2$ population near the divertor plate becomes the Boltzmann distribution, which is consistent qualitatively with the estimation by emission spectroscopy of the Fulcher-$\alpha$ band of H$_2$ in LHD.
Plasma detachment is one of the expected phenomena to decrease the heat flux to the divertor plate. It is pointed out that the MAR plays a significant role to decrease the heat flux in the detached plasmas. To analyze the MAR, Sawada, et al.[A] developed the NT-CR code and they found that the rate coefficients of H$_2$ MAR reactions in the divertor plasmas strongly depend on the initial rovibrational state of H$_2$. Neutral transport codes are widely used to analyze divertor plasmas. However, the H$_2$ rovibrational population has not been calculated in other neutral transport simulation codes prior to our NT-CR code. By combining the NT-CR code; EMC3-EIRENE code; and the MD code of C and W divertor plates, we successfully calculated each H$_2$ population with the vibrational state ($v$) and rotational state ($J$) at any position in plasma. In our combining code, the roles of the three codes are as follows. (R1) The spatial distributions of the electron temperature (T$_\textrm{e}$) and density (n$_\textrm{e}$) in plasma are calculated using the EMC3-EIRENE code[B]. (R2) The initial H$_2$ rovibrational state is calculated for carbon or tungsten divertor plates by the MD code[C]. (R3) Using T$_\textrm{e}$ and n$_\textrm{e}$ from R1 and the initial H$_2$ rovibrational state from R2, the NT-CR code calculates the H and H$_2$ populations in the plasma. Here we comment that the initial H$_2$ states depend on the material structure of the divertor plate, especially in tungsten. As the first attempt, we adopted the amorphous carbon or the crystal tungsten, respectively (see Fig.1).
Using our combining codes, we calculated the H$_2$ rovibrational populations for the following three cases: (case 1) a certain LHD plasma with carbon divertor plates; (case 2) the divertor plasma region for detach plasma model with carbon divertor plate; and (case 3) the divertor plasma region for detach plasma model with tungsten divertor plate.
In case 1, we evaluated the H$_2$ rovibrational population produced in the LHD plasma as Fig.2[D]. The H$_2$ populations on the surface of the divertor plate (the purple line in Fig.3) reduce to the population at the point X (the red line), which means that the high J-components of the H$_2$ population disappear by MAR reaction while the H$_2$ is transposing from the divertor to the point X. After all, the J-dependence of each H$_2$ population at Point X for $v$ = 0,1,2 becomes the Boltzmann distribution as the red lines in Fig.3, which is consistent qualitatively with the estimation (Fig.4) by emission spectroscopy of the Fulcher-$\alpha$ band of H$_2$ in LHD.
In case 2, we focus on the divertor plasma region where the MAR is expected to be more dominant than in the core plasma region. Using the combining code, we discovered that the excitations of H$_2$ between the vibrational states remove the kinetic energy from the plasma under a certain detached plasma condition. Moreover, the excitations for $J$=0 is the most dominant. This discovery cannot be achieved by other simulation codes because of the following two reasons: (a) MD can analyze the rovibrational state of H$_2$ generated from the plate. On the other hand, other simulations of the plate ($\textit{e.g.}$, binary collision approximation, phenomenological wall model) are quite difficult to consider the rovibrational level of H$_2$. (b) Our NT-CR code can treat the reactions related to the rovibrational level of H$_2$, whereas the other NT codes seldom include the rotational level processes.
Last, in case 3, we also consider tungsten bcc-crystal Fig.1 for the divertor plate. We perform the MD and NT-CR simulations similar to that in the carbon target. Thus we obtain the rovibrational population of H$_2$ for the tungsten, which is smaller than the carbon case. The reason is as follows: the tungsten crystal generates H$_2$ much less than the amorphous carbon. However, the tungsten surface becomes more complicated as the plasma irradiation time becomes longer.
[A] K. Sawada and M. Goto, Atoms 4 (2016) 29.
[B] G. Kawamura et al, Plasma Phys. Control. Fusion 60 (2018) 084005.
[C] S. Saito, et al., 17th PET, 19-21 Aug., 2019, USA; Contrib. Plasma Phys. (accepted).
[D] K. Sawada, et al., 17th PET, 19-21 Aug., 2019, USA; Contrib. Plasma Phys. (accepted).
Speaker: Prof. Hiroaki Nakamura (National Institute for Fusion Science) -
Here we report the mechanism of plasma heating by magnetized fast-isochoric (MFI) heating scheme (1). The mechanism was visualized experimentally by combining spectroscopic and spatially- and temporally-resolved X-ray imaging techniques. The MFI scheme employs an external magnetic field for guiding a high-intensity relativistic electron beam (REB) generated by relativistic laser-plasma interactions to a small and dense fuel core along magnetic field line. Our studies clarified the followings; (i) the REB guiding resulted in doubling the drag heating efficiency up to 8% (2), (ii) ultra-high-energy-density region having 2.0 keV of temperature and 2.2 Peta-Pascal of the pressure was created with 12% of energy coupling efficiency through the diffusive heating process, namely by heatwave propagating in the core with 10$^{8-9}$ cm/s of the speed (3), (iii) 8% of the current drag coupling efficiency is scalable to >15% of the efficiency in an ignition-scale high area-density core (> 0.5 g/cm2) and heating laser intensity (1 x 10$^{20}$ W/cm$^2$) based on a model (4) that are fairly consistent with the experimental observations. This efficiency is >3 times higher than that of the currently pursed central ignition scheme.
The fast-isochoric heating may produce the ignition spark with avoiding ignition quench caused by the fluid mix between the spark and a cold fuel layer. This mix is a crucial problem of the adiabatic compression heating in the central ignition scheme. In the fast-isochoric heating scenario, a relativistic intensity laser, namely heating laser, generates the high-intensity REB. The REB carrying a significant fraction of the heating laser energy travels in an over-dense plasma from its generation zone to the pre-compressed fuel core. Part of the REB's kinetic energy is deposited in the pre-compressed core, and then the heated plasma core becomes the ignition spark. There are two major mechanisms in the fast-isochoric heating process as illustrated in Fig. 1. One is the drag heating mechanism, where energy is transferred from REB to the fuel core by electron-electron binary collisions. The other is the diffusive heating mechanism, where energy is transferred by thermal conduction to a cold plasma from a high-temperature plasma that is heated by the resistive return current induced by the forward REB current. Kilo-tesla magnetic field is required to guide the forward REB current and also the return current within a diameter of the pre-compressed core. A laser-driven capacitor-coil target was used to generate a kilo-tesla magnetic field (5,6).
A solid ball was compressed by six beams of fuel compression laser in this experiment. The Cu and Ti atoms contained in the ball were tracers for the measurement of laser-to-core coupling and electron temperature of the core. The solid ball was attached to a gold cone. The heating laser was focused on the tip of the cone to produce the REB. The center of the coil was located at 230 $\mu$m from the center of the solid ball. The drag mechanism heats uniformly the entire pre-compressed plasma because of long range (0.5 g/cm$^2$ for 1 MeV REB) of the REB. The diffusive mechanism heats a smaller part of the pre-compressed plasma to higher temperature than those heated by the drag mechanism due to shorter range of thermal electrons than that of REB. The visualization of He_alpha X rays emitted from the heated region enables us to evaluate the efficiency of the diffusive heating process in addition to the drag heating efficiency that was evaluated from the absolute yields of electron collision-induced K$_\alpha$ X rays. The visualization of He$_\alpha$ X rays emitted from the heated region enables us to evaluate the efficiency of the diffusive heating process in addition to the drag heating efficiency that was evaluated from the absolute yields of electron collision-induced K$_\alpha$ X rays. Hydrodynamic motion of the heated plasma is negligible during the drag heating time-scale, on the other hand, the diffusive heating is slower than the drag heating. Time and spatial scale of the diffusive heating process are several tens ps and µm respectively, therefore, hydrodynamic motion can affect the evaluation of the diffusive heating efficiency. < 15 $\mu$m and < 10 ps of spatial-temporal resolutions are required for evaluating the diffusive heating efficiency. It was difficult to utilize a fast and high-spatial resolution x-ray imaging technique under the harsh radiation environment produced by high-intensity laser-plasma interactions because the intense radiation causes a noisy background in the images. We visualized the heating area correlated with a plasma density profile by using quasi-monochromatic x-ray imagers and/or a thin-photo-cathode x-ray image detector that is insensitive to hard X-rays (7).
A number of K$_\alpha$ photons emitted from the tracers correlates with energy deposited in the core via the drag heating mechanism. The internal energy of a heated plasma was obtained by combining two-dimensional (2D) electron temperature distributions obtained from 2D He$_\alpha$ x-ray images and 2D density ones that were separately measured with an X-ray backlight technique. X rays from He-like ions (He$_\alpha$) are emitted from a high-temperature region, and its yield depends on electron temperature of the region. The diffusive heating efficiency was calculated as a ratio between evaluated internal energy and heating laser energy. The measurements revealed moderate heating of the entire core by the drag mechanism with 8% of energy coupling (2) and 12% of the diffusive heating efficiencies for 0.1 g/cm$^3$ of areal density in our experiment, respectively. As shown in Fig. 2, the drag and diffusive heating processes were visualized with Fresnel phase zone plates (FPZPs), whose spatial resolution and gating time are 5 µm and 10 ps, respectively. The drag heating is the dominant process in the dense region because this is caused by electron-electron binary collisions, and the diffusive heating localized in the less compressed region near the cone because the heatwave propagation velocity is inversely proportional to the plasma density.
(1) T. Johzaki et al., Plasma Phys. Control. Fusion 59, 014045 (2017).
(2) S. Sakata et al., Nat. Commun. 9, 3937 (2018).
(3) K. Matsuo et al., , Phys. Rev. Lett. (accepted).
(4) S. Sakata et al., Plasma Fusion Res. 14, 3404138 (2019).
(5) S. Fujioka et al., Sci. Rep. 3, 1170 (2013).
(6) S. Fujioka et al., Plasma Phys. Control. Fusion 51, 124032 (2009).
(7) H. Shiraga et al., Rev. Sci. Instrum. 79, 1 (2008).Speaker: Shinsuke Fujioka (Institute of Laser Engineering, Osaka University) -
Here we report theoretical and numerical studies for efficient plasma heating by high intensity lasers with a lateral confinement of the laser-accelerated fast electrons in the laser spot area. Recent experiments using kilojoule (kJ) petawatt lasers show efficient particle accelerations and plasma heating, indicating that the lateral loss of fast electrons is considerably small. We here found that (i) the lateral motion of fast electrons in a relativistic laser-foil interaction has a 'random walk' nature due to the scattering by fluctuating fields at the plasma surface, so that the motion can be described as a diffusion. (ii) The slow, diffusive escape from the spot area corresponds to an effective confinement of fast electrons, by which the bulk electrons can be heated to keV temperatures via collisional processes. Particle-in-cell (PIC) simulations demonstrate that the diffusion model is applicable for the kJ lasers having large spot sizes. The present theory can provide a guideline for efficient isochoric heating for high energy density (HED) science including fusion studies.
Petawatt class high power lasers with relativistic intensities $> 10^{18}\,{\rm W/cm^2}$ can heat dense matters isochorically and create HED plasmas with giga-bar pressures [ 1] . Such a heating of bulk electrons is driven by the fast electrons accelerated by the laser at the target surface. Hence, for the efficient heating, confinement of the fast electrons is crucial. Many efforts for the confinement have been made, e.g., applying external magnetic fields, and using mass-limited or layered target with different resistivity. The confinement is also important for intense x-/gamma-ray generation and ion acceleration.
Recently, kJ class relativistic intensity lasers with picoseconds (ps) pulse durations are available such as LFEX, NIF-ARC, and LMJ-PETAL. In contrast to the ultraintense femtosecond (fs) pulse lasers with tight focusing down to the wavelength $1\,{\rm \mu m}$, the kJ lasers have large spot sizes exceeding $50\,{\rm \mu m}$. Experiments using the kJ lasers and thin foil targets demonstrated efficient ion accelerations [ 2] owing to a continuous heating of the fast electrons during the over-ps laser irradiation [ 3]. However, there has been a question why the fast electrons, which is expected to escape from the spot area almost at the speed of light, can benefit from the continuous heating.In this study, we found the mechanism of a lateral confinement of the fast electrons in the spot area, which can enhance the bulk electron heating. For relativistic laser and thin foil interactions, the majority of fast electrons are trapped by the sheath electric potential and recirculate around the foil as illustrated by blue lines in Fig.$\,$1. We here find that, for a large spot laser (a), the lateral spread of fast electrons can be described as a diffusion process, or a 'random walk'. Namely, in (a), the fast electrons recirculate in the foil experience multiple scatterings by fluctuating fields excited in the spot region at both front and rear surfaces which are modulated by instabilities, e.g., the Weibel instability accompanying magnetic field filaments. This is different from the ballistic escape in the interaction with a small spot laser, Fig.$\,$1$\,$(b), where fast electrons experience the scattering at most only a few times before they escape from the spot area.
The random walk behavior with the multiple scatterings by the fluctuating fields is confirmed in the corresponding two-dimensional (2D) PIC simulation with laser spot radius $w = 35\,{\rm \mu m}$ and foil thickness $L = 5\,{\rm \mu m}$ as Fig.$\,$2$\,$(a) and (b).
The diffusion velocity $u_{\rm dif}$ can be derived from the diffusion equation $\partial P/\partial t = D\partial^2 P/\partial y^2$ where $P$ is the probability for the electron initially located at $y = 0$ to be at position $y$ at time $t$, and $D$ is the diffusion coefficient. At $y = w$, we have
$\begin{equation} u_{\rm dif}=\frac{D}{w}=\frac{L}{w}\frac{\tan\theta}{2}u_{\rm bal}, \end{equation}$where $\theta = 1$\,rad is the average scattering angle for the isotropic scattering by the surface magnetic field, and $u_{\rm bal} = c \sin\theta$ is the lateral velocity for the ballistic motion. For the large spot interaction as in Fig.$\,$1$\,$(b), $u_{\rm dif} \ll u_{\rm bal}$ is derived from Eq.$\,$(1).
Owing to the slow lateral escape due to the random walk, the fast electrons accumulate in the spot area. The achievable density of fast electrons at the steady state $n_h$ is determined by the balance between the number of electrons escape from the spot and that injected from the laser-plasma interface as
$\begin{equation} 0=D\frac{\partial^2 n_h}{\partial y^2}+S, \end{equation}$where the source term $S = \gamma n_cc/L$ indicates the inflow of fast electrons, and $\gamma n_c$ is the relativistic critical density. By solving Eq.$\,$(2), we see $n_h/(\gamma n_c)$ is greater than 1 and proportional to $(w/L)^2$, which we confirmed by a series of PIC simulations. Figure 3 shows the density evolution for the fast electrons with energies $>100$\,keV. In the large spot case (b), $\sim 20$ times higher density than the small spot case (a) is achieved, which clearly shows the effect of confinement, leading to the energy density of 9\,Gbar as in Fig.$\,$2$\,$(c). The peak density, $\simeq 12 n_c$ in Fig.$\,$3 (b), can be explained by the diffusion model with considering the extension of $L$ during the interaction.
For a solid Al target, the stopping power for an electron with energy 100 keV is $\sim 1$ keV/${\rm \mu m}$. Hence, when the electron travels a distance of $100\,{\rm \mu m}$ in the circulation, which takes several ps, without escaping from the spot area, the electron is capable to transfer energy to the bulk electrons via collisions, i.e., the drag heating. The lateral confinement keeps fast electrons inside the spot area over ps. When all the accumulated fast electron energy is distributed equally to the bulk electrons, the temperature becomes several keV. Collisional PIC simulations show such an energy cascade process in picoseconds, which is enabled by the lateral confinement discussed above.
[ 1] K. Matsuo, N. Higashi, N. Iwata et al., Phys. Rev. Lett. 124, 035001 (2020).
[ 2] A. Yogo et al., Sci. Rep. 7, 42451 (2017); D. Mariscal et al., Phys. Plasmas 26, 043110 (2019).
[ 3] N. Iwata, Y. Sentoku et al., Plasma Phys. Control. Fusion 62, 014011 (2020); J. Kim, A. J. Kemp, S. C. Wilks et al., Phys. Plasmas 25, 083109 (2018).
Speaker: Natsumi Iwata (Institute of Laser Engineering, Osaka University) -
Erosion of plasma-facing components (PFC) due to sputtering by impinging ions and neutrals is one of the key challenges on the road to fusion power. Erosion will be a contributor to the overall PFC lifetime estimation, and to impurity production that can potentially lead to radiative collapse. Moreover, PFC erosion is directly linked to the key issues of fuel retention by co-deposition and dust production. For predictive modelling of these effects in ITER and other fusion devices, the impurity transport and plasma-surface interaction (PSI) Monte-Carlo code ERO2.0 [1] has been developed. ERO2.0 is a massively parallel 3D code, capable of performing global erosion and deposition simulations of all relevant PFCs installed in the full-size reactor vessel. In this contribution, we provide an overview of recent and ongoing ERO2.0 applications to JET and ITER, highlighting the importance of 3D aspects of the PSI modelling.
ERO2.0 has been validated by applying it to model beryllium (Be) erosion and transport in the main chamber of the JET ITER-like wall (ILW) for different operating conditions covering limiter and divertor configurations. Parameter scans included a fuelling scan in limiter configuration and a power scan in divertor configuration resulting in ohmic, L- and H-mode plasmas shown in Figure 1(a). The corresponding 2D plasma backgrounds used as input to ERO2.0 were obtained mainly from EDGE2D-EIRENE simulations [2] extrapolated to the 3D contour of the JET wall.
We observe good agreement with experimental measurements from passive spectroscopy, including spectrally filtered wide-angle cameras with 2D resolution (see Figure 1(b)) as well as integrated lines-of-sight of spectrometer chords. We show that taking into account the 3D shape of the Be limiter tiles and the resulting complex magnetic shadowing patterns is crucial to reproduce the experimental measurements. Moreover, ERO2.0 qualitatively mimics the predominant Be transport into and Be deposition in the inner divertor, which was observed by post-mortem analysis of JET PFCs [3].
Following this code validation on JET, ERO2.0 has been used to predict the erosion of the Be first wall panels (FWP) in the ITER main chamber. Eight cases were studied, representing different plasma conditions or modelling assumptions, as summarised in Figure 2. The study covers ITER plasmas from the Pre-Fusion Power Operation (PFPO) and fusion power operation (FPO) phases and includes variations of density, scrape-off layer (SOL) flow velocity, heating power, magnetic configuration (separation between primary and secondary separatrices) and plasma species (H, He, D-T). The reference case #1 represents the baseline H-mode plasma in DT with Q=10 and 500 MW of fusion power.
The corresponding input plasma backgrounds, obtained from OEDGE simulations, were provided by the ITER organisation (IO). The OEDGE simulations are based on SOLPS solutions, and additionally use the ITER Heat and Nuclear Load Specifications (HNLS) [4] combined with an onion-skin model (OSM) to extend the computational grid to the wall [5].
In general, we observed that the regions of highest Be gross erosion are found on FWP 4-5 (inner wall), 8-9 (top) and 18 (outer wall near the divertor), as shown in Figure 3. An important result is that in all of the investigated FPO cases (#1-5), the major part of the eroded Be is redeposited in the main chamber. In case #1, this fraction amounts 90%, with the remaining 10% deposited in the divertor. This is in strong contrast to the comparable JET H-mode simulations, where more than 60% of the Be is deposited in the divertor. Comparably large fractions of up to 53% are however observed at ITER for the Pre-Fusion Power Operation (PFPO) cases #6-8, which have plasma parameters closer to the ones at JET. This suggests that Be long-range transport to the divertor is increasingly suppressed at high heating power and in particular high density plasma conditions in the far-SOL.
In case #2, a high imposed background SOL flow (Mach number 0.5) led in the ERO2.0 simulations to an entrainment of Be impurities and more than doubled their transport to the divertor, compared to case #1. In case #3, a lower density was assumed in the far-SOL plasma parameters, leading to a ~3x lower Be gross erosion and a ~2.5x higher Be deposition fraction in the divertor. Cases #4 and #5 have a different magnetic configuration that uses the minimum allowed distance between the primary and secondary separatrices, Δrsep=4 cm, in contrast to Δrsep=9 cm for the other cases. This leads to an elongated plasma shape and strong plasma-wall interaction around the machine apex, which for case #5 resulted in a 3x higher Be erosion compared to case #1. Cases #6-8 represent H and He plasmas from the ITER PFPO phase. These cases show a significantly lower (by one order of magnitude) Be gross erosion rate, compared to their FPO counterparts. Furthermore, the Be transport to the divertor is significantly increased.
One of the main modelling uncertainties in the ERO2.0 modelling are the plasma parameters near the surface, which determine the impact energy and angles of the plasma particles impinging on the surface, and thereby the sputtering yields. In particular, the HNLS assumes very high plasma temperatures (Te=10 eV and Ti=20 eV in the reference case #1) in the far-SOL of the entire main chamber area, leading to highly conservative predictions regarding the Be source strength. An exponential decay of the plasma parameters Te, ne, Ti would likely lead to a substantial reduction of the Be erosion. Quantification of this reduction is aim of upcoming ERO2.0 studies.
Moreover, in the present simulations, a perfectly smooth plasma-facing surface was assumed. However, micro-scale roughness is usually present in plasma-exposed surfaces, and can reduce the net erosion by re-deposition in the microstructures [6]. In addition, the smooth surface assumption leads to more shallow ion impact angles (and thereby increased erosion) when compared to a rough surface. Thus, the presented erosion and deposition rates represent upper limits for the expected values in ITER which will use Be PFCs with a roughness on the micrometer scale.
References
[1] J. Romazanov et al., Nucl. Mater. Energy 18 (2019) 331–338
[2] S. Wiesen et al., EDGE2D/EIRENE Code Interface Report, JET ITC-Report, 2006.
[3] A. Widdowson, et al., Nucl. Fusion 57 (8) (2017) 086045
[4] R. A. Pitts et al., J. Nucl. Mater. 2011, 415(1 Suppl), S957.
[5] S.W. Lisgo et al., J. Nucl. Mater. 438 (2013) S580–S584
[6] A. Eksaeva et al., Nucl. Mater. Energy 19 (2019) 13–18Speaker: Dr Juri Romazanov (Forschungszentrum Jülich) -
Anomalous plasma transport in the boundary region of a tokamak plasma is normally associated with density structures. These density structures are commonly termed as plasma blobs. Recently, a theory for a universal mechanism of plasma blob formation has been put forward that is based on a breaking process of a radially elongated streamer [1] in the presence of poloidal and radial velocity shear. The theory has been supported by numerical simulation results, but lacks experimental validation. In this work we report the first ever experimental validation of the universal mechanism by using NSTX and Aditya data to examine the dynamics of blob formation in that machines and comparing that to our theoretical and simulation results.
Figure-1(a)-(b) show snapshots of the formation of a plasma blob in one of the Ohmic L-mode plasma shots (shot #141745) in NSTX tokamak using images obtained from the gas-puff imaging (GPI) diagnostic [2]. The images have been superposed on quiver plots of the radial and poloidal velocities of the plasma obtained from velocimetry measurements [3,4]. It is to be noted that the visible images acquired by GPI are taken here as a proxy of the density contours as the intensity is mainly proportional to the density, and its temperature dependence has been neglected. The direction of the arrows indicates that the flow is sheared near the streamer structure in a manner such that the radially in and outward velocities can tear apart the plasma streamer structure. This is in accord with the theoretical picture of the universal mechanism of the plasma blob formation proposed in Ref[1]. Furthermore the mechanism requires that the following condition be satisfied,
${\frac{v_x^{'}}{\gamma}+\frac{v_{y}^{'}dx}{\gamma dy}\sim1,}$
where $v_{x}'$ and $v_{y}'$ indicate the radial gradients of the radial and poloidal velocities and $\gamma$ indicates the linear growth rate of the interchange instability. We will also check the validity of this condition from the experimental data. In Fig-1(a) one sees the initial plasma blob formation phase where a poloidal compression of the streamer structure is taking place due to the presence of the positive and negative poloidal velocities. In this phase, two opposite directions of the radial velocities are seen (inner part of the streamer moves towards lower-R side and the outer part moves towards the local blob center) and at $x\sim1465$ mm, $y\sim273$ mm the plasma velocity is zero. This corresponds to the radial stretching as indicated in Ref.[1]. Similar behavior has also been seen in experimental shots of Aditya tokamak [5] and is analogous to our past numerical simulations where the velocity is seen to vanish at the point where the magnitude of the electrostatic potential of the mode is maximum. Fig.-1(b) shows a fully developed plasma blob phase. Here, the poloidal compression is not seen but the radial plasma motion is similar to Fig-1(a). From our analysis of the experimental data we find that the magnitudes of $v_{x}'$ and $v_{y}'$ over the plasma blob are $v_{x}'\sim1\times10^{4}$/s and $v_{y}'\sim3\times10^{5}$/s. These values along with $\gamma\sim c_{s}/\sqrt{RL_{n}}\sim2\times10^{5}$/s and $dx\sim dy$ are found to nearly satisfy the universal condition given by above equation. An extended statistical validation of the universal condition carried out over a reasonably large blob data set is presently in progress and will be reported.
References
- N Bisai, Santanu Banerjee, and A Sen. Phys. Plasmas 26 :020701, 2019.
- S. J. Zweben, W. M. Davis, R. E. Bell, B. P. LeBlanc, S. M. Kaye, R. J. Maqueda, T. Munsat, J. R. Myra, Y. Sechrest, D. P. Stotler, and the NSTX Team. Nucl. Fusion, 55 :093035, 2015.
- Santanu Banerjee, H. Zushi, N. Nishino, K. Mishra, T. Onchi, A. Kuzmin, Y. Nagashima, K. Hanada, K. Nakamura, H. Idei, M. Hasegawa, and A. Fujisawa. Rev. Sci. Instrum., 86 :033505, 2015.
- A. Diallo, S. Banerjee, S.J. Zweben, and T. Stoltzfus-Dueck. Nucl. Fusion, 57 :066050, 2017.
- B. K. Joseph, R. Jha, P. K. Kaw, S. K. Mattoo, C. V. S Rao, Y. C. Saxena, and the Aditya Team. Phys.Plasmas, 4 :4292, 1997.
Speaker: Nirmal Kumar Bisai (Institute for Plasma Research, Bhat, Gandhinagar-382428, India) -
In magnetic confinement fusion devices, the plasma density is largely self-sustaining through internal recycling processes. While refueling the plasma, the recycling neutrals can also considerably affect the energy transport in both the plasma confinement and exhaust regions. A quantitative understanding of the neutral-related effects first requires a precise knowledge of the origin and source strength of the recycling neutrals, which is difficult to obtain in stellarators due to the inherent three-dimensionality. The W7-X stellarator is equipped with an island divertor with ten identical divertor modules. Most of them are monitored with $H\alpha$-cameras and neutral pressure gauges. In addition, two Langmuir probe arrays are mounted on one lower and upper divertor target, respectively. They are the most relevant diagnostics for the recycling process. Nevertheless, these diagnostics alone are unable to quantify the recycling neutrals either because of their local access to the plasma or because the interpretation of their results relies on physical quantities that are not readily available experimentally. One way to overcome these difficulties is to combine them with the EMC3-Eirene code in such a way that the code results are compared with the local measurements. Through experiment-modeling comparison, this paper attempts to make a first quantitative assessment of the source and transport of the recycling neutrals in W7-X. To this end, a feasibility and case study has been started. Most important here is the determination of the internal recycling flux.
The target probes can measure the local ion saturation current densities $J_{sat}$, but cannot provide sufficient information on the global recycling in the machine due to their limited number and location. An example is shown in figure 1, which compares the particle fluxes calculated by the 3D code and measured by the Langmuir probes for #20180814.25 at the time point of maximum $J_{sat}$. The standard iota = 5/5 divertor configuration is chosen for the comparison. The island control coils were not used, which would otherwise shift the deposition profile towards the divertor gap, i.e. further away from the probes. With a heating power of 5 MW and a diffusivity of 0.5 $m^2/s$, the separatrix density is varied in the simulations widely enough to fully cover the experimentally relevant range. The intrinsic carbon is assumed as the only impurity species and its yield is 4% of the hydrogen recycling flux. Although their positions are far from optimal, the target probe arrays, in combination with the IR- und $H\alpha$-cameras, are still useful for estimating the perpendicular transport coefficients. The up/down asymmetry in the measured ion saturation currents cannot yet be addressed by the 3D code due to the lack of drifts and to the ideal field and divertor geometry assumed in the simulation. In reality, the ten “identical” divertor units usually receive different heat and particle fluxes, as detected by the IR- and $H\alpha$-cameras, even under the action of the error-field compensation coils (trim coils). To facilitate a code-probe comparison under this circumstance, a modular averaging of the probe data is necessary, where the Jsat profiles should be aligned with the actual strike-lines estimated by the IR-camera and weighted according the relative $H\alpha$-intensities of all monitored modules.
The neutral pressure is measured with pressure gauges installed in the divertor chamber and in the main chamber as well. In general, they are remote from the plasma, especially from the recycling zones, so that the recycling neutrals have undergone several atomic reactions before reaching the gauge positions. Thus, the measured neutral pressure is a combined result of neutral source and transport, which could be a complex function of the background plasma parameters in the island. Figure 2 shows the correlation of the divertor-chamber-averaged neutral pressure $p_{0,div}$ with the total recycling flux, resulting from the numerical density scan described above. The recycling flux first increases with $p_{0,div}$, and then decreases when the plasma detaches from the targets driven by carbon radiation. During the detachment, $p_{0,div}$ remains at a high level, while the recycling flux sharply drops due to enhanced radiation. In fact, this desirable compatibility of the W7-X divertor between particle and power exhaust has already been identified and well analyzed in previous predictive studies$^1$, which was also confirmed in recent divertor experiments$^2$. Unfortunately, the decoupling of $p_{0,div}$ from the recycling flux under detached conditions means that the recycling flux, which is of particular interest for the topic addressed in this paper, cannot be determined by the pressure gauges.
In W7-X the recycling neutrals are well monitored with $H\alpha$-cameras. Nevertheless, evaluation of the line-integrated $H\alpha$-signals requires the local S/XB value (the number of ionization events per photon – a quantity that depends on the electron temperature and the densities of elections, ions, atoms, molecules and molecular ions), which is difficult to access experimentally. Numerical simulations are indispensable. For this purpose, a synthetic $H\alpha$-diagnostic has been implemented in the EMC3-Eirene code. Figure 3 shows the preliminary code results of the volume-averaged S/XB ratio as a function of the carbon radiation fraction from the density scan mentioned above. The strong dependence of the S/XB ratio on the divertor plasma state demonstrates the necessity of 3D simulations for decomposing the $H\alpha$-signals measured at W7-X.References
$^1$Y. Feng et al, Nucl. Fusion 56 (2016)
$^2$O. Schmitz et al, submitted to Nucl. FusionSpeaker: Dr Yuhe Feng (Max-Planck-Institut für Plasmaphysik, 17491 Greifswald, Germany) -
Introduction – The H-mode confinement regime will be the main operational scenario on ITER and also the current foreseen scenario for fusion reactors. A continuous effort towards better predictive capabilities of H-mode confinement is being pursued on both experimental and theoretical fronts. The H-mode is characterised by the formation of a pedestal near the plasma edge and as the fusion power scales as $p_{ped}^2$, it is advantageous to maintain a high pedestal pressure. This goal is challenged by the need to mitigate H-mode characteristic Type-I ELMs either by creating a highly radiating divertor using impurity injection [1] or via ELM-free or small ELMs regimes [2]. With a pedestal database, the H-mode physics studies performed on TCV are reviewed with emphasis on the comparison between the historically ECRH dominated scenario and the NBH H-mode regime explored more recently.
Pedestal database – The TCV pedestal database is one of several databases promoted by EUROfusion to stimulate the multi-machine comparisons, (JET, AUG, MAST-U and TCV), with common parameter definitions [3] and a common platform (IMAS: ITER integrated modelling and analysis suite). The pedestal structure is determined from the pre-ELM temperature and density profiles (75-99% of the ELM cycle) using TCV’s Thomson Scattering [4]. Pedestal parameters are obtained using a mtanh fitting function [5]. To reduce uncertainties in the equilibrium reconstructions, temperature and density profiles were systematically shifted such that $T_{e,sep}=50$ eV, estimated by using the two point model for the power balance at the separatrix. To enhance the overall pedestal data quality, entries were selected according to following rules: steady state intervals over at least 0.4s (~10$\tau_E$) and a reduced $R^2$ for the fit in the region 0.8<$\psi_N$<1.05 larger than 0.75. A new plasma equilibrium is computed using the CHEASE code from the fitted pedestal, accounting for the bootstrap current and, only then, is the pedestal stability analysed [6]. The TCV pedestal database currently contains $\sim$350 entries for about 100 implemented parameters. For TCV’s heating methods, (Ohmic, ECRH and NBH), the normalised average lost energy as a function of the ELM frequency is shown Fig.1-left with the pedestal temperature plotted against the pedestal density, Fig.1-right. The plasma shaping capabilities of TCV have been exploited in the range of parameters: 1.3<$\kappa$<1.8, 0.2<$\delta_b$<0.8 and -0.3<$\delta_u$<0.8. For similar H-mode plasmas except the top triangularity ($\delta_u\sim$-0.2 vs $\delta_u\sim$0.2), the pedestal width is increased by 25% while its top pressure is reduced by a factor of 2 for the negative triangularity cases, in line with EPED predictions [6].
ECRH ELMy H-modes – H-mode plasmas with Type-III ELMs are achieved in TCV with Ohmic heating only for $q_{95}$<3 (Fig.1 blue points). The central plasma density is too high for 2nd harmonic ECR heating, but central heating at third harmonic (X3) is possible. As the injected X3 power is increased (0<$P_{X3}$<0.5 MW), the type-III ELM frequency decreases to $f_{ELM}\sim$ 50 Hz. Additional X3 power doesn’t change the ELM frequency but increases significantly the normalised lost energy ($\Delta W/W\sim$17%). This is explained by some coupling of the ELM instability with the core MHD (m/n=1/1) that propagates the ELM crash to the plasma core region. Finally, for $P_{X3}$>0.8 MW, the ELM frequency increases, signifying a Type-I ELM regime, with a decrease in the normalised ELM losses [7-8]. Typical pedestal values for an ELMy H-mode heated with 1 MW of ECRH are $T_{e,ped}\sim$0.8 keV, $n_{e,ped }\sim3\times10^{19}$ m$^{-3}$, $n_{e,sep}\sim0.2n_{e,ped}$ and $T_e(0)/T_i(0)\sim$6. Such hot plasmas at low densities for $\psi_N$>0.8, including the pedestal, is still accessible to X2 heating and it was demonstrated that the ELM frequency can be controlled by ECRH modulation [9]. Interestingly, a steady state ELM free regime has been reached with 1.2 MW of X3 power with unfavourable $\nabla B$ configurations [10]. In today’s tokamaks, pedestal collisionalities relevant for ITER ($\nu_{\star,ped}\sim$ 0.1) might be achievable with an ECRH H-mode operational regime but conditions for partial detachment, $n_{e,sep}/n_G\sim$ 0.4 at the separatrix, are impossible.
NBH ELMy H-modes – The H-mode operational space extension towards high pedestal and larger separatrix densities has become possible with TCV’s neutral beam injector (1MW, 30keV), in operation since 2015. Moreover, ELMy H-mode can be achieved at lower plasma current ($q_{95}$>3) allowing the development of an ITER baseline scenario on TCV ($q_{95}\sim$3-3.6, $\kappa$=1.7, $\delta$=0.4) [11]. A scenario at $q_{95}\sim$4.5, $\delta$=1.5 is well established with accompanying Type-I ELMs ($f_{ELM}\sim$ 100 Hz, $\Delta W/W\sim$ 10%) and typical pedestal parameters $T_{e,ped}\sim$ 0.2 keV and $n_{e,ped}\sim4\times 10^{19}$ m$^{-3}$. The effects of $D_2$ fuelling and $N_2$ seeding on the pedestal stability and plasma confinement were investigated. Both induces an outward shift of the pedestal density relative to the pedestal temperature with a corresponding outward shift of the pedestal pressure that, in turn, reduces the peeling-ballooning stability, degrades the pedestal confinement and reduces the pedestal width [12], in line with AUG and JET results [13]. A small ELM regime with high confinement was achieved if, and only if, the separatrix plasma density was large enough ($n_{e,sep}/n_G\sim$0.3) and the magnetic configuration was close to a double null ($\delta$>0.4) [14]. The extension of the regime to $q_{95}$<4 was recently achieved. A second neutral beam (1MW, 50keV), planned by the end of 2020, will open new possibilities, not only for H-mode physics at large $\beta_N$, but also for fast particle physics.
[1] A. Kallenbach et al., “Partial detachment of high power discharges in ASDEX Upgrade” Nucl. Fusion 55 053026 (2015)
[2] J. Stober, et al., “Type II ELMy H-modes on ASDEX Upgrade with good confinement at high density” Nucl. Fusion 41 1123–34 (2001)
[3] https://users.euro-fusion.org/iterphysicswiki/index.php/Database
[4] P. Blanchard et al., “Thomson scattering measurements in the divertor region of the TCV Tokamak plasmas” JINST 14 C10038 (2019)
[5] R. Groebner et al., “Progress in quantifying the edge physics of the H mode regime in DIII-D”, Nucl. Fusion 41 1789 (2001)
[6] A. Merle, et al., “Pedestal properties of H-modes with negative triangularity using the EPED-CH model”, Plasma Phys. Control. Fusion, 59, 10, 104001, 2017.
[7] A. Pitzschke, “Pedestal Characteristics and MHD Stability of H-Mode Plasmas in TCV”, PhD thesis, EPFL, no 4917 (2011)
[8] A. Pitzschke et al., “Electron temperature and density profile evolution during the ELM cycle in ohmic and EC-heated H-mode plasmas in TCV”, Plasma Phys. Control. Fusion, 54, 015007 (2012)
[9] J.X. Rossel et al., “Edge-localized mode control by electron cyclotron waves in a tokamak plasma” Nucl. Fusion 52 032004 (2012)
[10] L. Porte et al., “Plasma dynamics with second and third-harmonic ECRH and access to quasi-stationary ELM-free H-mode on TCV” Nucl. Fusion, 47, 8, 952–960 (2007)
[11] O. Sauter et al., “ITER baseline scenario investigations on TCV and comparison with AUG”, this conference
[12] U. A. Sheikh et al., “Pedestal structure and energy confinement studies on TCV” Plasma Phys. Control. Fusion, 61, 1, 014002 (2019)
[13] L. Frassinetti et al, “Role of the pedestal position on the pedestal performance in AUG, JET-ILW and TCV and implications for ITER” Nucl. Fusion, 59, 7 076038 (2019)
[14] B. Labit et al., “Dependence on plasma shape and plasma fuelling for small edge-localized mode regimes in TCV and ASDEX Upgrade” Nucl. Fusion, 59, 8 086020 (2019)Speaker: BENOIT LABIT (Swiss Plasma Center (SPC) EPFL SWITZERLAND) -
The high coupling efficiency ($\eta_c$) can be achieved by using a shell target with holes in the direct fast ignition. Here, we have made it clear that a diffusive heating is not negligible compared with a collisional and a resistive heating. In order to obtain high $\eta_c$, it is important to maintain the low effective hot electron temperature ($T_{eff}$). In the shell target with holes, $T_{eff}$ can be suppressed by the prevention of the blow-off plasma, which comes from target irradiation by the compression laser (Gekko XII/$2\omega$:GXII) on the heating laser (LFEX/$1\omega$) path. Each heating contribution can be estimated from the electron/ion measurements. The maximum energy of 7±2 J is deposited to the core as the electron collisional heating in the heating LFEX energy of 343 J. 12±3 J can be estimated to be deposited to the target as the ion collisional heating. The resistive+diffusive heating is 22±10 J. The simulation also suggests similar tendency about the heating mechanism.
A fast ignition is performed by auxiliary heating to the imploded core. It is found that the heating laser does not provide the sufficient $\eta_c$ because the ablated plasma prevents the penetration of the heating laser in the direct fast ignition. Kitagawa \textit{et} $al.^{1)}$ has succeeded to realize low $T_{eff}$ even at the direct fast ignition, by using the spherical shell with two holes along to the LFEX axis. Here, the ablation plasma from the compression laser does not invade to the LFEX laser path. Since an areal density ($\rho R$) in our experiment is much smaller than the stopping by range of $T_{eff}$, the deposit energy is roughly the electron energy with the range equivalent to $\rho R$. When the number of electrons increases if $T_{eff}$ becomes low, total deposit energy as collisional also increases. The resistive heating also increases because the electron current increases. The diffusive heating comes through the diffusion from the plasma directly heated by the electric field of the heating laser. This means the source of the diffusive heating is deduced from super low energy component of the hot electron. Therefore, the diffusive heating also increases if $T_{eff}$ is low.
In the experiment, the electron energy spectrometers (ESM)$^{2)}$ to observe electron/ion spectra, Mandala neutron spectrometer to monitor the neutron energy spectrum, CR39 knock on proton measurement system to observe the target $\rho R$ and X-ray pinhole camera (XPHC) to monitor the core radius and the total absorbed energy, are used as shown in Fig. 1. X-ray streak camera (XSC) monitors the LFEX injection timing. A deuteride polystyrene shell of $500 \mu m^\phi$ and $7 \mu m^t$ with two holes of 100 $\mu m^\phi$ along the axis of LFEX is used as the target. Six beams of GXII with 1 ns Gaussian pulse, which are located at transverse direction to LFEX axis, compress the shell and creates the imploded core. LFEX auxiliary heats the core at the stagnation phase timing, which is
monitored by XSC. $T_{eff}$ is estimated at 0.86 MeV from the slope of ESM signal in Fig. 2(a). The amount of generated electrons can be estimated as follows; At the laser energy level used in this experiment, the reflection is assumed to be 25 \% from the previous experiment$^{3)}$. Therefore, it is estimated that 75 \% of the laser energy (343 J) is converted to hot electrons (here, ion number itself is assumed to be small). The maximum number of generated electrons can be estimated if this energy is divided by $T_{eff}$. However, hot electrons have a divergence angle. If the spatial distribution of the hot electrons is assumed to be uniform, the maximum number of electrons irradiated to the core is roughly half. On the other hand, the stopping energy at $\rho R$ = 0.0122 $g/cm^2$ is 0.077 MeV for electron. This means that each electron loses the energy of 0.077 MeV independent on the electron energy. Therefore, the maximum of 7±2 J is deposited to1 the core as the electron collision.
Heating by ions cannot be ignored in direct fast ignition. The deuterium ions accelerated by the ponderomotive force, hit the target deuterium and generate neutrons by the DD reaction (beam neutron). Here, the thermal neutrons are not dominant comparted with so-called beam neutrons. The average ion energy $T_{ion}$ can be estimated to be 0.33 MeV from the ion spectrum measured by ESM as shown in Fig. 2(b) (red line). This DD nuclear reaction cross section can be also obtained from $T_{ion}$. $\rho$ can be obtained from $\rho R$ and the core radius $R$ obtained from XPHC. The SRIM code can determine the ion stopping range. The number of target deuterium required for the reaction can be estimated. The number of neutrons is measured to be $3.7\times10^8$ by Mandala, thus the number of accelerated ions can be calculated. Since ions have a short stopping range, we assume that all relevant ion energy is deposited to the target. Therefore, 12±3 J can be estimated to be deposited to the target as the ion collision. On the other hand, the increase of the internal energy due to additional heating by LFEX can be obtained to be 41±9 J from the comparison between the brightness measured by XSC/XPHC assuming a blackbody radiation and the temperature without LFEX observed by the Mandala neutron spectrum assuming $T_e=T_i$. The high $\eta_c$ can be achieved in this configuration. The value is considerably larger than the sum of electron/ion collisional heating (19±4 J). Considering that resistive heating is comparable to collisional heating, the diffusive heating may be candidate of the discrepancy according to the simulation$^{4)}$.
$T_{eff}$ can be obtained from the slope of the spectrum. $T_{eff}$ keeps low especially at the counter direction against LFEX.
There are two different ion acceleration mechanisms of TNSA(Target Normal Sheath Acceleration) and ponderomotive. We assume that both accelerated ions can be observed when LFEX timing is far from the implosion, although only TNSA ions are observed near the implosion timing.
1)Y. Kitagawa, et al., Phys. Rev. Lett., 114 (2015) 195002.
2)T. Ozaki, et al., Rev. of Sci. Instrum., 83 (2012) 10D920-1.
3)T. Ozaki, et al., J. of Phys.: Conf. Ser., 717 (2016) 012043.
4)N. Higashi, et al., IFSA2019.Speaker: Dr Tetsuo Ozaki (National Institute for Fusion Science) -
Numerical simulations by the integrated divertor code SONIC show that the screening effect on the seeded high-Z impurity in the SOL plasma is improved through the enhancement of plasma flow induced by additional low-Z impurity injection. A single impurity injection of Ar into a steady-state high-beta plasma of JT-60SA results in a high Ar density at the top of SOL plasma, leading to an increase of core Ar density. This issue can be solved with even a small Ne seeding, which reduces Ar density in the SOL and the core plasmas. This is mainly caused by the enhanced friction force due to the higher D$^+$ parallel flow towards the inner divertor, which is originated from the strong Ne radiation around X-point. We show that the line emission of Ne$^{7+}$ has a key role for the generation of higher D$^+$ parallel flow.
SONIC: SOL and divertor transport code with kinetic treatment of multiple impurities
The divertor power handling is one of the critical issues for future magnetic fusion devices, such as JT-60SA and ITER. Seeding of impurities such as Ne and Ar into the plasma is regarded as one of the promising methods to mitigate divertor heat load. Impurities dissipate plasma energy by converting kinetic energy into radiation. The impurity radiation in SOL and divertor plasmas contributes to mitigate the heat flux towards divetor plates. Contrarily, the impurity radiation in a core plasma causes a deterioration of plasma performance especially for high-Z impurities. To establish a method for obtaining high impurity radiation power in SOL and divertor plasmas together with low core impurity radiation power by controlling the impurity transport is indispensable for the future fusion devices.
In order to obtain understanding of impurity transport processes and interactions between plasma and impurity, we have developed the integrated divertor code, SONIC $^1$. One of the unique features of SONIC is to compute the impurity transport processes kinetically, while most of other codes treats them fluidly. SONIC provides more precise impurity transport processes, e.g. ionization and recombination of impurities, a kinetic modelling of Coulomb collisions.
In this study, numerical simulations with SONIC on the JT-60SA steady-state high-beta plasma are discussed for the following two cases; Case A: Ar-only seeding (Ar seeding rate of 0.2 Pa m$^3$/s), and Case B: additional Ne injection (Ne seeding rate of 0.02 Pa m$^3$/s) to the Ar-only seeding plasma obtained in Case A. In both cases, intrinsic C sputtered from the first wall and the divetors is considered. The seeding locations of Ar and Ne (see Fig. 1), and the input power and particle flux from the core boundary are kept the same throughout the two cases. The simulations show that both Cases A and B result in partial detachment with a similar total impurity radiation power of ~13 MW.Mechanism of high-Z impurity screening by adding low-Z impurity seeding
A schematic view of the main results for Cases A and B is shown in Fig. 1. As shown in Fig. 2(a), a high Ar density (defined as the sum of all charge states $n_{\rm{Ar}}$) is seen around the top of the SOL plasma (TOP) in Case A. In contrast, even a small Ne injection at the seeding rate of 0.02 Pa m$^3$/s results in lower $n_{\rm{Ar}}$ in TOP in Case B. The low $n_{\rm{Ar}}$ in TOP in Case B is mainly due to the higher parallel D$^+$ flow velocity $\bf{u}$$_{\parallel,\rm{D^+}}$ towards the inner divertor (ID) region as shown in Fig. 2 (b). The sum of the friction and thermal forces parallel to the field line $F_{\parallel,\rm{fr}}+F_{\parallel,\rm{th}}$ of Ar$^{10+}$ is approximately balanced in the high field side in Case B as shown in Fig.2 (c) due to the enhancement of $F_{\parallel,\rm{fr}}$ by the higher $\bf{u}$$_{\parallel,\rm{D^+}}$ and thus the Ar ions can reach the ID region. On the other hand, in Case A, there exists $F_{\parallel,\rm{fr}}+F_{\parallel,\rm{th}}<0$ region between the inner midplane (IM) and the X-point (XP) due to $F_{\parallel,\rm{th}}$, which is exerted by the parallel temperature gradient in the detached plasma, i.e. the Ar ions moving towards the ID region are decelerated, or even reflected back towards TOP. Therefore, the additional Ne seeding reduces Ar ions at TOP, which is one of biggest sources for the core Ar ions. This result suggests that the radiation of the core plasma by Ar can be reduced by the small injection of Ne.
The higher $\bf{u}$$_{\parallel,\rm{D^+}}$ in Case B is originated from the Ne radiation with the following possible mechanism; First, Ne radiation dissipates the electron energy around XP. Then, the local electron temperature around the XP decreases, and therefore the local electron pressure also decreases. As a result, the electron pressure gradient towards TOP appears in Case B and thus the plasma can have higher $\bf{u}$$_{\parallel,\rm{D^+}}$ towards the ID region. We find that the line emission of Ne$^{7+}$, which is recombined from Ne$^{8+}$, is the major contributor for Ne radiation power around XP and plays a key role in the above process, resulting in the higher $\bf{u}$$_{\parallel,\rm{D^+}}$. The additional calculation without the Ne$^{7+}$ line emission (Case C) is carried out in order to evaluate its contribution to the higher $\bf{u}$$_{\parallel,\rm{D^+}}$. As seen from Fig. 2 (b), the high $\bf{u}$$_{\parallel,\rm{D^+}}$ cannot be seen in Case C and the Ar impurities are trapped at TOP.
This result suggests that the line emission of Ne$^{7+}$ can be used as a knob to control the Ar impurity transport. Both the essential role of Ne$^{7+}$ and the smaller radiation power in the SOL due to low $n_{\rm{Ar}}$ found in Case B are consistent with the bolometric and the spectroscopic observations in Ar+Ne seeding experiment in JT-60U $^{2, 3}$. To our knowledge, this study with SONIC is the first to have shown the improved screening of the seeded high-Z impurity in the SOL plasma with additional low-Z impurity injection by the high SOL D$^+$ flow and resultant strong friction force by numerical simulations. This work provides a solution to avoid the Ar accumulation at the top of the SOL plasma, which is one of the biggest sources into the core Ar ions. This screening effect will be examined in the future JT-60SA experiment.
$^1$ Kawashima H. et al 2006 Plasma Fusion Res. 1 031
$^2$ Asakura N. et al 2009 Nucl. Fusion 49 115010
$^3$ Nakano T. et al 2015 J. Nucl. Mater. 463 555Speaker: Dr Shohei Yamoto (Naka Fusion Institute, National Institutes for Quantum and Radiological Science and Technology) -
For a laser fusion reactor of fast ignition$^{1,2}$, we design an optimum implosion with solid spherical target and inserted conical gold target, where we achieve the maximum areal density $\rho R_{\rm max}=0.46$ g/cm$^2$. According to the hydro-equivalent, the results correspond to the re-quired laser energies for the implosions are 82 kJ for the ignition-scale-target ($\rho R_{\rm max}$=1.1 g/cm$^2$), and 1.3 MJ for the burning-scale-target (($\rho R_{\rm max}$=2.5 g/cm$^2$) respectively, which are significantly improved over our previous design work$^{3}$.
In the fast ignition of laser fusion, a reliable target design is required for an ignition and burning scale experiments. The fast ignition scheme consists of two phases, compression and ignition. In this paper, we discuss the first phase, the compression of the DT fuel using implosion dynamics. However, the first phase is strongly affected by the optimization of the second phase, ignition dynamic. In previous work, we have designed an optimum target which is highly compressed using multi-step laser pulse irradiation to a solid spherical target with gold conical target. For the FIREX-I scale implosion, two-dimensional radiation hydrodynamic simulation result shows that the $\rho R_{\rm max}$ was 0.38 g/cm$^2$ at most. Based on the hydro-equivalent, we estimate that the requirement of implosion laser energy for ignition-scale-target was 135 kJ $^{3}$. In the previous work, the effect of relativistic laser plasma interaction (LPI) caused by the heating laser in the ignition phase was not considered sufficiently. In order to save the computational resources, the FIREX-I scale size target was optimized, and larger targets were extrapolated based on the hydro-equivalent. In the hydro-equivalent implosions, the same shock velocity, adiabat, mass density and laser intensity lead the similar hydrodynamics, where an implosion laser pulse duration is proportional to the initial target radius, $t_{\rm L}$∝$R_{\rm 0}$, a total implosion laser energy is proportional to the target volume, $E_{\rm imp}$∝$R_{\rm 0}^3$. The requirement of the implosion process of the fast ignition is determined by an areal density of compressed fuel, which is proportional to the target radius, $\rho R$ ∝$R_{\rm 0}$∝$(E_{\rm imp})^{(1/3)}$, if the maximum densities are same in the both scales.
On the other hand, relativistic LPI is nonlinear phenomena, and the radius of the inserted cone tip is determined by the specification of the heating short pulse laser, especially by the laser intensity, pulse duration, and energy$^{4}$. For example, in the FIREX-I scale target, the stopping range of electron beam should be lower than 1.2 g/cm$^2$ (=0.6 g/cm$^2$ in $\rho R$). That is, the slope temperature of the electron beam should be less than 1 MeV, and the intensity of the heating laser should be lower than the 3$\times$10$^{19}$ W/cm$^2$ approximately. On the other hand, for the ignition and burning scale target, the tolerance of the laser intensity is increased. Ac-cording to the Ponderomotive scaling$^{5}$, $T_{\rm REB}$=3 MeV corresponds to the laser intensity of 2.3$\times$10$^{20}$ W/cm$^2$ for a 2nd harmonics pulse ($\lambda_{\rm L}$=0.53 $\mu$m). Therefore, the radius of the tip of the cone can be smaller than the previous design study.
Based on the previous optimized design, only the radius of the tip of the cone is varied as a design variable to maximize the areal density of compressed fuel core. For the numerical survey, we conducted the two-dimensional implosion simulation using a radiation hydrodynamic simulation code, PINOCO$^{6}$, in which gold cone target for guiding heating laser is attached to the solid spherical fuel target (initial target radius: $R_0$=263 $\mu$m). The open angle, and thickness of the gold cone are 30 degrees$^{6}$ and 15 $\mu$m$^{3}$ respectively. The material of the tip is CD. The laser energy is 8 kJ ($\lambda_{\rm L}$=0.35 $\mu$m), and the laser pulse is multi-step in order to minimize entropy increase.
The maximum areal densities ($\rho R$) with respect to the tip position ($d_{\rm tip}$) and tip radius ($r_{\rm tip}$) are plotted in Fig.1. In our previous work$^{3}$, we have concluded that the optimum position of the cone tip is $d_{\rm tip}$=60 $\mu$m, considering the tip-breakup-timing and the divergence angle of the energetic electron beam. Finally, the areal density of fuel core is significantly increased by 21% to $\rho R$=0.46 g/cm$^2$ using a cone whose tip radius is $r_{\rm tip}$=13.3 $\mu$m. Considering the hydro-equivalent, $r_{\rm tip}$=13.3 $\mu$m corresponds to $r_{\rm tip}$=28.8 $\mu$m for the ignition target with 82 kJ, and $r_{\rm tip}$=72 $\mu$m for the burning target with 1.3 MJ of implosion laser energy, respectively. In this configuration, the intensity of the heating laser does not exceed 2.3×10$^{20}$ W/cm$^2$. The estimated required laser energies for implosion and other specifications are summarized in Fig. 2 (table). This requirement may be reduced by sophisticated laser spatial and temporal pattern controlling in future work. Furthermore, in order to optimize the whole process of fast ignition, integrated simulation is necessary including the simulation of heating process with kinetic codes$^{6,8-9}$.
References
- M. Tabak et al., Phys. Plasmas 1, 1626-1634, 1994.
- K. Mima, et al., IAEA-CN-94/IF/03,2002.
- H. Nagatomo et al., Nuclear Fusion 59, 106055, 2019.
- T. Johzaki et al., submitted to High Energy Density Physics.
- S.C. Wilks, et al., Phys. Rev. Lett. 69 (9) 138, 1992.
- H. Nagatomo et al., Nuclear Fusion 55, 093028, 2015.
- T. Nakamura, et al., Laser and Particle Beams 26 207-212, 2008.
- H. Sakagami et al., Laser and Particle Beams 24 191-198, 2006.
- T. Johzaki, et al., Nuclear Fusion 55, 053022, 2015.
Speaker: Hideo Nagatomo (Osaka University) -
Electric fields in plasma plays a key role in understanding many plasma phenomena from confinement to particle flows. In fusion machines like tokamak, changes in the edge radial electric field are also correlated with changes in many edge phenomena such as L-H transition, rotation, transport, and the suppression of large magnetohydrodynamic (MHD) instabilities called ELMs (Edge Localized Modes) making it a very important parameter to be known with precise resolution both spatially and temporally. Gradients of plasma potential, gives the electric field present in the plasma, can be measured directly by emissive probes. Exponential region in emissive probes depends on Tw [wire temperature] while of collecting probes depends on Te [electron temperature]. Since in most plasmas Tw << Te, emissive probes can measure the plasma potential more accurately 1. Further, test experiments in CASTOR concludes the difference between the potential of cold probe and emissive probe is 1.3 times the electron temperature 2. Emissive electrode biasing experiments in ISTTOK shows its vital role in confinement of particles [3]. However, typical limitations of emissive probes such as the potential drop cross probe wire results in temperature non-uniformity across the filament rendering non-uniform electron emission. Requirement of separate heating and biasing electronic circuit increases make electronic circuitry more complex while need for filament replacement makes emissive probes less popular to be used in tokamaks as cited by Kella V P et al. These problems become more significant in presence of strong potential gradients and in magnetic fields as described by Schrittwieser R et al. Laser Heated Emissive Probes (LHEP) provides an excellent solution as suggested by Sheehan and Schrittwieser R. separately. Although LHEP eases out some of the limitations of emissive probes for plasma potential measurements, it brings new challenges regarding proper choice of probe-tip material, aligning the laser to the probe-tip inside the vacuum, etc. Owing the fact of high magnetic field environment, UHV, mechanical constraints and many tokamak exclusive challenges no LHEP, to our knowledge, is attempted to be used in tokamak. A LHEP system is developed and installed on ADITYA-U tokamak for obtaining first measurements from such probes in tokamaks. The radially movable LHEP system comprising of two probes of 6 mm diameter each, is mounted on radial port on the Aditya-U tokamak. LHEP system is designed to measure plasma potential and its fluctuations in the SOL region taking into account of the experimental and technical parameters of Aditya-U. Novelty in design is the compactness of system and continuous focusing of laser (cw CO2 laser at 10.6 µm, 55 W) onto probe surface irrespective of the radial movement of probes all without disturbing UHV of Aditya-U. Laser is shined on to the probetip using an air-cooled optical fiber. The crucial design facilitates up to 1 mm accuracy to fix radial positions of probes while up to 0.1 mm accuracy to focus laser onto probe surface. In house designed and developed electronics for LHEP ensures data acquisition with maximum signal to noise ratio along with probe biasing as per tokamak requirements. The photograph of the installed LHEP diagnostic system (both probe support system and laser focusing system) is shown in figure 1.
Two graphite probes of diameter 6 mm is placed at r ~ 26.5 cm (in the SOL region, 1.5 cm outside the limiter) in ADITYA-U tokamak edge region. Before, heating the probes, the cold probe characteristics is obtained for both the probes. The temporal profile of floating potential recorded from
both the probes in cold condition are plotted in figure 2.
The variation in floating potential with plasma column movement and with periodic gas puffs are studied in detail. Furthermore, applying a voltage sweep to the probes, electron density and temperature are estimated and their variation with gas puff are recorded for comparing the cold characteristics with LHEP characteristics. After these initial measurements, the LHEP probe-tip is biased with respect to vessel thereby recording V-I characteristics for several probe temperatures to ascertain the constant emission region of the probe. The plasma potential is then determined using the floating potential method [6]. Experiments have been carried out by measuring the plasma potential fluctuations by heating one probe with Laser and measuring floating potential fluctuations from the other probe in cold condition, fluctuations in electron temperature have been obtained in the edge region of ADITYA-U tokamak. In this paper, first measurements of LHEP from tokamak edge region is reported. The mean value of plasma potential and its fluctuations are studied in different plasma discharges during gas-puff applications of different gases such as neon, argon etc. is studied to show how different gases modify the edge plasma parameters in ADITYA-U tokamak.References:
- Sheehan J P and Hershkowitz N 2011 Emissive probes Plasma Sources Sci. Technol. 20 063001
- Roman Schrittwieser et al 2002 Measurement with the emissive probe in CASTOR tokamak Plasma Phys. Control. Fusion 44 567–578
- C. Silva et al 2004 Improved confinement events triggered by emissive electrode biasing on the tokamak ISTTOK Nucl. Fusion 44 799
- Kella V P, Mehta P, Sarma A, Ghosh J and Chattopadhyay P K 2016 A comparative study of sheath potential profile measurements with laser-heated and current-heated emissive probes Rev. Sci. Instrum. 87 043508
- Schrittwieser R et al 2008 Laser-heated emissive plasma probe Rev. Sci. Instrum. 79 083508
- J. P. Sheehan 2013 Accurately Determining the Plasma Potential Using Emissive Probes The 33rd International Electric Propulsion Conference, The George Washington University, USA.
- Payal M, Arun S and Joydeep G 2014 Laser heated emissive probes design and development under national fusion program and potential measurement Chin. Phys. Lett. 31 125201
Speaker: Abha Kanik (Gujarat University, Navrangpura, Ahmedabad) -
The contamination of core plasma by high-Z impurities, especially tungsten (W), is the main reason of very high level of radiated power in WEST [1] experiments. Determining the main sources of core contamination is indeed a key aspect in preparing a high confinement scenario for the second phase of WEST operation that will start at the end of 2020. Intrinsic light impurities, mainly oxygen and carbon, play a dominant role in the sputtering of W on plasma facing components (PFCs) and it is crucial to investigate their transport and spatial distribution in edge and SOL plasmas. In this contribution, we present a detailed analysis of WEST experiments supported by numerical modeling performed with the transport code SOLEDGE-EIRENE providing a clear picture of the impact of light impurities to W sources from divertor and main chamber PFCs.
SOLEDGE [2, 3] is a unique numerical tool for this kind of studies for two main reasons: firstly, it handles complex and realistic wall geometries thanks to the penalization technique allowing us to properly taking into account the interaction between the plasma and the multiplicity of objects located in the vessel. Secondly, with the recent implementation of multi-fluid collisional closure [4] it is now possible to estimate properly the parallel dynamics and poloidal distribution of light impurities in the edge and SOL plasmas, determined by the competition between thermal gradient and friction forces, without relying on the trace approximation.
We focus on a series of shots related to the experiment on “High power test of ITER-like plasma-facing components, exposure of pre-damaged PFC”, supported by the EUROfusion program. During these discharges, 4MW have been injected for about 5s, in lower single null, L-mode plasmas with 2.3MW of total radiated power, 4*10^19 m-2 of central line integrated density and plasma current of 500 kA with a height of the X point of about 115 mm above the lower divertor target plate. From experimental analysis it was clear that important contributors to W sputtering are oxygen but also carbon which comes probably from antennas limiters and lower divertor W-coated PFCs [5].
The input parameters for SOLEDGE simulations are the separatrix density at the outer mid-plane and the total injected power into the simulation domain that have been obtained combining measurements from reciprocating Langmuir probe, fast-sweep reflectometry and bolometry diagnostics. Radial transport coefficients for plasma density and temperature have been settled to D = 0.3m^2/s and χ = 1m^2/s respectively, values from which the SOL width recovered in the simulations is in agreement with the one measured in WEST experiments. The simulations results are then analyzed comparing density and temperature profiles at the lower divertor targets with experimental data from locally embedded Langmuir probes as well as the total radiated power with respect to the one obtained from bolometry measurements. In order to determine the percentage of light impurities present in the discharge, we consider simulations with oxygen as an effective medium Z charge impurity for sputtering. A parametric scan on oxygen concentration is performed, assuming from 1% to 4% of oxygen injected in SOLEDGE at the inner boundary of simulation domain. The simulations with 2% of oxygen are the ones matching quite well both divertor target profiles as well as total radiated power (see Fig.1).
To go further into the simulation/experiment comparison we look at the poloidal distribution of oxygen. The simulation results show a strong asymmetry between oxygen concentrations at the inner divertor target with respect to the outer one. This asymmetry has been measured in the experiments thanks to the WEST Vacuum UltraViolet (VUV) spectroscopy system with a moving line of sight in the poloidal plane, allowing one to retrieve the relative information on the angular position of the oxygen light emission.
The detailed force balance analysis on simulation results shows that thermal gradient forces are stronger at the outer divertor target pushing the oxygen toward upstream location and producing its depletion in this region. This result is in agreement with O IV (O^(3+)) line from VUV spectroscopy signals, as showed in Fig. 2. We have also computed the sources of W due to sputtering from both deuterium and oxygen ions in the series of simulations previously presented. One observers that the in-out asymmetry in W gross erosion source reverses for “oxygen-driven” vs “deuterium-driven” case. First data analysis from visible spectroscopy diagnostic confirms that for high input power the contribution from lower divertor inner target is much larger than that from outer target. Other key contributors are the baffle and antenna protection [6]. These findings are informing future plans to control these light impurities and W sources. First, to reduce the oxygen content in the plasma it will be important to continue to improve wall conditioning using both glow discharge cleaning as well as active boronization techniques. In addition, the development of a robust semi-detached divertor plasma scenario will permit to operate below the sputtering threshold of light impurities, reducing strongly W sources and opening the way to high confinement scenarios.
References:
[1] J. Bucalossi and the WEST team, 2014, Fusion Eng. Design 89, 907-912. [2] H. Bufferand et al. 2015, Nucl. Fusion 55, 053025. [3] G. Ciraolo et al, 2019, Nucl. Mat & En., 20, 100685 [4] H Bufferand et al, 2019, Nucl. Mat. & En., 18, 82. [5] G. J. van Rooij et al, 2020 Phys. Scripta, in press [6] A Gallo et al, to be submitted to Nuclear Fusion.Speaker: GUIDO CIRAOLO (CEA, IRFM) -
Executive summary: We report on the results of a quantitative study, based on experimental data, of the role plasma-atom and plasma-molecule interactions in power, particle and momentum balance during detachment. Important implications emerge for: 1) our understanding of detachment; 2) the interpretation of divertor spectroscopy measurements and 3) plasma-edge modelling, where the treatment of molecules may be immature. Our analysis of data from TCV shows detachment starts as the power required for ionisation approaches the power entering the ionisation region, limiting the ion source and thus the ion target flux. From that point onwards momentum losses, enhanced plasma-molecule interactions (with $H_2^+$ (and $H^-$)) and ultimately electron-ion recombination develop. The influence of these plasma-molecule interactions on power, particle and momentum balance are inferred to be significant compared to atomic interactions alone.
The development of detachment is affected by a range of atomic and molecular processes. There is currently no consensus regarding the relative importance of momentum and power loss, the role of molecular reactions nor the sequence of various processes observed during detachment. More specifically, it remains unclear whether the target ion current ($I_t$) reduction, observed during detachment, results from reductions in ion sources, increases in ion sinks, or both. In this work we investigate such issues experimentally through spectroscopy on TCV with emphasis on the effect of molecules.
For these investigations, we have developed and applied analysis techniques to the deuterium Balmer line emission, generating quantitative estimates of: (a) the atom-derived ion source and the related ionisation energy cost 1 (ion sinks/recombination have been addressed for over 20 years [2]); (b) the contributions of plasma-molecule interactions (with $H_2^+$ and $H^-$) to ion sources/sinks and hydrogenic radiative losses [3]. These techniques exploit the different sensitivities of each Balmer line to the various possible atomic and ‘molecular’ emission channels ($H$, $H^+$, $H_2$, $H_2^+$, $H^-$) to dissect each Balmer line into its emission channels and estimate the associated radiative loss and particle sink/source.
The results [4] of our analysis of a TCV core density ramp attaining detachment (see figure 1), indicate that, in the first detachment phase (red phase - target temperature, $T_t \sim 3 - 6$ eV), the observed saturation and decrease in It results from a saturation and reduction in the ion source. This is driven by both an increased energy cost of ionisation and a reduction in the power available for ionisation (due to increased upstream impurity radiation) [2,3]. As detachment develops further (green phase), the power available for ionisation decreases further, reducing $T_t$ ($ \leq 1.5$ eV) and increasing the electron-ion recombination (EIR, $<15 \%$ $I_t$) ion sink. This is consistent with SOLPS modelling results [4,5] as well as analytic divertor modelling [4].
The bifurcation between the measured $H\alpha$ emission and its atomic estimate indicates plasma-molecule interactions with $H_2^+$ (and possibly $H^-$) are enhanced near the detachment onset, resulting in excited H atoms and atomic line radiation losses. Our analysis results (see figure 1) indicate these plasma-molecule interactions with $H_2^+$ (and possibly $H^-$) result in: 1) significant ion sinks ($<35 \%$ $I_t$) through Molecular-Activated Recombination (MAR), which commence before EIR and remain higher; 2) excited atom radiation that can account for more than 50% of all hydrogenic line radiation; 3) significant enhancements to $n=3-6$ Balmer lines (not shown). The emission enhancement for the $n=5$ Balmer line (up to $30 \%$) result in reduced ion source estimates in the deepest detached phase, compared to previous work 1 that only accounted for atomic contributions to the Balmer line emission.
At least on TCV, we thus conclude that the influence of plasma-molecule interactions (with $H_2^+$ and possibly $H^-$) on particle and power balance can be significant. This has important implications for interpretation of divertor Balmer line spectra as well as divertor modelling codes where the present treatment of molecules may be incomplete and incorrect. The isotope dependence of the reactions as well as their dependence on vibrational states may not be properly accounted for.
With the inclusion of ion losses through MAR, TCV’s divertor particle balance implies an ion flow from upstream contributing to It in the deepest detached phases. This result will be compared further against SOLPS simulations [5] that indicate an increase in the ion flow from upstream towards the target during detachment. The MAR ion loss channel (in contrast to electron-ion recombination) leads to significant radiative losses. The implication of this (and the ion flow from upstream) will be investigated using analytic divertor modelling by extending the analytic Two Point Model (with Recycling - 2PMR [3]).
Excitation of $H_2$ through collisions between the electrons and $H_2$ occur in addition to reactions between the plasma and $H_2^+$ & $H^-$ that lead to the MAR and atomic line radiation losses explored above. Such collisions can result in significant power and momentum transfer from the plasma to the molecules according to TCV and MAST-U SOLPS simulations [6,7]. Through measuring the molecular ($H_2$) band emission, we find that excitation through these collisions occur in different regions of the plasma than reactions with $H_2^+$ & $H^-$.
Our techniques allow the estimation of strength, sequence and spatial distribution of the various plasma-molecule interactions during detachment. Application to divertors of dissimilar divertor geometries/chambers (e.g. TCV and MAST-U) may be useful in providing insight as to how geometry modifies the relative roles of plasma-atom and plasma-molecule interactions in detachment.
$[1]$ K. Verhaegh, et al. Plasma Phys. Control. Fusion (2019) 61 125018
$[2]$ B. Lipschultz, et al. Phys. Plasmas (1999) 6, 1907
$[3]$ K. Verhaegh, et al. Nucl. Fusion (2019) 59 126038
$[4]$ K. Verhaegh, et al. to be submitted
$[5]$ A. Fil et al, Plasma Phys. Control. Fusion (2020) 62 035008
$[6]$ A. Smolders, et al. submitted
$[7]$ O. Myatra, et al. to be submitted.
Speaker: Kevin Verhaegh (CCFE) -
Tokamak plasmas are mostly optically thin for visible radiation-emitting out of the plasma as the density of the emitter in the plasma is not sufficiently high to produced self-absorption of spectral lines in the visible region. However, in certain conditions, such as during pellet ablation inside plasma and massive gas injection signatures of absorption in the emission spectra might be observed, which can be very useful for comparison with radiative transfer modeling in fusion devices. One of the major outcomes of self-absorption is to determine the plasma opacity and to understand the self-absorption and prevailing density values. This is mostly happened for the spectrum having upper level with lower principal quantum numbers, which have greater line strengths [ref1]. Then, it is usually observed in the Ly-$\alpha$ emission as has been done in many high temperature plasma devices, such as large helical device [ref2].
However, self-absorbed spectral lines other than hydrogen from tokamak emission in the visible range are not reported till now as per our knowledge. In Aditya-U tokamak, self-absorbed spectral line from neutral lithium has been observed for first time during recently carried out disruption mitigation experiment using a novel Inductive Pallet Injection system design, built and installed on Aditya-U tokamak [ref3]. With this system disruption mitigation experiment was carried out successfully by injecting few hundred milligrams of Li2TiO3 particle and resulting the quenching of the plasma in a few milliseconds through the intense radiation (figure-1). In this work, the detailed analysis of the self-absorbed spectral line are quantitatively evaluated using the radiation transport model to get the ground state lithium density and the atom layer thickness.
The Lithium neutral (670.8 nm 2s2S–2p2P)spectral line shape has been monitored during the disruption mitigation experiment using a space resolved spectroscopic system having capabilities to record the emissions from different lines of sight passing through various plasma radial locations was used to record the spatial profile of the radiation. The measurement system is having a 1 m long multi-track Spectrometer (MTS) and CCD detector with a reciprocal linear dispersion 0.0056 nm/pixel. During the time of disruption after the injection of particles, the recorded spectral line of the Li neutral showed a dip in intensity near its central wavelength. The spectral line shape is plotted in figure2. After thorough analysis of the spectral line-shape of the neutral Li, it has been concluded that the observed dip in the intensity near the line center in the emission spectra of Li is most likely related to the opacity effect due to the high density of lithium neutral during the time of the measurement. This observed spectral line-shape has been fitted with two Gaussian line profile one corresponding to emission spectrum and another corresponding to absorption spectrum. Figure2 shows the fitting of the observed spectral line by using a Gaussian profile corresponding to Doppler and Instrumental Broadening width along with a negative Gaussian profile present at near to the line center.
In this paper, the detailed investigation of the self-absorption of LiI spectral line emission has been carried out to understand the opacity effect. From this quantitative analysis using the radiation transport modeling the ground state lithium density and the atom layer thickness has been obtained. The absorption peak is also analyzed for obtaining the Li densities and its radial profile at the time of particles injection into the plasma during the disruption mitigation experiments.
References:
- Ghosh, J., et al. "Vacuum ultraviolet measurements on hydrogen resonance lines in the maryland centrifugal experiment." Physics of plasmas 12.3 (2005): 034501.
- Goto M. et al, Plasma Fusion Res. 5, S2089 (2010)
- Paper in this conference by R. L. Tanna et al
Speaker: Nandini Yadava (Gujarat University) -
In the ADITYA-U tokamak 1, impurity seeding experiments were carried out to achieve transitions to radiative improved (RI) modes [2], which is usually characterized by the increased confinement along with increased plasma density, temperature and toroidal velocity shear profile as compared to the Ohmic mode. In this type of discharges, radiation from the edge region of the plasma is increased by a large fraction ( 50 %) by seeding the mid Z impurities, like neon and argon. The improved confinement in RI mode plasmas is believed to be related to the reduction of growth characteristics of the toroidal ion temperature gradient (ITG) mode due to the increase of plasma effective charge, Zeff. The suppression of turbulence through the increase of EB shear rotation in the impurity injected plasma [3] is another possible mechanism for improving the plasma properties in RI mode. In neon gas puffed discharges of ADITYA-U tokamak, it was observed that the reversal of toroidal rotation occurred at lower values of density as compared to the hydrogen puffed Ohmic discharges, indicating lower electron density threshold for toroidal rotation reversal with neon seeding. The radial profile of toroidal rotation in ADITYA-U tokamak has been systematically studied for both neon and argon gases seeding discharges to understand the mechanism responsible for lowering the threshold of density for rotation reversal to occur.
Impurities seeding experiments in ADITYA-U were carried by puffing the neon and argon gases in appropriate amount during the plasma current flat-top region using a gas fuelling system consists of piezo-electric valve and a programmable pulse generator. The puff duration, number of puffs and voltage applied to Piezo-electric valve were altered in the experiment to study the effect of amount and duration of gas injection on toroidal rotation. The impurities seeded plasmas of ADITYA-U tokamak [4] are characterized by the increase of line average electron density, ne and central electron temperature, Te(0), as measured by single channel microwave heterodyne interferometer and the soft X-ray diagnostics, respectively. Substantial change in plasma edge properties were observed with the increase of radiation as measured by AXUV detector and the reduction of hydrogen recycling illustrated by lower H signal monitored by optical fiber, interference filter and photo multiplier tube (PMT) based spectroscopic system. In these impurities seeded discharges, particle and energy confinement times were also increased significantly. These improved characteristics were initiated only after 10 to 15 ms of the gas injection depending upon the amount of gas injected into the plasma.
The radial profiles of toroidal rotation of the plasma, Ut, were thoroughly analysed for the discharges with neon and argon gases puffing. The toroidal rotation profile was measured by monitoring the Doppler-shifted passive charge exchange line emission at 529 nm from C5+ ion. This spatial profile of this spectral line was monitored using a high resolution spectroscopic diagnostics [5, 6] capable to obtain the spatial profile collecting the light simultaneously using seven lines of sight (LoS) with a spatial resolution of ~ 2.2 cm. To obtain the complete radial profile of rotation the light collection optics were placed inside a re-entrant tube mounted on a tangential port of ADITYA-U tokamak. The system was having a multi-track enabled 1.0 m long visible spectrometer and charge coupled device (CCD) with the reciprocal linear dispersion of ~ 0.0063 nm/pixel. The toroidal rotation radial profile was obtained by converting the spatial profile using ABEL-like matrix inversion technique. As shown in figure 1, the plasma core was observed to be rotating in the toroidal direction with a velocity of ~12 km.s−1 for a typical neon seeded discharge of ADITYA-U tokamak and the direction of rotation was in the co-current direction whereas the plasma rotates in the counter-current direction in the edge region. The reversal of the toroidal rotation profiles are analyzed by scanning ne for the impurities seeded discharges and results are compared with Ohmic discharges with hydrogen gas puff. It has been observed that the density threshold for rotation reversal can be brought down using impurity seeding in the ADITYA-U plasmas. Most importantly, the modification in the rotation is not accompanied by usually observed turbulence reduction in the plasma edge as no change in fluctuation in the ion saturation current is detected in the ADITYA-U tokamak's impurities seeding plasmas.
1. Tanna, R.L. et al 2019 Nucl. Fuion, 59, 112006 (2019).
2. Untenberg R. et al. Fusion Sci. Technol. 47 187 (2005).
3. Tokar M. Z. et al, 2000 Phys. Rev. Lett. 84 895 2000.
4. Chowdhuri M. B. et al., Proc. 27th IAEA Fusion Energy Conference, held during October 22-27, 2018 at Gandhinagar, India.
5. Shukla, G. et al. Rev. Sci. Instrum. 89, 10D132 (2018).
6. Shukla G. et al. submitted for presentation at 23rd Topical Conference on High-Temperature Plasma Diagnostics, going to be held during May 31- Jun 04, 2020 at Santa Fe, NM, USASpeaker: Joydeep Ghosh (Institute for Plasma Research) -
EX
The new results of the experimental study of the ion cyclotron emission (ICE) characteristic features in the NBI heated plasma of the TUMAN-3M tokamak 1 are presented. Figure 1 shows an example of the NBI ICE spectrum in deuterium plasma with deuterium NBI, comprising fundamental cyclotron resonance (IC) frequency and its harmonics. For the first time, a dispersion relation for the NBI ICE was investigated experimentally by two independent methods: using Doppler-shifted resonance condition (results presented in Fig. 2) or performing phase analysis of the signals of toroidally separated detectors. It is found that the observed NBI ICE spectra agree with the assumption that the underlying instability responsible for ICE excitation is magneto-acoustic one propagating nearly perpendicular to the magnetic field. NBI ICE spectra observed in deuterium and hydrogen plasma are different not only in terms of mass dependence of cyclotron frequency but also in relative amplitude and temporal dynamics of the lines comprising the spectra. Other characteristic features of the spectra, such as their fine structure and its dependence on isotope composition, are also discussed.
The emission of electromagnetic waves in the ion cyclotron range of frequencies is routinely observed at many magnetic fusion devices in the presence of fast ions 2. The non-equilibrium population of fast particles may be created by auxiliary heating systems, or appear as a result of fusion reactions [3]. As spectral properties of ICE are closely related to the fast ion population dynamics, it is considered to be a prospective diagnostic tool for fast ion component.
The experiments were performed in hydrogen and deuterium plasmas, with hydrogen, deuterium, or mixed beams with up to $\ E_0$=20 keV of energy and P = 0.3 MW of net power injected tangentially in co-current direction. The ICE was observed using in-vessel arrays of magnetic probes equipped with appropriate high-frequency electronics and ADCs. In all the cases, the frequency of the ICE observed lies in the range of 5 to 30 MHz and corresponds to the ion cyclotron resonance (ICR) frequency (and its harmonics) in the vicinity of the plasma center. This is remarkably different from observations on the other tokamaks, where ICE frequency corresponds, with rare exclusions, to resonance condition at the plasma periphery. When the mixed beam was injected (~70% D plus ~30% H), the observed emission frequency spectra correspond to the minority fast ion’s IC resonance and its harmonics (i.e. hydrogen ICR in D-plasma, and deuterium ICR in H-plasma). If the pure D or H beam was used, emission at the harmonics of the ICR frequency of the main ion was registered, with fundamental frequency strongly suppressed in D-plasma, but strong enough in H-plasma. Typically, NBI ICE spectral lines consist of 2 to 4 and more narrow (~ 50 kHz wide) lines separated by uneven intervals of 100 to 300 kHz. These spectral lines behave differently if saw-tooth oscillations or MHD islands are present in the discharge, indicating the different locations of the resonance region and/or fast ions causing them. The NBI ICE appears with a 2-5ms delay after a start of NBI pulse, obviously due to the necessity of a significant FI population to build up. On the other hand, ICE disappears not later than ~100 $\mu s$ after NBI termination, much faster than FI slowing down time. In some shots, NBI ICE terminates well before the end of the NBI pulse. This observation suggests that the presence of FI is a necessary but not a sufficient condition for the ICE to develop. Obviously, a balance between the FI drive and dumping of the underlying instability is a key factor regulating ICE excitation.
Calculations of the fast ion trajectories in the TUMAN-3M geometry revealed most probable candidate source of free energy for excitation of the instability observed – the stagnation fast ions (SFI) residing close to the magnetic axis or, in some cases, the potato ions (PFI), having much wider drift trajectories [4, 5]. When traveling along both types of drift trajectories, fast ions spend most of their flight time close to the plasma center, thus accentuating central ICR frequency. For both types of FI, Doppler-shifted resonance condition in a form $\omega = l\omega_{ci}+\textbf{kV}_{fi}$ was used for dispersion relation analysis [6]. Here, $\omega$ and $\textbf{k}$ are the ICE frequency and wavenumber, $\omega_{ci}$ is ICR frequency of the FI, $\textbf{V}_{fi}$ is its velocity, l is harmonic’s number. For SFI, $\textbf{kV}_{fi}\approx k_\parallel V_\parallel$, where index || denotes component of the wave vector and FI velocity parallel to the magnetic field. For PFI, $\textbf{kV}_{fi}\approx k_\perp V_d$, where $\ k_\perp$ is a component of wave vector perpendicular to the magnetic field, and $\ V_d$ is the gradient drift velocity of the PFI at the vertical part of the potato trajectory. Another factor influencing the ICE spectrum is the energy spectrum of the injected atomic beam. Fractions of atoms with energies $\ E_0$/2 and $\ E_0$/3 are usually present in the beam. The FIs with such energies form SFIs and PFIs with slightly shifted trajectories, with respect to the trajectory of FIs with energy $\ E_0$.
Dispersion relation analysis of the NBI ICE was performed using the above-mentioned Doppler-shifted resonance condition for fast ions residing on the stagnation trajectories. The latter were found from the equation of motion for FI with known total energy and in realistic experimental magnetic field configuration of the selected plasma shot. Figure 1 shows an example of the NBI spectrum observed in D-plasma with toroidal field $\ B_T$ = 0.96 T during injection of D-beam with energy $\ E_0$ = 17 keV. Fundamental central deuterium ICR frequency ~6.6 MHz and four harmonics are seen. The second harmonic is split into two lines separated by ~200 kHz, while the third harmonic is very weak and hardly seen in this graph, but can be resolved after appropriate filtering. Parallel wavenumbers corresponding to each spectral line are plotted as a function of frequency in Figure 2. The solid line indicates dispersion relation for the magneto-acoustic instability $\omega = kV_A = (k_{||}/\cos\alpha )V_A$, here $\alpha $ is a propagation angle (between $\textbf{k}$ and $\textbf{B}$), $\ V_A$ – Alfven velocity. As seen, a good agreement is reached for $\alpha=84^0$, i.e. nearly perpendicular propagation. With the help of toroidally and poloidally separated probes, it was possible to analyze the spatial structure of the observed NBI ICE – their mode numbers m and n, and direction of propagation as well. Results are found to give a close estimation of the propagation angle.
Similar in frequency range but different in underlying physics is the ICE observed on the TUMAN-3M in the absence of fast ions in the ohmically heated plasma. Some of the characteristic features of the ohmic ICE, with emphasis on their modifications under the NBI influence, are discussed.
This study of NBI ICE was supported by the RSF, project no. 16-12-10285-P. The experiments on the TUMAN-3M tokamak were performed under the state assignment of the Ioffe Institute, topic No. 0040-2019-0023.
$\textbf{References}$
1. L.G. Askinazi et al 2018 Nucl. Fusion 58 082003
2. McClements K.G. et al 2015 Nucl. Fusion 55 043013
3. Cottrell G.A. et al 1993 Nucl. Fusion 33 1365
4. L.G. Askinazi et al 2018 45th EPS Conference on Plasma Physics P5.1084
5. L.G. Askinazi et al Tech. Phys. Lett. 44, 1020–1023 (2018).
6. S. E. Sharapov et al, Phys. Plasmas 21, 082501 (2014).Speaker: Leonid Askinazi (Ioffe Institute) -
Under the auspices of EUROfusion, the ITER baseline scenario (IBL, [1]) is jointly investigated on AUG and TCV. While AUG results were presented at the last IAEA [2], this contribution focuses on recent results obtained in TCV. Such developments in TCV were only possible with the installation of an NBI heating source [3], allowing ELMy H-modes at ITER relevant $\beta_N$. The IBL scenario is mainly characterized by low $q_{95}$ (3-3.6), high positive triangularity ($\delta$>0.3) and relatively high elongation ($\kappa$>1.6). In AUG, these combinations lead to very steep and narrow edge transport barriers, when good confinement is obtained, with high pedestal pressure and therefore large Type-I ELM crashes. A similar behavior is also observed on TCV, since indeed the target plasma shape has been derived from the IBL AUG shape, as shown in Fig. 1.
The AUG shape (dashed red) has been rescaled to match TCV geometrical radius and further rescaled to match the minor radius, since there is a 20% difference in aspect ratio. As can be seen from Fig. 2, the TCV triangularity is slightly increased as compared with AUG, in order to approach the ITER design one, while the elongation is slightly smaller. It should be emphasized that a positive triangularity of 0.3-0.5 falls exactly in the steepest region regarding the sensitivity of the pedestal pressure versus (averaged) triangularity ([4], Fig. 10).
The global performance of the recent TCV IBL discharges is reported in the usual diagrams of H$_{98\text{y}2}$ vs $\beta_N$, as well as AUG results [2] (Fig. 3). Fig. 3a shows that ITER target values ($\beta_N$=1.8 and H$_{98\text{y}2}$=1, q$_{95}$~3-3.2) have been obtained, similarly to AUG (Fig. 3b). On AUG, a scenario with $q_{95}$~3.6, $\beta_N$=2.2 and H$_{98\text{y}2}$=1.2, to keep $\beta_N$H$_{98\text{y}2}$/($q_{95}$)$^2$ constant has been further studied as well.
![(a) H$_{98\text{y}2}$ factor vs $\beta_N$ at the time just before a large ELM crash for recent TCV ITER baseline scenario heated with NBI mainly and X3 in some cases. (b) AUG recent results with grey area representing previous results [2].](https://crppwww.epfl.ch/~sauter/IAEA2020/IAEA2020_fig2_120.png)
Currently such scenario has not been tried on TCV at maximum power, the developments at lower current ($\beta_N$<1.7) being focused on stationary discharges without MHD modes (Fig. 3a). The time traces of the discharge 64770, highlighted with a red arrow in Fig. 3(a), are shown in Fig. 4 (red). The various phases are representative of the studies that will be discussed in this work. The high current (275kA, $q_{95}$=3.2) phase starts at 1.3s and lasts about 150ms (dashed lines) with $\beta_N$~1.6 before a 3/2 mode is triggered at the 3$^\text{rd}$ ELM crash yielding $\beta_N$~1.3-1.4 and a 2/1 mode at 1.5s leading to $\beta_N$<1 at which value it self-stabilizes. The confinement time is about 35ms (H$_{98\text{y}2}$=0.95), the ELM period 50-65ms (15-20Hz) and the current redistribution time 100-150ms. The "stationary" time interval marked in Fig. 4 by the dashed lines is therefore about one current redistribution time, which is quite long and comparable to other tokamaks IBL scenarios, however only 4-5 energy confinement times and only 3 ELM periods. The high triangularity and TCV short current redistribution time may lead to significant magnetic perturbation at ELM crashes which tend to systematically trigger low m/n modes, sometimes directly a 2/1 mode eventually locking but not necessarily leading to a disruption (as in 64678, Fig. 4black). This will be compared to the relation between sawtooth period, resistive time and NTM onset shown in Ref. [5]. The sensitivity to NTM onset explains why the developments of the TCV IBL largely rely, at first, in the establishment of a stationary phase at $q_{95}$>4 and then the discharge can evolve towards $q_{95}$=3.6 or 3-3.2 for studying both IBL scenarios already achieved on AUG [2]. Initiating the H-mode phase at reduced $I_p$ has another advantage which is to avoid any problem with the L-H transition and potentially large 1$^{st}$ ELM. This was already observed related to NTMs triggered by the first long sawtooth period when entering in H-mode on JET. All these transient events are much better controlled when occurring at $q_{95}$>4.5, including the transition into H-mode and the final shape development. This is in a large part also used in the AUG IBL scenario.
The important role of 3$^{rd}$ harmonic (X3) ECH in preventing low m/n MHD modes onset in TCV is reported. This has been clearly demonstrated in TCV, but up to a maximum plasma current so far (Ip=240kA) corresponding to $q_{95}$=3.6-4. This is also seen in Fig. 4, where the first phase at lower plasma current has a nice stationary ELMy H-mode with no significant MHD activity, contrary to a similar shot without X3 (64678). Note that the latter has a 2/1 mode stabilizing and the discharge recovers to high $\beta_N$ values (1.5-1.6s). The possible reasons for this plasma current/$q_{95}$ dependence on MHD activity prevention will be analysed. On the one hand, it is more difficult to influence the global q profile with electron heating at high total plasma current, on the other hand higher $I_p$ leads to higher density on TCV (relatively open divertor and no pumping). For example, in #64770, X3 absorption is around 15-20% in the first phase, 0.8-1.2s (Fig. 4) but reduces to less than 5% at the maximum current when the line-averaged density reaches 1.1e20 m$^{-3}$. Since significant density peaking is observed in these discharges, the central density reaches values above 1.6e20 m$^{-3}$, the cut-off density of the X3 heating sources. The density peaking itself can lead to a stronger electron temperature flattening and to plasmas more prone to instabilities. Gyrokinetic studies and density peaking predicted by quasi-linear analyses will be presented.
It is worth mentioning that no significant carbon accumulation has been observed so far, however dedicated impurity seeding experiments have to be performed. The role of ECH versus NBI heating will be discussed, in particular at lower plasma current where X3 is still absorbed and with regards to the effects on density peaking. The latter might be due to improved core confinement. In most TCV IBL cases $T_e$~$T_i$ due to the relatively high density. Note that the Greenwald fraction obtained on TCV is still below about 0.6 despite the new baffles, while AUG obtains discharges up to Greenwald densities.
In AUG, recent discharges comparing heating mix, pellets vs gas puff and nitrogen seeding essentially provided similar performances as previously observed (grey area in Fig. 3b). Only the lower collisionality cases seem to recover the ITER design performance (green points). These low density discharges rely on magnetic perturbation and are quite difficult to obtain without locked mode.
Finally, controlled ramp-down phases, including safe H-L exit, were tested with the IBL scenario on AUG, guided by simulations [6], and reproducible, safe and relatively fast ramp-down have been obtained, showing how a combined control of current ramp-rate, shape and power can be beneficial.References
[1] A. C. C. Sips et al, Nucl. Fusion 58 (2018) 126010
[2] T. Pütterich et al, 2018 IAEA FEC, IAEA-CN-EX/P8-4
[3] A. Fasoli et al, Nucl. Fusion 49 (2009) 104005
[4] A. Merle et al, Plasma Phys. Control. Fusion 59 (2017) 104001
[5] I. T. Chapman et al, Nucl. Fusion 50 (2010) 102001
[6] A. A. Teplukhina, Plasma Phys. Control. Fusion 59 (2017) 124004Speaker: Olivier Sauter -
The theoretical model of the feedback instability is proposed to explain the mechanism of the correlation between the detachment and the cross-field plasma transport. It is shown that (1) the feedback instability on the detached divertor plasma can be induced in a certain condition in which the recombination frequency $\nu_{\textrm{rec}}$ is larger than the ion cyclotron frequency $\Omega_{\textrm{ci}}$ in the recombination region and the density gradient and the electric field in the direction perpendicular to the magnetic flux surface are not zero, and that (2) the feedback instability can provide the cross-field plasma transport in the boundary layer of magnetic fusion torus devices.
The correlation between the detachment and the cross-field plasma transport in the boundary layer has been reported in various magnetic confinement devices, that is, tokamak$^{1}$, helical$^{2}$, and linear$^{3}$ devices. Such a correlation is expected to expand the width of the heat flux to the divertor target, i.e., $\lambda_{\textrm{q}}$. However, the physical dynamics of the correlation has not been revealed. In this study, we investigate the cross-field dynamics in the detached plasma state with the coupling model between magnetized plasmas characterized by different current mechanisms. In the recombination region in front of a divertor target, $\nu_{\textrm{rec}}$ can be larger than $\Omega_{\textrm{ci}}$ because of the high density and the low temperature. In such a situation, the cross-field motion of ions is mainly in the direction of the electric field, while that of electrons is almost in the direction of the $E \times B$ drift. Thus, the difference in the direction of motion may provide the cross-field current in the recombination region. On the other hand, the cross-field current can be generated by only the polarization, the grad-$B$, and the diamagnetic drifts in the upstream plasma. We have considered whether such a difference between the current mechanisms in each region induces the cross-field plasma transport.
In this study, we have derived the linear dispersion relation from the continuity equations,
$\displaystyle \frac{\partial n^{\text{P}}}{\partial t} + \mathbf{\nabla}_{\perp} \cdot (n^{\textrm{P}} \mathbf{v}_{s \perp}^{\textrm{P}}) + \frac{\Gamma_{s \parallel}^{\textrm{P}}(z = L_{z}) - \Gamma_{s \parallel}^{\textrm{B}}}{L_{z}} = 0 \; \; \; \; \; \; \; (1)$
and
$\displaystyle \frac{\partial n^{\textrm{R}}}{\partial t} + \mathbf{\nabla}_{\perp} \cdot (n^{\textrm{R}} \mathbf{v}_{s \perp}^{\textrm{R}}) + \frac{\Gamma_{s \parallel}^{\textrm{B}} - \Gamma_{s \parallel}^{\textrm{R}}(z = -h)}{L_{z}} = -\alpha \left[(n^{\textrm{R}})^{2}-(n_{0}^{\textrm{R}})^{2} \right], \; \; \; \; \; \; \; (2) $
and the charge conservation equations,
$\displaystyle \mathbf{\nabla}_{\perp} \cdot \mathbf{j}_{s \perp}^{\textrm{P}} + \frac{j_{\parallel}^{\textrm{P}}(z = L_{z}) - j_{\parallel}^{\textrm{B}}}{L_{z}} = 0 \; \; \; \; \; \; \; (3) $
and
$\displaystyle \mathbf{\nabla_{\perp}} \cdot \mathbf{j}_{s \perp}^{\textrm{R}} + \frac{j_{\parallel}^{\textrm{B}} - j_{\parallel}^{\textrm{R}}(z = -h)}{L_{z}} = 0, \; \; \; \; \; \; \; (4) $
in the upstream plasma and the recombination region in the simple configuration as shown in Fig. 1. In this configuration, the magnetic field is parallel to the $z$ axis and the $x$ and $y$ directions correspond to the direction perpendicular to the magnetic flux surface and the toroidal direction in torus devices, respectively. Here, $n$ is the plasma density, $\mathbf{v}_{s \perp}$ is the flow velocity perpendicular to the magnetic field, $\Gamma_{s \parallel}$ is the parallel flux, $\alpha$ is the recombination coefficient, $n_{0}$ is the plasma density at the equilibrium state, $\mathbf{j}_{\perp}$ and $\mathbf{j}_{\parallel}$ are the perpendicular and the parallel currents, the superscripts P, R, and B indicate the quantities in the upstream plasma, in the recombination region, and at the boundary between those regions, respectively, and the subscript $s$ represents the particle species. The upstream plasma flow velocity $\mathbf{v}_{s \perp}^{\text{P}}$ is composed of the $E \times B$, the polarization, the grad-$B$, and the diamagnetic drifts, while the recombination region flow velocity $\mathbf{v}_{s \perp}^{\text{R}}$ includes each drift with the Hall mobility and the motion in the direction of the perpendicular electric field with the Pedersen mobility. Linearizing Eqs. (1)-(4), as a result, we obtain the cubic equation regarding the frequency $\omega$ as the dispersion relation. It is found that the one mode of them has a positive growth rate under a certain condition. Figure 2 shows the dependence of the growth rate $\gamma$ of the unstable mode, i.e., the feedback instability mode, on the wave number $k$ and the propagation direction $\theta$. In Fig. 3 we present the dependence of the group velocity $v_{\textrm{g}} = \partial \omega / \partial k$ of the unstable mode on $k$ and $\theta$. Here, the typical parameters for fusion torus devices are assumed as follows: $B = 5 \; \textrm{T}$, $\partial B / \partial x = -1 \; \textrm{T}/\textrm{m}$, $n_{0}^{\textrm{P}} = 5 \times 10^{19} \; \textrm{m}^{-3}$, $\partial n_{0}^{\textrm{P}} / \partial x = -1.67 \times 10^{21} \; \textrm{m}^{-4}$, the initial electric fields $E_{x0}^{\textrm{P}} = E_{x0}^{\textrm{R}} = -100 \; \textrm{V}/\textrm{m}$, the electron and ion temperatures $T_{\textrm{e}}^{\textrm{P}} = T_{\textrm{i}}^{\textrm{P}} = 50 \; \textrm{eV}$, $T_{\textrm{e}}^{\textrm{R}} = T_{\textrm{i}}^{\textrm{R}} = 0.3 \; \textrm{eV}$, $\nu_{\textrm{rec}} / \Omega_{\textrm{ci}} = 10$, $L_{z} = 10 \; \textrm{m}$, $h = 0.3 \; \textrm{m}$, the ion-to-electron mass ratio $m_{\textrm{i}} / m_{\textrm{e}} = 3.67 \times 10^{3}$, and the ion-to-electron charge ratio $q_{\textrm{i}} / |q_{\textrm{e}}| = 1$. In those figures, the area inside the red curve designates the unstable region in which the feedback instability can be induced. Thus, Figs. 2 and 3 indicate that the waves $k \rho_{\textrm{s}}^{\textrm{P}} > 0.8$ and $ \theta \sim 3\pi/4$ can transport the plasma lump with the speed $\sim 0.002 \; c_{\textrm{s}}^{\textrm{P}}$. The simple estimation shows that the maximum of heat flux density is reduced to $1 - (v_{{\textrm{g}}x}/c_{\textrm{s}}^{\textrm{R}})(\tilde{n}/n_{0}^{\textrm{R}})(h/\lambda_{\textrm{q}}^{\textrm{B}}) \approx 82$% of the initial value and that the heat flux width is expanded to $1 + (h/\lambda_{\textrm{q}}^{\textrm{B}})(v_{\textrm{g}x}/c_{\textrm{s}}^{\textrm{R}}) \approx 280$% of the initial width if $\tilde{n}/n_{0}^{\textrm{R}} \sim 0.1$ and $\lambda_{\textrm{q}}^{\textrm{B}} \sim 3 \; \textrm{mm}$. Here, $v_{{\textrm{g}}x}$ is the $x$ component of $v_{\textrm{g}}$ and $\tilde{n}$ is the time averaged density of the transported plasma lump.Furthermore, to verify the feedback instability model, the spiraling plasma ejection observed around the recombination front under the detached divertor condition in the NAGDIS-II linear device experiment$^{3}$ is analyzed, in which the NAGDIS-II contributes to the establishments of the detachment and the cross-field transport mechanisms for future fusion reactors such as ITER and DEMO. The radial speed $v_{r} \sim 80 \; \textrm{m}/\textrm{s}$ at $r \sim 20 \; \textrm{mm}$ and azimuthal speed $v_{\theta} \sim 200 \; \textrm{m}/\textrm{s}$ at $r \sim 5 \; \textrm{mm}$ obtained in the experiment are in good agreement with $v_{{\textrm{g}}x} \sim 800 \; \textrm{m}/\textrm{s}$ and $v_{{\textrm{g}}y} \sim 200 \; \textrm{m}/\textrm{s}$ estimated by the theoretical model with the NAGDIS-II parameters if $v_{r}$ is reduced as $r$ increases.
$^{1}$Potzel S et al. 2013 J. Nucl. Mater. 438 S285.
$^{2}$Tanaka H. et al. 2010 Phys. Plasmas 17 102509.
$^{3}$Tanaka H. et al. 2018 Plasma Phys. Control. Fusion 60 075013.
Speaker: Hiroki Hasegawa (National Institute for Fusion Science) -
In the last few years it was demonstrated experimentally on tokamak ASDEX Upgrade that with a big amount of seeded radiant (nitrogen) the outer target fully detaches, and the high radiation zone appears in the confined region up to 10 cm above the X-point. The location of such a spot may be controlled in real time by variation of the impurity seeding rate [1]. Such regimes might be promising for the next generation tokamak-reactors, where it is necessary to reduce the power crossing the separatrix in order to keep divertor heat loads within applicable limits. Recently similar regimes were obtained in JET too both for nitrogen and neon [2,3].
Attempts to reproduce such a radiative spot in the vicinity of the X-point with the SOLPS-ITER code modeling were not successful [4]: with a prescribed discharge power and an impurity seeding rate increasing no steady state solution could be found. Core plasma went into the radiative collapse in the modeling as impurity neutrals started to penetrate into the confined region near the X-point independently of how big the discharge power was. However, in the presented simulations it is demonstrated that stable solutions may be obtained if the number of particles in the system is fixed (or controlled by the feedback loops) and the temperature is prescribed rather than discharge power (i.e. if the different modeling strategy is used). Such solutions are analyzed in the present report.
In the SOLPS-ITER solutions the electron temperature falls down below the 5 eV in a wide (up to 15 cm in diameter) region above the X-point in the confined region (see Fig. 1) with a wide range of impurity content and discharge power – the more the amount of impurity in the system, the more power is required. If this region is mapped to the outer midplane (the OMP), its width is below 1 cm.
It is shown that this width is of the same order as the decay length of total energy flux through the flux surfaces in the confined region. The latter may be estimated from the assumption that the anomalous cross-field heat flux is turned into a parallel heat flux, which sinks above the X-point due to impurity radiation. Thus this decay length is of the order of $\lambda_q$ - the SOL width for the standard discharge with attached divertor – since approach to its estimate is the same. In extreme regimes with big amount of impurity the perturbed region becomes so wide that almost no energy flux reaches the separatrix: all available power is radiated from the cold region above the X-point.
The maximal radiation comes from the poloidally narrow layer of $ L_{pol} \approx 1.5 – 2$ cm in width, where temperature falls down to 5 – 10 eV and where big amount of not fully striped impurity exists (see Fig 2). Thus the poloidal heat flux fully decays on the scale $L_{pol}$, and no more heat penetrates further downstream, and no radiation is possible from the X-point itself.
Important feature of the regimes with radiative X-point is that neither electron temperature nor electron density nor electrostatic potential (see Fig 3) are flux surface functions in the perturbed region. It is worth mentioning the formation of the electrostatic potential peak with $e\varphi \sim 25 T_e$. Thus the plasma flow in the pedestal region changes significantly with respect to the regimes with attached outer divertor, which leads to the changes in the radial E-field profile and might affect the level of the turbulence.[ 1] F.Reimold et al, Nucl. Fusion 55 (2015) 033004
[ 2] A R Field et al., Plasma Phys. Control. Fusion 59 (2017) 095003
[ 3] S. Glöggler et al, Nucl. Fusion 59 (2019) 126031
[ 4] I Yu Senichenkov et al, Plasma Phys. Control. Fusion 61 (2019) 045013Speaker: Ilya Senichenkov (Peter the Great Saint Petersburg Polytechnic University) -
Magnetic equilibrium modeling using the FIESTA code shows that steady-state snowflake (SF) divertor (1) configurations can be created and maintained with the existing poloidal field coil set in the MAST-U tokamak. A full multi-fluid plasma transport model with a computational grid encompassing two poloidal magnetic field nulls, with charge-state-resolved carbon impurities sputtered at material boundaries, and fast convective SF plasma transport (mixing) has been applied to the standard and SF divertor configurations using the UEDGE code. Results show that the SF divertor 1) reduces peak heat loads to divertor plates by spreading particles and heat fluxes to additional strike points; 2) the additional strike points receive up to 10% of heat and particle fluxes (w.r.t. the outer strike point), with weak dependence on two-dimensional profiles of particle diffusivities and heat conductivities as well as SF geometry details; 3) SF divertors approach the outer strike point detachment conditions at lower upstream density w.r.t. the standard divertor. Experiments are planned to validate the model, as part of a broad advanced divertor research program in MAST-U (2). The SF divertor uses two nearby 1st order nulls separated by distance $d_{xx}$, for a larger region of low poloidal field $B_p$ (PF) that modifies geometry and transport resulting in heat and particle flux sharing and reduction (1).
Snowflake magnetic equilibria created with a free-boundary Grad-Shafranov equilibrium code FIESTA using realistic currents $I ≤ 4 kA$ and existing MAST-U poloidal field coils show significantly increased near-separatrix connection length (cf. the standard divertor). A three-coil SF divertor algorithm previously developed in NSTX and DIII-D experiments (3) was used and improved with additional PF coils. Based on the theoretical SF criteria (1), families of up-down symmetric SF configurations were created (as, e.g., in Fig. 1) with $d_{xx} ≤ a (\lambda_q /a)^{1/3} = 0.15 m$, where $a=0.65 m$ – plasma minor radius, $\lambda_q$ ~ $8 mm$ – SOL heat flux width. The core kinetic profiles from the H-mode plasma model with modest shaping, $I_p=1 MA$ and $P_{NBI}=2.5 MW$, were used. Two main SF variants were investigated: the SF-plus configuration with a 2nd null in the private flux region (Fig. 1 (b)) and the SF-minus with a 2nd null in the low field side common flux region (Fig. 1 (c)). Flux tube connection length from outer midplane to outer strike points within 1-2 mm from the separatrix were up to 50% higher in both the SF-plus and SF-minus configurations (cf. standard divertor).
Two transport models in UEDGE were used over a broad range of core-boundary densities $2·10^{19} m^{-3}$ to $2·10^{20} m^{-3}$ . In the standard transport model, SOL ion diffusivity and heat conductivity profiles were taken to match the MAST-U SOLPS/EIRENE model (4). A second model emulated the theoretically predicted fast plasma mixing (the “churning” mode) in the two-null SF region (5). The churning mode convective zone radius was estimated to be $r* = 0.81 a(B_p a/R)^{1/3}$ ~ $10 cm$ for the SF divertor, and $r* = 0.44 B_pa^2/R < 1 cm$ for the standard divertor. The enhanced SF transport was modeled as Gaussian profiles with $r* = 10 cm$ centered at the nulls (Fig. 1 (d), (e)), with peak values $140 m^2/s$ (cf. $1 m^2/s-2 m^2/s$ in the standard transport model).
With the standard transport, peak heat fluxes at primary strike points SP1 and SP4 were 30-50% lower in the SF-plus and SF-minus configurations, cf. standard divertor. Heat flux profiles were considerably broader, due to diffusion into the secondary strike points and the increased connection length (Fig. 2 (a)). Primary strike point heat flux profiles in the SF-plus and SF-minus configurations with a small $d_{xx} = 2.5-3 cm$ were similar. Heat fluxes showed weak dependence on wall albedo $\alpha_w$ (hydrogen atom wall reflection). However, heat and particle fluxes to secondary strike points SP2 and SP3 were found to be rather sensitive to the albedo. With $\alpha_w=0.98$, additional strike points received heat and particle fluxes of about 5-10% of those in the outer strike point (Fig. 2 (c,d)), as compared to only 1-3% with $\alpha_w=1$.As the core-boundary (upstream) plasma density ne was increased, primary strike points SP1 and SP4 in the SF divertors reached low temperatures (~ 1 eV) and low peak heat fluxes at lower ne than the standard divertor (Fig. 2 (a)).
A complex interplay of SOL geometry effects, transport, and magnetic configurations was observed with the enhanced SF transport. The additional plasma mixing led to spreading of ions over a wider divertor region thereby reducing peak ion and neutral densities (Fig. 1 (d-e)). Heat flux profiles in all strike points were broadened (Fig. 2 (b-d)), peak heat flues were reduced. At higher core density ($10^{20} m^{-3}$), heat fluxes to additional strike points SP2 and SP3 were increased. With the enhanced SF transport (mixing), small differences between SF configurations with various inter-null distance $d_{xx}=3-12cm$ were observed. Primary strike points SP1 and SP4 reached detachment conditions at lower core-boundary density ne than with the standard transport assumptions.
In summary, significant progress in numerical modeling of SF divertor configurations was made in preparation for MAST-U experiments. The results show that heat and particle fluxes were spread among additional strike points and reduced as compared to the standard divertor. The MAST-U modeling results are consistent with numerous findings in previous SF divertor experiments conducted in TCV, NSTX, and DIII-D tokamaks (1,3).
This work is supported by the US DOE under Contract DE-AC52-07NA27344 and the RCUK Energy Programme Grant Number EP/P012450/1 and EURATOM.
(1) D.D. Ryutov, Phys. Plasmas 14 (2007) 064502; Phys. Plasmas 22, 110901 (2015).
(2) J. R. Harrison et al. Nucl. Fusion 59 (2019) 112011
(3) V.A. Soukhanovskii et al., Phys. Plasmas 19 (2012) 082504; Nucl. Fusion 58 (2018) 036018
(4) Moulton D. et al 2017 Proc. 44th EPS Conf. (Belfast, UK, 2017)
(5) D. D. Ryutov et al., 2014 Phys. Scr. 89 088002; D. D. Ryutov D. et al., PPCF 54 (2012) 124050
Speaker: Dr Aleksandr Khrabryi (LLNL) -
After the IAEA-FEC 2016 presentation (1) showing that the midplane-mapped heat-load width $\lambda_q^{XGC}$ predicted by the XGC1 edge-gyrokinetic-code (2) for the full-current ITER is $\stackrel{>}{\sim}6\times$ wider than the experimental-data based formula $\lambda_q^{Eich}$ (PRL 2011) while the same code reproduces $\lambda_q^{XGC}\simeq\lambda_q^{Eich}$ for the present experiments, a major effort has been devoted to resolving this puzzle: Several new and higher-resolution extreme-scale predictive simulations and indepth physics studies have recently been performed. A machine learning program is then used to discover hidden kinetic parameters and produce a new predictive formula encompassing the Eich formula, the full-current ITER result and other anomalous XGC prediction results. It is found that the hidden parameters are associated with the low neoclassical $E\times B$ flow in the 15MA ITER edge (3,1) and the arousal of the trapped electron mode (TEM) turbulence across the magnetic separatrix.
For a full-current ITER H-mode operation at $I_p=15$MA with $q_{95}=3$, we have $\lambda_q^{Eich}(=0.63B_{pol,MP}^{-1.19}) \stackrel{<}{\sim}1mm,$ where $B_{pol,MP}$ is the poloidal magnetic field at outboard midplane separatrix. At this range of $\lambda_q$, which is about a factor of 5 narrower than the design value, the divertor operation must be at least in semi-detached conditions where the plasma pedestal density can easily reach the Greenwald density limit and the disruption-free operation becomes difficult.However, it is questionable if such a simple extrapolation from the present experimental data is valid since the 15MA ITER edge may obey fundamentally different physics, especially when the eddy shearing effect on the microturbulence by neoclassical ExB flow is expected to be much weaker due to negligible ion banana width compared to plasma size (3,1). As can be seen from Fig. 1, all $\lambda_q^{XGC}$ predictions on three US tokamaks (NSTX, DIII-D and C-Mod) agree with the Eich formula ($\lambda_q^{Eich}$, solid line) within experimental error bar (dashed lines). However, when the same code is applied to the 15MA ITER edge plasma, the XGC-predicted $\lambda_q^{XGC}$ was $\stackrel{>}{\sim} 6\lambda_q^{Eich}$. This anomalous result has been left as an unsolved puzzle since IAEA-FEC-2016 that required a major undertaking including new extreme-scale simulations and in-depth physics studies.
Several critical and higher resolution XGC1 simulations have recently been performed, which include the first H-mode scenario plasma in ITER with $I_p=5$MA. This plasma has $B_{pol,MP}=0.43T$ that is similar to a high field DIII-D plasma ($\ast$ mark in Fig.1), while the plasma size is the same as for 15MA ITER. We find $\lambda_q^{XGC}\simeq 2.2mm$, similar to $\lambda_q^{Eich}=1.7 mm$. This result denies the conjecture from some physicists that the pure size effect could be responsible for the large $\lambda_q^{XGC}$ in 15MA ITER. Comparison of the turbulence pattern between the 5MA and 15MA ITER cases reveals an interesting physics difference: In the 5MA case, turbulence patterns across separatrix are isolated and “blobby,” as in all the present tokamaks that obey the Eich scaling, while the patterns become a radially connected “streamer” type in the 15MA case (Fig.2).
In a JET 4.5MA plasma that is the closest experimental plasma to the 15MA ITER in the sense of both size (half the linear size of ITER) and the poloidal magnetic field $B_{pol,MP}$ (= 0.89T, compared to 1.21T), XGC1 finds $\lambda_q^{XGC}=0.64mm$, which is again within the experimental error bar from $\lambda_q({Eich}=0.72mm$ (see Fig.1). Also, the turbulence across the separatrix remaines blobby. The compact tokamak NSTX-U is under construction to raise the plasma current to 2MA, in which the edge electrons become weakly collisional. In this plasma, XGC1 finds that $\lambda_q^{XGC}\sim 2\lambda_q^{Eich}$ (Fig.1) with the edge turbulence becoming streamer type. However, in a 1.5MA NSTX-U model plasma, XGC finds $\lambda_q^{XGC}\sim \lambda_q^{Eich}$ (Fig.1). One noticeable difference in the dimensionless parameter of 1.5MA plasma from 2MA is that the plasma is collisional at separatrix. These NSTX-U results add to the puzzle that needs to be resolved together with the full-current ITER case.More influentially to this study, experiments at C-Mod raised $B_{pol,MP}$ values near and somewhat above the 15MA ITER value (4), and found that $\lambda_q$ still follows $\lambda_q^{Eich}$ (see two example experimental results in green stars in Fig.1). We have simulated one of theses high current discharges that has a similar $B_{\rm pol,MP}$ value (1.11T) to that of ITER (1.21T). We find $\lambda_q^{Eich}=0.4mm$ that is even somewhat smaller than $\lambda_q^{Eich}= 0.56mm$. As a result, the XGC1’s solution becomes double valued around the full-current ITER $B_{\rm pol,MP}$ if $B_{\rm pol,MP}$ is used as the sole parameter, and suggests hidden parameters that were missing in the regression study by Eich et al.
In an effort to find the hidden parameters and a scaling formula that will resolve the double-valuedness and encompass all the predictive simulation results including the 15MA ITER, the 2MA NSTX-U, the $\lambda_q^{Eich}$ formula, and the experimentally measured $\lambda_q^{Exp}$ values corresponding to the simulation points, we utilized the AI-based modeling engine Eureqa (5). Eureqa works by creating random equations from the input data through a technique called “evolutionary search” and suggest equations that fit the data with various accuracy.
Among many Eureqa-produced formulas, the simplest and physically meaningful formula is identified as
$\lambda_q^{ML}=0.63B_{pol,MP}^{-1.19} [1.0 + 1.08\times 10^{-10} (B_{pol,MP}\; a/\rho_{i,pol} \xi)^4]\;\;\; (A)$
with RMS error 0.19, where $\xi=1.0+\alpha \; \Theta [(a/R)^{1/2}/\nu_{e*}- \beta]$ with $\alpha=2.3, 1.7<\beta<1.9$, and $\Theta$ the heavyside step function. In this formula, there is a long distance between the C-Mod and the 15MA ITER $\lambda_q$ points, well-resolving the double-valuedness (Fig.3). A more definitive $\beta$ number or a smooth function for $\Theta$ could not be determined due the limited number of the off-$\lambda_q^{Eich}$ points. The form of the function $\xi$, a highly simplified criterion for collisionless TEM instability, has been manually instructed from the XGC observation that whenenver we have an enhanced $\lambda_q^{XGC}$, the edge turbulence across sepatrix becomes streamer-type TEM turbulences (Fig.2). Convincing XGC1 evidences for the TEM turbulence, instead of blobs, across separatrix in 15MA ITER plasma exist: For example, 1) a linear analysis without $E\times B$ shearing shows unstable TEMs just inside the separatrix, 2) a cross-correlation study of XGC1 data finds a distinctive non-adiabatic electron behavior, and 3) an unsupervised machine learning study shows a high-degree of correlation beteween trapped electrons and turbulence. The formula (A) is physically meaningful because the parameter $B_{pol,MP}\; a/\rho_{i,pol}$ is related to the ion banana width and the neoclassical $E\times B$ shearing rate ($\propto \rho_{i,pol}/a$). Both of them have sensitive influence on the turbulence structure. Also the electron collision frequency in all the existing tokamak edge plasmas that obey the $\lambda_q^{Eich}$ is not low enough to provoke the collisionless TEM turbulence.More number of predictive simulations (including the on-going study for 12.5MA ITER, to be presented at IAEA-FEC2020) and possible experimental data in the gap region could improve the accuracy of Eq.(A) (especially, of $\beta$ and $\Theta$). A future improved formula is, however, not expected to be much different from Eq. (A), which describes all the $\lambda_q^{XGC}$ results obtained so far, encompassing the full-current ITER, and is consistent with the kinetic physics discoveries and understandings.
This study is mostly funded by US DOE, with computing resources provided by OLCF.
(1) Chang C.S., Ku S., Loarte A. et al., Nucl. Fusion 57, 116023 (2017)
(2) Ku S., Chang C.S. et al., Phys. Plasmas 25, 056107 (2018)
(3) D.R. Hatch, M. Kotschenreuther, Nucl. Fusion 57, 036020 (2017)
(4) Brunner D., LaBombard B. et al., Nucl. Fusion 58, 094002 (2018)
(5) Product of Nutonian, now acquired by DataRobot.Speaker: Choongseok Chang (Princeton Plasma Physics Laboratory and KAIST) -
Control and/or mitigation of runaway electrons (REs) is necessary for the operation of larger fusion devices including ITER 1. The disruption generated REs, in particular, pose a serious threat for the plasma facing components in ITER as they are predicted to acquire very energy (~ tens of MeV). Many RE mitigation mechanisms like Massive Gas Injection (MGI) 2 and Resonant Magnetic Perturbations (RMPs) 3 are proposed and demonstrated in various tokamaks like DIII-D, JET, Tore Supra for mitigating the REs. However, the success of each method is limited and the search is on for a robust mechanism of RE mitigation. Due to lower values of plasma current, smaller tokamaks generally do not observe disruption generated REs, which limits the exploration of mitigation techniques. In order to confine the REs after the end of the discharge, i.e., the termination of plasma current, a novel method is attempted in a mid-size tokamak ADITYA-U. In this paper experimental observations of confinement of runaway electrons using an electrostatic field induced by biasing an electrode placed in the edge of ADITYA-U tokamak is presented. The REs are kept confined even after the termination of discharge current, which can very well be utilized for testing several RE mitigation techniques in smaller tokamaks where the disruption does not generate substantial RE population.
The experiment is performed in ADITYA-U tokamak, a medium sized, air-core tokamak with
R = 0.75 manda=0.25 m. The experiment is carried out withB_T = 1.0 - 1.3 T. The typical value of plasma current is I_p=80-160 kA and plasma duration is~ 80-330 ms. The line averaged central density is n_e ~ 1-4 10^{19} m^{-3}. A radial electric field can be induced inside the plasma by biasing an electrode placed inside the last closed flux surface (LCFS) [8]. In ADITYA-U tokamak, the electrode bias setup initially consisted of a double bellow assembly capable of translating a molybdenum electrode of 5 mm diameter inside the LCFS as well as varying its exposed length. It was later changed to a single bellow assembly containing an 8 mm tungsten electrode having an exposed length of 14 mm which can be translated inside the LCFS. A 450 V, 60 mF capacitor based biasing power supply is developed which can bias the electrode in multiple pulses employing an Insulated Gate Bipolar Transistor (IGBT) as a switch. For the purpose of this experiment, the electrode is placed 2.5 cm inside the LCFS and it is biased just before the disruption of the plasma. The typical waveform of voltage and bias current is shown in Fig 1. The REs are routinely observed in ADITYA-U tokamak for low density discharges. They are diagnosed with a 3 inch NaI (Tl) lead shielded scintillator detector coupled with a photo-multiplier tube (PMT), which is collimated to observe the Hard X-rays (HXR) produced when the REs interact with the limiter.
Figure 1 shows the multiple pulses of biasing applied to the electrode placed 2.5 cm inside the LCFS for the shot 33336. A biasing voltage of ~ 80 V draws typically ~ 20-30 A bias current depending on the plasma properties. The last pulse of biasing coincides with the plasma disruption as shown in figure 1. The plasma parameters for this shot are shown in figure 2. Figure 2a) represents the plasma current and loop voltage, 2b) shows the Hα signal, 2c) indicates the visible continuum, 2d) shows the HXR signal and plot 2e) shows the HXR flux. It can be seen clearly that in presence of biasing during the disruption, the HXR intensity as well as HXR flux persists for ~10 ms even after the termination of the plasma discharge as opposed to the case with no biasing. The visible continuum signal also persists after the plasma disruption. The zoomed picture is shown in figure 3. The observation has been repeated for various values of electrode bias voltages and currents.To obtain good insight into this new high confinement RE mode found in ADITYA-U Tokamak, a 2D, free boundary-like equilibrium is constructed for ADITYA-U tokamak profiles and parameters, which self-consistently solves a plasma plus an electron beam equilibrium using a newly developed mesh-free Grad-Shafranov solver [4]. Numerical results thus obtained support the idea that RE beam can attain a quasi-static equilibrium for a wide range of IRE/Ip, qualitatively supporting the experimental findings.
The disruption generated REs are not observed in ADITYA-U tokamak due to low plasma currents (< 200 kA) and stored energy (~ 2 kJ). The REs thus produced during the plasma discharge are generally lost as soon as plasma disrupts. Hence, the new observation suggests that the REs are confined with the help of electrode biasing even after the end of plasma discharge. The confinement of REs seems to be the result of the E_r×B_φ motion induced by the radial electric field. This result can be exploited to confine the REs for sufficiently longer times in smaller tokamaks for testing innovative concepts along with conventional mechanisms of MGI or RMPs for their effective mitigation.
References:
1 Boozer, Phys. Of Plasmas, 22, 032504, (2015)
2 Whyte et al, Phys. Rev. Lett., 89, 5, (2002)
3 Lehnan et al, Phys. Rev. Lett., 100, 255003, (2008)
[4] Dinesh Nath et al, “A new mesh-free Grad-Shafranov solver for Beam-Plasma Tokamaks'', Manuscript Under Preparation (Feb 2020)Speaker: Tanmay Macwan (Institute for Plasma Research) -
The high-power reconnection heating of merging spherical tokamak (ST) plasmas has been developed by TS-3U, TS-4U$[1]$, UTST, MAST$[2]$ and ST-40 experiments and PIC simulations$[3]$, leading us to direct access to burning and high-beta ST often with absolute minimum-B without using any additional heating like neutral beam injection (NBI). All of them confirmed, (i) the promising scaling of ion heating energy increasing with square of reconnecting magnetic field $B_{rec}$ ~ poloidal magnetic field $B_p$. (ii) ST-40 recently extended the maximum ion temperature $T_i$ up to 2.3keV. TS-3U, ST-40 and MAST experiments found that (iii) the reconnection heating forms interesting high-beta ST plasmas with hollow $T_i$ profile and often with absolute minimum-B (TS-3U) and that (iv) the toroidal field $B_t$ does not affect ion heating energy as high as 40-50% of poloidal magnetic energy for $q$>1 and also that (v) $B_t$ does not affect the global hollow $T_i$ profiles of the produced new ST plasma but its local $T_e$ profiles, forming the more peaked $T_e$ profile in the higher $q$ regime.
As shown in Figs. 1(a)(b), we axially merge two STs or spheromaks to form a new high-beta ST or the FRC which is transformed into a high-beta ST. In Fig. 1(d), all of TS-3U (ST, FRC: $R$=0.2m), TS-4U (ST, FRC: $R$=0.5m), UTST (ST: $R$=0.45m), MAST (ST: $R$=0.9m) and ST-40 (ST: $R$=0.5m) agree that the ion temperature increment $\Delta T_i$ increases with $B_{rec}^2$ ~$B_p^2$ up to 2.3keV under constant electron density $n_e$ ~ $1.5 \times 10^{19}$ $m^{-3}$, provided that the current sheets are compressed to the order of ion gyroradius. It is because the ion heating does not depend on plasma size but on the reconnection outflow whose speed is the order of poloidal Alfven speed $V_{Alf}$ = $B_{rec}$ / $\sqrt {\mu_0 m_i n_i}$ ). As shown in Fig. 1(c), this heating scheme with no temperature dependence enables us to realize the burning ion temperature (red line) over 5-10keV without using any additional heating, in sharp contrast with the conventional ohmic heating scenario with additional heating (blue line). TS-3U and MRX found that the reconnection converts about half of poloidal magnetic energy of the merging STs mainly to ion thermal energy of the new ST. This heating enables us to increase $T_i$ to the optimized value in Lawson diagram of Fig. 1(c), keeping $n \tau$ roughly constant or larger.![Radial profiles of (a) ion temperature $T_i$, (b) radial velocity $V_r$, (c) electron density $n_e$ and (d) absolute value of magnetic field $|B|$, during ($t$=$30\mu s$) and after ($t$=$40\mu s$) ST merging in TS-3U. Radial profiles of (e) $q$ value and toroidal current density $j_t$ in TS-3U, (f) ion thermal pressure $U_i$ for $B_t$=0.4T, 0.8T in MAST$[2]$, (g) ion temperature $T_i$ in ST-40, (h) R-Z contour of $T_i$ in PIC simulation after the new STs are produced by merging STs.](https://workshop.nifs.ac.jp/fec2020/image/20-Ono-image-fig_4.jpg)
Then the next important question arises as to what type of ST plasma this reconnection heating produces. Figures 2(a)-(d) show radial profiles of $T_i$, radial velocity $V_r$ measured by Doppler probe, $n_e$ by electrostatic probe and absolute value of magnetic field $|B|$ by magnetic probe at $t$=30$\mu$s (during reconnection) and at $t$=40$\mu$s (after merging) when we merge two ST plasmas with $B_t$/$B_p$~5 in TS-3U. We can confirm that the bi-directional reconnection outflow $V_r$~40km/s almost equal to 70% of poloidal Alfven speed dumps at the two “shock-like” downstream positions ($R$=0.14 and 0.21 m) where all of $T_i$, $n_e$ and $|B|$ peak. Similar hollow profiles of $T_i$ were observed both in MAST and ST-40, indicating that the common ion heating mechanism validates the $B_{rec}^2$-scaling in those merging experiments in Japan and UK. It is noted that the hollow profiles of $T_i$ and $|B|$ are maintained in the new STs after the merging is over. Especially, $|B|$ is minimized at around magnetic axis, indicating absolute minimum-B formation.
In present merging experiments, only TS-3U and TS-4U can measure internal magnetic field profile of absolute minimum-B configuration but MAST and ST-40 can measure similar double-peaked or hollow profiles of $T_i$ and ion pressure $U_i$. Figures 2(e)-(g) show radial profiles of $q$-value and toroidal current density $j_t$ in TS-3U and $U_i$ in MAST and $T_i$ in ST-40 measured by Doppler spectroscopy. The produced new high-beta ST has the double peaked $U_i$ or $T_i$, indicating its hollow thermal pressure profile of high-beta ST.
The recent PIC simulation of merging ST and spheromaks also confirmed the $B_{rec}^2$-scaling of reconnection (ion) heating and revealed that the 2D contour of $T_i$ has ring-type high $T_i$ region just when the merging is completed, as shown in Fig. 2(h). This fact indicates the double peaked $T_i$ profiles in Figs. 2(a)(f)(g) are due to the ring-type high $T_i$ area. All of those experiments and PIC simulation agree that the merging STs are transformed to the new high-beta ST with hollow and/or double peaked pressure profile. The reconnection proceeds from the outer flux surface to the core flux and the outflow speed of reconnection peaks in the middle of merging. It is probably the reason why $T_i$ becomes hollow in the produced ST configuration.
This hollow pressure profile is found to be connected with the absolute minimum-B configuration. Figure 3 shows 2D contours of $|B|$ (and poloidal flux) of STs for four different $B_t$, which are formed by type (b) merging in Fig. 1 together with that of low-beta ST without reconnection heating. While an FRC also has absolute minimum-B areas both at magnetic axis and at geometric axis, the high-beta ST has wider absolute minimum-B area at around magnetic axis due to strong $B_t$ around small $R$ area.
The series of merging experiments and PIC simulations confirmed the $B_{rec}^2$-scaling of reconnection heating up to 2.3keV and formation of the double-peaked or hollow $T_i$ and thermal pressure profiles sometimes with absolute minimum-B profile after the merging is over. Our Balloo code analysis indicates the absolute minimum-B is connected with the second-stable ST for ballooning instability. The reconnection heating is unique, cost-effective and useful to heat fusion plasmas directly to the alpha-heating region, to the optimized $T_i$ area in Lawson diagram without using any additional heating.$[1]$ Y. Ono et al., Nucl. Fusion 59, (2019), 076025.
$[2]$ H. Tanabe et al., Nucl. Fus. 57, (2017), 056037.
$[3]$ R. Horiuchi et al., Phys. Plasma 26, (2019), 092101.Speaker: Yasushi Ono (University of Tokyo) -
Accurate modelling of cross-field turbulent transport in tokamak’s edge plasma remains a challenge, many key experimental features such as edge transport barriers formation being still hard to simulate, especially for ITER size tokamaks. Being able to predict the SOL width or the power load imbalance between inner and outer divertor legs even for today’s JET size tokamaks is still an open issue. First principle modelling of edge plasma turbulence is thus today a very active topic in the fusion community driving many dedicated research projects such as the TSVV “European boundary code” project in Europe.
Despite this effort towards first principle modelling of the turbulent transport, 2D transport codes where the cross-field turbulent transport is emulated by ad-hoc diffusion remain today the most popular tool for experiments analysis or engineering applications. In these codes, the prescription of the diffusion coefficients rely on empirical basis and are usually adjusted to match simulation and experimental at a given location (usually mid-plane profiles). In this empirical procedure, the nature of the turbulence behind these transport coefficients is most of the time not taken into account. The main drawback of this approach is to lack predictability by missing the basic physics of turbulence mechanisms.
In order to address both the first principle modelling of edge turbulence but also to feed information about turbulence into transport codes, a dedicated effort has been made at IRFM in the last two years to develop the new code SOLEDGE3X which aims at encompassing a hierarchy of models from standard 2D transport code to 3D first principle turbulence modelling. The development of this new code comes from the merging of the transport code SOLEDGE2D [1] and the turbulence code TOKAM3X [2] previously developed within the French fusion community. Integrating features of the two latter codes, SOLEDGE3X is able to simulate tokamak edge plasma either in 2D or 3D, including: the realistic wall geometry, neutrals since it is coupled to EIRENE and impurities since the code is fully multi-species, based on the state-of-the-art Zhdanov collisional closure for multi-component plasmas. First results of turbulent plasma in WEST geometry including the sputtering of the wall have been produced in 2019 and represent a major step for first principle modelling of tokamak plasma in realistic conditions.
However, since full 3D simulations remain expensive in CPU time, it is still profitable to use SOLEDGE3X as a transport code in 2D to run fast interpretative simulations. An original idea implemented in SOLEDGE3X to improve turbulent transport predictability and go beyond the standard empirical approach is to use a reduced model for turbulence in the same fashion as the “k-epsilon” widely used in neutral fluid community. One or two equations are added to the standard mass, momentum and energy balance to describe the evolution of the turbulence intensity “k” and optionally the evolution of the turbulence dissipation “ε”. This model is used as a platform to include some ingredients of the turbulence physics - such as the interchange instability – in the framework of 2D transport codes. Of course, this reduced model is not first principle and require an empirical closure. We use the multi-machine scaling law for the SOL width $λ_q$ to do so, as a first approach of data assimilation. When SOLEDGE3X is used as a transport code with this reduced “k-epsilon” model for turbulence activated, the number of free parameters is drastically reduced since there is no need to prescribe transport coefficients, the “k-epsilon” model predicting a 2D map of cross-field diffusivities. Since the model is based on interchange instability, one recovers for instance the ballooning of radial transport in the low field side mid-plane. Also, since the model is closed using the multi-machine scaling law, the overall level of transport is automatically adjusted to get a SOL width compatible with this scaling law. This closure with the scaling law is thus powerful to obtain reasonable profiles, however it is also a weakness of the model since in principle, one could not reproduce an experiment where the SOL width does not follow the scaling law. This is the drawback of this kind of semi-empirical models which are not first principle and thus limited by the main assumptions behind their closure. In order to test the applicability of this reduced model for turbulence, a series of L-mode TCV shots have been simulated and the simulation results have been compared with experimental data [Figure 1]. Even if the simulation does not recover exactly the SOL width measured in the experiment, the overall agreement in term of peak heat flux, density and temperature on the target is quite remarkable. This “k-epsilon” model has shown to be a promising first step toward integration of turbulence physics inside transport simulations.
To go further, the advantage of the SOLEDGE3X code is to be able to run also first principle turbulent simulations where the turbulence intensity can be directly “measured”. The comparison between turbulent simulations and the reduced model for turbulence should in the future provide a clear path to improve these reduced models for turbulence. In that perspective, first results of full turbulent simulation in the same TCV geometry [figure 2] should help interpreting the results obtained with the reduced “k-epsilon” model and identify missing ingredients.
[1] H. Bufferand et al., Nucl. Fusion 55 (2015), 053025
[2] P. Tamain et al., J. Comp. Phys. 321 (2016), 606-623Acknowledgements:
This work was granted access to the HPC resources of CINES, under the allocation 2018-A0040510482 made by GENCI and to the HPC resources of Aix-Marseille University financed by the project Equip@Meso (ANR-10-EQPX-29-01). This work was supported by the EoCoE-II project, grant agreement 824158, funded within the EU’s H2020 program. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European CommissionSpeaker: Dr Hugo Bufferand (CEA) -
Fundamental mechanisms governing the prompt redeposition of tungsten impurities sputtered in tokamak divertors have been identified and analyzed to enable quantitative estimations and in-situ monitoring of the net erosion and lifetime of tungsten divertor plasma-facing components in ITER [1]. The net erosion of tungsten divertor PFCs is primarily determined by the prompt redeposition of tungsten impurities, which is very large in the high-density partially attached divertor plasma conditions expected in ITER operational divertor scenarios. Near the divertor targets, the width of the electric sheath is of the order of several main ion Larmor radii due to the magnetic field lines intersecting the material surfaces at grazing incidence, and a vast majority of tungsten impurities sputtered from divertor PFCs are thus ionized multiple times within the sheath region. The complex interplay between the successive ionizations of tungsten impurities sputtered from divertor PFCs and the sheath electric field is shown to ultimately govern the prompt redeposition of tungsten in the divertor region (fig. 1).
Consequently, the fraction of sputtered tungsten impurities which do not promptly redeposit and contribute to the net erosion of tungsten PFCs mainly depends on the ratio of the vertical ionization mean-free path of tungsten neutrals over the sheath width. A new scaling law quantifying the prompt redeposition of high-Z impurities in divertor was obtained (fig. 1). This new scaling law significantly differs from the scaling law conventionally used for low-Z impurities [2], since effects of the sheath electric field are much stronger on trajectories of high-Z impurities than on trajectories of low-Z impurities, due to the larger mass of high-Z impurities. Following from this new scaling law, tungsten prompt redeposition is shown to be predominantly governed by the tungsten ionization rates and the width of the sheath.
Furthermore, the net erosion of tungsten divertor PFCs can only be directly monitored through spectroscopic measurements of the emission of tungsten impurities in charge states 4+, 5+ or higher near the divertor targets [1]. The fraction of tungsten impurities which do not promptly redeposit corresponds to the fraction of tungsten impurities which are ionized outside of the sheath region into the higher charge states 4+ and 5+, as a consequence of the decrease of the tungsten ionization rate when the charge state increases. In contrast, the fraction of sputtered tungsten impurities ionized into the charge state 1+, 2+ and 3+ varies moderately with the divertor plasma conditions considered here and is therefore weakly correlated to the fraction of promptly redeposited tungsten impurities.
Besides, S/XB coefficients (number of ionization events per photon) for high‐Z impurities like tungsten are significantly reduced in high‐density divertor plasma conditions (fig. 2) when sputtered impurities are ionized within the sheath region, where electrons are repelled by the sheath electric field. The modification of the electron distribution in the sheath region must be taken into account to accurately estimate the ionization and emission rates and derive the S/XB coefficients used to determine the flux of impurities from divertor PFCs [3].
Reliable predictive modeling of the net erosion of tungsten divertor PFCs requires first- principles estimations of the critical physics parameters (tungsten ionization rates and sheath width) controlling tungsten prompt redeposition, and an experimental assessment of the validity of those parameters. New ground, metastable, and some excited state ionization rate coefficients for tungsten have been calculated using non-perturbative methods, with a range of possible scaling results that allow data for all of the excited states to be evaluated. New excitation and ionization data for neutral tungsten shows that the majority of the ionization at sheath temperatures and density conditions is due to excited state ionization. High wavelength resolution observations of neutral tungsten spectral line in the UV range in the DIII-D divertor support the benchmarking of the new dataset.
In parallel, dedicated experiments have been conducted in the DIII-D divertor to assess the modeling of tungsten prompt redeposition in divertor. The ratio of the net erosion rate of tungsten samples of different sizes exposed to the same attached divertor plasma conditions is only function of the fraction of tungsten promptly redeposited on each sample when the size of the tungsten samples is comparable to the characteristic distance of the prompt redeposition of tungsten (several mms). Ratios of the net erosion rate of millimeters size tungsten samples experimentally measured in the DIII-D divertor are well reproduced with a reduced model of tungsten prompt redeposition for a variety of attached divertor plasma conditions [4], and thus provide a robust benchmark for predictive models of tungsten prompt redeposition and net erosion in ITER.
Similarly, an experimental assessment of the sheath width has been conducted in DIII-D by measuring the angle of incidence of main plasma ions impinging on divertor targets, which is a function of the sheath width, from the erosion of spherical micrometers dusts (fig. 3) and from the impurity deposition pattern in micro-engineered trenches. Resulting estimations of the sheath width are in agreement with kinetic simulations of the electric sheath in the divertor.
This work was supported in part by the US Department of Energy under DE-SC0018423, DE-FG02-95ER54309, DE-FC02-04ER54698, DE-SC0019308 and DE-AC02-09CH11466.
[1] J. Guterl et al. “Predictive modeling of tungsten net erosion in divertor” in preparation
[2] G. Fussman et al. 1995 Proc. 15th IAEA Int. Conf. (Seville) vol 2 p.143 (1995)
[3] J. Guterl, et al. Contributions to Plasma Physics (2019)
[4] J. Guterl, et al. Plasma Physics and Controlled Fusion 61.12 (2019)Speaker: Dr Jerome Guterl (General Atomics) -
A mini-reactor called CANDY that based on kJ-class diode-pumped solid-state laser (DPSSL) is proposed to perform feasibility studies of the power plant in fast ignition scheme fusion. In order to implement CANDY, we have adressed two issues. First is a repetitive operation of pellet injection and laser illumination, second is target physics related to inertial confinement fusion plasma. First is achieved using repetitive laser system, (i) succeeded in 10 Hz injection of CD beads longer than 2 minutes, which was demonstrated for the first time using inserter that works at the same frequency of laser toward the reactor. Second issue is achieved using a world-class ultra-intense laser LFEX at ILE, Osaka University, here we have demonstrated (ii) an efficient heating of a counter imploded core of a density of 2.8±0.3 g/cc with temperature upto 1 keV with efficiency evaluated to be 9±0.8%. These results are encouraging and gives hints on the fast ignition scheme as a candidate toward future compact reactor developments.
A laser-driven inertial fusion energy (IFE) reactor achieves the fusion of injected fuel pellets, which are continuously delivered into the reaction chamber and engaged by laser beams at 10’s Hz. The target physics developments are expected to obtain fusion gain and developments of pellet injection and engagement are indispensable to realize laser fusion reactor. The Graduate School for the Creation of New Photonics Industries (GPI) have conducted laser inertial fusion program since 2008 with corroboration with industries and academia. Our strategy is to demonstrate integrated repetitive operation that scalable for future laser fusion power plant. We proposed a mini-reactor CANDY (1) that utilizes a few kJ/10 Hz laser driver with ~100 W fusion power (figure 1.). Toward its development, three key issues are conducted; (1) Repetitive laser development (2, 3), (2) Target physics related on counter illuminating fast ignition scheme (4, 5) and (3) Pellet injection & engagement with fusion reactions (6-8). Figure 2 represents specifications of achieve values in Hamamatsu, designed values of laser fusion mini-Reactor CANDY (1) and inertial fusion Test Reactor LIFT (Phase-IIII) (9). Here we report the present status of R&D toward the CANDY.
Repetitive laser development
In the Hamamatsu area, several repetitive laser systems those utilizing DPSSL are under operation. The first high repetitive inertial confinement fusion experiment laser system called HAMA (2) is under operation since 2008. The HAMA consists of a Ti : sapphire laser pumped by a 10J/1-10 Hz DPSSL. The HAMA utilizes counter fuel assembling beam with pulse tailoring and counter heating beam since 2013. In 2018, Hamamatsu Photonics have issued a press release of a 100-Joule class DPSSL based facility “TERU” (Trek on fusion Energy Roadmap toward Ultimate SDGs). That is under operation at 50 J/0.5 Hz with potential upto 10 Hz (3). Toward the fabrication of kJ-class DPPSL required for the CANDY, repetitive laser research and developments are on-going in Hamamatsu.Target physics related counter illuminating fast ignition scheme
We have demonstrated an efficient imploded core heating using a using a world-class ultra-intense laser LFEX at ILE, Osaka University by improving imploding laser beams from the previous experiments (4). Six green beams from GEKKO XII(GXII) laser (0.53 μm, 1.1 ns, 1.7 kJ) haveimploded a CD shell target to a density of 2.4 x solid density, that forms a 2.8±0.3 g/cc core. When two beams from LFEX laser (1.053 μm, 1.5 ps, 0.25 kJ) were axially illuminating to GXII symmetric axis, the bulk ion temperature increases beyond 1.0 keV from 0.6-0.7 keV evaluated using a multi-channel neutron spectrometer and the inferred heating efficiency was 9±0.8%. This result indicates that, as far as core density stays around a few g/cc, direct illuminating fast ignition scheme can expect a degree of core heating while the imploded core was surrounded by ablating corona plasma.Pellet injection & engagement
We have demonstrated 10 Hz operatin of bead pellete injection and engagemnt. The injection machine is upgraded from the previous 1 Hz injection system by increasing number of holes on disk rotor from 20 to 200 (6-8). As the results, we have achieved 10 Hz pellet injection and laser engagement that last beyond 2 minutes. The repetition rate is now upgraded to 10 Hz, the same frequency with laser repetition. Figure 3 represents experiments history of hit ratio per second for bead pellet injection & engagement system. The hit ratio beyonds 40% resulting in 5 times incremental of laser hit ratio per second from the previous results which indicates the first demonstration that the pellet injection frequency equals to that of laser amplified frequency, an important step to increase fusion power in the future power plant.
In conclusion, toward the laser fusion mini-reactor CANDY, we have made advances both on (i) target phsyiscs that the imploded core with density 2.8±0.3 g/cc are heated upto 1keV with heating coupling efficiency 9±0.8% using the ultra-intense laser LFEX at ILE, Osaka University, and (ii) pellet injection & engagement demonstrating a 10 Hz beads injection and laser illumination beyond 2 minutes with hit ratio beyond 40%.
(1) Y. Kitagawa et al., J. Phys. Conf. ser. 688, 012049 (2016).
(2) Y. Mori et al., Nucl. Fusion vol. 53, 073011 (2013).
(3) T. Watari et al., Proceedings of IAEA FEC2018 conference, IFE/P4-10 (2018).
(4) Y. Kitagawa et al., Phys. Rev. Lett. vol. 114, 195002 (2015), Nucl. Fusion vol. 57, 076030 (2017).
(5) Y. Mori et al., Phys. Rev. Lett. vol. 117, 005001 (2016), ibid. vol. 57, 116031 (2017).
(6) O. Komeda et al., Sci. Reports vol. 3, 2561 (2013).
(7) Y. Mori et al., Fusion Sci. & Technol. vol. 75, 36 (2019).
(8) Y. Mori et al., Nucl. Fusion vol. 59, 096022 (2019).
(9) T. Norimatsu et al., Nucl. Fusion vol. 57, 116040 (2017).Speaker: Yoshitaka Mori (The Graduate School for the Creation of New Photonics Industries) -
Scoping the possible operational regimes of a DEMO reactor requires reliable models of power exhaust processes. The scrape-off layer (SOL) and divertor in DEMO will need a higher radiated power fraction than present-day devices, and strong asymmetries in the SOL power fluxes entering the divertor regions may reduce the operational window in which the targets can be protected from excessive power loading. DEMO predictions will need to couple the transport and power dissipation in the SOL with the neutrals interaction in the divertor, which in practice requires using 2D plasma fluid codes. A bottleneck for the predictions are the long convergence times associated with the simulations, which is typically dealt with by performing various simplifications in the physics models used in the codes
[1,2].In 2D edge fluid codes such as SOLPS-ITER, the possible reductions in the physics models include using a fluid model for the neutrals instead of a kinetic model, bundling the charge states of the impurities, and neglecting the effects of cross-field drifts
[2]. Both cross-field drifts and kinetic neutrals have been shown in past benchmarking efforts to be important for reproducing various divertor regimes in present-day devices, but including both models simultaneously in DEMO simulations would likely lead to convergence times of several months. Kinetic neutrals increase the credibility of modelling divertor detachment and radiation due to the fuel neutrals, but for modelling impurity radiation and the SOL power fluxes that enter the divertors, the drifts and the level of detail included in the impurity models could be equally or even more significant, depending on how strong the gradients in the DEMO edge plasma parameters are. As divertor detachment largely depends on the power and particle fluxes which enter the divertor, we have decided to focus first on the physics ingredients which impact these upstream conditions in a DEMO-relevant, highly radiative SOL. In this contribution, the significance of the various physics models are studied in DEMO simulations and benchmarked against selected experimental results obtained from present-day devices.First predictions of drift effects in the EU-DEMO are emerging, and they are illustrated in Figure 1, which shows the power fluxes to the lower outer divertor modelled using SOLPS-ITER with a fluid neutral description for a symmetric double-null (DN) configuration. Both ExB and diamagnetic drifts are found to influence the solution by changing the in/out and the up/down asymmetries such that more power reaches the lower outer divertor. Although these results are obtained for an unseeded, attached scenario, our first results with Ar seeding indicate non-negligible effects of drifts in the radiated power fraction and in the power flux asymmetries also in conditions approaching detachment. Simulations with kinetic neutrals are planned as a next step to compare the significance of drift effects against that of a better neutral model.
Benchmarking the models used in DEMO predictions against experimental data is important in particular when a simplified physics set-up is used. The ASDEX Upgrade and JET tokamaks are both full-metal devices that allow disentangling the effects of seeded impurities from the intrisic impurities, which have low concentrations in the SOL. In this contribution we show comparisons of our simulations against low-density N-seeded L-mode experiments
[3], which have the benefit that the effects of ELMs on both the measured signals and on the physical conditions in the SOL are avoided. In such plasmas, the drift effects are typically noticeable[3]but strongly recombining regimes are avoided, which supports benchmarking our upstream models but give less value for validating the predictions for the DEMO targets. To illustrate the possible effects of using reduced physics models in the simulations, Figure 2 shows SOLPS results for JET using fluid neutrals, bundled N impurities and de-activated drift terms similar to[2]. The reduced physics simulations are able to capture some trends observed in these experiment, such as the relative strengths of core and edge radiation and the reduction of the outer target power load by the seeding. However, the dependence on Zeff is not reproduced, which is at least partly due to the missing drift effects, as observed in simulations using the full-physics set-up[4].Although a more complete picture of the importance of the various physics models for predicting radiation and power fluxes in the DEMO SOL is still being formed, our first results suggest that at least drift effects should be considered when modelling these processes, similar to the observations made when modelling present-day devices.
[1]F. Subba et al, Plasma Phys. Control. Fusion 60 (2018) 035013
[2]D. P. Coster, Contrib. Plasma Phys. 56 (2016) 790-795
[3]L. Aho-Mantila et al, Journal of Nuclear Materials 438 (2013)
[4]L. Aho-Mantila et al, Journal of Nuclear Materials 463 (2015)This work has received funding from the Academy of Finland (decision number 289726). This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
![Radiated power fractions (top) and outer target power loads (bottom) in an L-mode N seeding experiment in JET. Left: measured values in discharges #82291-#82299 [4], right: simulated values using fluid neutrals, bundled impurities and no drifts.](https://raw.githubusercontent.com/leena-a-m/fusion/master/synopsis_figure_2_small.png)
Speaker: Leena Aho-Mantila (VTT Technical Research Centre of Finland) -
Sustained fusion reactions have been measured in a quiescent deuterium Z-pinch plasma, wherein unity beta was achieved, and sheared flows alone provided stability. Measurements from multiple scintillator neutron detectors demonstrated that 2.45 MeV neutrons were emitted uniformly along the majority of the 50-cm plasma column and that the neutrons were produced from a thermonuclear process with negligible beam-target effects. Theory and simulations indicate that neutron yield will increase rapidly with pinch current, \propto I^{10}.
In a traditional Z-pinch equilibrium, an axial pinch current radially confines plasma pressure such that increasing the current results in higher densities and temperatures. While virulent pressure-driven instabilities are known to quickly destroy the traditional Z-pinch equilibrium, theory showed that introducing a sheared axial flow stabilizes the plasma1. Closely coupled with computational studies, a series of Z-pinch experiments at the University of Washington tested the theory of sheared-flow stabilization. Experimental measurements of the plasma equilibrium and stability confirmed that in the presence of a sufficiently large flow-shear, gross Z-pinch instabilities were mitigated, and radial force balance was achieved. Z-pinch plasmas of 50, 100, and 126-cm lengths were held stable for durations much longer than predicted for a static plasma, i.e. thousands of growth times2. Experimental results were combined with adiabatic scaling relations and detailed single-fluid, multi-fluid, and kinetic computational studies to explore the limits of plasma properties that can be achieved in a sheared-flow-stabilized (SFS) Z pinch.
The collaborative FuZE (Fusion Z-pinch Experiment) project between UW and LLNL scaled the SFS Z pinch to fusion conditions. Flow-shear stabilization was demonstrated to be effective even when a 50-cm long plasma column was compressed to small radii (3 mm). Improved understanding of the stabilization mechanism provided a means of increasing plasma parameters, e.g. ne > 1e17 /cc and Ti > 1 keV. Steady neutron production3 was observed for durations up to 8 microseconds (Fig. 1) during which the plasma was stable, and the current was sufficiently high to compress the deuterium plasma to fusion conditions. Measurements of neutron energy demonstrated a thermonuclear origin of the fusion process with negligible beam-target contributions. Neutron observations were not associated with MHD instabilities, and measured neutron yields[4] scaled with the square of the deuterium concentration and agreed with thermonuclear yields calculated with the measured plasma parameters.
Increasing the pinch current has demonstrated a corresponding increase in neutron yield, with yields of up to Yn = 5e6 neutrons per pulse. Neutron yield scales strongly with pinch current, as the Bennett relation (for fixed linear density N) gives T = I^2, and the d(D,3He)n fusion reactivity for T = 1-10 keV scales as sigma-v \propto T^4.
Initial experimental studies of neutron yield with pinch current show Yn dependent on the current to the 8-10th power. These results are supported by two-temperature MHD simulations of the FuZE device, which shows the neutron yield asymptotically increasing as I^{10}. See Fig. 2.
Kinetic simulations using the LSP code5 show stabilization of the m=0 mode with increasing flow shear for experimental conditions and continuing for fusion reactor conditions, shown in Fig. 3.
Experimental observations generally agree with theoretical and computational predictions, indicating that sheared flows can stabilize and sustain a Z-pinch equilibrium. If performance continues to improve with pinch current as experimentally observed and computationally predicted, the SFS Z pinch would make a compact fusion device for energy or neutron production applications.
Speaker: Uri Shumlak (University of Washington) -
We have recently developed a kinetic neutral code for understanding the hydrogen neutral gas dynamics in the divertor region. The Linear Divertor Analysis with fluid model (LINDA) code [A] has been successfully coupled with Kinetic Monte-Carlo Neutral Code (KMNC) in the present study. This is the first attempt for the GAMMA 10/PDX detached plasmas to perform such a self-consistent and detailed analysis using the fluid plasma and kinetic neutral code. We then investigate numerically the plasma and neutral particles transport behavior at the end-cell of GAMMA 10/PDX. The simulation results clarified that hydrogen (H) neutral particles can effectively reduce the plasma energy and can generate the detached plasma by enhancing the ionization, the charge-exchange, and the recombination loss. The obtained knowledge from the study contributes to understanding the physical mechanism of plasma detachment operation by H neutral particles in the divertor region of the fusion devices.
Fusion reactors cannot be realized without reducing and/or controlling the heat and particle loads to the divertor plate by plasma detachment. Therefore, understanding of the plasma detachment mechanism and of roles/effects of neutral particles on the detachment processes are indispensable in order to realize the long pulse operation of fusion reactors. In particular, the impact of atomic and molecular processes on the detachment is an unsolved critical research issue. Linear plasma devices greatly contribute to revealing plasma physics because the linear device has a much more simple structure than tokamaks. The impact of H Gas-Puffing (GP) on the plasma parameters has been investigated in the divertor region of the linear plasma confinement device GAMMA 10/PDX [B]. However, the physical mechanism of plasma detachment and energy loss processes during plasma-neutral interactions have not yet been clarified so far. In particular, the impact of atomic and molecular processes of neutral particles on the energy loss processes has not yet been understood. Consequently, the numerical simulation research has been started in order to clarify unsolved physical issues noted above.
In this simulation, the LINDA code and the KMNC code have been numerically solved in the self-consistent manner. More specifically, plasma transport is defined by the fluid equations while the neutral particles are defined by the kinetic neutral code. The recycling and puffed neutral particles are considered as $H_2$. The test particles are puffed into the bulk plasma with the cosine distribution. By applying a newly developed integrated code, LINDA+KMNC, we found that neutral particles concentrate near the target plate and reduce towards the upstream region. It was also found that the ion temperature ($T_i$) is reduced in the vicinity of the target plate when H$_2$ Gas-Puff (GP) is performed. The plasma ion density ($n_i$) is also increased near the target plate by the GP. The charge-exchange (CX) loss and the momentum loss increase significantly near to the target plate when GP is performed, which induces a significant reduction in the $T_i$. Moreover, the $n_i$ is increased while electron temperature ($T_e$) is reduced by the ionization of neutral particles. At the higher neutral injection cases, $T_e$ is reduced to below the threshold of ionization potential. Consequently, the $n_i$ is reduced by the recombination. These simulation results may solve the critical issues such as the role of atomic processes on the energy loss processes and on the detached plasma formation.
In the LINDA code, 2-D numerical grids have been created in the cylindrical coordinate system based on the axisymmetric condition. Figure 1 shows the mesh structure of the simulation region. A tungsten (W) target plate is designed at the end of the simulation space, as shown in Fig.1. Figure 2 shows the H neutral atom density ($n_{oH}$). For the case a (without GP), the neutral density comes from the recycling neutral particles and is shown to be very small. However, the $n_{oH}$ is enhanced significantly near the target plate when H$_2$ GP is performed as shown in Fig. 2(b)-(c). The $n_{oH}$ increases with increasing GP rate.
The axial profiles of $T_i$, $n_i$, and source terms of the fluid equations are shown in Fig. 3(a)-(d). The cases of the Fig.3 legends are explained in the Fig.2 caption. For without GP, the $T_i$ profile reduces slightly. Furthermore, the $T_i$ reduces drastically in the direction of the target plate when H$_2$ GP is performed. As shown in Fig. 3(b), the $n_i$ increases significantly near the target plate. Since neutral density is higher closer to the target plate, as shown in Fig. 2(b-c), the energy loss terms are also concentrated close to the target plate. As shown in Fig. 3(c)-(d), the ion energy and momentum loss are greatly enhanced, which leads to a strong reduction in $T_i$.
The fluid code has been compared with the Particle-In-Cell (PIC) code [C] for the first time to understand the validity of the fluid code LINDA. As shown in Fig. 4, the $x$-axial profile of $n_i$ and $u_{i||}$ of the two codes are found to be consistent, which indicates the validity of the LINDA code.
Finally, the LINDA code outcomes are compared with the experimental data and plotted in Fig. 5. In particular, the pattern of the allover in the $n_e$ qualitatively agrees with the simulation outcomes. In addition, for higher neutral injection cases, the $T_e$ is saturated for both the simulation and the experiment.
[A] M.S. Islam et al., Plasma Phys. Control. Fusion 59 (2017) 125010.
[B] M.S. Islam et al., Plasma Fusion Res. 11, 2402042 (2016).
[C] T. Pianpanit et al., Plasma Fusion Res. 11, 2403040 (2016).Speaker: Dr Md Shahinul Islam (National Institute for Fusion Science, Japan) -
It is critical to solve the power exhaust issue for future fusion devices, such as China Fusion Engineering Test Reactor (CFETR) [1]. In case of the steady-state operation, the divertor plasma is expected to be in the detachment regime. Thus, the heat load onto the targets can be reduced within the engineering limits and the sputtering can also be mitigated. Radiative snowflake divertor [2] is a promising way to effectively reduce the heat load onto divertor targets. Comparing to the single-null divertor configuration, the magnetic flux expansion are increased by introducing the second X point in the snowflake divertor configuration [3], where the detachment is believed to be promoted. Our previous work [4] also supported the positive effect on divertor detachment in the quasi-snowflake divertor. Nevertheless, we also find that the detachment sequence can be influenced, i.e. the low-field-side (LFS) target can be detached in prior to the high-field-side (HFS) one when the plasma density is relatively low. While the flux expansion can be controlled by moving the position of the second X point, it implies that there is an additional way to control the characteristic of divertor detachment. To support the design of snowflake divertor and the control of detachment in the future fusion device, the influence of flux expansion on the detachment sequence of LFS and HFS target is further studied by SOLPS [5] simulation in this work.
For the reason of the consistency to the previous work, this work is based on the 5.7 m major radius case of CFETR, where the fusion power is assumed to be 200 MW. By changing the position of the second X point (labeled by I, II, III and IV) along the line of plasma center and the main X point in EFIT, a series of the quasi-snowflake divertor configuration is generated, as shown in Fig. 1, where the flux expansion in the HFS scraped-off layer (SOL) is similar and the flux expansion in the LFS is decreasing with increasing distance between two X points. In the SOLPS simulation, because we are focusing on the effect of flux expansion, the divertor target is placed perpendicular to the separatrix to avoid too much complex geometry effects. Fully tungsten first walls and divertor are assumed and neon is seeded to achieve radiative divertor. The recycling rates at first walls and divertor targets are fixed to 100% for all species except for the location of pumping. The total power across the core-edge boundary is fixed to 100 MW and equally distributes to ions and electrons. Constant particle and thermal diffusivities are assumed, i.e. ${D_{\perp}}=0.3 \rm{\ m^2s^{-1}}$, ${\chi_{\rm{i}}}={\chi_{\rm{e}}}=1.0 \rm{\ m^2s^{-1}}$. The density scan is performed by $\rm{D_2}$ puffing.
When keeping the upstream density at ~$2.5 \times10^{19} \rm{\ m^{-3}}$, the neon impurity is seeded to achieve ~35 MW radiation power. The simulated radiation distribution are shown in Fig. 2 for four configurations with different LFS flux expansion. It can be seen that, the radiation are concentrated near the LFS target for the configurations with larger LFS flux expansion (I and II), while inversed situation is found for the smaller LFS flux expansion cases (III and IV). Thus, the LFS plasma is detached in prior when the LFS SOL flux expansion is large to a certain extent, which is not the case in the single-null divertor. The simulation results show clear influence of the flux expansion on the detachment sequence of HFS and LFS divertor targets.
A density scan is then performed for case IV. From the radiation distribution shown in Fig. 3, it can be seen that, for a certain LFS flux expansion, the detachment sequence can also be influence by the upstream density. With a higher upstream density (~$3.0 \times10^{19} \rm{\ m^{-3}}$), he HFS target detachment is achieved in prior, which is the normal case. When the upstream density is lower (~$2.0 \times10^{19} \rm{\ m^{-3}}$), the LFS target detached firstly. The simulation results show that the reverse of the detachment sequence should be a combined effect of the flux expansion and upstream density. In Fig. 4, the delicate results of the dependence of the detachment sequence on the upstream density (controlled by $\rm{D}_2$ puffing) is shown. An obvious critical upstream density can be identified, below which the detachment is achieved in prior for the LFS target.
In summary, we have performed simulations to the snowflake divertor configurations with different LFS magnetic flux expansions. The influence of the flux expansion on the detachment sequence of the LFS and HFS targets is clearly shown in the simulated results, where larger LFS expansion prefers the LFS divertor detachment. For the configuration with a certain LFS flux expansion, a critical upstream density is identified, below which the LFS target will be detached in prior. A 1D SOL model based derivation about the critical density and its relationship about the flux expansion will be also presented in the conference.
Reference
[1] Y.X. Wan et al., Nucl. Fusion 57 (2017) 102009.
[2] V.A. Soukhanovskii et al., J. Nucl. Mater. 463 (2015) 1191.
[3] D.D. Ryutov et al., Phys. Plasmas 2015, 22 (2015)110901.
[4] M.Y. Ye et al 2019 Nucl. Fusion 59 (2019) 096049.
[5] R. Schneider et al., Contrib. Plasma Phys. 46 (2006) 3.Speaker: Minyou Ye (School of Nuclear Science and Technology, University of Science and Technology of China) -
Power exhaust is considered as one of the most critical issues for future fusion devices. It is considered as an indispensable way for reducing divertor heat load to a tolerable level by seeding impurity and achieving a radiative divertor, especially for the future tokamaks with full metal walls. For China Fusion Engineering Test Reactor (CFETR) [1], which aims to achieve a fusion power ~1 GW, a high total radiation fraction is necessary. However, the impurity seeding is restricted to avoid degradation of the main plasma performance. In this work, the influence of the kind of the seeded impurity species (nitrogen, neon and argon) on the effective plasma ion charge $Z_\rm{eff}$, which is critical for further study on the impurity issues in core plasma and pedestal [2], is emphasized for the high radiative scenario.
The simulations are performed using SOLPS5.0 code package [3]. Fully tungsten as the first walls and divertor are assumed in simulations. The recycling rates at first walls and divertor targets were fixed to 100% for all species except for the location of pumping. Basic geometry configuration and computational mesh using in SOLPS simulations is shown in Fig. 1, as well as the locations of pumping and gas puffing. The total power across the core-edge boundary is fixed to 200 MW and equally distributes to ions and electrons. A specific set of radially varying cross-field transport coefficients which is identical for all ions account for the pedestal structure predicted by EPED model for CFETR H-mode discharges. A high radiation fraction $f_\rm{rad}$~0.85, which results in detached divertor plasmas is achieved in simulations by injecting a mixture of deuterium gas and different impurity gases. A density scan was performed by varying the deuterium ion density at the core-edge boundary in the range from $6.0 \times {10^{19}}{\rm{\ m}}^{\rm{-3}}$ to $1.0 \times {10^{20}}{\rm{\ m}}^{\rm{-3}}$. The separatrix density is in the range of $2.3{\rm{-}}3.3\times {10^{19}}{\rm{\ m}}^{\rm{-3}}$ and is well below one third of Greenwald density.
The simulated effective plasma ion charge $Z_\rm{eff}$, electron density $n_\rm{e}$ (both at the core-edge boundary) and the radiated power $P_\rm{rad}$ are fitted according to the Matthews’ scaling [4] from multi-machine database, which results ${Z_{eff}}=1+\left. {0.78509{{(Z-1)}^{0.30778}}{P_\rm{rad}}} \middle/{n_\rm{e}^2} \right.$ as shown in Fig. 2, where different trends can be found for different impurity species. Meanwhile, the fitting formula implies that, the the higher atomic number of the seed impurity $Z$ is, the better performance (i.e. lower $Z_\rm{eff}$ with same $P_\rm{rad}$ and $n_\rm{e}$ ) can be obtained, which is not the result according to the simulation. In order to figure out which impurity species has better performance, the radiative efficiency ${H(Z_\rm{eff})} = \left. {f_\rm{rad}} \middle/{(Z_\rm{eff}-1)} \right.$ is plotted in Fig. 3. For the three chosen impurity species, we have $Z_\rm{Ar} \gt Z_\rm{Ne} \gt Z_\rm{N}$, while for the radiative efficiency $H(Z_\rm{eff})$, it is ${H(Z_\rm{eff},\rm{N})} \gt {H({Z_\rm{eff}},\rm{Ar})} \ge {H(Z_\rm{eff},\rm{Ne})}$.
It is noteworthy that there is a relationship of ${E_\rm{ion}^\rm{Ne}} \gt {E_\rm{ion}^\rm{Ar}} \gt {E_\rm{ion}^\rm{N}}$ for the ionization potential $E_\rm{ion}$ of impurities, which follow the same sequences of the radiative efficiency $H(Z_\rm{eff})$. For different impurities, a larger ionization potential leads to ionizations closer to the upstream (according to the ionization source) and longer resident time $\tau$ in the plasma (based on the DIVIMP calculation). When obtaining the Matthews’ scaling, the non-coronal effect is omitted. However, in the SOL region, $\tau$ can be less than 1 ms, which leads to a non-coronal condition. The omitted non-coronal effects may be one of the reasons for the discrepancies with different seeding impurities.
Because it is beyond our ability to give a precise quantitative description of the non-coronal effect in the edge plasma, a simple relationship, i.e. the radiated power coefficient ${L_\rm{non-coronal}} \propto \left. {L_\rm{coronal}} \middle/ {n_\rm{e}^\beta} ({\left. {E_\rm{ion,imp}} \middle/ {E_\rm{ion,D}} \right.})^\gamma \right.$, is assumed. Thus, a modified formula is brought out to include the influence of non-coronal effect, ${Z_{eff}}=1+\left. {0.14882{{(Z-1)}^{0.39939}}{P_\rm{rad}}} \middle/{n_\rm{e}^{2-{0.54212}}} ({\left. {E_\rm{ion,D}} \middle/ {E_{\rm{ion},{imp}}} \right.})^{0.87085} \right.$. It can be seen that, the discrepancy for the trends of different impurity species does not appear. Furthermore, the result shows a dependence of ~$n_{\rm{e}}^{-1.5}$, which is in consistent with the experiment results of JET with high radiation and low core pollution [5].
In summary, we have performed SOLPS simulations for CFETR with different radiation impurity species, i.e., N, Ne and Ar. The radiative efficiency for different impurity species indicates that $H(Z_\rm{eff})$, it is ${H(Z_\rm{eff},\rm{N})} \gt {H({Z_\rm{eff}},\rm{Ar})} \ge {H(Z_\rm{eff},\rm{Ne})}$. It implies that, in case of nitrogen can not be seeded due to the tritium issues, Ar is preferred. Furthermore, a modified formula by introducing the non-coronal effect in the Matthews’ scaling is brought out for fitting the simulated $Z_\rm{eff}$, $n_\rm{e}$ and $P_\rm{rad}$. As a result, the discrepancy for the trends of different impurity species when fitted by Matthews’ formula is improved.
Reference
[1] Y.X. Wan, et al., Nucl. Fusion 57 (2017) 102009.
[2] N. Shi et al, Nucl. Fusion 57 (2017) 126046.
[3] R. Schneider et al., Contrib. Plasma Phys. 46 (2006) 3-191.
[4] G.F. Matthews, et al., J. Nucl. Mater. 241-243 (1997) 450-455.
[5] J. Rapp, et al., J. Nucl. Mater. 390-391 (2009) 238-241.Speaker: Shifeng MAO (University of Science and Technology of China) -
Hailong Du1*, Guoyao Zheng1, Xuru Duan1, Houyang Guo2, Jiaxian Li1, Lei Xue1 and Yue Zhou1
1Southwestern Institute of Physics, Chengdu, 610041, China
2General Atomics, PO Box 85608, San Diego, CA 92186-5608, United States of America
Email: duhl@swip.ac.cnA new mid-size device, HL-2M $[$ 1$]$, is being installed at SWIP to address critical physics and technology issues towards advanced divertor development for CFETR and future fusion reactors. The divertor structure of HL-2M has been designed to be compatible with various advanced divertor magnetic configurations, including snowflake (SF-, SF+ and Trip), as well as advanced divertor structure with a V-shaped target in the lower outer divertor, similar to the new small angle slot divertor (SAS) in DIII-D $[$ 2$]$. Furthermore, other advanced divertor configurations will be explored with the upper divertor in the near future. A major purpose of the optimization of divertor configuration is to facilitate access to highly radiative divertors, which are the operational regimes required for power load control in ITER and nuclear fusion reactors, In particular, the control of detachment has been identified recently as a serious challenge due to the rapid decrement of divertor target electron temperature (called detachment cliff). This work systematically examines the formation of necessary conditions of such a detachment cliff in HL-2M SF-divertor, and comparisons are made with the DIII-D conventional open divertor. Both HL-2M and DIII-D tokamas have almost the same size, as shown in figure 1. The modeling with SOLPS including drifts have uncovered that the root cause for the formation of detachment cliff is the strong interplay between the E×B drift and the carbon radiation loss under the condition of high confinement (H-mode) plasmas.
SOLPS modeling results showed that the E×B drift exhibits a strong influence on power dissipation and detachment in both HL-2M SF- and DIII-D open divertors. For the favorable Bt direction, i.e., with the ion grad-B drift toward the X-point, a large number of particles are driven by the Er×B drift from the outer divertor into the inner divertor through the private flux region. This reduces the plasma density and hence the power radiation loss in the outer divertor, thus driving the outer divertor plasma away from detachment. The detachment cliff occurs in both HL-2M SF- and DIII-D outer divertors, as marked by a pronounced drop in the electron temperature, Te, at the divertor target, when ne,sep, is ramped up to a sufficiently high level, as shown in Fig. 2 and Fig. 3, respectively. The SOLPS analysis reveals a strong coupling between the E×B drift-driven flows and carbon radiation in HL-2M SF- and DIII-D. The high confinement H-mode facilitates the formation of the detachment cliff due to the narrow SOL width and hence steep radial gradients near the separatrix, which enhances the polodial E×B drift,(Fig.2 (a), Fig. 3(a)). Furthermore, the SOLPS modeling indicates that carbon radiation also plays a key role in the formation of the detachment cliff in HL-2M SF- and DIII-D without additional impurity seeding [3, 4], thus uncovering the common physics mechanisms for this critical issue in both devices.
Acknowledgments
This work was supported by National Natural Science Foundation of China under Grant Nos. 11805057, 11775071; National Key R&D Program of China No. 2018YFE0301101.References
[1] Zheng G Y, et al. 2016 Investigations on the heat flux and impurity for the HL-2M divertor Nucl. Fusion 56 126013
[2] Guo H Y,et al. 2019 First experimental tests of a new small angle slot divertor on DIII-D Nucl. Fusion 59 086054
[3] Du H, et al. 2020 Exploring SF- in-out asymmetry and detachment bifurcation in HL-2M with E × B by SOLPS Nucl. Mater. Energy 22 100719
[4] Du, H, et al. 2020 SOLPS analysis of the necessary conditions for detachment cliff, Nucl. fusion (accepted online)Speaker: Hailong Du (Southwestern Institute of Physics) -
Finding an optimized Inertial Confinement Fusion1–3 experimental design is a challenge due to the large number of physical parameters that can be modified from experiment to experiment, and the inability of simulations to accurately
and rapidly a priori predict experimental results when these changes are made. Recently, a novel method[4] has been developed to address this issue by statistically coupling simulation and experimental outcomes, resulting in the first truly predictive models for the observables of ICF experiments. These models have been used to design the highest performing experiments on the OMEGA laser system[5], which are predicted to result in about 500 kJ of fusion yield at
energies typical of the National Ignition Facility (NIF)[6]. Analyzing the dependencies of the models have also resulted in an improved scientific understanding of the degradation mechanisms affecting implosions on the OMEGA laser system. and has led to facility upgrades that have increased performance and reproducibility for ICF experiments.The rather large parameter space over which design optimization takes place necessitates the existence of a rapid and accurate predictive tool to conduct any optimization scheme whose end goal is ignition[7,8], a necessity for inertial fusion energy (IFE) to become a reality. Historically, the primary tool in ICF design has been the radiation-hydrodynamic (RH) simulation[9–13]. Though recent advances in physics understanding has led to RH simulations achieving better
agreement with experimental observations[14–17], these simulations (regardless of spatial dimension) are not yet able to predict the results of a future experiment in which the initial conditions have been changed a priori.The inability of RH codes to accurately predict the effect of changes in their designs is likely a major obstacle in achieving ignition in ICF, since it precludes any effective implementation of iterative optimization methodologies that could be use to rapidly increase performance. This predictive deficit also
severely restricts the ability of scientists to identify degradation mechanisms directly from experimental data, as the effect of a degradation mechanism cannot be easily decoupled from varying initial conditions in the absence of an accurate predictive model (statistical, or otherwise).However, consider that if the outputs of an RH code $\mathbf{O}_{\textrm{1D}}^{\textrm{sim}}$ uniquely define the inputs to the code
$\mathbf{I}_{\textrm{1D}}$ (which are also the inputs to the experiment), then it follows that the outputs of the experiment
$\mathbf{O}_{\textrm{3D}}^{\textrm{exp}}$ are $ \begin{equation} \mathbf{O}_{\textrm{3D}}^{\textrm{exp}} = \mathbf{F}_{\textrm{3D}}^{\textrm{exp}}\left[ \mathbf{I}_{\textrm{1D}},\mathbf{S}_{\textrm{3D}}^{\textrm{sys}},\mathbf{S}_{\textrm{3D}}^{\textrm{ran}} \right], \end{equation}$where
$\mathbf{S}_{\textrm{3D}}^{\textrm{sys}}$ and $\mathbf{S}_{\textrm{3D}}^{\textrm{ran}}$ are systematic and random 3D perturbations in the experiment. For a repeatable experiment, $\mathbf{S}_{\textrm{3D}}^{\textrm{ran}} << \mathbf{S}_{\textrm{3D}}^{\textrm{sys}}$, and we have
$\begin{equation} \mathbf{O}_{\textrm{3D}}^{\textrm{exp}} =\mathbf{F}_{\textrm{3D}}^{\textrm{exp}}\left[ \mathbf{I}_{\textrm{1D}},\mathbf{S}_{\textrm{3D}}^{\textrm{sys}} \right]=\mathbf{F}_{\textrm{3D}}^{\textrm{map}}\left[ \mathbf{O}_{\textrm{1D}}^{\textrm{sim}} \right] \end{equation}$since $\mathbf{S}_{\textrm{3D}}^{\textrm{sys}}$ are constants. This implies that the results of an experiment can be related to the outputs of a 1D RH code by the function $\mathbf{F}_{\textrm{3D}}^{\textrm{map}}$. Ref. 4 approximates $\mathbf{F}_{\textrm{3D}}^{\textrm{map}}$ with power laws, and reconstructs it statistically by comparing a database of over 200 OMEGA direct-drive cryogenic implosions spanning a wide range of initial conditions, resulting in models for the neutron yield, areal density and hotspot radius (Fig.1). Ensemble averages of several semi-independent models constructed using this method can be taken to generate predictions. These predictions are typically accurate to within 10%, (Fig. 1) making them considerably more accurate than 1D simulation results. They are also pessimistic compared to 1D simulations, which fail to predict the rapid drop-off in yield when the adiabat decreases and convergence increases. Instead, the predictive models correctly expect low yields and areal densities for highly convergent, low adiabat implosions.
As the statistically inferred $\mathbf{F}_{\textrm{3D}}^{\textrm{map}}$ operates on 1D simulations, a large swathe of parameter space can be rapidly scanned for viable designs, enabling rapid and iterative design. Using these models, a performance improvement campaign was conducted on OMEGA, as reported in Ref. 4. By following the recommendations of the models, the neutron yield on OMEGA was tripled, and the areal density was increased by 60%. Due to OMEGA’s energy constraints, the ignition-relevant performance of these implosions was assessed by the theory of hydrodynamic scaling[18], and were predicted to produce fusion yields of about 500 kJ, with a normalized Lawson parameter[19] of about 0.7 when scaled to 1.9 MJ of symmetric drive (Fig. 2). Previous results on OMEGA[20] were expected produce approximately 100 kJ of fusion energy when scaled to 1.9 MJ of symmetric drive[21], and recent indirect-drive experiments at the NIF have demonstrated 56 kJ of fusion energy[22].Inspecting the exact form of a predictive model has also led to physics insight regarding the degradation mechanisms active on OMEGA. In particular, it was observed[23] that targets filled with tritium close to the shot date tended to be underpredicted, while targets filled well before the shot tended to be overpredicted (Fig. 3). Accounting for this improved the prediction accuracy, including for a number of ‘outlier’ implosions, prompting an investigation for the underlying physical mechanism. One hypothesis was the build-up of Helium-3
from the beta decay of tritium, which considerably increases the initial vapor pressure of the target. A new database of simulations that accounted for the initial vapor pressure of He3 was constructed, and it was found that the quality of prediction remained high even in the absence of ad-hoc variables to
account for the age of the fill. As this is a relatively small (10-20% for moderate adiabats) effect relative to changes in design, the statistical model was essential in providing a baseline expectation from which deviations could be identified. Recent controlled experiments have confirmed this dependence (Fig. 2), and changes to the OMEGA facility to minimize the fill age are underway to maximize future performance.Though the US ICF program has not yet achieved ignition, steady progress over the last few decades means that only comparatively modest improvements are required to demonstrate ignition (Fig. 2), which is necessary for the realization of IFE. While we cannot know the exact magnitude of the improvements the statistical model can provide, it is clear that applying the statistical model to generating new designs, and to investigate and eliminate degradation sources has the potential to push the ICF program to the high yields necessary for IFE.
- J. Nuckolls, et. al, Nature 239 (1972)
- S. Atzeni et. al,(Oxford Univ. Press, Oxford, 2004).
- S. Craxton, Phys. Plasmas 22 (2015)
- V. Gopalaswamy, et al. Nature 565, 581–586 (2019).
- R. Boehly, Opt.Commun. 133 (1997)
- E. M. Campbell and W. J. Hogan, Plasma Phys. Contr. Fusion 41 (1999)
- A. R. Christopherson, Phys. Plasmas 25 (2018),
- R. Betti, Phys. Rev. Lett. 114 (2015)
- J. Delettrez, et. al, Phys. Rev. A 36 (1987),
- P. B. Radha, Phys. Plasmas 12 (2005),
- G. Zimmerman, et. al, J. Opt. Soc. Am. 68 (1978).
- M. M. Marinak, Phys. Plasmas 8 (2001),
- S. Atzeni, Comput. Phys. Commun. 169 (2005)
- I. V. Igumenshchev, Phys. Plasmas 17 (2010),
- D. T. Michel, High Power Laser Sci. Eng. 3 (2015)
- V. N. Goncharov, Phys. Plasmas 13 (2006)
- S. X. Hu, Phys. Rev. E 92 (2015)
- R. Nora, Phys. Plasmas 21(2014),
- R. Betti, Phys. Plasmas 17 (2010)
- S. Regan, Phys.Rev. Lett. 117 (2016)
- A. Bose, Phys. Rev. E 94 (2016)
- S. Pape, Phys. Rev. Lett. 120 (2018)
- A. Lees, et. al, Bull. Am. Phys. Soc. 64 (2019).
Speaker: Mr Varchas Gopalaswamy (University of Rochester/Laboratory for Laser Energetics) -
To study the runaway electron (RE) dynamics during plasma discharge, as well as to develop scenarios for disruption mitigation, a gamma-spectrometric system has been developed and commissioned at the ASDEX Upgrade tokamak (AUG). The diagnostic system consists of two scintillation gamma-ray spectrometers based on the fast LaBr3(Ce) crystals. These spectrometers observe the AUG tokamak chamber quasi-radially at the equatorial plane [1, 2]. Both spectrometers are equipped with modern data acquisition systems based on fast digitizers recording detector signals with an ultra-high sampling rate (up to 400 MHz). The fast digital processing of the recorded waveforms from the detector allows obtaining the Hard X-Ray (HXR) energy spectrum at any given time of the discharge under investigation. In order to carry out the pulse-height analysis under conditions of high detector load and many piled-up events, up-to-date digital signal processing algorithms are used. Dynamics of REs has been studied using this diagnostic and reported in this paper.
The measurements were carried out in the regimes of RE beam generation by injection of argon or krypton gas into a deuterium plasma. In the interaction of a developed RE beam with a heavy gas target, powerful bremsstrahlung flux is induced, reaching energy close to 20 MeV. The electron energy distributions were reconstructed from the measured HXR spectra by deconvolution methods [3, 4]. For this purpose, the detector response functions for monoenergetic gamma radiation in the range 0.1–30 MeV with a step of 0.1 MeV were calculated using the MCNP code. Bremsstrahlung generation functions caused by the interaction of accelerated electrons with a gaseous target were simulated as well corresponding to the detector viewing geometry. An analysis is carried out of the evolution of the maximum RE energy (Emax) derived by reconstructing RE distribution functions (REDF) from the measured HXR spectra with DeGaSum deconvolution code [4]. The experimentally obtained maximum RE energies at different stages of the discharge were compared with the results of test particle simulations that include the effect of toroidal electric field, plasma collisional drag force, synchrotron deceleration force. Figure 1 shows the main signals of AUG discharges # #34084 and 34183. During the discharge #34183 a series of deuterium pellets was injected after the argon puff. Figure 1e represents the evolution of the maximum RE energies reconstructed from HXR measurements in comparison with the results of test particle calculations. It was observed that the electrons attain their maximum energies in 50-100 ms after the gas injection. Then it gradually decreases due to the drop in loop voltage, energy loss due to synchrotron radiation emission and collisions dissipation of energy with the background plasma. The test particle simulation suggests that the Emax calculated by DeGaSum code matches reasonably well with the simulated energy evolution of REs created in the initial phase of the current quench. The decreases in RE-energy after the massive gas-injection (MGI) may be attributed to collision dissipation of the RE energy with plasma and that also leads to pitch angle scattering of REs, the increase in the RE pitch angle again enhances the synchrotron radiation losses that combined effect results in decreases of the Emax. These results are confirmed and supported by the test particle simulations.
Furthermore, HXR measurements at the discharge with a multiple deuterium pellet injection (PI) allowed observing the effects of plasma cooling and argon ion recombination after PI. In Fig.2 red dots show the runway current restored with the DeGaSum code from the measured HXR spectra. This result shows the diagnostic measurements can also aid to reconstruct RE-current evolution during disruption mitigation scenarios and reported herein for the first time. The green line shows the estimated argon atomic density in the place of the RE beam localization. This argon concentration is needed to explain the observed RE current. Analysis of REDFs has shown that after PI the runaway beam interacts mainly with the neutral gas target, which concentration significantly exceeds the electron density. During the discharge with pellet injections, Emax of REs was by 3-5 MeV lower than for the discharge with an application of MGI only (see Fig 1e).
The experiments and diagnostic signal analysis along with simulations demonstrated the ability of gamma-ray spectrometry to provide the most important RE parameters, such as maximum RE energy and current that is the measurement requirement for ITER [5]. Measurements at AUG make it possible to test equipment and RE diagnostic techniques to use them at ITER and new generation tokamaks.
References:
1. M. Nocente, et al., RSI 89 (2018) 10I124;
2. A. Dal Molin, et al., 46th EPS Conf. Plasma Phys. (2019) P1.1015;
3. A.E. Shevelev, et al., Nucl. Fusion 53 (2013) 123004;
4. A.E. Shevelev, et al., NIM A 830 (2016) 102-108;
5. A. J. H. Donne, et al., Nucl. Fusion 47 (2007) S337–84.Speaker: Alexander Shevelev (Ioffe Institute) -
An increasing number of installations for magnetic plasma confinement implement tungsten as a material for the plasma-facing components. For example, tungsten plates are used in the ITER divertor. Tungsten is the heaviest plasma impurity and despite the fact that its content in plasma $c_{\rm{W}} = \sum{n_\rm{W}}/n_e$ is relatively low ($c_{\rm{W}}\approx10^{-4}–10^{-5}$), due to its high emissivity W ions can radiate from the plasma a significant amount of electron energy and thereby greatly affect the plasma parameters. Besides that, most installations conduct experiments with additional heating, which introduces the energy of fast atoms and high-frequency waves into the plasma. This dramatically affects the transport processes in the plasma. The neoclassical transport of impurities is defined theoretically, but their anomalous transport is not. The study of anomalous transport characteristics in regimes with additional heating is a highly relevant task.
The T-10 tokamak is an installation with a circle cross section and a limiter configuration (major radius $R = 1.5$ m, minor radius $a = 0.3$ m) where the carbon (poloidal) limiter was replaced by a tungsten one in 2015. T-10 is equipped with a gyrotron complex with a total capacity of up to 2.5 MW for heating electrons at the second harmonic of electron cyclotron frequency (ECR-heating). Three gyrotrons are used: “A” (1 MW, 140 GHz, co-injection, capable of rotation along the poloidal angle), “B” (0.5 MW, 129 GHz, with injection perpendicular to the plasma current) and “C” (1 MW, 140 GHz, capable of rotation along the toroidal angle, with co-/counter-injection). The electron temperature $T_e(r)$ is measured using ECE-diagnostics and the slope of the spectrum in the SXR-region. CXRS-diagnostics [1, 2] is used to measure the ion temperature $T_i(r)$ and the density of light impurity nuclei $n_Z(r)$. The electron density $n_e(r)$ is measured with a 16-channel interferometer.
The tungsten inflow into the discharge is estimated from the brightness of the WI atom line $\lambda= 400.1$ nm at the W-limiter region. The density of tungsten ions is determined from the integrated radiation losses power recorded by AXUV-diagnostics. In [3], it is shown that the AXUV signal in the central region of the column (within $a/2$) is almost exclusively generated by tungsten radiation. In [4] the preservation of coronal equilibrium is determined on ASDEX Upgrade. It is confirmed for T-10 conditions (in ohmic and ECR regimes) [5]. This makes it possible to determine the profile of the total tungsten ions density $\sum{n_i}(r)$ from the expression:
$$P_{\rm{W}}^{rad}(r) = {n_e}(r)L_{\rm{W}}^{eff}(r)\sum\limits_i {{n_i}(r)},$$ where $P_{\rm{W}}^{rad}$ is power of radiation losses in the central region, registered by AXUV detectors, $L_{\rm{W}}^{eff}$ is tungsten cooling factor [6] calculated in the coronal equilibrium approximation. In T-10 experiments with ECR-heating, tungsten is removed from the center of the plasma column [7, 8]. In discharges where the removal is exponential, the description of the AXUV-signal experimental dynamics in the region from 0 to $a/2$ in the transport model allows to estimate the transport coefficients of W ions. For this purpose, a system of two continuity equations (dynamic and stationary) is solved: $$\left\{ \begin{array}{l} \frac{{\partial {n_\rm{W}}(r,t)}}{{\partial t}} + {\rm{div}}\left[ { - D(r) \cdot \nabla {n_\rm{W}}(r,t) + V(r) \cdot {n_\rm{W}}(r,t)} \right] = {Q_\rm{W}}(r,t),\\ {\rm{div}}\left[ { - D(r) \cdot \nabla {n_\rm{W}}(r) + V(r) \cdot {n_\rm{W}}(r)} \right] = {Q_\rm{W}}(r), \end{array} \right.\ (1)$$ where $n_\rm{W}$ is the full density of tungsten ions, $D$ and $V$ are the desired diffusion coefficient and pinching velocity, $Q_\rm{W}$ is the sum of sources and sinks ($Q_\rm{W}\approx0$ for the central region of the plasma). The solution is carried out by choosing $D(r)$ and $V(r)$ that allow us to describe the temporal behavior of the tungsten radiation profile recorded by the AXUV detector in the best way. The scenario of such a discharge is shown in Fig. 1: at the ECRH stage, there is a quasistationary part (600 – 640 ms) where the electron density and temperature are nearly constant, but the decay of the AXUV-signal in the central region is pronounced. It indicates the removal of tungsten from the center of the plasma. The red line on last plot is an exponential approximation of the AXUV signal to find the experimental decay time $\tau_{exp}$. The system (1) is solved using the STRAHL code integrated in the ASTRA code. Transport coefficients are considered as the sum of neoclassical and anomalous components: $D = D_{neo} + D_{an}$ and $V = V_{neo} + V_{an}$. Neoclassic is calculated using the NEOART code. To conduct the complete study, the selection of discharges with different ECR heating parameters (on-axis / off-axis heating, presence/ absence of sawtooth oscillations, active / suppressed MHD mode $m = 2$) suitable for processing is carried out. The dynamics of tungsten removal from the central region of the cord is simulated for time intervals in which the exponential decay of AXUV-signals occurs without changes in $n_e$ and $T_e$. An example of an AXUV decay description in a discharge with central heating with a power $P_{EC} = 0.75$ MW is shown in Fig. 1 (red line). The calculation is made with the coefficients $D$ and $V$ shown in Fig. 2. The characteristic decay time is $\tau_{exp}\approx14$ ms. This time can be obtained by adding diffusion to the 1.5 – 2 m$^2$/s over the ohmic level. The pinch term in this case turns out to be practically zero at radii up to the $a/2$. Thus, the introduction of on-axis ECR-heating leads to the tungsten removal from the plasma center mainly due to an increase in its anomalous diffusion.References
- V.A. Krupin et.al. Physics of Atomic Nuclei, 2015, Vol. 78, No. 10, pp. 1164–1173
- Sushkov A. et.al. High-resolution multiwire proportional soft x-ray diagnostic measurements on TCV // Rev. Sci. Instrum. 2008. Vol. 79, No 2 Pt 1. P. 023506
- V.A. Krupin et.al. Nucl. Fusion 57 (2017) 066041 (9pp)
- Pütterich T. et.al. Modelling of measured tungsten spectra from ASDEX Upgrade and predictions for ITER // Plasma Phys. Controlled Fusion. IOP Publishing, 2008. Vol. 50, No 8. P. 085016
- Zemtsov I. et al. — In: 44th EPS Conf. on Plasma Physics, 2017, p. 4.175.
- Demura A.V. et.al. Tungsten Ions in Plasmas: Statistical Theory of Radiative-Collisional Processes // Atoms Peace Int. J. Multidisciplinary Digital Publishing Institute, 2015. Vol. 3, No 2. P. 162–181
- V.A. Krupin et.al. Nucl. Fusion 57 (2017) 066041 (9pp)
- Sertoli M. et.al. Local effects of ECRH on argon transport in L-mode discharges at ASDEX Upgrade // Plasma Phys. Controlled Fusion. 2011. Vol. 53, No 3. P. 035024
Speaker: Mr Ivan Zemtsov (National Research Centre "Kurchatov Institute") -
A D-D fusion reaction during the Alfvénic collisional merging formation process of a field-reversed configuration (FRC) plasma has been detected by the developed fast response neutron detector in the FAT-CM device at Nihon University [A]. The neutron radiation and its dependency on the translation velocity of FRC plasma have been observed as an experimental evidence of an excited shock heating in the collisional merging of the oppositely-translated FRCs. Two dimensional MHD simulation does not predict high enough ion temperature that produces fusion reaction, thus the observed significant amount of neutron radiation indicates the rapid acceleration of non-thermal ions through the excited shock as one of the regeneration channels from the kinetic to the internal thermal energy.
In an FRC, a high beta ($ \beta $ ~ 1) simply-connected confinement region purely with poloidal magnetic flux is isolated in a mirror-like open magnetic field structure (Fig. 1). Because of this unique feature of the FRC, heating/current drive technique is almost limited to a tangential neutral beam injection. Therefore, building-up the poloidal flux being sufficient to capture the tangentially injected fast beam ions is the most important development issue. The super-sonic/Alfvénic collisional merging FRC formation has been proposed and conducted as an effective formation technique that does not inhibit the uniqueness of FRC. Because a long-lived hot FRC, using the super-sonic/Alfvénic collisional merging formation, has been demonstrated in the C-2W device at TAE Technologies [B], FRC attracts attention again as a steady-state fusion reactor core. In the ideal MHD equilibrium of FRC, pressure gradient, poloidal magnetic flux and toroidal current must be balanced while keeping its beta value of almost unity. Therefore, these dynamic processes must be completed faster than the relaxation and magnetic diffusion times to avoid energy and flux losses during the translation and merging process, where the plasmoid experiences drastic change of its volume and shape. On FAT-CM, behavior of the super-sonic FRC translation/collision has been investigated.
The FAT-CM device has stainless-steel confinement chamber with quasi-steady state external magnetic field. Initial FRCs are formed by the field-reversed theta-pinch method in the two formation sections with deuterium gas-puffing. FRCs with electron density of ~1×10$^{21}$ m$^{-3}$ are typically generated by the main compression field of ~0.4 T, after which they are accelerated by the gradient of the external guide magnetic pressure. The translated FRCs collide in the middle of the confinement chamber at the axial velocity in the range of 100 – 200 km/s (Fig. 1).
Figure 2 (a) traces time evolutions of axial translation velocity and characteristic velocities of Alfvén speed $V_{A}$ and sound speed $V_{S}$ at each FRC position during the dynamic translation process, simulated by 2-D resistive MHD equilibrium code. Note that in the super- Alfvénic collisional merging process the 2-D simulation cannot reproduce discontinuous profile, non-thermal ion and non-axisymmetric components of the magnetic field caused by shockwave with reasonable accuracy. The FRC in the formation region is accelerated by the external magnetic pressure up to $V_{A}$ and then translated into the confinement region while keeping the axial velocity. Therefore, the translation velocity exceeds $V_{A}$ and $V_{S}$ (typically ~50 and ~100 km/s, respectively) when FRC enters the confinement region as shown in Fig. 2(b).
As the experimental evidence of the excited shock, neutron detection has been performed during and after the collisional merging process by the newly-built fast response neutron detector consisting of a columnar plastic scintillator and a photomultiplier tube. The neutron measurement indicates a quick rise of the neutron flux when FRC translation velocity is faster than the Alfvenic velocity as seen in red solid line of Fig. 3(a). Here, blue dotted line denotes the collisional merging case with sub-Alfvén velocity, while green dashed line is single-sided FRC translation case without collision/merging. The number of D-D reaction cannot be explained by the adiabatic heating process under the assumption of resistive MHD model (Fig. 1). Therefore, the observed fusion reaction appears to be caused by the accelerated particles in the excited shocks. Soft X-ray (SXR) radiation detected by a surface-barrier diode with 37.5 $ \mu $m Be filter also shows a pulsed rise around $t$ ~ 40 $ \mu $s and following thermalization as shown in Fig. 3(b) [C].
Gradual increase in the SXR intensity is consistent with the thermal equilibrium time of several tens of microseconds between ions and electrons of the FRC plasma. This also suggests the non-thermal ion component accelerated in the super-Alfvénic collision process.
REFERENCES
[A] T. Asai et al., Nucl. Fusion 59, 056024 (2019).
[B] H. Gota et al., Nucl. Fusion 59, 112009 (2019).
[C] J. Sekiguchi et al., Plasma Fusion Res. 14, 3402116 (2019).Speaker: Prof. Tomohiko Asai (Nihon University) -
In the context of the inertial fusion energy (IFE) programme, the researches have embarked on a new stage which provides an opportunity to produce pure fusion ignition and burn by using power laser facilities and appropriate scale targets. The ignition and high gain target design requires free standing cryogenic target with an isotropic hydrogen fuel on the inside surface of a spherical polymer shell. This is due to the fact that the progress in plasma implosion up to intensive fusion reactions lies in formation of a given fuel structure that must be isotropic for reaching fusion conditions. Our approach to this issue is based on the development of structure sensitive methods resulting in the bulk homogenisation of fuel (hydrogen isotopes and their mixtures). Guided by general rules of crystallography, they also take into account the peculiarities inherent in the hydrogen isotopes as quantum molecular solids. These methods allow forming isotropic hydrogen layers by using high cooling (q = 1−50 K/s) combined with fuel doping (tritium, neon, argon) which results in creation of stable ultimate disordered structures with a high defect density or isotropic medium (Figure1). As applied to cryogenic solid layering the conception of isotropic layer structure comprises a certain level of its dispersity providing the required quality of layer surface finish. This is nano layering technology for which the grain size is scaled back into the nanometer range.
A unique feature of the experimental procedure is the following: a batch mode is applied and high cooling rates are maintained in the FST layering module (Figure 2a) inside free rolling targets that has no analogue in the world’s practice of IFE Cryogenics. This is precisely the condition which, in many respects, has defined a successful course of research. The obtained isotropic ultrafine layers have enhanced mechanical strength and thermal stability. For D Т mixture, tritium T2 is considered as a high melting additive with respect to D2 and deuterium tritide. This is a significant factor for the processes: (1) layer quality survival during target delivery; and (2) regular propagation of the shock wave, the front of which has to be extremely smooth.
Thus, the technical approach based on the free-standing target (FST) layering method in line-moving spherical shells is a credible pathway to a reliable, consistent, and economical target supply. So fast fuel layering is a necessary condition in-line target production [a] due to the following reasons: tritium inventory minimization, the producing targets in the massive numbers, and obtaining fuel as isotropic ultrafine layers. Multiple target protection methods are also used in the FST approach [b]. Among them are outer protective cryogenic layers, metal coatings of different configurations and compositions, nano-coatings for specific applications, co-injection of a special protective cover ahead of the target, etc.
As fusion reactions must occur approximately ten times a second, then a free-standing target (FST) factory becomes an integral part of any IFE reactor. Additionally, the high-power laser experiments with repetition rates between 1 and 10 Hz require developing noncontact delivery systems for safe, stable and friction-free target transport at the laser focus. In this area our approach is the target transport with levitation. The operational principle is based on a quantum levitation effect of type-II high-temperature superconductors (HTSC) in the magnetic field 2.
Significant progress has been made in the development of the hybrid electromagnetic accelerator, which is a combination of the acceleration system (field coils generating the traveling magnetic waves) and the levitation system (permanent magnet guideway (PMG) with a magnetic rail or magnetic track).
The obtained results have shown that the HTSCs can be successfully used to maintain a friction-free motion of the HTSC-sabots (target carriers) over the PMG, and also to provide a required stability of the levitation height over the whole acceleration length due to a pinning effect. Additionally, using the driving body from MgB2 superconducting coils as a sabot component (critical current 5000 A at magnetic induction 0.25 T) allows reaching the injection velocities 200 m/s under 400g overload at 5-m-acceleration length. Recent results obtained at the LPI may help improving the actual design of HTSC-maglev linear accelerator be using a circular one. Significant reduction of the accelerator dimensions and the number of the field coils can be obtained in a circular accelerator, in which only several field coils are arranged in a circle in the PMG-system.
Our objective of the study was to develop conceptual designs of the FST factory including IFE target muss-manufacturing followed by their rep-rate delivery at the reactor chamber. In this report, the detailed modeling and proof-of-principle (POP) experiments are outlined and discussed.
Acknowledgements
The presented results have been obtained in the frame of the IAEA Research Contracts #11536, #13871, and #20344 as well as in the frame of the State Task of the government of Russian Federation.
References
a. Irina Aleksandrova, Eugeniy Koshelev, Elena Koresheva. In-line target production for laser IFE. Applied Sci. 2020, 10 (2), 686 (pр. 1-17)
b. I.V. Aleksandrov, E.R. Koresheva, E.L. Koshelev. Multilevel system for protecting the cryogenic target during its delivery to the focus of high-power laser facility at high repetition rate. Physics of Atomic Nuclei, 2019, 82 (7), 1060–1071
Speaker: Dr Elena Koresheva (P.N.Lebedev Physical Institute of Russian Academy of Sciences, Moscow, Russian Federation) -
The dependence of particle and energy confinement in fusion plasmas on the main ion mass still challenges the current theoretical understanding of tokamak physics, although it represents one of the aspects of paramount relevance for the extrapolation to a fusion reactor. Moreover, knowledge is required in order to allow the experience developed in hydrogen (H) operation to be directly transferred into a successful operation in deuterium (D), an essential step from the pre-nuclear phase to the nuclear phase in ITER and in any future nuclear fusion reactor.
Many effects, which are connected to different physical mechanisms $[1-6]$, can concur in determining the mass dependence. However, as these mechanisms are often interdependent and can result in non-linear coupling of the edge and the core of the plasma, clear comparisons with the theoretical predictions are hindered by the difficulty of separating their roles. This also prevents a solid progress in the physical understanding.
In this contribution a new method is demonstrated, which allows us for the first time to experimentally separate the effects of core transport and edge pedestal with respect to the mass dependence. This enables unprecedented analyses on the separate properties of these two plasma regions. The resulting knowledge is then applied to mitigate the degradation of confinement in H and investigate previously inaccessible parameter regimes.In experiments with hydrogen isotopes in the ASDEX Upgrade (AUG) and JET tokamaks a strong degradation of the pedestal is observed in H compared to D plasmas for similar heat and particle fluxes $[4,5]$.
In consequence, a strong interconnection between the core and edge plasma arises in these isotope studies. We demonstrate experimentally that an increase of the plasma triangularity $\delta$ for H plasmas results in a matched pedestal top with low $\delta$ D plasmas without affecting the core transport properties.
This breaking of the core and edge plasma interconnection allows an independent analysis of both regions.A data set of H and D H-mode plasmas on AUG (0.8 MA and -2.5 T) and low density hybrid plasmas on JET (1.4 MA and 1.7 T) shows a strong correlation of the core confinement with the fast-ion content.
The fast-ion energy in D plasmas with D-NBI is 2-3 times higher than for H plasmas heated with the same power by H-NBI (FIG. 1(c)).
Therefore, for standard operational conditions the fast-ion content is directly correlated with the isotope mass.
The reason for this is partly the reduced voltage typically used for H-NBI, but also the fast-ion slowing down time which is mass dependent.It is known that ITG turbulence can be stabilised by fast ions in L-mode $[6,7]$ and H-mode $[8,9]$.
For a discharge pair with different isotopes and matched pedestal in AUG non-linear GENE simulations quantify the fast-ion ITG stabilisation. While collisions and EM effects show differences with ion mass, the dominant contribution which allows steeper gradients with D main ions for the same heat fluxes is due to fast ions in AUG (FIG. 1(b)). This explains the possibility for improved core confinement with higher isotope mass when a significant fraction of fast ions is present - typically, $W_{\rm fast}/W_{\rm th}>1/3$.The comparison of simulations for AUG and JET plasmas allows to assess the relative importance of the different contributions to ITG turbulence stabilisation - collisions, EM effects, $E\times B$ shearing and fast ions. The similarities and differences found between the two tokamaks help to understand how each contribution scales.
E.g. the isotope dependence introduced via fast ions should vary with NBI heating power and will be mitigated with increasing density at constant heating power. This is indeed observed for AUG (FIG. 1) and JET.
The NBI power dependence enters because $W_{\rm fast}/W_{\rm th}$ is not constant with power, since $W_{\rm fast}\propto P_{\rm NBI}^\alpha$ with $\alpha > 1$ while $W_{\rm th}\propto P_{\rm heat}^\beta$ with $\beta < 1$.The H-mode pedestal is degraded with lower isotope mass independently of the fast-ion content in the plasma. The reasons for this are the ELM losses/stability or the heat/particle transport or a combination of both.
Linear GENE simulations at the edge of these plasmas reveal drift-wave turbulence to be dominant. Drift waves already exhibited an isotope dependence in L-modes on AUG and JET $[10]$. While decreasing the isotope mass enhances the linear growth rates of drift waves, stronger plasma shaping will reduce them again. While important for drift waves, little impact of $\delta$ on stability is found as expected, because, for the analysed collisionality range the pedestal is mainly ballooning limited.
Drift waves provide a theoretical explanation for the experimental observations of a matched pedestal between differently shaped H and D plasmas. This explanation is supported by measurements with Doppler reflectometers which show a strong reduction of the edge density fluctuation amplitude when increasing $\delta$ in H.Understanding mechanisms that reduce confinement in hydrogen H-modes allows us to counteract them. For the first time H discharges will be run with $\beta_{\rm N}$ approaching the maximum values achieved in D. Thereby, the $\beta$ dependence of $\chi_{\rm H}/\chi_{\rm D}$ as predicted by theory $[3]$ is tested.
Altogether, this will improve the accuracy of predictions towards fusion plasmas with higher $M_{\rm eff}$ and more dominant fast-ion population and will increase the confidence for transferring to D, the ITER results obtained in H.References:
$[1]$ BUSTOS A., et al, Physics of Plasmas 22 (2015) 012305.
$[2]$ SCHNEIDER P.A., et al, Nuclear Fusion 57 (2017) 66003.
$[3]$ GARCIA J., et al, Nuclear Fusion 57 (2017) 14007.
$[4]$ LAGGNER F.M., et al, Physics of Plasmas 24 (2017) 56105.
$[5]$ MAGGI C.F., et al, Plasma Physics and Controlled Fusion 60 (2018) 14045.
$[6]$ BONANOMI N., et al, Nuclear Fusion 59 (2019) 96030.
$[7]$ CITRIN J., et al, Phys. Rev. Lett. 111 (2013) 155001.
$[8]$ GARCIA J., et al, Nuclear Fusion 55 (2015) 53007.
$[9]$ RYTER F., et al, Nuclear Fusion 59 (2019) 96052.
$[10]$ BONANOMI N., et al, Nuclear Fusion 59 (2019) 126025.Speaker: Philip A. Schneider (Max-Planck-Institiut für Plasmaphysik) -
The Gas-Dynamic Multimirror Trap Project
D.V. Yakovlev, P.A. Bagryansky, A.D. Beklemishev, A.V. Burdakov, I.S. Chernoshtanov, A.A. Ivanov, M.S. Khristo, I.A. Kotelnikov, S.V. Polosatkin V.V. Postupaev, V.V. Prikhodko, V.Ya. Savkin, D.I. Skovorodin, E.I. Soldatkina, A.L. Solomakhin, A.A.Molodyk¹, V.I. Scherbakov², S.V.Samoilenkov², A.P.Vavilov²Budker Institute of Nuclear Physics SB RAS, Novosibirsk, Russia
¹S-Innovations/SuperOx, Moscow, Russia
²SuperOx, Moscow, RussiaThe Gas Dynamic Multimirror Trap (GDMT) project represents a next step for research of magnetic plasma confinement in linear axisymmetric systems led by the Budker Institute of Nuclear Physics (BINP). The project’s goal is to assess the feasibility and demonstrate the maturity of the gasdynamic mirror concept for various practical fusion applications ranging from volumetric neutron sources for fusion material tests [1,2] and subcritical hybrid reactor drivers [3] to the power-generating fusion reactors running on advanced weakly activating types of fuel: D-D, D-He³ and p-B¹¹. Among the strong arguments in favor of pursuing this approach are the simplicity of the magnet system, inherently steady state operation without large plasma currents, easy management of plasma heat and particle exhaust, efficient utilization of the confining magnetic field with experimentally demonstrated values of β = 0.6 in GDT experiments [4] and a value of β ~ 1 achieved on 2XIIB [5] and C-2W devices [6]. Recent experimental results and the emergence of new confinement concepts encourage further development of linear systems [7,8].
The updated version of the project, which is being actively developed since 2018, integrates some of the novel ideas about the plasma confinement in linear systems, such as high-β discharge regimes and application of ECRH plasma heating methods [7]. The current version of the project also aims to capitalize on the improvements in magnet and plasma heating technologies and sets higher targets for confining magnetic field (up to 20T) and pulse duration (up to 5s) of high-power-density NBIs developed by BINP [9,10]. The target parameters of the first stage of the machine, which is planned to be constructed by 2024, are summarized in Table 1.
The project is centered around an axisymmetric magnetic mirror trap, which in its basic configuration (fig.1, top) resembles the existing experimental device GDT [11] in BINP. The main difference to the existing experiment is the increased pulse duration of ~ 5s that will allow not only reaching the steady state conditions with regards to plasma processes, but also testing the long-pulse operation of essential subsystems, such as plasma heating systems and cryogenic pumping systems. With the help of up to 10 MW neutral beam injection (NBI) power, one of the main goals of GDMT will be the exploration of advanced high-β discharge modes [12], which are expected boost the energy confinement of the machine well above the basic gas-dynamic scaling of a GDT-based fusion device. The target plasma performance in GDMT ranges from Q ∼ 4% (DT) in the worst-case scenario, when new ideas have little impact, to Q ~ 1 (DT) if the advanced confinement with β ~ 1 proves to be successful [8].
GDMT is being designed as a flexible modular system, which further into the project will be upgraded to test dedicated multimirror and helical tail sections [8] for plasma flux suppression (fig.1, bottom). In its final form with multiple tail sections and increased confinement vessel length, the machine can have total length of up to 70m and extended pulse duration of more than 1000s. Table 1 lists the parameters of GDMT in its basic configuration.
Fig. 1. Initial configuration of GDMT (top). Upgraded configuration of GDMT (bottom).Table 1. Parameters of GDMT (2020)
Length, m 14
Central cell (CC) length, m 6
Plasma vessel diameter, m 1
Mirror field, T 20
CC field, T 0.3–3
CC field rampup rate, T/s 0.54
Mag. energy, MJ 80
NBI energy, keV 30–40
NBI power, MW 10
NBI duration, s 2–5
NBI angle, deg 45
ECRH power, MW < 6
ECRH frequency, GHz 90–130In order to achieve long-pulse (5s) and, eventually, steady state operation (> 1000s) the machine will use a fully superconducting magnet system to create the confining magnetic field. The system will consist of central a confinement solenoid (central cell), compact high-field magnetic mirror coils at both ends of the confinement zone and the additional tail sections for plasma flux suppression. Among the major requirements for the magnet system is the ramp-up operation of the central solenoid, which must be able to increase the magnetic field in the confinement zone from 0.3 to 3T within 5s. The duration of field ramp-up and plasma discharge was selected based on the expected superconductor performance during pulsed operation and the complexity of the power supply. Magnetic compression of a low-field-high-β discharge is proposed as a shortcut to a high-field-high-β discharge, which promises the highest energy confinement time if it can be achieved. In order to maximize the gas-dynamic confinement time in all modes of operation, the magnet system will also feature high-field (20T) magnetic mirrors with cold bore of ~ 240mm. Due to requirements for extremely high magnetic fields and fast ramp-up operation, the magnet system will make heavy use of second generation high-temperature superconductors (HTS) [13] to achieve the required operating parameters. Recent technology development of these superconductors enabled design of magnets operating at ultra high field at less stringent cryogenic requirements. Within this project, we propose the first ever large-volume use of high current HTS superconducting cable for a fusion research installation.
The paper describes these and other features of the machine, outlines the updated experimental program of GDMT and describes the planned near-term experimental and design activity that supports the project.References
1. P.A. Bagryansky, Z. Chen, I.A. Kotelnikov et al 2020 Nucl. Fusion 60 036005
2. B.V. Kuteev, P.R. Goncharov, V.Yu. Sergeev et al 2010 Plasma Phys. Rep. 36 281–317
3. D.V. Yurov and V.V. Prikhodko 2016 Nucl. Fusion 56 126003
4. P.A. Bagryansky, A.V. Anikeev, A.D. Beklemishev et al 2011 Fusion Sci. Technol. 59 1T
5. W.C. Turner, J.F. Clauser, F.H. Coensgen et al 1979 Nucl. Fusion 19 1011
6. H. Gota, M.W. Binderbauer, T. Tajima 2019 Nucl. Fusion 59 112009
7. P. A. Bagryansky, A. G. Shalashov, E. D. Gospodchikov et al 2015 Phys. Rev. Lett. 114, 205001
8. P. A. Bagryansky, A. D. Beklemishev, V. V. Postupaev 2019 J. Fusion Energ. 38 162–181
9. A.N. Karpushov, S. Alberti, R. Chavan et al 2015 Fusion Eng. and Des. 96–97 493-497
10. Yu.I. Belchenko, V.I. Davydenko, P.P. Deichuli et al 2018 Phys. Uspekhi 188 6 595–650
11. A.A. Ivanov and V.V. Prikhodko 2017 Phys.-Usp. 60 509
12. A.D. Beklemishev 2016 Physics of Plasmas 23, 082506
13. S.V.Samoilenkov et al 2016 Supercond. Sci. Technol. 29 024001Speaker: Dmitry Yakovlev (Budker Institute of Nuclear Physics) -
A large effort is currently under way to demonstrate thermonuclear ignition in the laboratory via inertial confinement fusion (ICF) 1. In laser driven inertial confinement fusion (ICF), a spherical capsule of deuterium and tritium (DT) is driven to high velocities by direct irradiation of laser energy (direct drive) or an x-ray bath of an irradiated hohlraum (indirect drive) [2]. At stagnation, the final fuel assembly consists of a low-density (30-100 g/cc), high-temperature (5 -10 keV) core -the hot spot- surrounded by a dense (300 -1000 g/cc), cold (200 - 500 eV) fuel layer - the compressed shell. Such plasma conditions are sufficient for initiating DT thermonuclear fusion when a deuteron and triton fuse to produce a 14.1 MeV neutron and a 3.56 MeV alpha particle. The alpha particle primarily deposits energy in the plasma by colliding with electrons, raising the hot spot temperature and further increasing the fusion reaction rate. This positive feedback cycle is called “alpha heating” and ignition is a direct consequence of the resulting thermal instability. Ignition has yet to achieved in a laboratory plasma and its demonstration is widely viewed as a major scientific achievement with important applications to fusion energy generation and to the stewardship of the nuclear stockpile [3]. Unlike in steady state plasmas, as those envisioned for magnetic confinement fusion [4], assessing ignition in ICF is greatly complicated by the transient nature of implosions and the fact that ignition starts from the central hot region (“hot spot ignition”) and then propagates to the cold and dense surrounding fuel (“burn wave propagation”).
Recent experiments at the National Ignition Facility (NIF) have demonstrated significant alpha heating leading to significant amplifications of the fusion yield close to 3 folds [5]. Despite much work on assessing and measuring the degree of alpha heating, there are two crucial questions still unanswered with regard to ignition:(1) what is ignition in inertial fusion and (2) what fusion yields are required in ICF to claim that ignition has taken place? In this work, we try to answer both questions. We first provide a physical definition of hot spot ignition in ICF and then provide an approximate formula for the fusion energy yield corresponding to the ignition point. The definition of ignition is of general validity for laser fusion and it identifies the onset of the thermal runaway within the hot spot of an ICF implosion just prior to the burn propagation in the dense fuel. It is shown in this Letter that the onset of burn propagation can be uniquely identified through the dimensionless parameter $f_{\alpha}$ which compares the deposited alpha particle energy to the hot spot’s internal energy:
\begin{equation}
f_{\alpha} \equiv \frac{1}{2} \frac{\theta_{\alpha} E_{\alpha}}{E_{hs}},
\end{equation}where $E_\alpha$ is the total alpha-particle energy, $\theta_{\alpha}$ is the fraction of alpha particles deposited into the hot spot, and $E_{hs}$ is the hot spot internal energy at bang time (when the neutron production rate is maximized). In Figure 1, the yield amplification is plotted as a function of $f_{\alpha}$ for a simulation ensemble of 1-D LILAC[6] simulations with different masses, convergence ratios, and hot spot temperatures (turquoise points). The rest of the colored points respectively represent 2-D DRACO[7] simulations of implosions where a density modulation has been applied to the shell's inner surface. Ignition occurs at the critical value $f_\alpha\approx 1.4$ corresponding to a yield amplification due to alpha heating of about 15x to 25x [8,9]. For $f_{\alpha}<1.4$, alpha-heating is mostly confined to the hot spot and the yield amplification depends uniquely on the level of alpha-heating within the hot spot (represented by $f_{\alpha}$). For $f_{\alpha}>1.4$, the burn front propagates into the dense shell which significantly amplifies the fusion producing mass and yield. In this regime, the shell's areal density plays a dominant role in determining the fusion yield enhancement.
The next step is to relate the fusion yield required for marginal ignition (at a yield amplification of 20) to the fuel mass and areal density. In Ref.[10], it was shown that a large enhancement in the fusion yield was correlated to the parameter
\begin{equation}
\chi_{no \alpha} \equiv \left(\rho R \right)^{0.61}\left(\frac{0.12 Yield_{16}}{M_{stag}}\right)^{0.34},
\end{equation}
where $M_{stag}$ is the stagnated DT mass in mg, $\rho R$ is the neutron averaged fuel areal density in $g/cm^2$, and $Yield_{16}$ is the neutron yield in units of $10^{16}$ neutrons. The parameter $\chi_{no \alpha}$ is computed from simulations without alpha transport and $\chi_{no \alpha} \simeq 1$ represents the plasma conditions due to pure hydrodynamic compression which are required for ignition. It follows that we can expect marginal ignition to occur in the neighborhood of $\chi_{no \alpha} \simeq 1$ which allows us to relate the fusion yields of marginally ignited implosions to their designed areal densities and mass. For marginally ignited targets, we obtain the following least squares fit:
\begin{equation}
Y_{ign} (MJ) \approx \left( \frac{M_{stag} (mg) }{0.21} \right)^{0.81} \left( \frac{1}{\rho R (g/cm^2)} \right)^{2.61} .
\end{equation}
This equations provides a formula relating the fusion yield required to claim ignition to the designed fuel areal density and stagnated mass. It is important to note that the lower the areal density implosions require higher fusion yields for ignition.In summary, the ignition condition for inertially confined plasmas has been identified as the transition from thermal instability of the hot spot to propagating burn in the shell. Using a large ensemble of 1D and 2D simulations, we show that this definition of ignition is valid in the presence of asymmetries and differences in shell adiabat and kinetic energy. Ignition corresponds to a yield amplification of 15x to 25x and a value of $f_{\alpha} \simeq 1.4$.
[$1$] J.H. Nuckolls, L. Wood, A. Thiessen, G.B. Zimmermann, Nature 239, 139(1972)
[$2$] S. Atzeni and J. Meyer-ter-vehn, The Physics of Inertial Fusion (Clarendon, Oxford, 2004); J. D. Lindl, Inertial Confinement Fusion (Springer, New York, 1998)
[$3$] J.D. Lawson, Proc. Phys. Soc. London B 70, 6 (1957)
[$4$] J. Ongena,et al, Nature Physics, 12, 398-410 (2016).
[$5$] S. Le Pape, et al, Phys. Rev. Lett. 120, 245003 (2018).
[$6$] J. Delettrez, et al, Phys. Rev. A 36, 3926 (1987)
[$7$] P. B. Radha, et al., Phys. Plasmas 12, 056307 (2005).
[$8$] A. R. Christopherson, R. Betti, and J. D. Lindl, Phys. Rev. E 99, 021201(R) (2019).
[$9$] J. D. Lindl, S. Haan, O. Landen, A. R. Christopherson, and R. Betti, Phys. Plasmas 25, 122704 (2018)
[$10$] P. Y. Chang et al, Phys Rev Lett, 104,135002 (2010).Speaker: Ms Alison Christopherson -
Neutrons along with x-rays emission have been reported in plasma focus (PF) devices, if the filled gas is deuterium 1. The origin of neutron emission is the subject of debate, due to occurring of complex physical phenomena during pinch phase. Most of the PF scientific community believe that neutron production takes place due to beam-target fusion mechanism [2]. Some investigators reported a fraction of thermonuclear neutrons [3]. Neutrons emitted in axial direction were reported with the higher energies than that in radial direction [4] that makes thermonuclear fusion reactions suspicious in PF devices. Both nuclear fusion reactions, the beam-target and thermonuclear are considered at the time of pinch. To estimate the neutron origin time, it is mandatory to take into account all the time delays that neutrons take to reach the detector. If neutrons would have been originated during the pinch phase, the beam-target and/or thermonuclear fusion reactions could be the possible mechanisms. Otherwise, other processes should be included.
To study the neutron origin time, in the present work, the temporal correlations between low, and high-energy x-ray and high-energy x-rays and neutron signals were obtained for a hundred joules plasma focus device (PF-400J). A method, time history analysis (THA) to estimate the time of origin of neutrons and low energy x-rays with respect to high-energy x-rays was developed and applied. Figure 1 shows the variations in neutrons kinetic energies with neutrons origin time (tn,o) – high-energy x-rays (HXR) origin time (txry,o) in axial and radial directions. Figure 1 reveals that the higher energies of neutron correspond to the neutrons that were originated after HXR (tn,o > txry,o). An interesting observation is that neutrons energies increase from the compression to after pinch phases, assuming that the HXR origin time coincide within the pinch. It seems that the longer the time of the acceleration of deuterons, higher the energies of neutrons. Figure 1 suggests that deuterons start to gain energy during the compression phase. It seems, before pinch, Fermi acceleration dominates. During pinch, main source of energy gain will be the induced electromagnetic fields. It can be imagine that during compression not all deuterons that are accelerated by Fermi acceleration participate in fusion processes. During pinch phase, these deuterons further accelerate by the induced electromagnetic fields along with Fermi acceleration inside the pinch. Similarly, not all deuterons that are gaining energy within pinch take part in fusion processes. The remaining deuterons accelerate further, at the moment of the decay of the pinch. The higher neutron energies after the decay of the pinch appeared because deuterons are already energized during compression and pinch phases. However, the dependence of neutrons energy on collision angle and deuterons energy has to be thoroughly study. Present study is limited to experimental observations that are obtained because of time history analysis (THA). Interpretations of THA reveals that in axial direction, neutrons originate mainly before HXR. On the other hand, in radial direction, neutron origin time is mainly after HXR. In addition, it was found that neutrons energies are lower that are produced before HXR and higher that are produced after HXR. This trend is observed in both directions radial as well as axial. In axial direction, neutron energies are fitted exponentially. While, in radial direction second order polynomial fits neutron energies. These fittings suggest that in axial direction, neutron energies increase faster from compression to after pinch phases. On the other hand, in radial direction this increment is relatively slower. In fact, after 10 ns the increment changes much faster, in axial direction. Nonetheless, in radial direction, this change is not visible instead keep growing linearly. At present, the reason for this behavior is not clear. A similar analysis was performed for temporal correlation between low-energy x-rays and HXR. The results of these findings will discussed during the conference.
References
1. J Jain, J Moreno, D Morales, S Davis, B Bora, G Avaria, M J Inestrosa-Izurieta, and L Soto, Laser and Particle Beams 35, 656-662.
2. S.V. Springham, S. Lee, and S.P. Moo Braz. J. Phys. 32, 172 (2002).
3. D. Klir, P. Kubes, M. Paduch, T. Pisarczyk, T. Chodukowski, M. Scholz, Z. Kalinowska, B. Bienkowska, L. Karpinski, J. Kortanek, J. Kravarik, K. Rezac, I. Ivanova-Stanik, K. Tomaszewski and E. Zielinska Plasma Phys. Control. Fusion 54, 015001 (2012).
4. J. H. Lee, L. P. Shomo, M. D. Williams, and H. Hermansdorfer Physics of Fluids 14, 2217 (1971).Speaker: Dr Jalaj Jain (Comisión Chilena de Energía Nuclear) -
Some years ago concepts such as European HiPER gave an opportunity to study in common the physics of the many components for an efficient IFE Reactor. The integral scheme then proposed allows understanding the interconnection among the different components when approaching a realistic design. A common vision for nuclear fusion in general.
The work presented here will remark new research in recent times to get solutions to the plasma facing components (PFC), neutronics assessment, tritium breeding, corrosion, and advanced materials, all of them linked in their responses. In addition, it will include some spin-off from the plasma physics of High Energy Density.
The conditions expected in the PFC materials depend on design and new advanced ones are being designed, manufactured and irradiated based in nanostructures. We have implemented in our laboratory in collaboration with Nano4Energy ® and CIEMAT a set up to cover by sputtering the inner wall of pipes using SiC, and we have optimized the procedure to obtain coating well adhered to the ODS steel (Eurofer). Corrosion and adhesion test are now underway [ref1].
We have proposed [ref2] a novel approach to dynamically manage the efficient tritium breeding ratio (TBR) based on a ceramic breeder (Li2TiO3) with a Be multiplier and a neutron reflector consisting in a water tank (or heavy water) which is responsible through a simple system or filling/emptying such tank to tune the TBR.
The irradiation of spherical gold nanoparticles with nanosecond laser pulses induces shape transformations yielding nanocrystals with an inner cavity. The concentration of the stabilizing surfactant, the use of moderate pulse fluences, and the size of the irradiated particles determine the efficiency of the process and the size of the void. Hollow nanocrystals are obtained when molecules from the surrounding medium (e.g., water and organic matter derived from the surfactant) are trapped during laser pulse irradiation. These experimental observations can have very interesting prospects for the development of hollow plasmonic nanoparticles with potential applications in materials [ref3].
A large effort is developed in the assessment and qualification of a sophisticated modelling system for neutronic calculations and the associated responses, which is actually applied successfully to ITER, IFMIF-DONES and DEMO facilities [ref4,ref5]. This is a goal to be applied also to next concepts in Inertial Fusion Energy.
![Integral Computational model for Neutronics calculations [ref4, ref5].](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/1365/2.jpg)
Suites of codes have developed to model multiphysics experiments related to such fields as laboratory astrophysics and plasma based soft X-ray lasers among others. The hydrodynamic behaviour of laser created plasmas for these purposes is modelled using ARWEN code. The amplification of XUV radiation through plasmas is modelled using our codes 1D-DeepOne and 3D Dagon that solve the Maxwell-Bloch equations in time domain, obtaining the resulting wavefront and temporal structure of the amplified radiation. Finally, EMcLAW is a 3D Maxwell solver with adaptive mesh refinement, which is perfectly suited to model the propagation of short pulses through long distances.
![Comparison of a 3D astrophysical simulation (ASTRO) with the code SPHINX and a 2D laboratory astrophysics simulation (LAB) with ARWEN [ref6].](https://nucleus.iaea.org/sites/fusionportal/Shared%20Documents/FEC%202020%20Images/1365/3.jpg)
REFERENCES
[ref1] T. Hernandez, R. Gonzalez-Arrabal, et al., submitted to Nuclear Materials and Energy (2020).
[ref2] A. Fierro et al., accepted in Fusion Engineering and Design (2020).
[ref3] G. González-Rubio, et al., J. Phys. Chem. Lett. 11 (2020) 670–677.
[ref4] P. Sauvan et al., IEEE Transactions of Nuclear Science 63 (1), 375-384.
[ref5] P. Sauvan et al., Fusion Engineering and Design 151, 111399.
[ref6] García-Senz et al, The Astrophysical Journal 871:177 (2019).Speaker: Jose Manuel Perlado (Instituto de Fusión Nuclear “Guillermo Velarde” / ETSII / Universidad Politecnica de Madrid)
-
-
-
This paper presents solutions for critical problems in Japan’s DEMO (JA DEMO), which include common DEMO design issues beyond ITER-relevant technologies. The highlights of this design study are (i) system design for electric power generation, (ii) remote maintenance concept to attain a high plant availability, and (iii) novel concept for water-cooled pebble bed blanket and tritium recovery. The proposed concept as JA DEMO will be the foundation for Japan’s DEMO that can be envisioned in the next stage of ITER.
1. Introduction
The pre-conceptual design of JA DEMO has been proposed by adding the technology outlook of ITER and JT-60SA. JA DEMO has a relatively large major radius Rp = 8.5 m for long-pulse operation and a relatively low fusion power Pfus = 1.5 GW for divertor heat removal [A]. The cooling water for DEMO is operated at the PWR (Pressurized Water Reactor) water conditions of 15.5 MPa and 290 ºC-325 ºC. In this paper, we state the following three issues of JA DEMO design; (i) in order to evaluate the electric power output, it is necessary to clarify the design of the cooling system and the heat balance of the primary heat transfer system (PHTS), and the power consumptions of the pumping power in the PHTS, the freezing power for the superconducting coil, the injected heating power for plasma operation, etc., (ii) in order to develop a remote maintenance scheme (RMS) for the replacement of large components in the high dose area, the movement line of remote handling (RH) device must be arranged to maintain the soundness of the device based on the shutdown dose map of JA DEMO, and (iii) in order to develop a blanket concept with the capabilities of high tritium (T) productivity and pressure tightness, it is necessary to obtain a high packing factor up to 80 % with binary packing and to recover the T generated in the packing area without retention. Approaches to resolve these design issues and remained challenges are presented in the following sections.2. Evaluation of electric power output
The electric generation systems for JA DEMO is developed without intermediate heat exchanger. This is because T permeation through the steam generator to the secondary cooling system in the PHTS is found to be less than the restricted amount of T disposal for pressurized water reactors in Japan [B]. Here, the generator output was evaluated to be 640 MW with the thermal efficiency 34.4 %. The power consumption was analyzed to evaluate the electric power output. To evaluate the pumping power in the PHTS, the intricate cooling system was designed in consideration of suitable configurations of the systems for the PHTS such as steam generator, pressurizer, etc. as shown in Fig. 1. The pumping power of the cooling system in the PHTS was evaluated to be 71 MW. Since the design concept of the superconducting coil is basically same with to that of ITER, it is determined by scaling the freezing power in ITER as 91 MW. In the NBI system of the heating and current drive (H&CD), the neutralization efficiencies of the photon and gas neutralizers are assumed to be 90 % (theoretical value) and 56 % (ITER technology), respectively. Here, the neutralizer method makes a difference in the power consumption of 118 MW. Hence, the total power consumption was found to be 386 MW with the photon neutralizer, and the electric output was evaluated to be 254 MWe as summarized in a table of Fig. 1.
3. Remote maintenance scheme in the high dose environment
The RMS is as follows for the JA DEMO: after cutting of a cooling pipes in the maintenance ports, the blanket segments and divertor cassettes integrated with shielding plugs (SP) are replaced through vertical upper ports and bottom ports, respectively, as shown in Fig. 2. The major tasks in RMS for JA DEMO have been how to replace large-scale components and how to implement the maintenance work in the high radiation environment. In the previous studies, we had focused mainly on the construction method for stably replacing large-scale components which weighs 100 tons and stands 10 meters high as shown in Fig. 2. We found that if plant availability was expected to reach about 70% which is the requirement of JA DEMO, the blanket segments must be replaced in parallel from 4 out of 16 ports. We also found out the characterizations of RMS in the high radiation environment [C]. By arranging SPs, the dose rate of the points for cutting and re-welding cooling pipes in the maintenance ports can be reduced to about ~ 0.1 Gy/h as shown in Fig. 2(a). However, when the outboard blanket segments and the divertor cassettes with the SP were removed together, the dose rate in these maintenance ports increases to 100 Gy/h at the maximum as shown in Fig. 2(b). Here, the dose rate in the vacuum vessel of ITER during maintenance is evaluated to be 250Gy/h according to the requirements for remote handling of the shielding blanket. Therefore, maintenance techniques developed for ITER such as pipe cutting or re-welding are applicable to the JA DEMO in terms of the spatial dose rate.
4. Blanket concept and tritium recovery
The conceptual design of the breeding blanket of honeycomb structure, which has pressure tightness against in-box loss-of-coolant accidents in a water-cooled solid breeder, has been developed as shown in Fig. 3(a). As honeycomb structure is higher in pressure tightness than square prism structure, the area for filling the mixed pebbles breeder of Li2TiO3 pebbles and Be12Ti ones can be enlarged. From the results of the 3D neutronics analysis, the target of the overall TBR (>1.05) would be achievable. To achieve the TBR target, it is necessary to fill the packing factor to 80% by binary packing. However, there is a concern that the amount of T retained in the breeding area may increase due to pressure drop. To avoid T retention in the breeding area, the flow of He-purge gas is arranged by using three inflow points of He-purge gas near the first wall. Then the retention of T was resolved as shown in Fig. 3(b).
References
[A] Y. Sakamoto et al., 27th IAEA Int. Conf. on Fusion Energy (2018) FIP/3-2
[B] R. Hiwatari et al., Fusion Eng. Des. 143 (2019) 259-266
[C] Y. Someya et al., Fusion Eng. Des. 124 (2017) 615-618Speaker: Youji Someya (National Institutes for Quantum and Radiological Science and Technology) -
The challenge
A power plant based on a tokamak architecture has a magnetic cage with thick shielding, so maintenance needs to be conducted through long narrow access ports (see figure 1). This means unprecedented dexterous handling of massive flexible components, which is difficult, slow and expensive. For this reason, industrial plant to date has always been arranged to allow the heaviest lifting to be done from above by a crane.
Power plants also need to achieve higher availability than has been achieved with existing experimental machines. This drives the need for simple and robust maintenance schemes on maintenance-compatible plant architecture. This makes maintenance design driving, which is why it must be understood early in the plant design.
Examples of design driving maintenance considerations are the segmentation of the in-vessel components and the layout of the pipework, components and transport corridor in the ports. These aspects are critical to maintenance, but the choice also impacts the design of the components and indeed the layout of the whole machine and ex-vessel systems.
The size and mass of the in-vessel components for DEMO are significantly higher than for existing machines, including ITER, and the radiation and temperature will be even more challenging. This means DEMO will require novel maintenance solutions and technologies with a different power plant architecture to existing machines. The ITER approach of assembling in-vessel maintenance equipment and handling a high number of components is not appropriate for DEMO and would not be able to meet the plant availability requirements. The same is true for the JET approach of performing all maintenance operations using booms deployed through equatorial ports.
However, there is a large technology overlap with ITER and with wider industry and this will be used wherever possible because of the cost savings, the reduction in technical risk, the data and demonstrated reliability and it is important to standardise the remote maintenance systems as far as possible to reduce the time and cost of design and testing. This applies to the handling systems and movers, but also to the tools and all the interfaces, including handling, mechanical attachment, electrical, signal and fluid (pipes).
Where there is no technology overlap, novel solutions and enabling technologies must be developed to meet the powerplant requirements. There is significant technical risk associated with enabling technologies that are essential to the chosen architecture because if the requirements cannot be achieved then changes to the architecture would be required. The technical readiness level for these enabling technologies must be increased in step with the developing design, to maintain an acceptable risk level. Examples of these enabling technologies for DEMO include the in-bore cutting and welding that is required for the closely grouped pipes on the plasma facing components (breeding blankets and divertor cassettes), and the dexterous handling systems for the massive breeding blankets and divertor cassettes.
In-bore cutting and welding
The EU has been developing miniature laser cutting and welding tools (see figures 2 and 3) that can be deployed down the bore of pipes to demonstrate the feasibility of the enabling technology [1., 2.]. Laser has the advantages of speed, and, as a non-contact process, it has better recovery and rescue options than existing technologies, such as TIG arc welding. Initial trials have been successfully conducted on proof-of-principle tools. However, further development is required to increase the laser power and ultimately to produce a qualified weld process.
Trials have also been conducted on pipe alignment (see figure 4) and tool deployment (see figure 5) which have shown that the alignment forces are high and would be very challenging for the pipe handling systems. Further work is required to develop an alignment system that is less demanding of the handling system.
Dexterous handling of massive components
The layout of DEMO requires a mover that can fit in the port and dexterously handle the in-vessel components that can weigh up to 80 tonnes. A design has been produced for a blanket handling system (see figure 6) but it has limited reserve factors on the working loads, cannot sustain the seismic load case and the stiffness and dynamic performance of the mover would make it difficult to control, leading to very slow operation and lack of accuracy.
New work is starting on the maintenance systems for different tokamak configurations that have split blankets and a double-null. The design process requires time intensive iterations between design tools such as structural analysis, kinematic evaluation and the dynamic and control models. Work is underway to allow these tools to communicate directly with each other and ultimately to integrate them into a single tool. This virtual engineering will allow much more rapid development and optimisation to support the evolving plant design.
Port layout
The layout of the port is also critical to the plant design. It is difficult to arrange all the components and services in a way that they can be maintained whilst retaining the transport corridor for the large plasma facing components that must be transferred out of the vessel. Work on DEMO has shown that modified industrial handling equipment can be used for maintenance. For example, robots, slideways and scissor lifts. This means the technology risk is low, but the integration risk remains high and the layout of the equipment is critical to success. A small change in the pipe layout can make it unmaintainable. A model of the port layout has been created to enable the visualisation required to create the maintenance strategy and design the equipment, and this has been animated to show the sequence of operations and communicate the constraints to the plant designers (see figure 7).
Looking ahead
The technical risks inherent with the novel maintenance solutions required for a fusion power plant have been well recognised across the world. In the EU, a seven-year programme has been proposed to analyse the maintenance implications of the EU DEMO plant design to enable plant option down-selection and to develop the enabling technologies that are not available from industry or ITER.
The focus of the programme is to develop the design of remote maintenance equipment integrated with the plant design, to reduce or remove key risks. Within this programme, designs are required to allow feasible maintenance of the breeder blankets and other in-vessel components. As the system designs become more detailed the programme will produce proof-principle equipment to validate designs and virtual engineering tools to reduce the key technical risks.This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 un-der grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
[1.] “Service Joining Strategy for the EU DEMO”, T. Tremethick et al, ISFNT-14 (2019).
[2.] “Laser cutting and welding tools for use in-bore on EU-DEMO service pipes”, K. Keogh et al, Fusion Engineering and Design, volume 136, part A, November 2018, pages 461-466.
Speaker: Mr Oliver CROFTS (UKAEA) -
According to the European strategy to fusion energy, the development and the operation of a demonstration power plant (DEMO) is foreseen as the single step between ITER and a commercial tokamak fusion power plant (FPP). DEMO is required to feature all key systems and components of an FPP and to comply with a set of general goals [Donné, 2018]. These goals include a few hundred megawatts electric power generation, a closed fuel cycle and long pulse (or steady state) plasma operation.
The conceptual design of DEMO begins with the quantitative definition of these main goals and shall proceed by selecting the major reactor parameters. The approach adopted in the EU-Programme [Federici, 2018] follows different steps, iteratively repeated until a certain grade of satisfaction, consistency and attractiveness is met (see Figure 1, black solid line path). Hence, the beginning of the actual engineering design hinges upon the verification of these conditions.
Dedicated computational tools - referred to as systems codes - are deployed to produce a reactor baseline, upon definition of reactor requirements, constraints and architectural features. The systems code output is transferred to the design codes, normally in form of a 2D reactor sketch and major reactor parameters (e.g. radial build, fusion power, major radius and magnetic field). Presently available fusion systems codes, such as PROCESS [Kovari, 2014], aim at exploring one (or more) reactor configurations that simultaneously fulfill the plasma physics operational limits, the engineering constraints and the plant general goals. In general, they rely on rather basic physics and engineering models (mostly at zero or one-dimensional level). The design codes, instead, are very detailed but run on much longer computing times and model limitations.
Due to the broad multi-physics and multi-scale spectrum involved within the design process, it is rather challenging to maintain data consistency and to keep an expedite design flow. In turns, wide modelling gaps between systems codes (0D/1D) and design codes (3D) might slow down and hinder the design loop, thereby increasing the number of iterations. To this end, the connections between system and design codes can be consolidated by complementing the systems codes by means of a more refined and intermediate system analysis tool (see Figure 1, red-dashed line path): this step is referred to as MIRA, Modular Integrated Reactor Analysis.
MIRA [Franza, 2019] is a high spatial resolution design tool developed at KIT, incorporating the physics and the engineering insights of the utmost domains of tokamak reactors. MIRA relies on a modular structure and provides a FPP baseline. With reference to the flowchart of Figure 2, it incorporates into a unique computing environment a mathematical algorithm for the following problems:
- 2D radial/vertical geometry and material characterization of physical components, i.e. plasma, breeding blanket (BB), vacuum vessel (VV) and toroidal field (TF) and poloidal field (PF) coils;
- 2D free-boundary plasma magnetic equilibrium;
- constrained gradient-based PF coil solver, incorporating plasma shaping requirements and coils engineering constraints for all plasma operational phases, i.e. breakdown, start of flat-top (SOF) end of flat-top (EOF);
- plasma and divertor physics modules based on one-dimensional ion, electron and impurity transport [Fable, 2018];
- quasi 2D neutron and photon core radiation transport for the evaluation of breeding (TBR), heating and shielding capabilities of all in-vessel reactor components;
- 3D magneto-static modelling of TF and PF coils elements and engineering characterization of superconducting cables;
- Plant integral modelling and power performances.
Compared to presently available system codes, MIRA is based on a higher mathematical sophistication, engaging systems analyses up to three-dimensional space resolution. This allows scoping multiple reactor configurations with a more consolidated modelling granularity at a component level and with a holistic view of the entire plant, thereby leaving room for design modifications at higher degree.
The MIRA approach has been applied to the DEMO baselines 2015 [Franza, 2019] and 2017, generated by means of the PROCESS code. The analyses have been carried out by taking an identical set of input assumptions and requirements (e.g. same fusion power, major radius and aspect ratio) and observing the response on the key design targets and the imposed operational limits. Based on more accurate MIRA analysis, both baselines have been found in conflict with some of these limits, in particular on plasma burn time (33 % below the two-hour goal) and maximum TF ripple on the plasma boundary (13 % above its upper limit). The major causes for these discrepancies between PROCESS and MIRA are attributed to the reduced space resolution, to the modelling simplifications and to limited engineering capabilities of PROCESS. In PROCESS, the burn time is based on simple 0D magnetic flux conservation and peak magnetic field in the central solenoid (CS), whilst the TF ripple is calculated from predefined scaling laws. In MIRA, instead, the burn time derives from 2D calculation of poloidal flux profiles $\Psi(r,z)$ at SOF and EOF magnetic equilibrium configurations (see Figure 3) with plasma shaping requirements ($\partial \mathcal{D}_p^t$ target shape) and coil technological limits. The TF ripple is computed via 3D magneto-static analysis.
The outcomes of the DEMO 2015 and 2017 baselines analyses have illustrated that modelling simplifications, affecting the state-of-the-art systems codes, affect considerably the overall design of the reactor. Therefore, the application of the MIRA approach to analyse a DEMO baseline can mitigate the lack of modelling instruments before engaging the design codes analyses (Figure 1). Accordingly, a set of active measures has been addressed to steer some identified reactor geometric variables in favour of a design point that fulfils the imposed constraining conditions. Such measures involved a parameter scan on the inboard BB radial width and outboard TF coil radial extension. Apart from fulfilling the lower and upper bounds on burn time and TF ripple, the addressed parametric studies have shown also non-trivial inter-parametric dependencies, never explored in fusion system analyses. For instance, a reduced thickness for inboard BB and TF coil leg has been identified in connection with plasma burn time, TBR requirements and coils technological limits (both in CS and TF coils superconducting cables).
In conclusion, this work poses new basis to designing a tokamak reactor and to parametrizing multiple technological solutions. Accordingly, a deeper and a more centralized multi-physics reactor analysis can speed up and improve the whole design process.
References
[Donne, 2018] T. Donne, W. Morris, X. Litaudon, C. Hidago, D. McDonald, H. Zohm et al. European Research Roadmap to the Realisation of Fusion Energy. EUROfusion, November 2018.ISBN: 978-3-00-061152-0.
[Fable, 2018] E. Fable, C. Angioni, M. Siccinio, H. Zohm. Plasma physics for fusion reactor system codes: Framework and model code. Fusion Engineering and Design 130 (2018) 131–136.
[Federici, 2018] G. Federici, C. Bachmann, L. Barucca, W. Biel et al. DEMO design activity in Europe: Progress and updates. Fusion Engineering and Design, vol. 136, pp. 729 – 741, 2018
[Franza, 2019] F. Franza. Development and Validation of a Computational Tool for Fusion Reactors System Analysis". PhD thesis, Karlsruhe Institute of Technology (KIT), June 2019. DOI: 10.5445/IR/1000095873, https://publikationen.bibliothek.kit.edu/1000095873
[Kovari, 2014] M. Kovari, R. Kemp, H. Lux, P. Knight et al. PROCESS: A systems code for fusion power plants - Part 1: Physics. Fusion Engineering and Design, 89:3054–3069, 2014.
Speaker: Dr Fabrizio Franza -
As the EUROfusion EU-DEMO design programme approaches the transition between the pre-conceptual and conceptual design phase the systems code PROCESS has been improved to incorporate more detailed plasma physics, engineering and integration models. Unlike many systems codes PROCESS combines the physics modelling with both technology and costs analysis. Key to the conceptual design phase are detachment, toroidal field magnet design, double-null power sharing, operational sensitivity and economic uncertainty analysis. All of these have been integrated into PROCESS [1.], [2]. During the pre-conceptual design systems codes are an essential tool for exploring fusion power plant concepts. They allow one to model the interaction of the plant systems and quickly perform reactor optioneering. To be able to carry out these large scoping studies the fidelity of the models can be restricted to reduce the computational time. The EUROfusion EU-DEMO baseline designs are created using the systems code PROCESS and the ability to measure these trade-offs has led to important design choices being examined during the DEMO pre-conceptual design phase. Ruling out unfeasible designs allows EUROfusion to efficiently identify where in the design space to carry out detailed design work. This contribution describes how PROCESS has been retooled for EU-DEMO conceptual design.
To allow optimisation of the underlying plant systems for a more fixed plant design greater detail had to be integrated into PROCESS. For EU-DEMO a 1-D scrape-off layer (SOL) model has be implemented to capture the power loss mechanisms in the SOL, to validate the core plasma power balance and to determine if the plasma is in a detached state – a key design requirement for EU-DEMO [3]. In combination with the SOL model PROCESS can now allow power sharing in a double-null configuration. The choice of single versus double-null is a fundamental choice for EU-DEMO so capturing the behaviour is essential for a systems code. A 1-D plasma transport solver has been integrated into the code to produce a self-consistent plasma model with plasma profiles for correctly calculating heating and current drive power deposition and determining the plasma radiation by integrating over the profile [4].
As the design space of EU-DEMO becomes smaller there is a need to understand what the sensitivity of the design is given some uncertainty on the performance and engineering parameters. PROCESS is an ideal tool for this analysis due to its breadth of scope and computational speed. It has given insight into the likelihood of a given EU-DEMO design achieving the high-level goals of EUROfusion, such as reaching the net electric power target [5] (the same analysis has been used on CFETR [6]). The PROCESS uncertainty tools have been used to analyse the cost sensitivity of DEMO designs to determine the primary cost drivers. This information will contribute to the decisions during concept down-selection.
One of the primary drivers of machine design, performance and cost are the superconducting magnets. Therefore, correctly calculating the space required, the achievable field and cost is essential for PROCESS. High temperature superconductors (HTS) can potentially offer a performance, engineering and cost benefits. A REBCO (rare earth barium copper oxide [7]) HTS model has been written for PROCESS for the TF coils. The operating temperature of the TF coil for both LTS and HTS is 4.5 K for the analysis presented here as it is often preferable to go to higher field to achieve large net electric power, as the fusion power is proportional to $\beta^2 B^4$. This is done rather than increase the operating temperature to save on electrical power needed for the cryogenic system or use an alternative to helium as a coolant.
PROCESS has been used to analyse the impact of toroidal field coil stress on machine design with LTS [8] and can now compare with HTS. Figure 1 shows the effect of the allowable Tresca stress in the TF coil steel. The LTS model includes a quench calculation with a variable copper fraction, while the HTS model imposes a maximum superconductor current per unit area of copper, chosen as 100 A/mm2 or 200 A/mm2. PROCESS was set to minimise the major radius and to produce 500 MW net electric power for 2 hours. Figure 1 shows that one can achieve higher fields at smaller machine size with HTS. The reduction in major radius depends on the copper requirement and is in the range 0.25-0.5m. At higher allowable stress the HTS PROCESS runs start to prioritise smaller machine size over further increasing the field.
Further detail will be added to reduce uncertainty in the models and allow more robust design scoping studies, such as an equilibrium solver. All systems codes will need to be comprehensively rebuilt to make them relevant for the conceptual design phase. The UKAEA power plant technology group is revising PROCESS to make improvements and collaboration easier.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053 and from the RCUK Energy Programme [grant number EP/T012250/1]. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
[1.] M. Kovari, R. Kemp, H. Lux, P. Knight, J. Morris, and D. J. Ward, “" PROCESS " : A systems code for fusion power plants-Part 1: Physics,” Fusion Eng. Des., vol. 89, no. 12, pp. 3054–3069, Dec. 2014.
[2] M. Kovari et al., “‘PROCESS’: A systems code for fusion power plants - Part 2: Engineering,” Fusion Eng. Des., vol. 104, pp. 9–20, Mar. 2016.
[3] J. Morris, M. Kovari, N. Asakura, and Y. Homma, “Comparison of the systems code PROCESS with the SONIC divertor code,” IEEE Trans. PLASMA Sci., 2020.
[4] E. Fable, C. Angioni, M. Siccinio, and H. Zohm, “Plasma physics for fusion reactor system codes: Framework and model code,” Fusion Eng. Des., vol. 130, pp. 131–136, May 2018.
[5] H. Lux et al., “Uncertainties in power plant design point evaluations,” Fusion Eng. Des., vol. 123, 2017
[6] J. Morris, V. Chan, J. Chen, S. Mao, and M. Y. Ye, “Validation and sensitivity of CFETR design using EU systems codes,” Fusion Eng. Des., vol. 146, 2019.
[7] R. Heller, P. V. Gade, W. H. Fietz, T. Vogel, and K. P. Weiss, “Conceptual Design Improvement of a Toroidal Field Coil for EU DEMO Using High-Temperature Superconductors,” IEEE Trans. Appl. Supercond., vol. 26, no. 4, pp. 1–5, Jun. 2016.
[8] J. Morris, R. Kemp, M. Kovari, J. Last, and P. Knight, “Implications of toroidal field coil stress limits on power plant design using PROCESS,” Fusion Eng. Des., vol. 98–99, 2015.Speaker: James Morris -
The 2018 National Academies of Sciences (NAS) Report of the Committee on a Strategic Plan for U.S. Burning Plasma Research and the more recent APS-DPP Community Planning Process (CPP) recommend that the U.S. should pursue innovative science and technology to enable construction of a Fusion Pilot Plant (FPP) that produces net electricity from fusion at reduced capital cost. Such a mission requires development and integration of multiple physics and technology innovations. A unique feature of the tokamak-based FPP approach in the U.S. is the integration of a high fraction of self-driven current with high core plasma pressure and high divertor parallel heat flux. The integration of a high-performance core and edge has not been previously accessed nor is it presently planned to be accessed by existing U.S. or international facilities. This integration is sufficiently challenging that construction and operation of a dedicated sustained-high-power-density (SHPD) tokamak facility is proposed by the U.S. community to close this integration gap. This presentation describes the performance of present and planned tokamak facilities, gaps between present/planned facility capabilities and the U.S. FPP regime, and presents pre-conceptual engineering studies for a SHPD facility to narrow/close the core-edge integration gaps to the FPP.
Figure 1 compares (left) pressure and (right) parallel heat flux between (blue) present/planned tokamaks and (red) proposed FPP/FNSF/SHPD devices (2,3,4). Figure 1 shows the FPP regime has plasma pressure = 0.2-0.7 MPa in the range of high-performance pulsed tokamaks (e.g. ITER baseline, EU Demo, Italian DTT, ADX, SPARC) but with approximately 3 times higher self-driven fraction enabling long pulse to steady-state operation. Present stellarators are shown in the left graph at f$_{self}$ = 1 and have q$_{||0} < $1 GW/m$^{2}$ so are not included in the right-hand graph. The FPP-regime pressure and heat flux are also 3 times higher than expected in present/planned facilities at high f$_{self}$ =0.6-0.8. SHPD aims to close these gaps through innovations such as: high field/current density/temperature superconducting magnets, advanced divertors, elevated confinement, and enhanced-efficiency current drive actuators.Systems studies have been carried out for a range of major radius R = 1 to 2m and aspect ratio A = 1.6 to 4 to identify SHPD facility sizes and powers that can access FPP-level plasma pressure and edge heat fluxes. These studies include aspect-ratio-dependent elongation and beta limits (4), assume toroidal field winding pack current density J$_{WP}$ = 70MA/m$^2$, and FPP-level elevated H$_{98}$ ≥ 1.5 sufficient to operate at the nominal n=1 no-wall kink stability limit (4).
The table at the top of Figure 2 lists the assumed variation of heating power versus R. The powers are chosen to have H$_{98}$=2 at A=2.4 as A is varied. Figure 2 shows (a) the magnetic field increases from 1.5-3.5T at A=1.6 to 7-9T at A = 4, (b) the Lawson parameter is maximized for A=1.8-2.4, (c) FPP-level pressures are accessed for R = 1.2m between A = 2-3, and (d) FPP-level heat fluxes are most readily accessed for R ≥ 1.4-1.6m. However, by increasing heating power to P$_{heat}$ = 50MW at R = 1.2-1.4m it also possible to access low-A FPP and A=4 FNSF-level (2) heat fluxes at both A = 2.0 and A=2.4 as indicated in the right-hand side of Figure 1. Accessing ARC-level (3) heat fluxes at A=3 requires more power and/or a larger major radius SHPD facility.
High-Z solid tungsten divertors and first-walls have the highest technical readiness level and are the near-term option for fusion development. However, high-Z solids have several challenges, including plasma-facing component damage from erosion and re-deposition and neutrons, and high-Z impurity accumulation and associated core plasma radiative collapse. Low-Z (Li) and high-Z (Sn) liquid metals (LM) are increasingly being studied as a means of addressing these challenges. Figure 3 shows a R=1.2m, A=2.4 double-null SHPD concept with a small central solenoid for IP ramp-up and sufficient in-vessel space and flexibility to study a range of divertor effects including: radiation, detachment, leg length, closure, advanced magnetic geometries, and solid and/or liquid metals all to facilitate the SHPD core-edge integration research mission.(1) R.J. Goldston et al, Plasma Physics and Controlled Fusion 59 055015 (2017)
(2) C.E. Kessel et al, Fusion Science and Technology 68 225-236 (2017)
(3) B.N. Sorbom et al, Fusion Engineering and Design 100 378 (2015)
(4) J.E. Menard et al, Nuclear Fusion 56 106023 (2016)*Research supported by the US DOE Contract No. DE-AC02-09CH11466.
Speaker: Jonathan Menard (Princeton Plasma Physics Laboratory) -
Spherical Tokamak reactor (STR) is attractive due to its inherent capabilities such as disruption avoidance, natural elongation, natural divertor and high beta capability, apart from a smaller size, with presumably lower costs [ 1, 2]. There has been an extraordinary evolution from the early concepts like SMARTOR [ 3] with devices like START, NSTX, MAST, GLOBUS-M and a number of others with the HTS based future devices like STEP [4]. Given the pace of development of the new superconducting materials [5,6] and the new divertor concepts [7,8,9], the STRs represent a rapidly developing front and may very well be realized not far in the future. Following an elegant paper by Peng et al. in 1986, a range of compact reactor designs ($R$ and $P_f$) has emerged, e.g. FNS-ST (0.5m, 10 MW), DTST (1.1m, 30-60 MW), ARC (3.3m, 525 MW), SlimCS (5.5m, 2950 MW), ARIES-ST (3m, 2980 MW) with a variety of objectives like, neutron source, component-test-facility (CTF) and power plant [10,11,12,13,14]. However, while the high neutron loads are welcome for reactor economics, the size reduction comes at a penalty of extreme heat loads on the divertor with concomitant engineering challenges [15]. Several designs of STRs are currently being developed around the world with scoping studies and available data from currently operating tokamaks as well as other experimental/dedicated test facilities and insights from experts [16]. This paper brings out the role of constraints arising from steady-state power balance and core-radiation. It is argued that the core-radiation plays a crucial role in the reactor design, as it not only restricts the accessible parameter-space but also determines the limits on impurity accumulation [17]. A comprehensive physics-design study [18] shows that about 50$\%$ of the heating power needs to be lost by core-radiation. Such considerations can impact stability as well [19]. In the following, the ST-parameter space ($R-B_t$) is analyzed to elucidate the limits posed by the various constraints. For $T_i$ from 6 to 20 keV, the fusion power (MW) may be approximated for analytic purposes as:
$$P_{F}=0.026 \frac{(S_n+S_T+1)^2}{(2S_n+2S_T+1)} \frac{\kappa {\beta_{N}}^2 {S_{\kappa}}^2}{q^2 A^4} R^3 {B_t}^4$$ where $q=5 R B_t S_k/(A^2 I_p)$ is the safety factor, $I_p$ is the plasma current in MA, $A$ is the aspect ratio and $S_k$ is the shape factor. $\beta_N = \beta a B_t/I_p$ and $S_n$, $S_T$ are the exponents for the parabolic profile of the density and temperature respectively. The stored energy in MJ can be expressed as: $$W_\beta = \frac{\pi}{8}\frac{\kappa S_\kappa}{q A^3} \beta_N R^3 {B_t}^2$$ In steady-state, where the power from $\alpha$-particles and the externally injected power are balanced by the transport losses, the power-balance is given by $W_\beta = P_L \tau_E $, where $P_L$ (defined as $P_H (1-f)$) is the power reaching the edge, after a fraction $f$ of the power deposited $$P_H = P_\alpha + P_{ext}= P_F (1/5+1/Q)$$ is radiatively lost from the core region. It is assumed that the ITER-IPB(98,y2) scaling holds good, although it is likely to be more favorable in reality [20]: $$\tau_E = 0.0562 H_h {I_p}^{0.93} {B_t}^{0.15} {n_{19}}^{0.41} R^{1.97} \kappa^{0.78} \epsilon^{0.58} M^{0.19} P_L^{-0.69}$$ The power-balance can then be written as: $$Q_{LF} = (f_\alpha /5 + 1/Q )(1-f)$$ where $f_\alpha$ is the fraction of $\alpha$-particles which transfer their energy to the plasma. The $Q_{LF}$ is actually the ratio $P_L/P_F$ and is an involved expression with fractional powers of plasma parameters. To understand its dependencies, it is best approximated as: $$\frac{\beta_N \ A^{14/5}\ q^{6/5}}{{B_t}^{92/35}\ H_h^3 \ f_G^{6/5}\ S_k^{16/5} \ M^{3/5}\kappa^{2/5} \ R^{9/5}}$$ where, the nearest integer ratios are used to approximate the exponents in the expression for $\tau_E$. The radiated power fraction $f$ can be expressed in terms of $Q_{LF}$. Its role in accessibility constraints in the $R$-$B_t$ space has been shown in Fig.1, where, the contours of constant $P_f$ are shown along with the limits on achievable $B_t$ assuming either copper or HTS peak current-density in the center-stack. The constant fusion contours intersect increasingly high divertor load curves as one makes the reactor more compact. The dotted curves ($f$=0, 0.5 and 0.94) correspond to the power balance constraint. The $f=0$ curve shows the limit of 'no core-radiation' and thus represents the lower boundary of physically acceptable solutions. Thus, for a given set of parameters as an example ($q$=3, $\kappa=2.5$, $\delta=0.3$, $\beta_N=5$, $Q=5$), there exists an upper limit on the value of $R$ (3m). The two $Q_{LF}$ curves that 'bracket’ the fusion power curve, define the accessible space until the limit on achievable $B_t$ is encountered. An example of a design point (R=1.25 m, Bt=2.8 T, Pf = 200 MW) has been shown (red dot). It may not be possible to meet it unless almost 60$\%$ of the heating-power is radiated from the core. Such constraints make it necessary to examine how much core concentration of impurities would be acceptable. Fig.2 shows impact of $Q$ in the parameter space -- higher values reduce the available space in the lower left-hand corner. This has implications for the reactors which may operate at modest values of $Q$ (CTF or fusion-fission hybrid, fissile material converters or radioactive waste processing, or just fusion-science devices). At the same time, the higher $Q$ demand from power reactors (to remain cost-competitive and investment-attractive), eliminates a large space and pushes accessibility points further up. An important consequence of the power balance constraint is that the divertor heat load (transported power) $P_{div} \approx B_t^{3/2}/R^{4/5}$. The gradients of $P_{div} \approx$ constant are in dramatic contrast to those of constant neutron load contours, so while the neutron load per unit area varies slowly as one moves towards the top left-hand corner, the divertor load builds up rapidly. Three case studies will be presented ($R$=1.75, 1.25 and 2.25m for $P_f$=100, 200 and 900 MW respectively) in detail. Fig.3 shows how the power balance constrains the $\kappa- \beta$ space for the case $R$=1.25m, $P_F$ = 200 MW. It can be seen that higher $\beta$ cases will need a higher $\kappa$. The sensitivity to different $\tau_E$ scaling, as well as impurity transport, the effects of neutron and particle loads on the center-stack, first-wall and divertor will be presented in detail.
References:
1. Y-K.M Peng et al., Nucl. Fusion 26 769 (1986)
2. Gi et al., Nucl. Fusion 55 063036 (2015)
3. D. Jassaby et al., Plasma. Phys. Control. Fusion 3, 151, (1977)
4. L. A. El-Guebaly et al., Fusion Sci. and Tech., 74:4, 340-369 (2018)
5. A. Sykes et al., Nucl. Fusion 58 016039 (2018) and references therein
6. Dennis Whyte, Phil. Trans. R. Soc. A 377 20180354 (2018)
7. M. Kotschenreuther et al. Nucl. Fusion 50 35003 (2010)
8. R. J. Goldston et al., Phys. Scr. T167 014017 (2016)
9. V. A. Soukhanovskii, IEEE Trans. Plasma Sci. 44 3445 (2016)
10. B. V Kuteev et al., Nucl. Fusion 51 073013 (2011)
11. Y-K.M. Peng et al., Nucl. Fusion 40 583 (2000)
12. B.N. Sorbom et al., Fusion Eng. and Design 100, 378–405 (2015)
13. Tobita K. et al., Nucl. Fusion 49 075029 (2009)
14. Najmabadi F. et al., Fusion Eng. Des. 65 143 (2003)
15. E. Surrey, Phil. Trans. R. Soc. A 377: 20170442 (2019)
16. M. Tillack et al., Nucl. Fusion 53 027003 (2013)
17. H. Takenaga et al., Nucl. Fusion 45 1618 (2005)
18. S. C. Jardin et al., Fus. Eng. Des. 65 165 (2003)
19. P. Kaw et al., Phys. Rev. Lett. 65 2873 (1990)
20. P F Buxton et al.,Plasma Phys. Control. Fusion 61 035006 (2019)Speaker: Dr Shishir Deshpande (Institute for Plasma research) -
The use of high temperature superconductors (HTS) in magnetic systems of fusion devices enables magnetic fields over 16 T, unachievable with low temperature superconductors (LTS), and promises significant reduction in cryogenic and energy budget [1-4]. HTS materials are considered by some researcher groups as the enabling material to make magnetic confinement systems more compact and more affordable. Until recently, some unresolved obstacles for using HTS materials in fusion applications existed, including little knowledge available on critical current in field and mechanical properties at cryogenic temperature, unknown behaviour in alternating magnetic field, limited availability, short piece length, lack of proven cable technology, and high price. Recent technology advances resolved many of these tasks bringing HTS materials to greater maturity for use in fusion devices.
Recently, S-Innovations company successfully developed and marketed a new product, particularly suitable for creating high magnetic fields for fusion devices: SuperOx second generation (2G) HTS wire with modified ReBa2Cu3O7 (ReBCO, Re—rare earth element) composition [5]. The new SuperOx wire demonstrates exceptionally high critical current and engineering current density in high magnetic field; for instance, engineering current density over 1000 A/mm2 at 20 K, 20 T and over 2000 A/mm2 at 4.2 K, 20 T has been repeatedly achieved in commercially produced wire (Fig. 1). The in-field performance of the new SuperOx wire in the 4.2-65 K temperature range is 1.5-2.5 times better than that of the previous product based on the GdBCO superconductor (Fig. 2).
This impressive result has been achieved by S-Innovations in hundreds of 300-600 m long pieces of routinely manufactured wire. The unique feature of the new wire is that the ReBCO layer in it does not contain c-axis correlated nano-columns, in contrast to the common opinion that only with nano-columnar defects is it possible to achieve enhanced in-filed performance in 2G HTS wire.
The mechanical properties of this HTS wire largely depend on the mechanical properties of the strong Hastelloy C276 substrate. In particular, the wire exhibits tensile strength of over 600 MPa and is fully stable at least up to 0.4% elongation. The mechanical properties examined at various temperatures are available for SuperOx 2G HTS wire and compare favourably well to that of other manufacturers.
An important issue of the technology is an understanding of losses appearing in superconductor in alternating magnetic fields. A large set of data was recently collected for AC loss behaviour of SuperOx 2G HTS wire in liquid nitrogen at wide range of current and frequency up to 800 Hz. The results confirm earlier observations that AC losses of 2G HTS wires are low as compared to the most other practical superconductors. The available experimental data can be scaled and AC losses at the operational conditions of large fusion magnets can be calculated.
Studies of SuperOx and other groups demonstrate that viable high current cables for fusion can be manufactured from these wires (see, e.g. [6]). More research is underway to perfect engineering approaches in this direction.
The new SuperOx wire has been routinely manufactured with reproducible quality and high yields at unprecedented volumes for 2G HTS wire industry. The manufacturing processes are highly automated and integrated into an intelligent production management system. A comprehensive quality management system comprises numerous in-line and off-line QC procedures. The production capacity was doubled during the past year, and 5-10-fold capacity increase scenarios are in place awaiting order commitments. The pending 5-10-fold expansion will result in a production volume of multiple tonnes per year, the level at which 2G HTS wire price will allow the construction of economical fusion power plants for future green energy [3,4].References
1. D. G. Whyte, J. Minervini, B. LaBombard, E. Marmar, L. Bromberg, M. Greenwald, “Smaller & Sooner: Exploiting High Magnetic Fields from New Superconductors for a More Attractive Fusion Energy Development Path”, J Fusion Energ (2016) 35:41–53 DOI 10.1007/s10894-015-0050-1
2. P. Bruzzone, W. H. Fietz, J. V. Minervini, M. Novikov, N. Yanagi, Y. Zhai and J. Zheng, “High temperature superconductors for fusion magnets”, 2018 Nucl. Fusion 58 103001
3. B. N. Sorbom, J. Ball, T. R. Palmer, F.J. Mangiarotti, J. M. Sierchio, P. Bonoli, C. Kasten, D. A. Sutherland, H. S. Barnard, C. B. Haakonsen, J. Goh, C. Sung, D. G. Whyte, “ARC: A compact, high-field, fusion nuclear science facility and demonstration power plant with demountable magnets”, Fusion Engineering and Design 100 (2015) 378–405
4. A. Sykes, A. E. Costley, C. G. Windsor, O. Asunta, G. Brittles, P. Buxton V. Chuyanov, J. W. Connor, M. P. Gryaznevich, B. Huang, J. Hugill, A. Kukushkin, D. Kingham, A. V. Langtry, S. McNamara, J. G. Morgan, P. Noonan, J. S. H. Ross, V. Shevchenko, R. Slade and G. Smith, “Compact fusion energy based on the spherical tokamak”, 2018 Nucl. Fusion 58 016039
5. A. Molodyk, S. Samolenkov, A. Markelov, P. Degtyarenko, S. Lee, V. Petrykin, M. Gaifullin, A. Mankevich, A. Vavilov, B. Sorbom, J. Cheng, S. Garberg, Z. Hartwig, S. Gavrilkin, S. Awaji, D. Abraimov, A. Francis, D. Larbalestier, C. Senatore, M. Bonura, O. Pantoja, S. Wimbush, N. Strickland, A. Vasiliev, High temperature superconducting wire based on ReBa2Cu3O7 with Re2O3 nanoparticles exhibits engineering current density over 1000 A/mm2 at 20 K, 20 T and over 2000 A/mm2 at 4.2 K, 20 T, to be published.
6. D.Uglietti, N.Bykovsky, K.Sedlak, B.Stepanov, R.Wesche, P.Bruzzone, “Test of 60 kA coated conductor cable prototypes for fusion magnets” 2015 Supercond. Sci. Technol. 28 124005Speaker: Dr Alexander Molodyk (S-Innovations, SuperOx)
-
-
-
Detailed new H-mode pedestal measurements in the inter-ELM periods of DIII-D discharges find that ion temperature gradient (ITG)-scale density fluctuations (ñ) can explain anomalous ion heat flux (Q
$_{i}$) during the ELM (Fig. 1), followed by Q$_{i}$becoming neo-classical$^{1}$until the next ELM, and with trapped electron mode (TEM) and microtearing mode (MTM) like ñ and magnetic fluctuations ($\tilde{B}$) able to explain anomalous electron heat flux (Q$_{e}$) between ELMs (Fig. 1). The observed ITG-scale, TEM-scale, and MTM-like turbulence (ñ,$\tilde{B}$) behaviors are in clear contrast to the often assumed suppression or near suppression of these modes in the pedestal during inter-ELM periods$^{2}$. This new information impacts our fusion power goals as it will lead to improved predictive capability when existing (and new) models take into account the transport caused by these modes and their various temporal phases during the inter-ELM period. The goal of predicting and controlling pedestal pressure evolution and improving ELM control (or elimination) is believed necessary to optimize fusion performance and to protect plasma facing components of ITER and future burning plasmas.
Inter-ELM behavior of electron pedestal density, temperature, and pressure gradients ( ∇n$_{e,ped}$, ∇T$_{e,ped}$,and ∇P$_{e,ped}$) estimated from Thomson scattering measurements and pedestal localized turbulence were studied in ITER similar shape (ISS) H-mode plasmas. The gradients of electron temperature, density, and pressure remain approximately saturated for more than 75% of the inter-ELM period (Fig 2). Immediately after an ELM, evolution of ITG-scale low-k (k$_{\theta}\rho_{s}$~0.2, k$_{\theta}$is the fluctuation poloidal wavenumber and$\rho_{s}$is the ion Larmor radius) ñ (Fig 2b), measured by Doppler backscattering at the foot of the pedestal, correlates strongly with a rapid increase and subsequent saturation of ∇n$_{e,ped}$(Fig 2d). This low-k ñ increases strongly when the pedestal profiles relax after an ELM crash, is correlated with a decrease in local E$\times$B shear and later is suppressed by increase in local E$\times$B shear (Fig 2c). The evolution of this ITG-scale ñ amplitude correlates well with the divertor D$_{\alpha}$light intensity (Fig 2a) and ∇T$_{e,ped}$(not shown) starts to increase after suppression of the ITG-scale ñ. These observations are consistent with ITG-scale ñ driving electron and ion heat flux as well as particle flux just after the ELM crash. Experimental results from ASDEX-U indicating the anomalous nature of both ion and electron heat fluxes right after an ELM crash, are consistent with our observations$^{1}$.
TEM-scale intermediate-k (k$_{\theta}\rho_{s}$~1) ñ measured by Doppler backscattering in the steep gradient region of the pedestal shows a critical ∇P$_{e,ped}$behavior with ñ increasing rapidly when a critical ∇P$_{e,ped}$is reached (Fig. 3). As the pressure gradient increases, increases in E$\times$B velocity shear (not shown here) leads to saturation of TEM-scale ñ (at ~120 kPa/m, Fig 3) indicating a balance between saturated values of gradient drive, shear suppression, and turbulence level. The approximate saturation of ∇P$_{e,ped}$that follows the large increase of TEM-scale ñ at critical ∇P$_{e,ped}$also suggests the related turbulence transport has a role in clamping ∇P$_{e,ped}$. Note however, although ∇P$_{e,ped}$becomes approximately constant within 25% of the ELM cycle there remains a slow increase in ∇P$_{e,ped}$as shown in Fig 2g and Fig 3. Magnetic turbulence$\tilde{B}$, measured by polarimetry and fast magnetic probes (Fig 2h), has been identified as MTM turbulence$^{3}$and increases with ∇T$_{e,ped}$(∇T$_{e,ped}$not shown but is very similar to the ∇P$_{e,ped}$shown in Fig. 2g ) in the inter-ELM period. Both TEM-scale ñ and MTM-like$\tilde{B}$can cause significant electron thermal transport and may explain the anomalous nature of Q$_{e}$in the pedestal as observed by transport calculations using experimental profiles (in ASDEX-U$^{1}$and DIII-D$^{4}$tokamaks).
Initial linear trapped-gyro-Landau-fluid (TGLF)$^{5}$calculations in the nearly saturated phase of the ELM cycle (60-99% of the ELM cycle, refer Fig 2) show that intermediate-k (k$_{\theta}\rho_{s}$~1) electrostatic fluctuations are the most unstable modes in the steep gradient region of the pedestal and propagate in the electron diamagnetic direction in the plasma frame. The growth rates are approximately 1.5-2 times more than the local E$\times$B shearing rates and so are relatively unaffected by the shearing. This is qualitatively consistent with the observation of these modes by the DBS system. These modes are still the most unstable modes when electromagnetic effects are included in TGLF calculations. Interestingly, scans of electron temperature and density scale lengths indicate that these modes are driven by both ∇T$_{e}$and ∇n$_{e}$in qualitative agreement with the observed ñ vs ∇P$_{e}$behavior (Fig 2 and 3). Further, the TGLF simulations indicate a critical gradient type response similar to experiment (Fig 3). Based on these above TGLF results we initially identify these modes as TEM type instabilities. The above TGLF results are consistent with TEM type instabilities leading us to initially identify the experimentally observed density fluctuations to be trapped electron modes.
The new scientific understanding reported here on the ELM and inter-ELM turbulence, thermal fluxes, and temporal behavior are critical to testing, validating, and improving current models and simulations. These improved models and simulations are in turn fundamental to predicting and controlling pedestal pressure and ELMs (or even ELM elimination) necessary for optimizing and protection of future fusion devices.This work was supported by the US Department of Energy under grants DE-FG02-08ER54984, DE-AC02-09CH11466, DE-FG02-08ER54999, DE-SC0019302 and DE-FC02-04ER54698.
$^{1}$E. Viezzer et al., Nucl. Fusion 57, 022020 (2017).
$^{2}$P.B. Snyder et al., Nuclear Fusion 51, 103016 (2011).
$^{3}$J. Chen et al., Submitted to IAEA FEC 2020.
$^{4}$S. Haskey et al., Submitted to IAEA FEC 2020.
$^{5}$G. Staebler et al., Phys. Plasmas 14, 055909 (2007).Speaker: Kshitish Kumar Barada (University of California-Los Angeles) -
Since the initial JET operations with the metal wall (JET-ILW), the experimental results have shown a pedestal pressure in baseline plasmas that tends to be 10-20% lower than in the corresponding earlier carbon wall operations (JET-C) [1]. While this degradation seems mainly correlated with the high fueling rates typical of JET-ILW [2,3] and/or the lack of carbon impurity [4,5], an exhaustive and comprehensive explanation for the lower pedestal performance has not been achieved yet. This work will address the role of fueling and its goals are:
- to prove that the lower pedestal performance in D fuelled JET-ILW plasmas are due to the higher separatrix density ($n_e^{sep}$) produced by the higher neutral pressure,
- to describe the corresponding physics mechanisms that lead to the pedestal degradation.
In the baseline scenario of JET-ILW, operations with no gas-fueling rate have been extremely challenging due to the problems related to tungsten influx and divertor heat loads. Since most of the JET-C plasmas have no gas fueling, a direct comparison of JET-ILW and JET-C pedestals obtained with identical engineering parameters is not possible. A further complication is related to the fact that the peeling-ballooning (PB) stability model (implemented with ideal MHD equations) does not describes correctly the experimental JET-ILW results (the experimental pedestal with high fueling rates does not seem to reach the stability boundary when the ELMs are triggered [1,2,3]). Therefore, the work is based on two levels. First, the work focuses on the empirical understanding of the pedestal behavior in JET. Then, based on these results, an investigation of the pedestal transport and an extension of the PB stability analysis is done with the GENE [6] and JOREK [7] codes.
Figure 1 shows the height of the electron pedestal pressure ($p_e^{ped}$) versus $n_e^{sep}$ for a set of JET plasmas with the same engineering parameters apart from fueling rate and divertor configuration. The JET-C dataset has higher $p_e^{ped}$ than the JET-ILW dataset. However, the two datasets align very well in the $p_e^{ped}$-$n_e^{sep}$ diagram. Moreover, the JET-ILW pulses with lowest $n_e^{sep}$ reach a pedestal pressure comparable to JET-C. This suggests that the separatrix density is one of the key parameters to understand the difference between carbon and metal wall. The higher $n_e^{sep}$ in JET-ILW is likely due to the higher neutral pressure, as recently discussed for JET-ILW [8] and AUG [9], produced by the higher gas fueling and/or different recycling. In figure 1, note that the subsets with different divertor configurations show no systematic difference, strengthening the hypothesis that the neutral pressure plays the key role.
The standard PB stability analysis performed with ideal MHD can only partially explain the empirical trend. This is shown in figure 1 by the red line, which represents the pressure predictions obtained with the Europed code [10]. The increase in $n_e^{sep}$ initially leads to a sharp reduction in the predicted $p_e^{ped}$. This is due to the fact that the increasing $n_e^{sep}$ is intrinsically linked to the outward shift of the $n_e$ position ($n_e^{pos}$), shifting the $p_e$ profile and destabilizing the PB modes [3,11]. While this explains rather well the JET-C trend, the effect saturates at high $n_e^{sep}$. The prediction significantly overestimates the experimental $p_e^{ped}$ for the JET-ILW pedestal with high $n_e^{sep}$.
Therefore, the next steps are to understand the mechanisms that (1) set the pedestal gradient and (2) trigger the ELMs at high $n_e^{sep}$. First of all, we note from figure 1 and figure 2(a) that the reduction of pedestal gradient is correlated with the increase in $n_e^{sep}$ and in $n_e^{pos}$-$T_e^{pos}$. The increase in these parameters leads to the increase of $\eta_e$ (ratio between $n_e$ and $T_e$ gradient length) [3], which in turn can destabilize microinstabilities, increase turbulent transport [12, 13] and hence reduce the pressure gradient. This hypothesis is under investigation with GENE and is supported by preliminary results shown in figure 2(b), where the growth rates (mainly of ETG modes) are higher in pedestals with higher $n_e^{sep}$ and higher $n_e^{pos}$-$T_e^{pos}$ [14,15].
Then, it is necessary to understand the ELM triggering mechanisms. The discrepancy between the experimental results and the ideal MHD results is quantified with the ratio $\alpha_{crit}/\alpha_{exp}$ (where $\alpha_{crit}$ is the normalized pressure gradient predicted by ELITE and $\alpha_{exp}$ is the experimental one). Figure 3 shows that $\alpha_{crit}/\alpha_{exp}$ increases with increasing resistivity. This suggests that resistivity might have a destabilizing effect on the PB modes, as theoretically discussed in [16]. This hypothesis is currently under investigation with the non-ideal MHD non-linear code JOREK.
The picture that is emerging is the following. Due to higher gas fueling rate / different re-cycling, JET-ILW has higher neutral pressure than JET-C. This leads to higher $n_e^{sep}$ and higher $n_e^{pos}$, producing higher $\eta_e$, increasing the turbulent transport and reducing the pedestal gradient. In turn, the lower pedestal gradient leads to a lower temperature inside the separatrix, increasing the resistivity and making resistive effects on the MHD stability non-negligible.
References
[1] Beurskens M. et al., NF 54 043001 (2014)
[2] Maggi C. et al., NF 57, 116012 (2017)
[3] Frassinetti L. et al., NF 59 076038 (2019)
[4] Giroud C et al., IAEA FEC, EX/3-4 (2018)
[5] Beurskens M. et al., NF 56 056014 (2016)
[6] T.Görler et al., JCP 230, 7053-71 (2011)
[7] Huysmans G. et al. NF 47, 659 (2007)
[8] Frassinetti L. et al, to be submitted to NF
[9] Kallenbach A. et al., NME 18, 166 (2019)
[10] Saarelma et al., PPCF 60, 014042 (2018)
[11] Dunne M. et al., PPCF 59, 014017 (2017)
[12] Hatch D.R et al., NF 57, 036020 (2017)
[13] Kotschenreuther M. et al., NF 57 064001 (2017)
[14] Chapman B. et al., to be presented at the 47th EPS conference
[15] Hatch D.R et al., NF 59, 086056 (2019)
[16] Wu et al., PoP 25, 092305 (2018)Speaker: Lorenzo Frassinetti (KTH, Royal Institute of Technology) -
$\qquad$ Characterizing and understanding the power threshold conditions for ITER to achieve H-modes ($P_{LH}$) is a major goal of a series of L-H transition experiments undertaken at JET since the installation of the ITER-like-wall (JET-ILW), with Beryllium wall tiles and Tungsten divertor [1,2,3,4]. In this contribution we report on results from L-H transitions studies in H, D and new almost pure $^4Helium$ plasmas, and compare the results with ITER predictions. The most notable result is that the density at which $P_{LH}$ is minimum, $n_{e,min}$, is considerably higher for $^4He$ than for D, and strongly influenced by shape.
$\qquad$ A detailed analysis of the pre-transition $E_r$ profiles across the ne scan in D and $^4He$ find matching qualitative changes in the $E_r$ profile. In high field NBI heated D plasmas, we report on power balance analysis and its impact on $n_{e,min}$. Modelling of the plasma SOL does show differences in the heat flux required to drive a transition between H and D (in the high $n_e$ branch), and $^4He$ plasmas are also being studied.Characterizing the L-H transition power threshold for H, D, $^4He$: $n_{e,min}$, ion heat flux, $E_r$
$\qquad$ The interest on $^4He$ plasmas is not purely academic, and our data brings surprises. The ITER Research Plan includes a low toroidal field Pre-Fusion Operating Power phase with either Hydrogen or Helium plasmas in order to study H-modes as early as possible, before the nuclear phase that starts with D plasmas. A prediction of $n_{e,min}$, was made inspired on the studies of Ryter [5], who observed in AUG that a sufficient edge ion heat flux is necessary to achieve a sufficient radial electric field (shear). Assuming pure electron heating in ITER, $n_{e,min}$ has been evaluated on the basis of 1.5-D transport modelling as the density at which the ratio of edge ion power flux to total edge power flux starts to saturate with increasing density. The result of this modelling is that $n_{e,min}$~$0.4 n_{GW}$, independent of the ion species [6]. The transition condition in that model is itself based on the assumption that the He power threshold, $P_{LH}(He)$, is 1.4 x $P_{LH}(D)$, while $P_{LH}(H)$=2x$P_{LH}(D)$ [7].$\qquad$ We find that for Horizontal Target plasmas the estimate of $n_{e,min} \sim 0.4 \times n_{GW}$ is in agreement with the D data, but $n_{e,min}$ is closer to $0.5 \times n_{GW}$ for H, and to 0.6 for Helium. NBI heated H plasmas have higher $P_{LH}$ than RF heated ones, the reasons are still being investigated. The data points with no black outline correspond to transitions during an NBI blip, so $P_{LH}$ is probably higher. In fact radiation is considerably higher for the dominantly RF-heated Horizontal Target Helium plasmas at low density, so the auxiliary power required for the L-H transition to take place is lower for H than for He below 3.3×$10^{19} e/m^3$. Above $n_{e,min}$(He), D and He have similar $P_{LH}$, below the Martin scaling, while H has a much higher $P_{LH}$.
In the 1.8 T Corner dataset $n_{e,min}$ is not so easily identified. Above 0.4$n_{GW}$, $P_{LH}$ in Corner is generally higher than in Horizontal Target for all species, approaching the Martin scaling for D and He, much higher for H.
$\qquad$ The strong shape effect shown in all L-H transition datasets at JET is in apparent contradiction with the ion heat channel determining $n_{e,min}$. A detailed study of the relation between ion heat flux and $n_{e,min}$ in 3T, 2.5 MA D plasmas, now with $T_i$ measurements, is underway [9]. We find the e-i exchange term is subdominant and unlikely to determine $n_{e,min}$.
$\qquad$ In a dataset with Horizontal Target, 2.4 T, 2 MA, NBI-heated plasmas (not shown), we find that $n_{e,min}(He) \sim 0.7 \times n_{GW}$, while $n_{e,min}(D) \sim 0.4 \times n_{GW}$. Above $n_{e,min}(He)$, D and He have similar $P_{LH}$. In this case we are attempting to reproduce the ITER transport models and $P_{LH}$ predictions and contrast them with the data. For these plasmas Doppler reflectometry shows that the $E_r$ profile of the low ne branch for both D and He plasmas has a modest $E_r$ well inboard of the separatrix and a sharp peak further out, while the high density branch has a clear $E_r$ well, but no peak near the separatrix.
$\qquad$ DIII-D results show a ~30% increase in $n_{e,min}$ of He plasmas relative to D [10], lower than our 50% shift. AUG studies show no difference in $n_{e,min}$ between H, D and He, and the same $P_{LH}$ for D and He plasmas[11]. In AUG $H+^4He$ mixtures [14], more than 20% $n_{He}/(n_{He}+n_D)$ is needed to see a change in $P_{LH}$(H), while <10% suffices in JET NBI heated plasmas [4]. C-Mod results [12] show He data in the low $n_e$ branch for $n_e<0.3 \times n_{GW}$, while in D $P_{LH}$ increases with density, indicating a shift in $n_{e,min}$. Above $n_{e,min}(He)$, $P_{LH}$ in JET-ILW is similar for D and He, therefore the increase in $P_{LH}$ due to higher $n_{e,min}$ is compensated by the lower power required to access it, since ITER had assumed $P_{LH}(He)$=$1.4\times P_{LH}$(D).
Simulations of L-H transitions for hydrogen isotopes with the HESEL[15] model find that $P_{LH}$ decreases with increasing mass number like $A^{-1.2}$. Results in $^4{He}$ plasmas are expected soon.
Summary and Outlook:
Our results question the logic that supports He for access to H-mode in the early operating phase of ITER, but not necessarily the final power estimate. Detailed analysis is on-going, to provide better understanding of the mechanisms involved and produce an improved prediction. Novel $E_r$ measurements will enable a more detailed understanding of L-H transitions in D and $^4He$ plasmas.
A Tritium campaign is planned at JET for summer 2020. We expect to obtain L-H transition data for pure Tritium, H+T and H+$^4He$ mixtures. This should inform future experiments in JET and ITER.
References:
[1] CF Maggi et al 2014 Nucl. Fusion 54 023007
[2] E Delabie, Proc. of the 25th IAEA FEC, Saint Petersburg, Russia, EX/P5 (2014)
[3] J Hillesheim et al Proc. of the 26th IAEA FEC, Kyoto, Japan, (2016)
[4] J Hillesheim et al Proc. of the 27th IAEA FEC, Gandhinagar, India (2018)
[5] F Ryter et al 2013 Nucl. Fusion 53 113003
[6] ITER Research Plan within Staged Approach, ITR-Report 18-003 (2018) p 351
[7] D McDonald et al, Plasma Phys. Control. Fusion 46 519 (2004)
[8] U. Kruezi et al 2020 JINST 15 C01032
[9] P. Vincenzi et al., 46th EPS conf., P2.1081 , Milano, Italy (2019);
[10] Gohil et al, Nucl. Fusion 51 (2011) 10;
[11] F Ryter et al, Nucl. Fusion 54 083003 (2014);
[12] C.E. Kessel et al 2018 Nucl. Fusion 58 056007;
[13] C. Silva, to be presented at EPS 2020;
[14] U Planck, submitted to NF
[15] Nielsen AH et al, Physics Letters A 379 (2015) 3097–3101.Speaker: Dr Emilia R. Solano (Laboratorio Nacional de Fusion, CIEMAT, SPAIN) -
Sudden increase of the energy confinement time by 38% is observed during divertor detachment operation with RMP application in deuterium plasmas in LHD. Edge transport barrier (ETB) is formed during the RMP induced H-mode phase at the inner separatrix of magnetic island created by the RMP, leading to steepening of pressure profile while the detachment is maintained. Such behavior has been observed neither without RMP nor in hydrogen discharges to date. During the RMP induced H-mode, MHD activity is detected in magnetic probe, and high charge state impurity emission decreases indicating decontamination of core plasma. Core plasma transport analysis shows clear decrease in the transport coefficient after the detachment transition and in the subsequent RMP induced H-mode phase. The present results have revealed, for the first time, the impact of 3D edge magnetic field structure and isotope effect on core plasma performance during divertor detachment operation that provides important perspective toward integrated performance in future reactors.
Compatibility of good core plasma performance with enhanced edge radiation to mitigate the divertor heat load is a crucial issue for magnetically confined fusion reactors. It is also not clear yet how 3D edge magnetic field structure affects the divertor heat load and the core plasma transport during detachment either in tokamaks or in stellarators. In LHD, stable detachment control is realized with RMP application of m/n=1/1 mode [1]. The present paper reports new observations on core plasma confinement during divertor detachment operation in deuterium and hydrogen plasmas.
The RMP application creates a magnetic island of m/n=1/1 in the edge stochastic layer, where the impurity radiation is enhanced due to increased volume of cold plasma region.
Figure 1 shows time traces of plasma parameters with RMP application in deuterium plasma operation. The total NBI heating power of three beam lines was ~5 MW with injection energy of 152~177 keV (3.6MW in co-direction and 1.4 MW in counter-direction, respectively). Density was ramped up with deuterium gas puff and the radiated power increased as well. The detachment transition occurred at t = 3.95 sec as indicated by the reduction of divertor heat flux in Fig.1 (e). It was found that there appears toroidal modulation in the divertor heat load pattern caused by the m/n=1/1 RMP field mode structure. During the detached phase, the density was increased even further beyond the density limit (nSudo, Sudo limit) (Fig.1 (a)). It was observed that plasma stored energy spontaneously increases at t ~ 4.5 sec by 38%, and spikes appears at the magnetic probe signal and divertor heat flux while the detachment is maintained. Such increase of stored energy during detachment is unique to the deuterium plasmas with RMP application. This phenomenon has never been observed without RMP and in hydrogen plasmas up to now.
The RMP induced H-mode is accompanied by the ETB formation with steep edge pressure gradient. Figure 2 shows the radial profiles of the plasma pressure at the edge region for deuterium plasmas with RMP application during the detached phase and the subsequent H-mode phase. Due to the strong edge cooling at the island where $T_e$ < 10 eV, the pressure profile is clamped by the edge island in the detached phase. Clear steepening of pressure profile is observed in the RMP induced H-mode phase (red, 5.1 sec), where the ETB is formed at the inner edge of the magnetic island induced by the RMP.
Figure 3 shows energy confinement time, $\tau_E$, scaled by gyro-Bohm dependence, $\tau_E^{GB} \propto (n_e / P)^{0.6} B^{0.8}$. $\tau_E / \tau_E^{GB}$ monotonically decreases with increasing density. The systematically smaller values in the case with RMP application are due to the shrinkage of the plasma volume caused by the edge magnetic island, while heat transport coefficients, $\chi_{eff}$, obtained by transport analysis, show similar values for both cases with and without RMP. It is observed that during the RMP induced H-mode phase the $\tau_E / \tau_E^{GB}$ increases by 38%.
The core plasma transport has been analyzed with TASK3D code by taking into account the NBI heating profile. Figure 4 shows the results of the deuterium plasma. After the detachment transition, the pressure profile as well as the NBI heating profile become peaked (t = 4.3 sec). The peaked heating profiles are attributed to deeper penetration of NBI in the detached phase with shrinkage of plasma volume due to the strong edge cooling. The transport coefficient decreases at almost entire plasma in the detached phase. In the RMP induced H-mode phase (t = 4.9 sec), the pressure profile becomes further steeper at the edge while the NBI heating profile also slightly peaked. The transport coefficient decreases by a factor of 1.6 to 2 at the edge region, $\rho$ ~ 0.8, which corresponds to the inner separatrix of the island as shown in Fig.2 where the ETB is formed.
[1] M. Kobayashi et al., Nucl. Fusion 59 (2019) 096009.
Speaker: Masahiro Kobayashi -
Due to the limited power exhaust capability of the divertor, a future DEMO reactor needs a high core/pedestal radiation level, controlled and tailored by an appropriate seed impurity mix. The core/pedestal seeding has to be integrated with substantial divertor radiative cooling and a no/very small ELM plasma regime. Required boundary conditions of the seeding scenario are sufficient energy confinement and low fuel dilution. Taking into account the radiative capabilities of the potential seeding species (N2, Ne, Ar, Kr, Xe), only Ar is expected to contribute substantially to both pedestal and divertor radiation for reactor conditions. Therefore, a likely scenario is the combination of Ar with Kr or Xe for core radiation at reduced dilution and with N2 or Ne for enhanced divertor radiation. For the divertor seeding species a high divertor impurity compression is also required to avoid excessive core dilution. This paper reports about recent studies in ASDEX Upgrade towards the development of corresponding DEMO seeding scenarios with emphasis on operation with no ELMs, good energy confinement, and compatibility with the tungsten plasma facing components.
One route to a no/small ELM scenario is the reduction of the edge pressure gradient in the steep gradient zone by radiative losses. Ar is particularly suited for AUG conditions since it creates a strongly radiating ring in the pedestal region. The radiation effect can be combined with additional pedestal transport effects as provided by turbulence or magnetic perturbations to assure ELM suppression. Recently, a new type of ELM-free H-mode with good energy confinement has been discovered at ASDEX Upgrade [1]. The scenario exhibits similarities with the EDA H-mode from Alcator C-Mod, a quasi-coherent electromagnetic mode with toroidal mode number n ~20 in the pedestal region results in complete stabilization of ELMs at good confinement. First attempts to combine this scenario with a feedback controller for the power flux over the separatrix via Ar seeding [2] showed ELM-free conditions at up to 5 MW injected NBI and ECRH combined power (figure 1, left panels).
At higher power and core radiation, Psep could be kept low, in the vicinity of the H-L transition, but ELMs remained, albeit at reduced size (Fig 1, mid panels). Future studies are planned to determine whether a refined seeding strategy will allow for full ELM suppression also at high heating power. Figure 1 (right panel) compares the electron pressure gradient in the pedestal region with and without Ar seeding for H-mode discharges with and without ELMs. Indeed, Ar seeding reduces the maximum pressure gradient, while global confinement is even improved due to increased pressure further inwards.
A prerequisite for efficient divertor radiative cooling is the achievement of a high divertor impurity density compared to its core density, namely a high divertor compression. To shed light on the underlying processes, calibrated gas puffs of N2, Ne and Ar have been injected into ASDEX Upgrade H-mode plasmas for different values of the divertor neutral pressure. The temporal development of core densities and recycling rates was measured by charge exchange recombination spectroscopy, Zeff variations and a SPRED VUV spectrometer, see figure 2. The fastest removal rates are observed for Ar, shortening from ~ 0.3 to ~ 0.1 s as the divertor neutral pressure is increased from 0.6 to 4 Pa. At high pressure, the shortening saturates and Ne and Ar show equal pumping times. Such a shortening is not observed for N2, which appears to be dominated by wall storage and release processes.
Simple particle balance analysis allows already some conclusions. At very low pressure, pumping rates are expected to scale with the inverse square root of mass. At very high pressure, collisions with D2 molecules lead to an entrainment of the impurities in the D2 flow and equal removal rates for all species. The observed removal rates are proportional to the product of the divertor compression and the pumping speed. Faster removal rates of Ar vs Ne give evidence of a higher divertor compression of Ar vs. Ne. SOLPS calculation including the effects of neutral collisions are required for further quantification.
In conclusion, Ar pedestal radiation is a promising tool for reduction of the pedestal electron pressure gradient, and thus an important element for a no-ELM scenario. Active control needs to be expanded towards the tailoring of the spatial profile of the radiation, rather than the pure radiated power. A step in this direction has recently been demonstrated for the X-point radiator regime in ASDEX Upgrade [3]. First attempts for combining core Ar radiation with RMP ELM suppression showed short (0.5 s) phases of good performance, but were hampered by the subsequent occurrence of a locked mode.
[1] L. Gil et al, Stationary ELM-free H-mode in ASDEX Upgrade’, EPS 2019, submitted to Nuclear Fusion
[2] A. Kallenbach et al., Nuclear Fusion 52 (2012) 122003.
[3] M. Bernert et al., this conference.Speaker: Dr Arne Kallenbach (Max-Planck-Institut f. Plasmaphysik) -
An optimized pedestal regime called the Super-H Mode (SH-mode) is leveraged to simultaneously couple a fusion relevant core plasma with a scrape-off layer appropriate for realistic reactor exhaust solutions. Recent DIII-D experiments have expanded the operating space from previous studies of the SH regime and investigated optimization of impurity seeding, deuterium gas puffing, 3D magnetic perturbations, and plasma shape. Experiments demonstrate gas puffing and impurity seeding lead to a radiative mantle and low divertor temperatures (< 15eV) that are compatible with maintaining SH-mode and have marginal impact on pedestal and core pressure. An important recent result is that access to the SH-mode has been achieved in shapes matching JET plasmas with moderate plasma triangularity ($\delta_{avg} \sim 0.4$), providing a pathway for increased performance for the JET D-T campaign as well as increased confidence in the EPED predictions for SH-mode access in ITER.
Plasma shape is a key parameter impacting pedestal stability, and when SH-mode access is marginal, small changes in triangularity and aspect ratio can lead to an increase in global metrics like plasma stored energy through pedestal optimization. Previous experiments maximized plasma triangularity and volume in the SH regime in order to maximize pedestal and core performance; however, recent experiments show SH access can still be obtained at moderate plasma shaping. Figure 1 shows the pedestal electron pressure and density in two JET similar shapes, with one having an increased plasma triangularity from 0.3 (gray) to 0.4 (red). This change in triangularity opens access to the SH-mode channel, allowing a higher pedestal at the same density, and higher stored energy even with a slightly reduced plasma volume. The relatively modest plasma triangularity compared to the double null SH experiments leads to pedestal pressures which are farther from ideal $\beta$ limits, allowing plasma trajectories deep in the SH channel to be maintained in a stationary state. Robust SH-mode access in lower single null shapes with intermediate levels of triangularity implies applicability for potential use as a target scenario in both JET and ITER. Figure 1 indicates that SH-mode is compatible with JET plasma shapes and could increase plasma stored energy in the upcoming D-T campaign at the same engineering parameters.
By employing a dynamic density trajectory and shape control, peeling and ballooning physics can be decoupled in the pedestal {1}. The peeling-limited pressure pedestal reaches ~20-30kPa on DIII-D and up to ~80kPa on C-Mod {2}, even with strong gas puffing. The pedestal maintains low collisionality with a high separatrix density {3}, which is important for achieving a low heat flux to the divertor plate without degradation of the pedestal pressure. Recent experiments on DIII-D have employed co-current beam injection at full magnetic field ($B_t=2.1-2.2T $) and current ($I_p=1.4-2.0 MA$) in both closed and open divertor configurations. Initial indications show that a slightly larger divertor volume with a longer leg between the x-point and strike points allows more power to be radiated in the scrape off layer and pedestal regions, and to be excluded from the core more effectively. Advanced control algorithms {4,5} simultaneously optimize the line average density and divertor radiative power. Introducing 3D magnetic perturbations that pump out particles actively controls the line average density and allows the SH-mode plasmas to enter an extended stationary phase for 2s with sustained pedestal pressure and controlled impurity content. Dual seeding with deuterium and nitrogen radiate power near the separatrix and reduce the divertor heat flux and temperature to promote integration with requirements of plasma facing components, as shown in Fig. 2 for a lower biased double null ($\delta_{avg} \sim 0.57$). The SH-mode plasma has a pedestal electron temperature of ~1keV with divertor temperature <15eV. Feedback control on the divertor radiation was employed for optimal nitrogen seeding to maintain a steady dissipation of ~42% of injected power contained to the divertor and pedestal regions while maintaining a pedestal with $\beta_N^{ped}~0.8$ and core plasmas with $\beta_N^{core}=2.2$ and $W_{MHD} \sim 2MJ$.
{1} W.M. Solomon, et. al., Phys. Rev. Lett. 113, 135001 (2014)
{2} J. Hughes, et. al., NF 58, 112003 (2018)
{3} P.B. Snyder, et. al., Nucl. Fusion 59, 086017 (2019)
{4} D. Eldon et al., Nucl. Mater. and Energy 18, 285-290 (2019)
{5} F. Laggner et al., submitted Nucl. Fusion 2019Speaker: Theresa Wilks (UsMIT)
-
-
-
-
-
A description of a micromodel of a salt blanket for conducting benchmark experiments with the melting of salt fluorides (0.52 NaF+0.48ZrF4) is given. The experiments are intended for verification of nuclear data libraries and codes applied to justify the nuclear and radiation safety of full-scale subcritical blankets of thermonuclear installations based on liquid-salt technologies,
Three dry experimental ducts are in the salt blanket micromodel. Canisters with experimental samples (natNi, natZr, natCd, natTi, 59Co, 63Cu, 65Cu, 64Zn, natIn, 27Al, natMg, natFe, 169Tm, 197Au, 232Th) are placed in the ducts. A neutron generator NG24M with a neutron yield of ~ 10 raized to the power of 11 n/s is used as a source of neutrons with an energy of 14 MeV.
The results of the initial stage of experimental research are presented (Fig .1). These results are compared with the calculated data obtained using the MCNP4 code with the ENDF/BVII.0 and FENDL-3 libraries as well as MCNP6.1 code with the ENDF / BVII.0 library.
The drawing of the micromodel is shown in Fig. 2 and 3. The micromodel is represented by the following main elements:
Neutron generator with an energy of 14 MeV – 1;
Aluminum-based tank with retarder (water, heavy water) – 2;
Liquid salt blanket – 3;
Steel support for the distillate tank – 4;
Jack – 5;
Mechanism for rotating the blanket – 6.
The composition of the materials in the model:
tank material (mass.%): Mg-24 – 0.7; Al-27 – 98,3; Si-28 – 1;
blank material (mass.% ) E-110 (Zr – 99, Nb – 1);
the composition of salts (wt.%): 0.522 NaF+0. 478ZrF4.
The micromodel was created for Monte Carlo calculations using the MCNP-4 code and reproduces the real geometry of the expert with the highest possible accuracy.
The work is carried out under the RFBR project 19-29-02028 " Benchmark experiments for verification of nuclear libraries used for calculating full-scale subcritical blanks with FNS (Fusion neutron source)" and is a continuation of the work under the RFBR project 14-08-90042 "Modeling of subcritical blanks of a fusion neutron source for benchmark experiments in justification and licensing of a full-scale project" [1-3].References
- Yu. E. Titarenko, V. F. Batyaev, K. V. Pavlov, A. Yu. Titarenko, P. N. Alekseev, M. I. Gurevich, A. A. Dudnikov, A. V. Zirkin, B. V. Kuteev, A. B. Koldobsky, A. I. Kievitskaya, and Yu. G. Fokov. Benchmark experiments for verification of calculational parameters of fusion neutron source blankets. ATOMIC ENERGY (Moscow), Vol: 120, is-sue 1, 2016, pp. 43-48.
- A.V. Zhirkin, B. V. Kuteev, M. I. Gurevich, B. K. Chukbar. Neutronics Analysis of Blankets for the Hybrid Fusion Neutron Source. NUCLEAR FUSION, Volume 55, Issue 11, September 2015, p. p. 1-10.
- A.V. Zhirkin, P. N. Alekseev, V. F. Batyaev, M. I. Gurevich, A. A. Dudnikov , B. V. Kuteev, K. V. Pavlov, Yu.E. Titarenko, A. Yu. Titarenko. Fusion neutron source blanket: requirements for benchmark calculation accuracy and precision of the experiment. NUCLEAR FUSION, Volume 57, Issue 6, June 2017, p. p. 1-7.
Speaker: Dr Alexey Zhirkin -
Here we report the result of validation of tritium recovery system for liquid PbLi breeding blanket with Vacuum Sieve Tray (VST), that is a tritium extraction from multiple droplets under a vacuum. The VST device is installed in Oroshhi-2 PbLi test loop at the National Institute for Fusion Science (NIFS) in 2019 followed by further modification, and the performance and reliability of the process is demonstrated in 24h continuous operation.
Liquid PbLi is expected as a candidate breeder for ITER-TBM and DEMO. Based on successful proof-of-principle (PoP) study, VST development is in engineering demonstration / maturity phase, while various other extraction methods are in PoP phase. The authors have previously reported that droplets of PbLi show approximately 200 times faster tritium transport that assists the tritium extraction significantly. Key issues are a) possible performance degradation by the dense distribution of multiple nozzle arrays, b) performance verification in continuous operation and risk of degradation in long term operation. To solve these issues, following targets are established and experimental campaigns are in progress.
1) The Tritium Extraction Efficiency (TEE) verification. The authors previously reported the estimated TEE as shown in Fig. 1 [A]. A TEE greater than 80% is the target of this campaign as a typical requirement for such as EU-DEMO.
2) The multiple nozzle effects verification. Previous estimation of TEE is obtained under one nozzle column that may not be extrapolated to multiple droplets. Conservative analysis suggested that the TEE under multiple nozzles might be 20% to 60% of that of the one-column condition [B].
3) Reliable operation without TEE degradation and unexpected stability problems. Regarding the reliable operation, this campaign selected at least one day as the minimum target.![Fig. 1. Estimated TEE as a function of the falling period. TEE results using four different nozzle diameters are plotted on a same chart. The quasi dispersion coefficient is 3.4×10-7 m2s-1 [A]. The TEE of 0.8 is the assumed target of this study as a typical requirement for such as EU-DEMO.](https://workshop.nifs.ac.jp/fec2020/image/45-Okino-image-fig_1.jpg)
A multi-nozzle experimental setup, shown in Fig. 2, is designed to have a DEMO relevant extraction efficiency of 80 % with the falling height of 0.5 m. The numbers of PbLi nozzle, 7 and 19, are approximately one thousandth of that of DEMO WCLL loop. It was, as shown in Fig 2, fabricated and integrated into the Oroshhi-2 PbLi test loop at NIFS which has competent capacity for this purpose. Experimental campaign is performed by the following procedures.
1) TEE is measured by comparing the difference in deuterium concentration in PbLi, Cin and Cout, as TEE ≡ (Cin – Cout)/Cin. To validate the results, the mass balance between Q(Cin-Cout) and Jext, mass flow of the dissolved gas, are measured. Experiments with the nozzle diameters of 0.6 mm and 1.0 mm and the falling period of 0.1 and 0.25 s are performed. 2) Continuous operation is monitored by TEE transition.
As shown in Fig. 3, a first stage of multi-nozzle VST trial is performed. The pressures of extracted deuterium (D2) gas as the function of two different falling periods, t1 and t2, are consecutively measured with four nozzles setup.
The pressures of extracted D2 gas are P1 at the falling period of t1, and P2 at t2, respectively. Background gas pressure is assumed as P0 at before falling. The pressure ratio (P2-P0)/(P1-P0) identifies the mass transfer coefficient of the VST by comparing the theoretical TEEs at t1 and t2 in Fig. 1.
The obtained ratio is 1.9±0.3, whereas Fig. 1 reads 1.8. This is, however, a preliminary value and too early for analyzing a result. More experiments with multiple parameters must be performed, which are underway now.
By this campaign the viability of VST for ITER TBM, as well as DEMO liquid PbLi blanket will be evaluated.
[A] F. Okino et al., Fus. Eng. Des. 109-111 (2016) 1748-1753,
[B] F. Okino et al., Fus. Sci. Tech., 71 (2017) 575-583.Speaker: Dr Fumito Okino (Institute of Advanced Energy, Kyoto University) -
The ADITYA (R0 = 75 cm, a = 25 cm) ohmically heated circular limiter tokamak has been designed and built to produce plasma in a toroidal vacuum chamber of rectangular cross-section with 20 numbers of toroidal field coils capable of producing ~ 1.5 Tesla toroidal magnetic field at plasma centre. The ADITYA tokamak has been successfully operated for 3 decades with more than 30,000 discharges. Recently ADITYA tokamak has been upgraded into a state-of-art machine with new plasma chamber of circular cross section for divertor operation and good plasma control to support the Indian Fusion program. To minimize the up gradation cost and schedule, the existing sets of toroidal magnetic field (TF) coils, poloidal magnetic field (PF) coils and the Ohmic coils has been retained in the ADITYA-U tokamak. The TF coil of 1260 x1030 mm overall dimensions are picture frame type, made of two “C” halves with each half having 6 machined turn made of ETP copper conductor of 12.5 mm thick with inter turn insulation of GFRP as shown in figure-1. Among two halves, one is large in size (in radial direction) and another one is small in size. The two halves designed is considered to facilitate in-situ integration/dis-assembly of the plasma chamber. Two halves of the each TF coil finger joints are bolted together using 16 nos. of M20 bolts. TF coils are designed to with stand in plane and out of plane EM forces arising during operation. The maximum in plane EM forces calculated due to 50 kA current is ~0.16MN resisted by the buckling cylinder. The TF coil system is analysed to ensure the structural reliability. The TF coil finger joint’s fasteners, which connect the two C’s (large and small), are impossible to tighten due to non-availability of access, in presence of the surrounding supporting structure and wedge blocks after initial assembly of the ADITYA tokamak. After successful operation of TF coils of ADITYA tokamak close to the design parameters, the TF coils were dismantled for ADITYA upgrade. During the dis-assembly of Toroidal magnetic field (TF) coils, it was found that many TF coil’s joints were damaged and finger joints fasteners were completely loosen. The loosened bolting connections, resulted into arcing lead to TF coil damage. To avoid this happening again in ADITYA upgrade, finger joints of TF coils are designed to ensure the structural integrity of the joints without any maintenance once it is assembled. The novel mitigation technique implemented, is the use of multifunction wedge lock washers. The wedge type lock washer prevents bolt loosening caused by vibration and dynamic loads, compensates for loss of preload due to settlement and relaxations. It facilitated accurate preload, increased operational reliability while reducing the risk of unplanned production stops and accidents. The support structure as shown in figure-2 is designed and implemented along with a pair of lock washers used in each bolted joints in the TF Coil.
The damaged coils are refurbished using in-house techniques. These 20 TF coils are assembled one by one on Test Stand by joining both C’s sections and the electrical parameter testing (Resistance and Inductance) of the coils have been carried out. The measured values of all coils are within satisfactory limits and found that they are in good condition to be reused again. After successful tests, TF coils are assembled on the ADITYA-U tokamak as shown in figure-3.
The EM forces, support structure design and implementation to ensure the structural reliability of Toroidal Field magnet system of ADITYA-U tokamak is presented in this paper.Speaker: bharatkumar doshi (Institute for plasma research) -
Previous studies of the European Demonstration fusion reactor concept (DEMO) have shown that the expected amounts of radioactive waste at reactor end of life (EOL) can be of the order of 10$^4$ tonnes [refs 1,2,3]. These studies also suggested that comparable amounts of waste will be classified as low level waste (LLW) and intermediate level waste (ILW) 100 years after DEMO EOL. Since these studies were performed, updated models for the DEMO reactor have been developed. To assess what effect these changes have had of the waste expectations from DEMO new waste assessments have been performed.
These followed the same methodology as the previous studies: Monte-Carlo neutron transport calculations were performed on the DEMO design geometry and the resulting cell tallied neutron energy flux spectra are used in high fidelity inventory simulations, to find the expected activation of reactor components. MCNP v6.2 [refs 4,5] was used for the transport calculations and FISPACT-II [ref 6] for the inventory simulations. Two blanket concepts have been used in this study: the Helium cooled Pebble Bed (HCPB) and the water cooled Lithium-Lead (WCLL) designs. The complete reactor model has been studied and assessed in accordance to the UK LLW criteria, with the possibility of material being Non-active waste (NAW) assessed on IAEA clearance index.
The results of the current assessments suggest that the waste performance of the DEMO reactor remains comparable to previous work, the waste mass evolution of the current DEMO model are shown in figure 1. In both models studied the majority of reactor material is expected to require disposal as radioactive waste, with 1-2$\times10^4$ tonnes being classified as ILW, possibly needing geological disposal, 100 years after EOL. The WCLL model has greater total mass due to the Pb content of the LiPb breeder material. The LiPb also provides greater levels of neutron shielding, which is the cause of larger amount of NAW in the WCLL model.
A significant proportion of the ILW mass from DEMO arises from activated structural components in the near plasma blanket region. Current plans use Eurofer steel which is expected to produce long lived activation products. These can include $^{14}$C, $^{53}$Mn and $^{94}$Nb, the presence of which can cause difficulties when attempting to achieve LLW waste criteria. The activation profile expected in DEMO Eurofer is shown in figure 2, revealing $^{14}$C as the major cause of failing to meet the LLW limit used in this work.
It has been suggested that the Carbon content of activated steels could be reduced to 1 weight part per million via a so called decarburization process [ref 7]. The process, where Oxygen is blown across the surface of molten steel to create CO which is eventually captured as solid CaCO$_3$, has been claimed capable of reducing carbon content of steels to 1 weight part per million. The affect of such a technique has been applied to the Eurofer results from the inventory simulations and the resulting CaCO$_3$ inventory has also been estimated, the resulting ILW masses are shown in figure 3.
It was found that decarburization could improve the expected waste evolution of DEMO Eurofer, but some secondary ILW CaCO$_3$ was produced. While this technique has the potential to improve the waste disposal prospects of in-vessel fusion steels, there are still a number of issues that must be resolved before it could be adopted as a DEMO waste mitigation strategy. These include proper assessment of the secondary waste burden from CaCO$_3$ and whether it can be safely applied to large volumes of activated steel (it has only previously been tested on non-active material).
The $^{94}$Nb content in activated Eurofer is most commonly a result of neutron capture reactions on Nb impurities. The reduction of these impurities may also improve waste performance. Unfortunately the global activity limits used by UK criteria mean a reduction $^{94}$Nb would not improve waste classification, as $^{94}$Nb is dwarfed by other activity sources, see figure 2. Nb reduction can have an affect when individual nuclide limits are applied, such as those in different waste management systems. For example the French LLW system, which is based on individual nuclide activities, allows 9.2$\times10^{7}$ Bq/kg of $^{14}$C and 1.2$\times10^{5}$ Bq/kg of $^{94}$Nb. Comparing these to the activities shown in figure 2 reveals that these criteria may make decarburization unnecessary, but Nb impurity reduction could provide a significant improvement in long term waste classification. The possibility of applying waste mitigation techniques, such as decarburization or Nb-reduction, need only be considered if relevant to the waste regulations in-force at the chosen site for DEMO.
The expected levels of radioactive waste is an ongoing issue for the DEMO reactor concept. The application of waste mitigation techniques could lower the amount of ILW, but it will remain on the order of 10$^4$ tonnes on decommissioning (approximately 100 years post EOL) time scales. It should be noted that any waste mitigation techniques applied after EOL will produce secondary wastes which need to be included in complete reactor waste assessments.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
- M. R. Gilbert et.al , "Activation, decay heat, and waste classi?cation studies of the European DEMO concept", Nucl. Fusion, vol. 57,p. 046015, 2017.
- M. R. Gilbert et.al , "Waste assessment of european DEMO fusion reactor designs", Fus. Eng. Des., vol. 136, pp. 42-48, 2018.
- M. R. Gilbert et.al , "Waste implications from minor impurities in european DEMO materials", Nucl. Fus., 2019.
- C. J. Werner et.al , "MCNP6.2 Release Notes", Tech. Rep. report LA-UR-18-20808, Los Alamos National Laboratory, 2018.
- C. J. Werner et.al , "MCNP Users Manual - Code Version 6.2", Tech. Rep. report LA-UR-17-29981, Los Alamos National Laboratory,2017.
- J. -Ch. Sublet et.al , "FISPACT-II: An advanced simulation system for activation, transmutation and material modelling", Nucl. Data Sheets, vol. 139, pp. 77-137, 2017.
- L. Di Pace et.al , "Feasibility studies of DEMO potential waste recycling by proven existing industrial-scale processes", Fusion Engineering and Design, vol. 146, pp. 107-110, 2019. SI:SOFT-30.
Speaker: Greg Bailey (CCFE) -
The China Fusion Engineering Test Reactor (CFETR) is the next device in the roadmap for the realization of fusion energy in China, which aims to bridge the gaps between the fusion experimental reactor (ITER) and the demonstration reactor (DEMO). For the fusion reactor, the neutronic and shielding assessment is indispensable for the performance and safety design of the reactor, which strongly impact the key design features, such as tritium self-sufficiency, heat generation, activation of structural materials, shielding and radioactivity safety, etc. Based on the latest design of CFETR, the preliminary neutronic analysis for the major components of CFETR tokamak have been performed. The key nuclear parameters, such as tritium breeding ratio (TBR), neutron wall load of first wall, neutron flux, nuclear heating, burn-up of breeding material, activation data, material damage (dpa), gas production rate, etc., have been obtained and analyzed, also the impact on TBR by the ports of neutral beam injection (NBI) is studied. Meanwhile, the shielding performance is assessed and the optimization measures are proposed to improve the shielding capability to the key components of the CFETR tokamak.
Based on the current physics design and major components design of CFETR, three-dimensional neutronic analysis model including tritium breeding blanket, divertor, vacuum vessel, thermal shield, toroidal field (TF) coils, poloidal field (PF) coils, central solenoid (CS), cryostat, bio-shield and port models, has been developed to be used for neutronic and shielding analyses of CFETR. The neutronic analysis model represents a 22.5° regular sector model of CFETR with reflecting boundary and standard D-T neutron source.
The tritium breeding blanket (TBB) is one of the most important components of CFETR with functions of tritium breeding, heat generating and nuclear shielding. Currently, the helium cooled ceramic breeder tritium breeding blanket (HCCB TBB) is the primary option for CFETR, which adopts high pressure helium gas as coolant, ceramic lithium silicate as tritium breeder and beryllium as neutron multiplier. After several rounds of neutronic design and optimization, the TBR of HCCB TBB is 1.166 which could well meet the target value. If consider 3 NBI ports of CFETR, the TBR decreases to 1.116. After 10 years operation with 1 GW fusion power and 50% duty factor, the TBR of HCCB TBB reduces by 1.8%. The total nuclear heating of HCCB TBB is 1420 MW, which is removed through the helium cooling system and used to generate electricity. For other major components of CFETR tokamak, nuclear heating with a detailed distribution has been obtained too, which is the important input parameter for thermo-hydraulics analysis and safety analysis. According to the preliminary operation plan and maintenance plan of CFETR, the activity and decay heat of each major component of CFETR have been evaluated by means of particle transport and activation coupling calculation. These activation data can provide reference and basis for rad-waste classification and disposal of CFETR.
The major components of CFETR tokamak, including vacuum vessel, tritium breeding blanket, divertor, thermal shield, shall provide adequate shielding to ensure that the TF coils will not be damaged by irradiation during the whole life of CFETR. In order to assess the shielding performance of TF coils, the overall shielding performance of CFETR has been evaluated based on the three-dimensional neutronic analysis model. Nuclear parameters such as fast neutron fluence and nuclear heating have been calculated and analyzed for TF coils at different positions. The results show that the shielding performance of TF coils nearby divertor, NBI port and high magnetic field areas could not meet the design requirements comparing with the irradiation limits. Combined with the engineering design requirements of CFETR, the preliminary shielding design and optimization for the divertor, NBI port and inboard areas have been performed. After optimization, the shielding performance of CFETR could basically meet the design requirements of TF coils.
Meanwhile, neutronic analysis for the components and areas outside cryostat of CFETR is ongoing, which will provide the radiation field distribution for the tokamak building and the input for the shutdown dose rate analysis, etc.
Acknowledgements
This work was supported by the National Key R&D Program of China with grant numbers 2017YFE0300501 and 2017YFE0300503, and the National Natural Science Foundation of China with grant number 11905046.Reference
[1] Y.X. WAN, J.G. LI, Y. LIU, et al., “Overview of the present progress and activities on the CFETR”, Nuclear Fusion, 2017, 57: 102009
[2] G. Zhuang, et al., Progress of the CFETR design. Nuclear Fusion, 2019, 59: 112010
[3] X.Y. WANG, X.H. WU, et al., “Design Progress of CFETR HCCB TBB”, presented at CFETR 2019 annual meeting, Huangshan, China, Sep.24-28, 2019
[4] X.H. Wu, et al., “Optimized Design and Engineering Analyses for HCCB Blanket of CFETR”, Presented at 27th International Conference of Nuclear Engineering, Tsukuba, Japan, May 19-24, 2019Speaker: Mr Qixiang CAO (Southwestern Institute of Physics) -
Experimental tests on a prototype system of a novel, rapid time-response disruption mitigation system (DMS) being developed as a back-up option for ITER, referred to as the Electromagnetic Particle Injector (EPI) have been able to verify the primary advantages of the concept, which are its ability to meet short warning time scales of <10ms while attaining the projected high velocities for deep radiative payload penetration in ITER-scale plasmas. Because the ITER plasma would have about two-orders of magnitude more energy than in present experiments, realistic 3d MHD simulations, benchmarked against present experiments is an essential step to project to ITER. In support of this requirement, new capabilities have been implemented in the M3D-C1 code to model radiative material injection into tokamak plasmas, and initial simulations for the NSTX-U configuration have been conducted.
Predicting and controlling disruptions is an important and urgent issue for ITER. While a primary focus is the early prediction and avoidance of conditions favorable to a disruption avoidance, some disruptions with a short warning time may be unavoidable. For these cases, a fast time response disruption mitigation method is essential.
Recent studies on DIII-D [Ref. 1] show that the overall response time of the Shattered Pellet Injector (SPI) is 25ms, and because of much larger pellet size and distances, this system is likely to be significantly slower in ITER. Additionally, as the plasma energy content is increased from 0.2MJ to 2MJ, the penetration depth shows a strong reduction [Ref. 1]. ITER, with 350MJ stored plasma energy, could benefit from a DMS that has both fast response time and higher-velocity such as from the EPI. Calculations [Ref. 2] suggest that the high-velocity pellets from EPI could penetrate deep into ITER grade plasmas. The present understanding, based on the theoretical work of Konavalov, et al., is that as little as 5g of Be, if it is deposited deep inside the plasma, may be adequate for both thermal quench and runaway electron mitigation in ITER [Ref. 3].
As shown in Fig 1, the EPI relies on an electromagnetic propulsion system to overcome the limitations of present gas-based systems, such as SPI, which are limited to 200m/s for large mass pellets. A metallic sabot is accelerated electromagnetically to the required velocities (> 1km/s) within 2ms, at which point it releases well-defined microspheres, or a shell pellet, of a radiative payload.
Initial experimental tests from the prototype system (EPI-1) have demonstrated 150m/s within 1.5ms, consistent with calculations, giving confidence that larger ITER-scale injector can be developed [Ref. 4]. Following these successful experiments, a new upgraded system, (EPI-2) in a tokamak deployment configuration has been built to increase the velocity to 1km/s. Initial results from the operation of this system at 2.1T have extended the attainable velocities to over 600m/s in the same 1.5ms, consistent with the projections for this system which indicate the attainment of 1km/s with the use of 3T boost magnetic field.
Fig. 2a shows experimental data from the operation of EPI-2. The magnetic field probe traces that record the expanding magnetic flux behind the sabot show continuous acceleration along the 30cm long acceleration region. The processed data from these signals (Fig. 2b) shows the attainment of over 600m/s in about 1ms. Near-term tests plan to extend this velocity to 1km/s by operating the boost coils at 3T. Basic aspects of payload separation from the sabot, and sabot capture have also been demonstrated on EPI-2 at 150m/s, and the method can be extended to over 2km/s.
Unlike the SPI case, the EPI injects a payload consisting of well-defined shape and velocity. Consequently, the 3d MHD modeling of EPI payload penetration should be easier, and more precise, permitting reliable benchmarking against present experiments.
In support of this EPI work, new capabilities have been implemented in the M3D-C1 code. We have implemented a carbon ablation and radiation model into the M3D-C1 code [Ref. 5] and used it to simulate EPI of solid carbon pellets into NSTX-U. (Fig. 3).
The ablation model is based on a neutral gas shielding approach (NGS)[Ref. 6,7] in which the key quantity is the shielding factor which is the ratio of the plasma heat flux that has reached the pellet surface and to the plasma heat flux before entering the pellet neutral cloud.
In the case of strong shielding pellets, an ablation rate equation that is related to the electron temperature, electron density, pellet radius, pellet charge and atomic mass number and the adiabatic index is used. In the weak shielding approximation, the ablation rate is related to the pellet radius, electron density, electron temperature, and the shielding factor and is given in the reference by Kuteev [Ref. 7]. In the intermediate shielding regime, Ref. [6] proposes an interpolation. Both intermediate and weak regimes agree well with a series of AUG shots with C pellets [Ref. 6]. The radiated power is calculated using the KPRAD [Ref. 8] module.
At the very least, it is highly desirable for the DMS to deposit a major portion of the radiative payload inside the q=2 surface in less than the thermal quench duration (~2ms) after the radiative material contacts the plasma edge. In addition, if the radiative payload transit time from the q=2 surface to the magnetic axis is less than the thermal quench duration, then additional advantages, such as the possible suppression of the initiation of runaway electrons becomes possible. The EPI is potentially capable of meeting these important needs for ITER. Because of the well-defined payload velocity and size, the 3d MHD modeling should allow more realistic benchmarking with experimental results from present tokamaks, which is an essential step for reliably projecting disruption mitigation to ITER plasmas.
*This work is supported by U.S. DOE Contracts: DE-AC02-09CH11466, DE-FG02-99ER54519 AM08, and DE-SC0006757.
- R. Raman et al., Nucl. Fusion (2020) https://doi.org/10.1088/1741-4326/ab686f
- R. Lunsford et al., Fusion Science Technol. (2019) DOI: https://doi.org/10.1080/15361055.2019.1629246
- S.V. Konovalov S.V. et al 2012 IAEAFEC 2012 Conf. ITR/ P1-38 (www-naweb.iaea.org/napc/physics/FEC/FEC2012/
papers/338_ITRP138.pdf) - R. Raman et al, Nucl. Fusion 59 (2019) 016021
- N. Ferraro, B. Lyons, C. Kim, et al, Nucl. Fusion 59 016001 (2019)
- Sergeev et al., Plasma Phys. Rep. 32 (2006) 363
- Kuteev et al., Sov. J. Plasma Phys. 10 (1984) 675
- D.G. White, et al, DIII-D Technical Report GA-A22639, General Atomics (1997)
Speaker: Mr Roger Raman (University of Washington) -
Effective thermal energy utilization from in-vessel components is important issue to develop the plant concept of a fusion DEMO, because there are several in-vessel components in a fusion power plant. How to use of divertor thermal energy is now a key for development of the plant concept for JA DEMO. Following two options are newly proposed as for the divertor thermal energy considering tritium permeation: its application to electric power generation and to hydrogen production. Furthermore, adaptability of those options against the future grid system is also evaluated in this paper.
1. Introduction
Thermal power outputs have been evaluated and the primary heat transfer system (PHTS) has also been developed for Japan’s fusion DEMO (JA DEMO) [11,12,13]. Thermal power outputs from in-vessel components are shown in Fig.1. Heat utilization from divertor with Cu-alloy pipe (Div-Cu-alloy) is remained to be applied to energy production system in JA DEMO. In this paper, we propose two options for heat utilization from Div-Cu-alloy and develop the overall plant concept of energy production in JA DEMO with careful consideration of tritium permeation into the coolant. Adaptability of those options against the future grid system is also evaluated.
![Thermal power from in-vessel components [11,12]](https://workshop.nifs.ac.jp/fec2020/image/22-Hiwatari-image-fig_1.jpg)
2. Two options for plant concept for energy production in JA DEMO
A world energy scenario compatible with Paris Agreement has suggested that application of fusion energy not only to power generation but to hydrogen production is significantly effective in the future Japan’s energy market [14]. Those options are pursued using the Div-Cu-alloy thermal power.
First, hydrogen production system is developed. The high-temperature steam decomposition method for hydrogen production is proposed. To get the high-temperature steam, thermal energy from the Div-Cu-alloy (Max. 172MWth) and backplate (16MWth) is applied. System diagram and thermal balance for hydrogen production are shown in Fig.2 and Fig.3.
Such effective use of waste heat for hydrogen production has been proposed in a commercial reactor design [15] This hydrogen production concept is applied to JA DEMO plant system, but the concept is improved to avoid the tritium permeation using heat exchanger and steam generator (Fig.2). The vaporized steam is heated up to around 1200K in the heat exchanger, and then hydrogen and oxygen are generated in the electrolysis cell of high temperature steam. Thermal power of 42MWth from the backplate &VV and Div-Cu-alloy and electricity 229MWe from generator enables to produce hydrogen 1.77kg/s at the rating operation of hydrogen production.
Second, application to electric power-up system is developed. Pre-heating system of turbine coolant by Div-Cu-alloy heat is installed into PHTS between steam turbine downstream and steam generator in Fig.2. Resultantly, the extracted steam from upstream of turbine for pre-heating can be reduced, ant then electric power generation can increase by 22.4MWe. When the coolant from Div-Cu-alloy is applied to pre-heating system, an intermediate heat exchanger is also installed to avoid the tritium permeation to the secondary coolant. The tritium permeation to the secondary coolant in PHTS from blanket is also evaluated, and the result shows the total permeated tritium less than Japan’s light water reactor limit. Because of comparison of permeated influx to coolant in blanket (3.0g/day) and divertor (1.25x10-2g/day) which are evaluated with TMAP code, permeated tritium from Div-Cu-alloy to the secondary coolant is considered to be negligible.3. Adaptability to partial load operation in the future grid system
Development of partial load operation is recommended in Japan’s Policy to promote R&D for a fusion DEMO reactor [16]. As a typical case, adaptability to Duck Curve [17] in the grid is examined for the above two plant options.
In case with the hydrogen production system, the basic operation concept of JA DEMO for Duck Curve is as follows: During ramp-down and over-generation phase, surplus electricity is applied to generate hydrogen. Even in the over-generation phase, electric power generation might be stopped by using turbine bypass in Fig.2. During ramp-up phase of Duck Curve, electricity for hydrogen generation reduces rapidly, and the rating electricity is transfer to the grid. Fig.4 shows the typical operation pattern adaptable to Duck Curve by production of both electricity and hydrogen.
In case with the electric power-up system, the turbine bypass system serves as the main role in Fig. 2. The time-scale difference between plasma ramp-up (~100 sec) and turbine ramp-up (~ hour order) has been made clear. We decided to install the turbine bypass to synchronize between plasma and turbine to compensate this time-scale difference. This turbine bypass is effectively applied to carry out the partial load operation similar to electric output in Fig.4. This system has enough resilience to play a role as a controller in the grid.
In conclusion, this paper reveals that fusion energy can play not only the role of the baseload, but also the role of ancillary service in the grid. To show this fact, first, the plant concepts for production of both electricity and hydrogen in JA DEMO are developed with careful consideration of tritium permeation. Second, operation methods adaptable to partial load operation such as Duck Curve by electricity and hydrogen production are proposed.
[11] R.Hiwatari et al.”Development of plant concept related to tritium handling in the water-cooling system for JA DEMO”, Fus. Eng. Des. 143(2019)259
[12] Y.Miyoshi et al., ” Cooling water system design of Japan's DEMO for fusion power production”, Fus. Eng. Des. 126 (2019)110
[13] Y. Sakamoto et al., “Development of Physics and Engineering Designs for Japan’s DEMO Concept,” IAEA 27th Fusion Energy Conf., FIP/3-2, Gandhinagar, India, October 22-27, 2018
[14] K. Gi et al., “Potential contribution of fusion power generation to low-carbon development under the Paris Agreement and associated uncertainties”, Energy Strategy Reviews 27 (2020) 100432
[15] K. Okano, Y. Asaoka, R.Hiwatari et al., Journal of plasma fusion research Vol.77 p.601 (2001) (in Japnaese)
[16] Japan’s Policy to promote R&D for a fusion DEMO reactor, MEXT, Japan, (2017)
[17] California ISO Fact sheet, “What the duck curve tells us about managing a green grid”, 2016Speaker: Dr Ryoji Hiwatari (National Institutes for Quantum and Radiological Science and Technology) -
Nuclear performance evaluation is the core basis of tritium breeding blanket design and also the key input for thermo-hydraulic and thermo-mechanic numerical analysis. Whereas, the tritium breeding blankets have the features of complex structure, heterogeneous neutron flux distribution and a long energy span of neutron, which are considerable challenges for neutronienter image description herecs modeling and transport calculation. Normally the homogeneous blanket models, especially for the pebble bed of tritium breeder and neutron multiplier, are used in worldwide for the neutronics calculation, however, the accuracy and applicability of this method need to be assessed considering the heterogeneous configuration of blanket. Referring to the blanket design of China Fusion Engineering Test Reactor (CFETR), the homogeneous model and high-fidelity model for neutronics calculation have been built. The neutronics effect of the homogeneous structure and the space self-shielding effect of pebble beds are studied separately considering two different types of coolant, helium and water. The calculation deviation of the homogenous model comparing with the high-fidelity model is obtained, which will provide strong support for neutronics design and optimization of tritium breeding blankets, such as Chinese ITER helium cooled ceramic breeder test blanket module (HCCB TBM) and CFETR blanket.
Firstly, the homogenous neutronics model (homogeneous both in structure and pebble beds, Model A), half homogeneous neutronics model (homogenous only in pebble beds, Model B) and high-fidelity neutronics model (heterogenous, Model C) are performed by using the McCAD code individually. In the homogeneous model, different materials of the breeding blanket are mixed together according to their volume weights in each functional region. The reflecting boundaries are applied in these three models, including both toroidal and poloidal directions. Significantly, a packing fraction of 52.36% (simple cubic packing) in both the Li4SiO4 and Be pebble beds is assumed. A general neutron source is adopted and it is a Gaussian fusion energy spectrum which is added in the front of the first wall (FW). ODS steel is selected as the structure material, and Li4SiO4 as the breeding breeder in the pebble bed regions and Be is utilized as the neutron multiplier.
Secondly, Model A and B are considered for studying the neutronics effect of homogeneous structure on the solid breeder blanket. The MCNP code is applied for the 3D neutronics transport calculation for the solid breeder blanket, and FENDL2.1 is used. Nuclear performance evaluation, including the TBR (Tritium Breeding Ratio), neutron flux and neutron energy spectrum of each tritium breeding region is performed. Also, the neutron mean-free paths of Li in each tritium breeder regions are calculated. The results indicate that the homogeneous structure has small impacts (~0.30% TBR overestimation from 1.2162 to 1.2194) on the performance of the helium-cooled blanket concept, but makes the TBR overestimated by ~2.48% (from 1.0173 to 1.0425) in the water-cooled blanket concept due to the moderation of water towards neutrons.
Thirdly, the diameter of the pellet is assumed to be 0.8-1.2mm and there are more than one hundred million (~1E8) pellets in a single solid breeder blanket module and the MCNP input file will exceed ~3GB , which demands a huge amount of computer memories making it difficult to get solution. Therefore, two simplified models, Model D1 (homogeneous in structure, but real structure for pebbles) and Model D2 (homogeneous both in structure and pebble beds), are adopted. The space self-shielding effect is assessed in Model C and Model D1 with pellets of 1cm in diameter individually. Results show that the TBR overestimations are ~0.33% and ~0.30% in helium-cooled concept and ~4.02% and ~3.78% in water-cooled concept separately, which indicates that these two results are in good coincidence with each other. Therefore, simplified models are verified to be rational for performing space self-shielding neutronics analysis with pellets of real size.
Finally, nuclear performance evaluation with pellets of real size is performed and results indicate that the space self-shielding effect caused by the pebble beds has little influence on the neutronics performance if it is helium-cooled, yet ~1.28% overestimation for the TBR if it is water-cooled with 1mm diameter pellets. Based on Model D1, the TBR vs the diameter of pellets is also studied which implies that the overestimation could be omitted if the pellet diameter gets smaller to ~0.1mm for water-cooled blanket concept.
In conclusion, the homogeneous model is rational for neutronic analyses if the coolant is helium. Yet, there is non-negligible overestimation for the TBR of water-cooled blanket concept with more than 1mm diameter pellets, and high-fidelity model should be adopted during the neutronics transport calculation.
ACKNOWLEDGMENTS
The work at SWIP (Southwestern Institute of Physics) was supported under National Natural Science Foundation of China Number 11905046 and National Key R&D Program of China Number 2017YFE0300503 and 2017YFE0300601. Also acknowledge to the KIT (Karlsruhe Institute of Technology) for the development of McCAD code.
REFERENCES
1. Y. X. WAN, J. G. LI, Y. LIU, X. WANG, V. CHAN, C. CHEN, “Overview of the Present Progress and Activities on the CFETR.” Nuclear Fusion 57.10(2017):102009.
2. G. ZHUANG, G. Q. LI, J. G. LI, Y. X. WAN, Y. LIU, X. WANG, “Progress of the CFETR Design.” Nuclear Fusion (2019).
3. X. Y. WANG, K. M. FENG, Y. J. CHEN, L. ZHANG, Y. FENG, X. H. WU, “Current Design and R&D Progress of CN HCCB TBS.” Nuclear Fusion (2019).
4. X. Y. WANG, Development Status of Helium Cooled Ceramic Breeder Tritium Breeding Blanket (HCCB TBB) in China [R], Budapest Hungary, 2019.*Speaker: Mr Shen QU (Southwestern Institute of Physics) -
The electrochemical technique is shown to be effective to develop liquid blanket system, not only for the molten salt one but also for the liquid metal ones. (i)Oxygen impurity in lead lithium eutectic alloy was removed as CO2 gas by electrolysis in the molten salt contacting with the liquid metal, as well as (ii)nitrogen impurity in liquid lithium. Considering the electrochemical compatibility of lithium alloy and salts, (iii)new liquid blanket system, lead lithium intermetallic compound / chloride molten salt blanket system, is also proposed and its basic characteristics are investigated.
In order to realize a liquid blanket, it is important to develop purification systems including the detritiation system and to clarify and control corrosion behavior. These are basically required for any liquid breeding material candidate, lithium, lead-lithium eutectic alloy, or molten salt. In the research on fluoride molten salts, such as FLiBe, removal of impurities and recovery of tritium using electrochemical methods, including redox control, have been widely performed. As well as those of FLiBe, those of liquid tritium breeding metals can be solved using molten salts. In this research, electrochemical removal of nitrogen impurity in liquid lithium and oxygen in lead-lithium eutectic alloy, both of them are very important form the viewpoint of corrosion reduction, are focused. Furthermore, a new liquid blanket system, Li-Pb intermetallic compound with chloride molten salt, is proposed considering its electrochemical characteristics.1. Removal of oxygen impurities from lead-lithium eutectic alloy
Oxygen impurity in lead-lithium eutectic alloy has negative effect in terms of corrosion to structural materials and deterioration of functional materials. The cold trapping method -removal of precipitates in cooled section- or the hot trapping method -reaction with much stronger oxide forming material- can be used, though the achievable concentration level or the difficulty in continuous processing are inevitable. However thermodynamically, oxygen impurity in the eutectic alloy is expected to be distributed to the molten salt as lithium oxide when the alloy is in contact with molten salts. The lithium oxide in molten salts can be removed from the system by electrolysis as well as the reduction of lithium which is fed back to the alloy. In our research, cyclic-voltammetry (CV) measurement was performed in the eutectic chloride molten salt LiCl-KCl contacting with the liquid eutectic alloy using a glassy-carbon electrode (Fig.1). The change of CV was obtained when oxygen impurity was directory fed into the liquid alloy as shown in Fig.2 through the tube 6 in Fig.1. In addition, the weight loss of the carbon electrode during the continuous electrolysis was equivalent to the value calculated from the amount of charge assuming the formation of CO2 gas. Based on these results, this electrochemical method is show to be effective for the removal of oxygen impurity from eutectic alloys.
Figure 1. Expetimental setup of electrochemical purification.
(1) counter/reference electrode (Li-Pb) (2) working electrode (glassy carbon) (3) LiCl-KCl eutectic melt (4) Li-Pb (5) stainless crucible with an electric heater (6) stainless tube (7) K-type sheathed thermo-couple
Figure 2. Change of CV depending on the addition of oxide impurity into liquid metal. Increase of oxide impurity increases the current, whih is consumed for the formation of CO2 gas.[A]2. Removal of nitrogen impurities from lithium
It is known that lithium has a specific reactivity with nitrogen unlike other alkali metals, and nitrogen in lithium is reported to degrade the compatibility with structural materials and the tritium trapping material such as yttrium. Several investigations have been done to remove nitrogen in lithium using titanium or titanium-containing alloys, but the method requires high reaction temperature and replacement of nitrogen trapping materials. The recovery of hydrogen from lithium in contact with molten salts has been studied in the United States in the 1980s, and the similar method is applicable to the removable of nitrogen. The method have advantages such as low-temperature operation of about 673K and the possibility of continuous operation. In our research, the CV change of a nickel electrode in LiCl-KCl salt contacting with liquid lithium is obtained, as shown in Fig. 3, which indicates that nitrogen impurity in lithium can also be electrochemically removed.
Figure 3. Proportional relationship between the gradient of CV and added Li3N. The current is increased as nitrogen impurity is added to lithium to be removed through electrolysis [B]3. Lead lithium intermetallic compound / chloride molten salt blanket
The lead and lithium alloy system is known to have high melting point intermetallic compound phases at around 75 at% lithium concentration. Although those are kinds of solid breeding materials, it is considered difficult to recover tritium by gas sweep due to the high lithium activity. However, it can be inferred from the example of lithium that the molten salt contacting with Li-Pb compound will contain LiT distributed from Li-Pb phase when tritium is generated. Here, as a molten salt, BeF2-based salts, such as FLiBe and FLiNaBe, usually considered as a candidate blanket material is not suitable because of the reactivity with lithium in the Li-Pb compound. In contrast, a molten salt such as LiCl-KCl is suitable because of the compatibility to the Li-Pb compound. In this case, the Li-Pb compound acts as an electrochemical redox controller (reduction of TCl), as metal Be in fluoride salt, and has a certain neutron multiplication effect and a long-life breeding material due to the high lithium concentration. TBR of the Li-Pb compound / chloride salt system varies depending on the isotope enrichment ratio of lithium and chlorine, volume fraction, etc., but is calculated to be 1.1 or more in some cases. The concept of an electrochemical impurity treatment system is being studied based on experimental data.To realize the liquid breeder blanket system, the electrochemical method is shown to be effective not only for the fluoride molten salt system, which has been focused on, but for liquid metal blanket systems such as lead-lithium eutectic alloy and liquid lithium.
[A] T. Okada et al., 14th Int’l. Symp. Fusion Nucl. Tech. (ISFNT-14 ) (2019), Budapest
[B] H. Miyagaki et al., Fusion Sci. Tech. (TBP), DOI: 10.1080/15361055.2020.1716457Speaker: Juro Yagi (Institute of Advanced Energy, Kyoto University) -
Experiments on DIII-D support a new approach, confirmed by transport modeling, to achieving Q=10 in ITER using a scenario with low plasma current (~ 8 MA), high $\beta_{\rm p}$, and line-averaged Greenwald fraction ($f_{\rm Gw}$) above 1. At 8 MA the disruption risk and ELM challenge are greatly reduced, with the possibility that uncontrolled ELMs may be acceptable$^1$. Due to the need of sufficient fusion power and low plasma current, this approach requires high density with $f_{\rm Gw}$>1.0 simultaneous with high confinement quality ($H_{\rm 98y2}$>1). Using impurity injection, the recent DIII-D experiments achieve and maintain these simultaneous conditions. Previously, high $\beta_{\rm p}$ plasmas with $f_{\rm Gw}$ up to ~1.0 and $H_{\rm 98y2}$>1 were obtained in JT-60U, albeit transiently and usually operated at low absolute density$^{2,3}$, which is not favorable to reactor plasma. For the first time in a tokamak, experiments demonstrate that a stationary ITB at large radius ($\rho$~0.7) is compatible with $H_{\rm 98y2}$>1, at reactor-level absolute density ($n_{\rm e0}>1.0×10^{20}\;m^{-3}$), $f_{\rm Gw}$>1, and reactor-relevant $q_{95}$ as well (fig. 1). Such ITB is a key feature of the ITER 8 MA Q=10 modeling. Comparison between the experimental DIII-D profiles and the predicted ITER profiles shows also a good match of ITB location and profile shape (fig. 2). The DIII-D experiments confirm that the high density ITB in ITER modeling is achievable at similar $q_{95}$ using the high $\beta_{\rm p}$ scenario.
For the first time, neon injection is observed to trigger the formation of a large radius ITB in the density channel at reactor level density on DIII-D, which provides an effective experimental approach to achieve line-averaged Greenwald fraction well above 1. Pedestal density feedback control is used in the experiment. Therefore, the pedestal density is kept below the Greenwald limit, e.g. $f_{\rm Gw,ped}$<0.7. The large radius ITB strongly elevates the plasma density inside the pedestal. The density profile is fairly flat inside the ITB with $n_{\rm e0}>1.0×10^{20}\;m^{-3}$ (fig. 2). These high $\beta_{\rm p}$ experiments on DIII-D also show that the ITB is sustained long after (>8×the energy confinement time, $\tau_{\rm E}$) the neon injection is turned off. Impurity transport modeling based on the experimental data shows that neon provides electron source at $\rho$~0.75, which is the location where the foot of the ITB emerges. One of the possible mechanisms for the trigger of ITB formation could be that the important electron source creates a seed of local density gradient. The locally increased density gradient strengthens the effect of $\alpha$-stabilization of turbulence by reducing the non-adiabaticity of electron response$^4$, and starts a positive feedback leading to ITB formation. The same technique may also work for ITER high $\beta_{\rm p}$ plasma, when higher Z impurity is employed.
Using this technique, new high $\beta_p$ experiments on DIII-D demonstrate the stationary large radius ITB at reactor level density and reactor relevant $q_{95}$. Fig. 1 shows that the stationary phases of $f_{\rm Gw}$~1.0-1.1 and $f_{\rm Gw}$~1.3 at $q_{95}$~8 are sustained for 21 and 8 $\tau_{\rm E}$’s, respectively. These plasmas have even higher density than ITER at its 9 MA Greenwald limit. With ITB, the energy confinement is well above standard H-mode in these plasmas, e.g. $H_{\rm 98y2}$ up to 1.4. The degradation of confinement in one of the discharges shown in fig. 1 ($f_{\rm Gw}$~1.3 case) is due to excessive neon injection causing very high core radiation. The averaged neon injection rates are $1.3×10^{20}$ /s ($f_{\rm Gw}$~1.3) and $4.8×10^{19}$ /s ($f_{\rm Gw}$~1.1). With further optimization of the impurity injection waveform, sustained high confinement is also expected in the $f_{\rm Gw}$~1.3 case. These high $\beta$ ($\beta_{\rm N}$≤3.5, $\beta_{\rm p}$≤2.7) plasmas have $q_{\rm min}$>2.0 and quiet MHD behavior. Meanwhile, a non-inductive current fraction up to 0.9 is achieved simultaneously in the high density phase.
1D Transport modeling suggests the goal of Q=10 on ITER can be achieved at low plasma current (~ 8 MA) using the high $\beta_{\rm p}$ scenario. The simulations are performed using the STEP module in the OMFIT framework, integrating sub-modules for transport, heating, current drive and equilibrium calculations. Compared to previous ITER simulation works, the innovative features in this work include: 1) Plasma density, temperature, current profiles and equilibrium are all evolved in the simulations, while pedestal heights for density and pressure are prescribed to values slightly below the Greenwald limit (for density) and EPED prediction (for pressure); 2) Use TGLF model for transport prediction; 3) In TGYRO, E×B shear effect on turbulence suppression is turned off. Simulations use ITER “Day One” heating and current drive power: NBI≤33 MW, EC≤20 MW. Calculations predict that the following parameters can be achieved: Q=9.5±2.5, $I{\rm p}$=7.85±0.35 MA, $q_{95}$=7.54±0.39, $H_{\rm 98y2}$=1.74±0.13, $f_{\rm Gw}$=1.48±0.13, $f_{\rm NI}$≥98% and $P_{\rm fus}$=350±50 MW. A large radius ITB is shown in both temperature and density channels for electron and ion species (fig. 2).
Density profiles experimentally achieved on DIII-D match the density profiles simulated for ITER Q=10 in shape (ITB radius) and absolute core density value (fig. 2), while the core confinement quality is maintained well above standard H-mode levels. Although the experimental electron temperature is low compared with ITER simulation results, its shape is still a very good match to the ITER simulation results, if a multiplier of 9 is applied to the DIII-D data. The results confirm that the required large radius ITB at $f_{\rm Gw}$ above 1 and reactor level density in ITER Q=10 modeling is achievable experimentally at similar q95 using the high $\beta_{\rm p}$ scenario. These new high $\beta_{\rm p}$ experiments on DIII-D strongly support the ITER Q=10 simulations and pave the avenue to a new low plasma current (~ 8 MA) approach for ITER’s Q=10 Goal.
This work was supported in part by the US Department of Energy under DE-FC02-04ER54698 and DE-SC0010685.
$^1$R. A. Pitts, et al., Nucl. Mater. Energy, 20 (2019) 100696
$^2$R. C. Wolf, Plasma Phys. Control. Fusion, 45 (2003) R1
$^3$N. Oyama and the JT-60 Team, Nucl. Fusion, 49 (2009) 104007
$^4$M. Kotschenreuther, et al., “Regimes of weak ITG/TEM modes for transport barriers without velocity shear”, UP10.00020, 61st APS-DPP, Oct 21-25, 2019, Fort Lauderdale, USSpeaker: Siye Ding (ORAU/ASIPP) -
The use of high temperature superconductors (HTS) in magnetic systems of fusion devices enables magnetic fields over 16 T, unachievable with low temperature superconductors (LTS), and promises significant reduction in cryogenic and energy budget [1-4]. HTS materials are considered by some researcher groups as the enabling material to make magnetic confinement systems more compact and more affordable. Until recently, some unresolved obstacles for using HTS materials in fusion applications existed, including little knowledge available on critical current in field and mechanical properties at cryogenic temperature, unknown behaviour in alternating magnetic field, limited availability, short piece length, lack of proven cable technology, and high price. Recent technology advances resolved many of these tasks bringing HTS materials to greater maturity for use in fusion devices.
Recently, S-Innovations company successfully developed and marketed a new product, particularly suitable for creating high magnetic fields for fusion devices: SuperOx second generation (2G) HTS wire with modified ReBa2Cu3O7 (ReBCO, Re—rare earth element) composition [5]. The new SuperOx wire demonstrates exceptionally high critical current and engineering current density in high magnetic field; for instance, engineering current density over 1000 A/mm2 at 20 K, 20 T and over 2000 A/mm2 at 4.2 K, 20 T has been repeatedly achieved in commercially produced wire (Fig. 1). The in-field performance of the new SuperOx wire in the 4.2-65 K temperature range is 1.5-2.5 times better than that of the previous product based on the GdBCO superconductor (Fig. 2).
This impressive result has been achieved by S-Innovations in hundreds of 300-600 m long pieces of routinely manufactured wire. The unique feature of the new wire is that the ReBCO layer in it does not contain c-axis correlated nano-columns, in contrast to the common opinion that only with nano-columnar defects is it possible to achieve enhanced in-filed performance in 2G HTS wire.
The mechanical properties of this HTS wire largely depend on the mechanical properties of the strong Hastelloy C276 substrate. In particular, the wire exhibits tensile strength of over 600 MPa and is fully stable at least up to 0.4% elongation. The mechanical properties examined at various temperatures are available for SuperOx 2G HTS wire and compare favourably well to that of other manufacturers.
An important issue of the technology is an understanding of losses appearing in superconductor in alternating magnetic fields. A large set of data was recently collected for AC loss behaviour of SuperOx 2G HTS wire in liquid nitrogen at wide range of current and frequency up to 800 Hz. The results confirm earlier observations that AC losses of 2G HTS wires are low as compared to the most other practical superconductors. The available experimental data can be scaled and AC losses at the operational conditions of large fusion magnets can be calculated.
Studies of SuperOx and other groups demonstrate that viable high current cables for fusion can be manufactured from these wires (see, e.g. [6]). More research is underway to perfect engineering approaches in this direction.
The new SuperOx wire has been routinely manufactured with reproducible quality and high yields at unprecedented volumes for 2G HTS wire industry. The manufacturing processes are highly automated and integrated into an intelligent production management system. A comprehensive quality management system comprises numerous in-line and off-line QC procedures. The production capacity was doubled during the past year, and 5-10-fold capacity increase scenarios are in place awaiting order commitments. The pending 5-10-fold expansion will result in a production volume of multiple tonnes per year, the level at which 2G HTS wire price will allow the construction of economical fusion power plants for future green energy [3,4].References
1. D. G. Whyte, J. Minervini, B. LaBombard, E. Marmar, L. Bromberg, M. Greenwald, “Smaller & Sooner: Exploiting High Magnetic Fields from New Superconductors for a More Attractive Fusion Energy Development Path”, J Fusion Energ (2016) 35:41–53 DOI 10.1007/s10894-015-0050-1
2. P. Bruzzone, W. H. Fietz, J. V. Minervini, M. Novikov, N. Yanagi, Y. Zhai and J. Zheng, “High temperature superconductors for fusion magnets”, 2018 Nucl. Fusion 58 103001
3. B. N. Sorbom, J. Ball, T. R. Palmer, F.J. Mangiarotti, J. M. Sierchio, P. Bonoli, C. Kasten, D. A. Sutherland, H. S. Barnard, C. B. Haakonsen, J. Goh, C. Sung, D. G. Whyte, “ARC: A compact, high-field, fusion nuclear science facility and demonstration power plant with demountable magnets”, Fusion Engineering and Design 100 (2015) 378–405
4. A. Sykes, A. E. Costley, C. G. Windsor, O. Asunta, G. Brittles, P. Buxton V. Chuyanov, J. W. Connor, M. P. Gryaznevich, B. Huang, J. Hugill, A. Kukushkin, D. Kingham, A. V. Langtry, S. McNamara, J. G. Morgan, P. Noonan, J. S. H. Ross, V. Shevchenko, R. Slade and G. Smith, “Compact fusion energy based on the spherical tokamak”, 2018 Nucl. Fusion 58 016039
5. A. Molodyk, S. Samolenkov, A. Markelov, P. Degtyarenko, S. Lee, V. Petrykin, M. Gaifullin, A. Mankevich, A. Vavilov, B. Sorbom, J. Cheng, S. Garberg, Z. Hartwig, S. Gavrilkin, S. Awaji, D. Abraimov, A. Francis, D. Larbalestier, C. Senatore, M. Bonura, O. Pantoja, S. Wimbush, N. Strickland, A. Vasiliev, High temperature superconducting wire based on ReBa2Cu3O7 with Re2O3 nanoparticles exhibits engineering current density over 1000 A/mm2 at 20 K, 20 T and over 2000 A/mm2 at 4.2 K, 20 T, to be published.
6. D.Uglietti, N.Bykovsky, K.Sedlak, B.Stepanov, R.Wesche, P.Bruzzone, “Test of 60 kA coated conductor cable prototypes for fusion magnets” 2015 Supercond. Sci. Technol. 28 124005Speaker: Dr Alexander Molodyk (S-Innovations, SuperOx) -
An accurate calculation of radial neoclassical transport is important for both tokamaks and stellarators. In tokamaks, deviations of the magnetic field from axisymmetry (caused, for example, by ripple due to the finite number of coils or by resonant magnetic perturbations) can result in significant neoclassical damping of the toroidal rotation [1]. In stellarators, their intrinsically three-dimensional configurations lead to specific neoclassical transport regimes (see e.g. [2, 3]) that produce radial energy transport comparable, and often larger, than its turbulent counterpart [4]. Although typically less demanding than gyrokinetic codes, the computational cost of neoclassical simulations is crucial for a thorough characterization of transport in three-dimensional configurations, especially at low plasma collisionalities.
In this work we present KNOSOS [5] (KiNetic Orbit-averaging SOlver for Stellarators), a freely-available [6] open-source code that provides a fast computation of low collisionality neoclassical transport in three-dimensional magnetic confinement devices by rigorously solving the radially local bounce-averaged drift kinetic equation coupled to the quasineutrality equation. Apart from its remarkable speed, KNOSOS includes physics often neglected in neoclassical codes, such as the effect of the component of the magnetic drift that is tangent to magnetic surfaces and the component of the electrostatic potential that varies on the magnetic surface, $\varphi_1$.
In the first part of this contribution, by characterizing plasmas of several devices, we show that, where applicable, KNOSOS reproduces the results of standard neoclassical codes, as illustrated in figure 1, being orders of magnitude faster. The examples provided include the calculation of $\varphi_1$, which is compared to Doppler reflectometry measurements in the stellarator TJ-II [7]. This quantity, $\varphi_1$, can have a strong impact on the radial transport of highly-charged impurities in three-dimensional magnetic configurations [8]. Only recently did stellarator neoclassical codes start to calculate $\varphi_1$, and at a large computational cost. The fast calculation of the bounce averaged main ion distribution with KNOSOS opens the door to a fast evaluation of neoclassical impurity transport using recently-derived analytical expressions [9].
In the second part of the contribution we illustrate how, by retaining the effect of the component of the magnetic drift that is tangent to magnetic surfaces, KNOSOS can describe the superbana-plateau transport regime of stellarators and non-axisymmetric tokamaks. An example is provided in figure 2. We also explain that KNOSOS keeps the dependence of the tangential magnetic drift on the magnetic shear, a relevant element for the calculation of the neoclassical toroidal viscosity in tokamaks with broken axisymmetry at low collisionalities [10]. We end by outlining several planned applications of KNOSOS for stellarators and tokamaks, including detailed validation activities in Wendelstein 7-X, LHD and ASDEX Upgrade, among others.
References:
[1] K C Shaing et al., Plasma Phys. Control. Fusion 54, 124033 (2012).
[2] C.D. Beidler et al., Nucl. Fusion 51, 076001 (2011).
[3] I. Calvo et al., Plasma Phys. Control. Fusion 59, 055014 (2017).
[4] A. Dinklage et al., Nucl. Fusion 53, 063022 (2013).
[5] J. L. Velasco et al., submitted to J. Comp. Phys., arXiv:1908.11615 [physics.plasm-ph]
[6] https://github.com/joseluisvelasco/KNOSOS
[7] T. Estrada et al., Nucl. Fusion 59, 076021 (2019).
[8] J.M. Regaña et al., Plasma Phys. Control. Fusion 60, 104002 (2018).
[9] I. Calvo et al., Nucl. Fusion 60, 016035 (2020).
[10] S. Satake et al., Plasma Phys. Control. Fusion 53, 054018 (2011).Speaker: Dr Jose Luis Velasco Garasa (CIEMAT) -
The challenge
A power plant based on a tokamak architecture has a magnetic cage with thick shielding, so maintenance needs to be conducted through long narrow access ports (see figure 1). This means unprecedented dexterous handling of massive flexible components, which is difficult, slow and expensive. For this reason, industrial plant to date has always been arranged to allow the heaviest lifting to be done from above by a crane.
Power plants also need to achieve higher availability than has been achieved with existing experimental machines. This drives the need for simple and robust maintenance schemes on maintenance-compatible plant architecture. This makes maintenance design driving, which is why it must be understood early in the plant design.
Examples of design driving maintenance considerations are the segmentation of the in-vessel components and the layout of the pipework, components and transport corridor in the ports. These aspects are critical to maintenance, but the choice also impacts the design of the components and indeed the layout of the whole machine and ex-vessel systems.
The size and mass of the in-vessel components for DEMO are significantly higher than for existing machines, including ITER, and the radiation and temperature will be even more challenging. This means DEMO will require novel maintenance solutions and technologies with a different power plant architecture to existing machines. The ITER approach of assembling in-vessel maintenance equipment and handling a high number of components is not appropriate for DEMO and would not be able to meet the plant availability requirements. The same is true for the JET approach of performing all maintenance operations using booms deployed through equatorial ports.
However, there is a large technology overlap with ITER and with wider industry and this will be used wherever possible because of the cost savings, the reduction in technical risk, the data and demonstrated reliability and it is important to standardise the remote maintenance systems as far as possible to reduce the time and cost of design and testing. This applies to the handling systems and movers, but also to the tools and all the interfaces, including handling, mechanical attachment, electrical, signal and fluid (pipes).
Where there is no technology overlap, novel solutions and enabling technologies must be developed to meet the powerplant requirements. There is significant technical risk associated with enabling technologies that are essential to the chosen architecture because if the requirements cannot be achieved then changes to the architecture would be required. The technical readiness level for these enabling technologies must be increased in step with the developing design, to maintain an acceptable risk level. Examples of these enabling technologies for DEMO include the in-bore cutting and welding that is required for the closely grouped pipes on the plasma facing components (breeding blankets and divertor cassettes), and the dexterous handling systems for the massive breeding blankets and divertor cassettes.
In-bore cutting and welding
The EU has been developing miniature laser cutting and welding tools (see figures 2 and 3) that can be deployed down the bore of pipes to demonstrate the feasibility of the enabling technology [1., 2.]. Laser has the advantages of speed, and, as a non-contact process, it has better recovery and rescue options than existing technologies, such as TIG arc welding. Initial trials have been successfully conducted on proof-of-principle tools. However, further development is required to increase the laser power and ultimately to produce a qualified weld process.
Trials have also been conducted on pipe alignment (see figure 4) and tool deployment (see figure 5) which have shown that the alignment forces are high and would be very challenging for the pipe handling systems. Further work is required to develop an alignment system that is less demanding of the handling system.
Dexterous handling of massive components
The layout of DEMO requires a mover that can fit in the port and dexterously handle the in-vessel components that can weigh up to 80 tonnes. A design has been produced for a blanket handling system (see figure 6) but it has limited reserve factors on the working loads, cannot sustain the seismic load case and the stiffness and dynamic performance of the mover would make it difficult to control, leading to very slow operation and lack of accuracy.
New work is starting on the maintenance systems for different tokamak configurations that have split blankets and a double-null. The design process requires time intensive iterations between design tools such as structural analysis, kinematic evaluation and the dynamic and control models. Work is underway to allow these tools to communicate directly with each other and ultimately to integrate them into a single tool. This virtual engineering will allow much more rapid development and optimisation to support the evolving plant design.
Port layout
The layout of the port is also critical to the plant design. It is difficult to arrange all the components and services in a way that they can be maintained whilst retaining the transport corridor for the large plasma facing components that must be transferred out of the vessel. Work on DEMO has shown that modified industrial handling equipment can be used for maintenance. For example, robots, slideways and scissor lifts. This means the technology risk is low, but the integration risk remains high and the layout of the equipment is critical to success. A small change in the pipe layout can make it unmaintainable. A model of the port layout has been created to enable the visualisation required to create the maintenance strategy and design the equipment, and this has been animated to show the sequence of operations and communicate the constraints to the plant designers (see figure 7).
Looking ahead
The technical risks inherent with the novel maintenance solutions required for a fusion power plant have been well recognised across the world. In the EU, a seven-year programme has been proposed to analyse the maintenance implications of the EU DEMO plant design to enable plant option down-selection and to develop the enabling technologies that are not available from industry or ITER.
The focus of the programme is to develop the design of remote maintenance equipment integrated with the plant design, to reduce or remove key risks. Within this programme, designs are required to allow feasible maintenance of the breeder blankets and other in-vessel components. As the system designs become more detailed the programme will produce proof-principle equipment to validate designs and virtual engineering tools to reduce the key technical risks.This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 un-der grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
[1.] “Service Joining Strategy for the EU DEMO”, T. Tremethick et al, ISFNT-14 (2019).
[2.] “Laser cutting and welding tools for use in-bore on EU-DEMO service pipes”, K. Keogh et al, Fusion Engineering and Design, volume 136, part A, November 2018, pages 461-466.
Speaker: Mr Oliver CROFTS (UKAEA) -
According to the European strategy to fusion energy, the development and the operation of a demonstration power plant (DEMO) is foreseen as the single step between ITER and a commercial tokamak fusion power plant (FPP). DEMO is required to feature all key systems and components of an FPP and to comply with a set of general goals [Donné, 2018]. These goals include a few hundred megawatts electric power generation, a closed fuel cycle and long pulse (or steady state) plasma operation.
The conceptual design of DEMO begins with the quantitative definition of these main goals and shall proceed by selecting the major reactor parameters. The approach adopted in the EU-Programme [Federici, 2018] follows different steps, iteratively repeated until a certain grade of satisfaction, consistency and attractiveness is met (see Figure 1, black solid line path). Hence, the beginning of the actual engineering design hinges upon the verification of these conditions.
Dedicated computational tools - referred to as systems codes - are deployed to produce a reactor baseline, upon definition of reactor requirements, constraints and architectural features. The systems code output is transferred to the design codes, normally in form of a 2D reactor sketch and major reactor parameters (e.g. radial build, fusion power, major radius and magnetic field). Presently available fusion systems codes, such as PROCESS [Kovari, 2014], aim at exploring one (or more) reactor configurations that simultaneously fulfill the plasma physics operational limits, the engineering constraints and the plant general goals. In general, they rely on rather basic physics and engineering models (mostly at zero or one-dimensional level). The design codes, instead, are very detailed but run on much longer computing times and model limitations.
Due to the broad multi-physics and multi-scale spectrum involved within the design process, it is rather challenging to maintain data consistency and to keep an expedite design flow. In turns, wide modelling gaps between systems codes (0D/1D) and design codes (3D) might slow down and hinder the design loop, thereby increasing the number of iterations. To this end, the connections between system and design codes can be consolidated by complementing the systems codes by means of a more refined and intermediate system analysis tool (see Figure 1, red-dashed line path): this step is referred to as MIRA, Modular Integrated Reactor Analysis.
MIRA [Franza, 2019] is a high spatial resolution design tool developed at KIT, incorporating the physics and the engineering insights of the utmost domains of tokamak reactors. MIRA relies on a modular structure and provides a FPP baseline. With reference to the flowchart of Figure 2, it incorporates into a unique computing environment a mathematical algorithm for the following problems:
- 2D radial/vertical geometry and material characterization of physical components, i.e. plasma, breeding blanket (BB), vacuum vessel (VV) and toroidal field (TF) and poloidal field (PF) coils;
- 2D free-boundary plasma magnetic equilibrium;
- constrained gradient-based PF coil solver, incorporating plasma shaping requirements and coils engineering constraints for all plasma operational phases, i.e. breakdown, start of flat-top (SOF) end of flat-top (EOF);
- plasma and divertor physics modules based on one-dimensional ion, electron and impurity transport [Fable, 2018];
- quasi 2D neutron and photon core radiation transport for the evaluation of breeding (TBR), heating and shielding capabilities of all in-vessel reactor components;
- 3D magneto-static modelling of TF and PF coils elements and engineering characterization of superconducting cables;
- Plant integral modelling and power performances.
Compared to presently available system codes, MIRA is based on a higher mathematical sophistication, engaging systems analyses up to three-dimensional space resolution. This allows scoping multiple reactor configurations with a more consolidated modelling granularity at a component level and with a holistic view of the entire plant, thereby leaving room for design modifications at higher degree.
The MIRA approach has been applied to the DEMO baselines 2015 [Franza, 2019] and 2017, generated by means of the PROCESS code. The analyses have been carried out by taking an identical set of input assumptions and requirements (e.g. same fusion power, major radius and aspect ratio) and observing the response on the key design targets and the imposed operational limits. Based on more accurate MIRA analysis, both baselines have been found in conflict with some of these limits, in particular on plasma burn time (33 % below the two-hour goal) and maximum TF ripple on the plasma boundary (13 % above its upper limit). The major causes for these discrepancies between PROCESS and MIRA are attributed to the reduced space resolution, to the modelling simplifications and to limited engineering capabilities of PROCESS. In PROCESS, the burn time is based on simple 0D magnetic flux conservation and peak magnetic field in the central solenoid (CS), whilst the TF ripple is calculated from predefined scaling laws. In MIRA, instead, the burn time derives from 2D calculation of poloidal flux profiles $\Psi(r,z)$ at SOF and EOF magnetic equilibrium configurations (see Figure 3) with plasma shaping requirements ($\partial \mathcal{D}_p^t$ target shape) and coil technological limits. The TF ripple is computed via 3D magneto-static analysis.
The outcomes of the DEMO 2015 and 2017 baselines analyses have illustrated that modelling simplifications, affecting the state-of-the-art systems codes, affect considerably the overall design of the reactor. Therefore, the application of the MIRA approach to analyse a DEMO baseline can mitigate the lack of modelling instruments before engaging the design codes analyses (Figure 1). Accordingly, a set of active measures has been addressed to steer some identified reactor geometric variables in favour of a design point that fulfils the imposed constraining conditions. Such measures involved a parameter scan on the inboard BB radial width and outboard TF coil radial extension. Apart from fulfilling the lower and upper bounds on burn time and TF ripple, the addressed parametric studies have shown also non-trivial inter-parametric dependencies, never explored in fusion system analyses. For instance, a reduced thickness for inboard BB and TF coil leg has been identified in connection with plasma burn time, TBR requirements and coils technological limits (both in CS and TF coils superconducting cables).
In conclusion, this work poses new basis to designing a tokamak reactor and to parametrizing multiple technological solutions. Accordingly, a deeper and a more centralized multi-physics reactor analysis can speed up and improve the whole design process.
References
[Donne, 2018] T. Donne, W. Morris, X. Litaudon, C. Hidago, D. McDonald, H. Zohm et al. European Research Roadmap to the Realisation of Fusion Energy. EUROfusion, November 2018.ISBN: 978-3-00-061152-0.
[Fable, 2018] E. Fable, C. Angioni, M. Siccinio, H. Zohm. Plasma physics for fusion reactor system codes: Framework and model code. Fusion Engineering and Design 130 (2018) 131–136.
[Federici, 2018] G. Federici, C. Bachmann, L. Barucca, W. Biel et al. DEMO design activity in Europe: Progress and updates. Fusion Engineering and Design, vol. 136, pp. 729 – 741, 2018
[Franza, 2019] F. Franza. Development and Validation of a Computational Tool for Fusion Reactors System Analysis". PhD thesis, Karlsruhe Institute of Technology (KIT), June 2019. DOI: 10.5445/IR/1000095873, https://publikationen.bibliothek.kit.edu/1000095873
[Kovari, 2014] M. Kovari, R. Kemp, H. Lux, P. Knight et al. PROCESS: A systems code for fusion power plants - Part 1: Physics. Fusion Engineering and Design, 89:3054–3069, 2014.
Speaker: Dr Fabrizio Franza -
The 2018 National Academies of Sciences (NAS) Report of the Committee on a Strategic Plan for U.S. Burning Plasma Research and the more recent APS-DPP Community Planning Process (CPP) recommend that the U.S. should pursue innovative science and technology to enable construction of a Fusion Pilot Plant (FPP) that produces net electricity from fusion at reduced capital cost. Such a mission requires development and integration of multiple physics and technology innovations. A unique feature of the tokamak-based FPP approach in the U.S. is the integration of a high fraction of self-driven current with high core plasma pressure and high divertor parallel heat flux. The integration of a high-performance core and edge has not been previously accessed nor is it presently planned to be accessed by existing U.S. or international facilities. This integration is sufficiently challenging that construction and operation of a dedicated sustained-high-power-density (SHPD) tokamak facility is proposed by the U.S. community to close this integration gap. This presentation describes the performance of present and planned tokamak facilities, gaps between present/planned facility capabilities and the U.S. FPP regime, and presents pre-conceptual engineering studies for a SHPD facility to narrow/close the core-edge integration gaps to the FPP.
Figure 1 compares (left) pressure and (right) parallel heat flux between (blue) present/planned tokamaks and (red) proposed FPP/FNSF/SHPD devices (2,3,4). Figure 1 shows the FPP regime has plasma pressure = 0.2-0.7 MPa in the range of high-performance pulsed tokamaks (e.g. ITER baseline, EU Demo, Italian DTT, ADX, SPARC) but with approximately 3 times higher self-driven fraction enabling long pulse to steady-state operation. Present stellarators are shown in the left graph at f$_{self}$ = 1 and have q$_{||0} < $1 GW/m$^{2}$ so are not included in the right-hand graph. The FPP-regime pressure and heat flux are also 3 times higher than expected in present/planned facilities at high f$_{self}$ =0.6-0.8. SHPD aims to close these gaps through innovations such as: high field/current density/temperature superconducting magnets, advanced divertors, elevated confinement, and enhanced-efficiency current drive actuators.Systems studies have been carried out for a range of major radius R = 1 to 2m and aspect ratio A = 1.6 to 4 to identify SHPD facility sizes and powers that can access FPP-level plasma pressure and edge heat fluxes. These studies include aspect-ratio-dependent elongation and beta limits (4), assume toroidal field winding pack current density J$_{WP}$ = 70MA/m$^2$, and FPP-level elevated H$_{98}$ ≥ 1.5 sufficient to operate at the nominal n=1 no-wall kink stability limit (4).
The table at the top of Figure 2 lists the assumed variation of heating power versus R. The powers are chosen to have H$_{98}$=2 at A=2.4 as A is varied. Figure 2 shows (a) the magnetic field increases from 1.5-3.5T at A=1.6 to 7-9T at A = 4, (b) the Lawson parameter is maximized for A=1.8-2.4, (c) FPP-level pressures are accessed for R = 1.2m between A = 2-3, and (d) FPP-level heat fluxes are most readily accessed for R ≥ 1.4-1.6m. However, by increasing heating power to P$_{heat}$ = 50MW at R = 1.2-1.4m it also possible to access low-A FPP and A=4 FNSF-level (2) heat fluxes at both A = 2.0 and A=2.4 as indicated in the right-hand side of Figure 1. Accessing ARC-level (3) heat fluxes at A=3 requires more power and/or a larger major radius SHPD facility.
High-Z solid tungsten divertors and first-walls have the highest technical readiness level and are the near-term option for fusion development. However, high-Z solids have several challenges, including plasma-facing component damage from erosion and re-deposition and neutrons, and high-Z impurity accumulation and associated core plasma radiative collapse. Low-Z (Li) and high-Z (Sn) liquid metals (LM) are increasingly being studied as a means of addressing these challenges. Figure 3 shows a R=1.2m, A=2.4 double-null SHPD concept with a small central solenoid for IP ramp-up and sufficient in-vessel space and flexibility to study a range of divertor effects including: radiation, detachment, leg length, closure, advanced magnetic geometries, and solid and/or liquid metals all to facilitate the SHPD core-edge integration research mission.(1) R.J. Goldston et al, Plasma Physics and Controlled Fusion 59 055015 (2017)
(2) C.E. Kessel et al, Fusion Science and Technology 68 225-236 (2017)
(3) B.N. Sorbom et al, Fusion Engineering and Design 100 378 (2015)
(4) J.E. Menard et al, Nuclear Fusion 56 106023 (2016)*Research supported by the US DOE Contract No. DE-AC02-09CH11466.
Speaker: Jonathan Menard (Princeton Plasma Physics Laboratory) -
As the EUROfusion EU-DEMO design programme approaches the transition between the pre-conceptual and conceptual design phase the systems code PROCESS has been improved to incorporate more detailed plasma physics, engineering and integration models. Unlike many systems codes PROCESS combines the physics modelling with both technology and costs analysis. Key to the conceptual design phase are detachment, toroidal field magnet design, double-null power sharing, operational sensitivity and economic uncertainty analysis. All of these have been integrated into PROCESS [1.], [2]. During the pre-conceptual design systems codes are an essential tool for exploring fusion power plant concepts. They allow one to model the interaction of the plant systems and quickly perform reactor optioneering. To be able to carry out these large scoping studies the fidelity of the models can be restricted to reduce the computational time. The EUROfusion EU-DEMO baseline designs are created using the systems code PROCESS and the ability to measure these trade-offs has led to important design choices being examined during the DEMO pre-conceptual design phase. Ruling out unfeasible designs allows EUROfusion to efficiently identify where in the design space to carry out detailed design work. This contribution describes how PROCESS has been retooled for EU-DEMO conceptual design.
To allow optimisation of the underlying plant systems for a more fixed plant design greater detail had to be integrated into PROCESS. For EU-DEMO a 1-D scrape-off layer (SOL) model has be implemented to capture the power loss mechanisms in the SOL, to validate the core plasma power balance and to determine if the plasma is in a detached state – a key design requirement for EU-DEMO [3]. In combination with the SOL model PROCESS can now allow power sharing in a double-null configuration. The choice of single versus double-null is a fundamental choice for EU-DEMO so capturing the behaviour is essential for a systems code. A 1-D plasma transport solver has been integrated into the code to produce a self-consistent plasma model with plasma profiles for correctly calculating heating and current drive power deposition and determining the plasma radiation by integrating over the profile [4].
As the design space of EU-DEMO becomes smaller there is a need to understand what the sensitivity of the design is given some uncertainty on the performance and engineering parameters. PROCESS is an ideal tool for this analysis due to its breadth of scope and computational speed. It has given insight into the likelihood of a given EU-DEMO design achieving the high-level goals of EUROfusion, such as reaching the net electric power target [5] (the same analysis has been used on CFETR [6]). The PROCESS uncertainty tools have been used to analyse the cost sensitivity of DEMO designs to determine the primary cost drivers. This information will contribute to the decisions during concept down-selection.
One of the primary drivers of machine design, performance and cost are the superconducting magnets. Therefore, correctly calculating the space required, the achievable field and cost is essential for PROCESS. High temperature superconductors (HTS) can potentially offer a performance, engineering and cost benefits. A REBCO (rare earth barium copper oxide [7]) HTS model has been written for PROCESS for the TF coils. The operating temperature of the TF coil for both LTS and HTS is 4.5 K for the analysis presented here as it is often preferable to go to higher field to achieve large net electric power, as the fusion power is proportional to $\beta^2 B^4$. This is done rather than increase the operating temperature to save on electrical power needed for the cryogenic system or use an alternative to helium as a coolant.
PROCESS has been used to analyse the impact of toroidal field coil stress on machine design with LTS [8] and can now compare with HTS. Figure 1 shows the effect of the allowable Tresca stress in the TF coil steel. The LTS model includes a quench calculation with a variable copper fraction, while the HTS model imposes a maximum superconductor current per unit area of copper, chosen as 100 A/mm2 or 200 A/mm2. PROCESS was set to minimise the major radius and to produce 500 MW net electric power for 2 hours. Figure 1 shows that one can achieve higher fields at smaller machine size with HTS. The reduction in major radius depends on the copper requirement and is in the range 0.25-0.5m. At higher allowable stress the HTS PROCESS runs start to prioritise smaller machine size over further increasing the field.
Further detail will be added to reduce uncertainty in the models and allow more robust design scoping studies, such as an equilibrium solver. All systems codes will need to be comprehensively rebuilt to make them relevant for the conceptual design phase. The UKAEA power plant technology group is revising PROCESS to make improvements and collaboration easier.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053 and from the RCUK Energy Programme [grant number EP/T012250/1]. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
[1.] M. Kovari, R. Kemp, H. Lux, P. Knight, J. Morris, and D. J. Ward, “" PROCESS " : A systems code for fusion power plants-Part 1: Physics,” Fusion Eng. Des., vol. 89, no. 12, pp. 3054–3069, Dec. 2014.
[2] M. Kovari et al., “‘PROCESS’: A systems code for fusion power plants - Part 2: Engineering,” Fusion Eng. Des., vol. 104, pp. 9–20, Mar. 2016.
[3] J. Morris, M. Kovari, N. Asakura, and Y. Homma, “Comparison of the systems code PROCESS with the SONIC divertor code,” IEEE Trans. PLASMA Sci., 2020.
[4] E. Fable, C. Angioni, M. Siccinio, and H. Zohm, “Plasma physics for fusion reactor system codes: Framework and model code,” Fusion Eng. Des., vol. 130, pp. 131–136, May 2018.
[5] H. Lux et al., “Uncertainties in power plant design point evaluations,” Fusion Eng. Des., vol. 123, 2017
[6] J. Morris, V. Chan, J. Chen, S. Mao, and M. Y. Ye, “Validation and sensitivity of CFETR design using EU systems codes,” Fusion Eng. Des., vol. 146, 2019.
[7] R. Heller, P. V. Gade, W. H. Fietz, T. Vogel, and K. P. Weiss, “Conceptual Design Improvement of a Toroidal Field Coil for EU DEMO Using High-Temperature Superconductors,” IEEE Trans. Appl. Supercond., vol. 26, no. 4, pp. 1–5, Jun. 2016.
[8] J. Morris, R. Kemp, M. Kovari, J. Last, and P. Knight, “Implications of toroidal field coil stress limits on power plant design using PROCESS,” Fusion Eng. Des., vol. 98–99, 2015.Speaker: James Morris -
This paper presents solutions for critical problems in Japan’s DEMO (JA DEMO), which include common DEMO design issues beyond ITER-relevant technologies. The highlights of this design study are (i) system design for electric power generation, (ii) remote maintenance concept to attain a high plant availability, and (iii) novel concept for water-cooled pebble bed blanket and tritium recovery. The proposed concept as JA DEMO will be the foundation for Japan’s DEMO that can be envisioned in the next stage of ITER.
1. Introduction
The pre-conceptual design of JA DEMO has been proposed by adding the technology outlook of ITER and JT-60SA. JA DEMO has a relatively large major radius Rp = 8.5 m for long-pulse operation and a relatively low fusion power Pfus = 1.5 GW for divertor heat removal [A]. The cooling water for DEMO is operated at the PWR (Pressurized Water Reactor) water conditions of 15.5 MPa and 290 ºC-325 ºC. In this paper, we state the following three issues of JA DEMO design; (i) in order to evaluate the electric power output, it is necessary to clarify the design of the cooling system and the heat balance of the primary heat transfer system (PHTS), and the power consumptions of the pumping power in the PHTS, the freezing power for the superconducting coil, the injected heating power for plasma operation, etc., (ii) in order to develop a remote maintenance scheme (RMS) for the replacement of large components in the high dose area, the movement line of remote handling (RH) device must be arranged to maintain the soundness of the device based on the shutdown dose map of JA DEMO, and (iii) in order to develop a blanket concept with the capabilities of high tritium (T) productivity and pressure tightness, it is necessary to obtain a high packing factor up to 80 % with binary packing and to recover the T generated in the packing area without retention. Approaches to resolve these design issues and remained challenges are presented in the following sections.2. Evaluation of electric power output
The electric generation systems for JA DEMO is developed without intermediate heat exchanger. This is because T permeation through the steam generator to the secondary cooling system in the PHTS is found to be less than the restricted amount of T disposal for pressurized water reactors in Japan [B]. Here, the generator output was evaluated to be 640 MW with the thermal efficiency 34.4 %. The power consumption was analyzed to evaluate the electric power output. To evaluate the pumping power in the PHTS, the intricate cooling system was designed in consideration of suitable configurations of the systems for the PHTS such as steam generator, pressurizer, etc. as shown in Fig. 1. The pumping power of the cooling system in the PHTS was evaluated to be 71 MW. Since the design concept of the superconducting coil is basically same with to that of ITER, it is determined by scaling the freezing power in ITER as 91 MW. In the NBI system of the heating and current drive (H&CD), the neutralization efficiencies of the photon and gas neutralizers are assumed to be 90 % (theoretical value) and 56 % (ITER technology), respectively. Here, the neutralizer method makes a difference in the power consumption of 118 MW. Hence, the total power consumption was found to be 386 MW with the photon neutralizer, and the electric output was evaluated to be 254 MWe as summarized in a table of Fig. 1.
3. Remote maintenance scheme in the high dose environment
The RMS is as follows for the JA DEMO: after cutting of a cooling pipes in the maintenance ports, the blanket segments and divertor cassettes integrated with shielding plugs (SP) are replaced through vertical upper ports and bottom ports, respectively, as shown in Fig. 2. The major tasks in RMS for JA DEMO have been how to replace large-scale components and how to implement the maintenance work in the high radiation environment. In the previous studies, we had focused mainly on the construction method for stably replacing large-scale components which weighs 100 tons and stands 10 meters high as shown in Fig. 2. We found that if plant availability was expected to reach about 70% which is the requirement of JA DEMO, the blanket segments must be replaced in parallel from 4 out of 16 ports. We also found out the characterizations of RMS in the high radiation environment [C]. By arranging SPs, the dose rate of the points for cutting and re-welding cooling pipes in the maintenance ports can be reduced to about ~ 0.1 Gy/h as shown in Fig. 2(a). However, when the outboard blanket segments and the divertor cassettes with the SP were removed together, the dose rate in these maintenance ports increases to 100 Gy/h at the maximum as shown in Fig. 2(b). Here, the dose rate in the vacuum vessel of ITER during maintenance is evaluated to be 250Gy/h according to the requirements for remote handling of the shielding blanket. Therefore, maintenance techniques developed for ITER such as pipe cutting or re-welding are applicable to the JA DEMO in terms of the spatial dose rate.
4. Blanket concept and tritium recovery
The conceptual design of the breeding blanket of honeycomb structure, which has pressure tightness against in-box loss-of-coolant accidents in a water-cooled solid breeder, has been developed as shown in Fig. 3(a). As honeycomb structure is higher in pressure tightness than square prism structure, the area for filling the mixed pebbles breeder of Li2TiO3 pebbles and Be12Ti ones can be enlarged. From the results of the 3D neutronics analysis, the target of the overall TBR (>1.05) would be achievable. To achieve the TBR target, it is necessary to fill the packing factor to 80% by binary packing. However, there is a concern that the amount of T retained in the breeding area may increase due to pressure drop. To avoid T retention in the breeding area, the flow of He-purge gas is arranged by using three inflow points of He-purge gas near the first wall. Then the retention of T was resolved as shown in Fig. 3(b).
References
[A] Y. Sakamoto et al., 27th IAEA Int. Conf. on Fusion Energy (2018) FIP/3-2
[B] R. Hiwatari et al., Fusion Eng. Des. 143 (2019) 259-266
[C] Y. Someya et al., Fusion Eng. Des. 124 (2017) 615-618Speaker: Youji Someya (National Institutes for Quantum and Radiological Science and Technology) -
Spherical Tokamak reactor (STR) is attractive due to its inherent capabilities such as disruption avoidance, natural elongation, natural divertor and high beta capability, apart from a smaller size, with presumably lower costs [ 1, 2]. There has been an extraordinary evolution from the early concepts like SMARTOR [ 3] with devices like START, NSTX, MAST, GLOBUS-M and a number of others with the HTS based future devices like STEP [4]. Given the pace of development of the new superconducting materials [5,6] and the new divertor concepts [7,8,9], the STRs represent a rapidly developing front and may very well be realized not far in the future. Following an elegant paper by Peng et al. in 1986, a range of compact reactor designs ($R$ and $P_f$) has emerged, e.g. FNS-ST (0.5m, 10 MW), DTST (1.1m, 30-60 MW), ARC (3.3m, 525 MW), SlimCS (5.5m, 2950 MW), ARIES-ST (3m, 2980 MW) with a variety of objectives like, neutron source, component-test-facility (CTF) and power plant [10,11,12,13,14]. However, while the high neutron loads are welcome for reactor economics, the size reduction comes at a penalty of extreme heat loads on the divertor with concomitant engineering challenges [15]. Several designs of STRs are currently being developed around the world with scoping studies and available data from currently operating tokamaks as well as other experimental/dedicated test facilities and insights from experts [16]. This paper brings out the role of constraints arising from steady-state power balance and core-radiation. It is argued that the core-radiation plays a crucial role in the reactor design, as it not only restricts the accessible parameter-space but also determines the limits on impurity accumulation [17]. A comprehensive physics-design study [18] shows that about 50$\%$ of the heating power needs to be lost by core-radiation. Such considerations can impact stability as well [19]. In the following, the ST-parameter space ($R-B_t$) is analyzed to elucidate the limits posed by the various constraints. For $T_i$ from 6 to 20 keV, the fusion power (MW) may be approximated for analytic purposes as:
$$P_{F}=0.026 \frac{(S_n+S_T+1)^2}{(2S_n+2S_T+1)} \frac{\kappa {\beta_{N}}^2 {S_{\kappa}}^2}{q^2 A^4} R^3 {B_t}^4$$ where $q=5 R B_t S_k/(A^2 I_p)$ is the safety factor, $I_p$ is the plasma current in MA, $A$ is the aspect ratio and $S_k$ is the shape factor. $\beta_N = \beta a B_t/I_p$ and $S_n$, $S_T$ are the exponents for the parabolic profile of the density and temperature respectively. The stored energy in MJ can be expressed as: $$W_\beta = \frac{\pi}{8}\frac{\kappa S_\kappa}{q A^3} \beta_N R^3 {B_t}^2$$ In steady-state, where the power from $\alpha$-particles and the externally injected power are balanced by the transport losses, the power-balance is given by $W_\beta = P_L \tau_E $, where $P_L$ (defined as $P_H (1-f)$) is the power reaching the edge, after a fraction $f$ of the power deposited $$P_H = P_\alpha + P_{ext}= P_F (1/5+1/Q)$$ is radiatively lost from the core region. It is assumed that the ITER-IPB(98,y2) scaling holds good, although it is likely to be more favorable in reality [20]: $$\tau_E = 0.0562 H_h {I_p}^{0.93} {B_t}^{0.15} {n_{19}}^{0.41} R^{1.97} \kappa^{0.78} \epsilon^{0.58} M^{0.19} P_L^{-0.69}$$ The power-balance can then be written as: $$Q_{LF} = (f_\alpha /5 + 1/Q )(1-f)$$ where $f_\alpha$ is the fraction of $\alpha$-particles which transfer their energy to the plasma. The $Q_{LF}$ is actually the ratio $P_L/P_F$ and is an involved expression with fractional powers of plasma parameters. To understand its dependencies, it is best approximated as: $$\frac{\beta_N \ A^{14/5}\ q^{6/5}}{{B_t}^{92/35}\ H_h^3 \ f_G^{6/5}\ S_k^{16/5} \ M^{3/5}\kappa^{2/5} \ R^{9/5}}$$ where, the nearest integer ratios are used to approximate the exponents in the expression for $\tau_E$. The radiated power fraction $f$ can be expressed in terms of $Q_{LF}$. Its role in accessibility constraints in the $R$-$B_t$ space has been shown in Fig.1, where, the contours of constant $P_f$ are shown along with the limits on achievable $B_t$ assuming either copper or HTS peak current-density in the center-stack. The constant fusion contours intersect increasingly high divertor load curves as one makes the reactor more compact. The dotted curves ($f$=0, 0.5 and 0.94) correspond to the power balance constraint. The $f=0$ curve shows the limit of 'no core-radiation' and thus represents the lower boundary of physically acceptable solutions. Thus, for a given set of parameters as an example ($q$=3, $\kappa=2.5$, $\delta=0.3$, $\beta_N=5$, $Q=5$), there exists an upper limit on the value of $R$ (3m). The two $Q_{LF}$ curves that 'bracket’ the fusion power curve, define the accessible space until the limit on achievable $B_t$ is encountered. An example of a design point (R=1.25 m, Bt=2.8 T, Pf = 200 MW) has been shown (red dot). It may not be possible to meet it unless almost 60$\%$ of the heating-power is radiated from the core. Such constraints make it necessary to examine how much core concentration of impurities would be acceptable. Fig.2 shows impact of $Q$ in the parameter space -- higher values reduce the available space in the lower left-hand corner. This has implications for the reactors which may operate at modest values of $Q$ (CTF or fusion-fission hybrid, fissile material converters or radioactive waste processing, or just fusion-science devices). At the same time, the higher $Q$ demand from power reactors (to remain cost-competitive and investment-attractive), eliminates a large space and pushes accessibility points further up. An important consequence of the power balance constraint is that the divertor heat load (transported power) $P_{div} \approx B_t^{3/2}/R^{4/5}$. The gradients of $P_{div} \approx$ constant are in dramatic contrast to those of constant neutron load contours, so while the neutron load per unit area varies slowly as one moves towards the top left-hand corner, the divertor load builds up rapidly. Three case studies will be presented ($R$=1.75, 1.25 and 2.25m for $P_f$=100, 200 and 900 MW respectively) in detail. Fig.3 shows how the power balance constrains the $\kappa- \beta$ space for the case $R$=1.25m, $P_F$ = 200 MW. It can be seen that higher $\beta$ cases will need a higher $\kappa$. The sensitivity to different $\tau_E$ scaling, as well as impurity transport, the effects of neutron and particle loads on the center-stack, first-wall and divertor will be presented in detail.
References:
1. Y-K.M Peng et al., Nucl. Fusion 26 769 (1986)
2. Gi et al., Nucl. Fusion 55 063036 (2015)
3. D. Jassaby et al., Plasma. Phys. Control. Fusion 3, 151, (1977)
4. L. A. El-Guebaly et al., Fusion Sci. and Tech., 74:4, 340-369 (2018)
5. A. Sykes et al., Nucl. Fusion 58 016039 (2018) and references therein
6. Dennis Whyte, Phil. Trans. R. Soc. A 377 20180354 (2018)
7. M. Kotschenreuther et al. Nucl. Fusion 50 35003 (2010)
8. R. J. Goldston et al., Phys. Scr. T167 014017 (2016)
9. V. A. Soukhanovskii, IEEE Trans. Plasma Sci. 44 3445 (2016)
10. B. V Kuteev et al., Nucl. Fusion 51 073013 (2011)
11. Y-K.M. Peng et al., Nucl. Fusion 40 583 (2000)
12. B.N. Sorbom et al., Fusion Eng. and Design 100, 378–405 (2015)
13. Tobita K. et al., Nucl. Fusion 49 075029 (2009)
14. Najmabadi F. et al., Fusion Eng. Des. 65 143 (2003)
15. E. Surrey, Phil. Trans. R. Soc. A 377: 20170442 (2019)
16. M. Tillack et al., Nucl. Fusion 53 027003 (2013)
17. H. Takenaga et al., Nucl. Fusion 45 1618 (2005)
18. S. C. Jardin et al., Fus. Eng. Des. 65 165 (2003)
19. P. Kaw et al., Phys. Rev. Lett. 65 2873 (1990)
20. P F Buxton et al.,Plasma Phys. Control. Fusion 61 035006 (2019)Speaker: Dr Shishir Deshpande (Institute for Plasma research) -
In this paper, we present a novel approach to simulate plasma turbulence in tokamak geometry. Even with most advanced supercomputers in these days, it is still very challenging to simulate turbulence and transport globally in the realistic tokamak geometry. One of challenges is the time scale disparity between ion and electron due to the large mass ratio between them.
We propose a new kinetic-fluid hybrid model to simulate ion scale electromagnetic turbulence (i.e. fluctuations with $k_\perp \rho_i \le 1$) in efficient ways retaining essential kinetic electron physics. All ionic species are modeled by gyrokinetic equations. The evolution of electron density perturbation $\delta n_e$ is described by the fluid continuity equation. Perpendicular and parallel pressure are divided into the contributions from passing and trapped electrons as $\delta P_{e\perp,\parallel} = \delta P_{e\perp,\parallel}^P + \delta P_{e\perp,\parallel}^T$. For passing electrons, we assume isotropic pressure perturbation only from the density perturbation:
$\delta P_{e\perp}^P = \delta P_{e\parallel}^P = T_{e0} \delta n_e^P = T_{e0} (\delta n_e - \delta n_e^T)$.
The pressures $\delta P_{e\perp,\parallel}^T$ and the density perturbation $\delta n_e^T$ of trapped electrons are evaluated by solving the bounce-averaged kinetic equation:
$\frac{\partial f_e^T}{\partial t} + \frac{d\beta}{dt}\frac{\partial f_e^T}{\partial \beta} + \frac{d\alpha}{dt}\frac{\partial f_e^T}{\partial \alpha} = C(f_e^T) + S^{P,T}.$
Here, $\beta$ and $\alpha$ denote the radial location and toroidal angle of the bounce center of trapped electron at outer mid-plane. $C(f_e^T)$ and $S^{P,T}$ represent operators for the Coulomb collision and the source-sink due to trap-passing population changes. The details of the equations for the bounce motion and the operators can be found in Ref.$[1]$. To describe the evolution of the vector potential, the parallel Ohm’s law for passing electrons is combined with the Faraday’s law.
Though the fluid model alleviates the requirement of time step for the simulation of parallel electron dynamics, the model contains the physics in Alfven time scale, and is still burdensome to simulate with explicit methods. To resolve this, we introduce an Implicit-Explicit (IMEX) Runge-Kutta scheme$^{[2]}$, which separates simulation variables into two groups – those with fast temporal variations and the remaining – and all linear terms and time derivatives of the fluid equations involving the former variables are implicitly discretized. We demonstrate that this IMEX discretization can significantly accelerate the simulations and also resolve the cancellation issue caused by directly evaluating the time derivative of the vector potential $\partial A_{\parallel} / \partial t$. This new hybrid model is implemented in the global gyrokinetic code gKPSP $^{[1]}$.
We demonstrate that the new model allows very efficient simulations of key ion scale instabilities such as ITG, TEM, and KBM. Figure 1 shows linear simulations of ITG and KBM. With a fixed toroidal mode number $n_\varphi=19$, the beta value is changed by increasing the electron density. For low beta values, ITGs are excited, but as the beta increases larger than 1.5%, the mode is changed to KBM. For comparison, global linear GENE simulation results are also plotted$^{[3]}$. We can see that the new model can accurately reproduce the critical $\beta_e$ for the ITG $\rightarrow$ KBM transition, while the growth rates and frequencies of KBM from gKPSP are slightly smaller than those from GENE.
Next, we perform a scan of toroidal mode numbers with a fixed beta value $\beta_e = 1.0\%$. For low toroidal mode numbers, ITGs are excited, but as the mode number increases, ITG<->TEM transition occurs as the result in Figure 2 shows. Again, GENE simulation results with the same conditions are plotted, and we can confirm that the results from the new model agree well with the GENE results.
As a final benchmark test, we perform global linear simulations of TAE with energetic ions. We choose the simulation conditions in Ref.$[4]$ for the TAE simulations. Figure 3 shows the growth rates of TAEs driven by varying fast ion energies. The growth rates agree well with the results in Ref.$[4]$. In the figure, the electrostatic and electromagnetic potential are also presented.
From the analysis on computing costs of the new hybrid model, it is found that the overall computing costs for electromagnetic simulation are roughly 5 times of those for electrostatic ITG simulation with adiabatic electrons.
Currently, we are performing global nonlinear simulations using the new hybrid model. We will present that, with the new model, gKPSP can efficiently simulate various global physics of electromagnetic turbulence in realistic tokamak plasma conditions.References:
$[1]$ J.M. Kwon et al, Comput. Phys. Commun. 177, 775 (2017).
$[2]$ L. Pareschi et al, J. Sci. Comput. 25, 129 (2005).
$[3]$ T. Görler et al, Phys. Plasmas. 23, 072503 (2016).
$[4]$ A. Konies et al, 24th IAEA FEC ITR/P1-34 (2012).Speaker: Dr Janghoon Seo (National Fusion Research Institute) -
As the ITER staged approach plan is finalized, Korean fusion community has recently constructed a roadmap toward the Korean demonstration reactor (K-DEMO). Eight core technologies essential for the K-DEMO design are identified: 1) Core plasma technology, 2) System design and safety, licensing technology 3) Fusion materials technology, 4) Wall components technology, 5) Superconducting magnet technology, 6) Heating device technology, 7) BOP technology and 8) Fuel cycle technology. Detailed R&D plans for the core technologies and K-DEMO design activities are being discussed in the community.
One important project to involve the 8 core technologies in the K-DEMO roadmap is the development of integrated numerical simulations of the K-DEMO reactor, so called, V-DEMO (Virtual DEMO) project. The virtual reactor system is proposed to be a “digital twin” of K-DEMO reactor, and hence will aid the design activities of K-DEMO and safety analysis and licensing procedures to begin with. In the reactor operation stage, the V-DEMO system will help code validation, operator training and advanced reactor designs.
The critical role of V-DEMO project is to provide optimized K-DEMO designs upon given input conditions. If a K-DEMO is designed to be built now, options are limited by the current technologies. On the other hand, a K-DEMO is built by market forces 20 years in the future, that is, after the demonstration of burning plasmas by ITER, a broad spectrum of options can be considered.
The main goal of the V-DEMO project can provide the necessary design options when the decision is made to build a K-DEMO in the future. The digital twin of K-DEMO reactor will be based on integrated knowledge of core (burning plasma) and boundary plasma physics, plasma-material interaction physics, blanket and power plant technologies of BOP (Balance of Plant). Particularly, recent developments of information technologies such as supercomputer-aided simulations, artificial intelligence (AI), big data (BD) and machine learning (ML) are expected to advance current tokamak simulations and power plant simulations to assimilate K-DEMO reactor at an unprecedented level.
In this study, plans and strategies to develop fusion simulation technologies for V-DEMO are reviewed. First, the scope and concept of V-DEMO system are defined. The goal is to bring the V-DEMO system to represent a fusion power plant on virtual space aided by supercomputing and virtual reality (VR) technologies. It is important to define the applications of the V-DEMO project and required fidelity and complexity to meet this goal. The numerical modules and units, that is, components of V-DEMO system should be identified. Secondly, the key technologies to realize the V-DEMO system should be determined. It will involve the survey and analysis of the current status of simulation technologies in the fusion community. Critical simulations and virtualization technologies should be determined.
Finally, the strategy and plans for the development of V-DEMO should be established. A staged development plan is prepared and the strategy for collaboration between power plant engineering community and fusion community is studied. A strategy to accelerate V-DEMO development with artificial intelligence and digital twin technologies should be established by engaging software engineering community. The critical part of the project is the validation and verification by comparisons of existing experimental data obtained by KSTAR and ITER. The superconducting tokamak KSTAR will provide high performance plasma data of DEMO relevant density and DEMO relevant scenarios. ITER will provide most critical burning plasma data to benchmark the core plasma simulations as well as the engineering data of tokamak systems including TBM (test blanket module) data. Additionally, neutron sources are critical to the development of V-DEMO system as it provides the neutron related data for materials, structures, tritium breeding and many other fusion reactor systems.
Speaker: Hyun Kyung Chung -
Good vacuum health of any fusion machine is a stringent requirement for a quality plasma discharge. A pumping system of a fusion machine has to remove the gas load during two scenarios efficiently: the first one involves preparing the machine by baking of vessel and plasma-facing components to ~150 degree C to 200 degree C to achieve base vacuum level of < 1 E-8 mbar and the other is during plasma operation. The gas load during the baking process mainly comprises water vapor, nitrogen, Co, and during plasma operation, it is Helium and hydrogen isotopes.
To address the two requirements, a liquid nitrogen-based and liquid helium based cryopump development was undertaken at the Institute for Plasma Research, India. Both the pumps are capture pumps and use coconut shell charcoal as sorbent. The novel aspect of both kinds of pumps is the use of highly microporous activated coconut shell charcoal with pore size < 2 nm [1., 2.]. Optimized thickness of the adhesive coating was used. An epoxy-based adhesive, which is thermally conducting, vacuum compatible with low outgassing rate, and able to withstand thermal cycling, was developed for the purpose [3.,4.]. The charcoal bonded with the metal plate can withstand 150 degree C, and this helps in quality regeneration removing the gases clogged in pores. Generally, commercial cryopumps are baked up to 50 degree C. The developed pumps are portable liquid cryogen based pump having no moving parts, and hence requires less maintenance. Fig 1(a) shows the adsorption and desorption isotherms of the activated charcoal granule and desorption part is reversible in nature. The type of isotherm relates to microspores nature of coconut shell charcoal. Fig 1(b) shows the pores surface area of ~ (1400 ±5 %) m2/g at LN2 temperature (using BET method).
Liquid Nitrogen-based pump can provide a solution to evacuate large gas loads in a fusion machine during its preparation targeting achieving base vacuum. During the process of pumping, when the pressure level is in the range of < 1E-5 mbar, water desorption from the surfaces under vacuum takes place along with other gases like Nitrogen, and it hinders achieving the UHV level. Commercial cryopumps operating at ~20 K are generally used during the process. The developed concept of a bath kind of cryopump with cryopanels coated with coconut shell charcoal can pump Nitrogen at liquid nitrogen temperature apart from water vapor. It provides an easy and economical solution to handle immense gas load released during baking of components in a fusion machine. The pump comprises a liquid nitrogen bath with attached conical cryopanels of copper, forming a pine tree kind of structure and are conduction-cooled. The surface area of the panels is 0.7 m2. Fig.2 shows the pumping speed found on an experimental set up as per American vacuum society ( AVS) standard is > 4500 l/s at a throughput of 1 E-2 mbar l/s for continuous pumping mode of > 50 hours.
The results are discussed in detail in this paper. The results of the prototype led to the design of cryoadsorption cryopump, which can find application in a fusion machine to handle gas load during baking process. The design parameters target the pumping speed of > 14000 l/s for Nitrogen and 20,000 l/s for water vapour. It is a 500 mm opening pump. Fig.3 shows the design of this pump and it is discussed in this paper.
Liquid helium-based cryopump used the same sorbent. A 300 mm opening pump with a cryopanels of 0.1 m2 was used to evaluate the pumping speed. A relative study of adsorption isotherm for hydrogen and helium was carried out at < 10 K. It was observed that the relative pore surface area at < 10 K for gases like hydrogen and Helium was higher than the standard pore surface area at 80 K. The pump delivered a pumping speed of > 2 l/s/cm2 of pumping speed for Hydrogen and Helium (Fig-4).
Simulation of the experimental set up was carried out using Molflow analysis (Ref-5). Results of simulation and experiment are discussed in the paper. The concept with a large number of cryopanels and thus large surface areas to pump hydrogen and helium provides a solution to handle the considerable gas load during plasma operation.
From simulation studies, the inference is that pumping speed for Hydrogen and Helium gases is in the range of 2500 to 3000 l/s, and the corresponding sticking coefficients are in the range of 0.1 to 0.2. Fig 5 shows the results obtained using MOLFLOW simulation and Fig 6 shows the variation of pumping speed over the sticking coefficient.
REFERENCES
1.Satish M Manocha, "Porous carbons", Sadhana”, Vol. 28, Parts 1 & 2, February/April 2003, pp. 335–348
2.Bansal R C, “ Active carbon”, (New York: Dekker) 1988
3.W. J. Parker, R. J. Jenkins, C. P. Butler, and G. L. Abbott, “Flash Method of Determining Thermal Diffusivity, Heat Capacity, and Thermal Conductivity”, Journal of Applied Physics Volume 32, Issue 9
4.R.Gangradey et.al, “OGMS: A Facility to Measure Out-gassing Rate of Materials,” Procedia Materials Science Volume 6, 2014, Pages 272-277.
5.R.Gangrdey et.al. “Activated Charcoal cooled to Liquid Helium temperature Exhibiting Pumping For Hydrogen and Helium gases”, Vacuum Journal, DOI: 10.1016/j.vacuum.2019.109026Speaker: Ranjana Gangradey (Institute for Plasma Research) -
Extraction of heat from the breeding blanket in tokamaks requires a network of heat carrying pipes from the blanket modules (BM) to the steam-generator (SG). Regardless of the type of the blanket concept, the operational requirements will mandate the RH compatibility and remote maintenance. The pipe-network will need to connect all BM in a given sector by sector-manifold, which in turn will itself integrate with a common ring header after it exits from reactor vessel. While a number of concepts for DEMO and blanket modules are available, there is not sufficient information on how is the pipe routing planned for the case when the entire vessel is covered by BM. In the case of ITER, as the tokamak is progressing in construction, there are good reference layouts for the piping inside the vessel for the shield blankets. But, DEMO being an electricity producing reactor, it is necessary to examine various options for heat extraction while observing thermal constraints arising from interfaces with vessel, ports, cryostat, building, etc. Given the expertise and investment made for ITER, it would not be surprising to imagine that the next step tokamak will closely match the major parameters of ITER, hence it is of interest to examine the compatibility of present piping and manifolds of ITER (for water cooled shield BM) with that of a possible helium-cooled BM reactor.
In this work, we carry out an analysis of the hydraulic parameters assuming three different fusion reactor configurations. Configuration A is of 500MW with about 1.25 MW power to be extracted from each BM and in this case the layout within the vessel is assumed similar to ITER. Configuration B and C are similar to DEMO and a compact power plant like ARIES-ST. It is assumed that all three configuration used helium as a coolant for the BM. A sketch of the overall scheme for all the three configurations has been shown in Fig. 1. The supply ring header (SRH) connects the SG to the 18 sectors (shown as circles) from where the extracted heat is taken to the SG by the return-ring-header (RRH). Each sector will itself have a manifold to connect all the BM within that sector by a sub-header. With regards to a real electricity producing power plant like DEMO e.g. the design proposed in 1, the requirement of power extraction per module may be ~4 times that of ITER. The complexity of piping for DCLL blanket has been partially elaborated by Federici et al.2. A greater challenge will be encountered while analyzing the pipe-routing for future compact reactors proposed, e.g. ARIES-ST3 where a DCLL concept has been elaborated by Tillack et al.[4]. In general, the space limitations for the sector-manifold and pipe-entry and exit through the ports (of vessel, cryostat, bio-shield etc.) are already tight, so it becomes even more challenging for designing as the available space is less but the power-handling capacity needs to be almost an order of magnitude higher than configuration A.
Table 1 shows the parameters for the 500 MW scenario (Configuration A), with 440 BM, like it has been considered for ITER [5][6]. The inlet/outlet pressure and temperature of helium is taken to be 80/79.9 bar (to get adequate heat-capacity) and 300C/500C [7], and the diameters of the different headers have been evaluated in order to keep the velocity ~ 100 m/s (well below the 10% of the sonic speed). The pressure-drop seems to be nominal for the required flow parameters. The length of the pipes from BM to RRH and from RRH to SG are considered as ~40m and ~20m respectively. Following Ref. [8], we consider the diameter to be of ~ 48 mm for all the BM outlet pipes going up to RRH. The pressure drop in the above routing is found to be approximately ~ 2.25 bar. The power loss by radiation from all the hot pipes (500C) to the vacuum vessel (~200 C) seems negligible (1.23 MW). The detailed results of thermal and stress analysis of the piping network will be presented.
The hydraulic parameters for the three configurations have been given in Table 2. Detailed results will be presented for all three cases.
In summary, the attempt to analyze piping from BM to SG is in itself an important step because the present designs (whether of ITER TBM[9] or DEMO TBM [10]) have considered only equatorial dedicated ports. The blanket piping design for considering entire vessel area covered by breeding blankets will present even bigger challenges of compact machines like DEMO-FNS [11].
Keywords: Fusion power plant design, Piping layout, Blanket module
References
1 D. Maisonnier et al., “DEMO and fusion power plant conceptual studies in Europe,” Fusion Eng. Des., vol. 81, no. 8–14 PART B, pp. 1123–1130, 2006.
2 G. Federici et al., “Overview of the design approach and prioritization of R&D activities towards an EU DEMO,” Fusion Eng. Des., vol. 109–111, pp. 1464–1474, 2015.
3 F. Najmabadi, “Spherical torus concept as power plants—the ARIES-ST study,” Fusion Eng. Des., vol. 65, pp. 143–164, 2003.
[4] A. Team, M. S. Tillack, X. R. Wang, J. Pulsifer, and S. Malang, “ARIES-ST breeding blanket design and analysis,” vol. 50, pp. 689–695, 2000.
[5] A. R. Raffray et al., “The ITER blanket system design challenge,” Nucl. Fusion, vol. 54, no. 3, 2014.
[6] M. Merola, F. Escourbiac, R. Raffray, P. Chappuis, T. Hirai, and A. Martin, “Overview and status of ITER internal components,” in Fusion Engineering and Design, 2014, vol. 89, no. 7–8, pp. 890–895.
[7] G. Aiello, J. Aubert, N. Jonquères, A. L. Puma, A. Morin, and G. Rampal, “Development of the Helium Cooled Lithium Lead blanket for DEMO,” Fusion Eng. Des., vol. 89, no. 7–8, pp. 1444–1450, 2014.
[8] A. Furmanek, P. Lorenzetto, and C. Damiani, “Modified blanket cooling manifold system for ITER,” Fusion Eng. Des., vol. 84, no. 2–6, pp. 793–797, Jun. 2009.
[9] G. Federici, L. Boccaccini, F. Cismondi, M. Gasparotto, Y. Poitevin, and I. Ricapito, “An overview of the EU breeding blanket design strategy as an integral part of the DEMO design effort,” Fusion Eng. Des., vol. 141, pp. 30–42, Apr. 2019.
[10] G. Aiello et al., “HCLL TBM design status and development,” Fusion Eng. Des., vol. 86, no. 9–11, pp. 2129–2134, 2011.
[11] B. V Kuteev, Y. S. Shpanskiy, and D. Team, “Status of DEMO-FNS development,” Nucl. Fusion, vol. 57, no. 76039, 2017.Speaker: Piyush Prajapati (Institute for Plasma Research) -
The most effective numerical treatment of turbulence within a transport model of a tokamak plasma is as a multi-scale, multi-physics problem. Multi-scale since the typical time- and space-scales associated with turbulence are usually much smaller than the time- and space-scales of the plasma as a whole, which is the domain of a transport code; multi-physics since a range of possible physics needs to be included in the problem, but these can usually be treated separately.
The work described here reduces the full problem to the coupling of three separate codes: a transport code that evolves the macroscopic one dimensional fields based on geometric information provided by an equilibrium code and transport coefficients derived from running multiple instances of a flux-tube turbulence code; the changing profiles from the transport code are inputs to the equilibrium and turbulence code forming a time loop.
The MUSCLE2[REF1] coupling framework is used to implement this workflow in a tightly coupled, modular and extensible manner, while the unit of exchange between the codes is the appropriate Consistent Physical Object (CPO)[REF2]. The workflow is similar to that described in Falchetto et al.[REF3] and some information about the workflow development can be found in Hoenen et al.[REF4], Luk et al.[REF5] and Luk et al.[REF6].The EasyVVUQ library developed as part of the VECMA project[REF7] is being used to incorporate the calculation of uncertainty intervals into the workflow. As identified by the VECMA project, different approaches to implementing UQ are possible. The easiest to implement is to treat the application as a black-box and then to apply one of the standard UQ techniques to the inputs and outputs of the black-box. In this case the black-box is the entire workflow run to (quasi) steady-state. This has been implemented with a simpler proxy for the turbulence code in order to make development of the techniques faster.
Work has now started on opening up the workflow so that UQ is applied at the code level within the workflow, and rather than just passing profile between codes, distributions of profiles are passed.
The initial application has been to an ASDEX Upgrade Standard H-mode shot with the density profile taken from the experiment and the electron and ion temperature profiles predicted. The figure on the left shows the measured ion temperature from four AUG standard H-mode shots (with varying heating schemes) and the results of predictive simulations based on a simple model (labelled GEM0) and the results from the workflow using the GEM code at 8 flux-tube positions.The future challenge will be to bring the pieces of UQ together with the stochastic nature of the turbulence code to examine the feasibility of producing profile predictions incorporating a turbulence code together with uncertainty intervals. If this proves to be feasible with the used gyro-fluid turbulence code, the extension to gyro-kinetic turbulence codes will be an obvious next step.
References
[REF1] J. Borgdorff, M. Mamonski, B. Bosak, K. Kurowski, M. Ben Belgacem, B. Chopard, D. Groen, P.V. Coveney, A.G. Hoekstra, “Distributed multiscale computing with MUSCLE 2, the Multiscale Coupling Library and Environment”, Journal of Computational Science, 2014, Volume 5, Issue 5, pp.719-731. doi:10.1016/j.jocs.2014.04.004.[REF2] F. Imbeaux, J.B. Lister, G.T.A. Huysmans, W. Zwingmann, M. Airaj, L. Appel, V. Basiuk, D. Coster, L.-G. Eriksson, B. Guillerminet, D. Kalupin, C. Konz, G. Manduchi, M. Ottaviani, G. Pereverzev, Y. Peysson, O. Sauter, J. Signoret, P. Strand, ITM-TF work programme contributors. “A generic data structure for integrated modelling of tokamak physics and subsystems”, Computer Physics Communications, Elsevier, 2010, 181, pp.987 - 998. doi:10.1016/j.cpc.2010.02.001
[REF3] Falchetto, Gloria L., David Coster, Rui Coelho, B. D. Scott, Lorenzo Figini, Denis Kalupin, Eric Nardon et al. "The European Integrated Tokamak Modelling (ITM) effort: achievements and first physics results." Nuclear Fusion 54, no. 4 (2014): 043018.
[REF4] Hoenen, Olivier, Luis Fazendeiro, Bruce D. Scott, Joris Borgdoff, Alfons G. Hoekstra, Pär Strand, and David P. Coster. "Designing and running turbulence transport simulations using a distributed multiscale computing approach." In 40th EPS Conference on Plasma Physics, EPS 2013; Espoo; Finland; 1 July 2013 through 5 July 2013, vol. 2, pp. 1094-1097. 2013.
[REF5] O. O. Luk, Olivier Hoenen, Alberto Bottino, Bruce D. Scott, and D. P. Coster. "ComPat framework for multiscale simulations applied to fusion plasmas." Computer Physics Communications 239 (2019): 126-133.
[REF6] O. O. Luk, O. Hoenen, O. Perks, K. Brabazon, T. Piontek, P. Kopta, B. Bosak, A. Bottino, B. D. Scott, and D. P. Coster. "Application of the extreme scaling computing pattern on multiscale fusion plasma modelling." Philosophical Transactions of the Royal Society A 377, no. 2142 (2019): 20180152.
[REF7] https://www.vecma-toolkit.eu/
Speaker: Dr David Coster (Max Planck Institute for Plasma Physics) -
The tokamak progress in 20th century accompanied by the replacement of their plasma-facing components (PFC) from high Z materials to low (Ref. 1). This made it possible to obtain a relatively pure (Zeff <1.5) fusion plasma. However, superconducting tokamaks of the 21th century discovered a new obstacle for fusion as limiting of the duration of the discharges (PH/S - limit 1) due to the accumulation of wall erosion products inside the discharge chamber, which finishes by disruption. As solution for steady-state tokamaks would be the removal of erosion products from the vacuum chamber during discharges or between them while maintaining PFC with small Z. It is desirable that they have a low melting point for removing them as liquid. The first candidate is Li. The purpose of T-11M research is a creation of lithium PFC protection in the form of a thin lithium film covering the wall and lithium plasma filling SOL for reducing the plasma-wall electric field and suppressing unipolar arcs (Ref. 1). For this, a lithium emitter (limiter) based on capillary porous structure (CPS) (Ref. 2), injecting lithium during discharge into plasma and the Li-collectors system in SOL between the T-11M wall and the plasma edge were used. The function of the collector is to prevent lithium withdrawal to the wall, and in the future, to remove the trapped lithium outside without violating the vacuum conditions. The Li collection is based on the fact that the ionized lithium flows along the edge magnetic surfaces. A collector crossing them becomes a limiter for this flow. Until the collector temperature exceeds 300° C, it is able to accumulate incoming lithium (Ref. 1). However, its excessive immersion into the plasma periphery with plasma current can perturb the boundary magnetic surfaces (“magnetic islands”), which will activate the undesirable transport of lithium to the wall and into the plasma. We analyze hear the various combinations of the three lithium limiters at T-11M tokamak, which were used as lithium emitters and collectors.
The experiments were conducted at T-11M tokamak: R/a=0.7 /0.27 m, BT = 1-1.5 T, Ip = 70-90 kA, Δt ~ 250 ms, <Ne> from 0.5 to 8 1019 m3. Figure 1 shows the scheme of T-11M with system of its movable lithium limiters: a vertical lithium CPS emitter with a operated surface coated by W-felt with lithium 2 (Figure 1, 3), and two similar longitudinal limiters (Fig. 1 1,2) using as lithium CPS-based collectors installed at a small angle to the toroidal magnetic field in the SOL region created by the vertical emitter.
To record the temperature of the longitudinal Li-collectors during the discharge, two high-speed infrared (IR) cameras (Fig.1, IR-1,2) VarioCam HD were used, which made it possible to simultaneously observe the temporal surface temperature variation of both limiters during the experiment. In visible light, two cameras (Figure 1, VC - 3 and 4) recorded their behavior. A Mach probe (figure 1, 4) was used to control the plasma parameters in SOL.
In the course of development of the T-11M limiter system, they were used as separate emitter-collector combinations. Figure 2 shows a scheme of the respective combinations.
The vertical lithium emitter (figure 1, 3) was used in two combinations II and IV as the main lithium emitter. In other cases, it was removed into the port.
The power of the heat flux P coming to the surface of the limiter during the discharge was calculated according to ΔТ(t) taking into account "grayness".
As mentioned above, the main function of the collector system is to prevent excessive lithium flux to the wall. Comparison of different schemes of lithium emitters and collectors (I-IV figure 2), showed that the best scheme with cleanest and hottest plasma was IV. Mach probe and LiI measurements revealed that scheme IV is characterized by the steepest decay λ of the lithium distribution in the SOL shadow from the hot plasma boundary to the chamber wall, which demonstrates a decrease in the radial transport of lithium in this region. Li decay λ decreased from λ=3.1cm (scheme I), λ=3cm (scheme II) and λ=3.65cm (scheme III) up to λ=1.75cm (scheme IV figure 2 ). One of the explanations of success of scheme IV could be the assumption that the asymmetry of the collectors (emitters) provides the formation of magnetic islands that destroy magnetic surfaces in limiters regions. The IR measurements confirmed the validity of this assumption. For this, an experiment was conducted with the vertical displacement of the plasma column on two symmetric lithium limiters 1 and 2 (case III ). A small (~ 1.5cm) initial vertical shift made it possible to divide them into a “lead” (1) and a “slave” (2). Under the conditions of cylindrical symmetry small shift (1cm) of the plasma column to the “lead” limiter should not qualitatively change the ratio between the heat fluxes arriving at such slightly shifted symmetric limiters (figure 2 scheme III ). It was true while the shift was small. When it reached 2cm, (figure 3 the ratio between the IR signals changed dramatically - the heat flux to the “slave” 2 times exceeded the flux to the “leading”. Figure 3 sequentially shows the dynamics of changes in heat fluxes on 1 and 2 limiters as the vertical shift increases (Δ=+1cm up to Δ -1cm) . Such transition should have been expected when a magnetic island with isolated O point and an open (symmetrically of 180 ° along the torus) X-point concentrating heat flux into this zone formed at the plasma periphery.
The conclusion that can be made on the basis of the T-11M experiments is that the local lithium emitters and collectors of the lithium circuit should be placed strictly symmetrically (with accuracy higher 1cm) radially and along the torus, thereby suppressing the possibility of the formation of magnetic islands and increased lithium escape to the tokamak walls during the creation of lithium PFC protection.References
1.Mirnov 2019 S.V. Nucl. Fusion 59 015001
2. Evtikhin V.A. et al. 1996 16 IAEA Fusion Energy Conf. (Montreal Canada) IAEA Vienna 3 659Figures:
Speaker: Prof. Sergey Mirnov (TRINITI) -
We have analysed cross-scale interactions between trapped-electron-mode (TEM) and electron-temperature-gradient (ETG) turbulence by means of gyrokinetic simulations. We find that (i) TEM turbulence suppresses ETG turbulence and dominates electron heat transport, and (ii) the suppression of ETG by TEM can ubiquitously happen even when TEM-driven zonal flows are subdominant. Additionally, from the comparison with hydrogen and deuterium plasmas, TEM-stabilized deuterium plasma is affected by ETG, resulting in simultaneous enhancement of TEM and ETG fluctuations. This brings a new perspective on the isotope effects of anomalous electron heat transport via cross-scale interactions between TEM and ETG turbulence.
Cross-scale interactions between electron- and ion-scale fluctuations are one of unresolved issues in turbulent transport. Our previous studies carried out direct numerical simulations of ion-temperature-gradient-mode (ITG) or micro-tearing-mode (MTM) and ETG turbulence, and revealed the mechanism of cross-scale interactions: e.g., ITG turbulent eddies suppress ETG streamers, and ETG turbulence damps short-wavelength zonal flows or localized current sheet of MTM (Maeyama 2017). Besides the theoretical studies, a possible impact of cross-scale interactions in ITER is explored via Tokamak experiments such as Alcator C-Mod, DIII-D (Holland 2017), and JET (Mantica 2020). On the other hand, TEM, which is known as one of important ion-scale instabilities, is not well analyzed in the context of the cross-scale interactions with ETG turbulence. A simulation study for a low magnetic shear plasma reported that TEM-driven zonal flows suppressed ETG turbulence (Asahi 2014). However, it is known that TEM turbulence have a variety of saturation mechanisms depending on plasma parameters where zonal flows are not necessarily dominant. Therefore, one of our key questions is whether TEM coexists with or suppresses ETG, if there is no dominant zonal flow.
To deal with this problem, we set up local flux-tube gyrokinetic simulations resolving electron and ion scales simultaneously. Magnetic field is a circular torus geometry with $q=1.4$, $s=0.8$, $r/R=0.18$. Plasma parameters are $R/L_{Ti}=1$, $R/L_{Te}=9.3$, $R/L_n=3$, $T_e/T_i=3$ and $m_i/m_e=1836$. In these parameters, ITG modes are stable since the ion temperature gradient is low. From the linear analysis, ion-scale fluctuations at $k_y\rho_{ti}<1$ are dominated by TEM, while the electron-scale fluctuations are at $k_y\rho_{ti}>1$ are ETG. Most unstable wavenumbers of these instabilities are roughly separated by the ratio of ion and electron gyro radii.
Figure 1 plots the comparison of electron energy flux spectra. At the early phase of nonlinear simulation ($t=9-12R/v_{ti}$), ETG modes rapidly grow up and saturate. Then ETG driven streamers peaks around $k_y\rho_{ti}=3$ and dominate electron energy flux. Simultaneously, long-wavelength TEM slowly grows up. After the growth of TEM ($t=20-40R/v_{ti}$), TEM turbulence around $k_y\rho_{ti}=0.3$ dominates electron energy flux. One finds that the peak of ETG at electron scale disappears in the presence of TEM. The energy flux level is comparable to that in a low-resolution (single-scale TEM turbulence) simulation plotted in Fig. 1.
Figure 2 plots the time evolution of electrostatic potential fluctuation of zonal flows ($k_y\rho_{ti} = 0$), TEM ($0 < k_y\rho_{ti} < 1$) and ETG ($1 < k_y\rho_{ti}$). Linear growth of ETG is observed before $t<7R/v_{ti}$. After the saturation of ETG, long wavelength TEM grows with smaller growth rate than the maximum linear growth rate of TEM (plotted by a slope in Fig. 2). The growth of TEM also saturates around $t=16R/v_{ti}$. At the time, both of zonal flows ($k_y\rho_{ti}=0$) and high-wavenumber fluctuations ($1 < k_y\rho_{ti}$) are generated via inverse and forward cascades. It should be noted that the suppression of ETG appears at the same time of TEM growth ($8 < t v_{ti}/R < 14$), but not at the zonal flow generation ($16 < t v_{ti}/R$). Additionally, zonal flow is not the dominant fluctuation of TEM turbulence in this case (e.g., $\langle |\varphi_{k_y}|^2 \rangle =46$ for $k_y\rho_{ti}=0$ and $\langle |\varphi_{k_y}|^2 \rangle = 251$ for $k_y\rho_{ti}=0.3$). This is in contrast with the previous study (Asahi 2014) employing low magnetic shear $s=0.4$ and $T_e/T_i=1$, where ETG is suppressed by TEM-driven strong zonal flows. Our study reveals that strong zonal flows is not necessarily required for the suppression of ETG by TEM. It resembles the suppression of ETG by ITG (Maeyama 2017), where ITG-driven turbulent eddies distort ETG streamers and suppress them.
We have also carried out a multi-scale TEM/ETG turbulence simulation for the deuterium plasma ($m_i/m_e=3672$), where linear growth rates of TEM are reduced by the collisional isotope effect (Nakata 2017). Simultaneous enhancement of TEM and ETG fluctuations with stronger electron heat transport than that in the ion-scale TEM turbulence simulation is observed, suggesting cross-scale interactions of TEM and ETG turbulence.
Our study expands the understanding of cross-scale interactions between ion- and electron-scale turbulence. It is found that ETG is suppressed by TEM turbulence even when zonal flows are subdominant. Comparison with hydrogen and deuterium plasmas brings a new insight into isotope effects via cross-scale interactions. Experimental measurement of electron-scale fluctuations will be desired future in context of isotope effects.
References
(Maeyama 2017) S. Maeyama et al., Nucl. Fusion 57 (2017) 066036; Phys. Rev. Lett. 119 (2017) 195002.
(Holland 2017) C. Holland et al., Nucl. Fusion 57 (2017) 066043.
(Mantica 2020) P. Mantica et al., Plasma Phys. Control. Fusion 62 (2020) 014021.
(Asahi 2014) Y. Asahi et al., Phys. Plasmas 21 (2014) 052306.
(Nakata 2017) M. Nakata et al., Phys. Rev. Lett. 118 (2017) 165002.Speaker: Shinya Maeyama (Nagoya University) -
Generally, DEMO requires larger toroidal field (TF) coils than ITER, resulting in two major difficulties, the tolerance in TF coil fabrication and installation and an increased inductance. This paper presents the possible solutions these on the basis of the design study on Japan’s DEMO (JA DEMO). It was confirmed that, in the case of adopting a mitigated tolerance by a factor of 2.5-5 compared with that of ITER, the resulting error field of TF coils is correctable to an acceptable level in terms of locked mode avoidance. The validity of a fast discharge scheme of TF coil current at the discharge time constant of less than 30 sec was confirmed, leading to a reasonable terminal voltage of each TF coil and the consistency with the electromagnetic forces acting on the vacuum vessel due to the induced eddy current. The problems resulting from large-scale TF coils commonly used in DEMO can be resolved with these approaches.
1. Introduction
Japan’s DEMO (JA DEMO) requires larger toroidal field (TF) coils than ITER to attain higher fusion output (about 1.5 GW) with larger plasma volume and to allow the installation of breeding blanket inside the TF coil bore. The adoption of such large TF coils results in two major difficulties, (1) the tolerance in TF coil fabrication and installation (2) an increased inductance. Main design parameters of the JA DEMO are a plasma major radius of 8.5 m, fusion output of 1.5-2 GW, the net electricity of 0.2-0.3 GW, and magnetic field on the plasma axis of 6 T [A]. The superconducting coil system of the JA DEMO consists of 16 toroidal field (TF) coils, a central solenoid (CS) and 7 poloidal field (PF) coils, as shown in figure 1. To demonstrate steady-state electric power generation in a power plant scale, a higher magnetic field strength and a 1.5 times larger TF coil bore than those in ITER are necessary [B].
2. TF coil tolerance issue and error field correction
The design concept of TF coil is basically similar to that of ITER, that is usage a radial plate, double pancake and cable-in-conduit conductor. One of the issues specific to the large-sized steady-state tokamak DEMO concept is the increasing technical difficulties due to the larger toroidal field coils than those in ITER. The fabrication tolerance has an effect on engineering difficulties of large coil system. JA DEMO adopts the design strategy of TF coils of mitigating the error field requirement and the error field is corrected by using an EFCC if needed to avoid locked modes. The target error field is set at $\delta$B$_{\rm TMEI}$/B$_{\rm T}$ ~10$^{-4}$ in JA DEMO design. Here, $\delta$B$_{\rm TMEI}$ is three mode error index (TMEI) defined by Fourier components B$_{1/1}$, B$_{2/1}$ and B$_{3/1}$ of poloidal/toroidal mode number m/n = 1/1, 2/1 and 3/1 on q = 2 surface. The $\delta$B$_{\rm TMEI}$ of 0.6 mT is targeted for an error field correction with respect to the TF coil, PF coil and CS coil error fields in DEMO. The manufacturing and assembly tolerances could produce non-axisymmetric magnetic fields, leading to error fields. The relation between the tolerances and the resulting error fields have been calculated with a Monte-Carlo approach. The previous calculation results showed the mitigated target of error field of about $\delta$B$_{\rm TMEI}$/B$_{\rm T}$ ~10$^{-4}$ provides a mitigation in the tolerance in coil fabrication and installation by a factor of 2.5-5 compared with that of ITER, contributing to realize the fabrication of larger coils for DEMO. In order to determine the EFCC currents, we have applied the least square method to minimize three error field components of m = 1, 2 and 3 with n = 1 at the q = 2 surface. In the estimations, anti-series connections between coils located at the opposite side toroidally are assumed. In order to avoid interference with maintenance ports and NBI ports, the EFCCs has two types which are different toroidal angle (67.5$^\circ$ and 45$^\circ$) located on plasma outboard side, where would be an effective position for error field mitigation are newly adopted, as shown in figure 2. As a result, the EFCC current required to correct $\delta$B$_{\rm TMEI}$ up to 0.1 mT was 200 kAT even when a set of EFCCs are arranged outside the vacuum vessel in a toroidally non-periodic manner. The result indicates that the error field caused by mitigated tolerances of TF coil fabrication and installation can be corrected with ex-vessel EFCCs at realistic coil currents.
3. Fast discharge scheme of TF coil current
The large TF coil size of the JA DEMO gives a 3 times greater self-inductance if the conductor current and the magnetic field strength are the same magnitudes as those of ITER and then generates a 3 times terminal voltage that is quite capable of losing reliability of TF coils. Focusing on issues related to increase in the coil self-inductance due to increase in the TF coil size, identification of discharge time constants and fast discharge scenario [C] that match other reactor systems were investigated. The increasing in the discharge time constant reduces the terminal voltage, however the temperature acceptable limit of the conductor must be not exceeded in the coil quench event. In order to evaluate the upper limit of discharge time constant, the time evolution of conductor temperature on the fast discharge event is calculated, as shown in figure 3. It is found that the discharge time constant requires less than 30 seconds, which is consistent with a vacuum vessel design (less than allowable stress of SUS316L, 143 MPa@100$^\circ$C), as shown in figure 4. The TF coils are divided into serially connected segments that are electrically isolated from each other and only the coil segment having a failed coil is rapidly discharged. The circuit current analysis indicates that this discharge scheme enables to reduce the terminal voltage with a factor of ~0.6 or less and which would contribute to ensure reliability of the turn insulations. These results shown that the validity of a fast discharge scheme of TF coil current at the discharge time constant of less than 30 sec was confirmed, leading to a reasonable terminal voltage of each TF coil and the consistency with the electromagnetic forces acting on the vacuum vessel.
References
[A] Y. Sakamoto et al., 27th IAEA Int. Conf. on Fusion Energy (2018) FIP/3-2
[B] K. Tobita et al., 2019 Fusion Sci. Technol. 75 372-383
[C] Y. Itoh et al., Plasma and Fusion Res., 14 (2019) 1405167Speaker: Hiroyasu Utoh (National Institutes for Quantum and Radiological Science and Technology) -
The Russian concept of steady-state operating plasma-facing components (PFC) based on the use of a stagnant or slowly flowing liquid metal (LM) as a plasma-facing material enclosed in a capillary-porous system (CPS) (Ref. 1) integrates all the advantages of LM with the possibility of uniform distribution of its layer on the surface, regardless of its orientation in space, with high resistance to splashing under electromagnetic force. Heat removal is provided by thermal conductivity through the PFC structure to the flowing coolant. Thus, the ability of a CPS-based PFC to withstand high-density stationary heat flux and not pollute the plasma strongly depends on the selection of design, materials and coolant. The first experience in creation of such limiter type PFC with a LM and an active surface temperature stabilization system were successfully implemented for T-11M and FTU tokamaks (Ref. 2).
The development program for liquid metal PFCs, including the creation and model testing of prototypes, is aimed at solving the following key tasks: 1 - providing continuous refilling and even distribution of LM on the CPS surface; 2 - achieve periodical replacement of liquid metal (in particular lithium) in the CPS structure with the aim to decrease of the concentration of dissolved tritium; 3 - study the ability of LM surface cleaning from the deposited films and products of interaction of LM with the residual gases through a providing of LM flow on the surface and in the structure of CPS; 4 - development of design and operation parameters definition for the LM supply systems; 5 - study of the process of removal of energy flux of high density from the PFC surface; 6 - development of design and operation parameters definition for the systems of effective heat exhaust. Solving these problems will allow creating reliable and capable PFC for a stationary tokamak reactor.
The solution of the problems according to program points 1-4 is achieved by studying the created prototypes of the lithium limiter of T-11M (Fig.1) and the mockup of the lithium divertor target of T-15MD (Fig.2) both in the conditions of the vacuum pilot facility and in the conditions of the tokamak. Both prototypes consist of an internal part and an external liquid lithium supply/replacement system (Fig.3). The in-vessel elements of the prototypes, whose plasma-facing surfaces (cylindrical for T-11M and flat for T-15MD) are made of stainless steel and covered with a layer of CPS made of molybdenum mesh, are equipped with an upper collector for liquid lithium supply and distribution and a lower collector for LM lead out. The maintenance of lithium in the liquid state is provided by an electric heater. The mockup of the divertor target allows to change the angle of inclination of the plasma facing surface to the horizon within 30-90o. The refilling, retention and uniform distribution of lithium on the plasma-facing surface are due to capillary forces in the CPS structure. When an excessive amount of LM is fed to the upper collector, it slowly drifts in the CPS structure under the gravity to the lower collector. Cleaning the surface of the CPS from films of lithium reaction products with residual gases and deposited materials realized by the flow of LM on the surface of the CPS with an excess of LM supplied to the upper collector.
Liquid lithium supply and discharge tubes connect the prototype of in-vessel element to the external lithium supply / replacement system. The supply of lithium to the PFC is carried out under the effect of the hydrostatic pressure of the lithium column in the supply tank or by creating an excess pressure of lithium in the tank by supplying argon gas. The lithium stream leaving the PFC enters the drain tank and returns to the supply tank using an electromagnetic pump. The system provides a wide range of experimental possibilities for organizing the supply and replacement of lithium and is additionally equipped with measurement, monitoring and service elements. The first experiments with liquid lithium limiter of T-11M tokamak in pilot facility demonstrated the possibility of creating a uniformly distributed flow of lithium both in the structure and on the surface of the CPS. The rate of Li exchange in the CPS was ~0.4 cm3/s and the surface flow rate was ~ 1 cm/s, which was determined by the change in the level of LM in the supply tank and visually by the speed of the special marker on flowing surface respectively.
The problems in program points 5 and 6 are solved using thermal model (Fig.4), providing a study of the process of cooling the PFC at heat flux in the range of 1-15 MW/m2 in the modeling conditions of the experimental facility. The thermal load on the mockup surface is applied using a scanning electron beam. Heat removal from the heat receiving surface of the model is provided by evaporation of water droplets coming to the inner side of the receiving wall with a gas-drop stream generated by the atomizer (Ref. 3). The operation of the atomizer is provided by the system for supplying the coolant components (air, water) at various adjustable values of pressure (0.1-0.3 MPa) and flow rate (for water of 0-250 l/h; for air - 0-10 Nm3/h). The evaluation of the heat exhaust efficiency is controlled by the value of the receiving surface temperature and using a calorimeter. Experiments are conducted starting from the end of 2017. To date, it has been found that at power flux up to 12 MW/m2 the effective heat removal is provided at a coolant pressure not higher than 0.2 MPa with a heat transfer coefficient of 70-100 kW/m2K. Thermal model to study the process of heat transfer with the coolant flowing perpendicular to the cooling wall was developed and prepared for testing in 2020-2021.
The results of creation and tests of the prototypes and models made it possible to develop and justify the design of the complex steady-state operating lithium limiter of the tokamak T-15MD, ensuring the implementation of a closed circuit of atoms of lithium in the tokamak chamber without it uncontrolled accumulation, to optimize the operating parameters of the cooling system and properties of gas-water coolant.- Evtikhin V.A., Lyublinski I.E., A.V. Vertkov, et al., Liquid Lithium Tokamak Reactor, Fusion Energy 1996 (Proc. 16th IAEA Conf. on Fusion Energy, Montreal, 1996). Vol. 3, IAEA, Vienna (1997) 659-665.
- Vertkov A., Lyublinski I., et al., Experience in the development of liquid metal plasma facing elements based on capillary-porous structures for a steady-state tokamak (survey). — Physics of Atomic Nuclei, 2018, vol. 81, № 7, p. 1000—1007.
- S.V. Mirnov, A.T. Komov , A. N. Varava et al., Features of heat transfer with cooling by dispersed flow, Now. 2019, Journal of Physics Conference Series 1370:012002 DOI10.1088/1742-6596/1370/1/012002.
Speaker: Alexey Vertkov (JSC “Red Star”) -
WEST is an actively cooled full W tokamak aiming at power exhaust studies in long / steady-state pulses. WEST can operate in lower single null (USN), upper single null (USN) and double null configuration with an aspect ratio of 5-6. The lower divertor was partially made of ITER-like target in phase 1 (2016-2019), it will be fully made of ITER-like PFC in phase 2 starting autumn 2020 [1, 2]. For this purpose two lower hybrid current drive (LHCD) launchers and three ion cyclotron resonance heating (ICRH) antennas, all actively cooled, have been installed and commissioned.
In the 2019 campaign, the LHCD and ICRH coupled powers have both reached ~5MW/1s separately and 8.8MW/0.5s when combining the two RF systems. Long pulse operation was also carried out with LHCD (PLH=3.0MW) extending the pulse length to 55 seconds (Figure 1). The experiments were performed at high magnetic field (Bt=3.6-3.7T) in a large range of plasma configurations: X-point plasmas (R~2.5m, a~0.45m, kappa~1.3) in LSN and USN configuration, plasma current in the 0.3-0.7MA range (q95~3-6), electron density in the 2.5-8.5×1019 m-3 range (ne/nGW=0.3-0.8).
The fraction of radiated power Prad/Ptot is generally 50-55% but boronization of the vessel walls reduces this fraction to 30-40%. The effective charge Zeff is also significantly reduced to ~2 at high power. Tungsten is, in most cases, the major radiating species but no sign of tungsten accumulation in the core is observed in MHD-free discharges. Nitrogen injection during the early phase of the discharge is found to be beneficial for increasing the edge radiation and peaking the electron temperature. It results a weaker MHD activity and higher performance of the RF-heated plasma.
In L-mode, the stored energy, WMHD, increases according to the H96 L-mode scaling law up to 350kJ, L-H transition was observed, after fresh boronization, when combining 4MW of LHCD with 1MW of ICRH (Figure 2). It results a significant increase of the particle confinement time (30% increase of plasma density with gas injection turned off) but the plasma radiation increases leading to an oscillatory regime.
LHCD allows to reduce the loop voltage to ~0.1V up to ne=4x1019 m-3 indicating a current drive efficiency of 0.5-0.6×1019 A.W-1m-2 for 0.5MA discharges (Figure 1). Higher efficiency (~0.7-0.8×1019 A.W-1.
m-2) can be achieved after a fresh boronization or at lower plasma current (0.4MA) when the LH power deposition profile, deduced from the hard X-ray diagnostic, is more peaked.
Although the stored energy of ICRH-heated discharges increases accordingly with the confinement scaling law, moderate central electron heating is found in particular when ICRH is combined with LHCD. Tungsten sources from an ICRH antenna were investigated from visible spectroscopy [4, 5]. When the antenna is powered with 1MW, the tungsten flux (resp. sputtering yield) increased by a factor 10 (resp. 5) near the antenna mid-plane. However, the increment of fraction of radiated power is generally small when ICRH power is added on a LHCD discharge for well-tuned antennas [4].
In addition to the experimental results, simulations with the 1.5D METIS code [6] and the 3D C3PO/LUKE code [7] to verify the coherency between electron temperature, loop voltage, radiation , plasma composition are presented.
Acknowledgements: This work has been carried out within the framework of the French Research Federation for Fusion Studies and of the EUROfusion Consortium.
References
1 J.Bucalossi and the WEST team, see http://west.cea.fr/WESTteam, Fusion Eng.and Design 89 (2014) 907-912
2 C.Bourdelle et al., Nucl. Fusion 55, 063017 (2015)
[3] C.C.Klepper et al., 47th Conf. on Plasma Phys., Sitges, June 22-26 2020, invited paper
[4] J.Hillairet et al., this conference
[5] L.Colas et al., this conference
[6] J.F.Artaud et al., Nucl. Fusion 58 (2018) 105001 (25pp)
[7] Y.Peysson et J.Decker, FST, 65 (2014) 22Speaker: Marc Goniche (CEA, IRFM) -
To resolve concerns for heat/electromagnetic loads from vertical displacement events (VDEs), we developed an equilibrium controller, which can predict and control VDEs to an arbitrary direction and allows us to cope with either upward or downward unmitigated VDEs. Here we improve the prediction rate by parameterizing a power supplies voltage saturation rate with a newly developed adaptive voltage allocation (AVA) scheme. Furthermore, the AVA also broadens accessible elongation ($\kappa_X$) by 4% (red dashed to green curve in Fig. 1) in the presence of power supplies voltage saturation due to hybrid plasma current ($I_p$) and shape/position control and unknown plasmas disturbances. In preparation for coming JT-60SA integrated commissioning, the VDE predictor and the AVA scheme are implemented in our equilibrium controller, and its performance is demonstrated by using MHD Equilibrium Control Simulator “MECS” [e.g.,$\sim$1].
Highly elongated plasmas are essential to achieve high fusion output in tokamak devices, while high $\kappa$ makes VDE control difficult. Once VDEs are triggered, which causes significant asymmetric heat loads on first walls and electromagnetic torques on conducting materials such as especially in ITER [2]. Although disruption prevention including VDEs has been extensively explored to reduce the disruption rate to achieve ITER requirements [3], it may still not be easy to mitigate all disruptions perfectly, because it requires forecasting disruptions before delay time of actuators such as control coils, gus-puff, and pellets, etc. In large superconducting devices such as ITER, DEMO, and JT-60SA, number of control coils is limited, and high inductance of each control coil causes saturation of power supply voltages to control coil currents. In the absence of in-vessel coils such as in DEMO and JT-60SA during the integrated commissioning, we should achieve hybrid control of $I_p$, plasma shapes, and VDEs in the presence of the voltage saturation.
We first developed an equilibrium controller with a new AVA scheme, which measures the voltage saturation rate to control position/shape and $I_p$. The developed controller has well been tested in MECS [1] with realistic experimental conditions of JT-60SA, such as, noises of diagnostics and its filters, control delays, coil current and voltage limits, etc. Notably, axisymmetric passive/active structures are assumed in MECS, whose VDE growth rate has no significant difference from that calculated with 3D structures and eddy currents [4]. Here, we briefly explain our new equilibrium controller with the AVA scheme, which is updated from [1]. First, control values are
$~~~\delta\vec{\psi}_{\rm{S,PID}}=G_{\rm{SP}}\delta\vec\psi_{\rm{S}}+G_{\rm{SI}}\int\delta\vec\psi_{\rm{S}}dt+G_{\rm{SD}}\frac{\partial\delta\vec\psi_{\rm{S}}}{\partial{t}},\\ ~~~\delta\vec{\psi}_{\rm{X,PID}}=G_{\rm{XP}}\delta\vec\psi_{\rm{X}}+G_{\rm{XI}}\int\delta\vec\psi_{\rm{X}}dt+G_{\rm{XD}}\frac{\partial\delta\vec\psi_{\rm{X}}}{\partial{t}},\\ ~~~G_{\rm{SP}},G_{\rm{SI}},G_{\rm{SD}},G_{\rm{XP}},G_{\rm{XI}},G_{\rm{XD}}:{\rm{PID~gains}},$
where the $\delta\vec\psi_{\rm{S}}$ and $\delta\vec\psi_{\rm{X}}$ are the residual flux for position/shape and $I_p$ control. The PID gains are optimized by frequency response analyses and are fixed during simulation. Control coil currents are determined by $\delta{\vec{I}}=\vec{M^\dagger}\delta\vec\psi$, where $\vec{M^\dagger}$ is a pseudo-inversed control matrix calculated from the Green function of the coils. From the circuit equations, the power supply voltages can be obtained by $\vec{V}=\vec{F}(\delta\vec{I})$, where $\vec{F}$ is a vector function to calculate the circuit equation. Here, we newly introduce AVA gains ($G_{\rm{S,AVA}},G_{\rm{X,AVA}}$), which are determined so that voltages calculated from $\vec{F}(G_{\rm{S,AVA}}\vec{M^\dagger}\delta\vec\psi_{\rm{S,PID}})$ and $\vec{F}(G_{\rm{S,AVA}}\vec{M^\dagger}\delta\vec\psi_{\rm{X,PID}})$ are not saturated by using binary search. The resultant $G_{\rm{S,AVA}}$ and $G_{\rm{X,AVA}}$ represent voltage saturation rates ($\ge{1}$) to compensate the residual flux for position/shape and $I_p$ control. Finally, output voltages are
$~~~\vec{F}(\vec{M^\dagger}\delta\vec\psi_{\rm{S,PID}}+G_{\rm{X,AVA}} \vec{M^\dagger}\delta\vec{\psi}_{\rm{X,PID}}).$
Here, the $G_{\rm{X,AVA}}$ only is used to reduce the influence from the $I_p$ to the position/shape control from the machine protection point of view. The $G_{\rm{S,AVA}}$ only is used as a proxy to judge the VDE probability.A novel VDE direction control algorithm is developed with the AVA scheme, which guides the VDE to an arbitrary direction and halves the concerns to protect the device from either upward or downward unmitigated VDEs, although the perfect mitigation/avoidance of VDEs is favorable. The typical VDEs detection and control behavior is shown in Fig. 2, where we assume 1% minor-collapse of total $I_p$, which is frequently observed in tokamak plasmas such as with magnetic islands. As shown in Fig. 2(a) and (b), the plasma vertically oscillates and finally a downward VDE takes place. Our controller detects this downward VDE at the timing shown by a filled circle in Fig. 2(c). After the detection, the downward VDE is controlled to the upward VDE by fixing coil currents. In JT-60SA, this downward to upward control is planned to protect lower walls from heat loads. As shown in Fig. 2(d), the $G_{\rm{S,AVA}}$ decreases towards the VDE, which is a powerful proxy for classification.
We developed the VDE predictor by a support vector machine (SVM) classifier with three parameters: the newly introduced $G_{\rm{S,AVA}}$, vertical velocity of current centroid $v$, and decay-index $n$, which is curvature of externally applied vertical field $B_z$ and defined by $-R_p/B_z\partial{B_z}/\partial{R_p}$ with a plasma major radius $R_p$. The stability is determined by the balance between the driving term and the suppression term from the controller, which is evaluated by $G_{\rm{S,AVA}}$. The $G_{\rm{S,AVA}}$ is newly adopted to evaluate the position/shape control performance in our SVM classifier, while the vertical velocity is a typical classification parameter [5], and the decay-index $n$ is a classical index for the VDE growth rate, where the more $n$ decreases, the more VDE growth rate increases. In Fig. 3, we control VDEs to opposite direction from Fig. 2, i.e., upward to downward. The classifier is trained by 17 simulation with 7 upward VDEs and 378/460747 trained/test data. The VDE predicted region is projected to ($n,v$) and ($n,G_{\rm{S,AVA}}$ ) plane as shown in Fig. 3(a) and (b). Here, the typical ramp-up scenario without VDEs are also shown by dot-points, which intersects the VDE predicted regions in Fig. 3(a). Thus, if we predict VDEs only by $n$ and $v$, which could cause many false alarms. On the other hand, by introducing the $G_{\rm{S,AVA}}$ as shown in Fig. 3(b), our SVM classifier successfully decouples the VDE predicted region from data without VDE, which perfectly eliminates the false alarms. Our equilibrium controller operates with ~100$\mu$s including the AVA scheme and the VDE prediction, which will satisfy the JT-60SA requirement.
An accessible elongation ($\kappa_X$) is explored by also using MECS during the $I_p$ flat-top (2.5 MA) in the presence of $I_p$ minor-collapses, as shown in Fig. 4(a). Here, drops of $I_p$ are 0-3% of total $I_p$, which can be considered as a margin of the accessible $\kappa_X$ in the presence of accidental disturbances. An improvement of the controllability against the minor-collapses can be seen when the $\delta{I_p}$ is not less than 2%, and the accessible $\kappa_X$ increases by 4% from 1.55 to 1.61 in the 3% $I_p$ drop case (see also red dashed and green curves of Fig. 1). Detailed temporal behaviors of two cases with 3% $I_p$ drop, that are with/without AVA and indicated by arrows in Fig. 4(a), are compared in Fig. 4(b). A sudden minor-collapse occurs at $t=15~$s, and in the absence of AVA, the oscillation cannot be suppressed even with lower $\kappa_X$ in comparison with the AVA case as shown in Fig. 4(b2). In the case with AVA, after the minor-collapse, the $G_{\rm{X,AVA}}$ is automatically reduced to around 0.1 as shown in Fig. 4(b3), which improve the VDE controllability. After 16.5 s, the oscillation is successfully suppressed and the plasma reaches steady states. We note the AVA scheme also improves the controllability during $I_p$ ramp-up/down, $\beta$ collapse, etc.
[1]$~$Y.$~$Miyata$~\it~et~al.$,$~$Plasma$~$Fus.$~$Res.$~$9$~$(2014)$~$3403045.
[2]$~$A.R.$~$Raffray$~\it~et~al.$,$~$Nucl.$~$Fus.$~$54$~$(2014)$~$033004.
[3]$~$E.J.$~$Strait$~\it~et~al.$,$~$Nucl.$~$Fus.$~$59$~$(2019)$~$112012.
[4]$~$N.$~$Cruz$~\it~et~al.$,$~$Contr.$~$Eng.$~$Prac.$~$63$~$(2017)$~$81.
[5]$~$S.A.$~$Sabbagh$~\it~et~al.$,$~$in$~$proceedings$~$of$~$IAEA$~$FEC$~$2018,$~$EX/P6-26.
Speaker: Shizuo Inoue (QST) -
Here we report an innovative divertor pumping system using proton conductor pump, fuel proceesing system with isotopic recycling by this device, and its impact on the tritium fuel sufficiency of DEMO, particularly that enables commissioning without initial loading of significant quantity of tritium prior to its operation. Proton conductor pump is developed by the authors that actively compress and transport hydrogen isotope gases from the divertor region. Pure DT mixture can immediately be recycled to the fuel feed that significantly shorten the cycle time and tritium inventory. On the other hand, modelling of fuel cycle of DEMO and its commissioning operation suggests plant can manage tritium self sufficiency with this direct recycling path.
It is well known that ITER will consume majority of world tritium supply in the near future, and any DEMO reactor following it will have to acquire its own fuel. The authors have reported that fusion DEMO can produce its tritium fuel to be needed in the series of tests by its own operation from initial DD discharge until reaching long pulse or steady state full power burning with no external supply[1].
Figure 1 shows the simplified primary fuel cycle that continuously process plasma exhaust and return it as a fuel. It is generally understandable that closed tritium fuel plant can collect tritium in the vacuum vessel and other components in the primary fuel cycle within reasonable dwell operation time, or in steady state operation. Such a fuel cycle requires some specific features related to the throughput and response time that has not been given sufficient consideration in the past. Typically DEMO will have the range of a few % of burnup and throughput of 200 – 500 Pam3/s in steady state operation. Fuel cycle evacuates and pressurizes plasma exhaust, chemically purifies hydrogen isotopes followed by an isotope separation process to prepare the fuel supply to the plasma. In the early stage of commissioning, fuel cycle will essentially be operated also in steady state, while tritium is supplied to the plasma in the pulses with dwell period that allows tritium recovery after the discharge.
Primary fuel cycle must process all the exhaust and return pure DT to the main plasma with minimal cycle time. Throughput is given by the fusion power and burn fraction. Since inventory is estimated from throughput multiplied by residential time constant, if total recycle time constant is too long, required tritium inventory will not be acceptable. For self-sufficiency, tritium has to be available for the plasma operation by the processing within the cycle time. In realistic operation scenario of the DEMO, cycle time of primary fuel cycle and each discharge pulses followed by dwell will have to be consistent with dynamic operation cycle. Primary fuel cycle involves condensed phase of hydrogen such as liquid in the distillation column, adsorption on catalysts or cryopump, and solid in the pellets, those requires considerable residential time and inventory of tritium fuel.
In this study, impact of the direct fuel recycle in the primary processing loop with minimal purification is considered and analyzed. Fuel system was described by a dynamics model, considering realistic commissioning of DEMO, that will be mainly pulsed DD and low concentration DT with full breeding blankets.
For the direct recycling of the bypassed DT fuel, an innovative concept of Proton conductor plates that selectively extract, pump and compress the purified DT fuel, is proposed and evaluated. Fig.2 shows the principle of the pump. Proton conductor membrane with both side with hydrogen permeable electrode can transport hydrogen isotopes electrochemically at large compression ratio. As experimentally confirmed and shown in the Fig.2, pumping characteristics is described by the Nernst’s Law, indicating high efficiency, typically 1 order of compression for 50 mV. For the divertor pumping, required compression will be 1 Pa to 100kPa, that is feasible with reasonable voltage. This pump has unique and particularly suitable characteristics for divertor pumping, because it does not have any mechanical action while active and controllable compression is possible.
Fig.3 shows the unique and attractive features of for divertor pumping for this purpose. Proof of principle experiments in Fig.2 showed the practical pumps for immediate recycling from divertor will be feasible. Pure hydrogen isotopes are selectively compressed for the fuel recycling, while residues that includes He ash and other impurities are sent to the plasma exhaust processing of the fuel cycle.
Major advantage of the application of this pump is a reduction in processing time and inventory. Typical time constant derived from the protonic impedance was around 20 seconds, because hydrogen ions are the main carrier in the conductor, as a kind of condensed phase. Since other part of the direct recycle path shown in Fig.1 has typical space velocity of orders of second, it significantly reduces the cycling time and fuel inventory by 2 orders of magnitude. The fuel loop which involves condensed phases such as distillation column, typical have time constant and invories of hours and kgs respectively.
This pump can be attached to the duct close to the divertor exhaust, for instance, approximately 1m2 plate for each divertor cassets. It will provide sufficient conductance and thus pumping capability. Further, because of the isotopic difference in the permeability, proton conductor pump will provide slight isotope separation for adjusting DT contents in the fuel.
This study suggests the practical and important fuel cycle strategy for fusion development. Bypassing DT fuel with practical pump will reduce the entire tritium inventory in the plant and shorten the recycling of tritium fuel. In the dynamics of fuel cycle, this feature means small amount of tritium injected to the pulsed plasma can be recovered in significantly shorter intervals and be ready for the next discharge. Such an efficient use of minimal tritium with self-sufficiency provides practical and feasible solution for the commissioning of the entire DEMO plant without external source. This result poses major impact on the fuel strategy for DEMOs after ITER.[1] Satoshi Konishi, , Ryuta Kasada, Fumito Okino,,“Myth of initial loading tritium for DEMO—Modelling of fuel system and operation scenario”, Fus. Eng. and Design, 121, October 2017, 111–116.
Speaker: Dr Satoshi Konishi (Institute of Advanced Energy, Kyoto University) -
One of the main goals of WEST is the assessment of power handling capabilities and lifetime of tungsten divertor components under high heat flux and high fluence operation in a full W tokamak environment [ 1 ]. In phase 1, covering the four experimental campaigns (named C1 to C4), WEST operated with a mix of actively cooled ITER like tungsten plasma facing units and inertial tungsten coated graphite components. The preliminary step towards testing the tungsten components is to determine the heat flux deposition pattern. For this purpose, a set of complementary diagnostics allowed characterizing properties of heat loads on the lower divertor: arrays of embedded thermocouples (TC) and Fiber Bragg Gratings (FBG), infra-red thermography on both strike point areas within the same field-of-view, and arrays of flush-mounted Langmuir probes (LP). The flux-mounted Langmuir probes are composed by two separate arrays of 29 probes each with 12.5mm spatial resolution. The embedded thermal measurements are composed by a set of 60 thermal probes such as TC and FBG embedded in the W-coated graphite components and deployed over the lower divertor as shown on figure 1. The strike point position (x0) and the intensity of the peak heat flux (qn), as well as the target heat flux decay length (λ_q^t) are computed simultaneously from the thermal measurements using inverse method based on finite element method solvers. A first heat flux calculation was performed during early diverted ohmic plasma experiments as depicted in [ 2 ]. On this basis, the study has been widely extended and heat flux has been computed for plasma experiments using additional heating power, such as Lower Hybrid (LH) or Ion Cyclotron Resonant Heating (ICRH). A heat flux data base covering the full phase 1 of WEST operation, with deuterium and helium plasmas, is presented in this paper. It includes about 300 plasma experiments achieving at least four seconds of steady state plasma to get significant heating and accurate heat flux calculation from embedded measurements.
In this paper, we report on the first heat flux database built from embedded thermal measurements as a function of the continuous progress achieved in WEST : from the first ohmic diverted plasma (obtained during the second experimental campaign C2 in 2018) up to the high power (up to 6 MW total injected power during 4 s) and high energy (up to 90 MJ total injected energy in lower single null configuration) steady state L-mode plasma experiments performed in the last experimental campaign (C4 in 2019). We report on the heat load pattern, position of the strike point position, peak heat flux and heat flux decay length, on outer and inner sides, as a function of the plasma parameters, heating power (Ptot), divertor power (Pdiv=Ptot-Prad-Pripple losses), plasma current (Ip), density (ne), and divertor magnetic equilibrium (far and close X-point). The consistency between the complementary diagnostics will be discussed on this database and compared with the different uncertainties corresponding to each diagnostics: IR (emissivity distribution evolving with space and time), TC/FBG (inverse method), LP (sheath heat transmission coefficient). Concerning the strike point position, a good agreement (<1cm) is obtained between thermal inversion, magnetic reconstruction and flush-mounted Langmuir probes measurement. The maximum heat flux currently reported on the W-coated graphite components is slightly above 5 MW.m-2. This was obtained with a conducted power on the divertor of ~2MW in two different scenario: 1/ with combined LH and ICRH power and optimized conditioning (low radiated fraction ≈40%) at low X-point height (dX=40mm), 2/ with LH power only and high X-point height to concentrate the scrape-off layer power on targets (dX=120mm).
For the same divertor power, the experimental heat fluxes are higher during C4 than C3 (see figure 2 red and blue points, respectively) while the heat flux decay length reported on the outer target remains similar (see figure 3). A clear improvement of the divertor performances in terms of heat flux capabilities is observed during this period, related to a possible change in the particle and heat transport dynamic. On the outer targets, the heat flux decay length on the targets remains the same and varies almost linearly with the magnetic compression of the field lines, from 60 mm for low compression down to 10 mm for high compression. The outer and inner heat flux ratio is found to be clearly asymmetric (in the 2/3 or two third range) as expected with the drift flows in forward magnetic field [ 3 ]. Those data are used to define WEST divertor operational domain and prepare the WEST phase 2 with an ITER-like fully actively cooled lower divertor allowing higher divertor heat flux (up to 20 MW/m2) and longer discharges (plus length up to 1000 s).
[ 1 ] C. Bourdelle et al, Nuclear Fusion 55 (2015) 063017.
[ 2 ] J. Gaspar et al, Fusion Engineering and Design 146 (2019) 757-760.
[ 3 ] R.A. Pitts et al, Journal of Nuclear Materials 337-339 (2005) 146-153.
Speaker: Dr jonathan gaspar (Aix Marseille Univ, CNRS, IUSTI, Marseille, France) -
The new Divertor Tokamak Test facility has taken off. DTT [ ] (Fig.1) is a superconducting tokamak with 6 T on-axis maximum toroidal magnetic field carrying plasma current up to 5.5 MA in pulses with total length up to 100 s. The D-shaped device is up-down symmetric, with major radius R=2.14 m, minor radius a=0.65 m and average triangularity 0.3. The auxiliary heating power coupled to the plasma at maximum performance is 45 MW, which allows matching the PSEP/R values with those of ITER and DEMO, where PSEP is the power flowing through the last closed magnetic surface. DTT is a divertor facility, i.e. it is designed to accommodate a variety of divertor configurations both in single and double null scenarios. In addition to this primary mission dedicated to plasma exhaust in regimes where plasma core and edge behaviors are integrated, DTT will provide a facility for high performance tokamak physics and to address core confinement and stability issues also in negative triangularity configurations, and the management of transient events like disruptions and ELMs. In the last eighteen months the design has progressed steadily, the legal entity for the DTT construction has been established, the 250 M€ loan from the European Bank of Investment given within the EU Juncker Plan has been activated, the first large procurement (superconducting strand) has been assigned, a strong scientific and managerial partnership with the Eni energy company has been established and the team is growing.
Progress in the engineering design. Advances in the design regarded the integration of machine components inside and outside the cryostat. The AISI 316L(N) vacuum vessel is segmented in 18 sectors, each with 5 access ports. The first wall is made by sprayed tungsten plasma facing units mounted on back structural supports that provide connection to the vessel and mechanical resistance against electromagnetic loads. The bulk tungsten single null divertor is made by 54 cassettes and is fully maintainable through remote handling. DTT will be equipped with 18 Toroidal Field (TF) modules, each made of a winding pack inserted in AISI 316L(N) casings. Each pack is made by cable-in-conduit Nb3Sn conductors operating at 43.5 kA with 12 T maximum magnetic field. The superconducting central solenoid and the poloidal field magnet system are supported by the TF magnet (total weight of the superconducting magnets to 440 tons). A layer-wound approach for the central solenoid is chosen and validated. The design of pre-compression bars and of support structures connected to the TF magnet is completed, together with the supports, inner joints and terminations of poloidal field magnets. The whole magnet system is enclosed in a thermal shield attached to the TF magnet. All interfaces are now well defined.
Heating The heating systems will be installed in two stages: a day-1 configuration with 24 MW coupled power and a second step to reach 45 MW. In the first stage 16 MW of ECH power, 3 MW of ICH and one 7.5 MW NNBI injector will be installed. The 170 GHz EC system is designed for a strong power localization and profile control, using fully independent front-steering antennas. ICRH (60-90 MHz) will be based on an antenna with 3.5 MW/m2 power density, shaped to fit the scrape-off layer and with an adjustable radial position and external matching. The NNBI is based on RF plasma sources capable to produce negative ion current of 40A accelerated up to 400-450 keV.
Physics Scenarios. The integrated modeling of DTT plasmas [ ] has been carried out with the JINTRAC suite of codes and covers the region inside the separatrix, calculating the pedestal pressure with EPED1 (Europed code) and fixing the pedestal density to achieve a volume averaged density ~ 0.43 nGW (normalized to Greenwald density). Heating is modeled self-consistently. The region inside the top of the pedestal is modeled with QuaLiKiz or TGLF quasilinear transport models and with NCLASS or NEO for neoclassical transport. Fig. 2 shows profiles obtained for the SN full power H-mode scenario with 32 MW ECRH, 15 MW NBI and 3 MW ICRH using QuaLiKiz for turbulent transport, which is mainly driven by ion-scale ITG/TEM. Core Te peaks at values above Ti due to strong central ECRH and stiff ion heat transport. Density profile is moderately peaked with central density ~ 21020m-3. The peaked rotation profile with core value of 50 krad/s does not provide significant EB stabilization of ion heat transport. Global plasma parameters for this scenario are βN=1.6, τE=0.28 s, total DD neutron rate ~1.41017 s-1 (30% thermal). Total radiation is 15 MW. The integrated modeling results are also validated against gyrokinetic simulations, to corroborate the validity of the quasi-linear models in the particular case of DTT.
One crucial issue for the DTT integrated physics approach, based on a proper weak similarity scaling to preserve the spatio-temporal scale hierarchy relevant to ITER/DEMO, is the role of energetic particles (EP) as mediators of cross-scale couplings [ ]. EP transport is a multi-scale process requiring a self-consistent, kinetic, treatment. Extending first-principle-based gyrokinetic simulations to transport time scales is a formidable task. This makes predictive analyses very challenging and calls for reduced descriptions which preserve the necessary physics ingredients. The paper will discuss how DTT is crucial for the validation of such reduced transport models.
Fast-ion losses due to trapped-precession resonance are estimated with the code ORBIT. Initial ion positions and pitch are calculated with METIS, using a full heating scenario and the geometry of the 400 keV NNBI. The LFS magnetic ripple of the reference SN scenario (δB⁄B~0.42% ) gives a small contribution, with up to 0.5% collisionless particle loss in the first ~1000 toroidal transits.
DTT MHD stability will be discussed. Stability of Alfvénic modes driven by energetic particles will be considered using hybrid MHD-Gyrokinetics simulations.
DTT is being designed with a high level of flexibility, in particular as far as divertor scenarios are concerned. From a magnetic point of view the external and internal coils allow to control and optimize the local magnetic configuration in the vicinity of the divertor target [ ]. The reference single null, double null and snowflake configurations can be produced at (or close to) the maximum current of 5.5 MA. A noticeable feature of DTT is that it allows for negative triangularity scenarios with proper divertor. Figure 3 shows a 5 MA single null scenario with δ=0.13 and δlower=0.16 and a double null at 3.5 MA with δ=0.38.
Preliminary 2D edge fluid-kinetic modelling of power exhaust is being done both in pure deuterium and with argon and neon seeding. In pure deuterium detachment is obtained with PSOL~10 MW for single null divertor and at higher values for snow flake configurations (PSOL~15-20 MW). Operation at maximum input power calls for high radiation fractions (Prad/Psol ~ 80-90%), which can be obtained with impurity seeding. Both at low and high separatrix density (ne,sep~0.51020 m-3 and ~110 m-3) Prad/Psol ~ 90% can be obtained in single null, double null and snow-flake configurations. Zeff at the separatrix =2.3 is needed for high density single null configurations.
References
1) DTT interim design report (2019) https://www.dtt- project.enea.it/downloads/DTT_IDR_2019_WEB.pdf
2) Casiraghi I., this conference
3) Falessi M. et al., this conference
4) Ramogida G., this conference
Speaker: Prof. Piero Martin (Consorzio RFX and DTT s.c.r.l.) -
A new drift-kinetic theory and computational approach to understand the plasma response to small magnetic islands associated with neoclassical tearing modes (NTMs) is presented. It demonstrates that drift effects associated with the passing particles support a pressure gradient across a sufficiently small magnetic island, leading to a suppression of the instability drive from the bootstrap current perturbation. This result is an important input to quantifying the control requirements of future reactor-grade tokamaks, such as ITER.
NTMs degrade the performance of tokamak plasmas. They may hamper progress towards high fusion power in planned JET deuterium-tritium (DT) plasmas 1, and they are a concern for ITER and future fusion power plants. The instability drive is understood: above a threshold in beta-poloidal, and if an initial “seed” magnetic island exceeds a critical width, the plasma pressure gradient is flattened within that island. This removes the bootstrap current there, resulting in a filamentation of the total current density that amplifies the initial seed island, removing more bootstrap current and driving the island to even larger widths. The resulting large saturated island degrades confinement.
A key question for theory is: What is the physics that determines the critical island width and threshold poloidal beta? This is important for understanding how to develop NTM avoidance strategies and for quantifying the requirements of control systems, such as the electron cyclotron current drive (ECCD) planned for ITER. Experimentally, the critical island width is found to be ~2-3 trapped ion banana orbit widths 2. Thus, a theoretical model must retain finite ion banana width effects, requiring a kinetic approach. This is the aim of the present paper – to develop a kinetic theory of the ion response to small scale magnetic islands, and explore the consequences for NTM threshold physics.
Drift-kinetic models in 5D space (3 spatial and 2 velocity coordinates) are challenging to solve, but simulations have shown that the ion contribution to the bootstrap drive is reduced for island widths below the ion banana width [3]. We extend that theory to also solve for the electron response and self-consistently determine the impact of the electrostatic potential required for quasi-neutrality. The problem is made tractable by exploiting the small ratio of poloidal ion Larmor radius to plasma minor radius, and developing an expansion to reduce the dimensions of the system. To leading order, one finds that the orbit trajectories for both trapped and passing particles are unperturbed by the presence of a magnetic island. Proceeding to next order, averaging over the orbits provides a solvability condition for the particle response in a reduced 4D space: the canonical toroidal angular momentum, the helical angle labelling magnetic field lines on a flux surface, and two velocity coordinates (speed and pitch angle).
We have previously solved this 4D system numerically, employing a simplified model for the electrons that exploits the small ratio of electron poloidal Larmor radius to island width [4,5]. The result for the passing ions is particularly interesting – their distribution function appears to be flattened across a “phase space” island structure (not the magnetic island) that is very similar to the magnetic island, but shifted radially relative to it (fig 1). When the flow of the passing particles along magnetic field lines is reversed, the radial shift of the phase space island relative to the magnetic island is also reversed. A consequence of this shifted island structure is that a pressure gradient is supported across the magnetic island (fig 2). When the island width is rather larger than the poloidal Larmor radius, the shift of the phase-space island relative to the magnetic island is negligible and the pressure is then flattened across the magnetic island as expected in the standard NTM model.
To understand the physics of the simulations, we further reduce the model by considering the limit of small collision frequency, much less than the island propagation frequency in the ExB rest frame [6]. Neglecting collisions to leading order, the orbit averaged equations reveal that particles follow streamlines that represent the combined effect of parallel streaming and cross-field drifts: grad-B, curvature and ExB. These streamlines, which we derive analytically, define “drift surfaces” (Fig 1). In the absence of the ExB drift we find that the drift surfaces for the passing particles are exactly the same as the magnetic island flux surfaces, but shifted radially by a few poloidal Larmor radii. When collisions are neglected, the particle distribution function is constant on these drift surfaces. Note in Fig 1 the agreement between the analytic drift surfaces and the “phase-space” island structures of the full, 4D numerical solution (colour contour plot). The origin of the pressure gradient that is supported inside the magnetic island is therefore a consequence of passing particle physics.
We have shown that the distribution function is constant on the drift surfaces, but we do not know how it varies across them. We re-introduce collisions into the model as a perturbative correction and average over the drift surfaces to derive a transport equation for the distribution function that balances collisional diffusion across the drift surfaces with pitch-angle scattering. This equation reveals two interesting boundary layers: one close to the separatrix of the island in the drift surfaces and one close to the trapped-passing boundary in pitch angle. The latter requires special attention as the collisions cannot be treated perturbatively there, so a new collisional boundary layer equation is derived. The solution of this layer equation connects trapped and passing particle solutions. An iteration procedure provides the self-consistent electrostatic potential and hence the full distribution function.
This semi-analytic calculation predicts a density profile that supports a gradient within magnetic islands that are sufficiently narrow (a few poloidal Larmor radii), confirming the conclusions of the 4D numerical solution. However, there remain some differences between predictions for the distribution function obtained using the two approaches, and these differences influence the current density distribution in the island vicinity. We show how the part of this current density that is in phase with the magnetic island influences whether it grows or decays (ie the conditions for a threshold). On the other hand, dissipation, which in our model is dominated by the effect of the trapped-passing boundary collisional layer, drives the part of the current density which is out of phase with the island; this determines the island propagation frequency.
Supported by UK Engineering and Physical Sciences Research Council, grant EP/N009363/1.
1 L. Garzotti, et al Nucl Fusion 59 (2019) 076037
2 R.J. la Haye, et al Nucl. Fusion 46 (2006) 451
[3] E. Poli, et al Phys Rev Letts 88 (2002) 075001
[4] K. Imada, et al, Phys Rev Letts 121 (2018) 175001
[5] K. Imada, et al Nucl. Fusion 59 (2019) 046016
[6] A. Dudkovskaia, University of York PhD thesis “Modelling neoclassical tearing modes in tokamak plasmas” (2020)Speaker: Prof. Howard Wilson (University of York) -
Removal of corrosion impurities from the molten Pb-16Li, which is a candidate breeder and coolant for the liquid metal breeding blanket system, is indispensable for the safe and efficient operation of the future fusion reactors. The corrosive nature of Pb-16Li in high temperature environment is responsible for the generation of corrosion impurities from the structural material. The deposition of corrosion impurities in the non-isothermal Pb-16Li loop may pose serious challenges (e.g. flow restriction due to plugging of flow path which further lead to the reduced heat exchange, generation of long lived activated corrosion products etc.) to the performance and safety of Pb-16Li breeding blanket system. In order to remove those corrosion impurities and activated corrosion products from the molten Pb-16Li, a purification component “Cold Trap” is generally recommended in the Pb-16Li loop system.
As a part of the development and testing of a Pb-16Li purification system, a wire mesh packing based prototype Cold Trap (CT) was designed, fabricated and installed in a purification loop with an objective to remove the impurities from ~ 260kg of Pb-16Li. In present work, we have assessed the functionality of CT in molten Pb-16Li system, continuously for a long duration of ~3500 hrs (~ 5 months). This long duration operation of CT in Pb-16Li system was imperative to test its reliability, since the recurrent replacement of the CT in the Pb-16Li system may not be feasible due to stringent operating conditions of fusion reactors. During the present experiment, we have achieved the reduction in concentration of impurities in Pb-16Li to their saturation solubility by passing the impure Pb-16Li through CT. The reduction in impurity level was also confirmed by the deposition of those dissolved impurities inside CT mesh packing in the needle shaped crystal formation. Some oxide depositions were also found inside the CT. Different characterization techniques such as SEM-EDX, XRD and ICP-AES were used for the characterization of the samples taken from the CT. The characterization results of deposited needle shaped crystal suggest the formation of Fe-Ni intermetallic while the XRD analysis of oxide deposition samples suggests the formation of Pb2O, Li2O and PbO.
Experimental details and main results: The prototype CT was installed in a bypass line of purification loop and a bleed (~10% nominal flow rate of the system) of Pb-16Li was passed through it for removal of dissolved corrosion impurities. In order to avoid freezing of Pb-16Li inside CT, Pb-16Li temperature inside the CT was maintained at ~260-270 oC which was ~25-35oC higher than the melting point of Pb-16Li ~235oC. The molten Pb-16Li inside the CT was cooled down to desired temperature using a temperature feedback controlled air cooling. During the experiment, samples of Pb-16Li were collected in-situ at different intervals to monitor the decrease in impurity concentration in Pb-16Li without disturbing the loop operation. In order to check the performance of CT for high solubility impurity like Ni, an impurity saturator was installed in bypass line of the purification loop to increase the concentration of Ni in Pb-16Li during the operation of system.
After completion of experiment, the corrosion impurities were found to be deposited in the needle shape crystalline form (~6-9 mm long needles and needle clusters) on the surface of the wire mesh packing of CT (refer Fig 1a). Small depositions of needle shaped crystal were also observed on the surface of CT outlet pipe (refer Fig 1b). The oxide impurities were found to be deposited on the internal walls of CT (refer Fig 1c).
The samples taken from the loop were analysed using ICP-AES techniques to determine their concentration in Pb-16Li. In the Pb-16Li samples taken after completion of experiment, the concentrations of Iron (Fe) and Nickel (Ni) which are the major corrosion impurities are found to be ~25ppm and ~925ppm, respectively. The concentration of these impurities are found to be near their saturation solubilities (~21 ppm for Fe and ~977ppm for Ni [refer 1, 2]) in the operating temperature range of CT (~260-270oC) and hence suggesting the deposition of excess impurities inside CT. In order to confirm this, the ICP-AES analysis of deposited crystals were performed. The concentration of Iron (Fe) and Nickel (Ni) in crystals deposited on wire mesh of CT are found to be ~74000 ppm and ~22000 ppm, respectively, which confirms the deposition of dissolved Fe and Ni impurities. The XRD and SEM-EDX analysis (refer Fig 2a and 2b) of crystals, deposited on the wire mesh packing of CT, also suggests the deposition of impurities in form of mainly Fe-Ni intermetallic.
In addition to this, the ICP-AES analysis results of Pb-16Li samples also shows the decrease in the concentration of Copper (Cu) from ~190 ppm to ~50 ppm during the experiment. The oxide samples were analysed using XRD technique and the analysis results (refer Fig 2a) suggests the presence of PbO, Pb2O and Li2O.References
[Ref 1] Investigation of models to predict the corrosion of steels in flowing liquid lead alloys, F. Balbaud-Celerier, F. Barbier, Journal of Nuclear Materials 289 (2001) 227-242
[Ref 2] The solubilities of nickel, manganese and chromium in Pb-17Li, Marten G. Barker and Tony Sample, Fusion Engineering and Design 14 (1991) 219-226Speaker: ANKUSH DEOGHAR (INSTITUTE FOR PLASMA RESEARCH) -
CEA has been involved for seven years (2014-2020) in the pre-conceptual design study of the magnet systems EU DEMO [1, 2] and for 2 years (2018-2020) in the industrial exploratory study for the cryogenic system and cryo-distribution for EU DEMO. In the framework of these Eurofusion workpackages, significant engineering and technological results have come out from the studies and they have paved the way towards the conceptual design of the cryogenic users and the associated cooling system for the EU DEMO future tokamak reactor (2021-2027).
The design of the cooling systems for the superconducting magnets at about 4 K, the High Temperature Superconducting Current Leads at 50 K and the thermal shields at about 80 K, has been investigated whereas the cooling parameters are not yet fixed at this stage of pre-conceptual design. It allows some iteration loops and discussions for optimal cooling parameters at the interfaces (temperature, pressure, mass flows, heat loads) [3]. This design methodology, which has been conducted before for previous cryogenic systems (K-STAR, JT-60SA, ITER), is an innovative engineering challenge in close collaboration with the cryogenic user designers, the cryogenics specialists and the industry. Indeed, it allows designing an optimal overall cooling system for the superconducting magnets and the thermal shields. The optimization relies mainly on minimization of the refrigeration loads with respect to acceptable cooling requirements, hence optimization of the general layout (volume and mass of the equipment), leading to significant reductions of investment and operation costs. Moreover, the implication of the industry at the early stage of the pre-conceptual design has shown to be relevant and has brought some fruitful insights for the future EU DEMO cryogenic system and cryo-distribution in terms of engineering aspects and R&D developments.
The specifications of the future cryogenic system and cryo-distribution for EU DEMO have led to the study and the comparison of several process options in order to meet the cooling requirements. The estimation of the overall refrigeration loads for EU DEMO cryogenics users is around 100 kW equivalent at 4.5 K, with a large contribution of the 80 K loads for the thermal shields (48% of the total cryogenic user loads in case the vacuum vessel is at 373 K, or 64% at 473 K for a continuous baking option). This interactive work between designers and industry enables to explore potential ways to reduce the total refrigeration loads, by enlarging the cooling parameter ranges and by studying the impact on the cryogenic users.
Integrated modelling tools [4] (thermal hydraulic, cryogenics, thermal-mechanics, electromagnetic) and iterative calculations were developed to address and to assess the cooling requirements.
Trade off process studies and estimations on the expected performance of the cryogenic equipment such as compressors, circulators and turbines are very useful to identify the critical issues related to the cooling requirements for the nominal operation, but also for the off design modes (cool down and turndown operation with reduced refrigeration loads). The technical gaps and the requested R&D developments are highlighted. Their identifications allow the industry to address them in the near future to be able to supply a suitable cryogenic system and cryogenic distribution for EU DEMO tokamak reactor.Finally, one of the main result of the industrial study was the general layout of the cryogenic system and its distribution, relying on the baseline solution with the helium and the mixed gas refrigeration processes (Fig.1). The choice of studying this process is a reasonable innovative approach, in favor of potential gain in terms of efficiency of about 5% compared to conventional helium process for large cryoplants. The resulting general layout of this process is of primary importance for engineering aspects and integration plant for EU DEMO reactor. Fig. 2 shows the preliminary layout of the cryogenic building, with the distribution box, the Helium cycles and Mixed gas Refrigeration cycles with 2 units each for redundancy. Safety aspects related to a fusion nuclear reactor are addressed and their impacts on the design are analyzed.
Acknowledgment
This work has been carried out within the framework of the EUROfusion Consortium for the EU-DEMO work package WPPMI. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
Speaker: Christine Hoa (CEA) -
The confinement of fast particles in fusion devices is shown to be strongly affected by Alfvén activity 1 and in order to model the associated wave particle interaction, a proper identification of the excited Alfvén Eigenmodes (AEs) is needed in the first place. The shear Alfvén spectrum is highly dependent on the iota profile of the magnetic configuration 2, which in turn depends on the net current present in the plasma. It has been proven that, in the TJ-II stellarator (R = 1.5 m, a = 0.2 m) 3 small changes on the iota profile, caused by the plasma current, can have a strong impact on the observed spectrum of shear Alfven waves [4, 5]. Ideally, a precise measurement of the iota profile, or failing this, an approximate knowledge of the different contributions to the plasma current is mandatory when it comes to obtaining reliable estimations of the shear Alfvén spectrum. While current drive in tokamaks has been extensively studied both experimentally, theoretically and numerically including comparisons with experiments for NBCD model validation, there have been only very few studies such as [6] of current drive in 3D non-axisymmetric configurations including stellarators so far.
The goal of the present work is the calculation of the neutral beam current drive (NBCD) for non-axisymmetric configurations and its validation in TJ-II NBI heated scenarios. Figure 1 shows the shear Alfven spectrum calculated with the STELLGAP code [7] for different plasma current profiles corresponding to different heating scenarios to illustrate the wide variety of conditions available in TJ-II for these studies.
In co-NBI scenarios (no current compensation), the dominant source of current is NBCD. This current is carried by the fast particles created during the NBI injection and its subsequent slowing down process. Theoretical models [8-10] that calculate the NBCD are available in the literature for different regimes and approximations and all of them rely on the assumption that the steady state fast ion distribution function is known. In order to obtain the fast ion distribution, numerical simulations must be performed. NBCD analysis in W7-AS [6] was carried out by means of a Fokker-Planck simulation code using first a simplified model to obtain the slowing down distribution function and then taking this distribution as input for theoretical NBCD models in the collisional [11] and collisionless [12] regimes. Preliminary estimations of NBCD (Fig. 2) were carried out for TJ-II plasmas [13] considering a more accurate calculation using Monte Carlo simulations. There, beam current drive was calculated using first the Monte Carlo code ASCOT [14, 15] to obtain the fast ion distribution function and NBCD was obtained using the model valid for the long mean free path regime [12].
The simulations performed with ASCOT did not include the charge exchange processes (CX), which have been shown to be very important in the range of energy of TJ-II plasmas; therefore, the NBCD results were corrected with simulations performed with the Monte Carlo code FAFNER [16], which is able to calculate the percentage of lost power due to CX. The calculations showed a reasonable agreement with the experimental data (Table 1).
In this work, neutral beam current drive is thoroughly investigated from numerical and experimental point of view, by carrying out simulations for different plasma parameters obtained in different heating scenarios (NBI with and without ECRH) and comparing with the large available data set. The comparison of the calculated NBCD with the one induced in the plasma when NBI is injected needs an estimation of the bootstrap current [17] contribution to the total measured plasma current (I_NBCD = I_P – I_BOOT). Where possible, a benchmark between FAFNER and ASCOT, both codes using now CX processes, will be presented in order to address the consistency of the theoretical-computational framework.
References
1 W.W Heidbrink, Plasma Phys. Control Fus., 56 ,095030 (2014)
2 Ya. I. Kolesnichenko, Physics of Plasmas, 8, 491 (2001)
3 E. Ascasibar et al., Nuclear Fusion, 59, 112019 (2019)
[4] A. Cappa et al., P4.1040, 45th EPS, Prague, July 2018
[5] A. Rakha et al., Nuclear Fusion, 59, 056002 (2019)
[6] N. Marushchenko et al., D-5.003, 29th EPS, Montreux, June 2002
[7] D. Spong et al. Phys. of Plasmas 10, 3217 (2003)
[8] T. Ohkawa, Nuclear Fusion, 10, 185 (1970)
[9] D. F. H. Start et al., Plasma Physics, 22, 303-316 (1980)
[10] Y. J. Hu et al., Physics of Plasmas, 19, 034505 (2012)
[11] S.P. Hirshman, Physics of Fluids 21, p1295 (1978)
[12] N. Nakajima and M. Okamoto, NIFS-27, Japan, (1990)
[13] S. Mulas et al., P90, 22nd ISHW, Madison, September 2019
[14] J. Varje et al., Submitted to Comput. Phys. Commun. (2019)
[15] O. Asunta et al., Comput.Phys. Commun. 188, 33 (2015)
[16] G. Lister, Max-Planck-Institut für Plasmaphysik Technical Report IPP 4/222, 1985
[17] J.L. Velasco et al., Plasma Phys. Control. Fusion 53, 115014 (2011)Speaker: sadig mulas -
The research on the physical and mechanical properties of structural and functional ma-terials of future fusion reactors has been actively conducting for many years in the Republic of Kazakhstan [1, 2].
Specialized KTM tokamak is being developed in Kazakhstan to research the behavior of candidate materials of the first wall under the conditions of high heat flows comparable to those in future thermonuclear reactors [9-11]. This is the tokamak with aspect ratio equal to 2, single-zero divertor plasma configuration, maximum plasma current of 750 kA, toroidal mag-netic field of 1 T, and duration of a purely inductive discharge scenario with basic parameters of τpulse≤1 second, and up to 5 seconds using an additional RF plasma heating system with a maximum heating power of 5 MW. The maximum design capacity of the thermal load on the receiving divertor plates is 20 MW / m2, which is comparable to the expected loads in the di-vertor area of the ITER thermonuclear reactor [12].
The main qualitative difference between the KTM tokamak and similar installations is the presence of transport-gateway and receiving-divertor devices. This allows replacing the test samples in the shortest possible time, without depressurizing the vacuum chamber (VC), which increases the experimental capabilities of the KTM tokamak.
Figure 1 shows the KTM tokamak appearance.
At the end of 2019, an experimental campaign was conducted to implement the KTM tokamak physical start-up. The experiments resulted in the plasma discharge of working gas hydrogen with the discharge duration of 65 ms and the maximum plasma current of about 100 kA with the toroidal field of 0.9 T. The KTM Tokamak physical start-up of was carried out using standard sources of pulsed power supply of the KTM tokamak electromagnetic system (EMS) coils [5, 6].
The analysis of the plasma initiation stage in the KTM tokamak was performed using TRANSMAK code [7]. According to the calculations the area of a breakdown has size of about 0.5 m with the magnetic field modulus value equal to 5 GS and is situated at the inner bypass with the center at R=0.6÷0.7 m radius. The value of the electric field intensity in the breakdown area is about 1.6 V/m, which at a radius of 0.7 m equals to 7 V voltage at bypass.
Before the experiments on obtaining plasma the KTM tokamak VC was lined with graphite tiles of FP 479 brand in two diametrically opposite poloidal sections.
The VC preparation for the experiments included vacuuming using two turbomolecular pumps with a pumping speed of 2000 l/s each and heating the VC at a temperature of 130 ºС for 8 days with subsequent treatment for 48 hours with a glow discharge in a hydrogen, helium and argon medium. Cleaning the VC with a glow discharge was also carried out at nights during the inter-start-up period. The maximum level of the residual gases pressure ac-cording to the results of the VC preparation was 2·10-7 Torr.
Starting systems, registering and processing the data during discharges was carried out using the standard KTM tokamak experiment automation system [8, 9].
Figure 2 shows parameters of one of the plasma discharge (discharge No.3669) with the maximum plasma current value of about 100 ka and the average electronic plasma density of about 1.5·1019 m-3 (the linear plasma density nedl is no more than 8·1018 m-2). In the discharge No. 3669 the plasma current growth rate is about 2.5 MA/s.
Figures 3 shows video frame of the plasma discharge No. 3669 observed by the KTM video camera system [10] and frame of the plasma position and shape reconstructed from magnetic measurements. The current thread method with a fixed position [11] was used to reconstruct the current, position, and shape of the plasma. Eddy currents induced in the con-ductive structures of the installation were taken into account in the process of solving the re-construction problem.
The synopsis describes the main experimental results of obtaining plasma discharges in the KTM tokamak. The KTM tokamak physical start-up was carried out in November 2019. The hydrogen plasma was obtained with maximum plasma current of about 100 kA, discharge duration of 65 ms and an average electronic plasma density of about 1.5·1019 m-3. Plasma dis-charges was carried out in the ohmic mode without the use of additional methods of preioni-zation and additional heating. The synopsis describes the work on preparing the KTM toka-mak for the physical start-up, the experiments conditions, and achieved results. The physical start-up has demonstrated the performance of the main KTM tokamak systems and the plasma obtaining possibility.REFERENCES
1 I. Tazhibayeva, E. Kenzhin, V. Shestakov, Y. Chikhray, T. Kulsartov, E. Azizov, O. Filatov, V. Chernov. Material Science Activities for Fusion Reactors in Kazakhstan, J. Nuclear Materials, 2009, p. v.386-388, pp.15-18.
2 I.L. Tazhibayeva, E.G. Batyrbekov, M.K. Skakov, G.V. Shapovalov, D.B. Zarva, T.V. Kulsartov, Y.V. Ponkratov. Fusion activities in republic of Kazakhstan, Reports of IV Inter. Sci. and Tech. Conf. “Innovative Designs and Technologies of Nuclear Power, Sept. 27-30, 2016, Moscow”, NIKIET, 2016, Vol. 2, рp. 208-218.
3 I. L. TAZHIBAYEVA, et al. KTM Experimental Complex Project Status, Fusion Science and Technology, vol.47, April 2005, pp.746 – 750.
4 N. Holtkamp, et al. An overview of the ITER project. Fusion Engineering and De-sign. Volume 82, Issues 5–14, 2007, pp 427-434. https://doi.org/10.1016/j.fusengdes.2007.03.029.
5 D.B. Zarva, А.А. Deriglazov, E.G. Batyrbekov, I.L. Tazhibayeva, V.M. Pavlov, А.М. Li, А.А. Mezentsev, S.V. Merkulov, Yu.N. Golobokov. The Electrotechnical Complex of the KTM Tokamak Pulsed Power Supply System, Problems of Atomic Science and Tech-nology, Ser. Thermonuclear Fusion. 2018, Vol. 41, No. 2, pp. 59–70. DOI: 10.21517/0202-3822-2018-41-2-59-70.
6 D.B. Zarva, А.А. Deriglazov, E.G. Batyrbekov, I.L. Tazhibayeva, V.M. Pavlov, А.М. Li, А.А. Mezentsev, S.V. Merkulov, Yu.N. Golobokov. Emergency protection system of electrotechnical tokamak KTM complex. Algorithmic and information support, Problems of Atomic Science and Technology, Ser. Thermonuclear Fusion. 2019. Vol. 42, No. 1, pp. 74-88. DOI: 10.21517/0202-3822-2019-42-1-74-88.
7 M. Lobanov, L.P. Makarova, A.B. Mineev, V.I. Vasiliev. Plasma initiation stage analysis in tokamaks with TRANSMAK code. Plasma Devices and Operations. Volume 11, Issue 3, 2003.
8 V. Pavlov, et. al. KTM Discharge Modelling in Program Control Mode, The 9th In-ternational conference «Nuclear and Radiation Physics» – Almaty, 2013. – p. 18.
9 K.I. Baystrukov, V.M Pavlov, A.V. Sharnin, A.V. Ovchinnikov, I.L. Tazhibaeva. Control and data acquisition system of Tokamak KTM, AIP Conference Proceedings, 2008, 996, pp. 297-306.
10 M. Skakov, B. Chektybayev, G. Shapovalov, A. Kolodeshnikov. Visible wide angle view imaging system of KTM tokamak based on multielement image fiber bundle, Review of Scientific Instruments.– May 2015.– Vol. 86, Issue 5.– doi: 10.1063/1.4921475.
11 D.W. Swain and G.H. Neilson. An efficient technique for magnetic analysis of noncircular, high-beta tokamak equilibria, Nucl. Fusion, vol. 22, no. 8, pp. 1015–1030, 1982.Speaker: Anuar Sadykov (Institute of atomic energy of National nuclear centre of Republic of Kazakhstan) -
The shear of the radial electric field at the edge is widely accepted to be responsible for turbulence reduction in edge transport barriers and thought as a key ingredient of the improved confinement of H-mode plasmas. Nevertheless, a full understanding of how its profile builds up at the edge is still lacking. It can either be formulated as the result of a competition between several mechanisms that generate and damp flows or as the result of a non-ambipolar particle flux, which enhances radial charge separation imposing a radial electric field profile. Among possible mechanisms, one can think of turbulence generated flow via Reynolds stress, ion orbit losses, toroidal magnetic ripple and the effect of neutral friction at the edge. Based on previous results obtained on Tore Supra, the radial profile of the perpendicular flow expected in WEST can be separated in three spatial areas. Inside $\rho=0.8$, the radial electric field is dominated by losses of thermal ions in the magnetic ripple (1) while between $0.7<\rho<0.95$, a competition between this latter and the generation of large scale flows by turbulence appears as a possible explanation of the measured poloidal asymmetry of the mean perpendicular velocity (2). In addition, edge conditions such as contact points and parallel dynamics in the scrape-off-layer (SOL) influence the edge profiles beyond $\rho=0.9$ (3). This contribution presents unexpected differences in the radial electric field profile observed in WEST between Lower Single Null (LSN) and Upper Single Null (USN) configurations and a study of the competition between specific mechanisms playing a role in the formation of this profile.
WEST is a large aspect ratio machine $(A=5-6)$ with a high level of magnetic ripple around $3\%$ at the plasma edge, symmetric divertors, producing RF heated plasmas (using both Low Hybrid and Ion Cyclotron Resonance Heating systems) with active X-point either at the bottom or at the top and a $B\times\nabla B$ drift always pointing down (as visible in Fig.1). In the context of H-mode scenario development, measurements of the $E \times B$ velocity using Doppler Backscattering system (DBS) in LSN configuration exhibits radial profiles with a well just inside the separatrix. This shape appears similar to standard observations in tokamak plasmas : a positive velocity outside the separatrix, changing sign across the separatrix forming a well in the edge of the confined plasma.
Comparing the radial profiles of the $E \times B$ velocity between LSN and USN plasmas in L-mode regime shows an impressive difference : the radial profile for USN does not contain a well and exhibits a smooth decay from the edge to the core (see Fig.1). This tendency is consistent with the fact that the configuration with the $B\times\nabla B$ drift pointing toward the active X-point, designated as the "favorable configuration", is more favorable to reach the H-mode as compared to the opposite case when $B\times\nabla B$ drift points away from the active X-point. Nevertheless, there are multiple proposed explanations without a clear consensus on the reason of such favorable versus unfavorable configuration and systematic experimental comparisons of velocity profiles in USN and LSN are scarce. Regarding the results from the Fig.1, the WEST tokamak emerged as a particularly interesting machine to study this difference since it appears exacerbated between these two configurations.
In this spirit, experiments have been performed to compare the DBS velocity profiles at the edge in both USN and LSN in different plasma conditions. Among the results, it is found that the plasma current impacts significantly the velocity profile in the USN discharges (Fig.2). At low current, the $V_{E\times B}$ profile does not exhibit a well but a slow decay leading to the striking difference with the LSN discharges. When increasing the plasma current, the $E \times B$ velocity starts to form a well to end up with a deeper profile than in LSN at high current. Indeed, in LSN configuration, the increase of the plasma current also deepens the radial electric field well ; however, the effect is less pronounced than in USN and the profile shape remains similar. Interestingly, for fixed plasma current and moderate to high heating power, the velocity profile ends up deeper in the USN configuration for same plasma conditions.
This result leads to an unexpected situation in which the "unfavorable configuration" seems more favorable, at least from the point of view of turbulence shearing. Conjointly, the dynamics of the density fluctuations measured with DBS is clearly affected in the inner branch of the velocity well in the case of USN while not in the case of LSN discharge.
The influence of the edge topology on the radial electric field has been addressed numerically using the gyrokinetic GYSELA code in which a limiter, modelled through a penalization technique, is placed at the bottom or at the top of a circular plasma (4). Figure 3 shows the resulting $V_{E \times B}$ velocity obtained in such simulations. The radial profile is found in qualitative agreement with the experiments, with a negative velocity in the confined region about few km/s. The presence of the limiters is associated with the formation of a well in the profile near the separatrix. In addition, the depth of the well is found to be sensitive to the edge configuration and limiter position as observed in both Tore Supra and WEST tokamaks. The shear layer appears related to both ion orbit loss through the separatrix and vortex-flow interaction. The difference between both configurations is studied regarding the sign of the Reynolds stress terms coming from magnetic shear and radial electric shear tilting.
In summary, in WEST, the shape of the radial electric field profile is found to strongly depend on the magnetic configuration. Changing the plasma conditions, such as plasma current, density, heating power, appears to modify the balance between mechanisms playing a role in the formation of the radial electric field. In USN, the profile evolves from an "unfavourable" shape, without any well, to a strongly sheared shape when increasing the plasma current or the heating power while in LSN the velocity profile appears less sensitive to such plasma conditions. The USN/LSN difference is captured by first-principles simulations including core-edge coupling with neoclassical and turbulent physics.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom esearch and training programme 2014 - 2018 and 2019 - 2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
(1) E. Trier et al. Radial electric field measurement in a tokamak with magnetic field ripple. Nuclear Fusion, 48(9), 2008.
(2) L. Vermare et al. Poloidal asymmetries of flows in the Tore Supra tokamak. Physics of Plasmas, 25(2):020704, 2018.
(3) P. Hennequin et al. The effect of SOL flows on edge and core radial electric field and rotation in Tore Supra. 37th EPS Conference on Plasma Physics, Dublin, 34A, P1.1040, 2010.
(4) E. Caschera et al. Immersed boundary conditions in global, flux-driven, gyrokinetic simulations. Journal of Physics: Conference Series,1125:012006, 11 2018.Speaker: Laure Vermare (CNRS, LPP Ecole Polytechnique) -
Abstract:
During the operation of the fusion reactor, the tritium breeder pebbles packed into the tritium breeding blanket formed a tritium breeder pebble bed. Under the influence of the severe environment such as irradiation swelling, thermal expansion, alternating stress, and so forth, the tritium breeder pebbles will be broken and pulverized, accompanied by changes in the thermomechanical properties and packing structures of the tritium breeder pebble bed. With the increase of the operation time, the fragmentation behaviors will become more and more serious, which may affect the tritium breeding efficiency and the heat transfer performance of the tritium breeding blanket. Therefore, in this work, the packing structures, fragmentation behaviors and mechanical properties of tritium breeder pebble bed were investigated utilizing experiments and numerical simulations, as well as the flow characteristics of the purge gas in the pebble bed.
The packing behaviors of tritium breeder pebble beds and neutron multiplier pebble beds are important to estimate the thermal properties of pebble bed, the thermal-mechanical of pebble bed, the flow characteristics of purge gas and the tritium breeding ratios of the blanket. In this work, the packing experiments and the DEM simulations were conducted to investigate the packing behaviors and mechanical properties of the mono-sized and binary-sized pebble beds, respectively. Owing to the tritium breeding zone of the typical solid breeder blanket can be simplified to prismatic containers. Thus, we investigated the widths' influences (Lx/d=5~60) on the average packing factors and porosity distributions in pebble bed with and without vibration. The results obtained in this study show that the packing factors and the porosity can be significantly influenced by walls and widths as shown in Fig.1. The average packing factors increase with the ratio of bed widths to pebble diameter and the vibration process can increase the average packing factor. In the area near the wall of the pebble bed, the porosity gradually oscillates and damping with the increasing distance from the wall. In further, the effect of pebble size distributions on the packing structures of pebble bed was explored experimentally and numerically. By optimizing the pebble size component, a higher packing factor can be obtained by using binary-sized pebbles and polydisperse pebbles. The maximum packing efficiency state appear at the volume fraction 70% of larger pebbles.
Besides, the mechanical properties of compressed pebble bed and the fragmentation characteristics of ceramic pebbles inside the pebble bed were investigated by the DEM simulation and the X-ray computed tomography (x-ray CT). The fragmentation characteristics of pebbles inside pebble bed under different compression loads were investigated experimentally. The Schematic of the pebble crushing experiment based on the x-ray CT and the Uniaxial compression test (UCT) as shown in Fig.3. The results show that the breakage rate of the tritium breeder gradually increased with the increase of the compressive load. The x-ray CT can characterize the fracture characteristics of pebbles inside the pebble bed.
Moreover, the packing behaviors of pebble beds under vibration were investigated experimentally. The influences of the vibration amplitude and frequency on the packing fraction were analyzed to optimize the pebble packing techniques for the helium-cooled ceramic breeder blanket. In addition, the flow characteristics of the fluid distribution and the pressure drop along the pebble bed were analyzed by the DEM-CFD method. The effects of the bed lengths, the diameter ratios of the cylinder to the pebbles, and the contact point treatment between the pebbles on the velocity distribution and pressure drop of the purge gas were investigated. The results in this work will provide some technical support for the optimal design of the solid tritium breeder blanket.
Keywords: Packing behaviors; Fragmentation behaviors; Mechanical properties; solid tritium breeder blanket; Discrete element method; X-ray computed tomography
Acknowledgments
This work is financially supported by the National Key Research and Development Program of China under Grant No. 2017YFE0300602, the Sichuan Science and Technology Program from Science & Technology Department of Sichuan Province of China under Grant No. 2018JZ0014, and National Natural Science Foundation of China under Grant No. 11905047.
Reference
1. A. Ying a, et al., Status of ceramic breeder pebble bed thermo-mechanics R&D and impact on breeder material mechanical strength, Fusion Eng. Des. 2012, 87:1130–1137
2. X.Y. Wang, et al., Current design and R&D progress of the Chinese helium cooled ceramic breeder test blanket system, Nucl. Fusion, 2019, 59:076019
3. J. Reimann, et al., X-ray tomography investigations of mono-sized sphere packing structures in cylindrical containers, Powder Technol. 2017, 318:471-483
[4]. J. Reimann, et al., 3d tomography analysis of the packing structure of spherical particles in slender prismatic containers, Int. J. Mater. Res., 2020, 111(1):65-77
[5]. B.P. Gong, et al., Discrete element modelling of pebble bed packing structures for HCCB TBM, Fusion Eng. Des., 2017, 121:256-264
[6]. B.P. Gong, et al., Numerical investigation of the pebble bed structures for HCCB TBM, Fusion Eng. Des., 2018, 136 1444-1451
[7]. R. K. Desu, et al., High temperature oedometric compression of Li 2TiO3 pebble beds for Indian TBM, Fusion Engineering and Design, 2018, 136:945–949
[8]. M. Lotfy, et al, Study on the Thermally-Induced Stress and Relaxation of Ceramic Breeder Pebble Beds, Fusion Sci. Technol. 2017, 72:3, 255-262Speaker: Mr Baoping Gong (Southwestern Institute of Physics) -
Technology Readiness Levels (TRLs) give a good idea of maturity of new technology and are used by industry and government organisations worldwide. Nine TRLs exist ranging from initial ideas at Level 1 where basic principles are observed to fully robust technologies validated for application in industry at Level 9. TRLs can be further subdivided into Readiness Levels for system, materials, methods, manufacturing and instrumentation as shown in Figure 1. The main TRLs were originally devised by NASA and also used by ESA for aerospace industry. DOE 1 in USA produced a guide on TRL assessment for wider application. EU H2020 2 and NDA (UK) [3] also produced TRL guides out which the one defined by the NDA(UK) is perhaps the most relevant to the nuclear industry but it was specifically written for nuclear decommissioning sector. This paper presents a case to develop definitions of fusion specific TRLs.
The 1 to 9 TRL scale is an ordinal scale. The effort or time needed to move from one point to another may not be linear. TRLs are time specific and most importantly context specific. TRLs should not be used to measure progress, risk, cost or duration. Technologies with low TRL can mature more quickly and technologies with high TRL can stagnate and never mature as is shown in Figure 2. TRLs of individual items say nothing about whether the whole system will work together. The integration and interfaces need to be assessed separately. TRLs are for individual technologies not for a system requiring integration of many. There will be a need to breakdown a complex system into sufficient number of sub-systems to apply TRLs. Furthermore, a technology which is mature in one sector may not be mature for fusion application and vice versa. All this has to be analysed in the case of Fusion technologies
Benefits of TRLs are that they can help decide whether a technology is ready for implementation and can help plan its further development. For cost-effective development of a technology, a better definition of TRLs is needed as it will help focus the resources. The example shown in Figure 1, indicates the target TRLs to be achieved by a project and a stage gate review indicates the current status which highlights the areas where development effort needs to be invested. A good TRL definition and assessment will help produce a cost-effective development plan, and it not available for fusion so far.
The need for fusion specific TRLs arises from the fact that technology requirements in the nuclear sector are different from other industry sectors. Furthermore, there are some notable differences between the fission and fusion. The IAEA Tecdoc 1851 [4] highlights these differences. For example, in the fusion applications, there
is no reactivity control or emergency cooling requirements and no core melt conditions to be addressed. Prevention of core meltdown is not a safety related function for fusion reactors. Most of the main parameters like the safety functions, consequences of failure, fault frequencies, contributing to the safety classification process in fission applications are different for the fusion applications. Yet, the fusion industry has mostly relied on existing codes, standards and industrial practices developed for fission. Fusion technology challenges are also different if not more demanding. In the fusion nuclear facilities there are many kinds of accidents that can be postulated due the fact that a tokamak is a complex and dense zone with many different energy source terms. In addition, the first confinement barrier, which is surrounded by those energy source terms, has a very complex boundary and is subjected to a combination of extreme loads not seen in other industry sectors. E Surrey et al [5] have acknowledged inadequacy of the existing TRLs for fusion and proposed an alternative methodology that allows a quasi-numerical analysis by a combination of three quantities: unmitigated probability of failure, severity and probability of failure detection.To help fusion technology achieve full industrialisation, it is prudent to consider internationally harmonised definitions for fusion specific TRLs and provide fusion specific definitions for all the 9 levels of TRL for system, materials, methods, manufacturing and instrumentation.
The new definitions can also take into account new approaches to design and validate a system, structure or component. For example, risk informed performance based probabilistic design methods can help produce cost-effective designs without compromising safety. The current design approach is to use allowable stress codes to design components that rely on experience-based safety factors which gives no idea of reliability or risk in design of the full system. The new approach can produce cost-effective design and give a good idea of the risk and probability of failure. This new approach is going to have wider scope and will be applicable to next generation of reactors.
References:
- U.S. Department of Energy, Technology Readiness Assessment Guide, DOE 413.3-4, 2009.
- Technology Readiness Levels (TRL), European Commission, HORIZON 2020, Work Programme 2014-2015, General Annexes, Extract from Part 19 – Commission Decision C (2014) 4995.
- Guide to Technology Readiness Levels for the NDA Estate and its Supply Chain, Nuclear Decommissioning Authority (NDA), 2014.
- Integrated Approach to Safety Classification of Mechanical Components for Fusion Applications, IAEA Tecdoc 1851.
- Assessing component suitability and optimizing fusion plant design – Alternative approaches to TRLs, E Surrey, J. Linton and M. Sadler, TPS 10398.
Speaker: Prof. Nawal Prinja (Technology Director, Wood Group (Nuclear)) -
We carried out global calculations of neoclassical (NC) impurity transport in helical plasmas including the effect of variation of electrostatic potential on the flux surface, $\Phi_1$, for the first time. An impurity hole plasma found in LHD experiment is analysed by the global simulation. Contrary to the conventional local approximation models, which indicate negative ambipolar radial electric field ($E_r$) for the ambipolar condition, the global NC simulation predicts positive ambipolar-$E_r$ and impurity particle flux is driven outwardly, which is even enhanced by $\Phi_1$. If $E_r$ is negative in the global simulation, on the other hand, $\Phi_1$ enhances the inward impurity flux. Our result indicates that the global effect is the key to correctly predict the $E_r$ profile and impurity transport in impurity hole plasmas.
In LHD, the formation of extremely hollow density profiles of impurity ions, such as carbon C$^{6+}$, in the core region of high $T_i$ plasmas has been observed. This phenomenon is called ``impurity hole''. Local NC simulations, which neglect radial drift in the guiding-centre equations of motion, find ion-roots (negative $E_r$) as the ambipolar conditions for such plasmas. These results indicate that highly charged impurities are expected to be driven inwardly, which are contradicting to the observation. Gyrokinetic simulation$^1$ also predicts that the impurity particle flux is inwardly directed in the impurity hole plasma. In recent studies, it has been shown that the $\Phi_1$ potential, which had commonly been neglected, has significant impact on NC impurity transport$^{2-4}$. Yet, all of those calculations have been performed with local simulation codes, and the necessity of global calculation has been suggested$^{3,5}$. We thus have extended a global NC code FORTEC-3D to include $\Phi_1$.
Besides the studies of the $\Phi_1$-effect, we find an electron-root of the ambipolar condition for the impurity hole plasma by the global simulation. In the high-$T_i$ and low-collisionality regions in which impurity hole is usually found, $E_r$ is relatively weak and the radially local approximation tends to be invalidated by the global effects. Thus, we have investigated the structure of $\Phi_1$ and its impact on particle transport for two different cases, one with an ion-root (local solution) and the other with an electron-root (global solution) $E_r$, respectively (see Figure 1).
In Figure 2, the radial carbon particle fluxes for the cases with and without $\Phi_1$ are compared. For the ion-root case, the C$^{6+}$ flux is driven further inwardly by $\Phi_1$. This result is analogous to the local results$^{3,4}$. In contrast, the radial flux is outwardly enhanced by $\Phi_1$ for the electron-root case. The reason behind the contrast between those cases can be found in the phase structure of carbon density variation. In Figure 3, the phase structure of $\Phi_1$ and the density variation of carbon, $n_{C1}$, on the flux surface at $r/a=0.5$ are shown, where $r$ is the radial coordinate and $a$ is the minor radius of the plasma. Comparing $\Phi_1$ and $n_{C1}$ between positive and negative $E_r$ cases, respectively, it can be seen that the phase structures of $\Phi_1 $ are not largely different but the structures of $n_{C1}$ are inverted in $\theta$-direction between the two cases. The sign of the radial particle flux driven by $\Phi_1 $ is determined by the product of the Fourier spectra of $n_{C1}$ and the radial $E\times B$-drift generated by $\Phi_1$. The phase inversion of $n_{C1} $ in $\theta$-direction reflects the sign inversion of the leading modes in the $n_{C1}$ spectrum. Thus, in order to avoid the negative enhancement of the flux by $\Phi_1$ for the ion-root case, either the phase structure of $\Phi_1$ or $n_{C1}$ has to be strongly modified, but our analysis implies that such radial modifications are unlikely to occur unless strong particle or torque source exists. Thus, our conclusion is that it is difficult for $\Phi_1$ to mitigate the inward C$^{6+}$ flux in ion-root plasmas.
Although impurity hole has been observed in several different shots, a measurement on the core $E_r$ profile has been reported only for a single case$^6$. Plasma profiles used in our study are close to but not the same as those in the measured case. The experimental study shows that while the sign of $E_r$ is indeed negative near the magnetic axis, $E_r$ transits to positive at some outer radius $r/a \sim 0.5$. Our code can be applied to such a case in which $E_r$ changes its sign, which is another advantage over the local codes. The sign of $E_r$, which is the key to predict the impurity flux, can change in the conditions which are not included in the present simulations. The result of this study suggests the necessity of extension of our code for self-consistent calculation of multi-ion-species transport and the ambipolar condition including $\Phi_1$.
- M. Nunami, et al., In 22nd ISHW (2019).
- J. M. García-Regaña, et al., Nucl. Fusion 57 (2017) 056004.
- J. L. Velasco, et al., Plasma Phys. Control. Fusion 60 (2018) 074004.
- A. Mollén, et al., Plasma Phys. Control. Fusion 60(2018) 084001.
- K. Fujita, et al., Plasma and Fusion Research 14 (2019) 3403102.
- S. Sudo, et al., Plasma Phys. Control. Fusion 58 (2016) 043001.
Speaker: Keiji Fujita (The Graduate University for Advanced Studies) -
The possibility to exploit the properties of medium and light impurities in contaminating plasmas is explored in many tokamaks devices [2-5]; since is crucial to fix the conditions that favor a strong increase of particle confinement while minimizing the amount of impurities needed, as well as to favor the so called "plasma detachment". The light doping action represents a good method to increase the core electron density, often without any undesirable central impurity accumulation; meanwhile the amount of impurities, and the relative edge radiation, have to be kept below the threshold where a disruptive MHD activity is generated.
In order to complete experimental observations regarding the electron density peaking in doped plasma referred in FTU [4], a series of experimental sessions have been performed in the last two campaigns by injecting Helium on the L-mode plasma scenarios. As first results reported in this work, it has been revealed that not only the total amount of inflated Helium, but also speed of injection, as well as edge conditioning, can influence the impurity effects. Examples are exposed in the figure 1.
VUV spectroscopy measurements help to evaluate the Helium presence, that triggers the particles inflow; indeed, only a small fraction of electrons could be attributed to the ones stripped by Helium ionization in the total contribution of the electron density rise. In the same phase, the attempt to interpret the radiation losses and effective charge traces, in respect to concurrent different impurities, is treated too. Further observations regard the behavior of the temperature profiles, also in comparison with previous works in which Neon impurity was used [4,6]. Generally, these plasmas, as consequence of the Helium injection, frequently excite some instabilities, such as MARFEs and MHD, that affect measurements and provoke strong perturbations, in some cases they appear simultaneously.
Considerable results have been achieved in the search of the best conditions to obtain a very high electron density peaking, by reaching the value of 5, all are detailed in this document.
A JETTO analysis adds interesting considerations about particle transport and improvement in confinement. Furthermore, a set of different scenarios has been experimentally performed. One of them, consisting in a scan of plasma current (principal parameters: $I_p$=250-500 kA, $B_T$=5.3 T, $n_{e_0}$=0.2-1 $10^{20}$ m$^{-3}$, $T_{e0}$=1-4 keV), has confirmed that the highest peaking is reached at low current. Another one explores the relationship between the increase of the electron density and the plasma position inside of the wall chamber, by varying plasma shape and magnetic configuration.
References
[2] A. Messiaen et al 1996 Phys. Rev. Lett. 77 2487
[3] G. Telesca et al 2000 Nucl. Fusion 40 1845
[4] C. Mazzotta et al 2015 Nucl. Fusion 55 073027
[5] K. A. Razumova et al 2017 Plasma Phys. Rep. 43 1043–1051
[6] N. A. Kirneva et al Effect of the impurity injection on plasma confinement in T-10 tokamak 44th European Physical Society Conference on Plasma Physics, June 2017, Belfast, United Kingdom. P4.169Speaker: Dr Cristina Mazzotta (ENEA, Fusion and Nuclear Safety Department, C. R. Frascati, via E. Fermi 45, 00044 Frascati, Roma, Italy) -
High-Temperature Superconducting (HTS) magnet is considered to be applied to the next-generation helical experimental device. Three types of large-current (6-18 kA) HTS conductors are being developed. One of the crucial requirements is to secure the high current density of 80 A/mm2 at a condition of 20 K temperature and 10 T magnetic field. In the first phase of the development, short samples of each conductor are being fabricated and tested in liquid nitrogen at 77 K with no external magnetic field. The observed critical current has been improving by optimizing the fabrication method, showing an extrapolation to satisfy the magnet requirement. Samples of the three conductors are also being fabricated to be tested at ~20 K and > 7 T. Quench protection is another crucial requirement, and the hot-spot temperature is examined with different protection scenarios for the three conductor options.
At National Institute for Fusion Science (NIFS), following the successful plasma confinement experiments conducted in the Large Helical Device (LHD), the post-LHD project is being discussed. One of its two candidates is to build a new LHD-similar but more optimized heliotron device having high magnetic field equipped with a HTS magnet system. Three candidate conductors named STARS (Stacked Tapes Assembled in Rigid Structure), FAIR (Friction Stir Welding (FSW), Aluminum alloy jacket, Indirect cooling, REBCO), and WISE (Wound and Impregnated Stacked Elastic tapes) are presently designed and developed.
The STARS conductor has been originally developed for the purpose of applying to the future helical fusion reactor FFHR-d1 [A, B]. In this conductor, REBCO HTS tapes are simply stacked in a copper stabilizer and stainless-steel (SS) jacket. A prototype sample (with SS jacket sustained by bolts) had achieved 100 kA at 20 K and 5.3 T [C]. A scaled-down conductor with a 18-kA operation current (Fig. 1(a)) is presently designed by applying laser beam welding (LBW) to the SS jacket. A preliminary testing of the laser beam welding has confirmed that the maximum temperature rise during the welding was ~40 degrees centigrade, which was well below the requirement of 200 degrees centigrade that guarantees no degradation of REBCO tapes. By having a flattened structure of the conductor cross-section, the helical winding property is supposed to be not very different from that of the NbTi/Cu conductor used for the present LHD, despite the enhanced mechanical strength using SS. The edgewise strain can be minimized to be almost zero by adjusting the inclination angle of the conductor in the winding package [D]. A 3-m-long straight conductor sample is being fabricated, which will be tested in March 2020 at 77 K and 0 T.
The FAIR conductor (Fig. 1(b)) has a stack of REBCO tapes imbedded in a circular aluminum-alloy jacket which is sealed by FSW [E]. A slight twisting (2 rotations per meter) is included to have uniform current distribution among REBCO tapes. A number of 1-m-long prototype samples have been fabricated and tested at 77 K, 0 T. Though some degradation was initially observed with the measured critical current by cyclic excitations, steady improvement is foreseen by optimizing the fabrication process, as is shown in Fig. 2.
The WISE conductor (Fig. 1(c)) is formed by inserting a stack of REBCO tapes into a flexible metal tube, and the winding package is impregnated with low-melting point metal alloy U-Alloy78 (melting point: ~78.8 degrees centigrade) [F]. A small solenoid coil was fabricated and 0.16 T was achieved with 800 A conductor current in liquid nitrogen. Although normal-transitions were observed from some tapes, the coil was stably excited. The reason is being investigated by testing short samples. A small helical coil was also fabricated by winding the WISE conductor into a helical coil-can which was fabricated by additive manufacturing.
The major specifications of the three conductors are summarized in TABLE 1. For all these three conductors, after having completed the present tests in liquid nitrogen and no external magnetic field, short samples will then be tested in the superconductor testing facilities under high magnetic field (>7 T) and cryogenic temperature (4-50 K). Resultantly, some suitable conductors will be selected to go into the next phase of long-conductor development also by manufacturing model coils.
The quench protection is another crucial issue for the HTS magnet design. The STARS conductor employs the conventional dump resistor method with a short discharging time constant of <3 s to limit the hot-spot temperature to <200 K. Since the aluminum-alloy in the FAIR conductor possesses less heat capacity, a secondary winding is imbedded in the conductor to be used as a heater for enhancing normal-zone propagation. The feasibility of this option is being examined through numerical simulations. The WISE conductor is wound as a Non-Insulation (NI) coil, which is supposed to have self-protection capability. The feasibility of this option is also being examined through numerical simulations. The fabrication method of the helical coils is also examined. In addition to the continuous winding technique employed for the present LHD, the “joint-winding” method associated with a bridge-type mechanical lap joint technique has been developed for the STARS conductor [G].References
[A] A. Sagara et al., Fusion Eng. Des. 89 (2014) 2114.
[B] N. Yanagi et al., Journal of Fusion Energy 38 (2019) 147.
[C] Y. Terazaki et al., IEEE Transactions on Applied Superconductivity 25 (2015) 4201205.
[D] Y. Narushima, et al., submitted to Plasma and Fusion Research.
[E] T. Mito et al., to be published in Journal of Physics Communications.
[F] S. Matsunaga et al., to be published in IEEE Transactions on Applied Superconductivity.
[G] S. Ito et al., Fusion Engineering and Design 136 (2018) 239.Speaker: Prof. Nagato Yanagi (National Institute for Fusion Science) -
The aspect ratio $A = R/a$ is a central parameter for tokamak design in many aspects. In particular, it is expected to critically impact plasma vertical stability and heating, disruption forces, tritium breeding, maintenance and cost [1]. Yet, there is some freedom in its choice. Most of present tokamaks, including ITER, operate at $A\sim3$, with notable exceptions like MAST ($A\sim1.5$) and WEST ($A\sim5$) operated at CEA Cadarache. This paper reports on three major new results regarding the impact of $A$ on confinement: (1) how does it translate in terms of tokamak design? (2) What do we learn from global flux-driven gyrokinetic simulations? (3) What is the current input of WEST data in this debate?
(1) On the basis of empirical scaling laws (SL) of the energy confinement time $\tau_E$, the benefit of operating at large or small $A$ is uncertain. Indeed, two of the most refined SL using large tokamak data bases, IPB98(y,2) [2] and DS03 [3] – valid for ELMy H-mode regimes – exhibit roughly the same variance ($16\%$) of experimental data. However, when expressed in dimensionless variables, the scaling exponent of $A$ is of opposite sign in both SL: DS03 predicts an increase of $\tau_E$ with A ($\omega_c \tau_E \sim A^{1.3}$, with $\omega_c$ the ion cyclotron frequency), the opposite trend is observed for IPB98(y,2) ($\omega_c \tau_E \sim A^{-0.73}$). This uncertainty reflects the lack of data from tokamaks with sufficiently different A values in the database. Forthcoming data from the WEST tokamak will likely help alleviating the uncertainty [4]. Furthermore, there likely exists a bias for compact tokamaks which usually operate at larger normalized beta $\beta_N$ values than more conventional ones, so that $A$ and $\beta_N$ may not be completely independent parameters in the database. We have explored the sensitivity of the European DEMO design [1] $-$ targeting the amplification factor $Q = 40$ and the fusion power $P_{fus} = 2.037$ GW $-$ with respect to $A$ when considering the two equally satisfying SL. Results are plotted on Fig.1 in terms of major radius $R$ and magnetic field $B$.
The different values of $\beta_N$ account for the different operational spaces for these two SL: DS03 predicts larger $\beta_N$ (and lower plasma current) to achieve similar performance, which can reveal challenging in a reactor aiming at zero disruption [5]. Importantly, the variations with $A$ of $(R,B)$ solutions are opposite for the two scaling laws: $R$ decreases (resp. increases) with $A$ if IPB98(y,2) (resp. DS03) holds. The opposite is true for $B$. Especially, DS03 predicts that similar performance can be achieved at lower $R$ (and slightly larger $B$) when reducing $A$.
(2) The issue of confinement and aspect ratio has also been addressed by means of flux-driven gyrokinetic simulations with GYSELA [6] in the ion temperature gradient driven regime of turbulence with adiabatic electrons. On the basis of first expected then observed SL of $\tau_E$, the heating source was tuned to keep the profiles close to their initial state. A database of about 40 close-to-steady-state simulations has been obtained, covering the parameter range $A=[3.2, 4.4, 6]$ and the collisionality $\nu_*=[0.004, 0.02, 0.1]$. Also, the normalized ion gyro-radius $1/\rho_*=[150, 190, 250, 380]$ for $A=4.4$ and $1/\rho_*=[190, 250, 380]$ for $A=6$. Using standard regression [7], the dimensionless energy confinement time is found to scale like (Fig.2) $\omega_c \tau_E = 1.5 A^{0.88} \rho_*^{-2.40} \nu_*^{-0.14}$.
The positive exponent of $A$, as for DS03, may suggest $\tau_E$ scales like the parallel transit time $\omega_c \tau_\parallel \sim ~ qA/\rho_*$ ($q$=safety factor). Also, increasing $A$ tends to reduce curvature and grad-$B$ drifts, which are the generic drives for ITG. Note finally that the scaling exponent of $A$ is not expected to change sign in regimes dominated by trapped electron modes since the trapped particle density scales like $A^{-1/2}$. This possible multiple dependency may actually result in a non-monomial SL with respect to $A$.
(3) WEST data have been collected in 2019 during the 4th experimental campaign. All considered shots are in L-mode, either purely ohmic $P_\Omega$ or with ion cyclotron resonance $P_{IC}$ and lower hybrid $P_{LH}$ heating. Radiative losses are in the range $45\%$ to $65\%$ of the input power $P_{in} = P_\Omega + P_{IC} + P_{LH}$ likely due to the increase of the impurity source with power. The energy confinement time has been estimated with 0.3s time window averages of the MHD energy divided by the heating power. We find the following SL: $\tau_E \sim I_p^{0.86} P_{ef\!f}^{-0.84}$, with $I_p$ the plasma current. The scaling exponents turn out to be relatively insensitive to the considered heating power, either the input or the effective $P_{ef\!f} = P_{in} - P_{rad}$ power, although the scattering is lower with $P_{ef\!f}$. Note that this SL is not dimensionally correct, in part due to the narrow parameter space. Yet, although operating at much larger aspect ratio than the tokamaks considered to construct the experimental database in L-mode, it exhibits similar scaling exponents with respect to both plasma current and heating power as the standard L-mode SL, respectively equal to 0.96 and $-0.73$ [2]. As a result, $\tau_E$ in WEST is in good agreement with the value predicted by the L-mode SL (Fig.3).
The agreement of WEST confinement with the standard L-mode SL, which predicts almost no dependency in $A$, would suggest that this law can indeed be extrapolated to large $A$ tokamaks. GYSELA results seem to qualitatively agree with DS03 SL in H-mode, favoring large $A$ machines. These are potentially good news for WEST, which would then exhibit even better performance in H-mode than initially expected.
References: [1] R. Wenninger et al., Nucl. Fusion 57 (2017) 016011; [2] ITER Physics Basis, Nucl. Fusion 39 (1999) 2175; [3] Sips A.C.C. et al., Nucl. Fusion 58 (2018) 126010; [4] Bucalossi J. et al., Fusion Eng. Des. 86 (2011) 684; [5] Y. Sarazin et al., Nucl. Fusion 60 (2020) 016010; [6] V. Grandgirard et al., Comput. Phys. Commun. 207 (2016) 35; [7] E. Caschera, PhD thesis, Aix-Marseille University ED352, France (2019)
Speaker: Yanick SARAZIN (CEA, IRFM) -
Whereas it is widely believed that velocity shear could suppress plasma instabilities and stimulate the transition from low (L-) to high (H-) confinement modes, the underlying physics of plasma instability suppression is still not clear. Often it is assumed that the stabilization of plasma instability characterized by the growth rate $\gamma _{inst} $ occurs when by the velocity shear $| V'_{0}|$ exceeds $\gamma _{inst} $ (e.g. see Refs. 1, 2). One of the complications of the analysis of the velocity shear effect on plasma instabilities is the non-Hermitian nature of the differential equations describing an impact of velocity shear on plasma/fluid instabilities [3]. However, we find that the situation is more complex and just effective Richardson number $Ri=(\gamma _{inst} /| V'_{0} |)^{2} $ cannot describe overall impact of $|V'_{0} |$. Employing
radial'' density profile $n(x)=\bar{n}+(\delta n/2)tanh(x/w$) (where $w$ is the effective width of the density profile and $\bar{n}\gg\delta n$ are some constants) and analyzing the localized modes, we find [4] that for $\kappa =| k_{y} w|\gg 1$ ($k_{y} $ is thepoloidal'' wave number) the growth rates of both fluid Rayleigh-Taylor (RT) and plasma interchange (I) modes could be significantly reduced even for $Ri\gg 1$ (see Fig. 1a).
On the contrary, the resistive drift waves (RDW) are not stabilized even for $Ri\ll 1$ (see Fig. 1b, where $\hat{\omega }_{*} =k_{y} \rho _{s} C_{s} /L_{n} $, $\nu _{\parallel} /\hat{\omega }_{*} =50$, $\nu _{\parallel} =k_{z}^{2} T_{e} /m\nu _{ei} $ is the effective parallel electron diffusion frequency, and $w/\rho _{s} =30$). However, the localized RDW modes cease to exist at $| V'_{0} | >| V'_{0} |_{loc} \approx 0.66(1+k_{y}^{2} \rho _{s}^{2} )^{-1} (\rho _{s} /w)C_{s} /L_{n} $. In addition, we find that, whereas the eddies of both RT and I modes in the presence of $| V'_{0} |$ become tilted into y-direction, Fig. 2a, those of the RDW become just shifted into radial direction, Fig. 2b, The results of numerical analysis of non-modal solutions of the RDW for $| V'_{0} |>|V'_{0} |_{loc} $ will be presented.
Unlike the effect of velocity shear, the results of the studies of an impact of neutrals on edge plasma instabilities and turbulence are somewhat controversial. Whereas some experiments show no effect of neutrals on edge plasma turbulence, others demonstrate an importance of neutrals for L- to H-mode transition (e.g. see Refs. 5, 6). Similarly, whereas some simulations show that neutrals result in increasing edge plasma turbulence, some others claim opposite effect (e.g. see Refs. 7, 8). One of the complexities of the incorporation of neutral effects into plasma instabilities, turbulence, and transport is the wide range of neutral-plasma interaction regimes (from kinetic to fluid) defined by the ratio of the wave length (frequency) of different plasma modes to neutral-ion collision mean free path (neutral-ion collision frequency).
We report here the results of a careful analysis of the effect of neutrals (ranging from kinetic to fluid transport regimes) on interchange, RDW, and grad($T_{e}$ ) instabilities [9] and find that in practice neutral make a very minor impact on these instabilities, although in dense divertor plasma an impact of neutrals on plasma stability could be important (see Ref. 10 and the refernces therein). However, we find that neutrals can significantly alter the generation of zonal flow by plasma turbulence (e.g. by DW turbulence [11]) and by that modify edge plasma turbulence and transport.[1] W. Horton, ``Turbulent transport in magnetized plasmas'' (World Scientific, Second Edition, 2018); [2] J. Kinsey, R. Waltz, and J. Candy, Phys. Plasmas 12 (2005) 062302; [3] L. N. Trefethen, et al., Science 12 (1993) 578; [4] Y. Zhang, S. I. Krasheninnikov, and A. I. Smolyakov, Phys. Plasmas, 27 (2020) 020701; [5] M. A. Pedrosa, et al., Phys. Plasmas 2 (1995) 2618; [6] D. J. Battaglia, et al., Nucl. Fusion 53 (2013) 113032; [7] D. P. Stotler, et al., Nucl. Fusion 57 (2017) 086028; [8] N. Bisai, R. Jha, and P. K. Kaw et al., Phys. Plasmas 22 (2015) 022517; [9] Y. Zhang, S. I. Krasheninnikov, submitted to Phys. Plasmas, 2020; [10] A. Odblom, et al., Phys. Plasmas 6 (1999) 3239; [11] A. I. Smolyakov, et al., Phys. Rev. Lett. 84 (2001) 491.
Speaker: Prof. Sergei Krasheninnikov (University California San Diego) -
Data assimilation techniques are applied to the integrated transport simulation (TASK3D[A]) in Large Helical Device (LHD). We use the ensemble Kalman filter (EnKF) and the ensemble Kalman smoother (EnKS) as data assimilation methods for the estimation of state variables. The time series data of experimentally measured temperature and density profiles are assimilated into TASK3D[B]. The obtained electron and ion temperature profiles and temporal variations agree well with measured ones owing to the optimization of the employed turbulent transport model and the heat deposition. These results indicate the effectiveness and validity of the data assimilation approach for accurate prediction of the behavior of fusion plasmas and the possibility of advanced turbulence modeling.
Integrated simulations for fusion plasmas have various uncertainties in each of the employed simulation models, specifically turbulent transport models. Because of these model uncertainties, the simulation results are not totally confident to predict the behavior of fusion plasma. To solve this problem and to predict the behavior of fusion plasmas with high accuracy in real-time, data assimilation techniques are applied to the integrated transport simulation by TASK3D for a plasma in LHD. As data assimilation methods, we use the EnKF for the prediction and the EnKS for the estimation of the models which can reproduce experimental data. The EnKS optimizes the state variables using both past and future data, while the EnKF using only past data.
For the LHD plasma, the gyro-Bohm model for electron turbulent transport coefficient, $\chi_{\rm e}^{\rm TB}$ and the gyro-Bohm gradT model for ion one, $\chi_{\rm i}^{\rm TB}$ are employed based on previous TASK3D simulations:
electron : $\chi^{\rm TB}_{\rm e}=C_{\rm e}\left(\frac{\rho_i}{a}\right)\left(\frac{T_{\rm e}}{eB}\right)$, ion : $\chi_{\rm i}^{\rm TB}=C_{\rm i}\left(\frac{\rho_i}{a}\right)\left(\frac{T_{\rm i}}{eB}\right)\left(\frac{T_{\rm i}'}{T}_{\rm i}a\right)$,
where $B$ is the magnetic field strength, $\rho_i$ is the ion Larmor radius, $a$ is the minor radius, and $C_{\rm e}$, $C_{\rm i}$ were constant factors. The state vector is composed of the electron and ion temperature ($T$), density ($n$), numerical coefficients of turbulent transport models ($C$), and heat deposition ($P$). The conventional calculation of the NBI heat deposition by the GNET-TD takes several hours for a simulation run of two seconds in actual plasma. The EnKF can accelerate the simulation by using the reduced simulation model for NBI while ensuring the prediction accuracy. The time series data of experimentally measured temperature and density profiles are assimilated into TASK3D.Figure 1 shows the simulation results of $T_{\rm e}$ and $T_{\rm i}$ by EnKF, that have been performed for the cycle of assimilation, $\tau_{\rm DA}=80$ msec. The obtained $T_{\rm e}$ and $T_{\rm i}$ profiles and temporal variations by the data assimilation system have agreed well with measured ones. Figure 2 (a) shows the filtered estimate of $C_{\rm e}$. The radial profile has been optimized to reproduce the experimental temperature data, but this estimate is based on only the past data. The EnKS corrects the filtered estimate using future data. The smoothed estimate of $C_{\rm e}$ by the EnKS is shown in Fig. 2 (b). Because the future observations are reflected in the filtered estimates, the position of peak moves in the direction of the time axis (form 1.2 sec to 1.0 sec). We have confirmed that the TASK3D simulation (without the EnKF) using the smoothed estimates of $C_{\rm e}$, $C_{\rm i}$, $P_{\rm e}$ and $P_{\rm i}$ can reproduce the experimental temperature data with the high accuracy by Fig. 3.
If the estimates of $C$ for various plasmas can be reproduced by a parametric or nonparametric model, a more appropriate turbulent transport model for the prediction can be obtained with relevant physical interpretation. The data assimilation system based on an integrated simulation code is expected to be a powerful tool to analyze, predict and control the behavior of fusion plasmas in high accuracy.
[A] S. Murakami et al. Plasma Phys. Control. Fusion 57, 119601 (2015).
[B] Y. Morishita et al. Nuclear Fusion (2020) doi: 10.1088/1741-4326/ab7596.Speaker: Mr Yuya Morishita (Kyoto University) -
One of the objectives of the JET DTE2 experimental campaign will be related to investigation of alpha particle physics and alpha-driven instabilities (1) along with identifying their effect on plasma performance. Recent JET experiments have shown that the 3-ion ICRF heating schemes are an efficient way for generating a fast ion population in the plasma as a consequence of beam ions acceleration by ICRF waves and can also be applied for maximizing beam-target neutron rate (2, 3). To use such a scheme for alpha heating demonstration in DT plasmas, one should find experimental conditions that allow to accelerate beam ions to energies at which DT reactivity is maximized. Interpretative analysis of existing experiments is an essential step in validation and improvement of physical models required for predictive modelling.
We focus on modelling a multi-ion species plasma discharge with significant fraction of fast ions and improvement of the existing models of thermal and fast ion transport. For interpretative analysis with the TRANSP code (4) we have chosen JET mixed H-D plasma discharges, in which combined D-NBI and ICRF heating was applied using the 3-ion D-(DNBI)-H scheme (3). Modelling such JET plasmas is complicated by describing NBI-ICRF synergy and presence of plasma MHD activity, like sawtooth crashes. Fig. 1a shows a heating scenario for the JET plasma discharge #91257, where beam ions are accelerated from 100 keV to the energies of ~1 MeV by absorbing RF power in the vicinity of the ion-ion hybrid layer in the plasma core, and a high energy tail is formed in the fast ion distribution. These discharges also feature large sawtooth crashes with a long time period (Fig. 1b). A significant difference between the time evolution of the measured neutron rate and the neutron rate computed by TRANSP is observed in the simulations assuming same diffusivity for electrons and thermal ions, i.e. using the standard approach (Fig. 1c). We have found that uncertainties in the terms that enter in the TRANSP particle balance equation are crucial for modelling mixed-plasma discharges, and that increased D transport is necessary to get the TRANSP relative concentration of thermal D ions closer to the experimentally measured values. Our numerical analysis has been improved with including fast ion transport associated with sawtooth crashes. The resulting TRANSP simulations reproduce qualitatively the time evolution of the neutron rate in #91257 (Fig. 1c), although the computed neutron rate still overestimates the measured value.
We have performed sensitivity studies on thermal ion transport models used by TRANSP and on particle balance, along with quantitative assessment of their impact on plasma performance parameters, in particular on the neutron rate. Because of lack of ion density profile measurements, the standard approach in TRANSP is to assume the same transport coefficients for thermal ions and electrons, computed from the measured electron density profiles. If TRANSP solves the particle balance equation using this approach, it leads to a negative ion outflux in the plasma core. Simulation results demonstrate continuous accumulation of D ions in the plasma core and, consequently, an overestimated neutron rate. Growth of D ion density is expected and naturally related to thermalization of D beam ions. However, increase of the relative concentration of thermal D ions in the simulations significantly exceeds the isotopic ratio measurements at the plasma edge. Uncertainties in the TRANSP particle balance equation are related to the ion outflux model settings. With modified settings that increase D transport, the ion outflux becomes positive, leading to lower D density and the neutron rate. In line with experimental observations, we observe a slow growth of D density, calculated by the particle balance equation taking into account the NBI particle source and charge exchange losses provided by NUBEAM (5). The problem of choosing appropriate settings for the thermal ion transport models becomes particularly important for simulations of mixed plasmas with ICRF heating, for which the location of the RF absorption region depends on the plasma composition.
Fast ion distribution is significantly modified in the presence of sawtooth activity. During the sawtooth crash, fast ions are redistributed radially (Fig. 2), causing a reduction in the neutron rate. The redistribution of fast ions due to a sawtooth crash is usually modelled without taking into account their energy and orbit types. A reduced model can be used to estimate the effect of low-n MHD instabilities like the sawtooth activity on fast ion transport (6, 7) and to include fast ion orbital dependence. With transport coefficients obtained from the ORBIT code (8), this model improves physical representation of fast ion redistribution. In our numerical analysis, we show not only the ability of TRANSP to reproduce main trends in the neutron rate, but there is also an indication that fast ion transport induced by sawtooth crashes has an impact on the neutron rate, and thus more sophisticated models are necessary to describe modifications in the fast ion distribution function.
There are a few possible reasons why the computed neutron rate is overestimated in our TRANSP simulations. With sensitivity studies we estimate influence of TRANSP input parameters on the computed neutron rate. Depending on the quality of experimental data and fitting methods, some variation can be expected in these parameters as well as in the computed neutron rate. Thus, validated diagnostic data are essential for reliable interpretative analysis. Another limitation is related to the fact that we do not have a complete set of physical models in TRANSP, and it does not allow us to take into account self-consistently effect of MHD activity on the fast ion transport. Further numerical analysis also requires advanced description of wave propagation in the plasma with large fraction of fast ions. In the presented TRANSP simulations, the TORIC code (9) assumes a Maxwellian distribution for the fast ion species. Development of a consistent model with a non-Maxwellian distribution function in TRANSP is in progress (10) and aims to improve the description of RF wave propagation and absorption in plasmas, containing a significant fraction of fast ions.
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Science, under contract number DE-AC02-09CH11466. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
(1) R.J. Dumont et al, Nucl. Fusion 58 082005 (2018). (2) Ye. O. Kazakov et al, 2017 Nature Phys. 13 973. (3) J. Ongena et al, 2017 EPJ Conf. 157 02006. (4) R. J. Hawryluk, 1980 Phys. of Plasma Close to Thermonucl. Cond. 1 19. (5) A. Pankin, D. McCune, R. Andre et al., 2004 Computer Phys. Comm. 159 157. (6) M. Podestà et al, 2014 Plasma Phys. Control. Fusion 56 p. 055003. (7) D. Kim et al, 2019 Nucl. Fusion 59 086007. (8) R. B. White and M. S. Chance, 1984 Phys. Fluids B 2455. (9) M. Brambilla, 1999 Plasma Phys. Control. Fusion 41 1. (10) N. Bertelli et al, 2017 Nucl. Fusion 57 056035.Speaker: Dr Anna Teplukhina (Princeton Plasma Physics Laboratory) -
ITER will be equipped with a tungsten divertor, which is planned to operate over a significant period of time, from the Pre-Fusion Power Operation phase into the Fusion Power Operation phase 1. As was shown in a number of linear devices as well as in tokamaks, tungsten (W) exhibits pronounced surface morphology changes under helium (He) plasma exposure, which can affect its thermal and mechanical properties (see 2 and references therein). In particular, He induced nanostructures, such as He nanobubbles or the so-called W fuzz, can be formed at the W surface, depending on plasma conditions and surface temperature. He operation is planned in the ITER Pre-Fusion Power Operation phase, in order to demonstrate H mode operation without activating the in-vessel components 1. He will also be present during the Fusion Power Operation phase, as ash from the D-T reaction. Investigating interactions between W plasma facing components and He plasmas in a tokamak environment is therefore a key point to consolidate predictions for the ITER divertor performance and lifetime.
This paper reports on a dedicated He campaign performed to investigate these issues in the full tungsten tokamak WEST. The experiment was run at the end of the first phase of operation of WEST, before dismantling the lower divertor components for post-mortem analysis. During this phase, WEST lower divertor was composed of a mix of inertial W coated plasma facing units (PFU) and ITER like actively cooled PFU. The experiment was designed in order to reach the conditions for W fuzz formation in the low field side outer strike point (OSP) area, namely: incident particle energy Einc>20 eV, He fluence >1024/m2, surface temperature > 700°C 2. For this purpose, a robust long pulse scenario was developed in He (L mode operation with plasma current Ip = 300 kA, Lower Hybrid (LH) power = 4 MW, average density 4 1019 m-3). The plasma flattop duration was adjusted to ~30 s, in order to reach a surface temperature above the threshold for W fuzz formation over a significant area around the OSP on the inertial PFUs. Repetitive He long pulses (in the range 20-30 s) were run, cumulating ~2000 s of plasma exposure and 4.4 GJ of energy coupled to the plasma over 1 week of operation.
Local divertor plasma conditions were recorded with an array of Langmuir probes (LP), while the divertor surface temperature was monitored by embedded thermocouples (TC) / Fiber Bragg Gratings (FBG) and infrared (IR) cameras. It is shown that the 3 criteria mentioned above for W fuzz formation were met in the OSP area. Indeed, typical electron temperatures measured by the Langmuir probes were Te~20 eV at the OSP as shown in figure 1, which corresponds to an
incident He energy Einc > 100 eV. The surface temperature is estimated from IR measurements (see figure 2), assuming an emissivity of 0.15± 0.03 at the OSP, as deduced from the TC/FBG measurements 3. As shown in figure 2, the temperature threshold for W fuzz formation is typically reached after 20 s at the OSP. The total He fluence reached at the OSP is assessed from cumulated Langmuir probe measurements to be ~4 1025 He/m2. Combining IR and LP measurements, it is shown that even in the worst case scenario (highest emissivity assumption), the threshold for W fuzz formation in terms of He fluence (~1024/m2) is reached in an area of ~1 cm width around the OSP, with surface temperature exceeding 700°C for ~100 s.
Electron temperature profile (1a) and cumulated He fluence (1b) measured by Langmuir probes as a function of monoblock number along the divertor. The location of the inner and the outer strike points are shown.*
Infrared view of the lower divertor for shot #55953 (t = 27s), with the maximum temperature in the OSP area circled in black (2a) and time evolution of the maximum OSP surface temperature during the same shot (2b). Temperatures are shown for black body (=1) and real OSP emissivity = 0.15± 0.03. The threshold for W fuzz formation (700 °C) is indicated with a dashed line.*In-vessel inspections using the WEST Articulated Inspection Arm (AIA) were performed before and after the He campaign. They did not reveal any macroscopic signs of W surface modification in the OSP area of inertial PFU, such as blackening of the surface reported when thick W fuzz is formed. Post mortem analysis of the components is however necessary before concluding on the type of He induced nanostructures formed during the experiment, and is ongoing at the time of writing. It will be reported in the paper. These results underline that in tokamak conditions, the complex balance between erosion (in particular from impurities) /redeposition (from W eroded from the main chamber) and W fuzz formation needs to be taken into account, as was shown in other tokamak experiments 4 [5] [6]. The data obtained will be used to consolidate the experimental database supporting the modelling effort for predicting W fuzz formation and growth in ITER, such as in [7].
1 ITER Research Plan within the Staged Approach, ITR-18-003, 2018, 2 G. De Temmerman et al., Plasma Phys. Control. Fusion 60 (2018) 044018, 3 J. Gaspar et al., submitted to PSI2020, 4 S. Brezinsek et al., Nuclear Materials and Energy 12 (2017) 575–581 [5] A. Hakola et al, Nucl. Fusion 57 (2017) 066015, [6] S. Brezinsek et al., submitted to PSI2020, [7] G. De Temmerman et al, Nuclear Materials and Energy 19 (2019) 255–261256
Speaker: Emmanuelle Tsitrone (CEA) -
Liquid metal blankets are advanced and have many attractive features such as low operating pressure, design simplicity, and a convenient tritium breeding cycle. However, the magnetohydrodynamic (MHD) effects are a key issue remaining to be solved, one of them is the coupling MHD effect[1-3] which normally exists in liquid metal blankets. The coupling MHD effect is that the flow state in one duct will be affected significantly by the leaking electrical currents from the neighboring duct because the coupling duct wall is conductive. Previous results[4,5] indicate that MHD flow states could be modified obviously by external inclined magnetic fields for one duct flows. However, a systematic study on what is the influence of the inclined magnetic fields on the coupling MHD duct flow states is still lacking. This work aims at clarifying the influence of the inclined magnetic fields on the MHD flow states such as MHD pressure drops and velocity distributions through two coupling ducts with conducting walls.
One of the most important results[3] for the coupling MHD effect is that the pressure gradient in the coupling ducts will be several times bigger than that in one single duct if the initial flow directions in the two neighboring coupled ducts are opposite. The external magnetic fields in Reference [3] have no inclination with one pair of walls and the two-dimensional (2D) simulations are based on a fully developed modeling. The three-dimensional (3D) simulation results of this work confirm the 2D results as shown in figure 1. The 3D numerical simulations are using a self-developed code based on OpenFOAM environment. The Ha and Cw in figure 1 are the Hartmann number and the wall conductance ratio respectively. The Ha denotes the ratio of the Lorentz force to the viscous force in the fluid and the Cw represents the non-dimensional conductivity of the duct wall. The comparison results in figure 1 indicate that the 2D results agree well with the 3D results. In the case of Cw=0.01, the coupling MHD effect is significant and the pressure gradient is about 7 times bigger than that of one single duct when the inclined angle of the external magnetic field is zero, the pressure gradients decrease with the increases of the inclined angles of the external magnetic fields. In the case of Cw=0.1, the coupling MHD effect is weaker than that in the case of Cw=0.01 and the pressure gradient is about 2.4 times bigger than that of one single duct when the external magnetic field has no inclination, the pressure gradients increase firstly and then decrease with the increase of the inclined angles of the external magnetic fields. The other important result of this study is that there are reversal velocity distributions in the corners of the duct when the external magnetic field has inclination such as the inclined angles is 22.5° as shown in figure 2. The big reversal velocity distributions are harmful to the heat transfer in liquid blankets and should be considered in the future blanket design because the external magnetic fields normally have inclination with one pair of duct walls. Other 2D and 3D numerical simulation results on the influence of the external inclined magnetic fields on the MHD flow states in the two coupling ducts are also included in this paper. The above-obtained results are important and helpful for future liquid metal blanket designs, and some new observed results such as the reversal velocity distributions update the related liquid MHD knowledge.
Acknowledgements: This work is supported by the Sichuan Science and Technology Program (Grant No. 2019YJ0297) and the Innovation Program of SWIP (201901XWCXRC002).
References:
(1) S. Molokov et al. J. Fluid Mech., vol. 299, pp. 73–95, Sep. 1995.
(2) Mistrangelo, C et al. Fusion Eng. Des.109–111 (2016) 1452–1457.
(3) Xiujie Zhang et al. IEEE TPS, VOL. 42, NO. 6, JUNE 2014, 1764-1769.
(4) Hua T.Q et al.1995, Fusion Eng. Des. 27 703–710.
(5) Xiujie Zhang et al. 2019, Nucl. Fusion 59 056018.Speaker: Xiujie Zhang (CnSWIP) -
We investigate ion temperature gradient (ITG) mode in hydrogen-isotope plasmas under a radial electric field in Large Helical Device (LHD) using the global gyrokinetic code, XGC-S. The radial electric field is taken into account by additional $E \times B$ drift motion in the poloidal direction. The present multi-mode linear simulations indicate the following properties of ITG mode and resulting heat flux. (i) In the absence of the radial electric field, mass-number dependencies of maximum growth rate and heat flux are in agreement with the quasi-linear theory. (ii) The radial electric field stabilizes the ITG mode and also affects the dominant wavelengths. The heat flux in heavy-hydrogen plasmas is lower than that predicted from the theoretical mass number dependency because of this modification of the mode structures. (iii) The radial electric field selectively stabilizes the ITG modes relevant to the light-hydrogen component in multi-component plasma. This selective contribution might limit the influence of the radial electric field on heavy-hydrogen heat flux.
ITG mode has been considered a primary cause of anomalous transports in magnetic confinement devices. Neoclassical transport accompanied by a radial electric field also plays a critical role in LHD. The radial electric field generates plasma rotation in the poloidal direction, and potentially affects the ITG mode and resulting turbulent transport. Recently, isotope effects in the anomalous transport are considered in deuterium experiments in the LHD. In this work, we investigate ITG mode properties in hydrogen-isotope plasmas under a radial electric field in a magnetic equilibrium relevant to LHD. The present work is the first step of global simulation study on the interaction between turbulent and neoclassical transport phenomena, which is not fully understood for hydrogen-isotope plasmas in stellarators.
We employ the global gyrokinetic code, XGC-S, developed for non-axisymmetric fusion devices$^{a}$. Benchmark studies on basic plasma phenomena in stellarators have been carried out for this code$^{b,c}$. Linear ITG simulations for a light-hydrogen plasma in the present equilibrium are also benchmarked with the previous EUTERPE simulations$^{d}$. An additional $E \times B$ drift term in a given radial electric field, $E_{r}$, is included in the ion equations of motion. We assume a uniform number density, $n=10^{19} m^{-3}$, and radial electric field related to the ion-root branch. In Figure 1(A), the temperature and radial electric field profiles are shown. We use unstructured meshes to represent perturbed electrostatic fields. Typical poloidal and toroidal resolutions are $\sim 1200$ / toroidal cross section / flux surface and $144$ / toroidal periodicity, respectively. Approximately $50$ delta-f marker particles are initially loaded per one mesh vertex. Only adiabatic response is assumed for the electron dynamics. The hydrogen mass number, $A$, is varied from $A=1$ to $A=3$. We also consider a multi-component plasma with $A=1$ ($50\%$) and $A=3$ ($50\%$) components.
In Figure 1(B), the maximum growth rates are summarized as functions of average mass number with ($\times$) and without ($+$) $E_{r}$. The growth rates are normalized to $V_{ti}/a$, where $V_{ti}$ and $a$ are ion thermal velocity for average mass number ($i=H,D,T$ for $A=1,2,3$) and typical minor radius, respectively. The green and red points indicate the growth rates in single- and multi-component plasmas, respectively. In the absence of $E_{r}$, the normalized growth rates are almost constant, exhibiting a dependency on the square root of mass number. The same property is observed in the presence of $E_{r}$ with lower growth rates. Therefore, the poloidal rotation is considered to stabilize the ITG mode with a similar stabilization rate independent of the hydrogen mass.
In Figure 2, mode amplitudes as a function of wavenumber in the early linear phase are shown for the $A=1$ (red), $3$ (blue), and multi-component (green) cases without (A) and with (B) $E_{r}$. The mode amplitudes are normalized to their maximum values. In the absence of $E_{r}$, the dominant wavenumber $k_{max}$ depends on the square root of average mass number in both single- and multi-component plasmas. The wavenumbers normalized to ion Larmor radius $\rho_{i}$ are roughly constant ($k_{max}\rho_{i}\sim0.3$). The theoretical linear growth rate, $\gamma_{k}$, is proportional to the wavenumber, independently of the hydrogen mass$^{e}$. Therefore, the maximum growth rate $\gamma_{k_{max}}$ is expected to be proportional to $V_{ti}/a$, namely, $(1/A)^{1/2}$. The maximum growth rates shown in Figure 1(B) agree with this theoretical prediction.
In the presence of $E_{r}$, as shown in Figure 2(B), $k_{max}$ are small compared to those in the absence of $E_{r}$. This reduction of wavenumber tends to be evident in the light-hydrogen plasma. In the multi-component plasma, the poloidal rotation mainly stabilizes higher wavenumber modes relevant to the light-hydrogen component. Consequently, the mode profile becomes similar to that in the heavy-hydrogen plasma. This result indicates selective stabilization of higher wavenumber modes due to poloidal rotation in multi-component plasmas.
The graphs in Figure 3 represent heat flux, $Q$, divided by the squared amplitude of (A) electrostatic potential $\Phi$, and (B) perturbed electric field $E$, in single-component plasmas as a function of the mass number. These values are normalized to that obtained for $A=1$ without $E_{r}$. Points $+$ and $\times$ represent the results obtained without $E_{r}$ and with $E_{r}$, respectively. $Q/\Phi^{2}$ decreases in heavy-hydrogen plasmas, inversely proportional to the square root of mass number. This result agrees with the estimate of heat flux in the quasi-linear theory$^{e}$, i.e., $Q/\Phi^{2}\propto k_{max}\sim (\rho_{i}/a)^{-1}$. The observed mass number dependencies of ${k_{max}}$, $\gamma_{k_{max}}$, and $Q/\Phi^{2}$ are also consistent with Gyro-Bohm scaling, where spatial and temporal scales are characterized by $\rho_{i}$ and $a/V_{ti}$, respectively.
In contrast, $Q/E^{2}$ increases as the mass number increases in the absence of $E_{r}$, as shown in Figure 3(B), because $E/\Phi\sim k_{max}$ decreases in the heavy-hydrogen plasmas. In the presence of $E_{r}$, however, $Q/E^{2}$ is roughly constant because the decrease in $k_{max}$ due to poloidal rotation is more evident in light-hydrogen plasmas. In the multi-component plasma, $Q/E^{2}$ of the heavy-hydrogen component (not shown in the graphs) is observed to be $\sim 0.42$ (without $E_{r}$) and $\sim 0.39$ (with $E_{r}$) in the same normalization as that in Figure 3(B). $E_{r}$ has little effect on these values, likely because of the selective contribution of poloidal rotation to higher wavenumber modes relevant to the light-hydrogen component.
Although these simulations do not include nonlinear evolutions, the mixing length estimation gives typical amplitudes of electrostatic potential, namely $e\Phi/T \propto\lambda_{max}/a$, where $\lambda_{max}$ is the wavelength of the dominant mode. According to this estimation, $Q/E^{2}$ has the same mass number dependency as $Q$, because $Q/E^{2}\sim Q/(\Phi/\lambda_{max})^{2}\propto QA^{0}$. The present simulation results indicate that the heat flux due to the ITG mode under the effects of the poloidal rotation in heavy-hydrogen plasmas might be small compared to that predicted from the mass ratio dependency in Gyro-Bohm scaling, $Q\propto A^{1/2}$.
[a] T.Moritaka ${\it et~al}$., Plasma ${\bf~2}$ (2019) 179-200.
[b] M.Cole ${\it et~al}$., Phys.Plasmas ${\bf~26}$ (2019) 032506.
[c] M.Cole ${\it et~al}$., Phys.Plasmas ${\bf~26}$ (2019) 082501.
[d] J.Riemann ${\it et~al}$., Plasma Phys.Control.Fusion ${\bf~58}$ (2016) 074001.
[e] F.Romanelli, Phys.Fluid ${\bf~B1}$ (1989) 1018.Speaker: Toseo Moritaka (National Institute for Fusion Science) -
Runaway electrons (RE) are one of the major concerns for ITER operations. In tokamak devices where fusion reactions take place at high rate and with large total current, the loss of plasma confinement can lead to runaway electrons formation (possibly up to 12 MA in ITER) via primary and secondary generation mechanisms [1,2]. The major issue with runaway electrons that form after the disruption is their high energy, mainly increased by the large electrical field at current quench (CQ), and the small pitch angle that could deposit unsustainable power on the plasma facing components causing deep melting in the tokamak structure. The main strategies on RE mitigation rely on increasing collisionality to avoid RE beam formation or to quickly dissipate its energy if already formed [1,8]. The latter action is unavoidably associated with undesirable fast-growing vertical displacement events, i.e. quick RE energy dissipation by heavy-Z material injection is linked with fast current decay leading to uncontrollable VDEs: it is a race among vertical displacement, energy dissipation and electromechanical loads that might not lead to ITER feasible solutions. Researchers are providing further techniques that might be used in combination with MGI/SPI such as a dedicated control strategy [3,4], 3D stochastic fields by resonant magnetic perturbation [7] and further instabilities [9]. Recent results obtained at DIII-D and further investigated at JET show that large RE currents, driven by the central solenoid after deuterium injection (SPI) to quickly reduce the drag, induce current-driven (low safety factor) kink instabilities with extremely fast RE beam loss and no sign of localized energy deposition, opening the path for an alternative RE mitigation strategy [3]. Continuing the studies of past years at FTU [4,5] we tested a number of alternative solutions for RE mitigation. In FTU large population of quiescent REs in steady state current ohmic discharges have been created and interaction with (multiple) deuterium pellets and Laser Blow Off (LBO) injection have been studied. In quiescent scenarios, it has been observed with a significative number of tests that D2 pellet on quiescent RE population might lead to REs fast growth up to the RE beam formation, if RE population or MHD activity is above a given threshold, meanwhile multiple pellets injection fairly close enough (1-20 ms) can produce quick REs total expulsion and or dissipation restoring a “safe” steady-state discharge with no REs as shown in Fig. 1 and Fig. 2. Different ablation rates of pellets with different size (1-2E20) depending on the REs quiescent population and inter-time pellet injection have been registered by fast H-alpha acquisition channels and the fast CO2 scanning interferometer. Such data can be important for pellet ablation models with REs that can be useful for ITER predictions. It is the worth to mention that pellet injections are as well the subject of studies at FTU for large MHD stabilization, possibly providing a solution to discharge recovery and RE preemptive dissipation at once. There have been also few cases in which pellets have been launched at CQ and, interestingly and somehow expected, the REs formed have lower energy. Indeed, one possible dissipation methodology would consist into providing a large number of electrons (deuterium injections) that could absorb at least part of the electrical field produced at the CQ preventing high level energy increase of RE seeding meanwhile flashing out all high-Z species at CQ would decrease the current drop and increasing then the RE beam controllability. It has also been observed, for the first time at FTU, density increase after D2 pellet injection as well as LBO ionization (and drag effects) on a post disruption RE beam with clear signs of increased background plasma temperature: fan-like instabilities seem to play an important role on such temperature increase. Modulated ECRH has been used in order to further increase background plasma density and temperature and a surprisingly synchronization with fan-like and MHD driven instabilities has been found. Data on RE energy evolution have been acquired with the new REIS [5,6] during RE quiescent flat-top as well as Ip ramp and post-disruption RE beam to provide data for RE energy model validation.
To conclude, the FTU last experimental campaign provided data on D2 pellet
effectiveness for preemptive RE mitigation technique, pellet ablation studies for ITER, as well as its effect on post-disruption RE beams and data for RE model validation.
Acknowledgement: This work has been carried out within the framework of the EUROfusion Consortium. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References:1 Lehnen M et al, J. Nucl. Mater. 463 39 (2015)
[2] Connor J W and Hastie R J, NF, 15 415 (1975).
[3] C. Paz-Soldan et al, A novel path to runaway electron mitigation via current-driven kink instability, submitted to IAEA 2020.
[4] D. Carnevale et al., Plasma Phys. Control. Fusion 61 014036 (2019).
[5] Esposito B. et al., PPCF, vol. 59, ISSN: 0741-3335 (2016).
[6] F. Causa et al, Review of Scientific Instruments 90, 073501 (2019).
[7] M. Gobbin et al, PPCF 60 1 (2017).
[8] G. Papp et al, RE generation and mitigation on the European medium sized tokamaks ASDEX Upgrade and TCV,26th IAEA (2016).
[9] F. Causa, NF 59 4 (2019).Speaker: daniele carnevale (University of Rome Tor Vergata) -
Heat load control on plasma facing components (PFC) is a critical issue for fusion devices like ITER: power injection and exhaust systems should be regulated and high radiating plasma scenario at the edge are investigated. Future fusion reactors will have to rely on impurity seeding to enhance the radiative fraction in the range of ~90% in order to cool the edge plasma and prevent PFC damages. Among the different gases tested in present fusion experimental devices, nitrogen (N2) is a viable seeding candidate for the size and conditions of present day divertors. However, the potential reactivity of N2 with hydrogen isotopes can lead to tritiated ammonia (NT3) formation which should be considered for the re generation of cryo pumps and processes in de-tritiation plants. Using N2 in the ITER facility as radiator in the edge plasma and particularly in the divertor is still nevertheless considered as a viable option. In the WEST tokamak [1, 2], designed for long pulse operation with diverted magnetic configuration and tungsten (W) PFC, long discharges were seeded with N2. The objectives of these experiments were to study plasma surface interaction (PSI) and ammonia formation in a full tungsten device in order to improve the understanding of physics of ammonia production, decomposition and transport in a magnetically confined plasma devices. This paper reports on various aspects of these discharges: power balance, impact of N2 injection on plasma parameters, and ammonia formation. In WEST, the upper divertor is equipped with actively cooled copper based PFC, coated with ~20µm of tungsten allowing long discharges in steady state conditions in upper single null configuration (USN). Although no active pumping has been installed, long discharges in USN configuration are now routinely operated.
In this context, 15 repetitive long and steady state discharges (up to 55s) have been performed in USN plasma configuration, cumulating 8 min of plasma and 1.18GJ of injected energy. The discharges were achieved in L-mode, with a plasma current of Ip= 400 kA, lower hybrid (LH) power of PLH= 1.5 to 3.2 MW, a plasma density of ne=3.3x1019m-2, a radiated fraction of ~55% and a diamagnetic energy Wdia ~200kJ. During the four last discharges cumulating more than 3 minutes of plasma, N2 has been seeded through a toroidal ring located below the divertor in the vicinity of the outer strike point (OSP) at a rate ranging from 0.1 to 0.22 Pam3s-1 for a total duration of 100s for investigating both the potential legacy of N2 injection on plasma performances and the formation of ND3 in steady state conditions. For all these discharges, the main radiative impurities have been identified as W, O, Cu (mainly during the LH power phase) including some traces of C whilst a rather weak contribution of N in the overall radiation is observed during the N2 seeding phases. The edge plasma characteristics are documented through density and temperature (ne and Te) measured with the set of Langmuir probes in the divertor and the reciprocating probe, whilst the potential ND3 formation during and after the discharges was monitored by a residual gas analyser (RGA) located in the pump duct at the outer mid plane. Performing long discharges also allows assessing more accurately the global power balance and distribution on various PFC. The overall power balance (conducted and radiated power) is evaluated through the main diagnostics of infrared thermography, bolometry and calorimetry, showing the complementarity of these diagnostics for closing the power balance within a range better than 90%. A typical discharge is displayed on [figure 1][1] showing the main plasma parameters. The plasma temperature at the OSP region is in a range of 5-10eV in the ohmic phase prior the LH power whilst it increases around ~15 eV during the heating phase. Several plunges have been performed with the reciprocating Langmuir probe. The profiles of both Te and ne have been measured up to a normalised radius of =r/a~1.05 corresponding to about Te~75eV and ne~1.6x1019m-3 at the separatrix. Over the entire heating phase, no effect of N2 can be detected all along the N2 injection on the Langmuir probe signals. Consistently with these results, the maximum surface temperature at the OSP region is around 300°C and no significant drop of the surface temperature is observed during the N2 injection. The N2 effect is observed in the bulk plasma through a weak increase of the N VII signal exhibiting an increase from pulse to pulse although the N2 seeding rate over these consecutive pulses is maintained in a range between 0.1 to 0.22 Pam3s-1. However, this increase is too weak to be detected through the bolometry track probing the bulk plasma whilst at the OSP region, the edge bolometry exhibits an increase by ~8%. It is worth noting that there is no significant/measureable change on the bolometrer tracks monitoring the lower part of the plasma. This shows that the radiation pattern is only weakly modified as confirmed by both the Langmuir probes and the IR measurements. Finally, and even for the discharge containing the largest nitrogen injection (# 55792 with 6.5Pam3of N2 injected @ ~0.22 Pam3s-1 from 10 to 40 s and 11Pam3 of D2), no ND3 is detected in the RGA during the discharge, at the end of the discharge when the plasma recombines and during the ~25min of outassing between pulses.
In these conditions, over such long time scales and in the absence of active pumping in USN configuration, these experiments suggest that the majority of the injected N2 sticks to the wall in the vicinity of the injection point and that the produced ammonia also sticks to the walls and is then released at a low rate below the sensitivity of the RGA system between pulses.
Acknowledgements: “This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.”
[1] J Bucalossi et al., Fusion Engineering and Design, 89 (2014) 907–912.
2 C Bourdelle et al., Nuclear Fusion 55 (2015) 063017 (15pp).
[3] T Dittmar et al., accepted for publication in Physica Scripta.Speakers: Dr Thierry Loarer (CEA-IRFM), Dr Timo Dittmar (Forschungszentrum Jülich - IEK-4) -
Here we report progress in electrical and mechanical performances of mechanical joints of high-temperature superconducting (HTS) conductors and methods to evaluate those quality. Our research activities showed that (i) Joint resistance for bridge-type mechanical lap joint and mechanical edge joint have been reduced to be acceptable value for HTS helical coils in FFHR-d1 helical fusion reactor, (ii) The bridge-type mechanical lap joint with indium insertion is preferable for use in joint-winding of the HTS helical coil because of its mechanical behaviour, and (iii) X-ray computer tomography (CT) scan and contact-probing current transfer length (CTL) method are promising to predict joint resistance, and quality control of the joints during fabrication process at room temperature before applying current.
Two designs of segment-fabrication of HTS helical coils have been proposed for FFHR fusion reactors, joint-winding of the HTS coils wound by connecting conductor segments, and the “remountable” HTS magnet (here “remountable” means being able to mounted and demounted repeatedly) assembled from coil segments with remountable joints. Fig. 1 shows schematic illustrations of the two designs. Bridge-type mechanical lap joint and mechanical edge joint are planned to be applied to those two designs, respectively. The broad surface of the stepwise stacked REBCO HTS tapes are joined mechanically via the joint piece in the bridge-type mechanical lap joint whereas the side edges of the stacked REBCO tapes are joined in the edge joint. The electric and mechanical performances of the joints and those quality assessment are important technical issues for the designs.
Hashizume and Ito of Tohoku University started to develop the mechanical lap and butt joints in 2000 and they also proposed bridge-type mechanical lap joint and mechanical edge joint with indium-foil insertion based on previous two joints, respectively [1]. Fig. 2 shows progress in joint performances for mechanical lap joint including the bridge-type joint and edge joint for this decade. In 2012, Tohoku University and NIFS started to develop the Stacked Tapes Assembled in Rigid Structure (STARS) conductor using REBCO tapes and its joint. We achieved a joint resistance of 1.8 nΩ at 100 kA, 4.2 K using bridge-type mechanical lap joint [1, 2]. We additionally introduced two techniques for these five years, heat treatment at around 100°C [1, 2] and integrated joint piece [3]. These techniques have reduced the joint resistivity (product of joint resistance and contact area) to those in single-tape joints and fabrication time of the large-scale joint from over a half day to within three hours. The mechanical edge joint has been also progressed and its performance reached to the same level as the bridge-type joint. The target value of joint resistivity is less than 10 pΩm2 from a viewpoint of electric power for cooling the magnet with joint length of <1 m in the FFHR fusion reactor, and the achieved joint resistivity satisfies the required performance.
A structural analysis of continuous HTS helical coils assuming joint-winding indicated that a maximum shear stress of 32–35 MPa is induced in the plane of the joint surface [1]. Fig. 3 shows shear strength of mechanical lap joint with indium-insertion as a function of contact resistivity (product of contact resistance between REBCO tape and indium foil, and contact area) and it is almost the same as the maximum in-plane shear stress of 35 MPa even with 100 MPa joint pressure. In the cohesive failure case, the joint resistance only increased slightly with increased displacement whereas soldered lap joints showed sudden resistance increase after its failure due to stress concentration [1]. REBCO tapes can only be displaced by less than 50 μm in the actual magnet due to the presence of the jacket. Therefore, the proposed joint is preferable for use in joint-winding of the HTS helical coil.
REBCO tape has interface resistance between its superconductor and stabilizers. We developed contact-probing CTL method [4] for nondestructive evaluation of the interface resistance before REBCO tape selection. Quality control for the joints during fabrication at room temperature is another important issue, and we also developed a method to evaluate contact area at the joint section based on X-ray CT scan [5]. Fig. 4 shows relationship between contact area evaluated by X-ray CT scan and contact resistance between HTS tape surface and indium foil. The contact resistivity was between 2 and 6 pΩm2 after introducing the heat treatment during the fabrication process, which was less influenced by configuration of the joint.
References
[1] S. Ito et al., Fusion Eng. Des. 136 (2018) 239.
[2] H. Hashizume et al., Nucl. Fusion 58 (2017) 026014.
[3] S. Ito et al., Fusion Eng. Des. 146 (2019) 590.
[4] R. Hayasaka et al., IEEE Trans. Appl. Supercond. 29 (2019) 9000805
[5] W. Chen et al., Fusion Eng. Des. 148 (2019) 111284.Speaker: Satoshi Ito (Tohoku University) -
As a promising strategy, quiescent H-mode (QH-mode) offers tokamak plasma scenarios without edge localized modes (ELMs) but with constant plasma density and radiated power. This type of ELM-free H-mode regime is often achieved at relatively low plasma density in experiments, and is found to be accompanied by the presence of edge harmonic oscillations (EHOs), which are believed to provide necessary transport to eliminate ELMs by the plasma self-regularization dynamics. The saturated kink-peeling mode has been suggested as a possible candidate for EHO. According to this scenario, an edge localized kink-peeling instability, with the dominant toroidal mode number typically below 3 ($n\leq3$), is non-linearly saturated by the plasma flow, thus leading to EHO.
In this work, we perform a systematic numerical investigation of various physical effects on the kink-peeling instabilities in a DIII-D QH-mode plasma from discharge 157102, utilizing the MARS-F/K/Q suite of code $[1][2][3]$. These include the wall eddy current effect, the influence of the plasma resistivity on the mode stability, the kinetic effects associated with thermal particle drift motions, the effect of plasma pedestal toroidal flow and flow shear, and finally the dynamics of self-consistent interaction between the plasma flow and the kink-peeling mode. Quantitative comparison has also been made between modeling and experimental data.
With the Spitzer model assumed in this study, we find that the computed $n=1$ and $n=2$ kink-peeling growth rate is well approximated by an analytic scaling of $S^{-1/3}$, similar to that for the edge localized infernal mode $[4]$.
With all the drift kinetic physics included in this study, namely (i) the precessional drift resonance of trapped ions and electrons ($\alpha_D$), (ii) the bounce resonance of trapped ions ($\alpha_B$), and (iii) the transit resonance of circulating ions ($\alpha_C$), MARS-K modeling finds destabilizing effect on the low-n kink-peeling mode (Fig.1(a)). Kinetic effects also modify the mode eigenfunction, by reducing the core mode components (Fig.1(c)) and by shifting the peak location of the perturbation near the plasma edge (Fig.1(d)).
The computed mode eigen-structures are also compared with experimental measurements. Figure 2 shows the amplitude and toroidal phase of the n=1 magnetic fluctuations associated with poloidal magnetic sensors from both low and high field sides of the torus. The modeled results agree reasonably well experiments, except for a slight systematic phase shift for the high field side magnetic signal.
The MARS-Q initial value quasi-linear simulations reveal that several plasma parameters sensitively affect the non-linear interaction between the plasma flow and the kink-peeling instability. These include the momentum diffusion coefficient $\chi_M$ as well as the amplitude $A_0$ of the initial perturbation. One example is shown in Fig.3, where we fix $\chi_M$=1m$^2$/s and vary $A_0$ $[5]$. The time traces of the perturbed $m/n=6/1$ radial field component, at the $q=6$ rational surface, show that different initial perturbation amplitudes result in similar level of the saturated mode amplitude (Fig.3(a)). The simulated saturation time scale, however, is different (Fig.3(a-b)). The time behavior of the flow damping near the plasma edge (at the q=6 surface as shown in Fig.3 (b)) shows a similar pattern independent of the saturation time scale (Fig.3 (b)). The edge flow always decreases to a minimal value of about -10 krad/s, before recovering to about 0 krad/s level.
Significant change of the plasma edge rotation profile (and rotation shear) occurs during the saturation of the kink-peeling mode, as shown by Fig.3(c) where $A_0=5\times10^{-8}$ is assumed. Further comparison of the time traces of the net torque amplitude (Fig.3(d)) shows that the dominant contribution to the flow variation comes from the net Reynolds stress torque and resonant electromagnetic ($j\times b$) torque.
Reference
$[1]$ Y. Q. Liu, et al, Phys. Plasmas 7, 3681 (2000).
$[2]$ Y. Q. Liu, et al, Phys. Plasmas 15, 112503 (2008).
$[3]$ Y. Q. Liu, et al, Phys. Plasmas 20, 042503 (2013).
$[4]$ G. Q. Dong, et al, Phys. Plasmas 24, 112510 (2017).
$[5]$ G. Q. Dong, et al, Nucl. Fusion 59, 066011 (2019).Speaker: Guanqi Dong (SWIP) -
We report the results of experimental characterization of the depth dependent helium concentration in single crystal tungsten specimens, as compared to modeling predictions, following repeated helium plasma exposures in the reciprocating collector probe in WEST during the C4 helium campaign. This study was motivated by the opportunity to experimentally validate modeling predictions of the helium retention and sub-surface helium spatial profiles and bubble concentrations following helium plasma exposure conditions in WEST, which is an all tungsten metal long pulse tokamak. WEST has a reciprocating collector probe that can be inserted into the far scrape off layer plasma (Fig. 1a) and provides repeatable exposure conditions during reproducible plasma discharges that limit sample heating and yet still provide significant helium plasma exposure that is above experimental detection thresholds. This enables experimental validation in well controlled conditions of a high fidelity, integrated modeling capability for plasma surface interactions that has recently predicted the performance of the ITER tungsten divertor to initial operation conditions relevant to helium plasma operation as well as for burning plasma conditions 1.
Six single crystal tungsten samples were loaded in the reciprocating collector probe and exposed to repeated plasma discharges in November 2019 to reach an accumulated helium fluence ranging from approximately 2 to 7x1023 m-2. Figure 1 shows a schematic illustration of the probe location in the WEST tokamak with the magnetic field lines and plasma facing surfaces, respectively, along with a photograph of the three lower slots containing tungsten single crystal samples in the collector probe. Two different tungsten single crystal orientations, either (100) or (111), were used in this experiment, in order to test modeling predictions that (111) oriented surfaces retain significantly more helium than (100) oriented tungsten surfaces 2.
These exposure conditions have been modeled using our coupled, high-fidelity plasma-surface interaction codes, including the evolution of the plasma-facing component (PFC) surface layer that is continually modified by contact with the fusion plasma. This approach to modeling plasma surface interactions is shown in Figure 2, and involves a wide range of physical phenomena: our current model includes components for a) the scrape-off layer plasma including helium and extrinsic impurities (using SOLPS[3]), b) the implantation of plasma ions into the material, including electromagnetic sheath affects, and subsequent wall erosion (using hPIC [4] and F-TRIDYN, and extension of TRIDYN [5]), and c) the dynamics of the subsurface (Xolotl, a continuum cluster dynamics code 2). These components are integrated to yield predictive capability for the changes in surface morphology in addition to the sub-surface helium retention and spatially dependent concentrations. The SOLPS modeling predictions of the background plasma conditions during the WEST helium plasma discharges have been benchmarked to Langmuir probe data obtained by the collector probe, and provide input to hPIC, F-TRIDYN and Xolotl on the helium implantation conditions (flux, ion energy).
The experimental measurements include laser ablation based techniques to quantify the depth dependence of the helium, using laser induced breakdown spectroscopy and laser ablation mass spectrometry. These laser based measurements build upon prior measurements of the helium and deuterium spatial distributions following low-energy plasma exposure in PISCES/A [6]. Additionally, thermal desorption measurements are performed to quantify the helium release kinetics and to provide additional experimental data to compare with the coupled plasma surface interaction modeling predictions of Xolotl. This presentation will discuss the agreement between the modeling predictions and experimental observations, possible reasons for any discrepancies, and plans for future experiments.
*Research supported by the US Department of Energy through the Scientific Discovery through Advanced Computing (SciDAC) program, jointly sponsored by the Fusion Energy Sciences (FES) and Advanced Scientific Computing Research (ASCR) programs within the U.S. Department of Energy, Office of Science, and partially supported by the DOE under DE-SC0020414 and DE-AC05-00OR22725.
REFERENCES CITED
1 A. Lasa, J.M. Canik, S. Blondel, T.R. Younkin, D. Curreli, J. Drobny, P. Roth, M. Cianciosa, W. Elwasif, D.L. Green, and B.D. Wirth, Physica Scripta T171 (2020) 014041.
2. S. Blondel, D. E. Bernholdt, K. D. Hammond, L. Hu, D. Maroudas, and B. D. Wirth, Fusion Science and Technology 71 (2017) 84-92.
[3]. W. Miller et al, Comp. Phys. Comm. 51 (1988) 355.
[4]. Khaziev R. and Curreli D. 2015 Phys. Plasmas 22 043503.
[5]. Drobny J. et al. 2017 J. Nucl. Mat 494, 278–283.
[6]. G. Shaw, W. Garcia, X. Hu, and B.D. Wirth, Physica Scripta T171 (2020) 014029.Speaker: Prof. Brian Wirth (University of Tennessee, Knoxville) -
We clarify the toroidal flow generation mechanism by electron cyclotron heating (ECH) in stellarator/heliotrons comparing the HSX and LHD experiment results. Radial diffusion of energetic electrons by ECH produces a canceling return current, which then generates a $j_r\times B$ torque that can play an important role in the toroidal rotation in the ECH plasmas. We investigate the energetic electron distribution by ECH by applying GNET code, which can solve the 5D drift kinetic equation for the energetic electrons. We evaluate the $j_r\times B$ torque and the collisional torque due to the friction of the toroidal drift motion of the energetic electrons. As a result, we obtained a significant torque due to the ECH and found that the torque becomes larger in the Mirror configuration than that in the quasi-helically symmetric (QHS) configuration in HSX. Solving the momentum balance equations and Maxwell's equation with the $j_r\times B$ torque, we evaluate the toroidal flow velocity and compare simulations with HSX experiments. The obtained flows have good agreement with ones in HSX experiments. Moreover the simulated toroidal flows driven by ECH are consistent with that of the LHD experiment retults.
Introduction
Recently, spontaneous toroidal flows have been observed in electron cyclotron heating (ECH) plasmas in many tokamak and helical devices such as JT-60U, LHD and HSX. To clarify the underlying mechanism, many experimental[A] and theoretical studies have been undertaken. The effects of the magnetic configuration on plasma flow are intensively investigated in HSX, where two typical magnetic configurations are considered. One is the Quasi-Helically Symmetric (QHS) configuration, which has a quasi-helical symmetry in $|B|$ and is dominated by the $(m,n)= (1,4)$ mode. The other is the Mirror configuration, where a set of auxiliary coils adds toroidal mirror terms, the $(0, 4)$ and $(0, 8)$ modes, to the magnetic field spectrum to break the helical symmetry. The parallel neoclassical viscosity of the QHS configuration is smaller than that of the Mirror configuration, so we expected that the toroidal flow velocity in the QHS configuration would be more significant than that of the Mirror configuration. However a smaller toroidal flow was observed in the QHS configuration. The mechanism of the toroidal flow generation has not been understood well yet.
Simulation model
ECH can drive the radial electron current $j_e$ due to the radial motion of suprathermal electrons [B]. The net current in the steady state should be canceled to maintain the quasi-neutrality, so the return current, $j_r(=-j_e$), flows by the bulk ions by ambipolar condition. Therefore, the bulk plasma feels the $j_r\times B$ torque due to the return current. On the other hand, the suprathermal electrons drift toroidally due to the precession motion. During the slowing down of the suprathermal electrons, they transfer their momenta to the bulk plasma due to collisions[C].
In this study, we investigate the behaviors of energetic electrons by ECH, which can generate the radial current and thus make the $j_r\times B$ torque in the HSX and LHD plasmas. Also, we evaluate the collisional torques, by collisions between energetic electrons and bulk plasma. We apply the GNET code, which can solve a linearized drift kinetic equation for energetic electrons by ECH in 5-D phase space[B]
$ \frac{\partial\delta f}{\partial t}+(\mathbf{v}_d+\mathbf{v}_\|)\cdot \frac{\partial\delta f}{\partial\mathbf{r}}+ \dot{v}\cdot\frac{\partial\delta f}{\partial\mathbf{v}}-C(\delta f)-L(\delta f) =S^{\rm ql}(f_{\rm Max}), $
where $C$, $L$, and $S^{\rm ql}$ are the collision operator, the orbit loss, and the ECH heating source, respectively. We solve the momentum balance equation to evaluate the toroidal flow velocity, and we introduce the $j_r\times B$ torque effect with using Maxwell's equation[D]. we evaluate the toroidal flows driven by ECH and compare them with experimental ones.
HSX plasma
In the perfectly symmetric configuration, the forces in the symmetry direction cancel each other. As seen in Fig. 1, the $j_r\times B$ torque and the collisional torque cancel each other, and the component parallel to the helical symmetry direction is very small in the perfectly helically symmetric configuration[E]. However non-symmetric magnetic modes enhance the radial current $j_e$. Thus, even in the QHS configuration the $j_r \times B$ torque is dominant and there is a net force in the symmetry direction due to other small non-symmetric modes. The force in Mirror configuration is more than twice as large as that in QHS configuration with the same input power. The collisional torque is so small as being negligible in QHS and Mirror configurations. Solving the momentum balance equations with $j_r \times B$ torque, the obtained flow velocity is shown in Fig.2. Here the absorption power calculated by ray-tracing code is 24kW in QHS configuration and 16kW in Mirror configuration. The integrated torque in Mirror configuration is about 4 times larger than QHS configuration even though the less heating power. The obtained flow in QHS configuration has a narrow peak, but the total toroidal flow in Mirror configuration is larger than that in QHS configuration.
LHD plasma
We evaluate the toroidal torques by ECH and the toroidal flow in the balanced NBI heating plasma of LHD, where the torque by the NBI heating is relatively small. Figure 3 (left) shows the toroidal torque by ECH and balanced NBI heating. Here NBI torque is evaluated with FIT3D code, which is a module for NBI heating in TASK3D, the integrated transport code for helical plasmas. We evaluate the toroidal flows driven by ECH torque, solving the 1D radial diffusion equation, and compare them with the experimental observations. As shown in Fig.3 (right), the obtained flows have good agreements with the experimental ones. The toroidal flow velocity is around zero with the balanced-NBI torque, while the flow velocity can reach 20km/s with the additional ECH torque.
[A] S.T.A. Kumar, et al., Nucl. Fusion 60 (2018) 054012.
[B] S. Murakami, et al., Nucl. Fusion 40 (2000) 693.
[C] M. N. Rosenbluth and F. L. Hinton, Nucl. Fusion 36 (1996) 55
[D] M. Coronado, et al., Phys. Fluids B 5 (1993) 1200.
[E] Y. Yamamoto, et al., Plasma Fusion Res. 14 (2019) 3403105Speaker: Yasuhiro Yamamoto (Kyoto University) -
Modest effects of deuterium line radiation trapping in MAST Upgrade tokamak Super-X and snowflake (SF) divertor plasmas are found using SOLPS-EIRENE and UEDGE code divertor plasma modeling (1, 2), and CRETIN code (3) radiation transport and collisional-radiative modeling. In MAST-U, both the Super-X and SF divertor plasmas are predicted to reach highly radiative (detached) regimes at lower upstream densities as compared to standard divertor configurations (1,2). Lyman series line radiation trapping in divertor increases deuterium ionization rate, reduces recombination and radiation volumetric rates. This may increase divertor detachment density threshold and diminish the advanced divertor geometry benefits (4-7). The CRETIN and UEDGE modeling of the detached Super-X and SF divertors shows that 1) divertor plasma properties are modified insignificantly ( $\le 10$ \%) when opacity effects are included; 2) however, deuterium Lyman radiation trapping is non-negligible and may be experimentally detectible. Divertor radiation trapping is likely to play an important role in future reactor-scale divertor tokamaks, including ITER (8). Model validation experiments are planned in MAST-U, as part of a comprehensive advanced divertor research program (9).
Plasma opacity $\tau=L / \lambda_{MFP}$, (where $L$ is spatial scale and $\lambda_{MFP}$ is a photon mean free path (MFP)) scales with $(L n_0)$, where $n_0$ is neutral (atomic) deuterium density.
In the SOLPS and UEDGE models of a 2.5-5 MW NBI-heated 1 MA H-mode plasmas, high $n_0 \le 10^{19}-10^{21} m^{-3}$ are found in detached regimes in Super-X and SF divertor configurations (1,2).
An expanded tightly baffled outer divertor leg plasma volume with a radial spatial scale of $L \sim 60$ cm is present in the Super-X divertor (1).
In the SF divertors, the primary and secondary strike points land on vertical targets in the divertor throat, leading to high recycling neutral fluxes and densities (2).The impact of Lyman line radiation transport on divertor plasma parameters was studied with UEDGE and found to be weak.
A Lyman-$\alpha$ escape factor model in UEDGE (10) introduces corrections to ionization, recombination and electron heating rates calculated by CRETIN based on parametrized optical depth $r_{\tau} \sim \int n_0 \; dx$ (integration is to the nearest boundary), as shown in Fig. 1 (a-c) for the Super-X, SF-plus and SF-minus configurations.
In the Super-X configuration, it is the outer leg that is mostly affected, where the optical depths reach $r_{\tau} \le 1000$.
In the SF-plus and SF-minus, insignificant opacity is predicted in the strike point regions and between the X-points, with the optical depths up to 100.
At detached Super-X divertor conditions in MAST-U, radiation trapping changes neutral and plasma properties, however, these changes appear to be insignificant.
Fig. 1 (d) shows $T_e, n_e$ and $n_0$ changes on the divertor plate normalized to the no radiation trapping case properties.
Fig. 1 (e) shows $T_e$ profiles along the separatrix (along the connection length) between the X-point and the plate for different densities, with and without impurities (carbon), and with and without radiation trapping.
Divertor heat and particle fluxes and detachment trend are weakly affected as well.Detailed continuum and line radiation transfer was studied with CRETIN, and the results demonstrate that whereas Lyman series line radiation absorption, scattering, and reemission take place, the impact on kinetic atomic rates and level populations is modest.
CRETIN was used with plasma and neutral backgrounds modeled by SOLPS and UEDGE for the Super-X configuration (1), and UEDGE for SF configurations (2).
A deuterium atom model with $n \le 20$ and sublevels was used.
In the SOLPS-EIRENE detached Super-X divertor model with 2.5 MW NBI heating, small Lyman-$\alpha$ optical depth $\tau_{\nu} = \int \kappa_{\nu} \rho dx \le 2$ (here $\kappa_{\nu}$ is absorption coefficient and $\rho$ is mass density) were observed in radial and poloidal directions.
In the SOLPS detached Super-X divertor model with 5 MW NBI heating, neutral densities in the outer leg were higher, hence the Lyman-$\alpha$ photon MFP was $\le 10$ cm, leading to some trapping in the Lyman-$\alpha$ line center with modest Lyman-$\alpha$ optical depth $\tau_{\nu} \le 10$ above the outer target plate.
In the UEDGE detached Super-X model with 2.5 MW NBI, higher recycling coefficients and no pumping, neutral densities were higher than in the SOLPS-EIRENE model, and the trapping was stronger ($\tau_{\nu} \le 1000$), leading to visible dips in Lyman series profiles.
In the SF-plus and SF-minus divertor models with detached strike points, the plasma, neutral and radiation fields were highly non-uniform because of the multiple dissimilar magnetic regions.
Lyman-$\alpha$ photon MFP was down to several cm in the strike point regions, resulting in weak line center trapping (optical depth between divertor throat sides (targets) $\tau_{\nu} \le 2$). In both Super-X and SF divertor simulations, the inclusion of radiation transport led to the increased $n\ge2$ atomic level populations (up to 20-100), and a modest reduction (30-50 \%) of Lyman series radiative flux on the divertor plates.The detailed CRETIN radiation transfer calculations used three models for the line shape due to thermal Doppler broadening, Stark broadening due to plasma electron and ion microfields, and the Zeeman splitting due to magnetic field.
At MAST-U divertor conditions, radiation transfer was weakly affected, the differences in Lyman intensities and radiated power were less than 5-10 \% .
However, the Lyman line shapes were significantly altered: while Doppler broadening was less than 0.005 eV, and Zeeman splitting at mostly toroidal field $\sim 0.5$ T was small, Stark broadening at detached divertor densities $\le 10^{20} m^{-3}$ could be potentially measured with spectroscopic diagnostics.In summary, if the present CRETIN and UEDGE predictions for the Super-X and SF divertors are validated in MAST-U experiments, detached plasma properties are not expected to be modified significantly as Lyman-$\alpha$ optical depths are modest. However, if neutral densities are increased, an increased opacity may lead to stronger non-linear plasma effects.
This work is supported by the US DOE under Contract DE-AC52-07NA27344 and the RCUK Energy Programme Grant Number EP/P012450/1 and EURATOM.
(1) D. Moulton et al. 2017 Proc. 44th EPS Conf. (Belfast, UK, 2017)
(2) A. I. Khrabry et al. IAEA FEC 2020 (this conference)
(3) H.A. Scott, Journal of Quantitative Spectroscopy \& Radiative Transfer 71 (2001) 689
(4) H.A. Scott et al. Journal of Nuclear Materials 266-269 (1999) 1247
(5) A.S. Wan et al. Journal of Nuclear Materials 220-222 (1995) 1102
(6) D. Reiter et al. Journal of Nuclear Materials 313–316 (2003) 845
(7) A.A. Pshenov et al. Nucl. Fusion 59 (2019) 106025
(8) B. Lipschultz, Nucl. Fusion 47 (2007) 1189–1205
(9) J. R. Harrison et al. Nucl. Fusion 59 (2019) 112011
(10) H. A. Scott and M. L. Adams, Contrib. Plasma Phys. 44, No. 1-3, 51 (2004)
Speaker: Vsevolod Soukhanovskii (Lawrence Livermore National Laboratory) -
A neural network version of the Multi-Mode Model (MMM) (1), known as MMMnet, has been trained to reproduce the calculation of plasma turbulent diffusivities needed for transport simulations. Model-based control applications require response models with fast (e.g. for closed-loop offline simulations) to extremely fast (e.g. for real-time control and estimation) calculation times, making well-established physics-oriented predictive transport codes challenging or impossible to use in these applications. The control-oriented predictive transport code COTSIM ($\underline{C}$ontrol-$\underline{O}$riented $\underline{T}$ransport $\underline{Sim}$ulator) has been developed to capture the most relevant plasma-response dynamics from a controls perspective while running at a speed useful for control-design applications. In order to achieve this calculation speed, COTSIM relies on simplifying assumptions and scenario-specific models, often leading to a limited range of validity and a lower level of accuracy. Recently, neural-network versions of a number of physics-based codes for both transport (2, 3) and sources (4) have been developed, which significantly reduce the calculation time required while maintaining relatively high prediction accuracies across many different scenarios. Inspired by these recent developments on machine-learning-based plasma-response modeling, a neural network that reproduces the predictions of the Multi-Mode Model for anomalous transport has been developed for DIII-D and is being integrated into COTSIM to enhance its fast prediction capabilities.
The MMMnet model has been trained to reproduce the calculation of the ion thermal ($\chi_i$), electron thermal ($\chi_e$), impurity particle ($D_z$), and toroidal momentum ($\chi_{\phi}$) turbulent diffusivities. A simple artificial neural network structure was used with three hidden layers and 100 nodes per hidden layer. Five separate networks were trained with different initializations of the weights, and the final prediction returned is the average of the five predictions. A low standard deviation between the five predictions indicates that the training is effectively eliminating any randomness introduced by the initial weights. Training data was generated by calling 1000 predictive TRANSP runs based on 83 different DIII-D shots and using MMM as the transport model. Instead of relying on a convolutional neural network to handle spatially-varying data, principle component analysis was applied to the profile data to reduce the number of data points used to describe each profile. This reduced the number of inputs and outputs to the network, thereby limiting the network complexity and calculation time. This version of MMMnet has an average correlation of 86% (see Figure 1) between the neural network prediction and the actual Multi-Mode Model values for data the network did not see during the training process. Figure 2 shows a comparison between the MMM data and MMMnet predictions of all four output profiles for a shot not used in training the network. The red lines are the MMM data and the dark blue lines are the average across the five neural network predictions. The blue shaded area represents one standard deviation above and below the average value predicted by the neural networks. A new version of the MMMnet model is being trained to also include ion particle and poloidal momentum turbulent diffusivities as outputs, making the network useful for many different applications (knowing the ion and impurity densities, the electron density can be computed from the quasi-neutrality condition).
COTSIM is a 1D transport code based on a prescribed MHD equilibrium, although the coupling with a nonlinear Grad-Shafranov solver is under development. It has a modular configuration, which makes adding or removing physics complexity as needed extremely simple. COTSIM can be configured to run by choosing transport and source models from a library of models ranging from empirical scalings to analytical models such as Bohm/Gyro-Bohm and Coppi-Tang. COTSIM is based on Matlab/Simulink, which makes it control-design friendly and capable of running closed-loop simulations and being wrapped by an external optimizer. Moreover, it is capable of simulating full discharges in a time ranging from a fraction of a second to several seconds, depending on the complexity of the models chosen for the simulation, which makes it suitable for effective iterative control design and real-time control applications.
Present efforts are focused on augmenting COTSIM’s model library by incorporating machine-learning- based models for transport and sources. These accelerated physics-based models will enhance the prediction capabilities of COTSIM without compromising its computational speed, while making it more adaptable to different plasma scenarios. In this work, MMMnet is integrated into COTSIM to compute the plasma turbulent diffusivities and enhance the prediction of the toroidal rotation profile and the electron temperature profile, which will in turn improve the prediction of both the resistivity and the toroidal current density profiles. These prediction capabilities play a critical role in rotation profile and current profile control applications. It is anticipated that the machine-learning-based model library will be expanded in the future by integrating into COTSIM additional neural-network models of codes such as TGLF (2), NUBEAM (5), and GENRAY.
This work has been supported by the US Department of Energy under DE- SC0010661, DE-SC0013977, DE-FC02-04E54698, and by the National Science Foundation Graduate Research Fellowship Program (GRFP) under Grant No. 1842163.
(1) RAFIQ, T., KRITZ, A. H., WEILAND, J., PANKIN, A. Y., and LUO, L., Physics of Plasmas 20 (2013) 032506.
(2) MENEGHINI, O., LUNA, C. J., SMITH, S. P., and LAO, L. L., Physics of Plasmas 21 (2014) 060702.
(3) CITRIN, J., BRETON, S., FELICI, F., et al., Nuclear Fusion 55 (2015) 092001.
(4) BOYER, M., KAYE, S., and ERICKSON, K., Nuclear Fusion 59 (2019) 056008.
(5) MOROSOHK, S., BOYER, M., and SCHUSTER, E., Neural-network version of nubeam for real-time control and scenario optimization in DIII-D, in APS-DPP Annual Meeting, 2018.
Speaker: Ms Shira Morosohk (Lehigh University) -
Predicting the dynamics of a burning plasma over long time scales, i.e. comparable with the energy confinement time or even longer, is essential in order to understand modern fusion experiments. Extending first-principle-based gyrokinetic simulations to these time scales is a formidable task from a computational resource point of view. This makes predictive analyses very challenging and calls for reduced descriptions which preserve the necessary physics ingredients. Most of the works addressing this problem for the study of core plasma transport are based on a systematic separation of scales between the reference equilibrium and fluctuations. Meanwhile, energetic particle (EP) transport in fusion devices is a spatiotemporal multi-scale process 1 which can invalidate this assumption leading the system toward the formation of long-lived phase space structures that account for the deviation of the EP distribution function from the local balance between sources and collisions. Furthermore, spatio-temporal mesoscales can be observed in drift wave plasma turbulence simulations. In a recent work [2] we have emphasized the fundamental importance of the self-consistency of the adopted description, including the determination of the characteristic spatiotemporal scales of the reference state. This is mandatory to understand present-day magnetic confinement experiments. In the present contribution we propose a theoretical framework to describe transport in the phase space based of the theory of Phase Space Zonal Structures (PSZS). In particular we extend the usual definition of plasma equilibrium in the presence of a residual level of electromagnetic fluctuations [3] introducing the concept of zonal state and deriving its governing equations. PSZS are long-lived formations in the particle phase space; that is, PSZS are undamped by (fast) collisionless dissipation mechanisms due to wave- particle interactions. Together with zonal structures, i.e. toroidally symmetric structures in, e.g., the scalar potential, they define the zonal state [2]. This definition is particularly important in collisionless burning plasmas, where (as, e.g., for EPs) one cannot readily describe transport via evolution of radial profiles of a reference Maxwellian. In this work, we derive the governing equations for the zonal state using gyrokinetic transport theory. The gyrocenter distribution functions evolve according to the gyrokinetic equation in the presence of sources and
collisions:The gyrocenter distribution functions evolve according to the gyrokinetic equation in the presence of sources and collisions:
$ \frac{\partial}{\partial t}\left(D_{a} F_{a}\right)+\frac{\partial}{\partial \mathbf{Z}} \cdot\left(D_{a} F_{a} \dot{ Z}_{a} \right)=D_{a} \left (\sum_{b} C_{a b}^{g}\left[F_{a}, F_{b}\right](\mathbf{Z}, t)+\mathcal{S}_{a}(\mathbf{Z}, t) \right).$
We assume an axisymmetric reference magnetic field that can be described adopting toroidal flux coordinates $(\theta,\phi,\psi)$. In the absence of fluctuations, collisions and sources, the particle motion is integrable and characterized by three invariants, i.e. the particle energy (per unit of mass) $\mathcal{E}$, the magnetic moment $\mu$ and the toroidal angular momentum $P_{\phi}$. For this reason, we will use $(\theta,\phi,P_{\phi},\mu,\mathcal{E})$ as phase space coordinates. Particle velocity can be decomposed as the sum of two contributions, i.e. $ \dot{ Z}_{a} = \dot{ Z}_{a0} + \delta\dot{ Z}_{a}$, representing, respectively, integrable motion in the reference state and the effect of fluctuations. The operator annihilating the advection in the phase space due to the integrable motion is given by: $ \overline {(...)} = \tau_b^{-1} \oint d \theta (...)/\dot\theta $, with $ \tau_b = \oint d \theta/\dot\theta $. The governing equation for PSZS dynamics is obtained by applying this operator to the toroidally symmetric component of the kinetic equation and extracting its low frequency component with respect to the hydrodynamic time scale which is indicated by the $S$ subscript:
$ \frac{\partial}{\partial t}\overline{F_{z0}}+\frac{1}{\tau_{b}}\left[\frac{\partial}{\partial P_{\phi}}\overline{\left(\tau_{b} \delta \dot{P}_{\phi} \delta F\right)_{z}}+\frac{\partial}{\partial \mathcal{E}}\overline{\left(\tau_{b} \delta \dot{\mathcal{E}} \delta F\right)_{z}}\right]_{S}= \overline{ \left(\sum_{b} C_b^{g}\left[F_, F_{b}\right]+\mathcal{S}\right)_{z\,S}} $
This expression describe transport processes in the phase space due to fluctuations and collisions/sources. Transport equations can be obtained taking its moments in the velocity space. As an example, in Ref. [2] we have calculated the governing equation for the radial density profiles. It can be shown that this approach is consistent with the usual evolution of macroscopic plasma profiles under the action of fluctuation induced fluxes, when the deviation of the reference state from local Maxwellian response is small. In particular, classical and neoclassical transport regimes are recovered in the proper limits. In the general case intermediate spatio-temporal scales can be developed by the zonal state. As an application, we calculate PSZS dynamics for an EPM (collisionless) simulation by the HMGC code. In particular, in the left panel of the following figure, we show a “slice” with fixed $\mu$ of $\overline{F_{z0}}$ during a phase space avalanche produced by an EPM at a given instant. In the right panel also $\mathcal{E}$ is fixed, consistently with the fact that, for the quasi-coherent spectrum typical of Alfvénic fluctuations, transport on a short time-scale is mainly a $1D$ (radial) process. In this case, PSZS have been depicted before and after the avalanche respectively in black and in blue.
The usefulness of this formulation becomes clear for long time scale calculations, in particular those related with gyrokinetic or hybrid simulations of EP transport, where the non-Maxwellian features and the role of wave-particle resonances are most important[@biancalani2019gyrokinetic]. We will discuss how nonlinear equilibrium evolution on transport time scale can be calculated by coupling a transport code solving the previous equation in a reduced $3D$ phase space to a $5D$ gyrokinetic code directly evaluating the expressions for nonlinear fluxes.
1 L. Chen and F. Zonca. Reviews of Modern Physics, 88(1):015008, 2016.
[2] M. V. Falessi and F. Zonca. Physics of Plasmas, 26(2):022305, 2019.
[3] M. Rosenbluth and F. Hinton. Physical Review Letters, 80(4):724, 1998.
[4] A. Biancalani et al. arXiv preprint arXiv:1912.09409, sub. to Physical Review Letters.Speaker: Dr Matteo Valerio Falessi (ENEA) -
The lower hybrid current drive (LHCD) system in WEST plays a key role for achieving long pulse operation and high performance plasmas. Up to 5 MW LHCD power has been coupled in WEST plasmas in both Lower Single Null and Upper Single Null configurations, and pulse duration of 55 s has been sustained with 3 MW of LHCD [1]. The LHCD power is launched into the plasma by two multijunction launchers (Full-Active-Multijunction and Passive-Active-Multijunction), previously used in Tore Supra.
Prior to the start of WEST, the front face of the Full-Active-Multijuntion (FAM/LH1) launcher, was reshaped in the toroidal direction in order to fit the toroidal curvature of the WEST plasmas [2], since the toroidal ripple in WEST is smaller than that in the former Tore Supra. This contribution now presents a detailed experimental analysis of the LH wave coupling in each multijunction module on the two launchers, i.e. the reshaped FAM/LH1 launcher and the non-reshaped Passive-Active-Multijunction (PAM/LH2) launcher, in different plasma configurations and at different edge electron densities. Modelling with the ALOHA coupling code [3] has been carried out, based on the experimental values of power and phase in each multijunction module. Fig. 1 shows the power reflection coefficient (RC) in each module on the upper half of FAM/LH1 versus the density measured by a Langmuir probe on the upper part of the LH1 launcher, for a database of 67 pulses. The ALOHA modelling for the upper part of LH1 using two density gradients ($\lambda_1$= 1 mm; $\lambda_2$ = 15 mm) is also plotted. Fig. 1 illustrates that low RC are obtained on all modules for the same edge density, meaning that the edge density is homogeneous along the toroidal direction and thus that the reshaping of the launcher mouth has been efficient.
In contrast, the RC on the individual modules on the PAM/LH2 launcher behave differently depending on their toroidal location (Fig. 2). An increase in RC with increasing edge density is seen on the lateral modules, while the opposite behavior is observed on the central modules. A comparison with ALOHA is not shown in Fig. 2, because there is no single ALOHA-run that gives agreement with the experimental values in all modules simultaneously. The ALOHA-calculations suggest that the edge density in front of the lateral modules is higher than that in front of the central modules, which is consistent with the fact that the LH2 launcher has not been reshaped. The toroidal asymmetry is also evident during long pulse operation (up to 55 s), where a local overheating of the edge modules is seen using the infrared monitoring [4]. A reshaping in-situ of the LH2 lateral modules is therefore foreseen in view of WEST phase 2 aiming at 1000 s pulses.
Nevertheless, good coupling and low RC (< 5%) on all modules on LH1 and LH2 can be obtained by adjusting the plasma and launcher positions and by tuning the plasma shaping with feedforward control of the upper and lower gaps (at ± 25 cm with respect to the mid-plane). Local gas puffing systems, located near the launcher side protections and radially retracted by ~ 20 cm, are available but have not been exploited significantly, since the weak pumping capability in WEST does not allow large flexibility gas fuelling. It can be noted that the LH wave coupling tends to improve when increasing the injected LHCD power, an effect that can be attributed to the ionization of the neutrals in the scrape-off-layer. SOLEDGE-EIRENE [5] simulations indicate that the neutral pressure at the radial location of the LH launchers in WEST is rather high (~ $10^{-3}$ Pa), which supports this hypothesis. This behavior is quite the opposite to that observed in Tore Supra, where the ponderomotive force effect could lead to a degradation in LH coupling when increasing the injected LHCD power [6].
References
[1] WEST overview paper, this conference.
[2] L. Delpech et al., Fusion Eng. Des. 96-97 (2015) 452-457.
[3] J. Hillairet et al., Nucl. Fusion 50 (2010) 125010.
[4] X. Courtois et al., this conference.
[5] G. Ciraolo et al., Nuclear Materials and Energy 20 (2019) 100685.
[6] M. Preynas et al., Nucl. Fusion 53 (2013) 013012.Speaker: Anshu Liang (CEA/IRFM) -
We investigate the merging processes of spherical tokamak (ST)-type plasmoids confined in a conducting vessel by means of a particle simulation. For this purpose, (i) we have developed a new particle simulation model to analyze physics not only near the contact point of two STs, i.e., the reconnection point, but also in the entire region of a poloidal surface. We find that (ii) plasma heating occurs in a wide region, and kinetic processes which can be expressed as compressional heating and viscous heating from macroscopic viewpoints play main roles. By performing simulation runs for different ion-to-electron mass ratio $m_i/m_e$, we clarify that (iii) thermal energy partition of ions and electrons approaches $\simeq 3:1$, and the component perpendicular to the magnetic field for ions and the parallel component for electrons are dominantly heated, respectively.
The merging of STs attracts attention as a candidate of future fusion reactors, since STs can confine higher-$\beta$ plasmas than standard tokamaks or helical devices. In the merging of STs, two STs with $\beta \sim 5 \%$ are merged through magnetic reconnection into a single ST with $\beta \sim 40 \%$ [A], So far, plasma heating in a local region near the contact point was focused on. Recently the advance of experimental methods such as 2D imaging measurement enables us to observe fine structures of heating and transport in a global region [B], but the mechanism remains to be unsolved. The elucidation of the detailed mechanism can lead to the higher performance of future ST reactors. For this purpose, particle simulation is one of the indispensable and powerful tools.
We investigate the merging processes of ST-type plasmoids by means of a particle simulation which we have newly developed [C]. Figure 1 shows the schematic diagram of our simulation model. The left part displays two plasmas in an ST device and the right part displays a poloidal plane simulated by the new particle code in a confinement system. The toroidal magnetic field $B_t$ exists in the simulation domain. The two plasmoids have the same plasma rotation in the poloidal plane to sustain the toroidal current density with the same sign. We set $\omega_{pe}/\omega_{ce} = 1.72$ and $\beta= 10 \%$ for the total magnetic field $B_0$, and $m_i/m_e=400$. The system size is $(7.46 \times 29.8) \rho_i$, where $\rho_i$ is the ion gyroradius for $B_0$.
The different points between the new [C] and previous models [D] are as follows. In previous works, particle simulations of reconnection have been carried out under an open boundary condition. Thus, energy for reconnection is supplied from the outside of the simulation domain [D]. Such models are suitable to simulation studies of magnetic reconnection in a local system. In contrast, the new model employs a closed system confined in a conducting vessel. In this model, most plasmas and energy are confined inside a newly formed ST, and energy available for merging process is limited. Thereby, magnetic reconnection is time-dependent and impulsive. In 2019, by the use of the new code, we have demonstrated a sharp peak impulsively appearing in the electron temperature profile[C], which can account for an experimental result in MAST[E].
In Fig. 2, we show the time evolution of poloidal magnetic field lines and toroidal magnetic field (color contours). Initially two STs satisfy an MHD equilibrium condition, but they are in an unstable high-energy state. Thus, the two STs begin to approach each other by an attractive force $J \times B$ (Fig. 2 (a)). Magnetic reconnection occurs at the almost center position between the two STs with oppositely directed velocities (Fig. 2 (b)). Through magnetic reconnection, the two STs are merged into one large ST, as shown in Fig. 2 (c).
We observe ion heating occurring in a wide region, which qualitatively supports an experimental result in TS-6 [B]. During merging of two STs, the conversion from plasma kinetic energy of high-velocity bulk flow $u$ into thermal energy takes place through kinetic processes. These kinetic processes are macroscopically expressed as the compressional heating term ($H_{\rm cmp} =-5/2 P \nabla \cdot u$) and viscous heating term ($H_{\rm vis}=- \pi_{i,j} \partial_i u_j$), where $P$ and $\pi_{i,j}$ are scalar and off-diagonal components of the stress tensor, respectively. Figure 3 shows the profiles of $H_{\rm cmp}$ and $H_{\rm vis}$. At $\omega_{ce}t=2391$ (Fig. 3(a)), compressional heating is dominant, because the confinement region shrinks by the merging of the two STs. At $\omega_{ce}t=3120$ (Fig. 3 (b)), viscous heating is dominant near the central point. The merging system has a net angular momentum around the geometric axis, and thus large velocity shear is produced near the contact point, in which a mixing process takes place between particles in different origins through a kinetic effect. As a result of the heating, in this case, the maximum of $\beta$ is increased to $\simeq 20\%$.
Lastly, we carry out various simulation runs for different values of $m_i/m_e$, where the initial ion parameters vary depending on $m_i/m_e$ while keeping the initial electron parameters constant in order to clarify the roles of the ion dynamics. Figure 4(a) shows that its partition rate approaches $\simeq 3:1$ as $m_i/m_e$ increases. Furthermore, we focus on the temperature anisotropy. According to Fig. 4(b), ions are heated dominantly in the direction perpendicular to the magnetic field, while electrons are heated mainly in the parallel direction. This is mainly because electrons moves along $B_t$ in the vicinity of the reconnection point, while two groups (STs) of ions with fast convergent flows $\pm u_y$ collide across $B_t$[C].
[A] Y. Ono et al., Plasma Phys. Control. Fusion {\bf 54} (2012) 124039.\
[B] H. Tanabe et al., Nucl. Fusion {\bf 59} (2019) 086041. \
[C] R. Horiuchi, et al., Phys. Plasmas {\bf 26} (2019) 092101. \
[D] S. Usami, et al., Phys. Plasmas {\bf 26} (2019) 102103. \
[E] H. Tanabe et al., Nucl. Fusion {\bf 57} (2017) 056037.Speaker: Shunsuke Usami (National Institute for Fusion Science) -
A natural way to control turbulence in magnetic fusion devices is to take advantage of zonal flows, which form spontaneously and can reduce the turbulence level. Zonal flows can even suppress turbulence completely in a certain parameter range where drift-wave instabilities would otherwise develop. But exploiting this effect, known as the Dimits shift, requires understanding of its physics, which has been unclear. Here, a generic understanding of the Dimits shift in electrostatic drift-wave turbulence is obtained, for the first time, by studying the tertiary instability of a zonal flow within reduced turbulence models (Ref. 1). We show that tertiary modes are localized near extrema of the zonal-flow velocity $U(x)$ with respect to the radial coordinate $x$. These modes can be described as quantum harmonic oscillators with complex frequencies, so their spectrum can be readily calculated. The corresponding growth rate $\gamma_{\rm TI}$ is derived within the modified Hasegawa--Wakatani model. We show that $\gamma_{\rm TI}$ equals the primary-instability growth rate plus a term that depends on the local flow "curvature" ${\rm d}^2U/{\rm d}x^2$; hence, the instability threshold is shifted compared to that in homogeneous turbulence. This shift is the Dimits shift, which we find explicitly in the Terry--Horton limit, and our analytic predictions agree well with results of numerical simulations. Our theory of the tertiary instability also extends to other turbulence models. For example, the key features of the tertiary instability of ion-temperature-gradient mode are reproduced by our theory and verified by gyrokinetic simulations.
Plasma microturbulence causes anomalous transport of particles and energy and sets the core plasma profile in tokamaks, where the underlying "primary" drift-wave (DW) instabilities develop if the plasma temperature and (or) density gradient exceed a certain threshold. However, having their linear growth rate $\gamma_{\rm PI}$ above zero is not enough to make plasma turbulent, because the "secondary instability" can suppress turbulence by generating zonal flows (ZFs); hence, the threshold for the onset of turbulence is modified compared to the linear theory. This constitutes the so-called Dimits shift (Ref. 2), which has been attracting attention for two decades. The finite value of the Dimits shift is commonly attributed to the "tertiary instability" (TI) (Ref. 3), and some theories of the TI have been proposed. However, basic understanding and generic description of the TI and the Dimits shift have been elusive.
Here, we propose a quantitative theory of the TI based on reduced models of DW turbulence; we also use this theory to explain the Dimits shift and, in some cases, even to calculate it semi-analytically (Ref. 1). We start by noticing that according to several numerical studies (Refs. 3-5), DW turbulence in the Dimits regime is localized near extrema of the ZF velocity $U(x)$. Assuming such localization, we derive an approximate equation that governs the localized DW modes. This equation coincides with that of a quantum harmonic oscillator with a complex frequency; hence, the mode structures and the growth rates are readily found analytically. Having $\gamma_{\rm TI} = 0$ corresponds to the threshold at which plasma becomes turbulent. This threshold is shifted relative to that of the primary instability in homogeneous turbulence, and that is precisely the Dimits shift.
Using the modified Hasegawa--Wakatani model, we calculate $\gamma_{\rm TI}$ analytically (Fig. 1) and find that the TI has features different from those that are commonly expected. First, the TI modes are different near maxima and minima of the ZF velocity and has a broad Fourier spectrum (in $x$); hence, they cannot be described adequately within the few-harmonic approximation, contrary to some previous theories. Second, the growth rates can be written as $\gamma_{\rm TI}=\gamma_{\rm PI}+\Delta\gamma(\mathcal{C})$, where $\mathcal{C}\doteq {\rm d}^2U/{\rm d}x^2$ is the local ZF "curvature", and $\Delta\gamma$ becomes zero in the absence of the ZF. Therefore, the TI can be considered as a primary instability modified by the ZF curvature. The fact that $\Delta\gamma$ is controlled by the ZF curvature is in variance with the conventional understanding that turbulence is regulated by the ZF shear $|{\rm d}U/{\rm d}x|$. Also contrary to what is commonly assumed, the TI is distinct from the Kelvin--Helmholtz instability. In the modified Hasegawa--Wakatani model that we use, Kelvin--Helmholtz modes are stabilized by the adiabatic electron response, while the TI is still present. Similar features of the TI are found within the ion-temperature-gradient-mode model introduced in Ref. 3. This has been tested in slab-geometry gyrokinetic simulations using the code $\mathsf{GS2}$ (Ref. 7).
Understanding the growth rate $\gamma_{\rm TI}=\gamma_{\rm PI}+\Delta\gamma(\mathcal{C})$ provides a generic explanation of the Dimits shift. We have calculated $\Delta \gamma(\mathcal{C})$ explicitly within the modified Terry--Horton model (as the adiabatic limit of the Hasegawa--Wakatani model) and used those results to determine the Dimits shift in this model. As seen in Fig. 2, our theory shows good agreement with simulations.
In summary, our work provides a generic qualitative understanding of the Dimits shift and also leads to (semi-) analytic predictions of this shift within a number of reduced models of DW turbulence.
This work was supported by the US DOE through Contract No. DE-AC02-09CH11466. This work made use of computational support by CoSeC, the Computational Science Centre for Research Communities, through CCP Plasma (EP/M022463/1) and HEC Plasma (EP/R029148/1).
(1) H. Zhu, Y. Zhou, and I. Y. Dodin, Phys. Rev. Lett. 124, 055002 (2020).
(2) A. M. Dimits et al., Phys. Plasmas 7, 969 (2000).
(3) B. N. Rogers, W. Dorland, and M. Kotschenreuther, Phys. Rev. Lett. 85, 5336 (2000).
(4) S. Kobayashi and B. N. Rogers, Phys. Plasmas 19, 012315 (2012).
(5) R. Numata, R. Ball, and R. L. Dewar, Phys. Plasmas 14, 102312 (2007).
(6) D. A. St-Onge, J. Plasma Phys. 83, 905830504 (2017).
(7) M. Barnes et al., GS2 v8.0.2, https://doi.org/10.5281/zenodo.2645150.Speaker: Hongxuan Zhu -
The Joint-Texas Experimental tokamak (J-TEXT), formerly named TEXT/TEXT-U in the University of Texas at Austin in USA, has been reconstructed and obtained the first plasma in the Huazhong University of Science and Technology in the spring of 20071. In J-TEXT tokamak, the upgraded resonant magnetic perturbation (RMP) system and the shattered pellet injection system were built, which allows flexible study of 3D effects and disruption mitigation in a tokamak2. At present, a typical J-TEXT discharge is a limiter configuration with three moveable poloidal rail limiter targets[3], which limits the extension of its research achievement to the advanced tokamak with a divertor configuration. At Austin, an inner poloidal divertor has been designed in TEXT for the interaction with the theoretical study of particle and energy studies in the H-mode[4]. Fortunately, the divertor windings are retained in the reconstructed J-TEXT and also the upgrade vacuum vessel for divertor. In order to upgrade the operation region of J-TEXT to the divertor configuration, and even H-mode, the recovery of the divertor system was put forward in 2016. Along with the establishment of divertor power supply, the construction of relevant diagnostics, and installation of the divertor target in high-field side, the divertor discharge has been tested from the end of 2018. And through the equilibrium calculation and position stability analysis, the control strategy has been evolved to be more stable.
Base on a series of preparations such as the divertor power supply, the divertor target and relevant diagnostics, the divertor configuration discharge has been realized for the first time in J-TEXT tokamak. There are two functional components of divertor windings: the divertor coils and the auxiliary bias coils. With different current direction in these coils, they can afford four baseline divertor configurations: middle single-null(MSN), double-null(DN), upper single-null(USN) and lower single-null(LSN), as shown in the figure 1. To realize these different configurations, the power system is designed as a four-quadrant operating power supply and the DV current has achieved 25kA with a flat top of 500ms. In order to handle the larger power in the future research, the inner chamber wall on the high-field side has been fully covered with tiles carefully aligned to toroidal symmetry.
With the help of the equilibrium EFIT code and spool model, both the equilibrium and the position instability have been analyzed. The result shows that, as a special baseline configuration with a divertor at the high-field side, the MSN case faces a serious in-out radial instability. Nevertheless, this kind of instability is not only relevant to the iron core but also to the equilibrium plasma position. The field decay index is closer to the stable region when the plasma is near the low field side. In the DV configuration experiment, the low-field limiter is moved outward firstly. Then in the discharge, the horizontal position(X) of plasma is pushed outward to 3cm away from the routine center of the limiter configuration step by step before the ramp-up of the DV current. Finally, according to the relevant diagnostics and an equilibrium flux distribution reconstruction, a controllable MSN divertor configuration is formed with a X-point at Z=0cm and R=26.6cm, 3cm away from the divertor target surface.
High density experiment and auxiliary heating experiment have been attempted in the divertor configuration. In the former, the central density of 3.8×10^19 "m" ^"-3" is achieved, which almost exceeds the Greenwald density without the consideration of the elongation. The MSN divertor configuration exhibits to be more stable in the density limit condition in this experiment. In the later, ECRH injection enhances the electron temperature and density, while more heat outflux is loaded on the divertor target tiles and causes intensive recycling and serious release of impurity, as shown in the figure 2.
The future plan of the divertor configuration discharge in J-TEXT tokamak includes three parts. The first part is to improve the ability to remain a stable DV configuration. A more reliable feedback control system is needed for the in-out radial instability. The compensation technology for the VF in the ramp-up phase of the DV current is being in our considering. And a comprehensive simulation is going to be designed and guide the optimization of the control parameters. The second part is the deeper investigation on the MSN configuration, especially the stable operation region, the confinement improvement and the transport in scraped-off layer(SOL). The disruption of density limit and related magnetohydrodynamic activity will be explored. The cooperation with ECRH and other auxiliary system will be carried out for the aim of enhancement of the confinement. Meanwhile, with the accessibility of J-TEXT, the turbulence transport evolution at the edge of plasma will be studied. The last but not least, the realization of other three baseline divertor configurations is on the way. A new biasing power supply would be test in the 2020 campaign.Reference:
1 G. Zhuang, Y. Pan, X. Hu, Z. Wang, Y. Ding, M. Zhang, L. Gao, X. Zhang, Z. Yang, K. Yu, The reconstruction and research progress of the TEXT-U tokamak in China, Nuclear Fusion, 51 (2011) 094020.
2 Y. Liang, N. Wang, Y. Ding, Z. Chen, Z. Chen, Z. Yang, Q. Hu, Z. Cheng, L. Wang, Z. Jiang, Overview of the recent experimental research on the J-TEXT tokamak, Nuclear Fusion, 59 (2019) 112016.
[3] Y. Jie, C. Zhipeng, L. Hai, W. Tong, Z. Mingcong, S. Zebao, W. Zhijiang, G. Zhuang, D. Yonghua, The application of limiter target electrostatic measurement system in J-TEXT tokamak, Plasma Science Technology, 21 (2019) 105105.
[4] P. Edmonds, E. Solano, A. Wootton, D. Gao, X. Mao, G. Li, W. Zhu, The design of an inner poloidal divertor for the TEXT Tokamak, Fusion technology, 1 (1989)
Speaker: Dr Zhipeng Chen (Huazhong University of Science and Technology) -
Understanding the properties of micro turbulence driven transport and the regulation mechanism is critical in magnetic fusion plasmas. The local and quasilinear theory demonstrates diffusive microscopic turbulent transport, which follows the Fick’s law relating the transport linearly to the gradient, i.e. Q=-nχ∇T and predicts the gyro-Bohm scaling of transport. However, this conventional localized turbulence and quasilinear calculation of transport fails to address deviations from the expected gyro-Bohm transport scaling observed in tokamak plasmas. To understand the breakdown of Fick’s law, non-local and non-diffusive transport mechanism such as avalanching is introduced1. Avalanching is a self-organized criticality (SOC) intrinsic to the systems exhibiting self-similarity, i.e., the spectral power law scaling S(f)~1/f. Here S(f) is the spectral density with f being the frequency. Avalanches are in mesoscopic scale, showing extended and collective fluctuation events, i.e., intermittent bursts.
Avalanches have been commonly observed in various flux-driven fluid and full-f gyrokinetic simulations. Recently, in δf gyrokinetic simulations of collisionless trapped electron turbulence with gKPSP code2, avalanches are observed in electron (ion) heat and particle transport channels. In those simulations, it shows that long stationary and mesoscale zonal flow staircase-like structures can form localized barriers to regulate transport avalanches. The mean zonal flow plays crucial roles in the regulations of avalanching events. Direct measurement of the mean zonal flow global profile is not trivial in experiments, and zonal flow staircase-like structure can only be inferred by the coherence length measurement of the turbulent fluctuations3.
On the other hand, avalanches have been rarely reported in tokamak experiments. This is partially due to the lack of diagnostic tools with sufficient spatio-temporal resolution, and partially due to the difficulty to suppress MHD instabilities which usually dominate the global transport. Recently, an experiment with MHD instabilities suppressed in KSTAR L-mode plasmas reports the non-diffusive avalanche like transport in electron heat channel[4] with a power law scaling S(f)~f^(-0.7). In this experiment, it shows that radial corrugations in the mean electron temperature profile are in a scale ∆~45ρ_i. Zonal flow staircase is believed to be responsible to generate the radial corrugations. There exist evidence from the experiment that implies dynamical interaction between avalanches and global mean flow structures, as shown in Fig. 1. In the figure, it shows before a large scale avalanching event occurs, the transport is regulated and the electron temperature fluctuation radial profile is corrugated. The mean zonal flow shearing could be responsible for this regulations. However, due to the difficulty to measure mean zonal flow from the experiment, regulations of avalanches and formation of radial corrugations in the temperature profile by mesoscale zonal flow shearing cannot be observed directly from experiments. To address this issue, we adopt the gyrokinetic simulations to study this L-mode plasma using gKPSP code[5,6].
With profiles and configurations from this KSTAR L-mode plasma as inputs, gyrokinetic simulation successfully reproduces the experimental observations. From the simulation, we obtain the power law scaling of electron temperature fluctuations as |δT_e (f)|^2~f^(-0.7) (Fig. 2) and corrugations in the radial profile of mean δT_e with a scale ∆~40ρ_i (Fig. 3), all of which have very good agreement with the experimental measurements. Calculation of the mean zonal flow profile from simulation directly demonstrates the generation of zonal flow staircase and its shearing effects in the regulation of avalanches and temperature corrugations.
1 Hahm T. S. and Diamond P. H. 2018 J. Korean Phys. Soc. 73 747-92
2 Qi L., Kwon J.M., Hahm T.S., Yi S. and Choi M.J. 2019 Nucl. Fusion 59 026013
3 Dif-Pradalier G., et. al., 2015 Phys. Rev. Lett. 114 085004
[4] Choi M. J., et al., 2019 Nucl. Fusion 59 086027
[5] Qi L., Kwon J., Hahm T.S. and Jo G. 2016 Phys. Plasmas 23 062513
[6] Kwon J.M., Qi L., Yi S. and Hahm T.S. 2017 Comput. Phys. Commun. 215 81–90Speaker: Lei Qi (National Fusion Research Institute of South Korea) -
Experiments to stabilize the vertical position of the plasma were carried out in a small tokamak device, PHiX. By the experiments, we demonstrated for the first time in the world that the combined magnetic field generated with saddle coils (SCs) and toroidal magnetic field coils (TFCs) could stabilize the vertically unstable positions of the plasmas with nearly circular cross-sections. The results provide a new avoidance scheme of VDE (vertical displacement event) caused by disruptions which is a long-standing problem.
Generally, to achieve a high beta tokamak, it is necessary to elongate the plasma cross-section. Tokamak plasmas with higher elongation ratio $\kappa$, however, require more precise feedback control. In case of disruptions, plasma control failure leads to VDE, which inevitably limits $\kappa$. It has been reported that the vertical position can be stabilized by applying three-dimensional helical magnetic fields to tokamak plasmas in CTH [1], JIPPT-II [2], and the tokamak device that tested at Bell Laboratories [3]. Furthermore, it has been reported that the horizontal position was stabilized by the helical magnetic field in TOKASTAR-2 [4]. When normal continuous helical windings are used on such as CTH and JIPPT-II, however, an inevitable increase in the aspect ratio, raises the device size, and drops the $\beta_{\rm t}$ limit. Also, it is difficult to install and manufacture continuous helical windings that pass through the inner side of the torus. Therefore, in this study, we have a helical magnetic field with SCs installed on upper, lower and outer side of the torus just like the semi-stellarator windings at Bell Laboratories to stabilize the vertical position of plasma. In contrast to the semi-stellarator windings with complicated shapes, such SCs are easy to install even on existing tokamaks and can be equipped on DEMO reactors.
PHiX is a small tokamak device with a small plasma minor radius of 0.09 m, a major radius of 0.33 m, and a toroidal magnetic field of 0.091 T at the vacuum vessel center. This device was manufactured to demonstrate the stabilization of the vertical position of the plasma by an external three-dimensional magnetic field. The vertical position of the plasma of PHiX is unstable even when $\kappa\sim1$. Since vertically long vacuum vessel is adopted in PHiX, shell effects against vertical movements are expected to be little. In addition, the iron core is used and the n-index $(=-R/B_{Z}\cdot\partial B_Z/\partial R)$ of the vertical magnetic field is negative. Figure 1 shows the arrangement of SCs used in this experiment on PHiX and the Poincare plot of the combined magnetic field by SCs and TFCs. The pseudo-stellarator field is produced by rotating the direction of the magnetic field generated by SCs. In this arrangement, the Poincare plot of the poloidal section shows a magnetic field line structure curved in the direction of positive n-index. The experiment was performed by comparing tokamak discharges with and without SCs excitation. No feedback control of plasma position and plasma current was conducted. The timing of energizing SCs was set 1 ms after tokamak breakdown, and any other experimental conditions were identical. Figure 2 shows the plasma responses in the experiment mentioned above. The discharge duration prolonged as SCs current, $I_{\rm SC}$ increased. When $I_{\rm SC} = 1.9 \:\rm{kAturns}$, the discharge could be sustained until the iron-core was saturated. The plasma position measured with magnetic probes in Fig. 2 (c) shows that the vertical upward displacement vacuum vessel was suppressed as $I_{\rm SC}$ increases. The line integrated electron density measured on the equatorial plane decreases as vertical position of the plasma moves from the center of the vacuum vessel. When $I_{\rm SC} = 1.9 \:\rm{kAturns}$, however, the value fluctuated little, indicating that the plasma remained at the center of the vacuum vessel. In addition, the high-speed camera images in Fig. 3 also support the suppression of vertical displacement of the plasma with an increase in $I_{\rm SC}$. From the above results, it was clearly shown that SCs have the effect of suppressing the vertical displacement when $I_{\rm SC}$ is large enough, and that the displacement could be moderated even when the $I_{\rm SC}$ was small. We also confirmed that the $I_{\rm p}$ direction opposite to that of TFCs destabilizes the plasma position on the contrary. And our computation with VMEC shows that the toroidally averaged elongation ratio $\kappa_\rm{ave}$ is 1.1 without SCs and 1.2 with $I_{\rm SC} = 1.9 \:\rm{kAturns}$ in the cases of discharges shown in Figs. 2 and 3.
[1] M. C. ArchMiller et al., Phys. Plasmas $\bf 21$, 056113 (2014)
[2] K. Sakurai and S. Tanahashi, J. Phys. Soc. Jpn. $\bf 49$, 2, 759 (1980)
[3] H. Ikezi and K. F. Schwarzenegger, Phys.Fluids $\bf 22$, 10, 2009 (1979)
[4] K. Yasuda et al., Plasma Fusion Res. $\bf 13$, 3402072 (2018)Speaker: Mr Shin Naito (Institute of Innovative Research, Tokyo Institute of Technology) -
The early phases of tokamak plasma current ramp-up are often of very short duration and of little concern on current devices. On ITER, the combination of costly, actively cooled plasma-facing components (PFC) and relatively long timescales ($\sim 10 s$) before the transition to X-point configuration, means that power flux management is key if PFC lifetime is not to be compromised. This paper will provide a comprehensive description of the strategies being put in place at ITER to ensure that this critical phase of each plasma discharge is properly managed.
As in many present tokamaks, early ramp-up on ITER will be performed in limiter configuration on the central column, benefiting from the proximity to the resonance location of Electron Cyclotron (EC) start-up assist and lower 3D stray fields produced by currents induced in the vacuum vessel, or due to port openings and ferritic inserts on the low field side $[1]$. Simulations of plasma magnetic control, performed with the DINA code, are used to design the ramp-up phase $[2]$. The transition to divertor configuration is typically made $\sim10 s$ after breakdown when $I_{p} \sim 3.5 MA$, and usually assumes some EC heating power at the level of a few MW $[2]$. Assuming a simple scrape-off layer (SOL) model for parallel heat flux, and taking into account the shaping of the beryllium first wall panels (FWP), this kind of scenario satisfies the constraints of acceptable FWP heat loads and minimizes poloidal flux consumption such that the burn duration will be maximized at high fusion gain.
It has, however, recently become clear that the near SOL heat flux channel width, $\lambda_{q,near}$, may be much narrower than previously thought $[1]$, posing a problem for wall heat loading if FWP alignment is not tightly controlled, and/or the power conducted into the SOL ($P_{SOL}$) is too high. Although multi-machine scalings of $\lambda_{q,near}$, and of the main SOL heat flux width, $\lambda_{q,main}$, have been used to optimize the inner wall FWP toroidal shaping $[1]$, so called longwave (LW) departures from lack of concentricity of the FW with the toroidal field (TF), significantly increase FWP heat fluxes over the expected values. The current engineering specification for blanket assembly assumes an n=1 LW misalignment and requires that the inboard FW be aligned to a target of $\Delta_{LW}$ = $\pm 5 mm$. In addition, the ITER Heat Load Specifications $[3]$ set maximum values of $I_{p} = 5 MA$ at $P_{SOL} = 5 MW$ for inboard limiter plasmas.
Detailed examination of this situation (see Fig. 1), using 3D field tracing, together with the expected radial SOL heat flux profile and properly accounting for power sharing on the inner wall, shows that for $\Delta_{LW}= 5 mm$, the maximum allowable stationary surface heat flux on the inner midplane FWPs will be largely exceeded when additional penalties are included for FWP front surface faceting and tilting. The situation can be further worsened if the assumed SOL transport turns out to be more severe than expected (corresponding to high values of parameter $R_{q}$ in Fig. 1). This analysis suggests the baseline LW (n=1) requirement may have to be tightened to a new, more challenging target of $\Delta_{LW} \sim \pm 3 mm$ to provide sufficient margin for heat loads.
To ensure that inboard FWP power loading during current ramp-up can be properly managed, ITER is adopting a 3-fold strategy:
- Continuous refinement of the accuracy with which the FW can be aligned to the Tokamak Axis Datum (TAD) on the basis of as built and as-assembled components.
- Development of a diagnostic system for precision measurements of the TF structure.
- Study of alternative current ramp-up schemes.
Concerning 1), statistical tolerance accumulation analysis based on a large ensemble of “virtual tokamak builds” finds that, for the case of an n=1 LW misalignment (which assumes the presence of a magnetic “centreline”), the baseline $\Delta_{LW} = 5 mm$ criterion can be met if the displacement between the centreline and the TAD is below $3-4 mm$. If the LW requirement is reduced to $\pm 3 mm$, alignment to TAD cannot be achieved. However, if a direct magnetic measurement of the TF structure at the inboard equatorial region can be made with an accuracy better than $\sim 1.5 mm$, then the tighter target can be met.
In fact, finite element (FE) simulations of the energization and locking of TF coils show that the real perturbation of the TF structure is likely to be more complex than a simple n=1. In which case the new alignment target will be local rather than global, making a measurement of the field structure even more important. Extensive design activities are now underway for the provision of such a measurement using an array of nuclear magnetic resonance (NMR) sensors (Fig. 2). They will be deployed during the First Plasma commissioning phase (before installation of the main blanket) and target a measurement at half nominal $B{\phi}$ (2.65 T). An analytic model has been developed to guide this TF Mapping diagnostic design. Using the cylindrical approximation, and tested against full 3D numerical simulations, it yields the perturbed field at given radial location produced by an ensemble of misaligned TF coils. Taking as an example the realistic field structure generated from the FE coil locking simulations, the model shows that a set of 18 toroidally distributed NMR probes in the region of the inboard vacuum vessel wall will be sufficient to reconstruct the TF structure within an error of $\pm 0.5 mm$ at the radial position of the key start-up FWPs. Estimates at this stage of the diagnostic design anticipate that the target $\sim 1 mm$ measurement accuracy in the spatial location of the chosen field magnitude can be satisfied.
Regarding alternative ramp-up schemes, an option is under study $[4]$ in which $I_{p}$ is increased up to $\sim 2 MA$ in circular plasma configuration at the same rate as in the standard scenario, but is then maintained constant for $\sim 10 s$, during which elongation is increased in preparation for the X-point transition. Reducing the level of EC power in this phase also helps to reduce FWP heat loads, at the expense of reduced burn duration.
References
$[1]$ M. Kocan et al., Nucl. Fusion 55 (2015) 033019
$[2]$ V.E. Lukash, et al., 38th EPS (2014), paper P5.010
$[3]$ R. A. Pitts et al., J. Nucl. Mater. 415 (2011) S957
$[4]$ Y. Gribov, et al., submitted to 47th EPS (2020)
Figure 1: Left: full bore limited equilibrium at $t \sim 10 s$, just before the X-point transition in standard ITER current ramp-up scenarios. Centre: corresponding 3D field line traces of surface heat flux on the inboard midplane FWP#4 where heat fluxes ($q_{\perp}$) are highest for cases with and without n=1 LW misalignment. Right: variation of ($q_{\perp,peak}$) with $R_{q}$ (ratio of power conducted in the narrow and main SOL heat flux channels) for $\Delta_{LW} = 0-7 mm$, $I_{p} = 5 MA$, $B_{\phi}=5.3 T$, $P_{SOL} = 5 MW$, $\lambda_{q,nearIW} = 4 mm$, $\lambda_{q,mainIW} = 50 mm$ (IW = inner wall).
Figure 2: Composite of CAD images for the conceptual TF Mapping diagnostic. Upper: NMR sensor boxes attached to the vacuum vessel wall inter-modular keys used to centre the Blanket Modules (BM). Lower: the cabling from port plug feedthrough to sensor, showing the toroidal distribution of the 27 probes, with 18 on the inner midplane and 9 vertically displaced upwards by 1 BM. These are largely foreseen as risk mitigation in case of failure of any probes on the midplane row.
Speaker: Dr R. A. Pitts (ITER Organization, Route de Vinon, CS 90 046, 13067 Saint Paul Lez Durance, France) -
Here we report a neutronics study in a blanket mock-up using a compact fusion neutron source. With the optimized discharge condition, the neutron source stably produces fusion neutrons with the production rates > 10^5 n/s in a deuterium?deuterium (DD) operation (neutron energy: 2.45 MeV). The operation showed a non-linear increase with the applied voltage, in consistent with the cross section of D(d,n)3He reaction. The tritium production rate (TPR) was measured by a single-crystal chemical vapor deposition (CVD) diamond detector with 6LiF film. 6Li(n,t)4He reaction rates in the experiments were evaluated by integrating triton peak appeared in the energy range of 2.1-2.5 MeV and then compared with the results of Monte Carlo simulations to obtain the calculated to experimental (C/E) values. As a deuterium?tritium (DT) operation largely increases the neutron yield and enables an experimental validation of tritium breeding ratio (TBR), the present approach using the compact neutron source is beneficial for design of fusion blankets.
TBR defined as the rate of bred tritium to burnt tritium in DT plasma must be greater than unity for a self-sufficient fueling of a fusion reactor. An experimental validation of TPR with spectral neutron flux is vital to guarantee the self-sufficiency as computational errors are introduced by uncertainties in the nuclear data, a discrepancy in the simulated and the actual geometries, and an unexpected neutron absorption by impurities. However, opportunities for conducting neutronics experiments have been limited by low availabilities of neutron irradiation facilities across the world. Herein, we propose a neutronics experiment using a discharge-type compact fusion neutron source, which facilitates evaluations of breeding performance of a blanket mock-up. We have developed a compact fusion neutron generator, housing a negatively biased cylindrical cathode and a grounded anode [A].
In previous studies, however, instability of glow plasma during operation and an inaccuracy in neutron production rate were observed [A] [B]. Thus, an optimum condition for a stable operation was investigated by altering discharge conditions as shown in Fig. 1. The neutron production rate was linearly increased with discharge current (Fig. 1a). The neutron production rate increased exponentially with voltage (V < 30 kV) and saturated above 60 kV (Fig. 1b), in consistent with the cross-section of D(d,n)3He fusion reaction. Consequently, the discharge condition was optimized to be 9 mA and 60 kV, which yielded the neutron production rate of 1.4 ×10^5 n/s. A rectangular blanket mock-up (20 × 40 × 55 cm) composed of graphite reflector, polyethylene moderator, and lithium carbonate breeder blocks were irradiated with DD fusion neutrons. TPR was measured by using the CVD diamond detector with 6LiF film (6Li enrichment: 95.6%). The detector was calibrated with an unsealed 241Am ?-source; the reaction rate per count was obtained to be 12.0 counts per reaction. The validity of the value was confirmed by neutron and triton transport calculations using PHITS code. Fig. 2a represents geometry of the experimental set up and the thermal neutron flux distribution calculated by MCNP 5.0 code with the FENDL 2.1 nuclear data. The results showed that DD fast neutrons were readily moderated in the blanket mockup. Fig. 2b shows the measured spectrum of energetic charged particle generated by 6Li(n,t)4He reaction using the diamond detector. The triton peak appeared in the energy range of 2.1?2.5 MeV, while broad helium counts were detected below the energy due to energy straggling in the film. The integration of 3H count in Fig. 2b gave the reaction rate per source neutron (s.n.) of 4.05×10^-7 (/s.n.). The calculated value was 3.16×10^-7 (/s.n.) with statistical error of <4% where the most of 6Li(n,t)4He reactions occurred by moderated neutrons in the energy range of 0.01-1 eV. The detector was positioned at two positions and different irradiation time; the C/E values in the experiments were evaluated to be 0.71-0.78.
The use of a compact fusion source is proposed for evaluating tritium breeding performance of a blanket mock-up. As a consequence of the developments, the neutron source can stably generate fusion neutrons with production rates greater than 10^5 n/s in a DD operation. Even greater neutron production rate (about a factor of two) will be achievable in a DT operation, enabling a DT neutron transport experiment in a blanket mock-up to validate its TBR. We believe the approach using the compact source can contribute to future design studies of fusion blankets.
References
[A] K. Noborio et al. J. Plasma Fusion Res. 9 (2014) 1306142.
[B] K. Mukai and S. Konishi, Fusion Eng. Des. 146 (2019) 1633?1636.Speaker: Dr Keisuke Mukai (Kyoto University) -
In November 2019, the physical start-up of KTM tokamak was carried out. One of the main research task of KTM tokamak is study of behavior of first wall candidate materials under influence of high heat flux, which is expected to be in future thermonuclear reactors [Ref.1]. KTM is an aspect ratio A=2, single null divertor configuration, plasma current Ipl=750 кА, toroidal magnetic field Bt=1 Т, plasma pulse duration in an ohmic mode τpulse≤1 seconds and up to 5 seconds with 5 MW ion cyclotron resonance additional heating. In the KTM the maximum estimated heat flux value on the divertor plates is expected to be 20 МW/m2, which is corresponds to expected heat load value in ITER [Ref.2].
Currently one of the optimal methods for the surface temperature measurements of first wall materials is noncontact infrared thermometry. It is possible to measure temperature distribution on a surface of research material with high spatial and temporal resolution by using thermographic camera According to physical principle of thermographic camera it is required to set correct body emissivity value to achieve precision temperature measurements.
In international experimental thermonuclear reactor ITER is planned to use metallic first wall made of beryllium and tungsten. In addition, studies of possibility to use lithium as first wall material for TNR and other materials are extensively carried out in the world.
Metallic first wall leads to the problem of precise determination of surface temperature by optical thermometry. This is due to the fact that metals are not black bodies, have low emissivity which in most cases depends on temperature. In addition, emissivity depends on condition of material surface and able to vary with time due to the both surface modification under effect of plasma emission and deposition material dust particle on the surface or, for instance, beryllium particle deposition on divertor plates made of tungsten. In such case, measurement error can achieve dozens of percent, especially in high temperature area.
For the correct thermographic measurements the original method was suggested [Ref.3,4]. This method allows to measure emissivity changes of research material during the plasma influence. The method based on using of IR laser. The method is based on the use of an infrared laser (IR) at the wavelength of the working range of the thermographic camera. Pulse-periodic laser radiation is used directly to determine emissivity changes of the sample during the plasma discharge and the corresponding correction of temperature measurements of the thermographic camera. The use of this method is designed to increase the accuracy of temperature measurements of the surface of the investigated materials with a thermographic camera in a wide temperature range.
The main idea of the proposed method is to use pulsed laser emission projected onto the body surface in the field of view of thermographic camera. Laser emission partially reflects from the surface of the body. It is possible to measure body reflectivity ρ or reflectivity changes by measuring power of reflected emission Wref. Since there is a direct dependence for opaque bodies between the emissivity (emissivity factor) ε and the reflectivity (the reflection coefficient) ρ:
ρ+ε=1, (1)
then when the value of one of the terms changes, for example, with increasing temperature, the second one also proportionally changes. Thus, it is possible to determine the value of the emissivity by controlling the value of body reflectivity
To control the reflection coefficient, it is proposed to apply periodic short pulses of the IR laser with a duration not exceeding one frame exposure and with a maximum repetition frequency ½ of the camera registration frequency. In this case, the laser pulses must be strictly synchronized with the exposure of the frame. Fig. 1 shows the proposed time diagram of the IR camera and laser. As can be seen from Fig. 1, application of such time diagram of the IR camera and IR laser operation, it is possible to measure the emitting power directly both from the body itself Wbody and from emission together with the reflected laser emission Wbody+ref= Wbody+ Wref. At the same time, the maximum effective operating frequency of the camera is reduced by half.
Depending on the speed of the heating the body surface, the frequency of laser pulses can vary from a peak equal to half of the maximum frame rate of the IR camera to the necessary minimum to control the heating rate.
Theoretically, the change of reflectivity Δρ can be defined from the change in the power of the reflected laser emission:
∆ρ(T_0-T_i )/ρ(T_0 ) =(W_ref (T_0 )-W_ref (T_i ))/(W_ref (T_0 ) ), (2)
where,
Wref (T0) – initial reflected laser emission at temperature T0
Wref (Ti) - reflected laser emission at temperature Ti
ρ(T0) – initial reflectivity at temperature T0
Thus, the change of body emissivity coefficient can be determined by measuring the change in the reflection coefficient. Since initially the value of ρ(T0) of the test sample is unknown, it is determined experimentally before plasma experiments.
The paper shows and discusses the results of testing the technique on a plasma beam facility [Ref.5]. Tungsten samples were heated by electron beams. Fig. 2 shows a layout of experiments on the plasma beam facility. During the experiments the samples were heated up to 1200 ºС.
Fig. 3 shows the curves of the relative changes of the reflected laser power radiation and reflectivity for the one of the tungsten samples. The emissivity of the sample determines by the measured temperature of the sample using a thermocouple, in accordance with the method in [Ref.6]. Then reflectance is calculated according to equation (1).
As can be seen from the data in Fig. 3, there is a fairly close coincidence of the measurement results for the proposed method and for the found values determined using thermocouple readings. Deviation is within the limits of the measurement accuracy.
Thus, testing results of the proposed thermographic method on the plasma beam facility has shown the possibility of using.References
1 I. L. TAZHIBAYEVA, et al., “KTM Experimental Complex Project Status”, Fusion Science and Technology, vol.47, April 2005, p.746 – 750.
2 N. Holtkamp, et al. An overview of the ITER project. Fusion Engineering and Design. Volume 82, Issues 5–14, 2007, Pages 427-434. https://doi.org/10.1016/j.fusengdes.2007.03.029.
3 Concept of a new approach in thermographic measurements for plasma-wall interaction studies on KTM tokamak / B. Chektybayev, E. Batyrbekov, M. Skakov, A. Sadykov // Program and Abstracts of 27-th IAEA Fusion Energy Conference, Ahmedabad, India, 22-27 October, 2018.
4 Active thermography method for metallic plasma-facing components temperature measurements in the thermonuclear fusion devices/ B. Chektybayev, E. Batyrbekov, M. Skakov, A. Sadykov // Abstract book of the 3-rd Quantitative InfraRed Thermography Conference Asia, Tokyo, Japan, 1-5 July, 2019.
5 V. Kurnaev, I. Vizgalov, K. Gutorov, T. Tulenbergenov, et al. , Investigation of plasma-surface interaction at plasma beam facilities, Journal of Nuclear Materials, Vol. 463, August 2015, Pages 228-232.
6 ASTM E 1933 – 97, “Standard Test Methods for Measuring and Compensating for Emissivity Using Infrared Imaging Radiometers”, 962-3 p, (1997).Speaker: Baurzhan Chektybayev (Institute of Atomic Energy of National Nuclear Center of Republic Kazakhstan) -
The breeding blanket will be the component in charge of extracting and amplifying the neutronic power in future Fusion Power Plants. In addition, it has to warrant the reactor tritium self-sufficiency through efficient breeding and recovery. An additional function is the protection of the vacuum vessel (VV) and magnets against radiation [1.]. Various blanket concepts have been proposed along the years based on different breeder materials, neutron multipliers and power extraction methods (coolants). Within the EUROfusion Power Plant Physics and Technology (PPPT) Program (2014-2018) four breeding blankets have been investigated [1.], among them, the Dual Coolant Lead-Lithium (DCLL) blanket. This concept uses the eutectic alloy of lead-lithium (PbLi) as tritium breeder, neutron multiplier and tritium carrier. In addition, the PbLi acts as main coolant, being a quasi-self-cooled concept. The structural material is EUROFER, while the secondary coolant, mainly used to refrigerate the first wall (FW) and the blanket structures, is helium (He) at a pressure of 8 MPa. The proposed program considered a “low temperature” version of DCLL, i.e. with a maximum operational temperature of 550 ºC, in order to allow the use of conventional materials and technologies. One of the major consequences in this limitation was that the potential thermodynamic efficiency, which is the main advantage of this blanket technology, is reduced [2.].
Activities coordinated by CIEMAT on the DCLL development started with the definition of system specifications, leading to an overall design of the blanket, which included a series of engineering analyses needed to prove the feasibility of this concept. Intensive CAD work was focused on the adaptation of the DCLL geometry to the space allocated for the blanket in different reactor models. The result was a pre-conceptual design of DCLL based on a multi-module segment configuration, consisting on a number of different blanket modules attached to a common back supporting structure. The last DCLL design, produced during 2018 for a DEMO reactor with a fusion power about 2 GW, consisted of 16 sectors distributed every 22.5° and including 5 segments each, 3 outboard and 2 inboard. Calculations in different engineering fields were needed to prove the good performances of the blanket. Firstly, neutronics computations were required to calculate responses such as the tritium breeding ratio (TBR), the nuclear heating distribution and the energy multiplication factor. This version of DEMO included a critical reduction in the radial thickness of the OB segments, with a negative effect on the TBR. However, the required TBR target (>1.1) was still achieved, maintaining also adequate shielding performances on the VV and Toroidal Field Coils [3.].
The thermalhydraulics (TH) performances of the DCLL have to assure a proper power extraction looking for the highest reactor efficiency. In this sense, different activities related to the coolants (PbLi and He) were also conducted to maximize their outlet temperature and, at the same time, maintaining a low value for the pressure drop. The PLATOON 1D code, developed at CIEMAT [4.], was employed to characterize key parameters which affect the performance of the primary heat transfer system and the power conversion system. CFD calculations (Ansys FLUENT) were also used to study specific issues in the He cooling system, e.g. the mass flow distribution. A trade-off between several He channel dimensions was carried out to select the most suitable to work under different operational conditions. Temperatures resulting from the TH analyses were used as input for a thermomechanical analysis which also included internal pressure loads.
Due to the interaction between the intense magnetic field of the reactor and the liquid metal (an excellent electrical conductor), magnetohydrodynamics (MHD) effects can appear. MHD phenomena may produce important pressure drops in the PbLi flows, impacting the blanket functionality; thus, MHD analyses are of primary interest [5.]. Apart from pressure drop estimations, specific MHD calculations were conducted to estimate the heat transfer between the He and PbLi circuits. Special components, the Flow Channel Inserts (FCI), are used to mitigate MHD pressure drops, but introduce important effects on the PbLi velocity profile that were also investigated.
The associated blanket R&D includes the development and characterization of ceramic components to mitigate MHD effects. CIEMAT has developed an ambitious workplan to fabricate ceramic components made of alumina, with different arrangements and geometries [6.]. An extensive characterization, including effects of ionizing radiation, has been performed. Results are promising although some issues are still unsolved and will be addressed in upcoming activities. In addition to FCI/ceramics development, some other activities have been developed in support of the design. The tritium extraction and removal system, which is a key component in charge of the tritium recovery, has been deeply investigated (technologies, materials, transport models). A consolidated design able of extracting tritium with a minimum 80% efficiency has been produced to fit with the DCLL design specifications [7.]. Finally, specific tritium transport models were developed to investigate tritium migration to the coolants and expected inventories in the blanket [8.].Recently, a working group supported by a panel of independent experts, made an extensive work to identify the best strategy to harmonize the ITER-TBM and the EU-DEMO BB programs, including their associated R&D. Thus, a ‘driver’ blanket will be installed in DEMO that should be in line with the ITER TBM Program, which now considers the HCPB and WCLL to cover all technologies (coolants, breeders) [9.]. It seems advisable to develop, in parallel, alternative concepts for the long term, aiming at higher efficiencies. In that sense, although the feasibility of the DCLL has been demonstrated (main issues are related to corrosion and unknown MHD phenomena), it has been suggested that activities will focus more on the R&D of the concept, excluding all the integration tasks. Among others, these activities could explore different configurations for the FW, operational windows of the coolants, strategies for TBR enhancement. At present, the efforts of the DCLL design team have been focused on exploring the adaptability of the low-temperature DCLL to the needs of more advanced breeding blankets, taking as basis the already developed DCLL concept. Thus, the activities are focused on optimizing the DCLL in terms of plant net efficiency, proposing solutions to solve the issues encountered during the period 2014-2018.
- L.V. Boccaccini et al., Fusion Eng. Des. 109–111 (Part B) (2016) 1199–1206
- D. Rapisarda et al., Trans. Plasma Sci. 44 (2016) 1603–1612
- I. Palermo et al., Fusion Engineering and Design 138 (2019) 217–225
- I. Fernandez-Berceruelo et al., Fusion Engineering and Design 124 (2017) 822-826
- F.R. Urgorri et al., Nucl. Fusion 58 (2018) 106001
- I. Fernandez-Berceruelo et al., Fusion Engineering and Design 136 (2018) 633–638
- B. Garcinuño et al., Nucl. Fusion 58 (2018) 095002
- F.R. Urgorri et al., Nucl. Fusion 57 (2017) 116045 (11pp)
- G. Federici et al., Fusion Engineering and Design 141 (2019) 30–42
Speaker: Dr David Rapisarda (CIEMAT) -
Runaway electrons (REs) are relativistic electrons produced during plasma breakdowns, ramp-ups and disruptions. They may reach several 10s of MeV and form multi-MA beams and pose a serious threat to the reliability and availability of future tokamak devices. During the initial commissioning phase of the WEST tokamak [Bucalossi 2014], a significant fraction of discharges contained runaway electrons created at breakdown. On discharge #52205, runaway electrons survived the current ramp-up and flat-top then hit the outboard limiter when the plasma disrupted. Shortly after the event, the quench of one of the toroidal field coil was observed.
The large number of runaway electrons in the initial commissioning phase of WEST was attributed to the narrow operational window in which breakdown and burn-through was achievable. Too low prefill pressure led to non-sustained breakdowns, and surprisingly, too high prefill led first to runaway electrons (see figure 1). Machine conditions can explain this behavior, as most of the runaway discharges had higher radiated fractions in the first 25 ms after breakdown. In those situations, the ohmic current failed to rise quickly enough, and the available flux change from the central solenoid was preferentially taken by runaway electrons seeds. RE beam scenarios ranged from very early dissipation after breakdown to slide-away discharges in which REs were sustained during the entire pulse (see classification on figure 1). The latter usually led to the most severe impacts on the plasma facing components, especially when happening on the toroidally localized outboard Antenna Protection Limiter as in the event described below.
Pulse #52205 was among the first X-point pulses attempts on WEST. Runaway electrons appeared during plasma current ramp-up then stayed during the whole current flat-top (350 kA). They were ultimately lost on the APL when the plasma disrupted 1.75 s after breakdown. The impact was visible during more than 1.5 s on the visible cameras, indicating strong in-depth heat deposition. A net voltage and a helium pressure rise appeared on Toroidal Field Coil #9 (TFC#9) 2 seconds after the runway loss event [Torre 2019]. Following the triggering of the safety discharge system, the current from TFC#9 was dumped in 20 s instead of the standard 120 s, indicating that the coil had quenched. The primary suspicion for the cause of the quench was the RE impact.
REs interacting with matter generate bremsstrahlung photons which produce photoneutrons by interaction with the nuclei of encountered atoms. Since REs are highly relativistic, their bremsstrahlung radiation is emitted in a forward-beamed cone. The emission cone of a runaway beam hitting tangentially the APL intersects TFC #9 and partially intersects TFC#10. The latter did not quench, but its temperature showed an anomalous increase after the RE event. This observation is a strong indication that REs were involved in the quench process. Furthermore, a similar quench event happened in 1989 [Duchateau 1991] after a disruption with REs. A retrospective analysis of the machine geometry at that time revealed that the relative location of the quenched coil and the outboard limiter was similar as in the event described in the present article.Simulations of the RE impact on the APL have been made using the GEANT4 code [Agostinelli 2003]. GEANT4 is a Monte-Carlo code simulating the passage of particles through matter. Physics models include electromagnetic, hadronic, and optical processes over an energy range from 250 eV to several TeV. A library of materials is also provided. The code is used in a wide range of applications, including particle physics detectors design, space engineering and medical physics. The runaway impact was simulated in a 55° section of the torus (see figure 2a and 2b). It comprises the carbon limiter upon which the runaways were lost, the stainless steel first wall and vacuum vessel. Three TFCs (#8, #9 and #10) were included in the geometry. The vertical and horizontal impact angles were set to 10° and 1° respectively.. A mono-energetic beam of 108 electrons was simulated as a compromise between statistical noise and computational time. Total energy was re-normalized to the real number of electrons.
The results confirm that most of the secondary particles (photons, electrons and neutrons) were emitted in a narrow forward beamed cone (see figure 2b). The energy deposited in various machine elements was computed for 15 MeV and 30 MeV electrons. Only their kinetic energy (300 kJ for 30 MeV) was considered. No conversion from magnetic to kinetic energy was assumed.
Simulation results show that half of the total energy was lost locally on the APL. The rest was spread over first wall, vacuum vessel and TFCs. The energy deposition map confirmed that TFC#9 was the most severely hit coil, with 6.8 kJ deposited in the casing, and 1.3 kJ in the winding pack. The deposition was peaked on the front side and corner of the winding pack (see figure 2c), with ~1500 cm3 receiving more than 100 kJ/m3. As a comparison, the winding pack of TFC#10 received only 0.6kJ. Due to the greater distance from the impact point, the heat deposition was also spread over a larger area on the side of the winding pack. This reinforces the hypothesis that REs are responsible for the quench of TFC#9 and the small temperature increase of TFC#10. Thermohydraulical simulations [Nicollet 2020] confirmed that the volume of high heat flux computed on TFC#9 was compatible with the temperature measurements on the coil casings as well as the dynamics of helium expulsion following the quench. An identical simulation with 15 MeV REs showed that the energy deposited in the TFC#9 winding pack was 6 times smaller, in worse agreement with thermohydraulical calculations. Finally, the nature of the radioisotopes detected on the impact point also confirmed that the maximum energy of the runaway beam was above 20 MeV. Measurements of material activation inside the torus showed that the highest count rate outside the direct impact point was observed on the path of the predicted gamma/neutron emission cone and in qualitative agreement with the map of heat deposition computed by the GEANT4.The much larger amount of shielding material (several tens of cm) makes such a quench scenario less likely on ITER compared to WEST where the toroidal field coils are closer to the plasma (only a few cm steel thickness). However, it sheds light on the importance of limiting the number of runaway events due their potential to deposit heat in remote and unexpected areas.
[Bucalossi 2014] J. Bucalossi et al., Fusion Engineering and Design, 89, 907912 (2014)
[Torre 2019] A. Torre et al., IEEE Transactions on Applied Superconductivity, 29, 5, 4702805 (2019)
[Duchateau 1991] J.-L. Duchateau et al., IEEE Trans. Mag., vol. 27, no. 2, pp. 2053-2056, (1991)
[Agostinelli 2003] S. Agostinelli et al., Nuclear Instruments and Methods in Physics Research Section A 506, 3 (2003)
[Nicollet 2020] S. Nicollet et al., Cryogenics, 106 103042 (2020)Speaker: Cedric Reux (CEA, IRFM, F-13108 Saint Paul-lez-Durance, France.) -
The Greenwald density limit ($\bar{n}_g$) defines a fundamental bound on the tokamak operating space, and so is of central importance to magnetic fusion. Recent experiments (1) reinforce the suggestion (2) that the density limit occurs due to an abrupt increase in edge particle transport, with edge cooling and MHD phenomena following as secondary consequences. Here, we present a theory of degraded particle confinement as $\bar{n} \to \bar{n}_g$ due to edge shear layer collapse. The crucial microphysics is a breakdown of the self-regulating turbulence–shear flow feedback loop, due to a drop in flow production. Electron adiabaticity characterizes parameter regimes. Favorable scaling of particle transport and density with current, like the Greenwald $\bar{n} \sim I_p$, emerges from the effect neoclassical screening on zonal flow production. Higher current strengthens zonal flow shear for fixed drive. Zonal flow screening physics has implications for the scaling of the density limit in different devices.
Theoretical work (3) has identified the transition from adiabatic ($\alpha = \frac{k_\parallel^2 V_{th}^2}{\omega \nu} >1$) to hydrodynamic ($\alpha < 1$) electrons with increasing $n$ as a cause of the drop in Reynolds stress-driven production of shear flows, consistent with fluctuation studies close to the density limit (1). These are shown in Fig. 1.
The physics mechanism is a transition from a regime of propagating drift waves, which generate a flow convergence and so a shear layer spin-up, to one of weakly propagating convective cells, with a consequently weak coherence of $\tilde v_r$ with $\tilde v_\theta$. In the latter, eddy tilting (symptomatic of flow generation) does not arise as a straightforward consequence of causality. Fig. 2 shows the scaling of transport fluxes with $\alpha$. 
Note that the particle flux increases for $\alpha <1$, while the vorticity gradient decreases there. This indicates that self-regulation fails in the hydrodynamic regime. Observe that the vorticity gradient ($\nabla u$) is a natural order parameter for the flow and superior to shear for prediction of suppression, since $\nabla u$ prevents local alignment of eddies with flow shear. Shear layer collapse is consistent with potential vorticity (PV—also total charge) conservation. However, for the $\alpha >1$ regime, PV fluxes of particles and vorticity (i.e. transport and zonal flow) are tightly coupled. For $\alpha <1$, the particles carry the PV flux. Thus, the different regimes manifest different branching ratios of the components of the PV flux, but the same total flux.Flow–fluctuation–transport feedback is shown in Fig. 3.
This suggests a unified picture of: the L-mode as a state of modest shear flow, the Density Limit as a state of weak flow, and the H-mode as a state of strong mean shear. The onset of the density limit by shear layer collapse emerges as a transport bifurcation.Favorable current scaling is a salient feature of the Greenwald Limit. Theoretical work has identified the neoclassical ‘screening’ of the sheared zonal flow (ZF) as the physical mechanism underpinning the favorable $I_p$ scaling. Since the effective ZF scale is set by $\rho_\theta$ (4), effective ZF inertia is lower for larger current. An approximate scaling
$ \tilde v_E^\prime \approx \frac{S_{k,q}}{\rho_i^2 + 1.6 \epsilon^{3/2}_T \rho_{\theta_i}^2} \sim \frac{ \sigma \left(\frac{e \phi}{T} \right)^2_{DW}}{\rho_{\theta_i}^2} \sim \sigma B_\theta ^2 \left(\frac{e \phi}{T} \right)^2_{DW} $follows. $\left(\frac{e \phi}{T} \right)^2_{DW}$ is the drift wave intensity and $\sigma \sim n^{-\alpha}$ represents production. Higher current strengthens ZF shear, for fixed drive. DW-driven nonlinear noise scales $\sim B_\theta^4$, ensuring persistent excitation of the edge shear layer with increasing $B_\theta$. We see that reduced screening at high $I_p$ can “prop-up” the shear layer vs. weaker production. Favorable current scaling persists in the (ion) banana and plateau regimes, but weakens deep into the Pfirsch–Schlüter regime, also consistent with shear layer collapse at high $n$.
These results have implications for devices other than tokamaks. RFP’s are known to exhibit ‘Greenwald-like’ scaling, as $\bar{n}_g \sim I_p$. This is not surprising, since in RFP $\rho_i$ is set by the poloidal field, i.e. $\rho_i = \rho_{\theta_i i}$, so classical zonal flow screening is weaker (ZF shear stronger) at high $I_p$. In stellarators, the principal correction to classical screening is due to helically trapped particles. This has no obvious length scale (5), so ZF screening is classical. This feature likely explains why attempts to link stellarator density limits to magnetic geometry have failed, and why stellarator density limits appear higher than those in tokamaks.
Ongoing work is concerned with analysis of perturbative experiments of shear layer collapse and with studies of zonal flow evolution in layers with variable $\alpha(r)$. Of particular interest are bias-driven shear studies, which attempt to enhance shear layer persistence beyond $\bar{n}_g$, and elucidate the local dynamics of the density limit.
Research is supported by the U.S. DOE, and CNNC and MOST, China.
(1) R. Hong, et al., Nucl. Fusion 58, 016041 (2018)
(2) M. Greenwald, PPCF 44, 2194 (2002)
(3) R. Hajjar, P.H. Diamond, M. Malkov, PoP 25, 062306 (2018)
(4) M.N. Rosenbluth, F.L. Hinton, Phys. Rev. Lett. 80, 724 (1998)
(5) H. Sugama, T.H. Watanabe, PoP 13, 012501 (2006)Speaker: P.H. Diamond (UCSD, USA) -
In drift-wave turbulence, cross-correlations between fluctuations produce turbulent fluxes, which have profound consequences for confinement and the underlying nonlinear turbulence dynamics. In this work, we use a new approach based on deep learning to obtain a reduced mean-field model for the drift-wave/zonal flow dynamics from simulations of the Hasegawa-Wakatani system. In particular, we uncover a non-diffusive particle flux driven by the vorticity gradient, which is a new mechanism for corrugation of the profile. In this system, this effect is stronger than that of the shear. The new method also explicitly recovers previous analytical models for zonal flow generation, based on negative viscosity.
Reduced models of turbulence and transport are essential to the predictive understanding of plasma profile formation. A useful reduced model should distill the essential physics understanding gleaned from large scale simulations and also enable wide-ranging explorations of parameter space. The key physics of anomalous transport, flow generation (and other structure formation processes), and virtually all aspects of the turbulent transport dynamics is ultimately encoded in the turbulent fluxes. Such fluxes are enabled by finite cross-correlations (cross-phases) between fluctuating quantities; computing these cross correlations is the central problem in plasma turbulence modeling. However, such calculations are challenging, since they require the use of several, frequently debatable, approximations.
In this work, we introduce a novel data-driven approach to determine the dependencies of the cross-correlations. The new method uses deep supervised learning to infer a mean-field model from numerical solution and/or experiment. This approach, a form of nonparametric regression, leverages deep learning’s resilience to the large amounts of noise inherent to turbulence, as well as its ability to model arbitrary nonlinear, multivariate functions.
As a test of concept, we numerically solve the 2-D Hasegawa-Wakatani (HW) system, and use computed mean field quantities to train a deep neural network (DNN) which outputs the local turbulent particle flux and Reynolds stress as a function of local mean gradients, flow properties, and turbulence intensity. Exact symmetries are exploited to select independent variables and constrain the model. Up to nonlinear saturation effects, the DNN infers a model for the turbulent particle flux of the form $\langle \tilde v_r \tilde n \rangle \simeq \varepsilon ( - D_n \partial_x n D_u \partial_x u)$ where $\varepsilon$ is the turbulence intensity, $n$ and $u$ are respectively the mean density and vorticity, and $D_n$ and $D_u$ are constants. The first term is the familiar quasilinear diffusion which tends to relax the driving gradient. The second, proportional to the gradient of vorticity (see Fig. 1), is non-diffusive and previously unreported. We recover this non-diffusive flux with a simple analytic calculation. Notably, the deep neural network finds that, in this system, the vorticity gradient effect on the particle flux is significant, and much stronger than the direct effect of the local vorticity (shear).
The non-diffusive flux originates from a shift in the drift-wave frequency induced by the nonlinear convection of vorticity. It has immediate implications for structure formation, as it tends to modulate the density profile in the presence of a quasiperiodic zonal flow, forming a staircase. This mechanism for staircase formation is independent from previous models based upon bistability. We show that the same shift in the drift-wave frequency also impacts the local linear growth rate, resulting in corrugations in the intensity profile localized where the density and vorticity gradients have the same sign.Using the new method, we also uncover a Cahn-Hilliard-type model for the generation of zonal flow via Reynolds stress, which agrees with and finds corrections to previous theoretical work 2. Explicitly, we find at leading order $\langle \tilde v_r \tilde v_\theta \rangle \simeq \varepsilon (-\chi_1 u+\chi_3 u^3-\chi_4 \partial_x^2 u)$ (see Fig. 2). The recovery of the small, but important hyperviscous term shows that the deep learning approach passes a sensitive test.
Finally, the learned models for the particle flux and Reynolds stress are combined with conservation of potential enstrophy, similar to [3], to form a 3-field, 1-D reduced model for the turbulent dynamics. This system is solved numerically. The solution of the simplified model correctly exhibits several features of the nonlinear dynamics of 2-D HW, including the spontaneous formation and stabilization of zonal flows, the modulation of the density profile, and the formation of corrugations in the turbulence intensity.While the 2-D HW system is a simple and computationally inexpensive example, our results suggest that the deep learning method can possibly be applied to other, more complex applications, such as more sophisticated reduced models, gyrokinetic simulations, or even experiment. This analysis thus opens a new avenue for model reduction, which circumvents some of the analytic intractability of such models. We speculate that training is likely to succeed if (a) the system exhibits a requisite degree of symmetry, (b) a finite set of important mean-field variables can be identified, and (c) data can be obtained spanning a broad range of those variables. Exact symmetries are especially useful, as model constraints. However, the underlying assumption of a mean-field model is valid only in the weak turbulence regime and breaks down when the dynamics become sufficiently intermittent. Moreover, the assumption of locality in space and time is a significant, ad hoc limitation.
This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation. It was supported by the U.S. Department of Energy.
References:
1 R. Numata, R. Ball, and R. L. Dewar. Phys. Plasmas 14, 102312 (2007).
2 P. H. Diamond, S.-I. Itoh, K. Itoh, and T.S. Hahm. Plasma Phys. Control. Fusion 47, R35 (2005).
[3] A. Ashourvan and P. H. Diamond. Phys. Plasmas 24, 012305 (2017).Speaker: Robin Heinonen (UCSD) -
Many tokamak experiments have demonstrated improved core confinement with high toroidal rotation shear. It is widely recognized that turbulence-driven zonal flow has a strong impact on confinement. It has been pointed out that zonal flow can be driven by the parallel compression$^{1,2}$, which is expected to be enhanced with equilibrium parallel rotation shear. Particularly, a gyrokinetic simulation study manifests an essential role of the parallel compression in the zonal flow generation and turbulence regulation in a realistic tokamak geometry$^3$. In this work, we report a gyrokinetic simulation study demonstrating zonal flow amplification in nonlinear saturation stage of rotating plasmas. The energy ratio of the zonal flow to the turbulence is found to increase with the parallel rotation shear.
We perform global $\delta\!f$ gyrokinetic simulations of electrostatic ion temperature gradient-driven turbulence with adiabatic electrons by using the gKPSP code$^4$. The radial profile of the ion temperature ($T_i$) gradient is given by $R_0/L_{Ti,eq}(r)=(R_0/L_{T0})\exp(-y^4)$, where $y=(r-r_c)/0.3a$. Here $R_0$, $a$, and $r_c$ are the major and minor radus and the most unstable radius, respectively. We apply the equilibrium sheared parallel rotation $U_{||,eq}(r)$ around $r_c$. We vary the values of $R_0/L_{T0}$ and the mean rotation shear at $r_c$, that is $U_{||0}'\equiv -(a/v_{T0})dU_{||,eq}(r_c)/dr$. Except for the equilibrium parallel rotation and the $T_i$ gradient, other plasma parameters are very similar to the CYCLONE base case. We measure the electrostatic field energy, which is defined as
$E_\phi = \frac{1}{8\pi} \int d^3x \left[ \left|\nabla\phi\right|^2 + \frac{\rho_{ti}^2}{\lambda_{Di}^2} \left|\nabla_\perp \phi \right|^2 + \frac{1}{\lambda_{De}^2} \left|\phi \right|^2 \right].$
Here $\phi$, $\rho_{ts}$, and $\lambda_{Ds}$ are electrostatic potential, the gyroradius, and the Debye length, respectively. We use the energy ratio of the zonal ($n=0$) to the non-zonal ($n\neq 0$) components as a figure of merit for the efficiency of zonal flow generation.
Figure1 shows the time histories of the field energy of the zonal and the fluctuation components with (a), and without (b), equilibrium parallel rotation shear $U_{||0}'$ for a fixed value of $R_0/L_{T0}=5.21$. The energy ratio of the zonal to the fluctuation components (the red curves) increases in the presence of $U_{||0}'$. We note that the linear growth rate of the case with $U_{||0}'$ is larger than that of the no rotation case, as shown by the earlier nonlinear saturation. In spite of the larger linear growth rate, turbulence is saturated at a lower level with the rotation shear due to the increased zonal flow.
The energy partition between the zonal flow and the turbulence for various values of $U_{||0}'$ and $R_0/L_{T0}$ is shown in Fig.2. The presented energy ratios are obtained by averaging over a time period during nonlinear saturation phase, which is denoted by the red broken lines in Fig.2, for example. The zonal energy portion increases with $U_{||0}'$ for the different $R_0/L_{T0}$ values. These results demonstrate zonal flow amplification by the mean rotation shear, being consistent with confinement improvement with high toroidal rotation shear in tokamak experiments.
In our previous work$^3$, we explained the generation of zonal flow as a consequence of the potential vorticity (PV) flux. And we evaluated contributions from the parallel compression and the perpendicular drift motions in the gyro-center density evolution, to the PV flux. According our findings, the two dominant contributions from the parallel compression and the grad-$B$ drift largely cancel out in the no rotational plasmas. This balance, but not exact, limits the PV flux and the zonal flow. On the other hand, with mean finite parallel rotation shear, the parallel compression-driven PV flux becomes dominant over the grad-$B$ drift-driven one. The asymmetry between the dominant contributions can produce larger PV flux, leading to the amplification of zonal flow. In the previous work, however, the physics of the zonal flow amplification and the turbulence saturation have not been fully elucidated. We will uncover these underlying mechanisms in detail in the future presentation.
References
$^1$ Wang L. et al 2012 Plasma Phys. Control. Fusion 54 095015
$^2$ Jhang H. and Kim S.S. 2019 Phys. Plasmas 26 112501
$^3$ Yi S. et al 2019 Nucl. Fusion 59 044002
$^4$ Kwon J.M. et al 2017 Comput. Phys. Commun. 215 81Speaker: Sumin Yi (National Fusion Research Institute)
-
-
-
ITB formation due to energetic particles. The performance of present-day and future fusion devices is largely determined by turbulent transport generated by plasma turbulence. Any mechanisms able to reduce the overall radial propagation of energy and particles is, therefore, crucial in view of scenario optimization. This contribution presents numerical results of turbulence suppression by supra-thermal ions and experimental results from ASDEX Upgrade, which support these numerical findings. More precisely, the simulations demonstrate, for the first time, the generation of an internal transport barrier (ITB) triggered purely by energetic particles in a monotonic safety factor configuration.
Physical mechanisms. These results are explained in terms of a resonant interaction between ion-driven turbulence and supra-thermal particles, recently identified via gyrokinetic flux-tube simulations [1]. Fast ions have been found to interact with the plasma micro-instabilities through a wave-particle resonance mechanism when the fast ion magnetic-drift frequency is close to the linear frequency of the ion temperature gradient (ITG) microinstability, thus amplifying an otherwise negligible interaction. A theoretical analysis and numerical simulations have shown that the flow of such a resonant energy exchange is determined by the fast particle temperature, density and their gradients and, in turn, sets the direction of the fast ion energy losses. Inward (outward) supra-thermal ion particle and heat fluxes are observed in correspondence with the strongest fast ion stabilization (destabilization). Therefore, in a radially global setup, stabilizing (inward fluxes) and destabilizing (outward fluxes) energetic particle effects on plasma turbulence occur at different radial positions - depending on the local values of the fast particle parameters - thus strongly affecting the bulk ion energy fluxes.
Gyrokinetic GENE simulations. In this contribution, we significantly extend the previous findings - based on flux-tube analyses - by means of global electromagnetic GENE [2] simula-tions with kinetic electrons (realistic proton-electron mass ratio) and a sophisticated bi-Maxwellian model for the energetic particle distribution. The main results are displayed in Fig.1, where the time evolution of the radial profile of the (surface averaged) ion heat flux is shown for the simulations without (left plot) and with (right plot) energetic particles.Fig.1 shows – to our knowledge - the first numerical evidence of an ITB generated solely by the energetic particles. In particular, a full suppression of ITG-driven turbulence is observed in the radial domain \rho_{tor}=[0.2,0.3]. As the turbulent heat flux drops in the ITB region, a corresponding increase of the neoclassical transport is observed. However, the resulting overall flux remains at significantly lower levels compared to the case without energetic particles. The same magnetic equilibrium and kinetic profile of the main plasma are employed in both the simulations regardless the presence of the supra-thermal ions, thus excluding e.g. any effects of the geometry on the generation of the ITB. These results are fully consistent with the physical picture of the wave-particle resonant interaction summarized above. More precisely, due to the local changes in the fast ion temperature and density profiles, the effect of supra-thermal particles on plasma turbulence turns from stabilizing in \rho_{tor}=[0.2,0.25] to destabilizing in \rho_{tor}=[0.25,0.3]. Thus, reverting the sign of the energetic particle energy flux from inward to outward. This sharp change in the direction of the fast ion heat flux strongly affects the shearing rate levels. Localized shearing layers are generated in correspondence of the negative inward and positive outward fast particle heat flux, thus tearing apart the radially elongated ITG eddies and resulting in the first ITB solely triggered by energetic particles.
Experimental evidence. These predict-first numerical results led to the design of an ASDEX Upgrade discharge [3] where this resonant interaction between ITG-driven turbulence and supra-thermal ions was maximized via gyrokinetic GENE simulations. The energetic particle profiles have been calculated with the code TORIC/SSFPQL [4]. A substantial increase of the peaking of the main ion temperature (Ti) profile of the order of ~80% is observed (as shown in Fig. 2) in the radial domain where the resonant interaction is predicted to be maximized. Such a relevant Ti is obtained via on-axis (\rho_{tor}=0) ion-cyclotron-resonant-frequency (ICRF) heating of the H minority species in D plasmas, with a large H concentration of n_H/n_e ≈ 0.11. Furthermore, no degradation of the energy confinement is observed during a ramp-up of the ICRF power, thus suggesting a substantial reduction of the anomalous turbulent transport. More specifically, a power balance analysis reveals a central region of improved confinement as the ICRH power is increased. The resulting radial heat flux profiles obtained with global GENE simulations exhibit the same features as in Fig. 1. The overall fluxes (including the neoclassical contribution) are in agreement with the experimental power balance only in the presence of energetic particles.Conclusions. This contribution provides, for the first time to our knowledge, numerical results that energetic particles can effectively trigger internal transport barriers in realistic tokamak configurations. These findings are supported by experimental evidence at ASDEX Upgrade and represent an essential step forward to access unique and still unexplored high confinement regimes.
References
-
A. Di Siena et al., NF 58, 054002 (2018); PoP 26, 052504 (2019).
-
F. Jenko et al., PoP 7, 1904 (2000); T. Görler et al., JCP 230, 7053 (2011); A. Di Siena et al., PoP 25, 042304 (2018).
- R. Bilato et al., to be submitted (2020).
- M. Brambilla and R. Bilato, NF 49, 085004 (2009); R. Bilato et al., NF 51, 103034 (2011).
Speaker: Mr Alessandro Di Siena (University of Texas at Austin) -
-
In the study of burning plasmas it is important to understand multi-scale interactions between energetic-particle-driven MHD mode and drift-wave turbulence for establishing good confinement of both energetic particles and bulk plasmas simultaneously. We investigate nonlinear multi-scale interactions between TAE, which is unstable at low $n$, and drift-wave turbulence, which is driven by micro-instabilities at high $n$, by means of a global gyrokinetic simulation code (GKNET [A]). We have revealed that TAE suppresses the most unstable drift-wave mode by violating the ballooning structure of the drift-wave mode, and the TAE transfers the energy from the most unstable drift-wave mode to lower $n$ modes to modulate turbulence, because the TAE has a finite $n$ in contrast to zonal flows ($n=0$). This modulation of drift-wave turbulence by TAE leads to an enhancement of both energy flux of bulk ions and particle flux of energetic ions. Hence, TAE and drift-wave turbulence synergistically enhance the transport of both bulk plasma and energetic particles.
Introduction: In order to realize good confinement of burning plasmas it is necessary to reduce both energetic particle transport and bulk plasma transport simultaneously. In burning plasmas drift-wave turbulence (DWT) and toroidal Alfven eigenmode (TAE) driven by energetic particles coexist, and they interact each other by nonlinear mode coupling, and thus the interaction may result in new transport phenomena, for instance, the transport of energetic particles may be influenced by turbulence and zonal flows which are active even in finite $\beta$ plasmas [B]. Since TAE is an MHD mode and drift-wave turbulence is electromagnetic at finite $\beta$, magnetic perturbations play an important role in the interaction and can increase turbulent transport by intensifying electrostatic potential perturbations [C]. In addition, interactions between TAE and drift-wave turbulence may excite stable low $n$ modes and increase turbulent transport as shown by the study of multi-scale interactions between magnetic islands and drift-wave turbulence [D].
Simulation model and linear stability: We investigate nonlinear interactions between TAE and drift-wave turbulence by means of a global $\delta$f gyrokinetic simulation code (GKNET) [A]. We consider a normal magnetic shear tokamak plasma which has energetic particle pressure gradient and bulk plasma pressure gradient with $\beta=1.28\%$, $T_f/T_i=25$, $m_i/m_e=100$, and $\rho_*=1/100$. Figure 1 (a) shows that this plasma is unstable against a TAE at low toroidal mode number $n=2$, which has real frequency in the gap of Alfven continuum indicated by yellow color. On the other hand, a drift-wave instability (kinetic ballooning mode: KBM) is unstable at high toroidal mode number $n \geq 6$. The TAE has global structure (Fig. 1 (b)), while the KBM has ballooning structure characterized by micro-scale (Fig. 1 (c)).
Turbulence modulated by TAE: We have performed a nonlinear simulation of the plasma which is unstable against both the TAE and drift-wave instability (KBM) (referred as "TAE+DWT"). In addition, we carried out nonlinear simulations of a plasma with flat energetic particle pressure profile to obtain drift-wave turbulence ("only DWT") and a plasma with flat bulk pressure profile to obtain TAE ("only TAE"), and then we compare them with "TAE+DWT" to understand the influence of interactions between TAE and drift-wave turbulence.
In "TAE+DWT", drift-wave turbulence (DWT) with zonal flows is established at first ($t=12$ and 16 in Fig. 2) because the growth rate of the drift-wave mode (KBM) is much higher than the TAE as shown in Fig. 1 (a). Then, in this turbulent state, the TAE ($n=2$) grows slowly and violates the ballooning structure of the turbulence to reach a quasi-steady turbulent state ($t=22$ and 36 in Fig. 2). We here compare time evolution of some main toroidal modes of perturbations in "TAE+DWT" and "only DWT" in Fig. 3 (a). The most unstable drift-wave mode ($n=12$) gets saturated by producing zonal flows ($n=0$) at $t=13$ for both "TAE+DWT" and "only DWT". Then, at $t=20$, the TAE ($n=2$) grows in "TAE+DWT", while $n=2$ mode decreases in "only DWT", resulting in much higher amplitude of $n=2$ mode in "TAE+DWT" as indicated by the red arrow. Following the growth of TAE ($n=2$) the most unstable drift-wave mode ($n=12$) further decreases in "TAE+DWT" compared to "only DWT" after $t=20$ as indicated by the blue arrow. Since the TAE has finite toroidal wavenumber in contrast to zonal flows ($n=0$), this nonlinear mode coupling between the TAE ($n=2$) and the drift-wave mode ($n=12$) enhances another lower toroidal wavenumber mode ($n=12-2=10$) as indicated by the green arrow. Hence, the TAE suppresses the most unstable drift-wave mode but enhances a lower toroidal wavenumber mode to modulate the drift-wave turbulence. Due to this modulation of turbulence by TAE, the energy flux of bulk ions $Q_i$ in "TAE+DWT" is enhanced at middle wavenumbers ($4 \leq n \leq 10$), and the peak of $Q_i$ in "TAE+DWT" is shifted from $n=12$ to $n=10$ compared to "only DWT" (Fig. 3. (b)). In addition, the particle flux of energetic ions $\Gamma_f$ is enhanced in "TAE+DWT" compared to "only TAE" (Fig. 3. (c)). Thus, the interaction between TAE and drift-wave turbulence enhances the transport of both bulk plasma and energetic particles.
[A] K. Imadera, Y. Kishimoto, K. Obrejan, T. Kobiki and J. Q. Li, IAEA-FEC, TH/P5-8 (2014).
[B] A. Ishizawa, K. Imadera, Y. Nakamura, and Y. Kishimoto, Phys. Plasmas, 082301 (2019).
[C] A. Ishizawa, D. Urano, Y. Nakamura, S. Maeyama, and T.-H. Watanabe, Phys. Rev. Lett., 025003 (2019).
[D] A. Ishizawa, Y. Kishimoto, and Y. Nakamura, Plasma Phys. Control. Fusion, 054006 (2019).Speaker: Akihiro Ishizawa (Kyoto University) -
Tokamak discharges with internal transport barrier (ITB) offer compelling features needed for steady state operation of burning plasmas for its improved energy confinement and high bootstrap current associated with large pressure gradient. To develop a stable ITB plasma usable for fusion reactors, it requires a deep physics understanding of ITB formation and its profile control. This paper reports on a novel simulation study, showing that magnetic islands can strongly change ${\bf E}\times{\bf B}$ shear flow structure and consequently influence turbulence, plasma transport, and confinement [1]. The results elucidate a long standing theoretical hypothesis based on experimental evidences that magnetic islands may trigger ITB formation inside a low-order rational magnetic surface. Practically, a low-n resonant magnetic perturbation (RMP) may be used as an ITB controller through its penetration to drive a controllable island at a desirable core location with a rational surface. The magnetic island perturbation is also found to largely change plasma self-driven current in tokamaks, impacting the steady state operation.
This simulation study employs a global gyrokinetic model that couples self consistent neoclassical and turbulent dynamics (neoclassical effects are often ignored in typical gyrokinetic turbulence simulations). This allows us to calculate a 3D neoclassical equilibrium electric field in 3D perturbed tokamak geometry with magnetic islands, which is shown to have a great impact on turbulence and the associated transport in the island geometry. This simulation feature is also critical for calculating plasma self-driven current in the presence of turbulence [2]. Simulation results reported here are obtained with experimental plasma parameters of a C-Mod L-mode Ohmic discharge which has ITG dominated turbulence in the core plasma. This discharge has a monotonic q profile. A prescribed static m/n=2/1 magnetic island is imposed on the $q=2$ surface located at the normalized minor radius $r/a=0.61$. To focus on the effects of magnetic islands, toroidal rotation is set to zero in the simulations.
It was shown that micro-turbulence can drive a strong electron current profile corrugation near a low-order rational magnetic surface with weak magnetic shear by residual stress, which may seed/drive a magnetic island at the surface [2]. Magnetic islands, by altering the topological structure of the confining magnetic field, have varied and complex impacts on plasma confinement in fusion experiments. Both the purely neoclassical simulation and the coupled neoclassical-turbulence simulation show a similar dynamic picture of radial electric field development from initial geodesic acoustic oscillation and radial propagation to the formation of a stationary $E_r$ well. Remarkably, the stationary $E_r$ well formed near the magnetic island is peaked at the inner boundary of the island (with a much lower peak at the outer boundary) (Fig.1). Compared to the case of axisymmetric tokamak (with no magnetic perturbations), the magnetic island is found to cause a change of plasma equilibrium from a weak negative radial electric field to a strong positive one. The electric field bifurcation is mainly due to the island-induced topological change in the magnetic field, which alters the ambipolarity condition of particle transport that determines the equilibrium electric field. Further, the $E_r$ well gets deeper as the island becomes wider (Fig.1). All these results are in qualitative agreement with experimental observations [3]. In addition to the strong zonal $E_r$ well, a stationary low-n potential structure dominated by a 2/1 mode is also formed at the same radial locations of inner and outer boundaries of the magnetic island (Fig.2). The amplitude of the $2/1$ potential is much higher than that of ITG turbulent fluctuations, and increases as the island width increases. The $2/1$ mode and its low-n harmonics (4/2, 6/3, 8/4 etc) form coherent vortex flows which are also observed in fluid simulation [4] and tokamak experiments [5] A remarkable feature revealed from our simulation is that the 2/1 modes are non-resonant, peaking at the boundaries of the island. The combination of zonal (0/0) and (2/1) mode (both defined with respect to unperturbed equilibrium geometry) produces a total potential well aligned with the perturbed flux surfaces, and correspondingly, a poloidal ${\bf E}\times{\bf B}$ shear flow on the perturbed surfaces. The island-induced radially localized ${\bf E}\times{\bf B}$ shear layer along with the sheared vortex flows due to the low-n potential can impact turbulence and associated transport in two-fold. First, the strong ${\bf E}\times{\bf B}$ shear is shown to effectively reduce and even suppress turbulent fluctuations in the inner core region next to the island through the well-known sheared advection of fluctuations (Fig.3). Second, the localized ${\bf E}\times{\bf B}$ shear layer can reduce and even block turbulence spreading from one side to the other [6], and then effectively decouple the plasma inside the shear layer from the outside. Depending on the width of the magnetic island, the two effects act together, naturally facilitating the formation of an ITB at the inner side of a rational magnetic surface where the island locates. The simulation further suggests that there may exist a critical island width which produces ${\bf E}\times{\bf B}$ shear that is strong enough to trigger ITB. The similar phenomenology of transport barrier triggered by the island-induced $E_r$ well may also play a role in the L-H transition.
Plasma self-driven electron current is also largely changed, not only locally in the island region but also globally. The current is changed in a way consistent with the island-induced electric field structure and pressure profile flattening. In particular, the parallel acceleration of electrons by the non-resonant 2/1 mode, which has intrinsically finite $k_{\parallel}$, is found to play a major role. As a result, axisymmetric current is largely reduced and a helical current is generated in the island region (Fig.4).
-
W. X. Wang, ``${\bf E}\times{\bf B}$ shear flow structure and plasma self-driven current generation in magnetic island", AAPPS-DPP 2019 (plenary talk), Hefei, Nov. 4-8, 2019.
-
W. X. Wang et. al., Nucl. Fusion ${\bf 59}$, 084002 (2019).
-
K. Ida et. al., Phys. Plasmas ${\bf 11}$, 2551 (2004).
-
A. Ishizawa et. al., Plasma Phys. Control. Fusion ${\bf 61}$, 054006 (2019).
-
M. J. Choi et. al., Nucl. Fusion ${\bf 57}$, 126058 (2017).
-
W. X. Wang et. al., Phys. Plasmas ${\bf 14}$, 072306 (2007).
Speaker: Dr Weixing Wang (Princeton Plasma Physics Laboratory) -
-
For over three decades, the observation of rapid core confinement improvement upon favourable modifications of edge operating conditions has been a nagging source of puzzlement for experimentalists investigating conditions for a lasting source of fusion energy in tokamaks. The transport properties of drift-wave turbulence and the interaction of the confined plasma with its material boundaries have long been recognised as essential to the resolution of this conundrum. Key aspects of the turbulent dynamics in the plasma edge are poorly quantified, owing to the disparity of temporal and spatial scales and the inadequacy of performing scale separations. Here we show, relaxing oft-made scale separation assumptions that a narrow region at the interface between open and closed magnetic field lines is central to explaining the transport properties of turbulence, globally. The proposed presentation, based on actual experimental parameters, discusses three main results. We unambiguously show (1) that turbulence is not only locally driven by local gradients but nonlocally controlled by fluxes of turbulence activity, primarily though not exclusively borne at the edge. This 'nonlocal' influence is mediated through vortex-flow localised interactions near the material boundaries and has two major consequences: (2) the nonlinear destabilisation of the linearly stable edge, providing a possible resolution for the so-called `shortfall’ conundrum and (3) the spontaneous emergence of a stable and localised transport barrier at the closed/open field line transition, possible prelude to the formation of a pedestal.
Low-frequency microturbulence in fusion plasmas is appropriately described by gyrokinetics. GYSELA[1] models ions and trapped electrons gyro-kinetically in the core and edge regions as well as the closed/open field line transition and the Scrape-Off Layer (SOL) through introduction of a simplified penalised limiter[2] mimicking the role of a heat and momentum sink, Fig.1-a. Importantly core, edge and SOL are treated on an equal footing. Specifically, GYSELA is built such that: (i) no scale separation is assumed as this simplification tends to break down towards the edge; (ii) the system is flux-driven implying that both fluxes and gradients are dynamic. Both properties are instrumental to the results here and have stringent implications for modelling as all spatial scales from the short ion-scale Larmor radius ($\rho_i$~1mm) to the global machine scale (L~1m) and intermediate gradient mesoscales typically~$\sqrt{\rho_i L}$, where turbulence self-organises, are self-consistently treated. Similarly, the fast turbulent motion (~$10^{-5}$s) and long collisional transport and energy confinement (~1s) scales are treated on an equal footing.
(1) Limiter-borne localised sources of nonlocal spreading. Spatial asymmetries spontaneously develop from the presence of the limiter. Starting from a poloidally-symmetric state, the combined action of magnetic drift and parallel losses leads to a poloidally-asymmetric kinetic equilibrium. Parallel gradients of density and temperature develop which lead to local accumulation (resp. depletion) of density (resp. temperature) in the vicinity of the limiter, though the pressure remains quasi-constant along the field lines. This asymmetry spans the SOL and extends in the confined edge plasma, typically within 10% of the last closed flux surface (LCFS). It leads to a poloidally-localised enhanced linear drive for ion-scale drift-wave instabilities, as confirmed through linear gyrokinetic analysis of the GYSELA profiles using the gyrokinetic code GKW[3] in the local approximation limit. Kelvin-Helmholtz like instabilities are weakly driven or stable. This instability source of is generic to the presence of the limiter and absent without, despite all other parameters being equal, as illustrated in Fig.1.
(2) Nonlocal SOL—edge—core interplay. In L-mode, key to apprehend the transition to H-mode, the situation of the edge is clouded. The universal observation in experiments that the relative level of fluctuations n/n monotonically increases from core to edge, n being the plasma density is often strongly under-predicted when pushing gyrokinetic approaches towards the edge region, bearing witness to the difficult understanding of its dynamics. This underprediction is commonly referred to as the ``transport shortfall''. Recently, further works have tested aspects of this shortfall; the issue, however, remains largely open and we will discuss it. Our work suggests a paradigm shift in how we model turbulence, suggesting that it is clearly not only locally driven. We find that an incoming front provokes ambient fluctuations to produce a controlled outward heat flux, first fuelling the otherwise linearly-stable edge turbulence then percolating through the transport barrier into the open field-line region and depositing its energy onto the boundaries. In our approach, these dynamics solves the shortfall conundrum.
(3) A spontaneous edge barrier. Only with the limiter and with an interface to the SOL, does a spontaneous radially-localised shear layer nucleate at the transition from closed to open field lines. Orbit loss through the LCFS results in a net radial current which equilibration is modified by the position of the limiter. Nonlinearly, the Er well is also sustained via vortex-flow positive feedback. This is shown in Fig.3, displaying the ExB flux of vorticity, proxy for the time increment in poloidal momentum deposition in the vicinity of the LCFS. Vortices of a definite vorticity are radially advected in a systematic direction. The resulting force acts such that each individual vortex in the vicinity of the shear layers either deposits its momentum into the flow or extracts the negative of its momentum from the flow, in both cases reinforcing it. A detailed discussion of the statistics of orbit loss and vortex-flow interaction is proposed. These nonlinear processes are important dynamical players for the organisation of turbulence in the edge region. Similarly to other non-equilibrium systems in nature, the tokamak plasma self-organises into a globally critical state, allowing a narrow 'tail' region at the edge—SOL interface to 'wag' the bulk core and edge plasma in a turbulence propagation time far faster than the plasma transport time[2].
[1] V. Grandgirard, et al., Computer Physics Communications 207, 35-68 (2016).
[2] E. Caschera, PhD dissertation, Aix-Marseille Univ. (2019) ; G. Dif-Pradalier et al. under review (2020).
[3] A.G. Peeters, et al., Comput. Phys. Comm. 180, 2650 (2009).Speaker: Guilhem Dif-Pradalier (CEA, IRFM, F-13108 Saint-Paul-lez-Durance cedex, France) -
This is the first report of spontaneous Internal Transport Barrier (ITB) formation in reversed magnetic shear plasmas by full-$f$ gyrokinetic simulations with kinetic electrons. We found that (1) a strong mean $E_r$ shear is formed near $q_{min}$ region in flux-driven Ion Temperature Gradient (ITG) turbulence, leading to spontaneous reduction of ion turbulent thermal diffusivity, while it is not observed in adiabatic electron case; (2) a robust co-intrinsic rotation is driven near $q_{min}$ surface and sustain the mean $E_r$ through the radial force balance, which physical mechanism is identified by momentum transport theory; (3) in the presence of electron heating, a counter-intrinsic rotation by Trapped Electron Mode (TEM) turbulence is selectively driven in negative magnetic shear region, leading to steeper $E_r$ shear and resultant larger reduction of ion turbulent thermal diffusivity. These results indicate that the co-existence of different modes can trigger the discontinuity near $q_{min}$, leading to spontaneous ITB formation.
Internal transport barrier, which acts as the shielding layer of particle and heat transport by suppressing turbulence, has a crucial key to achieve high-performance plasma confinement. In recent HL-2A discharge with co-NBI, ITB is formed in the almost flat $q$ profile region, in which mean flow triggered by toroidal rotation plays a dominant role in sustaining ITB. On the other hand, according to the NBI modulation test in JT-60U reversed magnetic shear discharge [i], they found that after the change from balance-NBI to co-NBI, ITB is suddenly collapsed because mean flow shear in outer half ITB region becomes weak in addition to the change of toroidal rotation shear. These indicate that not local but global pattern of mean $E_r$ is crucial for ITB formation. For the comprehensive study of such global effects, we have investigated the effect of momentum injection for ITB formation [ii] by means of our 5D full-$f$ gyrokinetic code GKNET [iii-iv], which can trace self-consistent profile evolutions coupled with mean $E_r$ through the radial force balance, even in weak/reversed magnetic shear plasmas not like $\delta f$ code. We found that toroidal rotation shear in outer region can change mean $E_r$ shear through the radial force balance, leading to ITB formation in which the ion thermal diffusivity decreases to the neoclassical transport level. However, enough large momentum injection, which is proportional to plasma size, was required for ITB formation in our previous work based on adiabatic electron model.
In this work, we have introduced a kinetic electron process based on hybrid electron model, in which trapped electrons are fully kinetic while passing electrons are adiabatic except $(m,n)=(0,0)$ one. Based on this model, we performed flux-driven ITG/TEM simulation for reversed magnetic shear configuration with $q(r=0.6a_0 )=q_{min}=1$. Figure 1 shows the time-spatial evolution of $E_r$ in (a) flux-driven ITG turbulence with adiabatic electrons, (b) ITG turbulence with kinetic electrons and (c) ITG/TEM turbulence with kinetic electrons. Here, we set only ion heating near magnetic axis in ITG case but both ion and electron heating near magnetic axis in ITG/TEM case. Figure 1 shows that the stable local maximum of $E_r$ is formed near $q_{min}$ surface only in the kinetic electron cases. This is because the kinetic electron dynamics can provide a finite phase difference between density perturbation and electrostatic potential, leading to more unstable linear ITG/TEM instability and resultant zonal flow generation.
Interestingly, we also found a robust net toroidal rotation around $q_{min}$ surface as is shown in Fig. 2 (a). According to the momentum transport theory, the residual stress part of momentum flux is given by $\langle\Pi_{RS}\rangle_{\theta\varphi}=\alpha I E_r'+\beta I'+\gamma \langle k_{\theta}k_{\varphi}\phi^2\rangle$, which first and second terms work to reduce diagonal momentum diffusion, which is confirmed by the correlation analysis. On the other hand, the steep electron temperature gradient selectively destabilize TEM in the negative magnetic shear region, where the ITG mode is relatively stable. This makes an opposite intrinsic rotation (see the red line among $0.2a_0[i] Y. Sakamoto et al., Nucl. Fusion, 41, 865 (2001).
[ii] K. Imadera et al., 26th IAEA-FEC, TH/P3-3 (2016).
[iii] K. Imadera et al., 25th IAEA-FEC, TH/P5-8 (2014).
[iv] K. Obrejan et al., Comput. Phys. Comm. 216, 8 (2017).
Speaker: Prof. Kenji Imadera (Kyoto University)
-
-
-
The role of the nonadiabatic electron drive in regulating the isotope mass scaling of gyrokinetic turbulence is assessed in the transition from ion-dominated core transport regimes to electron-dominated edge transport regimes. The scaling of the plasma energy confinement time with hydrogenic isotope mass is of critical importance, as most tokamaks operate with deuterium (D) as the main ion species, while ITER calls for dominant hydrogen (H) operation in the first phase, transitioning to 50:50 deuterium-tritium (DT) fuel composition at reactor-level operation. Experimental observations often show confinement improving with increasing ion mass {1}. Simple gyroBohm-scaling theoretical arguments (that ignore electron dynamics), however, predict that the turbulent ion energy flux scales with the square root of the ion mass, with the implication that the global confinement degrades with increasing ion mass. Using nonlinear gyrokinetic simulations of DIII-D, we illustrate a remarkable transition in the turbulent isotope scaling towards the plasma L-mode edge. The transition is controlled by finite electron-to-ion mass-ratio dependence of the nonadiabatic electron response, dominantly generated by the parallel motion, which represents a correction to bounce-averaging of the electrons. The nonadiabatic electron drive strongly regulates the turbulence levels and plays a key role in altering -- and in the case of the DIII-D edge, reversing -- the simple gyroBohm scaling rule. The finite electron-mass correction is larger for light ions and increases with increasing $q$ so that, while it is weak in the core, it dominates the mass scaling in the edge. Overall, these results may have favorable implications for global energy confinement and for the power threshold for the L-mode to H-mode transition in a reactor like ITER from H to D to DT, consistent with recent experimental observations comparing hydrogen and deuterium plasmas {2}.
Theoretical basis for gyrokinetic isotope scaling
The ion gyrokinetic equation together with the assumption of purely adiabatic electrons describes ion energy fluxes $Q_a$ that exhibit simple gyroBohm scaling:
$\qquad Q_a = C_0 \, Q_{\rm GBa} \quad\text{where}\quad Q_{\rm GBa} = Q_{\rm GBD} \sqrt{m_a/m_{\rm D}} \; .$
Here, the subscript $a$ is the species index, $Q_{\rm GBD} \doteq n_e T_e c_{sD} \rho_{*D}^2$ is the deuterium gyroBohm energy flux, $c_{sD}=\sqrt{T_e/m_D}$ is the deuterium sound speed and $\rho_{*D}=(c_{sD}/\Omega_{D,{\rm unit}})/ a$ is the normalized deuterium ion-sound gyroradius. Because $C_0$ is species-independent, we must always observe $Q_{\rm H} < Q_{\rm D} < Q_{\rm T}$, i.e. heavier isotopes should give rise to confinement degradation. When kinetic electron dynamics are fully retained, however, we expect the more complicated true gyroBohm scaling:
$\qquad Q_a = C\left(m_e/m_a\right) Q_{GBa} \; ,$
that contains an additional electron-to-ion mass-ratio dependence. We observe that the $m_e/m_a$ mass dependence of $C$ typically opposes the simple gyroBohm mass dependence, and can in some cases dominate and reverse the gyroBohm dependence so that $Q_{\rm H} > Q_{\rm D} > Q_{\rm T}$.
Reversal of simple gyroBohm scaling in electron-transport dominated edge regimes
Using CGYRO {3} we gauge the influence of kinetic electrons on the isotope scaling of energy flux in the transition from ion-dominated (core) transport regimes to electron-dominated (edge-typical) transport regimes. Simulation parameters are based on DIII-D #173147 at t=1705ms, an ohmically-heated L-mode discharge. Fig. 1 shows that CGYRO matches the total (e+i) experimental power-balance flux in both the ion-dominated core and the electron-dominated edge ($r/a \ge 0.9$), where $Q_e \sim 1.5 \, Q_i$ and TGLF underestimates the edge electron transport. The dominant linear mode in the core is ion-temperature-gradient (ITG) driven, whereas in the edge an electron temperature gradient-driven trapped electron mode (TEM) dominates. Fig. 2 compares the simulated ion energy flux for deuterium versus hydrogen versus 50:50 DT as the main ion species (with all other experimental parameters fixed). In the ITG-dominated regime, $Q_{\rm H} \sim Q_{\rm D} \sim Q_{\rm DT}$, meaning simple gyroBohm scaling is broken. This is dominantly due to electron collisions, which more strongly stabilize heavier species {4}, and weakly to the ${\mathbf E} \times {\mathbf B}$ flow shear {5}. However, a near gyroBohm scaling can be recovered by scaling the electron collision rate and the ${\mathbf E} \times {\mathbf B}$ shearing rate with the main ion thermal speed. In contrast, in the TEM-dominated edge regime, a strong reversal from the gyroBohm scaling is found, with $Q_{\rm H} \gg Q_{\rm D} \gg Q_{\rm DT}$. This implies that hydrogen confinement relative to deuterium is expected to be significantly worse than expected by the simple gyroBohm mass scaling. We demonstrate that this reversal is due to the nonadiabatic electron drive from the kinetic electron parallel response which acts to enhance the TEM turbulence for light ions.
Key role of the nonadiabatic electron response
The reversal from gyroBohm scaling in the electron transport-dominated edge is controlled by finite electron-to-ion mass-ratio dependence of the nonadiabatic electron response {6}. At fixed plasma gradients, this nonadiabatic effect is strongly enhanced at increased $q$, as shown in Fig. 3, and thus dominates in the plasma edge. This is consistent with the $q$-dependence of the electron parallel timescale relative to the ion drift timescale: $({\rm v}_i/a) \tau_e \sim q (R_0/a) {\rm v}_i/{\rm v}_e \sim q (R_0/a) \sqrt{m_e/m_i}$. For massless electrons, $\tau_e \rightarrow 0$, such that the passing nonadiabatic distribution vanishes and the trapped distribution is bounce-averaged and independent of mass ratio. At finite electron mass, the nonadiabatic correction increases with $q$ and decreases with $m_i$. Thus, for light species like hydrogen, deviation from the bounce-averaged limit is larger than for deuterium. Electron collisions provide a secondary mass-ratio correction to the flux scaling, so to recover simple gyroBohm scaling at low $q$ it is also necessary to reduce the collision frequency to eliminate $m_e$-dependence of the trapped-passing boundary layer width (shown in Fig. 3 for $q<2$ when ${\bar \nu}_e \rightarrow 0$).
Implications for global confinement and the L-H threshold in a reactor
For assessing the isotope scaling of global energy confinement in a reactor like ITER, it is essential to properly treat the precise nonadiabatic electron dynamics. Fluid or even bounce-averaged electron models are unlikely to recover the correct ion-mass scaling. For a full transport analysis, additional influences (e.g. impurities, heating, MHD) beyond the scope of this work must also be considered. However, plasma confinement is known to be sensitive to edge conditions. Tokamak L-mode edge conditions typically lead to electron transport-dominated turbulence regimes such as studied here, for which the nonadiabatic electron drive is enhanced, resulting in a favorable reversal of the simple gyroBohm scaling with ion mass from H to D to DT. This has implications for lowering the power threshold for the L-mode to H-mode transition in a reactor like ITER and could trend the theoretical turbulent-based global energy confinement isotope scaling toward agreement with experimental observations.
This work was funded by US DOE Grants DE-FG02-95ER54309 and DE-FC02-06ER54873.
{1} ITER Physics Basis Editors, Nucl. Fusion 39, 2175 (1999).
{2} C.F. Maggi et al., Plasma Phys. Control. Fusion 60, 014045 (2018).
{3} M. Nakata et al., Phys. Rev. Lett. 118, 165002 (2017).
{4} J. Garcia et al., Nucl. Fusion 57, 014007 (2017).
{5} J. Candy, E.A. Belli, and R. Bravenec, J. Comput. Phys. 324, 73 (2016).
{6} E.A. Belli, J. Candy, and R.E. Waltz, Phys. Plasmas 26, 082305 (2019).Speaker: Ms Emily Ann Belli (General Atomics) -
An accurate and predictive model for turbulent transport fluxes driven by microinstabilities is a vital component of first-principle-based tokamak plasma simulation. However, tokamak scenario prediction over energy confinement timescales is not routinely feasible by direct numerical simulation with nonlinear gyrokinetic codes. Reduced order modelling with quasilinear turbulent transport models provides significant computational speedup, and is justified in many regimes. The justification of the quasilinear approximation for transport driving spatial scales is a consequence of the underlying structure of tokamak microturbulence, and is validated by comparison to nonlinear simulations. This approach has emerged as a successful tool for prediction of core tokamak plasma profiles. We focus on significant progress in the quasilinear gyrokinetic transport model QuaLiKiz [1,2], and its application within flux driven integrated tokamak simulation suites.
To model 1s of JET plasma on order of 24 hours with 10 CPUs, QuaLiKiz employs an approximated solution of the mode structures to significantly speed up the computation time compared to full linear gyrokinetic solvers. Additional approximations include maintaining shifted-circle $(\hat{s}-\alpha)$ geometry, and the electrostatic limit. These approximations, together with optimisation of the dispersion relation solution algorithm within integrated modelling applications, leads to flux calculations $10^{6-7}$ faster than local nonlinear gyrokinetic simulations. This allows tractable simulation of flux-driven dynamic profile evolution over multiple confinement times including all transport channels: ion and electron heat, main particles, impurities, and momentum. QuaLiKiz is open source and available at www.qualikiz.com.
In this contribution, we will summarize the justification of the quasilinear approximation [3,4], sketch the basis of the QuaLiKiz transport model and its validity in comparison to nonlinear simulations, and illustrate validation of the model against experimental measurements at JET through flux-driven simulations within the JINTRAC integrated modelling suite [5,6], see figure 1 for an example. This capability 1) enhances the interpretation of present-day experiments, 2) enables “Predict First” simulations to aid with experimental optimization, and 3) allows theory-based extrapolation to future machine performance, at least with respect to core turbulence physics. While we focus here on JINTRAC simulations, QuaLiKiz is also coupled to the ASTRA [7,8], CRONOS [9] and ETS [10] integrated modelling codes.
Recent QuaLiKiz applications within integrated modelling include: W-accumulation interpretation and optimization, where the QuaLiKiz prediction of background kinetic profiles is critical for setting the neoclassical heavy impurity transport level [11-13]; modelling of multiple-isotope experiments at JET, where fast isotope mixing in the Ion Temperature Gradient (ITG) regime is crucial for experimental interpretation and has important implications for potential scenarios in JET DT, as well as for reactor burn control [14]; development of Uncertainty Quantification methods using Gaussian Process Regression to enhance statistical rigour in model validation, providing avenues for error propagation within QuaLiKiz simulations in integrated modelling [15]; predictive modelling for ITER scenarios, which predict the target Q∼10 when using a theory-based pedestal boundary condition [16]; and predictive modelling for DTT scenarios [17].
Beyond standard application within integrated modelling, QuaLiKiz has been leveraged for the development of realtime calculation capability for scenario optimization and realtime-oriented applications. This is based on machine learning methods, where a large database of pre-calculated QuaLiKiz runs is used to train feedforward neural networks to accurately reproduce model predictions. The neural network transport model provides a further 6 orders of magnitude speedup, 1 trillion times faster than the anchoring nonlinear simulations [18]. By coupling to the RAPTOR [19] control-oriented fast tokamak simulator, realtime-capable transport predictions are possible. This opens up a plethora of possibilities and innovation in realtime controller design and validation, scenario preparation, and discharge optimization.
While QuaLiKiz has had significant predictive success, continuously challenging and improving the model is a crucial component for instilling validity in wide parameter space. Beyond its role in experimental interpretation and prediction, reduced models such as QuaLiKiz are a key player in the multi-fidelity model hierarchy due to its feasibility for systematic comparison with experiments and identifying trends in model validation. This spurs further research, also incorporating higher fidelity linear and nonlinear models, ultimately improving our understanding of core tokamak turbulence physics.
We thus conclude with an overview of recent work dedicated to testing and improving the underlying QuaLiKiz assumptions. This includes: modification of the collisionality model, critical for obtaining the correct parameter dependencies of Trapped Electron Modes (TEM); validating the QuaLiKiz Electron Temperature Gradient (ETG) model versus multi-scale nonlinear GENE simulations; testing validity of QuaLiKiz towards the L-mode edge, where the standard ITG/TEM/ETG paradigm breaks down at high collisionality, due to the onset of modes with a drift-resistive nature, currently out of QuaLiKiz scope; testing the impact of s-α geometry on the turbulence regime, compared to full geometry, particularly at more outer radii where shaping effects are more prominent. Future work will extend QuaLiKiz to electromagnetic regimes.
References
[ 1] J. Citrin et al., Plasma Phys. Control. Fusion 59 124005 (2017), and http://qualikiz.com
[2] C. Bourdelle et al., Plasma Phys. Control. Fusion 58 014036 (2016)
[3] A Casati et al,. Nucl. Fusion 49 085012 (2009)
[4] J Citrin et al., Phys. Plasmas 19 062305 (2012)
[5] G. Cenacchi and A. Taroni, JET-IR , 84 (1988), eNEA-RT-TIB–88-5
[6] M. Romanelli et al., Plasma and Fusion Research 9 3403023 (2014)
[7] G. V. Pereverzev et al., IPP Report 5/42 (August 1991)
[8] E. Fable et al., Plasma Phys. Control. Fusion 55 124028 (2013)
[9] J.F. Artaud et al., Nucl. Fusion 50 043001 (2010)
[10] D. Kalupin et al., Nucl. Fusion 53 123007 (2013)
[11] S Breton et al., Nucl. Fusion 58 96003 (2018)
[12] F. Casson et al., submitted to Nucl. Fusion
[13] O Linder et al., Nucl. Fusion 59 016003 (2019)
[14] M. Marin et al., Nucl. Fusion 60 046007 (2020); and this conference
[15] A. Ho et al., Nucl. Fusion 59 056007 (2019)
[16] P. Mantica et al., Plasma Phys. Control. Fusion 62 014021 (2020)
[17] I. Casiraghi. P. Mantica et al., this conference
[18] K.L. van de Plassche et al., Physics of Plasmas 27, 022310 (2020) ; and this conference
[19] F. Felici et al., Plasma Phys. Control. Fusion 54 025002 (2012)Speaker: Jonathan Citrin (FOM DIFFER - Dutch Institute for Fundamental Energy Research) -
Outline A novel scheme to predict the plasma turbulent transport is developed by combining the machine learning technique and the first-principle gyrokinetic simulations. The machine learning technique is applied to find the relevant input parameters of the nonlinear gyrokinetic simulations which should be performed and to optimize the reduced transport model. The developed scheme can drastically reduce the computational costs to perform the quantitative predictions of the plasma profiles and the turbulent transport levels. Utilizing the scheme, the quantitative predictions for the turbulent transport can be realized by only one-time first-principle simulation for each radial position.
Conventional transport prediction schemes The first-principle simulation based on the gyrokinetics is powerful and reliable way to predict the turbulent transport or the plasma profiles. Indeed, it has been possible to validate the gyrokinetic simulations against the experimental observations within the experimental errors [1]. For the predictions, there are two main schemes. One is performing many first-principle simulations, which is called flux-matching technique [2]. The other is employing the reduced transport model, which is constructed by results of many gyrokinetic simulations. In the former scheme, by performing gyrokinetic simulations with changing the input parameters for the plasma profiles, we can find the transport fluxes that agree with the observations quantitatively. However, to obtain the matched transport fluxes, numerous gyrokinetic runs should be demanded. For example, even in the case of the single-species plasma, the conventional flux-matching demands 5 times or more runs for each radial position. Therefore, in particular, for the case of multi-species plasmas, we must perform a huge number of the simulations in the multi-dimensional parameter space which is exponentially expanding with increasing the number of the species. In the latter scheme, on the other hand, the reduced model enables us to obtain the turbulent transport fluxes without additional nonlinear gyrokinetic runs. However, since such reduced models are constructed by the limited parameters of the plasma profiles and include certain prediction errors, it is hard to reproduce the nonlinear simulation results precisely. In this work, by combining the first-principle simulations and the machine learning techniques via the reduced model, we reduce the number of the gyrokinetic runs and realize more reliable and efficient predictions.
A new transport prediction scheme We develop a new transport prediction scheme, which consists of three parts as shown in Fig.1. Here, we consider the ion heat transport as an example. To reduce the number of the first-principle simulations, we have to find the relevant input parameters of the simulations to realize the resultant transport fluxes which is close to the experimental results. First, in the developed scheme, for finding the input parameters of the simulation, we employ the reduced model for the ion heat diffusivity [3], which is constructed under the adiabatic electron assumption. The model includes the turbulent contribution $\cal L$ and the zonal-flow contribution $\tau_{\rm ZF}$ as $\chi_{\rm i}^{\rm model}/\chi_{\rm i}^{\rm GB} = A_1 {\cal L}^{\alpha_0} / (A_2 + \tau_{\rm ZF}/{\cal L}^{1/2})$. Here, ${\cal L} \equiv a(\rho) [R/L_{T{\rm i}} - \beta_0 R/L_{T{\rm i}}^{\rm cr}]$ with the critical temperature gradient $R/L_{T{\rm i}}^{\rm cr}(\rho)$, $\tau_{\rm ZF}$ is the zonal-flow decay time, and $A_1, A_2$, $\alpha_0$, and $\beta_0$ are constant numbers determined in the model. Using the machine learning numerical library [4] for the initial model, we find the initial guess of the temperature gradient which realizes the transport flux $Q_{\rm i} = -n_{\rm i} \chi_{\rm i} \nabla T_{\rm i}$ that agrees with the observations in the target plasma. Second, using the guessed parameter, one trial run of the first-principle simulation is performed for each radial position. Third, using the results of the trial run for each radial position, the machine learning is performed again to optimize the reduced model by tuning the parameters with $\alpha_0 \to \alpha$ and $\beta_0 \to \beta$ in the initial model. Then we can obtain the optimized transport model $\chi_{\rm i}^{\rm opt}$ suitable for the target plasma by only one gyrokinetic simulation at each radial position.
Figure 2(a) shows the comparison of the temperature gradient dependences of the ion heat diffusivities obtained by the initial reduced model and the optimized model for the ITG turbulent transport in the high-$T_{\rm i}$ plasma in the LHD [5]. Compared with the initial model, the optimized model with $\alpha/\alpha_0=0.89$ and $\beta/\beta_0=0.92$ quite agree with the nonlinear runs which are never used for the construction of the optimized model. In addition, it is confirmed that the optimized model can reproduce the weakening of the profile stiffness due to the nonlinear effects. Furthermore, as shown in Fig.2(b) for the convergence checks of the developed scheme, one trial run is enough to construct the optimized model for each radial position because the machine learning via the initial model can guess the relevant temperature gradient which is close to the guess from the nonlinear runs, independently. Therefore, the scheme can reduce the number of the first-principle simulation runs to only once for each radial position. The reduction of the computation time will be more remarkable in the multi-species case.
Using the optimized model, we can obtain the guesses for $R/L_{T {\rm i}}$ and predictions for $\chi_{\rm i}$ as shown in Fig.3. Although the temperature gradients guessed by the initial and the optimized models are not so different from each other, the optimized model can reproduce the turbulent diffusivities by the nonlinear simulations better than the initial model. At least in this application, we can obtain the transport levels which quantitatively agree with the nonlinear runs by using the developed scheme with performing just one-time first-principle simulation for each radial position. Since the developed scheme enables us to thoroughly reduce the number of the first-principle simulations, the scheme can be applied to the integrated transport code for quick and precise predictions of the plasma profiles under the operation scenarios with drastically saved computational resources.
[1] M. Nunami, et al., Phys. Plasmas 25, 082504 (2018).
[2] T. G$\ddot{\rm o}$rler, et al., Phys. Plasmas 21, 122307 (2014).
[3] S. Toda, et al., J. Phys. Conf. Ser. 561, 012020 (2014).
[4] M. Abadi, et al., arXiv:1603.04467 (2016).
[5] Y. Takeiri, et al., Nucl. Fusion 57, 102023 (2017).Speaker: Dr Masanori Nunami (National Institute for Fusion Science) -
Core transport in present tokamaks is mostly ascribed to micro-turbulence driven by the non-linear saturation of ion-scale ITG-TEM [1] instabilities ($k_\theta\rho_i\le1$, where $k_\theta$ is the poloidal wave number and $\rho_i$ the ion Larmor radius). It has been shown that electron-scale ETGs [2] ($k_\theta\rho_e\le1$) can also impact the heat transport, also exchanging energy with ITG-TEM turbulence by multi-scale coupling [3-9]. This topic of investigation gains a particular relevance due to its potential impact on devices like ITER, dominated by electron heating. ETG modes have been shown to play a role in plasmas with mixed ion and electron heating, since a proper balance of ion heating, decreasing the ETG threshold in $T_e$ gradient (which increases with increasing $T_e/T_i$ [10]), and electron heating (pushing $T_e$ gradient towards threshold while increasing the threshold due to $T_e/T_i$ increase), could destabilize them. Also all mechanisms that stabilize ITGs, such $E\times B$ or fast ions from neutral beams (NBI) and/or ion cyclotron resonance heating (ICRH), due to multiscale interactions open a window favourable for ETG destabilization.
The response of the $T_e$ profiles to the applied heating can be experimentally investigated by performing normalized electron heat flux scans and/or RF power modulation analysis. The two methods can be used in conjunction to extract information on the dependence of the gyro-Bohm normalized electron heat flux $q_{egB}$ on the normalised $T_e$ logarithmic gradient $R/L_{Te}$, yielding experimental values for the threshold $R/L_{Te,crit}$ for the onset of turbulent transport and for the ‘electron stiffness’ $\partial q_{egB} /\partial R/L_{Te}$. The experimental results can be compared with the output of gyrokinetic (GK) simulations, which infer both $R/L_{Te,crit}$ (fast linear runs), and the dependence of the saturated heat flux on $R/L_{Te}$ (more costly nonlinear runs). Resolving both ion and electron scales (i.e. performing nonlinear multi-scale simulations) is computationally very demanding and just became possible in the last years.
In order to access a broad range of parameters, a great effort is actually devoted to analyse different machines, comparing experimental and numerical results, within the framework of EUROfusion and of the ITPA Transport & Confinement group. In this paper, the analysis of plasmas of three different tokamaks, i.e. the Joint European Torus (JET, at Culham, UK), ASDEX Upgrade (AUG, at Garching, DE) and the Tokamak à Configuration Variable (TCV, at Lausanne, CH), is presented. Dedicated plasma discharges have been analysed experimentally and modelled numerically, by means of GK codes (GENE [11] and GKW [12]) and reduced quasi-linear models (TGLF [13] and QuaLiKiz [14]). The results of the different tokamaks concur to make a general picture indicating that ETGs could also be important for electron heat transport in fusion relevant conditions, in particular when $T_e\sim T_i$ with consistent fast ion density.
TCV is equipped with an NBI system, that allows the plasma to achieve $T_e\sim T_i$ in conjunction with high $R/L_{Te}$ (due to ECRH), allowing to access parameters compatible with ETGs. Two dedicated L-mode discharges, with $B_0=1.41$ T, $I_p=170$ kA have been performed with a different proportion of deposited ECRH power on- vs off-axis to perform a heat flux scan. Each pulse presented different phases corresponding to a different proportion of NBI/ECRH power to vary $T_e/T_i$, with ECRH both steady and modulated to allow a perturbative analysis. Both the experimental analysis and GK modelling (linear multi-scale and nonlinear ion-scale simulations) tend to indicate a possible role of ETGs at mid-radius when both ECRH and NBI are injected simultaneously, and at a larger toroidal radius $ρ_{tor}=0.7$ also when only ECRH is injected. In the former case, the main mechanism which explains the failure of ion scales alone to explain the experimental fluxes, is the stabilisation of ion-scales by the fast ions that are produced by the NBI. These results, published in [6], provide hints of a contribution of ETGs to electron heat transport in TCV plasmas.
Experiments on the AUG tokamak to study electron heat transport [5] have produced H-mode discharges with $B_0=2.5$ T, $I_p=0.8$ MA, injecting 2.5 MW of ECRH (steady and modulated to perform the perturbative analysis) and 5 MW of NBI in order to have $T_e\sim T_i$. Different discharges had different proportions of ECRH power deposition on- vs off-axis, in order to obtain the heat flux scan. At mid-radius, both the electron heat pulse diffusivity $χ^{HP}$ (from perturbative analysis) and $q_{egB}$ (from steady state scan), indicate strong turbulence levels above $R/L_{Te}\sim$ 6-7, leading to a moderate/high electron stiffness, consistent with the possible presence of ETGs. Both GENE and GKW linear-gyrokinetic simulations predict a role for ETGs for $R/L_{Te}>6$, based on an effective model for nonlinear turbulence saturation [15]. Both ion-scale and multi-scale simulations are being performed in order to analyse the most representative AUG pulse, to test the existence of an ETG ‘wall’ limiting the achievable $R/L_{Te}$. The preliminary multi-scale results are in agreement with TGLF in indicating a >30% contribution of ETGs to the electron heat flux for the cases close to threshold, setting high electron stiffness above it (see figure 1-2).
Following early results pointing to an important role of ETGs in JET [4], very recently dedicated sessions on ETGs have been performed at JET. Both L- and H-mode plasmas have been obtained, with $B_T=3.3$ T, $I_p=2$ MA, injecting 0-20 MW of NBI and up to 6 MW of ICRH (H minority, mainly heating electrons), achieving heat flux scans for a range of $T_e/T_i$ values. The preliminary analysis of the experimental data indicates that JET results are very similar to the AUG ones, with a strong increase of the electron stiffness for $R/L_{Te}>6$ (see figure 3). L-mode cases, in particular, allow to obtain sufficiently large values of $q_{egB}$ at large $R/L_{Te}>6$, giving the hint of a possible ETG ‘wall’. In parallel, high performance hybrid discharges are analyzed in order to study the ETG impact on these scenarios. Both sets of data are being modelled by means of single scale GK simulations and reduced models, in order to set the basis for heavier multi-scale GK simulations.
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014– 2018 and 2019– 2020 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This work was also conducted under the auspices of the ITPA Topical Group on Transport and Confinement. We acknowledge the CINECA award under the ISCRA initiative, for the availability of high performance computing resources and support.
[1] W. Horton, Rev. Mod. Phys. 71, 735 (1999)
[2] W. Dorland et al., Phys. Rev. Lett. 85, 5579 (2000)
[3] N.T. Howard et al., Nucl. Fusion 56, 014004 (2016)
[4] N. Bonanomi et al. Nucl. Fusion 58, 124003 (2018)
[5] F. Ryter et al., Nucl. Fusion 59, 096052 (2019)
[6] A. Mariani et al., Nucl. Fusion 59, 126017 (2019)
[7] S. Maeyama et al., Phys. Rev. Lett. 114, 255002 (2015)
[8] A. Marinoni et al., Nucl. Fusion 57, 126014 (2017)[9] C. Holland et al., Nucl. Fusion 57, 066043 (2017)
[10] F. Jenko et al., Phys. Plasmas 8, 4096 (2001)
[11] F. Jenko et al., Phys. Plasmas 7, 1904 (2000)
[12] A. Peeters et al., Comput. Phys. Commun. 180, 2650 (2009)
[13] G.M. Staebler et al., Phys. Plasmas 23, 062518 (2016)
[14] J. Citrin et al., Plasma Phys. Control. Fusion 59, 124005 (2017)
[15] G.M. Staebler et al., Nucl. Fusion 57, 066046 (2017)Speaker: Alberto Mariani (Istituto per la Scienza e la Tecnologia dei Plasmi, CNR, Milano (Italy)) -
Recent experiments in JET-ILW have been successfully exploring a high-performance H-mode scenario with no gas dosing at low $q_{95}$ ($I_p=3$ MA, $B_t$=2.8 T, $q_{95}=$ 3.2) and low triangularity, with peak neutron rates reaching values of 3.6$\times 10^{16}$ s$^{-1}$. This was enabled by operation at very low gas fueling, which is challenging in JET with the metal wall due the need to control the W influx into the core region. By starting H-mode operation at high density, applying a high level of gas injection early during the NBI heating phase to avoid ELM-free phases, it was possible to reduce the gas puffing to very low levels ($\approx 10^{21}$e/s), achieving a high performance, low density regime (called no-gas regime in the rest of the text) with averaged $n_e\approx3\times 10^{19}$ m$^{-3}$ and Greenwald fraction of ≈0.35, amongst the lowest ever achieved in JET-ILW. Operation at such low densities allows decoupling ions from electrons, resulting in higher $T_i/T_e$ than what obtained in conventional ELMy H-modes at higher densities and similar heating power.
One of the best examples of this no-gas scenario (#94900) is shown in Fig. 1, compared to the reference ELMy H-mode discharge (#94777) at similar heating power. Both discharges are heated by 20 MW of neutral beam injection (NBI) and up to 4 MW of ion cyclotron resonance heating (ICRH). In discharge #94900 the gas puffing is switched off at 8.5 s, resulting in a strong decrease in edge density (from 7$\times 10^{19}$ m$^{-3}$ to 2$\times 10^{19}$ m$^{-3}$) and a significant increase in the density profile peaking. This is accompanied by an increase in pedestal temperatures ($T_{e,ped}\approx$1.5 keV, $T_{i,ped}\approx$2 keV), enhanced toroidal rotation ($v_{tor,0}\approx$450 km/s), improved core ion confinement ($T_{i,0}\approx$15 keV, $T_{i,0}\approx 2\times T_{e,0}$) and ELMs substantially smaller and faster than those of the reference with gas fuelling ($f_{ELM}\approx$60 and 25 Hz respectively). In the no-gas phase, the ion temperature, stored energy and neutron rate continuously increase until the appearance of a core MHD mode ($n$=4) triggered by a sawtooth crash (see sharp drop in the neutron rate at 10.5 s, Fig. 1(e)), suggesting the performance is limited by MHD rather than transport. It must be noted that this MHD event does not lead to a disruption, the discharge survives and is landed safely. Density and radiation remain essentially constant after the gas puff is switched off, indicating particle transport is fast enough to provide adequate density and impurity control. This behavior differs from what observed in the hot-ion H-mode developed in JET-C(1) where density and radiated power increased constantly during long ELM free periods, eventually leading to a radiation collapse and back transition to L-mode.
Despite the absence of gas puffing and strong electron density peaking of the no-gas scenario, which typically lead respectively to an increase in W source and to strong inward impurity convection, central impurity accumulation does not take place, the temperature of the outer target and the total radiated power are comparable to the reference discharge. Both discharges also show very similar 2D radiation patterns, with strong localization on the LFS midplane at $\psi_N$>0.8 and very low central values. Due to the much lower electron density at the pedestal top, the no-gas discharge reaches similar radiated power to the reference with a factor 4 increase in mid-Z (Ni/Fe/Cr/Cu) and high-Z (W) impurity concentrations, which in turn are the cause for the increased LOS-integrated $Z_{eff}$ measurement (Fig. 1(c)). Due to the strong localization of these impurities on the LFS midplane, their increased concentration does not affect the plasma center where $Z_{eff}$ remains < 1.4 as in the reference, so core dilution is also kept under control.
The no-gas scenario exhibits remarkably good absolute and normalized performance, albeit transient, reaching peak values of $H_{98}\approx$1.4, $\beta_N\approx$2.2, $W_{MHD}\approx$9 MJ, and this is achieved at much lower collisonality ($\nu_{e,ped}^*$< 0.1, close to ITER values) than the reference ELMy H-mode at higher density. The electron pedestal pressure is also slightly smaller than the reference, albeit at lower density and higher temperature, but there is a significant increase in core electron and ion pressures, resulting in a 35% increase in global energy confinement at the maximum stored energy. Improved transport driven by the increase in sheared $E\times B$ flow [2] is thought to contribute to both the strong peaking of the density profile and the improved performance. Additional effects associated with the large population of fast ions present in these plasmas, as shown in [3], might also play a role in the overall improved thermal transport. The added impurity control is provided by the increased ion temperature screening enhanced by the extreme toroidal rotation [4].
An especially interesting feature of the no-gas regime is the marked reduction in ELM size compared to conventional ELMy H-mode plasmas. With the decrease in edge density the type I ELMs are replaced by very small ELMs at a much higher frequency. ELM size increases as the pedestal pressure increases, but remains significantly smaller than those obtained in the reference pulse at higher density and similar heating power. We note that in those conditions the link between ELM size and pedestal collisionality and/or edge density typically found for Type I ELMs is lost [5]. Both the electron density and temperature pedestals become wider, and there is a substantial reduction in the maximum $\nabla n_e$ as the edge density decreases, resulting in a lower maximum $\nabla P_e$. Pedestal stability analysis indicates that the edge operating point is below the peeling-ballooning boundary, which might explain the absence of large type I ELMs. The underlying physics mechanisms responsible of the onset of these small ELMs are still a matter of ongoing investigation.
The new no-gas H-mode regime recently demonstrated in JET-ILW provides a valuable opportunity to study the confinement properties and ELM dynamics of high temperature plasmas with temperature and density profiles substantially different from those obtained in the conventional scenarios. With the aim of improving our understanding and increasing the accuracy of extrapolations for ITER, this scenario allows validating existing transport models and investigating the role of different physics mechanism involved in the observed improved energy confinement and impurity control.
ACKNOWLEDGEMENTS
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the EURATOM research program 2014-2018 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This research was supported in part by grant FIS2017-85252-R of the Spanish Research Agency, including ERDF-European Union funding.
References
1. The JET team, Plasma Phys. Control. Fusion 37 (1995), p. A35
2. J. Garcia et al., Plasma Phys. and Control. Fusion 61 (2019) 104002
3. J. Garcia et al., Nucl. Fusion 55 (2015) 053007
4. C. Angioni et al., Physics of Plasmas 22 (2015) 055902
5. A. Loarte et al., Plasma Phys. and Control. Fusion 45 (2003) 1549-1569Speaker: Elena De La Luna (Laboratorio Nacional de Fusión, CIEMAT, 28040, Madrid, Spain)
-
-
-
High performance advanced tokamak scenarios are very attractive for future burning plasmas. They can be achieved by elevating the central $q$-profile to values around unity to stabilize the sawtooth instability, which would otherwise reduce performance and could trigger deleterious instabilities. High-$\beta$ plasmas can develop such a flat elevated central $q$-profile in the presence of MHD modes that modify the current profile [1]. The self-regulating mechanisms leading to this anomalous evolution of magnetic flux can be referred to by the general term "magnetic flux pumping". At DIII-D, flux pumping was observed in the presence of a 3/2 tearing mode, as well as when inducing a helical core via external perturbation coils [2]. In the work presented here, experimental evidence of anomalous current redistribution due to the dynamo effect produced by a 1/1 quasi-interchange instability [3] is discussed. It is shown that the ability of the mode to redistribute the centrally driven current, and thereby to suppress sawteeth, scales with the plasma pressure. This is potentially important for future non-inductive tokamaks, as it could provide a way to redistribute the current driven by electron cyclotron current drive (ECCD), which drives current most efficiently in the plasma centre. The flux pumping mechanism would redistribute current outward, maintaining a flat central q-profile around unity and maximizing both current drive efficiency and plasma stability at high $\beta_N$.
A theoretical model based on recent simulations suggests that flux pumping can occur in the presence of a saturated ($m=1$, $n=1$) interchange-like mode [4]. The flow in the convection cell combines with the perturbation of the magnetic field via the dynamo effect to generate an effective negative loop voltage in the plasma core. This prevents the central current density from peaking and thereby flattens the core $q$-profile. The mechanism is self-regulating such that the core q-profile is clamped to values close to unity. Figure 1a shows the central electric field resulting from a 1/1 mode predicted by simulations, plotted against the central loop voltage necessary to keep q$_0$ around unity. The latter depends on internal and external parameters that lead to central current peaking, like externally induced current drive. In the cases that lie above the line, the magnetic flux pumping mechanism is sufficiently strong to prevent sawtoothing, whereas in the cases below the line, $q_0$ is below unity and sawtoothing occurs. The simulation results suggest that the strength of the flux pumping mechanism depends on the core pressure. This dependency on $\beta_N$ stems from the pressure-driven nature of the 1/1 quasi-interchange mode. The simulations shown here use a generic tokamak geometry, but simulations based on ASDEX Upgrade (AUG) discharges are underway. The experimental results shown in figure 1b support the theoretical model and will be discussed below.
With the combination of the imaging motional Stark effect diagnostic (IMSE) [5] at AUG and the IDE equilibrium solver [6,7], changes of $q$ as small as 0.1 are measurable, even in the plasma center. Together with the current drive capabilities of the upgraded electron cyclotron resonance heating (ECRH) systems [8], AUG constitutes the ideal device to perform experiments that test these simulations. In discharges featuring a 1/1 mode, positive ECCD current was applied in several steps to decrease $q_0$ and trigger sawteeth. At the same time, an NBI power scan was performed to increase the $\beta_N$ value over the threshold necessary for the mode to suppress sawteeth at a given central current drive.
Figure 2 shows the heating power, $\beta_N$, ECCD, 1/1 mode activity and $q_0$ in an AUG discharge with 800kA plasma current, a $q_{95}$ of 5.2 and an $H_{98}$ factor of 1.1. Five phases are discussed in the following. The red shading indicates the presence of sawteeth, the blue shading their absence. The first phase starts with large sawteeth. After $t=1.55$s, the measured central $q$ profile remains flat and clamped around unity. The last large sawtooth is observed around 1.62s, but small sawteeth still remain. At 1.65s, 1/1 mode activity first appears. Once $\beta_N$ is increased above 2.4 the small sawteeth disappear as well, even when the central ECCD is increased. When the ECCD is increased further, above 130kA, sporadic small sawteeth reappear and become more frequent with more ECCD. In the last phase, the driven current is reduced and the sawteeth disappear again.
The bottom panel shows the modelled $q_0$ in red, resulting from an equilibrium reconstruction which takes external magnetic measurements, kinetic profiles, current diffusion and a sawtooth current redistribution model into account [7]. The blue curve shows the estimated $q_0$ when additionally taking into account the local measurements from the IMSE diagnostic, which will be referred to as "measured $q_0$". It can be seen that in phase III, without sawteeth, $q_0$ should drop well below unity if no other current redistribution mechanism were present besides neo-classical current diffusion. The modelled and measured $q$ profiles for this phase are shown in the right panel. The measurements show that the central safety factor stays stable around one, suggesting an anomalous modification of the current profile. At the beginning of phase IV, the modelled $q_0$ is sporadically increased to unity by sawteeth, but drops well below 1 between the sawteeth. Since such a low $q_0$ would immediately trigger a sawooth, this suggests that the flux pumping mechanism still plays a role, but is not strong enough to completely suppress the sawteeth. This can also be seen in the measured $q_0$, which remains closer to 1.
The measured $\beta_N$ and ECCD current from the different phases in this experiment are plotted in figure 1b (diamonds). The circles show the results from a similar discharge with more heating power, resulting in a higher $\beta_N$. For a comparison with the theoretical predictions (figure 1a), here $\beta_N$ is used as a proxy for the electric field that can be created by the 1/1 mode and the central ECCD current as a proxy for the electric field necessary to keep the central $q$ around unity. At a given ECCD current, $\beta_N$ needs to exceed a certain threshold to enter the sawtooth free regime. At a higher ECCD current, this threshold increases. This supports the simulation results from reference [4] where the flux pumping mechanism in the simulations is only able to prevent sawtoothing at sufficiently high $\beta_N$, and where the threshold is dependent on central current drive peaking.
In the proposed contribution, results from simulations based on the experimental data from AUG discharges will be presented. The qualitative and quantitative agreement with the electric field deficit in the experiment, calculated from the difference between the modelled and measured toroidal current, will be discussed.
[1] C. Petty et al, PRL, 102, 045005 (2009)
[2] P. Piovesan et al, NF, 57, 076014 (2017)
[3] S. C. Jardin et al, PRL, 115, 215001 (2015)
[4] I. Krebs et al, PoP, 24, 102511 (2017)
[5] O. Ford et al, RSI, 86, 093504 (2016)
[6] R. Fischer et al, FST, 69, 526-536 (2016)
[7] R. Fischer et al, NF, 59, 056010 (2019)
[8] J. Stober et al, PPCF, 62, 024012 (2020)Speaker: Dr Andreas Burckhart (Max Planck Institute for Plasma Physics) -
In recent years the multi-scale interaction between large-scale tearing modes and micro-scale turbulence has been found to be of paramount importance for thoroughly understanding the tearing mode physics and the island-induced transport, which will ultimately lead to developing a more effective method of the tearing mode control and optimizing the plasma performance in fusion devices, such as ITER [1-2]. In this work, the impact on micro-turbulence, meso-scale quasi-coherent modes (QCMs) and large-scale zonal flows as well as plasma rotations by naturally rotating $m/n$=1/1 and 2/1 islands and by externally magnetic perturbation-induced static $m/n$=2/1 island has been investigated using 2D ECE imaging and reflectometers in HL-2A and J-TEXT tokamaks [3-5]. The results indicate that for sufficient large islands there exist strong interactions between the tearing mode (TM) island and turbulence, including QCMs and zonal flows. The critical island width is in general consistent with the theoretical prediction [6].
In the case of rotating $m/n$=1/1 island generated in NBI heated plasmas at HL-2A, it is observed that for large islands ($W_c\geq$ 10$\rho_i\approx$ 4 cm) the electron temperature ($\tilde{T}_e$) and density fluctuations ($\tilde{n}_e$) are modulated periodically by the island rotation, with minimum amplitude at the O- and maximum at the X-point, respectively. The turbulence modulation is localized merely in the inner area of the island due to significant alteration of local profiles and gradient-driven turbulence. Evidence also reveals that for large islands turbulence spreading can occur across the island O-point from the region outside the island. The experimental observations of the turbulence modulation and spreading effects with large island are in general agreement with simulations for the trapped electron mode instability [3].
In the case of rotating $m/n$=2/1 island in HL-2A ohmic plasmas, it is found that for sufficient wide islands ($W_c\geq$ 4.5 cm) a quasi-coherent mode (QCM, peaked at $\sim$100-180 kHz) in density fluctuations is excited at the island boundary as the island O-point passes by, where the local $T_e$ profile is steepened, as shown in figure 1. Statistical analysis indicates that for the QCM excitation, a threshold value of $T_e$ gradient is needed and the QCM is solely observed in low density discharges, consistent with the linear ohmic confinement regime. These experimental evidences and also the linear stability calculations both suggest that the observed QCMs are driven by the trapped electron mode instability. Bi-spectral analysis further shows that there exists nonlinear coupling between the tearing mode and QCMs, whereas no nonlinear interaction is observed between the QCMs and ambient turbulence. The results verify that the observed QCMs are linearly driven by locally enhanced $T_e$ gradient with large islands, but not driven by ambient turbulence via nonlinear energy coupling [4].
In J-TEXT ohmic plasmas, the plasma rotations, geodesic acoustic mode (GAM) zonal flows and turbulence are found to be significantly modified by a static $m/n$=2/1 island induced by externally applied resonant magnetic perturbations (RMP). Whereas after the island formation the edge toroidal rotations shift from the counter-$I_p$ to co-$I_p$ direction and the perpendicular rotations from the electron to ion diamagnetic drift direction, both of rotation changes do not show much difference between the island O- and X-point. However, the turbulence level at the O-point is substantially lower than that at the X-point. The amplitude of the GAM zonal flows measured outside the island is damped and the peak frequency slightly increase after the RMP. In addition, at certain island size (e.g., $W\approx$ 3.8 and 4.5 cm in figure 2) the nonlinear coupling among ambient turbulence inside the island is considerably enhanced through the inverse energy cascading. However, this situation does not occur for a narrower or wider island, as depicted in figure 2. These results suggest a profound influence of the island size on the nonlinear interplay of turbulence and turbulent transport [5].
References:
[$1$] L. Bardóczi et al 2016 Phys. Rev. Lett. 116, 215001
[$2$] K. Ida et al 2018 Phys. Rev. Lett. 120, 245001
[3] M. Jiang et al 2019 Nucl. Fusion 59, 066019
[4] M. Jiang et al 2020 submitted to Nucl. Fusion
[5] M. Jiang et al 2019 Nucl. Fusion 59, 046003
[6] R. Fitzpatrick 1995 Phys. Plasmas 2, 825Speaker: Min Jiang (Southwestern Institute of Physics) -
A predictive 3D optimizing scheme in tokamaks is revealing a robust path of error field correction (EFC) across both resonant and non-resonant field spectrum. The new scheme essentially finds a way to deform tokamak plasmas in the presence of non-axisymmetric error fields while restoring a quasi-symmetry in particle orbits as much as possible. Such a “quasi-symmetric magnetic perturbation” (QSMP) has been predictively optimized by general perturbed equilibrium code (GPEC) {1} and successfully tested in DIII-D and KSTAR tokamak plasmas (Fig. 1). The QSMPs in experiments demonstrated no performance degradation despite the large overall amplitudes of 3D fields, as clearly compared with a resonant magnetic perturbation (RMP) or a typical non-resonant magnetic perturbation (NRMP) (Fig. 2). The results indicate that the tokamak EFC can be improved beyond the present resonant-overlap EFC approach alone, if the residual non-axisymmetry can be further compensated to a level of QSMP. The studies also validate that the torque response matrix, a unique product of the GPEC formulation, can be used to assess the degree of not only resonant but also non-resonant EF correction – which has been desired by ITER for a long time.
Successful EFC is critical in tokamaks for preventing disruptions, especially in next-step devices like ITER due to unfavorable scaling with high $B_T$ and $\beta_N$. Improved understanding of plasma response in the last decade allowed the development of a more reliable approach than earlier ones, using the resonant-overlap field {2}, i.e. error field component that triggers the dominant RMP response. This led to the successful multi-machine scaling of resonant EF thresholds against disruptive locked modes across wide operational regimes and 3D field spectra, including the n=1 and n=2 toroidal mode numbers {3}. Recent highlights include the successful EFC against 2/1 locked modes due to the EF by the high-field-side (HFS) using the low-field-side (LFS) coils in COMPASS and NSTX-U, despite an overall increase of non-axisymmetry in both cases. These experiments, however, also posed an important question that must be addressed; how to quantify and compensate the next key mode, or simply NRMPs. The residual EFs after the dominant RMP correction remained disruptive during L-H transitions in COMPASS {4}, and significantly degraded performance in evolving discharges in NSTX-U {5}. Such residual EFs also generally drive rotational damping through uncorrected NRMPs as shown in DIII-D {6}.
It turns out that the residual EF effects can be almost entirely suppressed if the additional coils are available to minimize the neoclassical toroidal viscosity (NTV) simultaneously with the dominant RMP correction. To experimentally demonstrate this, it is necessary to have 3 rows of coils – one coil to generate a strong proxy EF and RMP response, another coil to compensate the dominant resonant response while leaving a NRMP response, and yet another to minimize the remaining NRMP-driven NTV and leave only QSMPs. Figure 1 shows the n=1 coil configurations designed to test QSMPs in DIII-D and KSTAR, with the predicted torque profiles compared to RMPs and NRMPs. One can see that the local torque near resonant layers in RMP is significantly reduced in NRMP, and the global non-resonant torque in NRMP is minimized in QSMP. These fields were applied with maximum coil currents to high performance ($\beta_N\sim3$) DIII-D plasmas with ~9MW NBI power. As shown in Fig. 2, there was no change observed in performance and confinement during the QSMP, compared to the clear rotation braking with the NRMP or the strong density pumping and rotation braking with the RMP which eventually led to a locked mode. Surprisingly, the NRMP remained disruptive during L-H transition when tested with marginal DIII-D H-modes, but the effect could be eliminated in QSMP, suggesting a potential resolution to the aforementioned issue in COMPASS with the dominant RMP correction alone. The QSMP optimization was also performed in KSTAR discharges ($\beta_N\sim2$), again showing no performance changes in contrast to the RMP and NRMP (Fig. 3). Note that KSTAR can make a pure NRMP with strong NTV {7} as can be seen by complete elimination of local resonant torque in Fig. 1, by taking advantage of it’s 3 rows of in-vessel coils and also low intrinsic EF.
EFC will never eliminate all small 3D fields in a tokamak and it, in fact, commonly increases them when the correction coils are shaped very differently from the intrinsic error field sources. Instead, one can identify a safe and robust 3D state such as a QSMP and see if it is accessible in the course of EFC. The self-consistent perturbed equilibrium calculations with neoclassical transport in GPEC offer a torque response matrix $T(\psi)$, from which one can immediately predict the torque profile by quadratic operation $\Phi^\dagger\cdot T\cdot\Phi$, where $\Phi$ is a field spectrum vector on a toroidal surface or coil vector representing its amplitude and phase by complex numbers {1}. The minimum eigenstate of $T(\psi)$ then represents the best possible way to deform the plasma while sustaining the minimum variation in the field strength or action variation ($\delta B_L \approx 0$, $\delta J \approx 0$) as well as minimum resonant parallel current and corresponding resonant torque, i.e. achieving quasi-symmetry.In summary, new experiments demonstrated that a QSMP could be an ideal EFC state without any performance degradation, offering a new and complementary EFC approach in addition to the present resonant-overlap method and also a possible resolution on non-resonant error field correction problem. The torque response matrix in GPEC will enable the QSMP optimization in more complicated 3D tokamak environments such as ITER with many more 3D coils and potential EF sources. This QSMP is also an interesting concept in and of itself, as it holds a sizable local perturbation at least near the divertor and its possible utility will be further discussed. *This research was supported by U.S. DOE contracts #DE-AC02-09CH11466 (PPPL), #DE-FC02-04ER54698 (DIII-D), and also by the Korean Ministry of Science, ICT and Future Planning (KSTAR).
{1} J.-K. Park and N. C. Logan, Phys. Plasmas 24, 032505 (2017)
{2} J.-K. Park, N. C. Logan et al., “Assessment of EFC criteria for ITER”, ITPA MDC-19 Report (2017)
{3} N. C. Logan, J.-K. Park et al., “Scaling of the n=2 error field threshold in tokamaks”, submitted to Nucl. Fusion (2020)
{4} T. Markovic et al., the 45th EPS DPP in Prague, Czech Rep. (2018)
{5} N. Ferraro, J.-K. Park et al., Nucl. Fusion 59, 086021 (2019)
{6} C. Paz-Soldan, N. C. Logan et al., Nucl. Fusion 55, 083012 (2015)
{7} Y. In, Y. M. Jeon et al, Nucl. Fusion 59, 056009 (2019)Speaker: Jong-Kyu Park (Princeton Plasma Physics Laboratory) -
Full suppression of Edge Localized Modes (ELMs) by using n=4 resonant magnetic perturbations (RMPs) has been demonstrated for ITER for the first time (n is the toroidal mode number of the applied RMP). This is achieved in EAST plasmas with low input torque and tungsten divertor, thus also addressing significant scenario issues for ITER. In these conditions energy confinement does not drop significantly (<10%) when ELM suppression is achieved compared to the ELMy H-mode conditions, while core plasma tungsten concentration is clearly reduced. The target plasma for these experiments in EAST is chosen as close as possible to the ITER type-I ELMy H-mode operational scenario with low torque input. In these experiments the NBI torque is around $T_{\mathrm{NBI}}$ ~ 0.3-0.4Nm, which extrapolates to 11 – 16 Nm in ITER (compared to a total torque input of 35 Nm when 33 MW of NBI are used for heating). The observed ELM suppression window is consistent with the peak in the modeled edge stochasticity using the MARS-F code. ELM suppression is maintained up to 60% $N_{\mathrm{GW}}$ (the Greenwald density) by feedforward gas fueling after suppression is achieved. These results expand physical understanding and demonstrate the potential effectiveness of RMP for reliably controlling ELMs in future ITER high Q plasma scenarios.
Plasma energy confinement during ELM suppression in these EAST plasmas decrease very slightly (< 10%) compared to the Type I ELMy H-mode with no RMPs applied, as shown in Fig. 1. This is very different from previous observations using low n (n=1 and 2) RMPs in EAST with high $q_{95}$ (≥4) [Ref.1]. As shown in Fig. 1, ELMs are completely suppressed by using n =4 RMPs with odd parity (opposite phases in the upper and lower rows of coils current) in EAST but not when even parity is applied to a type-I ELMy H-mode plasma with $q_{95}$ ≈3.65 and low input torque (TNBI ~ 0.3-0.4Nm) leading to a plasma beta $\beta_N$ ~ 1.5-1.8, similar to that in the ITER Q =10 scenario. The electron and ion temperature are very similar with $T_{i0}$ ≈ $T_{e0}$ ≈2keV. Significant density pump out (20% reduction) takes place during ELM suppression, while the drop of stored energy is negligible (5%). The tungsten concentration (lower subgraph) is also reduced by a factor of 2 compared to type I ELMy H-mode when ELMN suppression is achieved, despite the lower plasma density. The threshold n=4 RMP current for full ELM suppression is around 2kA (⨉ four turns). Striations of the heat and particle fluxes at the divertor target are observed during ELM suppression and are consistent with the modeled magnetic footprint by TOP2D.
Suppression windows in both $q_{95}$ and plasma density are observed; in addition, lower plasma rotation favours access to ELM suppression. ELM suppression is achieved in a narrow $q_{95}$ window in [3.6, 3.8] with odd n=4 RMP configuration with a continuous q95 ($I_p$ ramp-up). ELM suppression can be maintained up to 60% of Greenwald density by feedforward gas fueling after suppression. It is interesting to note that there is not only an upper density but also a lower density threshold for ELM suppression of 40% $N_{\mathrm{GW}}$. Outside the $q_{95}$ and density window only ELM mitigation, not suppression, is observed.
The modelled magnitude of edge stochasticity taking into account the linear MHD plasma response evaluated with MARS-F [Ref.2] is found to be strongly linked to the observed ELM suppression effect. Although vacuum modeling by the MAPS code shows that the edge resonant harmonics are stronger for the even coil configuration, MARS-F shows that the plasma shielding effect is stronger for the even coil configuration than for the odd one. The edge resonances with plasma response shown in Fig. 2 are stronger for the odd coil configuration; this applies not only the resonant harmonic just near the pedestal top (m = -13, n =4) but to the other edge resonant harmonics as shown in Fig. 2. Therefore, the edge stochasticity, i.e. edge Chirikov parameter ($\sigma$) or stochastic layer width ($\delta_{\mathrm{ergodic}}$) might be a better figure of merit for ELM control optimization (previous studies for this parameter [Ref. 3] did not account for plasma response). Including plasma response leads to edge magnetic field stochasticity for odd parity configuration being higher than that for the even one, which is consistent with the experimental results in Fig. 1.
The observed ELM suppression window in this experiment provides a good opportunity to test the effectiveness of different figure of merits for ELM control optimization. MARS-F modelling shows that both the Chirikov parameter near the pedestal top and the normalized edge stochastic layer width have a peak at $q_{95}$~3.6-3.7 for the odd parity configuration as shown in Fig. 3. The resonant window modeled by $\delta_{\mathrm{ergodic}}$ agrees well with the suppression window observed in this experiment. Comparison between different criteria including pedestal top island width and edge X-point displacement to this experimental observation will also be presented. Further detailed modelling studies and experimental analysis of the 3D physics processes in these experiments will be addressed in this presentation, with the objective to expand physics understanding for ELM suppression and to provide a solid physics basis for its extrapolation to future burning plasma devices such as ITER.
This work is supported by the National Key R&D Program of China under Grant No. 2017YFE0301100, the National Natural Science Foundation of China under Grant No. 11875292 and U.S. DOE under DE-SC0020298.References:
(1) Sun Y et al, Phys. Rev. Lett. 117, 115001 (2016).
(2) Liu Y et al, Phys. Plasmas 17, 122502 (2010).
(3) Evans T et al, Phys. Rev. Lett. 92, 235003 (2004).Speaker: Prof. Youwen Sun (Institute of Plasma Physics, Chinese Academy of Scienses) -
ITER high Q operation requires the integration of high performance plasmas with high plasma density, good fast ion confinement with acceptable stationary and transient power fluxes to plasma facing components (1). To control transients associated with high performance plasmas (Edge Localized Modes or ELMs), ITER is equipped with a set of in-vessel coils that modify the edge magnetic field structure from 2-D to 3-D (so called Resonant Magnetic Perturbations or RMPs) thereby achieving the required control. While ELM control using this approach has been demonstrated in many tokamak experiments, there remain open issues regarding its integration with other requirements essential to achieve high Q operation in ITER, such as its effect on the achievable energy, particle and fast ion confinement, compatibility with semi-detached/radiative divertor operation, low input torque operation, pellet fueling, etc.
To address these integration issues a series of experiments has been carried out at the EAST tokamak in ITER-like H-mode plasmas with q_95 ~ 3.7, normalized plasma beta β_N ~ 1.7 and with NBI injected torque T ~ 0.3-0.4 Nm, which is similar (somewhat lower) to the normalized torque input of ITER plasmas with 33 MW of NBI. Control of ELMs has been explored with n = 2 and n = 4 symmetry for the toroidal waveforms of the currents applied to the EAST in-vessel ELM control coil set, which is composed of two rows of 8 coils above and below the midplane at the low field side of the tokamak (2). The magnitude of the applied current and the toroidal phase of the applied perturbation (for n = 2) has been scanned and their effects on energy and particle confinement, NBI fast ion losses and divertor power fluxes has been characterized.
As shown in Fig. 1 ELM suppression can be achieved with both n = 2 and n = 4 with similar levels of current in the RMP coils (12 kAt). However, the decrease of particle and energy confinement associated with the achievement of ELM suppression is much larger for n = 2 (30%) than for n = 4 (< 10%). This supports the use higher n RMP perturbations to optimize ELM suppression with regards to energy and particle confinement as required for ITER, where n = 3 or 4 are considered. It is important to note that ELM suppression with n = 4 RMPs can be achieved in a range of density 〈n_e 〉=0.4-0.6 n_Greenwald thereby having an upper and a lower density level.
Because of the lower effects on energy and particle confinement of n = 4 RMP, as required for ITER, the compatibility of pellet fueling with ELM suppression was investigated in this configuration. As shown in Fig. 2 the injection of pellets leads to sudden increases of the plasma density and to spikes in the D_alpha signal from the divertor. However, unlike ELMs, these are not accompanied by an increase in the divertor heat fluxes, unlike the ELMs that are triggered when no RMPs are applied. This is consistent with non-linear MHD simulations of pellet injection in H-modes that always show triggering of MHD activity when pellets are injected but not growing exponentially to trigger ELMs in all cases (3). When successive pellets are injected (10 Hz was used in experiments) and the plasma density increases beyond 0.65 n_Greenwald, ELM suppression between pellets is lost in a similar fashion as with gas fueling described below.
Studies have also been performed to investigate the integration of ELM suppression with radiative/semi-detached divertor operation for n = 2 and 4 RMPs. It has been found that divertor power flux reduction can be achieved while maintaining ELM suppression either by gas fueling or increasing intrinsic impurity radiation by Ne seeding. The reduction of the power flux is larger for the former approach as shown in Fig. 3, although ELM suppression is lost when the density increases to high values.
Interestingly, the divertor power flux and temperature decrease in both near separatrix and off-separatrix lobes, which is in contrast with previous EAST experiments in which ELM control was achieved by LHCD (4). This is consistent with the field line mapping shown in Fig. 3 (with TOP2D in the vacuum approximation) indicating that the outer lobes field lines in this n = 4 are not connected to plasma regions deep inside the separatrix (the pedestal plasma corresponds to 1-ρ_min ~0.05).
The paper will describe the detailed analysis of the wide range of experimental measurements obtained (including NBI losses) and will discuss implications for ITER.
This work is supported by the National Key R&D Program of China under Grant No. 2017YFE0301100 and the National Natural Science Foundation of China under Grant No. 11875292. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization.
References
(1) A. Loarte et al., Nucl. Fusion 54, 033007.
(2) Y. Sun et al. this conference.
(3) S. Futatani et. al. Nucl. Fusion 54, 073008.
(4) J. Li et. al. Nature Phys 9 (2013) 817.
Speaker: Alberto Loarte (ITER Organization) -
Next step fusion devices such as ITER will need a reliable method for controlling the quasi-periodic expulsion of a large amount of heat and particles onto the plasma-facing components caused by edge-localized modes (ELMs). Several options are currently being considered to achieve the required level of ELM-crash control in ITER; this includes operation in plasma regimes which naturally have no or very small ELMs and suppression of ELM-crashes by active control of edge pedestal with resonant magnetic perturbations (RMPs). Many experiments at different tokamak devices have been dedicated to providing a solution that can be applied to ITER by the application of different approaches. However, due to a lack of understanding of the ELM control mechanism, a reliable ELM control is far from perfect yet.
The characteristics of two different non-ELM-crash plasmas have been studied using 2D fluctuation imaging systems on the KSTAR; (1) RMP-driven ELM crash suppressed H-mode, (2) ELM-free (or ELM-less) H-mode plasmas.
The ELM-crash suppressed H-mode plasma by the RMP is characterized by the coexistence of the filamentary mode and smaller-scale turbulent eddies at the plasma edge $[1]$. We have identified filamentary mode similar to ELM filament that still maintained, even when the ELM crash has been completely suppressed on $H_{\alpha}$ signal by RMP. Correlation analysis among the ECEI channels showed that the RMP would keep enhancing turbulent fluctuations at the plasma edge toward the ELM-crash suppression. A cross-phase analysis showed that such edge turbulence has a rather broad dispersion with a wide range of wavenumber ($k_\mathrm{\theta}<1.1$ cm$^{-1}$) and frequency ($f<100$ kHz). A detailed analysis suggests that energy exchange between filamentary mode and RMP-driven turbulent fluctuations would be responsible for the ELM-crash suppression (Fig. 1). Also, the plasma perpendicular rotation and its local shear estimated by the movement of the turbulent eddies decreases rapidly at the transition of the ELM-crash suppression $[2]$. Such reduction of the perpendicular rotation and its local shear could affect the turbulent fluctuation level by RMP.
In contrast, the ELM-free H-mode plasma was accompanied by benign edge harmonic oscillations (EHOs) near the separatrix. The EHOs had long-wavelength (toroidal mode number, $n\le4$) and remained stable during the ELM-free phase, and the ELM crashes rarely occurred at this stage. The observed EHOs appeared discontinuously and synchronized with quasi-periodic RF ($f\sim500$ MHz) spikes, providing an enhanced particle transport from the core plasma. This could avoid an increase in plasma edge pressure and prevent ELM crashes. The bicoherence analysis revealed that there is a strong nonlinear interaction between EHOs, and the nonlinear interaction of EHOs has a significant effect on the ELM structure and dynamics (Fig. 2). In addition to the EHOs, rotational shear found to play a significant role in the ELM-free phase. If the rotational shear is sufficiently large enough compared to the typical ELMing H-mode plasma, the ELM crashes disappeared, and the fluctuation level from the EHO increased, resulting in an intense transport event.
* This work supported by the Korea Ministry of Science and ICT under NFRI R&D programs (NFRI-EN2001-11) and National Research Foundation of Korea (NRF) (No. 2019R1F1A1057545).References:
$[1]$ J. Lee et al, Phys. Rev. Lett. 117, 075001 (2016).
$[2]$ J. Lee et al, Nucl. Fusion 59, 066033 (2019).Speaker: Dr Jaehyun Lee (National Fusion Research Institute (NFRI)) -
KSTAR has clarified a set of unresolved 3-D physics issues that could be addressed in the ITER-like in-vessel 3-row, resonant magnetic perturbation (RMP) configurations. In particular, considering that one of the most critical metrics of RMP ELM-crash control would require the compatibility with the divertor heat fluxes under the given material constraints, a series of intentionally misaligned RMP configurations (IMC)$^{1,2}$ have been explored to reveal the relationship between RMP ELM control and divertor heat fluxes. Specifically, taking advantage of the time-resolved IR camera, each rotating IMC in either 3-row or a combination of 2-row IMCs helped us diagnose the ‘wet’ area of divertor in the vicinity of ELM-crash-suppression; ELM-crash-mitigation, ELM-crash-suppression, and mode-locking.
First of all, we have articulated the contrasting effect of kink (i.e. “away” phasing) vs anti-kink (i.e. “toward” phasing) responses on the ELM-crash suppression, as shown in Figure 1. Starting from a sub-marginal level of RMP current in a typical n=1, 90 deg phasing, the 3-row IMC in kink phasing (in red) becomes more kink-influenced, demonstrating the synergistic benefit of ‘kink’ phasing in ELM-crash-suppression. In contrast, the 3-row IMC in the anti-kink phasing (in green) becomes more insensitive to ELM-crashes at the sub-marginal level of RMP. In a way, this helps us recast the “away” and “toward” phasing as kink and anti-kink phasing respectively, as schematically shown in the lower left inset of Figure 1. Such experimental observation is in excellent agreement with what ideal MHD theory predicts $^{3}$. Previously, we had shown the divertor heat flux broadening with 3-row IMC-driven ELM-crash-suppression in both “away(kink)” and “toward (anti-kink)” phasings, while no such broadening was observed in the 2-row IMCs with top/bottom coils $^{1}$. Now, we have newly observed that the ‘wet’ area of ELM-crash-mitigation got more broadened than that of ELM-crash-suppression (not shown here), based on these 3-row IMC discharges.
Also, we have further investigated whether or not 2-row IMCs would be fundamentally deficient in divertor heat flux broadening during ELM-crash-suppression. In the earlier study, no mid-row was involved in the 2-row IMCs, although the mid-row is much more influential than the other off-mid rows. To clarify this issue, a set of 2-row IMCs, including mid-row, has been explored. Figure 2 shows the time evolutions of various plasma parameters and ‘wet’ area, where each phasing of 2-row IMC varies by the denoted angle in shades incrementally from a typical n=1, 90 degree phasing angle in the anti-kink direction. Throughout the whole IMC application period, the 2-row IMC-driven, ELM-crash-suppression has been accomplished, as shown at the bottom of Figure 2 (a). At the same time, no evidence of the divertor heat flux broadening can be found on the ‘wet’ area in this combination of middle/bottom row IMCs, as shown in Figure 2(b). Even for kink phasing in the 2-row IMCs with middle/bottom coils (as shown in Figure 3 (b)), a similar outcome has been obtained. Thus, it is a fair conclusion that the divertor heat flux broadening would require a third row, suggesting that the dispersal of the divertor heat flux in 3-row IMCs cannot be driven by helically structured 2-row IMCs alone. Nonetheless, no physics mechanism of the 3-row IMC-driven, divertor heat flux broadening during ELM-crash-suppression has been understood yet, while several hypotheses are being assessed$^{1}$. Interestingly, we have found that middle/bottom rows are much more effective in suppressing the ELM-crashes than top/mid rows, revealing strong up/down asymmetry in lower-single-null (LSN) plasmas, as shown in Figure 3.
Considering that the 3-row IMC-driven, ELM-crash-suppression in kink phasing have been securely obtained at 2.3 kA, a set of 2-row IMCs have been designed to compensate a missing off-mid row current (mid: 2.3 kA, off-mid: 4.6 kA). Surprisingly, such conditions led to a vastly contrasting outcome, proving a much more effective coupling of middle/bottom rows in ELM-crash-suppression than that of top/middle rows. In fact, there was a much lower threshold of ELM-crash-suppression in the combination of middle/bottom rows, even suggesting no need of top row (not shown here). This is reminiscent of the critical influence of X-point on RMP ELM control studied in MAST$^{4}$, though it was related to ELM-crash-mitigation, rather than ELM-crash-suppression.
Overall, the KSTAR has established a new holistic understanding of ITER-like RMP ELM control, elaborating various subtle points in the vicinity of ELM-crash-suppression and ‘wet’ area on divertor. These new findings in 3-D physics is expected to help us further reduce the uncertainty associated with 3-row ITER RMP.Acknowledgements
This work was supported by the Korean Ministry of Science and ICT for National Research Fund (NRF-2020M1A7A1A03007919), the UNIST research fund (1.180056.01), and the KSTAR project (NFRI-EN1901-11).References
1. Y. In et al., Nucl. Fusion 59 (2019) 126045
2. Y. In et al., Nucl. Fusion 59 (2019) 056009
3. J.K. Park et al, Nature Physics 14 (2018) 1223
4. A. Kirk et al, Nucl. Fusion 55 (2015) 043011Speaker: Yongkyoon In (Ulsan National Institute of Science and Technology) -
Injection of boron powder in the EAST X-point showed edge-localized mode (ELM) suppression with no confinement degradation over a wide range of operations. The work shows that ELM suppression was achieved by making the pedestal marginally "leaky” via an edge-localized GAM-induced particle transport. This approach potentially opens up novel methods for increasing the ELM suppression toolbox for future devices, such as ITER.
The power load onto plasma-facing components caused by type-I ELMs is a critical issue for the lifetime of in-vessel components in fusion reactors such as ITER. Significant effort has been invested in controlling ELMs, and the mainstream method to suppress ELMs is with the use of edge resonant magnetic perturbations (RMPs). A characteristic of ELM suppression by RMPs is the existence of a narrow q95 window of ELM suppression. Furthermore, such ELM suppression generally occurs with confinement degradation. ELM suppression coupled with optimum confinement over a wide range of operations are highly desirable for operational flexibility in ITER and future reactors. In addition to demonstrating ELM suppression without confinement degradation, this work investigates the physical mechanism leading to ELM suppression using boron powder injection on EAST.
Experimental results - ELM suppression was demonstrated on the EAST tokamak by injecting boron impurity in the upper X-point to investigate the effects of boron powder on tungsten core accumulation. Figure 1 displays the time history of relevant discharge parameters without and with boron injection (20 mg/s - a fraction of which produces localized density perturbation). Here, a proxy of boron injection is shown using the B-V emission (Fig. 1a). The presence of ELMs is indicated using the D$_{\alpha}$ signal (Fig.1b). It is clear from this figure that ELM suppression is associated with a slight increase in the stored energy (Fig. 1c) due to an overall increase of electron temperature for the same input power, as well as a reduction of carbon impurity (Fig. 1d). While in this discharge the core tungsten (W) is maintained albeit at a higher level than without B injection (Fig. 1f), other discharges have shown a reduction of core W accumulation.
ELM suppression is not caused by a cumulative/compound effect of boron on the divertor but induced by the interaction of boron powder with the background plasma. To rule out the cumulative effects of boron, a causality test was performed. Specifically, the injection timing was varied for repeated discharges to show that ELM suppression occurs after the boron emission reaches a threshold on the B-V monitor. In addition, we also showed that when boron injection is halted, ELMs return promptly. The direct effect of boron on ELM suppression provides a clear contrast from lithium induced ELM suppression, which is characterized by cumulative conditioning effects and a small reduction in stored energy [R. Maingi et al. 2018 Nucl. Fusion 58 024003].
We then turn to the interaction between the ablated boron powder and the background plasma by examining the fluctuations. ELM suppression is correlated with the onset of modes with multiple harmonics (see Fig 2). These n=1, 2,3 modes are detected in upper X-point (localization of the boron-induced density perturbation) using XUV (proxy for impurity and electron density fluctuations) detectors as well as on magnetic probes. Using multiple complementary diagnostics (BES, interferometers), we showed that these modes are radially localized in the edge near the separatrix.
A clear onset of the mode is observed when the boron emission reaches a certain threshold. Subsequently, the mode’s amplitude increases until saturation is reached. We observe that the mode onset is associated with a reduction of the core W signal, suggesting that the mode drives particles out. Evidence of the fluctuation at this mode’s frequency has been observed on Langmuir probes and divertor D$_{\alpha}$ signals corroborating the fact that particles are kicked out (though orbit loss) from the confined region to the SOL. In addition, the reduction of W is associated with the increase of mode amplitude, which is reminiscent of the edge-harmonic modes in Q-H mode being responsible for particle transport. A question remains: What is the nature of the mode?
Theory - The injection of boron occurs at the upper X-point and the ablated boron creates a local density perturbation. The driving mechanism of the GAM-like observations is described below.The local perturbation induced by the ablated boron leads to a poloidal asymmetry of charges. Such asymmetry will generate an oscillation akin to a Geodesic Acoustic Mode (GAM). While GAMs do not usually cause a net particle flux on their own, it is conceivable that this might change due to orbit loss for oscillations near the separatrix. Simulations are being performed to assess the existence of GAMs and kinetic calculations are in progress to investigate the GAM effectiveness of driving particles out if localized near the separatrix. The GAM frequency for radial wavenumber k$_r$=0 is calculated by solving a two-fluid eigenmode problem [K. Hallatschek, PPCF 49, B137 (2007), R. Hager et al., PPCF 55, 035009 (2013)] in realistic geometry. Because of the complex coupling between the GAM and the sound wave spectrum close to the separatrix, several GAM candidates exist that are classified according to their ratio of perpendicular to parallel kinetic energy E$_\perp$/E$_{\parallel}$ (Fig. 3). The mode with the largest fraction of perpendicular energy is the GAM with fGAM≈10 kHz at normalized poloidal flux $\psi_n$=0.995. But other (more soundwave-like) modes with non-negligible coupling to the radial electric field (E$_\perp$/E$_{\parallel}$∼10%) exist at lower frequencies between 2.8 and 6.6 kHz at $\psi_n$=0.995. The initial two-fluid calculations are consistent with the modes observed in the experiment. However, calculations of the accurate kinetic GAM frequency with nonlocal effects, in the absence of turbulence, are being performed using XGC to compare with the two-fluid calculations. Results presented could open new research opportunities in controlling ELMs by making the pedestal marginally “leaky” via edge-localized GAM-induced particle transport. Work by US DoE Work DE-AC02-09CH11466.Speaker: Ahmed Diallo (PPPL)
-
-
-
Scaling laws for the energy confinement time using engineering parameters as input are of great importance for fusion research. While they provide only limited insight in the actual physics determining the energy confinement, they are necessary tools for system studies and control schemes. It has become clear for both tokamaks and stellarators, however, that the scaling behavior is fundamentally different in various operation regimes. It is, hence, important to distinguish and characterize regions in the operational space according to their scaling behavior.
For stellarators, there are two almost identical cross-machine scalings available, called ISS95 and ISS04. In early low-density experiments of the optimized stellarator Wendelstein 7-X (W7-X), the energy confinement time scaling was well described by those scalings. W7-X has been designed and optimized to demonstrate fusion-relevant plasma performance during in stellarators during long-pulse operation. A positive density dependence of the energy confinement time would have favorable consequences to achieve this goal, since it is believed that fusion reactors would operate at densities between $1$ and $2\cdot10^{20}\,\mathrm{m}^{-3}$. Recent W7-X experiments at higher densities, however, show clear deviations from these scalings, especially concerning the density dependence, which is found to be weaker than suggested by the empirical scalings. This is visualized in Fig. 1, where the scaling exponent of the density ($\alpha$) is shown as a function of line-averaged density and ECRH power. The scaling exponent $\alpha$ is highest below $4\cdot10^{19}\,\mathrm{m}^{-3}$ and then remains relatively constant up until $6$ to $8\cdot10^{19}\,\mathrm{m}^{-3}$, depending on the heating power.
Both the plasmas used to derive the ISS04 scaling and the early plasmas of W7-X are relatively far away from fusion-relevant conditions. Hence, without understanding the exact origin of the favorable density scaling in stellarators, it cannot be assessed whether it extrapolates to fusion-relevant plasmas or not. Understanding the difference in the scaling behavior between the low-density and high-density plasmas in W7-X will help to elucidate this question. One issue with present energy confinement time scalings is that during the data selection process, plasmas with high radiation losses are excluded. This is done in order to scale as accurately as possible the energy transport with engineering or physics quantities without other influences that cannot be expected to scale accordingly. While this approach in principle makes sense, it introduces a bias towards low densities, since commonly the relative importance of radiation losses also increases with density. This is not only an issue because it is expected that power exhaust concepts of future fusion reactors include high levels of edge radiation but also because it potentially masks other effects that could occur while raising the density in today’s experiments. Up until now, two effects have been identified that likely contribute to the weaker scaling at higher densities: The characteristics of the turbulent transport in W7-X and a pressure drop in the edge, likely caused by the increasing impurity radiation and charge-exchange losses.
The first effect is a change in the relative importance of the electron and ion transport channel as Ti approaches Te. There are indications that in the standard ECRH plasmas of W7-X, turbulent ion-transport is stronger than the electron one (see M. Beurskens et al. at this conference) and, hence, a larger transfer of energy from the electrons to ions at higher densities leads to an increase of the overall losses. In W7-X this effect starts to play a role at densities between $3$ and $4\cdot10^{19}\,\mathrm{m}^{-3}$, coinciding with the first decrease of the density-scaling exponent seen in Fig. 1. The second effect is a pressure loss in the plasma edge, affecting the stored energy by effectively reducing the volume contributing to the total pressure. This pressure drop is visualized in Fig. 2 and includes the effect of the radiation losses. There are, however, indications that interactions with neutrals due to the increased fueling required for higher densities may play a role as well. Regarding fusion relevance, a pressure drop exclusively in the edge is less harmful than in the core. Yet, due to the larger volume in the edge it has a stronger impact on the stored energy. As can also be seen in Fig. 2, the core pressure is indeed not affected in the analyzed experiments and illustrates that profile effects also play a role in determining the density scaling of the stored energy. This has to be kept in mind for extrapolations to fusion plasmas and is currently not reflected in the simple scaling laws for the energy confinement time.
Since the radiation losses occur predominantly in the edge of the plasma, simply correcting the heating power, entering the scaling, by the radiation losses is not feasible and, hence, it is not easy to quantify how strong the impact of these losses on the energy confinement is. Hence, quantifying the relative importance of the aforementioned effects (radiation losses, charge-exchange losses and changes in transport behavior) is challenging and work in progress. It is clear, however, that simply excluding plasmas with high levels of radiation systematically removes effects from energy confinement time databases which are important when increasing the density towards fusion-relevant plasma conditions. Accordingly, the empirical scalings derived from such databases have to be revisited and new tools have to be developed to deal with this issue.Speaker: Golo Fuchert (Max-Planck-Institut für Plasmaphysik, Greifswald, Germany) -
Understanding the scrape-off layer (SOL) physics in current fusion devices is essential to solve the exhaust problem for future reactors. The SOL configuration determines the intersections between the magnetic field lines and the device targets, opening up parallel transport channels for plasma heat and particles, which represent the most direct way for the exhaust to reach the divertor targets$^1$. The interplay of parallel and perpendicular transport, along with the distribution of sources, sets the SOL dynamics for both particles and energy, and consequently defines the effectiveness of a particular divertor configuration. Quantifying the parallel particles flows can therefore give insights into the overall SOL behavior$^2$, allowing to assess the predominant flow directions of the main plasma ions and impurities, and to have a direct representation of convective heat transport. Despite the fact that ion convection is commonly considered negligible in the description of the total divertor heat deposition$^1$, recent results at DIII-D$^3$ point out how the standard model based on electron conduction dominance is not always valid. For the W7-X stellarator, a more significant role of convection is to be expected, as the characteristic SOL temperature gradients are weaker and the connection lengths longer than typically in tokamaks. A quantitative investigation of the parallel particle flows is hence confirmed to be a pre-requisite to the understanding of the general SOL transport properties, even for heat deposition.
In order to perform successful studies of the SOL parallel transport, it is necessary to rely on dedicated edge diagnostics. A powerful tool that has emerged recently is the Coherence Imaging Spectroscopy (CIS) diagnostic, a camera-based interferometric system capable of measuring Doppler particle flows associated with a selected visible emission line from the plasma$^{4, 5}$. In contrast to other flow diagnostics such as standard spectroscopy and Mach probes, CIS is distinguished by providing high time and flow velocity resolution, and high spatial coverage, all at the same time. Its 2D spatially resolved measurements led to the first direct detection of the complex 3D counter-streaming parallel flows structure in the SOL of W7-X$^6$ (see Fig.1). The W7-X application involves passive measurements of C$^{2+}$ impurities during hydrogen plasma experiments, with a view over the entire SOL. The C$^{2+}$ line-integrated measurements are restricted to the edge of the machine by the temperature dependence of the charged state photoemission coefficients, which peak in the range of 10-20 eV$^1$. This range coincides with the one necessary for hydrogen ionization. The vicinity to a source of H$^{+}$, combined with low temperature gradients from the radiation location to the divertor targets, leads to friction dominated scenarios$^7$ which imply that C$^{2+}$ impurities are a good proxy to examine the hydrogen ion parallel flow behavior in the SOL.
CIS measurements show a clear dependence on electron density variations. This is visible in Fig.2, in which C$^{2+}$ flow velocity and intensity (top) are plotted together with the peak heat load of one divertor module and the line integrated density (bottom). The plasma experiment can be divided into three domains: attached (before 6 s, Fig.1 (a)), transition to detachment (6-7.5 s, Fig.1 (b)), and detached (7.5-13 s, Fig.1 (c)). When the plasma is in the attached phase, $|v_{||}|$ increases with increasing density. This tendency reverses once the transition to detachment starts, as well as during the detached phase: $|v_{||}|$ drops even if the density keeps increasing, reaching values that are more than 4 times lower than in the attached case. The sensitivity of the velocity measured by CIS to density changes is also clear in the intervals 8.5-10.5 s and 11.5-13 s, which show variations not detectable in the peak heat load trace. The sharp decrease in $|v_{||}|$ is reliably correlated to the detachment transition and can therefore be used as a detachment signature. A velocity drop by at least a factor of 2 has been recorded for all the explored routes to detachment: density scans with main and divertor gas fueling systems (as shown in Fig.2 and Fig.3 respectively), and impurity seeding as in$^8$. The change in $|v_{||}|$ is always accompanied by an increase of the total C$^{2+}$ radiation (Fig.2-3 (top)), and a decrease of the heat flux to the divertor targets (Fig.2-3 (bottom)). The behavior shown here for a small amount of CIS lines of sight is shared by the rest of the measured data as well, confirming the tendency for the overall SOL.
These results suggest that particle parallel transport in the W7-X SOL robustly follows simple fluid dynamics, in which the ions stream towards the nearest target. With a direct measurement of the 2D distribution of a key SOL quantity (the ion fluid velocity), the CIS measurements allow a comprehensive test of this approximation. Although CIS has these powerful features, there are subtleties in the interpretation of the measurements, which require further work before a full quantitative understanding of the convective contribution to the divertor heat loads can be gained.
References
$^1$ P. C. Stangeby, The Plasma Boundary of magnetic fusion devices, IoP publisher (2000)
$^2$ B. LaBombard et al., Nuclear Fusion 44(10) 1047-1066 (2004)
$^3$ A. E. Jaervinen et al., Contributions to Plasma Physics, e201900111 (2019)
$^4$ J. Howard, J. of Physics B: Atomic, Molecular and Optical Physics 43, 144010 (2010)
$^5$ V. Perseo et al., Review of Scientific Instruments 91(1) 013501 (2020)
$^6$ V. Perseo et al., Nuclear Fusion 59(12) 124003 (2019)
$^7$ Y. Feng et al., Nuclear Fusion 49(9) 095002 (2009)
$^8$ F. Effenberg et al., Nuclear Fusion 59(10) 106020 (2019)Speaker: Mrs Valeria Perseo (Max-Planck-Institut für Plasmaphysik) -
Injection of sub-millimeter-sized powder grains of boron composites is performed in the Large Helical Device (LHD) showing wall conditioning effects in real time during long plasma discharges. This alternative boronization technique has been recently employed in both tokamaks (1-4) and stellarators (5), showing beneficial effects on plasma performance both through the modification of the edge plasma and through first wall conditioning.
The superconducting magnets of LHD allow performance of long plasma discharges (up to one hour), offering the opportunity to test this technique in steady-state operation (SSO) conditions, contributing to the assessment of its viability in future fusion reactors.
The Impurity Powder Dropper (IPD) (6), developed and built in PPPL, is mounted on the top of LHD. The selection of the upper port for the installation of the IPD was guided by predictive simulation of powder trajectories coupling the EMC3-EIRENE and DUSTT codes (7), evaluating the deflection of the injected powder grains by the plasma flow in the unique magnetic configuration of LHD. Here, a thick stochastic layer connects the helical confined region to four divertor legs, where the magnetic field is almost completely poloidal. The chosen position for powder injection relative to the LHD cross section in shown in Fig. [1]a.
First, dedicated short-pulse experiments are performed both before and after a standard glow discharge boronization using diborane gas (B$_2$H$_6$), with the aim of i) demonstrating effective impurity powder injection in the LHD geometry and ii) to characterize the response of the plasma to different injection rates in preparation of long pulse experiments. Controlled amounts of boron (B) and boron nitride (BN) powder were injected into hydrogen (H) plasmas, that were heated by 8MW NBI or $\leq3$ MW ECH, for up to 4 s.
Finally, a set of long pulse experiments was performed. Here, 10 s B powder bursts are injected into helium (He) and H plasmas with <800kW ECH heating of a duration between 37 s and 9 minutes.
Visible camera imaging (Fig. [1]c) and spectroscopy measurements of B lines (Fig. [1]b) confirm that the injected impurities effectively penetrate into the plasma, qualitatively confirming the predictive simulation results of Ref. (7). As a prompt effect of powder injection in short pulses, the density of the plasma is generally increased, together with the total radiated power and the stored energy. The electron temperature of the confined plasma is observed to increase for low density plasmas, while this effect is reversed for higher densities.
Control discharges featuring a density ramp-up with no powder injection were performed before and after the impurity powder injection experiments. An example of such discharges is shown in Fig. [2] (blue and orange). As an effect of cumulative B and BN injection, the oxygen and carbon level in LHD are shown to decrease by factors of 5 and 2 respectively. Consistently with a decrease of the impurity concentration, the total radiated power is also reduced after cumulative powder injection, indicating conditioning of the plasma facing components due to B injection. After the powder injection, the gas puff to obtain the same plasma density (up to intermediate densities) is increased, suggesting a decrease of recycling due to B wall conditioning. Similar results hold for the experiments performed at $R_{ax}=3.9$ m instead of the usual $R_{ax}=3.6$ m. For the experiments performed after a conventional boronization, no substantial reduction of the impurity level is observed when comparing control discharges (black and yellow in Fig. [2]), since the impurity level was already minimized by the previous B$_2$H$_6$ boronization. Still, an identical gas puff feeding results in, after cumulative B and BN injection, 18% lower plasma density, suggesting reduction of recycling at the strike points.
Leveraging the results of short pulse experiments, it was demonstrated how the IPD can be used in long discharges for both real time wall conditioning and plasma density control. In the top panel of Fig. [3] we consider a 40 s He discharge with constant gas fueling after t = 20 s, featuring a 10 s B powder injection (green shade). A few seconds after the usual increase in plasma density due to B injection (as observed in 4 s - short pulse experiments) the plasma density starts to decrease with respect to an identical discharge without powder injection (dashed lines), and correspondingly the plasma temperature is increased (since the input power is unvaried). This can be considered evidence of reduced recycling after powder injection. The variation of plasma density and temperature can be minimized when an active feedback control is used on the gas fueling (bottom panel of Fig. [3]). The decrease of recycling can still be observed from the increase of gas puff after powder injection with respect to the reference case, and on a shot-to-shot basis from the comparison of repeated control discharges with no powder injection alternating the ones whit B injection. Results from interpretative EMC3-EIRENE simulations of the LHD edge plasma will be presented.(1) R. Maingi et al., 2018 Nucl. Fusion 58 024003
(2) R. Lunsford et al., 2019 Nucl. Fusion 59 126034
(3) E. Gilson et al., in preparation
(4) A. Bortolon et al., 2019 Nucl. Mater. and Energy 19 384-389
(5) R. Lunsford et al., in preparation
(6) A. Nagy et al., 2018 Review of Scientific Instruments 89 (10):10K121
(7) M. Shoji et al., 2019 Contributions to Plasma PhysicsSpeaker: Federico Nespoli (Princeton Plasma Physics Laboratory) -
A notable improvement in the plasma performance has been observed after the injection of a series of frozen hydrogen pellets into ECH heated plasmas in the stellarator W7-X [1]. In these experiments, the pellet series rise the plasma density considerably and the central fuelling results in a peaked density profile. At the end of the pellet series, the ion and electron temperatures rise and almost equilibrate, and the plasma stored energy increases by more than 40\% reaching values above 1 MJ. This enhanced confinement phase lasts for several confinement times and terminates as the density and its peaking decay [2]. The stabilisation of this enhanced confinement will be of the utmost importance for exploring reactor-relevant scenarios in the device’s next operational phase with an actively cooled divertor.
A candidate to explain these observations is the reduction of the turbulent transport due to both, the stabilization by centrally peaked density profile of the ion temperature gradient (ITG) driven instabilities, and the specific stability properties of the electron-density-gradient driven trapped electron mode (TEM) in the W7-X magnetic geometry [3, 4]. This theoretical description has been to some extent supported by the drop in the line-integrated density fluctuation level measured with a phase contrast imaging (PCI) system during the enhanced confinement phase [5]. The present work reports, for the first time, radially-resolved measurements of density fluctuations and radial electric field by Doppler reflectometry (DR) that can provide new insights into the nature of this transient suppression of turbulence. The study encompasses several high-performance phases, observed under different heating power levels and magnetic configurations, along the 2018 W7-X experimental campaign. A comparison with both, neoclassical and gyrokinetic simulations is also presented in an attempt to test the ability of our most sophisticated tools to reproduce these observations.
Measurements are obtained with a DR system working in the 50-75 GHz frequency range in O-mode polarization. The reflectometer front end uses a single antenna and a set of mirrors for launching and receiving the signal at fixed probing beam angle of $\theta = 18\circ$ [6]. Under these conditions, perpendicular wave-numbers of the turbulence in the range $k_\perp \sim 7-10$ cm$^{-1}$ are measured at the accessible local densities in the range from 2.8 to 6.3 $10^{19}$ m$^{-3}$.
Profiles of radial electric field, $E_r$, and density fluctuation level have been measured in discharges with different pellet sequence reaching different plasma densities and heated with different ECH power levels, in two magnetic configurations, one with low ripple (called standard) and one with high iota. A pronounced $E_r$-well is measured with local $E_r$ values as high as $-40$ kV/m in the radial range $\rho \sim 0.7-0.8$ during the post-pellet enhanced confinement phase in both magnetic configurations. The maximum $E_r$ intensity at the $E_r$-well scales with both density and ECH power level following a similar trend as the plasma energy content. For comparison, in plasmas with similar ECH heating power and density but fuelled by gas puffing, the $E_r$ profile is rather flat with values close to $-15$ kV/m within the range $\rho \sim 0.5-0.9$. The density fluctuation level decreases from the plasma edge toward the plasma core in all plasma scenarios but the drop is more pronounced in the post-pellet enhanced confinement phase than in the puffing fuelled plasmas. In the post-pellet phase the fluctuation level decreases by about $15 - 20$ dB within the radial range from $\rho \sim 0.9$ to $\rho \sim 0.5$, while it decreases only by about 10 dB in the puffing fuelled plasmas. As an example, figure 1 shows the profiles measured in the standard magnetic configuration at two different ECH power levels and densities.
Despite the qualitative similarity of profiles in the two magnetic configurations, the fluctuation levels are found to be significantly lower in the high iota compared to the standard one. This dependence appears to be in agreement wit their linear stability properties [4]. Indeed, both configurations exhibit a reduction in the growth-rates of electrostatic instabilities for certain combinations of temperature and density scale-lengths, that are achieved during the post-pellet phase, but the predicted reduction is more pronounced in the high iota configuration.
Encouraged by these observations, neoclassical as well as gyrokinetic simulations are in progress to conduct a systematic comparison with the experimental results. First neoclassical simulations using DKES and KNOSOS [7] show changes in the $E_r$ profile comparable to the experimental ones. Regarding the density fluctuations, gyrokinetic simulations using EUTERPE [8] and stella [9] are being carried out to disentangle the effects of $E_r$, plasmas profiles, and magnetic configuration on the fluctuation level. The effect of plasma profiles and magnetic configuration on the linear stability properties is studied using both EUTERPE and stella. The specific effect of $E_r$ is studied with EUTERPE while non-linear simulations with stella will provide information on saturated fluctuation levels under different plasma conditions.[1] T. Klinger et al., Nuclear Fusion 59, 112004 (2019)
[2] S. A. Bozhenkov et al., Submitted
[3] J.H.E. Proll, et al., Phys. Rev. Lett. 108, 245002 (2012)
[4] J.A. Alcuson, et al., Plasma Phys. Control. Fusion 62, 035005 (2020)
[5] A. von Stechow et al., Density Turbulence Reduction by Equilibrium Profile Gradient Control in W7-X. 22nd Int. Stellarator / Heliotron Wksh. – ISHW (Madison, September 2019)
[6] T. Windisch et al., Proc. 14th Intl. Reflectometry Wksh. - IRW14 (Lausanne, May 2019) O.207
[7] J.L. Velasco, et al., submitted to J. Comp. Phys., arXiv:1908.11615 [physics.plasm-ph]
[8] E. Sánchez et al., Nuclear Fusion 59, 076029 (2019)
[9] M. Barnes, F. I. Parra, and M. Landreman, J. Comp. Phys. 391, 365-380 (2019)
Speaker: Teresa Estrada (CIEMAT) -
The stellarator W7-X has conducted, so far, experiments under both limiter and divertor conditions$^{1-2}$. While in the former case, smooth flux surfaces at the plasma edge, being free of low-order resonances, are cut by five local limiters, in the latter case, a lower-order island chain is intersected with ten sophisticated divertor units. Here, an interesting question emerges as how and to what extent the different plasma boundary conditions affect the impurity radiation and thereby the plasma performance, especially through radiation–driven thermal instability. They are the main topics addressed in this paper.
Figure 1(a) shows the data points collected from the quasi-stationary state of the limiter plasmas. They are all hydrogen plasmas generated with ECRH. Most of the stationary limiter plasmas have a radiation fraction $f_{rad}$ less than ~45%. At higher radiation fractions, the limiter plasmas are usually unstable and thermal instabilities develop$^{3}$. In this case, the stored plasma energy drops and the plasma shrinks, driven by an inwards movement of the radiation zone (see figure 2(c)). In contrast, most of the divertor plasmas can be stably operated at clearly higher radiation levels up to $f_{rad}$ ~ 0.9$^{4-6}$ without significant loss of the energy content. Examples ($P_{heat}$ ~3 MW) are shown in figure 1(right), gathered from both pre- and post-boronisation experiments.
In general, concerning impurity radiation, the island divertor has shown two beneficial effects in contrast with the limiter: 1) intensive radiation is located at the edge (r/a > 0.8) even at high radiation levels, and 2) the plasma remains stable up to a radiation fraction of ~90%.
Using two comparative examples, figure 2 shows the 2D radiation intensity distributions obtained by tomographic reconstructions based on the bolometer measurements. One is a limiter discharge (#20160309.7) and the other (#20171109.45) is a pre-boronisation divertor plasma to ensure that the comparison is made under similar wall-conditions. These reconstructions are performed by Gaussian Process Tomography method in the Minerva Bayesian modelling framework$^{7,8}$, which usually gives results similar to those of another algorithm, minimum Fisher information tomography$^{9,10}$. It clearly shows that a radiation zone, which peaks at r/a ~ 0.8 inside the LCFS (the red dashed line) of the limiter plasma, shifts outwards to the separatrix (see the magnetic islands with Poincare plot in black) in the diveror plasma. A common feature of all the discharges studied is that the radiated power loss fraction, defined as $f_{rad} =P_{rad}/P_{heat}$, increases with plasma density, while it decreases by increasing the heating power at a fixed plasma density.
Boronisation significantly reduces the carbon and oxygen yield in the machine$^{9}$. After wall-boronisation, the plasma densities required for accessing high-radiation regimes increase by a factor of ~3. The radiation distribution in the divertor plasma is less affected, and still localized at the plasma periphery. Poloidal asymmetry in radiation distribution is observed before and after boronisation. The radiation distribution becomes more up/down-asymmetric towards high-density, high-radiation, as shown in figure 2 (d). It is interesting to mention that this up/down asymmetry reverses under certain plasma conditions after boronization. The reason is unclear and needs to be investigated. It is often seen that a detachment transition is accompanied with an abrupt rise in radiation with respect to the density increment. At the same time, the radiated power from the SOL, $P_{rad,SOL}$, which can be estimated based on the line-integrals of bolometer channels purely viewing the SOL decreases. The portion inside the separatrix , i.e. $f_{rad,core} = 1- P_{rad,SOL}/P_{rad}$, jumps to a higher level, leading to a visible drop of the stored energy. An example is shown in figure 3, which shows to what extent this can happen. This sharp increment of the radiation level at a critical density was usually observed in limiter plasmas at detachment transition as well. However, after the transition, most of them are unstable and even collapse, accompanied with an inward movement of the radiation zone$^{3}$. By contrast, most of the divertor plasmas can withstand the abrupt transition, with the radiation layer stabilized around the separatrix.
As described, the island divertor concept on W7-X offers a large operating window up to high-density and high-radiation scenarios. Understanding the plasma radiation behavior and the impurity transport involved are essential for finding optimized operation windows of W7-X. This requires the knowledge of the background plasma parameters (of sufficient quality) and also a suitable tool, such as the EMC3-Eiren code$^{12,13}$, which is a 3D numerical tool capable of self-consistently treating the plasma, impurity and neutral transport in realistic 3D SOLs. Comparisons between the experimental results of bolometers with code simulations are ongoing.
References:
$^{1}$T. Klinger, et al Plasma Phys. Control. Fusion 59 014018 (2017)
$^{2}$T.S. Pedersen, et al Nucl. Fusion 55, 126001 (2018)
$^{3}$D. Zhang et al, to be submitted (2020)
$^{4}$D. Zhang, et al, Phys. Rev. Lett. 123, 025002(2019)
$^{5}$O. Schmitz, et al submitted to Nucl. Fusion (2020)
$^{6}$M. Jakubowski, et al IAEA-FEC 2020
$^{7}$J. Svensson, JET Internal report, Tech. Rep. (EFDA-JET-PR (11) 24, 2011).
$^{8}$J. Svensson and A. Werner, in 2007 IEEE International Symposium on Intelligent Signal Processing (IEEE, 2007)
$^{9}$M. Anton et al., Plasma Phys. Control. Fusion 38 (1996) 1849
$^{10}$D. Zhang, H. Thomson et al., 40th EPS Conf. on Plasma Physics,Espoo, Finland (2013)
$^{11}$T. Wauters, et al Plasma Phys. Control. Fusion 62 (2020)
$^{12}$Feng Y et al 1999 J. Nucl. Mater. 266-269 812
$^{13}$Reiter D et al 2005 Fusion Sci. Technol. 47 172Speaker: Dr Daihong Zhang (Max-Planck Institut fuer Plasmaphysik, D-17491 Greifswald, Germany) -
The perpendicular transport in the Scrape-Off Layer (SOL) of toroidal fusion devices is widely assumed to be dominated by turbulent transport. In tokamaks, a particularly important role in this regard is played by blob-filaments exhibiting coherent radial propagation. These filaments have been shown to significantly contribute to the perpendicular SOL transport and impact fundamental SOL properties such as the SOL width, plasma recycling at the first wall, and can furthermore cause high intermittent heat loads in plasma facing components [1,2]. In stellarators, the role of such turbulent filaments is less clear and is yet to be explored in the island divertor of the optimized stellarator Wendelstein 7-X. While turbulent transport is in general expected to be play an important role in the W7-X SOL due to the long connection lengths (several 100m), the particular characteristics are only beginning to be assessed. Both numerical simulations and experimental analysis are particularly challenging given the complex three-dimensional magnetic topology, non-uniform magnetic curvature, and three-dimensional plasma profiles and gradients.
Here, we employ reciprocating Langmuir probe arrays in conjunction with fluid model simulations to investigate the SOL turbulence in W7-X with a focus on turbulent filaments [3]. To avoid the additional complexity introduced by the magnetic islands, we first focus on experiments in the shadowed SOL region outside the islands, see Fig. 1a). The turbulent structures are in the experiment identified by conditional averaging of poloidally distributed electric probes measuring plasma density and potential fluctuations, yielding the well-known interchange mode picture of a dipolar potential distribution around a positive density perturbation which results in a radial ExB convection (Fig. 1 b/c). These structures are shown to be field aligned filaments due to their high parallel coherence along the magnetic field between reciprocating and target Langmuir probes. The delay between reciprocating and target probe in Fig. 1b) is consistent with the poloidal distance of the flux tubes (~cm) and poloidal propagation velocities (~km/s), while the parallel dynamics inside the filaments are very fast (on the electron thermal velocity time scale).
The experiments have been accompanied by drift plane simulations of seeded filaments, where the filament charge separation drive is due to the averaged magnetic curvature along the field lines at the measurement position of the reciprocating probe. Simulations and experiment agree remarkably, see Fig. 2. Compared to typical results in tokamaks, the radial velocitiy of filaments is about an order of magnitude smaller in W7-X. Larger filaments (20-30mm) show velocities as predicted by the fundamental sheath limited scaling [1,2] despite the long connection length, while smaller filaments are much slower than both sheath-limited and inertial scaling law. The modeling indicates that the slow radial velocity is due to the small curvature drive in W7-X which is a consequence of the large major radius and the particular magnetic geometry.
![Scaling of filament velocity vs size in two different connection length regions. The solid and dashed gray lines represent the fundamental filament scalings [1,2].](http://carsten-killer.de/fig2.jpg)
The slow radial velocity further implies that the turbulent filaments in W7-X do not travel significant radial distances within their life-time and therefore only cause local turbulent transport. In particular, a radial turbulence spreading as seen in tokamaks in not observed in W7-X, where the entire turbulence is local, i.e. the filaments are born and decay on the same flux surface. This conclusion that the turbulence observed in the W7-X SOL has a local character is further supported by the flat skewness and kurtosis profiles, i.e. by the absence of strongly peaked intermittent events.
These observations are an important first step to understand the complex SOL plasma parameter distribution in W7-X. A quantitative analysis of the experimentally determined turbulent transport is of course an obvious next step but requires additional modeling developments given that the measurements are 1-dimensional in a three-dimensional system. As a further next step, similar joint investigations of experiment and simplified simulations are being analyzed inside the magnetic islands, which adds further complexity to the analysis [4,5].
References
1 D’Ippolito et al., PoP 18 060501 (2011)
2 Carralero et al., NF 58 096015 (2018)
[3] Killer, Shanahan, et al., submitted to PPCF (2020)
[4] Killer et al., NF 59 086013 (2019)
[5] Zoletnik et al., PPCF 62 014017 (2020)Speaker: Carsten Killer (Max-Planck-Institute for Plasma Physics, Greifswald, Germany) -
Divertor detachment was successfully sustained using higher-Z (krypton, Kr) and lower-Z (neon, Ne) superimposed seeding. Plasma radiation could be enhanced at the upstream region in the edge plasma with suppression of impurity accumulation toward the core plasma. Here, the pre-seeded Kr emission was drastically enhanced after the subsequent Ne seeding. Formation of negative radial electric field, $E_r$, in the edge plasma due to the Ne seeding and inward diffusion due to a hollow electron density, $n_e$, profile should be a key for the enhancement of the Kr emission and the sustainment of the detachment.
Divertor detachment using impurity seeding is one of the effective operation scenarios to reduce divertor heat loads to lower than 10 MW/m$^2$ in ITER and fusion reactors. Steady-state sustainment is required for the detachment. Furthermore, to manage the power exhaust, radiation enhancement is required not only in the divertor region but also in the upstream region with suppression of dilution. Therefore, multi species impurity seeding is proposed in JT-60SA [A]. Moreover, it is predicted by the COREDIV code that additional Kr seeding is effective for divertor detachment [B]. In LHD, we have investigated detachment using Ne or Kr seeding individually [C]. Thus, in this study, we attempted to superimposed seeding using Kr and Ne in anticipation of experiments on tokamak devices. Since the cooling rate of Ne reaches a maximum at the electron temperature, $T_e$, ~ 30 eV while the rate of Kr has a local minimum at $T_e$ ~ 30 eV, it is expected that these impurities could enhance plasma radiation complementarily.
Fig. 1-3 show the plasma behavior using Kr+Ne, Ne, and Kr seeding, respectively. Using Kr+Ne superimposed seeding, the detachment could be sustained for ~ 1 s. On the other hand, using only Ne seeding, the detachment was short-lived. In the case of only Kr seeding, reduction of divertor heat flux, $q_{div}$, and enhancement of total plasma radiation, $P_{rad}$, was not significant.
As shown in Fig. 1, Kr was seeded 0.4 s before Ne seeding since the response of Kr is slower than Ne. After the Ne seeding, the plasma was detached and the detachment was successfully sustained ~ 1 s. This detachment was terminated by the limitation of the pulse length of the NBI. The $q_{div}$ decreased by ~ 85% of $q_{div}$ before Kr seeding. The radiation fraction ($f_{rad} = P_{rad}/P_{NBI, abs}$) was ~ 40% which is 10% higher than the case with only Kr seeding and previous sustainment using multi-pulse Ne seeding [C]. Here, $P_{rad}$ and $P_{NBI, abs}$ are total radiation power and absorbed NBI power, respectively. Line-averaged electron density, $n_{e, bar}$, increased due to the ionization of impurities and change of wall recycling. Although NBI port-through power is constant, $P_{NBI, abs}$ increased with increasing $n_{e, bar}$ due to the increase of heating efficiency. $\tau_{E, exp}/\tau_{E, ISS04}$ is ~ 0.8 which is 10% higher than the case with only Ne seeding. Here, $\tau_{E, exp}$ and $\tau_{E, ISS04}$ are energy confinement time evaluated by the experiment and ISS04 scaling, respectively. Kr emission remains very low until Ne seeding, which then triggers KrXIX to increase together with NeVIII.
The direction of $E_r$ plays an important role in this detachment. Using only Ne seeding, positive $E_r$ was formed in the edge plasma after Ne injection as shown in Fig. 2. The positive $E_r$ can exhaust the seeded Ne. Radiation profile and $T_e$ profile indicate that NeVIII, which has the ionization energy of 239 eV, or lower charge states are the dominant radiator. After the Ne seeding, edge $T_e$ decreased due to the ionization of the seeded Ne. Then, $P_{rad}$ in the edge region was enhanced and the plasma was detached. Since the decreased edge $T_e$ recovered with increasing the heating efficiency of NBI, the edge $T_e$ increased. Due to the minor-radially outward shift of the radiation region, the radiative cooling was weakened. Finally, the plasma using only Ne seeding was reattached. Owing to the positive $E_r$ and increase of the edge $T_e$, the detachment using only Ne seeding was short-lived. In the case of Kr seeding, $E_r$ gradually increased after the Kr seeding and positive $E_r$ was formed as shown in Fig. 3. Moreover, since higher-Z impurities have a lower first ionization energy, Kr is ionized in the more downstream region in the ergodic layer where the friction force is dominant compared with Ne. Therefore, Kr could not penetrate deeply and $P_{rad}$ enhancement was quite small. The moderate $q_{div}$ decrease clearly indicated that Kr only seeding was not effective. On the other hand, in the case of Kr+Ne seeding, negative $E_r$ was formed at the edge region after 4.9 s and a clear decrease in $T_e$ at the plasma edge region was observed due to Ne seeding as shown in Fig. 1. Therefore, Kr could penetrate deeper than the case with only Kr seeding. Finally, radiation was successfully enhanced in the upstream region in the edge plasma compared with the same $f_{rad}$ using only Ne seeding with suppression of impurity accumulation toward the central plasma. Here, $E_r$ increased after 5.4 s. The reason for the increase of $E_r$ can be considered that the measurement position with a Doppler backscattering system was affected by a termination of deuterium gaspuff at 5.3 s.
The direction of diffusion due to a hollow $n_e$ profile can contribute the sustainment of the detachment using Kr+Ne seeding. The $T_e$ at the local maximum of $n_e (r_{eff}/a_{99} ~ 0.94)$ was 400 eV after the Ne seeding as shown in Fig. 1. The ionization energies of NeVIII and KrXIX are 239 eV and 785 eV, respectively. Therefore, NeVIII emits outside the position of the $n_e$ local maximum and KrXIX emits inside the position of the $n_e$ local maximum. Here, the direction of the diffusion due to $\nabla n_e$ is outward for NeVIII and inward for KrXIX. The inward diffusion can sustain the detachment by holding the Kr slightly inside the LCFS in the case of Kr+Ne seeding while the direction should be outward in the case of only Kr seeding due to inefficient reduction of the edge $T_e$. The outward diffusion for Ne can exhaust the Ne. It can support the termination of the detachment in Ne only seeded plasmas. Multi-pulse seeding of Ne is a candidate to sustain the detachment, but it is not easy since the reattached plasma shown in Fig. 2 is different from the background plasma, e.g., in $n_{e, bar}$. Density control using further pumping may be helpful. Multiple pulses using moderate Ne seeding could sustain the detachment with $f_{rad}$ ~ 30% 3.[A] K. Gałązka et al., Contrib. Plasma Phys. 58 (2018) 751.
[B] R. Zagórski et al., Nucl. Fusion 57 (2017) 066035.
[C] K. Mukai et al., Nucl. Fusion 55 (2015) 083016.
Speaker: Dr Kiyofumi Mukai (National Institute for Fusion Science) -
Energy confinement and energetic-particle (EP) driven MHD modes have been experimentally studied from the viewpoint of magnetic configuration effect in the Heliotron J helical device. Experiments scanning the bumpiness component in magnetic Fourier spectra show that (i) maximal kinetic energy in electrons is achieved when the medium bumpiness is chosen where the neoclassical (NC) diffusion is reduced, while it is degraded when the bumpiness component is too high or too low, and (ii) the excitation and damping of energetic-particle (EP) driven modes depend on the magnetic configuration, and they are mitigated with increasing on-axis ECH power at low bumpiness configuration. These results suggest that the magnetic configuration has an important role in global energy confinement and EP driven modes in helical plasmas.
It is an issue to optimize magnetic configuration for reduction of NC and turbulent transport, good energetic particle confinement and MHD stability in helical devices. The Heliotron J device has the advantage of flexible control of magnetic configuration by combination of a helical coil, toroidal coils and vertical coils. The bumpiness, which is the toroidal mirror ratio defined by εb = B04/B00, is a key component to control plasma performance in quasi-omnigeneous optimization. Here B04 and B00 are the toroidal mirror and uniform magnetic fields of the Fourier components in Boozer coordinates, respectively. The previous experiments showed a favourable energy confinement was obtained in a high εb configuration in NBI-only plasmas. A nonlinear gyrokinetic simulation qualitatively reproduces reduction in thermal transport coefficients with increasing the bumpiness, which is related to reduction in the ITG mode [A]. EP-driven MHD modes can also be controlled by magnetic field configuration such as rotational transform and its magnetic shear [B].
We have recently extended the experimental study on the dependence of energy confinement on the magnetic configuration in ECH-only plasmas where electron heating is dominant. Four configurations are chosen, the ultrahigh-, high-, medium-, and low-bumpiness configurations (εb = 0.22, 0.15, 0.06 and 0.01 at r/a = 2/3, respectively) with fixed toroidicity, helicity, rotational transform and plasma volume. Plasmas are produced and heated by 2nd X-mode 70-GHz ECH of about 0.25 MW and the ne and Te profiles are measured with a Nd:YAG laser Thomson scattering diagnostic. Figure 1 shows the electron kinetic energy as a function of electron density. The electron kinetic energy is maximal at high and medium e_b configurations, while it is low when the bumpiness is too high and too low, suggesting that optimization of NC transport is compatible with anomalous transport although anomalous transport plays a dominant role rather than NC transport. Figure 2 shows the ne and Te profiles in the three configurations. The central Te is almost the same, while the edge Te is low at the low εb configuration. The ne profile is hollow with the peak position more inside, and the edge ne is lower at the low εb configuration. This means that the plasma profile shape particularly at the edge region affects the global energy confinement. The difference in plasma profiles may be connected to the change in turbulent transport similarly as observed in the NBI plasmas. The bumpiness could affect the growth rate of TEM or ETG modes particularly at the edge region.
The magnetic configuration also affects EP-driven MHD modes such as global Alfvén eigenmodes (GAEs) and energetic particle modes (EPMs). Figure 3 shows the amplitudes of some observed modes for three configurations. The mode structure of GAEs in the medium εb configuration has been reproduced by a hybrid simulation code, MEGA. For the low εb configuration, the modes are mitigated with an increase in ECH power. On the other hand, in the high εb configuration, the GAE at 90 GHz is monotonically suppressed as the EC power increases, and the EPM at 100 kHz is destabilized. In the medium εb configuration, the GAE at 100 kHz is suppressed, and the EPM at 120 kHz is not a monotonic function. These results indicate the mode stabilization is related to the change in shear Alfvén spectra and population of trapped electrons which depend on the bumpiness. The non-monotonic behavior of the EPM may also be related to the balance between stabilization effect by continuum damping and destabilization effect by energetic particle pressure.
[A] A. Ishizawa, et al., Nucl. Fusion 57 (2017) 066010
[B] S. Yamamoto, et al., Nucl. Fusion 57 (2017) 126065Speaker: Prof. Kazunobu Nagasaki (Institute of Advanced Energy, Kyoto University) -
Visible magnetic-dipole (M1) lines can serve as a novel diagnostic mean of tungsten in ITER and future DEMO reactors. Here we report that a local tungsten density in core plasmas of the Large Helical Device (LHD) is successfully assessed with the measurement of visible M1 lines emitted by W$^{26+}$ and W$^{27+}$. By such a measurement, (i) the radial profile of total tungsten density in the LHD core plasmas with a line-averaged electron density of $\sim$ 4 $\times$ 10$^{19}$ m$^{-3}$ and central electron temperature of $\sim$ 1 keV is found to be a hollow, and (ii) the total tungsten density is found to be decreased in the whole region with an increase in the electron temperature due to the onset of reheat mode. The scheme with visible M1 lines used in this work can be also applied for the tungsten density measurement at the ITER edge plasma.
Tungsten transport and its density control in plasmas are key issues in order to maintain high-performance plasmas in ITER with a tungsten divertor, because it is well known that a tungsten has a strong radiation cooling power. Electron temperature in ITER can range from 0.1 keV at the edge up to about 20 keV at the plasma center. In such a plasma, tungsten can be highly ionized ($q$ = 20 - 60), and can emit strong line emissions in extreme-ultraviolet (EUV) and soft X-ray regions. Until now, only a few direct measurements of tungsten ion density with spectroscopic diagnostics have been reported in the EUV range [A]. Recently, we observed visible lines emitted from highly charged tungsten ions due to magnetic-dipole (M1) transitions in the LHD core plasmas after a tungsten pellet injection [B]. Such visible M1 lines are useful particularly for fusion plasmas, because mirrors and optical fibers can be used for avoiding a direct neutron radiation on detectors. Currently, the W$^{26+}$ ion density can be also assessed by using the measurement of two M1 lines (335.7 nm and 333.7 nm), in addition to the W$^{27+}$ ion density assessed by the measurement of a M1 line (337.7 nm) [B], and then a total tungsten density in the LHD core plasmas can be assessed based on such ion densities.
Figure 1 shows time traces for the discharge of interest. In this discharge, a polyethylene tube containing a tungsten wire (0.6 mm long and 0.15 mm diameter, i.e., the number of tungsten particle is about 6.8 $\times$ 10$^{17}$ /pellet) is injected at the time of 4.03 s. After the tungsten injection, the central electron temperature is lowered down to 1 keV due to the strong radiation by the tungsten ions. Then, the central electron temperature remains at an almost constant value until $t$ = 5.4 s, while the line-averaged electron density gradually increases due to a hydrogen gas puff. About 0.1 s after the hydrogen gas puff is stopped at $t$ = 5.3 s, the electron temperature is increased, meanwhile the line-averaged electron density is decreased. Such changes in electron temperature and density can be considered due to the onset of a reheat mode.
Figure 2 (a) shows the radial profile of ion densities of W$^{26+}$ and W$^{27+}$ at $t$ = 5.0 s, which are obtained by dividing measured local emissivity of M1 lines by theoretical photon emission coefficients (PECs) of these M1 lines. Photons emitted along 44 line-of-sights (LOSs) on a horizontally elongated poloidal cross section of the LHD plasma are guided to a UV-Visible spectrometer by an optical fiber array. The local emissivity is inferred by an Abel inversion of vertical profiles of line-integrated intensities at the 44 LOSs. The PECs are calculated using a collisional-radiative (CR) model in which proton collision effects are also considered. Density dependence of the PECs is significant at high electron densities in the core, because collisional excitation and de-excitation among excited levels become important. Such density dependence, omitted in the corona model which is valid at low densities, can be correctly calculated with the CR model. The calculated PECs at $t$ = 5.0 s in Fig. 2 (d) are almost constant in the core due to a flat electron density profile. The ion abundance ratios of W$^{26+}$ and W$^{27+}$ (see Fig. 2 (b)) agree with theoretical values of the ionization equilibrium model where CR ionization and recombination rate coefficients are used. In the ionization equilibrium, fractional ion abundance is determined by rates of ionization and recombination of tungsten at the local temperature and density, regardless to transport. Thus, the agreement with the ionization equilibrium model indicates that tungsten transport time in the core is so long that the fractional ion abundance instantaneously equilibrate at local temperature and density via the fast ionization and recombination processes.
Based on this result, a total tungsten density is evaluated from the W$^{27+}$ ion density divided by the theoretical fractional ion abundance of the ionization equilibrium model. It should be noted that W$^{24+}$ - W$^{28+}$ have a dominant abundance in the plasma of interest. Figure 3 shows time evolution of the total tungsten density from $t$ = 4.8 s to $t$ = 5.6 s. Until $t$ = 5.4 s, the electron temperature and its profile change little. In this phase, the total tungsten densities gradually decrease with time, however the densities at $\rho$ = 0.3 do not change much leading to a hollow profile of the radial distribution at $t$ = 5.4 s. On the other hand, after the onset of reheat mode (from $t$ = 5.4 s), in which the core electron temperature increases significantly, a rapid reduction of the total tungsten density is observed. The rapid reduction of the total tungsten density in the whole region during the reheat mode can be considered as the enhancement of the radial transport of tungsten.
[A] Y. Liu et al, Jpn. J. Appl. Phys. 57 (2018) 106101
[B] D. Kato et al., 26th IAEA FEC (17-22 Oct., 2016, Kyoto, Japan), EX/P8-14
Speaker: Dr Daiji Kato (National Institute for Fusion Science) -
Turbulence properties against the variation of isotope ratio and zonal flow activity are elucidated in Heliotron J. The turbulence amplitudes for density and potential fluctuations reduce as the hydrogen/deuterium(H/D) gas ratio is varied from H to D dominant plasmas and zonal flow activity is enhanced. Two-point correlation analysis reveals that the correlation of the fluctuations decreases in D plasmas, although the turbulence scale size increases as D gas fraction increases. A statistical analysis using a joint probability density function technique also indicates that the density and potential fluctuations are decoupled in D plasmas, which should contribute to the suppression of turbulence-driven transport and the confinement improvement in D plasma. These observations suggest that the isotope effect can emerge through the reduction and decoupling of density/potential fluctuations, which is attributed to the enhanced zonal flow activity in D plasmas observed in the experiment.
Confinement improvement in deuterium plasmas, called “isotope effect”, has been a long-standing issue in the study of magnetic confinement fusion. The isotope effect contradicts a fundamental model of transport, because an increase of characteristic scale (ion Larmor radius or turbulence scale size here) simply gives the increase of transport, in other words, D plasmas should have a degraded performance compared with the H plasmas, incompatible with the experimental observations. A hypothesis is proposed to explain the isotope effect recently, which is attributed to isotope dependence of turbulence system including a zonal flow activity. A couple of experimental works, including our past work, also report the dependence of zonal flow activity exists on the H/D isotope ratio. However, turbulence responses behind the isotope dependence of zonal flow have not yet been studied in detail so far.
In this study, the isotope dependence of local turbulence properties is characterized in a helical device, Heliotron J. The machine has major/minor radii of $R/a =1.2/0.17 \mathrm{m}$ with the magnetic field strength $\mathrm{B} = 1.25 \mathrm{T}$ on axis. In this experiment, the plasma was sustained with electron cyclotron heating with the power of $< 0.3 \mathrm{MW}$, and the H/D ratio was carefully controlled to keep line-averaged density constant ($~0.2 \times 10^{19} \mathrm{m}^{-3}$) and to reproduce the same plasma conditions. Two Langmuir probes, located at different toroidal sections, were used by fixing them at the same flux surface of $\rho\sim 0.8$ to measure the local turbulence and zonal flow.
The frequency spectra of floating potential and ion saturation current indicate that the turbulence level gradually increases against H/D ratio, as hydrogen is more dominated and zonal activity is enhanced as shown in Fig. 1(a) and (b). Interestingly, the small difference in the spectra can be seen; higher frequency components of $> 100 \mathrm{kHz}$ emerge in the case of floating potential, and fluctuation level in all the frequency range increases in ion saturation current, as hydrogen becomes dominant. This observation suggests that turbulence-induced transport increases in H plasmas.
Dependence of turbulence correlation length on the H/D ratio was also evaluated with a two-point cross-correlations technique, by using a pair of adjacent probe tips with a distance of $\sim 5 \mathrm{mm}$. The cross-correlations for density and potential are plotted against the H/D ratio in Fig. 1(c) and (d), respectively. In this analysis, the correlation is evaluated with the dominant fluctuation in the frequency range from $10$ to $40 \mathrm{kHz}$. This is because the fluctuation component less than $> 50 \mathrm{kHz}$ is dominant in edge plasmas, and noises such as cross-talk in the frequency range $> 50 \mathrm{kHz}$ are not large but non-negligible for the correlation analysis. Fig.1(c) shows that no clear dependence against HD ratio exists for potential fluctuation, and however significant dependence is found for density fluctuation (ion saturation current). The correlation for density fluctuation decreases as hydrogen gas becomes dominant, which implies that the scale size of density fluctuation is smaller in H but larger in D plasmas. This observation is qualitatively consistent with the past experimental results, however, it is unfavourable for transport. The correlation between density and potential fluctuations, which is essentially important to determine turbulence-induced transport, is then characterized, as shown in Fig. 1(e). The correlation decreases as the D gas is dominated in Fig. 1(e). This decrement suggests that the potential and density fluctuation are more decoupled in D plasmas, and the decoupling contributes to the reduction of turbulence-induced transport, even if the turbulence scale size increases in D plasma as mentioned in the previous paragraph.
Furthermore, the decorrelation of density and potential is also demonstrated from a statistical viewpoint using a joint probability density function(joint-PDF) technique, as shown in Fig. 2(a-c). The joint-PDF is an extension of one-dimensional PDF, and the distribution shows each PDF of each quantity and indicates a degree of correlation between two different quantities. If two quantities are strongly correlated with a linear relation, the distribution has a distorted, linear shape. A more randomized, rounded distribution shape indicates a weaker coupling, and the distribution should be symmetric in X-Y axes in the case of completely random variables. The joint-PDF for potential and density fluctuations can be seen to be a more elliptic shape in the H plasma, compared with the case in D plasmas, which shows that the fluctuations are correlated more in H plasmas while are decoupled in D plasmas. The difference between the two cases is also shown in Fig. 2(c), representing that the correlated components with an asymmetric shape reduce at the higher fluctuation level（$>$ the standard deviations $\sigma_{is}$ and $\sigma_{Vf}$), while the uncorrelated components increase at the lower fluctuation level ( < $\sigma_{is}$ and $\sigma_{Vf}$) in D plasmas. The reduction of the correlated component corresponds to the decoupling between density and potential fluctuations and the distortion of the fluctuation PDF, which could reduce particle/heat fluxes in D plasmas.
Speaker: Shinsuke OHSHIMA (Kyoto University) -
Over the past few years, the need to develop reliable external actuators to control Alfvén Eigenmodes (AEs) in burning plasmas has triggered a substantial amount of experimental activity in several tokamaks and stellarators [1]. This contribution summarizes the experiments related to the AEs control topic that have been carried out in the TJ-II stellarator, paying special attention to the latest experimental developments and the recent comparison with theoretical predictions. In this context, considerable effort has been devoted to investigate the impact of ECRH and EC driven current (ECCD) on NBI driven AEs observed in TJ-II plasmas [2, 3, 4]. Other actuators, as NBI injection energy, pellet injection, plasma density, magnetic configuration scans [5] or ohmic current induction have also been considered but the clearest impact on the shear Alfvén waves spectrum, while at the same time minimizing the induced changes on the main plasma parameters, has been obtained in ECCD experiments [6]. While moderate ECCD acts mainly on the rotational transform of the device barely changing the plasma parameters (ECRH power deposition profile for oblique injection is very similar to the perpendicular one), ECRH power itself strongly modifies the temperature and the density of the background plasma. Recent examples illustrating both situations are presented in figures 1 and 2.
In the first case, changes in activity are mainly related to changes in the spectrum of modes issued from the modified magnetic configuration. In the second case, the EC power induces density and temperature variations that have a direct impact on the energetic particles drive and the damping rates of the excited modes and, to a lesser extent, on the spectrum of modes. In Figure 1, we may observe the effect of a small amount of EC driven current when comparing the magnetic fluctuations spectrograms measured in two NBI+ECRH shots for which the ECRH power is being injected with and without associated ECCD.
The changes in plasma current trigger the destabilization of a very clear frequency-locked mode (see figure 1(b)). Despite of the uncertainties related to the precise determination of the iota profile (that was estimated using theoretical calculations of the induced currents, i.e. NBCD, ECCD and bootstrap current contributions), the simulations using the FAR3D code predict the destabilization of several GAEs and HAEs whose frequencies match the ones observed experimentally [7].
The effect of ECRH power is better studied by adding extra power in AEs scenarios that have already been achieved using a combination of NBI and ECRH, thus avoiding the strong impact that ECRH power has when applied to a pure NBI heated plasma. For instance, adding further power in the ECCD scenario described above results in temperature variations that only produce an increase of the mode frequency without modifying its amplitude (see figure 1(c)). On the other hand, a similar experiment, but using a chirping mode scenario, demonstrated a clear mode amplitude mitigation, again accompanied by an increase of the mode frequency. This effect was previously documented in [2].
Figure 2 shows the result of scanning the ECRH power deposition location. Two ECRH beams at full power (keeping the first one on–axis while modifying the direction of the other one on a shot to shot basis) were used as target plasma for co-NBI heating. No pure NBI heated plasma is used here. In the NBI phase, the on-axis beam was shut-down, thus leaving the off-axis beam. Five different off-axis position were employed in the launching direction scan while keeping a more or less constant density during the NBI phase ($\overline{n}_e\approx 0.6\times10^{19}$ cm$^{-3}$) as the temperature decreased. The main mode shows a more or less pronounced chirping character whose behaviour could be tied to the modifications of the plasma profiles. However, slight changes in plasma current should also be considered to understand the changing mode pattern. In these experiments, provided no ECCD is being generated when moving the beams off-axis, only NBCD and bootstrap current, both of them affected by the changes in the temperature profiles,are at play. On the other hand, off axis heating modifies the fraction of ECRH power that goes to trapped or circulating electrons and this could have an influence on the mode pattern, particularly on the chirping mode structure since this later depends on the equilibrium between the impact of electron dynamical friction on beam ions and their diffusion in velocity space.
Independently of the more or less correct understanding of the theoretical reasons behind the observed behaviours, the experiments show that ECRH and, in particular ECCD, are efficient tools to modify AEs activity. Whether or not an effective control of AEs, which means developing capabilities enabling us to switch between stable and unstable mode conditions, can be achieved, is still under investigation.
References
[1] M. Garcia-Munoz et al., Plasma Phys. Control. Fusion 61, 054007 (2019)
[2] K. Nagaoka et al., Nucl. Fusion 53, 072004 (2013)
[3] Á. Cappa et al., 25th IAEA FEC (2014)
[4] E. Ascasíbar et al., 20th ISHW (2015)
[5] A. Melnikov at al., Nucl. Fusion 56, 076001 (2016)
[6] Á. Cappa et al, 45th EPS (2018)
[7] Á. Cappa et al, 16th EPPI (2019)Speaker: Álvaro Cappa (Laboratorio Nacional de Fusión CIEMAT) -
New experimental evidence based on spectroscopy, see Fig. 1, and other edge diagnostics indicate the existence of a high-recycling regime in W7-X. The high-recycling regime has not been accessible in the predecessor W7-AS and was predicted by modeling [1] for W7-X. The island divertor stellarator W7-X has also demonstrated stable, detached, high density divertor plasmas with good particle and power exhaust conditions, such that steady-state operation with particle removal by the pumps and sufficient power dissipation have been achieved [2,3,4,5]. Experimental measurements show also high divertor densities, complete heat flux detachment from the target, high neutral compression and beneficial impurity transport with indications of divertor retention of low-Z impurities [6] and benign core impurity transport [7]. This is a very important finding with respect to future power plants as the ability to exhaust the fusion power compatible with the limits of the plasma facing components is a paramount requirement for such a device. Therefore, the W7-X program develops and investigates exhaust scenario based on a detached divertor plasma.\
Two approaches of achieving heat flux detached divertor targets have been followed in the OP1.2b campaign. First, density ramp experiments using a feedback controlled gas puff system [2,3]. Second, impurity seeding of N and Ne were used to increase the radiated power and achieve detachment at lower plasma densities [5]. We focus here mainly on density ramp experiments.
\pagebreak
Fig. 2 shows the timetrace of a density ramp discharge that leads to completely heat flux detached targets after 3.0s. Consistent with the modeling expectations [8], above a threshold of $f_{rad} > 50\%$, the radiation increases strongly with a small density increment and the radiation front detaches from the target and moves toward the X-point. The movement of the radiation front is indicated by the movement of the N$^+$-line (399.5 nm) radiation and the high density region in the divertor shown in Fig. 3.b and 3.a, respectively.Currently, the exact localization of the radiation front with bolometry is not conclusively possible with respect to its position around and along the separatrix, but upstream profile diagnostics show a significant reduction of the electron temperature from about 150 eV to 50 eV around the separatrix up to 3.4 cm inside the nominal separatrix position. Keeping in mind the large uncertainties in the latter, this is still indicative of increased power losses in the very edge of the confined region and could be comparable to the very localized radiation features in X-point radiation regimes in tokamaks [9]. As N$^{+}$ pre-dominantly radiates at temperatures of $T_e = 3 \rm{\ eV}$, its location can be taken as a reference for the ionization front. This implies that the ionization front lifts off the target concomittent with the radiation front during the detachment transition. Fig. 2.c shows that at the same time a drop in neutral compression(black line) by about 50 % with only marginally reduced divertor neutral pressures is observed in experiment. Stark broadening of Balmer lines is used to assess the electron density in the divertor, see Fig. 1. High densities on the order of $10^{20} m^{-3}$ with $n_{e,div} > n_{e,sep}$ indicate the existance of a high recycling regime in W7-X experiments. Line ratio analysis of the Balmer lines ($\textrm{H}_\epsilon\ \&\ \textrm{H}_\delta$) shows that the divertor plasma in W7-X detachment conditions is still ionization dominated. The absence of strong recombination and the low neutral pressures of order 0.1 Pa in the experiments thus imply that the perpendicular transport is an important if not the dominant pressure loss channel in W7-X.
As a consequence, the dominant detachment physics in W7-X seems to be different to the traditional picture in tokamaks: In tokamaks large pressure losses and particle flux reduction across wide parts of the target are not easily obtained [9,10]. Pressure conservation naturally leads to the onset of a high recycling regime with lowering divertor temperatures. The high divertor densities then enable large power losses from impurity radiation. In contrast, the long connection length of island divertor stellarators, such as W7-AS and W7-X, provides large pressure losses from the main particle flow channel via perpendicular transport. The pressure losses can actually suppress the high recycling regime altogether, as observed in W7-AS [9], or limit the achievable divertor densities. In combination with the increased importance of perpendicular energy transport in the Scrape-Off Layer this prohibits large dissipated power fractions in the divertor by radiation [8]. As a consequence, a particular focus of tokamak detachment studies is the pressure loss, whereas in W7-X to characterize the available Scrape-Off Layer and divertor power losses and to achieve high divertor densities have a higher priority. The latter will be addressed in this contribution. The experimental data will be used to validate EMC3-Eirene modeling and the physics processes that lead to the radiation movement, the high recycling regime and the neutral pressure build up in the divertor will be analyzed.[1] Y. Feng, Nucl. Fus. 46 (2006)
[2] M. Jakubowski, Phys. Rev. Let. (submitted)
[3] O. Schmitz, Nucl. Fus. (submitted)
[4] D. Zhang, Phys. Rev. Let. (accepted)
[5] F. Effenberg, Nucl. Fus. 59 (2019)
[6] V. Winters, PSI 2020
[7] A. Langenberg, NF (2018)
[8] Y. Feng, Nucl. Fus. 44 (2002)
[9] F. Reimold, Nucl. Fus. 55 (2015)
[10] A. Kukushkin, Contrib. Plasma Phys. 56 (2016)
[11] Y. Feng, Plasma Phys. Control. Fusion 53 (2011)Speaker: Dr Felix Reimold (Max-Planck-Institut für Plasmaphysik) -
Both mean and oscillatory components of radial electric field Er play important roles in the formation of the turbulent and transport processes in toroidal devices, tokamaks and stellarators. Previous experiments [1, 2] and gyrokinetic modeling [3] have shown that the electrostatic potential φ (and Er) may vary along the flux surfaces that may affect the transport. In addition to the collisional neoclassical mechanisms, electrostatic turbulence may affect strongly the radial transport in some regions of stellarator plasmas. On top of that, the shear of ExB flow may suppress the turbulence and improve the plasma confinement. Therefore, measurements of 2D distributions of mean values and fluctuations for potential φ and density ne may shed a new light to the physical mechanisms of radial transport.
Four basic plasma scenarios were explored with heavy ion beam probe diagnostics in TJ-II. They are (i) low line-averaged density ne=0.45x1019 m-3 plasmas with higher PEC=470 kW (Te(0) = 1.6 keV) and (ii) lower PEC=250 kW (Te(0) = 1.4 keV) on-axis ECR-power, (iii) higher density ne=0.8x1019 m-3 plasmas with lower PEC=220 kW (Te(0) = 1.2 keV) (iv) and high-density ne=1.0-1.2x1019 m-3 co-NBI plasma with PNBI=550 kW (Te(0) = 0.6 keV).
Profiles of potential and density at for scenario (i) and (iii) are shown in Fig. 1. We see that growth of density is accompanied by evolution of the potential shape from conical to “Mexican hat”, while the density profiles are flat or slightly hollow. The potential has the positive peak at the ECRH power deposition area and LFS-HFS symmetry. HIBP and Langmiur probe (LP) data (lines and points at Fig. 1(a)) are consistent at the edge, 0.85<ρ<1. For scenario (iv) with high density potential profile is fully negative with a minimum value up to - 400 V at the plasma centre.
The poloidal 2D map (Fig. 2a, scenario (i)) shows that the local maximum of potential (up to 1 kV) coincides with the plasma center, and potential contours are consistent with the vacuum magnetic flux surfaces demonstrating the poloidally symmetric structure. Such poloidal symmetry holds for all four explored scenarios.
The contour plot for RMS density fluctuations with (f<300 kHz, k< 3 cm-1) presents a more complicated shape with poloidal anti-ballooning asymmetry, Fig. 2(b). Such poloidal asymmetry holds for both plasma potential and density perturbation for all four explored scenarios. Asymmetry is stronger for low-densities: at the mid-radius, where the density has maximum, RMS φ ~ 15 V at LFS vs ~ 20 V at HFS; RMS δne/ne ~ 2% at LFS vs ~ 3% at HFS, as presented in Fig. 2b, scenario (i). For scenario (i) the level of the potential and density perturbations was about 1.5-2 times higher than for scenario (iii). The lowest level of fluctuation RMS takes place for NBI plasma, scenario (iv).
Plasma density turbulence (f < 150 kHz, k < 3 cm-1) tends to rotate towards ExB drift velocity in ion diamagnetic drift (iDD) direction; Vturb = VExB = 10-15 km/s in low-density plasma in scenario (i). For scenario (ii) the feature is similar with the reduced velocity value. For scenario (iii) with higher densities and Mexican hat potential profile both Vturb and VExB, rotate to iDD in the core and to electron diamagnetic drift (eDD) direction at the edge, but Vturb exceeds VExB. For scenario (iv) with high density NBI plasma and negative monotonic potential profile both Vturb and VExB, rotate to eDD direction at the whole radial interval.
Obtained results show that TJ-II, equipped by the unique set of the diagnostics became a unique platform for validation of theoretical models and gyrokinetic simulations [3], predicting the poloidal symmetry breaking for plasma potential and density turbulence on magnetic flux surfaces in 3D devices.References
1. Sharma R. et al. 45th EPS Conf. on Plasma Physics (2018) P5.1061.
2. Estrada T. et al. Nucl. Fusion 59 (2019) 076021.
3. Sánchez, E. et al. Nucl. Fusion 59 (2019) 076029.
Speaker: Mr Leonid Eliseev (NRC Kurchatov Institute) -
As magnetic fusion devices operate at higher power, it is increasingly important to achieve an improved theoretical understanding of edge plasma turbulence and transport in order to control core energy confinement and to maintain the necessary limit on plasma heat flux to surrounding material components. Making use of advanced numerical methods from the computational fluid community and building on the success of continuum core-region codes (e.g., GYRO, GENE) the continuum finite-volume gyrokinetic code COGENT has been developed for edge plasma simulation. The code is unique in that it is based on a consistent high-order discretization and interpolation of the underlying equations, and hence the error (in particular near the X-point) can be bounded. It utilizes multiblock grid technology, in which logically distinct blocks are smoothly mapped from rectangular computational domains and a high-order interpolation is used to provide data communication between the blocks. To facilitate simulations of highly anisotropic plasma turbulence and transport the spatial grids are typically chosen to be aligned (or nearly aligned) with the magnetic flux surfaces in 4D or magnetic field lines in 5D. To facilitate time integration of physics problems involving multiple time scales an Additive Runge-Kutta (ARK) Implicit-Explicit (ImEx) method is used, such that selected stiff-terms (e.g., diffusion-like) can be advanced implicitly. The COGENT code solves full-F gyrokinetic equations for an arbitrary number of plasma species, which can also be coupled to a set of lower-dimensionality fluid equations in cases where a reduced fluid model is adopted to describe electrons or neutrals. Presently, the long-wavelength electrostatic limit is utilized and the options for self-consistent electrostatic potential variations include the gyro-Poisson equation, which supports kinetic electrons, and the quasi-neutrality equation, $\nabla \cdot j=0$, which can be coupled to a fluid electron model. The code offers a number of collision models, ranging from the simple Krook operator to the fully nonlinear Fokker-Plank (FP) operator, and includes an ad-hoc anomalous transport model that can be utilized for the case of 4D axisymmetric transport calculations.
COGENT is the only continuum gyro-kinetic code that can simulate 4D/5D edge plasma dynamics in a divertor geometry, which spans the magnetic separatrix of a tokamak. After a series of detailed verification studies, the first 4D (axisymmetric) full-F continuum gyrokinetic calculation adopting a simple (isothermal) model for electrons and an ad-hoc model for neutral density was performed for parameters characteristic of the DIII-D tokamak [1]. The simulations demonstrated the values of intrinsic rotation and radial electric field qualitatively similar to those observed in DIII-D experiments. More recently, a full-F 5D version of the COGENT code has become operational. A numerical algorithm utilizing locally a field-aligned multiblock coordinate system has been implemented to facilitate simulations of highly-anisotropic microturbulence in the presence of a strong magnetic shear. In this approach, the toroidal direction is divided into blocks, such that within each block cells are field-aligned and a non-matching (non-conformal) grid interface is allowed at block boundaries. The toroidal angle is playing the role of the “coarse” field-aligned coordinate, whereas the poloidal cross-section, comprised of the radial and poloidal directions, is finely gridded to resolve short-scale perpendicular turbulence structures and to support accurate re-mapping (interpolation) at block boundaries. In order to (a) facilitate verification activity and (b) demonstrate the efficiency of the field-aligned approach, the 5D single-block version of the code that uses the standard (non field-aligned) coordinate systems such as slab, cylindrical, or toroidal is maintained as well. The multiblock field-aligned version of 5D full-F COGENT code was successfully verified in simulations of the collisionless (universal) instability in a sheared slab geometry and ITG instability in a sheared toroidal annulus geometry by comparing against the Cyclone Base Case test results [2].
The present work reports on the first continuum gyrokinetic simulations of the ion-temperature-gradient (ITG) driven instability in a single-null geometry. The simulation model solves the long wavelength limit of the 5D full-F gyrokinetic equation for ion species coupled to the 3D quasi-neutrality equation for electrostatic potential variations $\nabla \cdot j=0$, where a fluid model is used for an electron response. The fast electron time scales that are present in the fluid electron model (e.g., time scales due to rapid electron conductivity), are treated implicitly by making use of the ARK method. Figure 1 shows self-consistent potential perturbations for the ITG mode obtained in illustrative COGENT simulations for the case of a model single-null geometry. The grid resolution corresponds to $N_ψ=160,N_ϕ=8,N_θ=640, N_{v_\parallel}=32,N_\mu=24$, and the simulation uses 1344 Haswell cores on the NERSC cluster. A single time step corresponds to $3 \times 10^{-8} s$ and takes 3 s of wall time.
[1] M. Dorf and M. Dorr, Continuum kinetic modelling of cross-separatrix plasma transport in a tokamak edge including self-consistent electric fields, Contr. Plasma Phys., 58, 434 (2018).
[2] M. Dorf and M. Dorr, Progress with the 5D full-F continuum gyrokinetic code COGENT, Contr. Plasma Phys., (2020); available online DOI: 10.1002/ctpp.201900113.
Speaker: Mikhail Dorf (Lawrence Livermore National Laboratory) -
The transport of energetic particles induced by the toroidal Alfvén eigenmode (TAE) excited by multiple super-Alfvénic particles was for the first time in stellarator/helical devices studied using comprehensive neutron and energetic particle diagnostics in Large Helical Device (LHD). It was clearly shown that both super-Alfvénic hydrogen and deuterium beam ions are transported to the outer region of the plasma owing to TAE. The total neutron emission rate ($S_n$) and stored energy ($W_p$) show that the degradation of global energetic particle confinement due to TAE is less than 5%. In addition, a study of energetic particle transport owing to the energetic particle mode (EPM) excited by the helically-trapped energetic particle pressure coupled with bulk plasma pressure (i.e., EIC) was performed. The neutron emission profile, $S_n$, and orbit following simulation show that the global confinement of helically-trapped energetic particles degrades by up to 50% due to EIC.
In a tokamak fusion reactor, two types of super-Alfvénic ions exist, e.g., fusion-born $\alpha$ particles for self-heating and beam ions for driving the current. Therefore, the combination of energetic particle pressure can excite TAE. Then, TAE can cause the transport of $\alpha$ particles and beam ions. In addition, energetic particle transport owing to EPM (e.g., off-axis fishbone/energetic particle-driven wall mode) is of considerable interest [A]. The understanding of energetic particle transport induced not only by TAE under multi super-Alfvénic particle condition but also by EPM in LHD is extremely helpful to clarify the fundamentals of energetic particle driven mode physics not only in stellarators/helical devices but also in tokamaks.
TAE excitation experiments were performed using intensive super Alfvénic hydrogen ($v_{bH}\sim1.5v_A$ where $v_A$ represents the Alfvén velocity) and deuterium ($v_{bD}\sim v_A$) beams in the relatively low field ($B_t = 0.6$ T) and high energetic particle beta ($\sim$1%) conditions (Fig. 1). The excited TAE composed of poloidal mode number/toroidal mode number of 1+2/1 has a relatively wide radial structure whose peak is at a normalized minor radius of ~0.6 [B]. A neutron flux detector, vertical neutron cameras, a tangential E||B neutral particle analyzer (NPA), and a fast ion loss detector (FILD) were employed at the same time to measure confined and escaping energetic particles. The TAE burst induces a slight decrease in $S_n$, a decrease in $W_p$ owing to the loss of heating source, an increase in the co-going transit fast hydrogen/deuterium flux, and an increase in the escaping fast ion flux corresponding to co-going (40$^\circ$ pitch angle) and transition orbits (60$^\circ$ pitch angle) near the injection energy. The short decay time of transition particles compared with co-going particles may result in shorter confinement time of transition particles. Although we observed a considerable increase in the fast neutral flux and escaping fast ion flux, $S_n$ and $W_p$ showed that the degradation of global energetic particle confinement due to TAE was less than 5%. In this experiment, the TAE amplitude decreased with an increase in the plasma density, and the considerable transport or loss of energetic particles were not observed at the line-averaged density of 2 $\times$ 10$^{19}$ m$^{−3}$.
By the intensive perpendicular neutral beam injection into the relatively low-density plasma ($\sim1\times10^{19}$ m$^{-3}$), the steep pressure gradient of helically-trapped energetic particles excites EIC classified as EPM. EIC located at the plasma edge ($r/a\sim$0.85) induced a decrease in $S_n$ by $\sim$50%, a decrease in $W_p$, and a change in the neutron emission profile. The orbit following simulation using the DELTA5D code [C] that included EIC fluctuation was performed to understand the energetic particle transport. A change in the neutron emission profile owing to EIC was successfully reproduced (Fig. 2). We determined that a change in the line-integrated neutron emission profile due to EIC indicated the considerable loss of helically-trapped beam ions, which excited EIC. The steep pressure gradient of beam ion immediately excites EIC by overcoming strong damping, whereas the pressure gradient of beam ion gradually excites TAE under weak damping conditions [D]. Therefore, the energetic ion transport owing to EIC is probably larger compared to the transport due to TAE. With an increase in the plasma density ($\sim1.5\times10^{19}$ m$^{-3}$), EIC became weaker, and no significant energetic particle transport was observed.
The effect of TAE and EIC on global energetic particle confinement in the high-density region in LHD is weak because the energetic particle pressure became lower due to the shorter slowing down time. This result showed the possibility of TAE and EIC control by a slight increase in the plasma density. This approach allows to us achieve the high-density operation of a helical-type fusion reactor.
[A] Okabayashi M. et al 2011 Phys. Plasmas 18 056112.
[B] Ogawa K. et al 2010 Nucl. Fusion 50 084005.
[C] Spong D. A. et al 2011 Phys. Plasmas 18 056109.
[D] Zheng L.-J. et al 2000 Phys. Plasmas 7 2469.Speaker: Dr Kunihiro Ogawa (National Institute for Fusion Science, National Institutes for Natural Sciences) -
Edge Localized Modes (ELMs) comprising of repetitive plasma eruptions from the edge of a tokamak plasma are a very common feature of high confinement mode (H-mode) operation in advanced tokamak devices. Each ELM outburst is associated with an expulsion of a large amount of energy and particles in a very short time that can potentially cause serious damage to plasma facing components and hence their control and mitigation are a major concern for safe tokamak operation. One of the most successful approaches for ELM control is by the use of resonant magnetic perturbations (RMPs) introduced from the edge of the plasma. A large number of experimental studies as well as theoretical investigations devoted to this topic are currently in progress worldwide.
The dynamics of ELMs is quite complex. The linear characteristics of the mode are well modeled in terms of the peeling/ ballooning instabilities$^1$ but their nonlinear evolution is not yet fully understood. Introduction of RMPs further complicates the dynamics and are best studied using numerical simulations. In an actual operational scenario additional physical factors such as plasma rotation can introduce significant modifications in the dynamics of ELMs. A few past studies$^{2,3}$ have shown that toroidal and/or poloidal rotation can significantly influence the stability properties of ELMs. In particular a sheared toroidal rotation is found to stabilize a Type-I ELM whereas a poloidal rotation can be either stabilizing or destabilizing depending on the direction of the flow. Some of these effects have also been experimentally identified. In the light of the above findings it is important to investigate what happens to the efficacy of RMPs in the presence of plasma rotation. Our present study is devoted to such an investigation.We have carried out detailed nonlinear simulations to study the dynamics of ELMs in the presence of a n=2 resonant magnetic perturbation. The simulations have been done on the Culham Transporter of Ions and Electrons (CUTIE) code$^{4,5}$ - a two-fluid initial value electromagnetic nonlinear global code which solves the full set of model two fluid equations at a scale intermediate between the device size and the ion gyro-radius and takes account of classical and neoclassical transport effects. A periodic cylinder model of the tokamak geometry is adopted with the toroidal curvature effects of the magnetic field lines kept to first order in the inverse aspect ratio. Poloidal mode coupling is implemented through additional curvature terms. ELMs are simulated by introducing a particle source in the confinement region and a particle sink in the edge region. CUTIE is capable of simulating multiple edge relaxation periods and has in the past also successfully modeled the L-H transition behavior in COMPASS-D$^4$. More recently CUTIE has been used to study ELM dynamics in the presence of RMPs$^6$ as well as pellet injection$^7$.
To study the effect of plasma rotation, we have introduced a toroidal momentum source to generate an equilibrium plasma rotation. An external magnetic perturbation of n=2 is also present to simulate the effect of RMPs. Our principal findings are as follows. The presence of flows significantly impacts both the magnitude and frequency of the ELM modes. In Fig.1 we show the time evolution of the ion temperature fluctuations (dTi/Ti0) for various situations e.g. with or without the presence of RMPs and flows. We see that the presence of a toroidal flow in the co-current direction has a positive impact on the mitigation process. In particular, the amplitudes of the ELM bursts are seen to be significantly diminished and their frequencies increased. This also leads to an overall improvement of the energy confinement and an increase in the plasma beta. This is shown in Fig. 2 for the various cases. The flow values that we have used at the axis are M= 0.011, 0.027 and 0.042 where M the Mach number is defined in terms of the Alfven velocity. Such flows are likely to be present in tokamaks either due to external driving from neutral beams or spontaneous generation from turbulence effects. Other noticeable changes (not shown) are in the profile modifications at the edge and the spectra of energy transfer. In general we also find that a co-current flow is more effective than a counter-current flow – in fact a large counter-current flow actually reduces the stabilizing influence of the RMP. Our results could be of direct relevance to present day experiments on RMP control of ELMs on tokamaks like DIII-D where substantial sheared flows can exist due to unbalanced NBI injection$^8$.References
$^1$P.B. Snyder, H.R. Wilson, T.H. Osborne et al, Plasma. Phys. Contr. Fus. 46 (2004) A131.
$^2$F. Orain, M. Becoulet, J. Morales et al, Plasma Phys. Contr. Fus. 57 (2015) 014020.
$^3$N. Aiba, M. Furukawa, M. Hirota et al, Nuclear Fusion 51 (2011) 073012.
$^4$A. Thyagaraja, Plasma Phys. Contr. Fus. 42 (2000) B255.
$^5$A. Thyagaraja, M. Valovic and P.J. Knight, Phys. Plasmas 17 (2010) 042507.
$^6$D. Chandra, A. Thyagaraja, A Sen et al, Nuclear Fusion 57 (2017) 076001.
$^7$D. Chandra, A Sen and A. Thyagaraja, Plasma Phys. Contr. Fus. 61 (2019) 085019.
$^8$T.E. Evans, M.E. Fenstermacher, R.A. Moyer et al, Nuclear Fusion 48 (2008) 024002.
Speaker: Dr Debasis Chandra (Institute for Plasma Research) -
Wendelstein 7-X (W7-X) is an optimized stellarator of the HELIAS type. ASDEX Upgrade is a medium sized tokamak. In electron cyclotron heated (ECRH) hydrogen plasmas the central ion temperature is clamped at Ti,0 ~ 1.5 keV in both W7-X (Figure-1) and AUG. These findings are found to be virtually independent of heating power and electron density, and for W7-X, independent of magnetic configuration. In both devices ion scale turbulence (ITG/TEM) is thought to be responsible for the enhanced turbulent transport. In combination with the off-axis ion heating profile that stems from the electron-to-ion energy transfer-profile, the enhanced turbulent transport leads to the observed clamping of Ti,0. For future fusion reactors with dominant electron (i.e. $\alpha$-particle) heating, of either stellarator or tokamak type, these W7-X and AUG experiments expose the potential performance limiting ion-temperature-“clamping” issue. Active turbulent suppression may then become a necessity.
The clamped ion temperature has been studied in an AUG deuterium plasma with high central density (1.10^20m-3) and high plasma current (1 MA) in H-mode, and Pecrh=4 MW. Central ECRH deposition yielded Te,0 = 3.4 keV and Ti,0 =1.9 keV. Moving Pecrh = 3.3 MW off-axis to r/a =0.6 and keeping only 0.7 MW in the core, resulted in unchanged pedestal parameters Te = Ti = 1 keV. However, central Te,0 = 2.3 keV significantly reduced while Ti,0 remains at 1.9 keV within error bars, whereas the total heat flux to the ions is reduced by a factor of two. This behaviour has been predictively modelled with the trapped-gyro-Landau-fluid model (TGLF) for turbulent transport, and relates to the result that in tokamaks the ratio Te/Ti has a strong effect on ITG stability. In a direct comparison to W7-X, in AUG experiments in hydrogen were conducted with 122 different heating and density variations between Pecrh = 0.5-5MW and ne,0 = 2-8 10^19m-3, in which Ti clamped at 1.5 keV, independent of confinement H- or L-mode. Turbulence suppression in these scenarios would be required to obtain enhanced Ti.
In W7-X, neoclassical transport losses have been minimized through a reduction of the effective ripple down to εeff ~ 0.8%. Its design aims are steady state operation up to a pressure <β> ~ 5% with mainly dominant electron heating. Neoclassical (NC) transport predictions show that this may be achievable with Pecrh ~ 10MW, and would have an energy confinement fISS04 = $\tau$e / $\tau$ISS04 > 2, compared to the ISS04 confinement scaling. Assuming Pecrh ~ 4.5MW in such simulations, ion temperatures of up to Ti=2.8-3.5 keV may be achieved, depending on the configuration chosen (Figure-1). In experiments after wall conditioning by means of boronization and with divertor operation, a wide operational window with Pecrh = 0.5-6MW and ne,0 = 0.2-1.4 10^20m-3 in four different magnetic configurations with <$\varepsilon$eff> 0.8% - 2.5%, was obtained. In contrast to the NC predictions, a confinement of only fISS04 = $\tau$e / $\tau$ISS04 < 0.7 was found, virtually independent of Pecrh and εeff, and degrading with density compared to the scaling. Most notably, in W7-X the achieved central ion temperature across the standard ECRH database was at maximum Ti ~1.5 ± 0.2 keV, independent of configuration, whereas the electron temperature could vary more widely as Te ~ 1 to 10 keV (Figure-1). Radiation losses and charge exchange losses have been excluded beyond reasonable doubt as the cause of the increased core transport losses.
For W7-X, various candidate turbulent mechanisms are investigated using non-linear and linear gyrokinetic flux surface averaged simulations. At low-density gradients with a/Ln $\leq$ 1, ion temperature gradients (ITG) are thought to dominate the ion heat transport, whereas trapped electron modes (TEMs) drive the electron heat transport. At high-density gradients a/Ln > 3-4 and low temperature gradients a/LT $\leq$ 1, TEMs dominate the overall transport. However, the so-called maximum-J property of W7-X entails that when the temperature and density gradients align like a/LTi ~ a/Ln, the turbulence growth rates are strongly reduced and improved confinement may be the result. This leads to the so called “stability valley”, as is thought to be observed in e.g. the post-pellet plasmas.
A power balance on a selected dataset (Pecrh = 2-6MW) and plasma density (3.5-7·10^19 m-3) shows that, focussing on r/a = 0.5-0.7, the electron heat diffusivity remains virtually unchanged at χe,exp ~ 0.7 ± 0.1 m^2/s despite a large variation of Pecrh. The electron transport therefore has a low degree of stiffness as seen in Figure-2a, as well as seen in separate heat pulse propagation experiments which show Se= $\chi$$\epsilon$$^{HP}$/$\chi$$\epsilon$$^{PB}$ $\leq$ 2. The ion heat diffusivity ranges from 0.5-2 m^2/s, with an average $\chi$i,exp ~ 1.1 ± 0.5 m2/s, see Figure-2b. The low variation of a/LTi and moderate variation of χi,exp for a given radius, may imply some degree of ion profile stiffness S=$\Delta$$\chi$/$\chi$ consistent with ITG turbulence, but the variation of ion heatflux Qi in gyrobohm units, is too small compared to experimental errors to be conclusive. Also, although the variation of Te/Ti =1-5 in the plasma center and Te/Ti =1-2 at mid radius, is expected to affect ITG turbulence, we have thus far not found any evidence for this in our heat transport studies. The possible role of ITGs on transport, and its suppression, has however been experimentally shown by introducing a significant density gradient by e.g. hydrogen ice-pellet injection. In the post pellet phase, density and ion temperature gradients transiently increase and align as a/Ln ≈ a/LTi =3-5. Consistent with our gyrokinetic modelling given the maximum-J property of W7-X, transport is reduced and higher central Ti,0 $\approx$ Te,0 $\sim$ 3keV are achieved (green stars in Figure-1). A power balance analysis at the highest Ti indeed shows that the ion heat transport drops to neoclassical levels $\chi$i,exp ≈ $\chi$i,NC (stars in Figure-2b), implying (ITG) turbulence suppression. Simultaneously a reduction in density fluctuations is seen from our phase contrast imaging diagnostic, also implying reduced turbulence levels.
Extrapolating the performance of W7-X plasmas for enhanced ECRH power is difficult. However, assuming the averaged experimental heat diffusivities, $\chi$e = 0.7 m2/s and $\chi$i = 1.1 m2/s, in our NTSS predictive transport model, one would require Pecrh > 50MW to approach the design aim of <$\beta$> = 5%. Therefore, turbulence suppression (through e.g. induced density gradients) is essential for future high performance electron heated scenarios in W7-X. New tools for scenario development in W7-X are: cryo-pumping; a continuous pellet injector; and enhanced ECRH power (8MW). The PNBI $\sim$ 7MW may yet reveal other venues to high performance.
Speaker: Marcus Nicolaas Arnoldus Beurskens (Max-Planck-Institut für Plasmaphysik, Greifswald, Germany) -
Due to the limited power exhaust capability of the divertor, a future DEMO reactor needs a high core/pedestal radiation level, controlled and tailored by an appropriate seed impurity mix. The core/pedestal seeding has to be integrated with substantial divertor radiative cooling and a no/very small ELM plasma regime. Required boundary conditions of the seeding scenario are sufficient energy confinement and low fuel dilution. Taking into account the radiative capabilities of the potential seeding species (N2, Ne, Ar, Kr, Xe), only Ar is expected to contribute substantially to both pedestal and divertor radiation for reactor conditions. Therefore, a likely scenario is the combination of Ar with Kr or Xe for core radiation at reduced dilution and with N2 or Ne for enhanced divertor radiation. For the divertor seeding species a high divertor impurity compression is also required to avoid excessive core dilution. This paper reports about recent studies in ASDEX Upgrade towards the development of corresponding DEMO seeding scenarios with emphasis on operation with no ELMs, good energy confinement, and compatibility with the tungsten plasma facing components.
One route to a no/small ELM scenario is the reduction of the edge pressure gradient in the steep gradient zone by radiative losses. Ar is particularly suited for AUG conditions since it creates a strongly radiating ring in the pedestal region. The radiation effect can be combined with additional pedestal transport effects as provided by turbulence or magnetic perturbations to assure ELM suppression. Recently, a new type of ELM-free H-mode with good energy confinement has been discovered at ASDEX Upgrade [1]. The scenario exhibits similarities with the EDA H-mode from Alcator C-Mod, a quasi-coherent electromagnetic mode with toroidal mode number n ~20 in the pedestal region results in complete stabilization of ELMs at good confinement. First attempts to combine this scenario with a feedback controller for the power flux over the separatrix via Ar seeding [2] showed ELM-free conditions at up to 5 MW injected NBI and ECRH combined power (figure 1, left panels).
At higher power and core radiation, Psep could be kept low, in the vicinity of the H-L transition, but ELMs remained, albeit at reduced size (Fig 1, mid panels). Future studies are planned to determine whether a refined seeding strategy will allow for full ELM suppression also at high heating power. Figure 1 (right panel) compares the electron pressure gradient in the pedestal region with and without Ar seeding for H-mode discharges with and without ELMs. Indeed, Ar seeding reduces the maximum pressure gradient, while global confinement is even improved due to increased pressure further inwards.
A prerequisite for efficient divertor radiative cooling is the achievement of a high divertor impurity density compared to its core density, namely a high divertor compression. To shed light on the underlying processes, calibrated gas puffs of N2, Ne and Ar have been injected into ASDEX Upgrade H-mode plasmas for different values of the divertor neutral pressure. The temporal development of core densities and recycling rates was measured by charge exchange recombination spectroscopy, Zeff variations and a SPRED VUV spectrometer, see figure 2. The fastest removal rates are observed for Ar, shortening from ~ 0.3 to ~ 0.1 s as the divertor neutral pressure is increased from 0.6 to 4 Pa. At high pressure, the shortening saturates and Ne and Ar show equal pumping times. Such a shortening is not observed for N2, which appears to be dominated by wall storage and release processes.
Simple particle balance analysis allows already some conclusions. At very low pressure, pumping rates are expected to scale with the inverse square root of mass. At very high pressure, collisions with D2 molecules lead to an entrainment of the impurities in the D2 flow and equal removal rates for all species. The observed removal rates are proportional to the product of the divertor compression and the pumping speed. Faster removal rates of Ar vs Ne give evidence of a higher divertor compression of Ar vs. Ne. SOLPS calculation including the effects of neutral collisions are required for further quantification.
In conclusion, Ar pedestal radiation is a promising tool for reduction of the pedestal electron pressure gradient, and thus an important element for a no-ELM scenario. Active control needs to be expanded towards the tailoring of the spatial profile of the radiation, rather than the pure radiated power. A step in this direction has recently been demonstrated for the X-point radiator regime in ASDEX Upgrade [3]. First attempts for combining core Ar radiation with RMP ELM suppression showed short (0.5 s) phases of good performance, but were hampered by the subsequent occurrence of a locked mode.
[1] L. Gil et al, Stationary ELM-free H-mode in ASDEX Upgrade’, EPS 2019, submitted to Nuclear Fusion
[2] A. Kallenbach et al., Nuclear Fusion 52 (2012) 122003.
[3] M. Bernert et al., this conference.Speaker: Dr Arne Kallenbach (Max-Planck-Institut f. Plasmaphysik) -
In this paper we will present nonlinear full-$f$ electromagnetic gyrokinetic simulations of turbulence in the pedestal and scrape-off layer (SOL) region of a tokamak. The algorithms in the Gkeyll code solve the electromagnetic gyrokinetic equations using a continuum high-order discontinuous Galerkin scheme. The equations are written in a sympletic form in which the particle parallel momentum is used as a coordinate. The details of our algorithms are presented in[1,2], with previous work in simulating electrostatic turbulence in in helical SOL plasmas in[3,4]. As far as we are aware, these are the \emph{first} fully nonlinear full-$f$ electromagnetic simulations in the high-$\beta$ regime on open field-lines.
Our setup is a helical, open field-line model of a NSTX-like SOL. We use a non-orthogonal field-aligned coordinate system. The simulation domain is a flux-tube on the outboard side that wraps helically around the torus, with the field-lines intersecting metal divertor plates on either ends. The interaction of the plasma with the divertor plate is modeled using a sheath boundary condition which allows current fluctuations into and out of the divertor plates.(The results presented in this abstract focus on the SOL with a helical model, but we have extended the code to general geometry sufficient for a pedestal and SOL with a limiter, with X-point capabilities available soon.) The simulation parameters approximate SOL conditions in an H-mode deuterium plasma: $B_\text{axis}=0.5$ T, $R_0=0.85$ m, $a=0.5$ m, $T_{e0}=T_{i0}=40$ eV. (The plasma $\beta$ is enhanced by a factor of 10 relative to the steady-state experiment to show the robustness of the algorithm, even for ELMs or or other high-$\beta$ transients.) The plasma core supplies a particle source for the SOL.
The bad curvature drive causes interchange instabilities in which blobs are ejected intermittently radially outwards. This is seen in the density contours in Fig.1 in which mushroom like structures appear. Significant magnetic fluctuations of around $2.5\%$ are seen in $|\delta B_\perp|/B_0$, greatly modifying the transport as compared to when the electromagnetic terms are neglected. In Fig.2 we show the radial profiles of density, temperature and beta for both electrons and ions. These are compared to the case in which the electromagnetic terms have been turned off. It is seen that in the ES case the profiles are shallower, indicating that the radial transport is less in the EM case. In general, the inclusion of the EM terms makes the turbulence more intermittent, significantly changing the transport in the high-$\beta$ regime. Even though the magnetic field-lines, Fig.3, are tied right at the divertor plates, they slip due to sheath resistance, and in the interior, fluctuate and reconnect as the blobs drag the field-lines in their outward motion. The results summarized here are for turbulence in the SOL.The extension of these results to include the pedestal physics also will be presented.
-
Ammar H. Hakim, Noah R. Mandell, T. N. Bernard, M. Francisquez, G. W. Hammett, E. L. Shi, "Continuum Electromagnetic Gyrokinetic Simulations of Turbulence in the Tokamak Scrape-Off Layer and Laboratory Devices", Submitted to Phys. Plasmas
-
Tess N Bernard, Eric L Shi, KW Gentle, Ammar Hakim, Gregory W Hammett, Timothy Stoltzfus-Dueck,and Edward I Taylor. "Gyrokinetic continuum simulations of plasma turbulence in the Texas Helimak". Phys. Plasmas, 26(4):042301, 2019
-
N R Mandell, A Hakim, G W Hammett, and M Francisquez. "Electromagnetic full-f gyrokinetics in the tokamak edge with discontinuous Galerkin methods", J. Plasma Phys., 86(1):905860109, 2020
-
Eric L Shi, Gregory W Hammett, Timothy Stoltzfus-Dueck, and Ammar Hakim. "Full-f gyrokinetic simulation of turbulence in a helical open–field–line plasma". Phys. Plasmas, 26(1):012307, 2019
Speaker: Mr Ammar Hakim (Princeton Plasma Physics Labaroratory) -
-
The capability to suppress edge localized modes (ELMs) is crucial for the success of ITER because the transient heat loads on the divertor due to ELMs would reduce the lifetime of plasma facing components to unacceptable levels. ELMs can be suppressed with the application of resonant magnetic perturbations (RMPs). But a side effect of RMPs is enhanced particle flux, or density pump-out, that reduces the plasma density by up to 50% [Evans 2006] and can significantly degrade fusion efficiency. Kinetic-level understanding of these RMP-driven phenomena is essential for predicting ITER's performance but is incomplete as of yet. One particular puzzle from experimental observation [Evans 2006] and transport modeling [Hu 2019] is how RMPs interact with neoclassical and turbulent transport to produce density pump-out even before magnetic islands penetrate the pedestal top while at the same time keeping electron heat well confined (or even improving confinement). Advanced MHD-assisted gyrokinetic simulations with the code XGC in realistic divertor geometry based on a DIII-D H-mode discharge with $n=3$, even parity resonant magnetic perturbations (RMPs) now reproduce these two experimental observations.
This study utilizes the global total-f gyrokinetic particle-in-cell code XGC [Ku 2018] with an RMP field from a linear two-fluid M3D-C1 [Ferraro 2009] calculation for arguably the highest-fidelity study of RMP-driven plasma transport to date. The combination of neoclassical and turbulent particle flux, in the presence of neutral particle ionization and charge exchange, explains a 50% density pump-out in less than 100 ms in the pedestal up to normalized poloidal flux $\psi_N\approx 0.96$ as well as the suppression of electron heat flux in the steep pedestal slope ($\psi_N\approx 0.97$). The majority of the enhanced particle flux at $\psi_N \ge 0.985$, where the magnetic field is stochastic, is from neoclassical transport (cf. [Hager 2019]). Turbulent transport is enhanced at $\psi_N \le 0.985$, where nested flux-surfaces remain intact. On the $\psi_N\simeq 0.97$ surface, both the turbulent particle and electron energy fluxes increase significantly due to the RMPs (see Fig. 1). But the electron heat flux is largely suppressed as the result of a cancellation between outward heat flux at longer wavelength ($k_\theta \rho_i \le 0.3$) and inward heat flux at shorter wavelength ($k_\theta \rho_i \ge 0.3$).
The RMS turbulence amplitude is enhanced due to the RMPs throughout the pedestal with a strong $n=3$ component [Fig. 2 (a) and (b)]. Closer examination of the turbulent fluctuations shows that ion drift modes (ITG) are active from $\psi_N\approx 0.94$ inward and electron drift modes (TEM) are active from $\psi_N\approx 0.94$ outward with a median poloidal mode number $k_\theta \rho_i \approx 0.25$, where $\rho_i$ is the ion gyroradius. The most influential (to the transport fluxes) RMP-induced modification is the amplification of the TEMs from the pedestal shoulder $\psi_N=0.94$ outward. This is shown at $\psi_N=0.97$ in Fig. 3, where the amplification of the TEMs is consistent with a reduction of the ExB shearing rate. The insights gained from these high-fidelity simulations suggest that a reduction in the TEM mode growth rate by an external means, e.g. enhanced edge rotation, could be a method to reduce the density pump-out. Our gyrokinetic understanding will also improve longer time-scale/lower-fidelity (e.g. [Hu 2019]) modeling to predict the steady state plasma profiles under RMP field. These discussions will also be presented.
This work is supported by the US Department of Energy (DOE) under contract nos. DE-AC02-09CH11466, DE-FC02-04ER54698. Computing resources: INCITE program at ALCF (DE-AC02-06CH11357) and ERCAP award at NERSC (DE-AC02-05CH11231).
[Evans 2006] T. E. Evans, R. A. Moyer, K. H. Burrell et al., Nature Physics 2, 419-423 (2006)
[Hu 2019] Q. M. Hu, R. Nazikian, B. A. Grierson et al., Physics of Plasmas 26, 120702 (2019)
[Ku 2018] S. Ku, C. S. Chang, R. Hager et al., Physics of Plasmas 25, 056107 (2018)
[Ferraro 2009] N. M. Ferraro, S. C. Jardin, Journal of Computational Physics 228, 7742-7770 (2009)
[Hager 2019] R. Hager, C. S. Chang, N. M. Ferraro, R. Nazikian, Nuclear Fusion 59, 126009 (2019)Speaker: Robert Hager (Princeton Plasma Physics Laboratory) -
E-mail: ysna@snu.ac.kr
Edge Localized Modes (ELM) are rapid MHD events occurring at the edge region of tokamak plasmas, which can result in damages on the divertor plates. Therefore, to fully suppress ELM via resonant magnetic perturbation (RMP) [1-4] is of great help to reach and sustain high-performance H-mode plasmas. It was found that certain conditions must met for the RMP-driven ELM crash suppression [5], so understanding its mechanism is crucial for reliable ELM control using RMP. The initial understanding of its mechanism was a stabilization of linear edge instability due to the pedestal gradient degradation by RMP. However, experimental observation showed that peeling-ballooning mode (PBM)-like mode structures remained in the ELM suppression phase [6]. In addition, the bifurcation of mode rotation in the edge region was found to be closely related to mode suppression [7]. Therefore, the initial understanding may have difficulties in explaining these experimental findings and it indicates that additional physics properties should be included to understand the mechanism. For this purpose, we have carried out nonlinear MHD simulations with 3D reduced MHD code, JOREK [8] for a recent n=2 RMP-driven ELM-crash-suppression in KSTAR [7]. We successfully reproduced [9] the natural ELM without RMP (Fig.1(a)), mode mitigation with small RMP strength ($I_{RMP}=2kA$, Fig.1(b)), and mode suppression by experimental RMP strength ($I_{RMP}=4kA$, Fig.1(c)). Also, such ELM-crash-suppression is attributable not only to the degraded pedestal but also to direct coupling between peeling-ballooning mode (PBM) [10] and RMP-driven plasma response. The coupling between PBM and RMP can 1) enhance the size of the island at the pedestal reducing the instability source by further pedestal degradation, and 2) increase the spectral transfer between edge harmonics preventing catastrophic growth and crash of unstable mode. Because of these effects, PBMs are nonlinearly saturated, and they persist during the suppression phase without a mode crash. This outcome is consistent with the previous studies [11,12]. In addition, the locking (or rotation bifurcation) of PBMs has been numerically simulated during the suppression phase. This mode-locking is a distinguishing feature of the mode suppression as rotating mode structure remains for the natural ELM and mode mitigated case (Fig. 1(d)). PBM locking may enhance the interactions between PBMs and RMP, and therefore, it is favorable to RMP driven ELM suppression. Here, slowly rotating PBM before RMP application can be easily locked by RMP. As $V_{E×B}$ is approximately equal to the initial PBM rotation in our case, $V_{E×B}\approx0$ in the pedestal will be advantageous to onset of ELM-crash-suppression. To test our hypothesis, we conduct additional RMP-ELM simulation with modified $V_{E×B}$, and confirm that the mode suppression is not achieved with enlarged $V_{E×B}$ on top of the pedestal, which may support the importance of $V_{E×B}\approx0$ on ELM-crash-suppression.
References:
- T. Evans et al., PRL 92, 235003 (2004)
- Y. Jeon et al., PRL 109, 035004 (2012)
- Y. Sun et al., Phys. Rev. Lett. 117, 115001 (2016)
- W. Suttrop et al., Plasma Phys. Control. Fusion 59, 014049 (2017)
- J. K. Park et al., Nature Physics 14 (2018), 1223
- J. Lee et al., Physics Review Letter 117 (2016), 075001
- J. Lee et al., Nucl. Fusion 59, 066033 (2019)
- G. T. A Huysmans et al., Plasma Phys. Controlled Fusion 51, 124012 (2009)
- S. K. Kim et al., Nucl. Fusion 60, 026009 (2020)
- J. W. Connor et al., Plasma Phys. Controlled Fusion 27, 621 (1985)
- M. Becoulet et al., PRL 113, 115001 (2014)
- F. Orain et al., Phys. Plasmas 26, 042503 (2019)
Speaker: SangKyeun Kim (Seoul National University) -
A fusion reactor based on a stellarator design has the advantage of an easier access to long pulse scenarios. In fact, one of the main goals of Wendelstein 7-X (W7-X), the largest advanced stellarator in the world, is to demonstrate the steady-state capabilities of the stellarator line. From 2022 onward, all plasma-facing components will be water-cooled, which should enable pulse duration of up to 30 minutes with up to 10 MW of ECRH input power. Therefore, in the recent campaign [1,2] a number of experiments were performed in order to prepare long pulse operation, addressing issues like the development of stable detachment, control of the heat and particle exhaust and the influence of leading edges on plasma performance. We have shown that an island divertor is a good concept for steady state heat and particle exhaust. A highlight of the recent campaign was a robust detachment scenario, which allowed removing power loads due to direct contact of divertor target plates with the plasma, while reaching neutral pressures at the pumping gap entrance yielded the particle removal rate close to the values required for stable density control in steady state operation.
The heat and particle exhaust in W7-X is realized with help of an island divertor [1], which utilizes large magnetic islands at the plasma boundary. In the so-called “standard configuration” with $n/m$ = 5/5 islands at the boundary attached plasmas showed very efficient heat flux spreading and favorable scaling with input power.
In Figure 1 an overview of peak heat loads to all 10 divertors is shown. At a given input power, there is a plasma density range which yields divertor heat flux below allowed 10 MW/m$^2$. For attached plasmas with $P_{\rm ECRH}\leq6$ MW, $q_{\rm max}$ was typically in the range of 4 to 6 MW/m$^2$, which, when extrapolated to 10 MW of input power, provides a safe operational range at line integrated plasma densities above $5∙10^{19}$m$^{-2}$.
Increasing ne leads to increasing plasma radiation and at radiation fractions $f_{\rm rad}>0.8$ to complete plasma detachment. The detachment regime led to a significant reduction of the divertor heat and particle fluxes. An example of a 30 second discharge [3] with peak heat flux reduced by factor of 9 is shown in Figure 2. Although a significant local up-down asymmetry due to edge drifts [4] is observed in the attached state at $t$ < 3 s, it vanishes when transitioning into detachment. In the downstream (near-target) region of the SOL, Langmuir probes show a drop of the electron temperature from ca. 30 eV to ca. 5 e V. Additionally, Coherent Imaging Spectroscopy shows that during detachment the parallel flow velocity of C2+ in the scrape-off layer is significantly reduced [5].
Before the onset of detachment, the values of the downstream electron density are significantly higher ($1.2-1.4\cdot 10^{20}$ m$^{-3}$) than electron densities near the separatrix (ne,sep is in the range $4-6\cdot 10^{19}$ m$^{-3}$. This difference between upstream and downstream density indicates that the divertor operates in the high recycling regime. This detachment regime is characterized by low impurity concentration ($Z_{\rm eff} = 1.5$) and high neutral pressure ($p_n \leq $ 0.1 Pa) in the subdivertor volume. Estimates of total pumping rate for detached discharges at ca. 0.6$\cdot 10^{21}$ [atoms/s] show that at these neutral pressures particle exhaust is at the level required for steady-state operation. As the radiation characteristics of carbon is similar to that of nitrogen, the results obtained here should to a large extent be valid for detachment driven by nitrogen seeding [6] . In steady state W7-X will operate with continuous pellet fueling, therefore also several discharges were performed to see if it is possible achieve stable detached scenarios with plasma density stabilized with low frequency pellet injection.
An example of such a discharge is presented in Figure 3, where peak heat flux is reduced from ca. 5 MW/m$^2$ to ca. 2 MW/m$^2$, neutral pressure at the pumping gap is kept at 0.06 Pa and Zeff stays at the level of ca. 1.2.
As W7-X is a low shear device, edge islands are sensitive to changes in toroidal current [7]. Several measures to counteract strike-line movements induced by the plasma current evolution were tested, e.g. by use of external coils or ECCD. An essential issue of long pulse operation is impurity control. A series of experiments were performed to study the behavior of intrinsic impurities as well as seeding low and highly recycling species to enhance plasma radiation. Overall W7-X shows good impurity control in low and high density discharges with Te/Ti > 1. We have found that despite high influx of carbon into the SOL during discharges with dedicated overloading of the leading edges, the plasmas remained stable. Line-of-sight averaged Zeff stayed below 1.5 throughout the discharge and radiation increased at the plasma edge only.
The results presented in this work form a promising outlook on the overall steady state compatibility of the detached island divertor concept in future experiments and a stellarator-based reactor.[1] T. S. Pedersen et al., Plasma Physics and Controlled Fusion 61, 014035 (2019).
[2] R. C. Wolf et al., Phys Plasmas 26 (2019).
[3] M. Jakubowski et al., Physical Review Letters (submitted).
[4] K. C. Hammond et al., Plasma Physics and Controlled Fusion 61, 125001 (2019).
[5] V. Perseo et al., in IAEA-FECNice, France, 2020).
[6] F. Effenberg et al., Nuclear Fusion 59 (2019).
[7] Y. Gao et al., Nuclear Fusion 59 (2019).Speaker: Marcin Jakubowski (Max-Planck-Institut für Plasmaphysik) -
Simultaneous high ion temperature ($T_i$) and high electron temperature ($T_e$) regime has been significantly extended due to the optimized heating scenario in the LHD deuterium (D) campaign. Such high-temperature plasmas were realized by the simultaneous formation of an electron ITB (e-ITB) and an ion ITB (i-ITB) by the combination of high power NBI and ECRH. In the high $T_i$ operation the EIC (Energetic ion driven interchange) modes, which causes the $T_i$ degradation, was clearly suppressed due to the ECRH superposition. Furthermore, the e-ITB was successfully formed without degradation of the ion thermal confinement for the high $T_i$ plasmas. This is due to the $T_{e0}$/$T_{i0}$ maintaining the moderate value of 0.66. Consequently, $T_{e0}$ could be increased up to 6.6 keV with the $T_{i0}$ ~10 keV (previous achievement at FEC2018 was $T_{i0}$ = 10 keV and $T_{e0}$ = 3.6 keV [1]).
In future reactors, the fusion reaction is expected to be sustained under the electron heating dominant condition, where both $T_i$ and $T_e$ are high. Thus not only the investigation of the confinement improvement but also the characterization of the thermal transport for the plasmas, of which $T_i$ and $T_e$ are simultaneously high, are necessary. In the present status, such a plasma condition is realized by the combination heating of a NBI and an ECRH. The effect of a $T_e$/$T_i$ and/or an ECRH on the ion thermal confinement has been studied in several devices. In recent years, an integration of high $T_i$ and high $T_e$ with the simultaneous formation of an i-ITB and an e-ITB has been achieved in the LHD [2]. The paper shows the successful extension of simultaneous high $T_i$ and high $T_e$ regime in the LHD deuterium operation due to the suppression of the EIC modes and control of the $T_e$/$T_i$ value.
The EIC is triggered by helically trapped ions at lower order magnetic resonant surface of $m/n$ = 1/1. The EIC causes loss of high energy ions both in the plasma core and the edge, leading to the decrease in $T_i$ [3]. Thus the mitigation of the EIC is a key for realizing higher $T_i$. Figure 1 shows the comparison of the time evolution of (a) the heating power, (b) the line-averaged-electron density $n_{e-fir}$, (c) the poloidal magnetic fluctuation amplitude $b_\theta$, (d) the neutron emission rate $S_n$, (e) the $T_{e0}$, and (f) the $T_{i0}$ without and with 2.5-MW ECRH superposition.
The NBI power was similar (~25 MW) and the operated magnetic configuration was same as $R_{ax}$ = 3.58 m and $B_t$ = 2.87 T. The C pellet was injected at 4.57 s to control the impurity and to increase the S/N of CXS measurement. The bursty increase of $b_\theta$ and the synchronized drop in the $S_n$ correspond to the EIC occurrence and the loss of the high energy ions, respectively. Due to the ECRH superposition the EIC event was clearly suppressed. Consequently, the $S_n$, which is reflected the amount of the confined high energy ions, increased and this contributed to realizing the higher $T_{i0}$. Note that the $b_\theta$ increase at $t$ = 5.1 s was due to the NBI breakdown.
Although the ECRH is effective to mitigate the EIC, the increase in $T_e$/$T_i$ causes the destabilization of the ITG mode [4], leading to the degradation of the ion thermal confinement. Figure 2 shows the dependence of the normalized scale length of $T_i$ on the $T_e$/$T_i$ at the effective minor radius of $r_{eff}$ = 0.1 m.
These data were obtained in the high $T_i$ operation with the superposed on-axis ECRH power from 0 to 5.4 MW. The significant degradation of the $R/L_{Ti}$ was observed both in the H and D plasmas when the $T_e$/$T_i$ exceeded ~0.7. These results show that not only the mitigation of EIC by ECRH but the control of $T_e$/$T_i$ in the moderate range is important for realizing high $T_i$ and high $T_e$ simultaneously.
Presently we could successfully obtain the high $T_i$ plasma (10 keV) with the increased $T_{e0}$ up to 6.6 keV by the on-axis ECRH with 1.8 MW (1/3 of the LHD-ECRH full power). The $T_{e0}$/$T_{i0}$ was 0.66, which value was slightly lower than the threshold of $R/L_{Ti}$ degradation shown in Fig. 2, and the EIC was also mitigated in the discharge. Figure 3 shows the radial profile comparison of (a) $T_e$, (b) the electron thermal diffusivity $\chi$$_e$, (c) $T_i$, and (d) the ion thermal diffusivity $\chi$$_i$ between the high $T_i$ plasmas without and with 1.8-MW ECRH.
The NBI power was similar (~30 MW) and the operated magnetic configuration was same as $R_{ax}$ = 3.6 m and $B_t$ = 2.85 T. The plasma W/O ECRH was obtained before FEC2018. The EIC events and the accompanying loss of the high energy ions were observed during the discharge W/O ECRH. Applying the on-axis ECRH superposition, $T_e$ gradient increased and the $\chi$$_e$ decreased especially in $r_{eff}$ < 0.2 m due to the e-ITB formation. For the ion, the reduction in the $\chi$$_i$ was observed around the plasma half radius associated with the EIC suppression and the peaked profile was maintained in the plasma central region with the $T_{i0}$ of 10 keV.
Figure 4 is the summary of the extension of the high temperature operational area in the LHD.
The simultaneous high $T_i$ and high $T_e$ regime was significantly extended from previous IAEA FEC. In the LHD deuterium experiment, we have attained the $T_i$ of 10 keV due to the better confinement accompanied with the stronger i-ITB compared with hydrogen plasmas [1, 5]. Present results of the extension of the high temperature regime reinforce the feasibility of the helical reactor and enable us to investigate the plasma confinement property close to ignition temperature condition. [1] H. Takahashi et al., Nucl. Fusion 58 106028 (2018).
[2] H. Takahashi et al., Nucl. Fusion 57 086029 (2017).
[3] K. Ogawa et al., Nucl. Fusion 58 044001 (2018).
[4] M. Nakata et al., Plasma Phys. Control. Fusion 58 074008 (2016).
[5] T. Kobayashi et al., Sci. Rep. 9 15913 (2019).Speaker: Hiromi Takahashi (National Institute for Fusion Science) -
Sudden increase of the energy confinement time by 38% is observed during divertor detachment operation with RMP application in deuterium plasmas in LHD. Edge transport barrier (ETB) is formed during the RMP induced H-mode phase at the inner separatrix of magnetic island created by the RMP, leading to steepening of pressure profile while the detachment is maintained. Such behavior has been observed neither without RMP nor in hydrogen discharges to date. During the RMP induced H-mode, MHD activity is detected in magnetic probe, and high charge state impurity emission decreases indicating decontamination of core plasma. Core plasma transport analysis shows clear decrease in the transport coefficient after the detachment transition and in the subsequent RMP induced H-mode phase. The present results have revealed, for the first time, the impact of 3D edge magnetic field structure and isotope effect on core plasma performance during divertor detachment operation that provides important perspective toward integrated performance in future reactors.
Compatibility of good core plasma performance with enhanced edge radiation to mitigate the divertor heat load is a crucial issue for magnetically confined fusion reactors. It is also not clear yet how 3D edge magnetic field structure affects the divertor heat load and the core plasma transport during detachment either in tokamaks or in stellarators. In LHD, stable detachment control is realized with RMP application of m/n=1/1 mode [1]. The present paper reports new observations on core plasma confinement during divertor detachment operation in deuterium and hydrogen plasmas.
The RMP application creates a magnetic island of m/n=1/1 in the edge stochastic layer, where the impurity radiation is enhanced due to increased volume of cold plasma region.Figure 1 shows time traces of plasma parameters with RMP application in deuterium plasma operation. The total NBI heating power of three beam lines was ~5 MW with injection energy of 152~177 keV (3.6MW in co-direction and 1.4 MW in counter-direction, respectively). Density was ramped up with deuterium gas puff and the radiated power increased as well. The detachment transition occurred at t = 3.95 sec as indicated by the reduction of divertor heat flux in Fig.1 (e). It was found that there appears toroidal modulation in the divertor heat load pattern caused by the m/n=1/1 RMP field mode structure. During the detached phase, the density was increased even further beyond the density limit (nSudo, Sudo limit) (Fig.1 (a)). It was observed that plasma stored energy spontaneously increases at t ~ 4.5 sec by 38%, and spikes appears at the magnetic probe signal and divertor heat flux while the detachment is maintained. Such increase of stored energy during detachment is unique to the deuterium plasmas with RMP application. This phenomenon has never been observed without RMP and in hydrogen plasmas up to now.
The RMP induced H-mode is accompanied by the ETB formation with steep edge pressure gradient. Figure 2 shows the radial profiles of the plasma pressure at the edge region for deuterium plasmas with RMP application during the detached phase and the subsequent H-mode phase. Due to the strong edge cooling at the island where $T_e$ < 10 eV, the pressure profile is clamped by the edge island in the detached phase. Clear steepening of pressure profile is observed in the RMP induced H-mode phase (red, 5.1 sec), where the ETB is formed at the inner edge of the magnetic island induced by the RMP.
Figure 3 shows energy confinement time, $\tau_E$, scaled by gyro-Bohm dependence, $\tau_E^{GB} \propto (n_e / P)^{0.6} B^{0.8}$. $\tau_E / \tau_E^{GB}$ monotonically decreases with increasing density. The systematically smaller values in the case with RMP application are due to the shrinkage of the plasma volume caused by the edge magnetic island, while heat transport coefficients, $\chi_{eff}$, obtained by transport analysis, show similar values for both cases with and without RMP. It is observed that during the RMP induced H-mode phase the $\tau_E / \tau_E^{GB}$ increases by 38%.
The core plasma transport has been analyzed with TASK3D code by taking into account the NBI heating profile. Figure 4 shows the results of the deuterium plasma. After the detachment transition, the pressure profile as well as the NBI heating profile become peaked (t = 4.3 sec). The peaked heating profiles are attributed to deeper penetration of NBI in the detached phase with shrinkage of plasma volume due to the strong edge cooling. The transport coefficient decreases at almost entire plasma in the detached phase. In the RMP induced H-mode phase (t = 4.9 sec), the pressure profile becomes further steeper at the edge while the NBI heating profile also slightly peaked. The transport coefficient decreases by a factor of 1.6 to 2 at the edge region, $\rho$ ~ 0.8, which corresponds to the inner separatrix of the island as shown in Fig.2 where the ETB is formed.
[1] M. Kobayashi et al., Nucl. Fusion 59 (2019) 096009.
Speaker: Masahiro Kobayashi -
EAST is typically operated on the radio-frequency (RF) waves heating scenarios, which is also highly related to ITER and CFETR. For the recent experiments of EAST H-mode operations, RF waves are found to be efficient to mitigate and suppress the edge instabilities, such as ELMs [1,2]. From the simulation point of view, the RF heating and driving effects can be studied from 2 aspects: directly on the equilibrium and directly on the instabilities. For the first aspect, the heating effects by RF waves are presented as the increase of the temperatures, and the current driving as the change of the current density profile, then those profiles are considered into the reconstruction of the magnetic equilibrium. In the previous work, the equilibriums including RF effects have already been used for the studies of the ELM behaviors [3,4]. Therefore, this presentation only focuses on the physics understanding for the active control of the edge instabilities by RF waves directly.
It is considered that there are several different mechanisms which are able to achieve the suppressions and mitigations on the edge instabilities by RF waves directly. For example, the low-hybrid wave (LHWs) and ion-cyclotron wave (ICW) show different capabilities on the edge instability control due to their different physics in the plasmas. LHW can excite the forced mode, and the strong nonlinear mode coupling between the forced and spontaneous modes leads to the mitigation of the edge turbulence. For ICW, the sheath potential of the ICW antenna is essential to change the radial electric field and able to suppress the linear growing of ELMs. Therefore, the six-field two-fluid module in BOUT++ framework [5] is extended based on the requirements to simulate the interactions between edge instabilities and RF waves [6].
The LHW is considered to drive the helical filamentary current (HFC) in the SOL region which can change the boundary topology by the radial magnetic field[1]. This radial magnetic field is treated as the initial perturbation in the simulations. The nonlinear evolution of the different modes of Te fluctuations are shown in figure 1 (a) and (b). The most obvious difference is the linear growing phase of the fluctuations which are changed by HCF. Figure 1(b) shows that all the modes of Te grow much more slowly than the case without HCF. Although the radial magnetic field induced by HCF could be much smaller than the perturbed field of the edge turbulence, it is still able to excite the perturbations with n = nHCF to grow up at the start of the linear phase. As shown by the red curve in figure 1(b). This forced mode is effective to compete with the spontaneous fluctuations and changes the spectrum of the eigenmodes even in the linear growing phase. There are strong nonlinear wave–wave interactions happened between those modes, and the phase coherence time gets decreased dramatically [7]. Therefore, none of the modes can grow to its amplitude in the no-HCF case. As shown in figure 1(c), the averaged total fluctuations of all the modes for the HCF case is able to be mitigated by 35% at least. This implies the edge turbulence is able to be mitigated by HCF. The mechanism gives a potential explanation to the active control on ELMs with LHW. This mitigation mechanism by HCF from LHW is quite similar to RMP coils, and could provide the more flexible method compared to the fixed RMP coil systems.
The ELMs are observed fully suppressed by the ICW heating during the H-mode discharge #77741 in EAST, as shown in Fig 2(a). The simulations with BOUT++ reveals that the key factor is the RF sheath on the ICW antenna. The RF sheath potential is able to increase the radial magnetic field inside the SOL and introduces a large flow shear across the separatrix [8]. The linear growth rate can be decreased to less than 0.02 with RF sheath effects in Fig. 2(b), and the ELM size is decreased from over 10% to 0.66%, which are mainly consistent with the experimental results in Fig. 2(c). The large flow shearing rate also enhances the nonlinear wave-wave interactions between the different modes, so the free energy is shared by these modes, which can decrease the energy loss and suppress the ELMs effectively. In the experiments, the probability of the ELM suppression by ICRF is low. In the simulations, it is also found that if the ELM is fully suppressed by ICRF, the requirements for the radial electric field Er is relatively high. A small sheath potential range are found for the good ELM suppression by ICRF, as shown in Fig. 2(d).
As the conclusion, this presentation exhibits two different mechanisms on the mitigation and suppression of the edge instabilities by two different waves. LHW excites the forced mode through HCF, and the strong nonlinear mode coupling between the forced and spontaneous modes leads to the mitigation of the edge turbulence. The sheath potential of the ICW antenna is essential to change the radial electric field and suppress the linear growing of ELMs. Both mechanisms can be explained by the enhancement of the nonlinear wave-wave interactions. Therefore, both LHW and ICW show a potential to be an effective method for the active ELM control besides RMP coils and pellet injections for the future tokamaks, such as CFETR and ITER.
[1] Liang Y.F. et al 2013 Phys. Rev. Lett. 110 115002
[2] Wan B.N. et al 2015 Nucl. Fusion 55 104015
[3] Wu Y.B. et al 2018 Plasma Phys. Control. Fusion 60 055007
[4] Huang Y.Q. et al 2020 Nucl. Fusion 60 026014
[5] Xia T.Y. et al 2013 Nucl. Fusion 53 073009
[6] Xia T.Y. et al 2019 Nucl. Fusion 59 076043
[7] Xi P.W. et al 2014 Phys. Rev. Lett. 112 085001
[8] Gui B. et al 2018 Nucl. Fusion 58 026027Speaker: Tianyang XIa (Institute of Plasma Physics, Chinese Academy of Sciences) -
Highlight of this work: This work predicts the optimal coil phasing, semi-empirical threshold coil current and ‘favorable’ $q_{95}$ window for ELM mitigation for HL-2M 1MA discharge scenario. It is found that pressure gradient may play an important role on determining the peeling-tearing displacement near X-point, due to the curvature effect (GGJ effect) of equilibrium magnetic field.
Resonant magnetic perturbation (RMP) generated by external coils is an effective method to suppress or mitigate edge localized mode (ELM) in H-mode toroidal plasma. Extensive efforts have been devoted to understand the mechanism of controlling ELM. It is demonstrated that edge-peeling response to RMP fields plays an essential role. The linear single fluid model (employed in MARS-F code) predicted results are in good agreement with experiential measurements in many cases $[1]$. MARS-F is widely applied to interpret the experimental observations and to optimize of RMP coil configuration.
This work focuses on the optimization of coil phasing for ELM control for the coming HL-2M tokamak device and on the influence of pressure profile at pedestal region on RMP fields, using MARS-F code. In the computations, plasma rotation and plasma resistivity are included. The former induces the screening of the applied RMP fields, while the latter yields the penetration of field. Moreover, the strong parallel sound wave damping term is also included, which moderately damps the core-kink response. The resonant radial perturbed magnetic field component b$^1_{res}$ at plasma edge or the plasma displacement ($\xi_X$) near the X-point is taken as the indicator to optimize coil phasing here. Actually, these two criteria are basically equivalent $[2]$.
We consider a equilibrium of HL-2M with 1.0 MA current. The key parameters are: major radius $R_0=1.75$ m, minor radius $a=0.65$ m, $B_T=1.8$ T, $q_0=1.07$, $q_{95}=3.25$ and plasma normalized pressure $\beta_N=1.63$ being much smaller than the no wall beta limit ($\beta^{no-wall-limit}_N \ \sim 3.6$). There will be two off-midplane rows of coils. Each row includes 8 coils, which allows the configurations with the maximum toroidal harmonic being $n=1, 2$ and $4$.For HL-2M, the coil basic parameters were already determined, such as the the coil width ($\Delta \theta =15^o$) and radial location ($\theta_c=\pm 40^o$). However, the off-midplane coils have one degree of freedom to choose : the coil phasing $\Delta \Phi$. The current on upper and lower coils is simply expressed as $I_{upper}\propto cos(n\phi)$ and $I_{lower}\propto cos(n\phi+\Delta\Phi)$, respectively. The numerical results indicate that the optimal coil phasing for n=1, 2 and 4 are $\Delta\Phi_{opt}=\pm 180^o , 100^o$ and $-50^o$, respectively. At the optimal phasing $\Delta\Phi_{opt}$, the edge-peeling response is dominant over the so-called core-kink response $[3]$. The maximum of $\xi_X$ for n=1 is about 1.5 and 10 times larger that of n=2 and 4, respectively. While the optimal phasing is not sensitive to the choose of toroidal rotation profile and pressure profile. During the variation of pressure profile, the normalized beta $\beta_N$ and q profile are fixed. More interesting, it is found that the amplitude of $\xi_X$ is generally reduced when the pressure gradient at edge increases. This is likely due to that the pressure gradient (GGJ effect) makes kink-tearing mode more stable. It is implied that the required minimum coil current for suppressing/mitigating ELM is enhanced when plasma pressure profile becomes more sharp at edge.
The comparison between linear response modeling and experiments in MAST $[4]$ yields a critical X-point displacement $\xi_X\sim 1.5 $ mm for achieving ELM mitigation. We simply assume the critical value $\xi_X\sim 2 $ mm as the guideline for controlling ELM on HL-2M, although there are difference in plasma configuration, coil geometry, and the actual threshold coil current between these two machines. In fig.1, the solid white curves represent the 2 mm level of X-point displacement. Clearly, at the designed coil geometry ( $\Delta \theta=15^o$ , $\theta_c=\pm 40^o$ ), the required coil current depends on the choice of coil phasing. With the bad choice of coil phasing (e.g. $0<\Delta \Phi<\sim 50^o$ ), the required coil current exceeds the the allowed maximum RMP coil current (=10 kAt) as designed. On the other hand, there is a wide region of ‘good’ coil phasing, which needs $I_c< 5$ kAt for achieving ELM mitigation based on the 2 mm X-point displacement criterion. Similar study will be carried out for other toroidal mode number.
Usually, the ELM mitigation/suppression is sensitive the $q_{95}$ value $[1]$. We predict the effective $q_{95}$ window for HL-2M as shown in fig.2. For $n=1$ case, the most effective $q_{95}$ window is in 3.1<$q_{95}$<3.2, in which the maximum (e.g. at the optimal coil phasing) of b$^1_{res}$ amplitude is about 8 times larger than that outside of this window. For $n=2$ case, the best window exists near $q_{95}~3$. Another ‘favorable’ $q_{95}$ window is 3.4<$q_{95}$<3.5. It is noted that the optimal coil phasing is not sensitive to the variation of $q_{95}$ for the studied equilibrium. Here, during scan of $q_{95}$, $\beta_N=1.63$ is fixed.
References
$[1]$ Y.Q. Liu, et.al, Phy. Plasmas 24, 056111, (2017);
$[2]$ L.N. Zhou et al. Nucl. Fusion, 58,076025, (2018);
$[3]$ Y.Q.Liu et.al. Nucl. Fusion 51, 083002, (2011);
$[4]$ A. Kirk et al. Nucl. Fusion, 55,043011, (2015)Speaker: Dr G.Z. Hao (Southwestern Institute of physics) -
Transition between isotope-mixing and non-mixing states in hydrogen-deuterium mixture plasmas is observed for the first time in the world in the isotope (hydrogen and deuterium) mixture plasma in Large Helical Device. In the non-mixing state, the isotope density ratio profile is un-uniform when the beam fueling isotope species differs from the recycling isotope species and the profile varies significantly depending on the ratio of the recycling isotope species, although the electron density profile shape is unchanged. The fast transition from non-mixing state to isotope-mixing state (nearly uniform profile of isotope ion density ratio) is observed associated with the change of electron density profile from peaked to hollow profile by the pellet injection near the plasma periphery. The transition from non-mixing to isotope-mixing state strongly correlates with the increase of turbulence measurements and the transition of turbulence state from TEM to ITG is predicted by gyrokinetic simulation.
Bulk charge exchange spectroscopy[A] has been applied to measure the radial profiles of hydrogen (H) and deuterium (D) density in the plasma from H$_\alpha$ and D$_\alpha$ lines emitted by the charge exchange reaction between the bulk ions and the neutral beam injected in Large Helical Device (LHD). Figure 1 shows radial profiles of electron density normalized by the line-averaged electron density and H and D density in the plasma with H beam fueling but without gas puff. Electron density and the ratio of H recycling increase shot by shot due to the H beam fueling in the previous shot. The electron density profile shapes are almost identical for these three discharges with different line-averaged density and different wall recycling isotope ratio. However, radial profile shapes of H and D density are quite different depending on the ratio of H recycling. The amount of H density increases as the H recycling is increased, although the amount of D density is similar for these three discharges. When the isotope recycling ratio is close to unity ($\Gamma_H/\Gamma_D$ = 0.8), there is almost no difference in profiles between H and D density as seen in Fig.1(b). In contrast, a significant difference in the profile shape (peaked or hollow) between H density and D density is observed in the lower density plasma where the H recycling is low enough ($\Gamma_H/\Gamma_D$ = 0.3), as seen in Fig.1(d).
The transition from non-mixing state to isotope-mixing state is observed after H and D pellet injections[B]. Because of the relatively shallow pellet deposition, pellet injections make the electron density profile more hollow. Before the pellet injection the H density profile is much more peaked than the D density profile due to the H beam fueling and D dominant recycling. After the pellet injection, the H density profile becomes similar in shape to D density profile regardless of the species of pellet as seen in Fig.2. Therefore, the flattening of H fraction profile both for the H and the D pellet is a clear evidence for isotope-mixing. If plasma is non-mixing state, the H fraction profile should be more peaked after the D pellet injection because of the edge pellet deposition. The transition from non-mixing state to isotope-mixing state occurs in a time scale shorter than the global particle confinement time (less than $\sim$15 ms), which implies the large ion diffusion coefficient in the isotope-mixing state.
Figure 3(a)(b) shows the electron density profile and, the density fluctuation spectrum integrated from edge to core along the laser beam line of the central chord of phase contrast imaging (PCI) for non-mixing and isotope-mixing states. As seen in Fig.3(b), the turbulence level increases by an order of magnitude in the isotope-mixing state. Electron temperature and its normalized gradient decreases significantly but ion temperature decreases slightly. The ratio of electron temperature to ion temperature ($T_e/T_i$ ratio) also decreases. Figure 3(c) shows the linear growth rates calculated with gyrokinetic simulation code GKV[C] for TEM and ITG turbulence, based on the radial profile of the density and the temperature measured. The non-mixing state is observed in the low-density plasmas with electron cyclotron heating (ECH) and neutral beam injection (NBI), where the beam fueling isotope species differ from the isotope species due to recycling. After the pellet injection, the isotope-mixing state is observed in higher density plasmas.
When the sign of density gradient changes from negative (peaked) to positive (hollow), the growth rates of both TEM and ITG decrease. The gyrokinetic simulation predicts that TEM propagating in the electron diamagnetic direction is unstable for the non-mixing state. However, the TEM is stabilized and ITG mode propagating in the ion-diamagnetic direction becomes unstable for the isotope-mixing state. In this experiment, the ratio of ion diffusion to electron diffusion coefficient, $D_i/D_e$, defined as $ - [ \Gamma_H/(\partial n_H/\partial r) + \Gamma_D/(\partial n_D/\partial r)] / [2\Gamma_e/(\partial n_e/\partial r) ]$ is evaluated from the quasilinear approximation in the gyrokinetic calculations, where the ambipolar condition of $\Gamma_H + \Gamma_D + Z_C \Gamma_C - \Gamma_e = 0$ holds in the simulation. The $D_i/D_e$ is 0.4 for the case of the TEM dominant state before the pellet injection, while $D_i/D_e$ is 2.5 for the case of the ITG-dominant state after the pellet injection. These simulation results are consistent with the non-linear GKW simulation results[D] where the isotope mixing is predicted to occur when ITG is dominant ($D_i > D_e$) and not to occur when TEM is dominant ($D_i < D_e$).
The fast transition between mixing and non-mixing state of isotope observed in this experiment stimulates the future development of non-linear global gyrokinetic simulation for multiple ion species, which is numerically challenging work. This results demonstrate that non-mixing and the isotope-mixing states depends on the turbulence state and give the important knowledge for predicting the isotope density profiles in the D-T mixture plasma in JET and ITER.
[A] K.Ida, et.al., Rev. Sci. Instrum. 90, (2019) 093503.
[B] K.Ida, et.al., Phys. Rev. Lett. 124, (2020) 025002.
[C] M.Nakata, et.al., Phys. Rev. Lett. 118, (2017) 0165002.
[D] C.Bourdelle, et. al., Nucl. Fusion 58, (2018) 076028
Speaker: Katsumi Ida (National Institute for Fusion Science) -
The feature of radiation collapse has been extracted from high-density NBI heated experiments in LHD using sparse modeling technique. Accessible density limited by radiation collapse has been evaluated using extracted features. The extracted feature is described by the combination of plasma parameters such as line averaged electron density $\bar{n}_\mathrm{e}$, impurity line emission intensity of OVI and FeXVI, and electron temperature $T_\mathrm{e,center}$. This expression of the feature enables detection of radiation collapse before its occurrence and identification of operational density regime with likelihood of radiation collapse much more practical than the Sudo scaling[1]. It is to contribute to the control of stellarator-heliotron plasma in the high density regime. Moreover, the extracted dependency of the electron temperature has been found to match the characteristics of the cooling rate of oxygen impurity, which suggests the key to radiation collapse.
In stellarator-heliotron plasma, radiation collapse is one of the most critical issues that limit performance of plasmas, while stable high-density operation is an advantage of a helical system over a tokamak. The Sudo scaling $n_\mathrm{e}^{\mathrm{Sudo}}=0.25\left(PB/a^{2}R\right)^{0.5}$[1] is well known as the scaling law of density limit, which suggests that the balance between heating power and radiated power loss is a key together with robust confinement capability such as plasma volume and magnetic field[2]. However, while other operational conditions such as wall condition and impurity concentration are thought to be related to radiation collapse, those conditions remain tacit knowledge and those effects are hidden behind the Sudo scaling.
In the present research, the support vector machine has been used to construct the classifier that distinguishes plasmas in the state where the radiation collapse is likely to occur and in the stable state. The dataset is based on experiment data using deuterium gas-puff plasma in LHD. The magnetic configuration is fixed at the magnetic axis position $R_\mathrm{ax}$ of $3.6\, \mathrm{m}$ with $B= 2.75\,\mathrm{T}$. The surveyed line averaged density and heating power range up to $1.1\times10^{20}\,\mathrm{m^{-3}}$ and $15\,\mathrm{MW}$, respectively. The experiment data was labeled according to density exponent $x=(\dot{P}_\mathrm{rad}/P_\mathrm{rad})/(\dot{\bar{n}}_\mathrm{e}/\bar{n}_\mathrm{e})$, which is a criterion derived from the relationship between radiation power and plasma density[3]. Using the constructed classifier, plasma parameters that are related to radiation collapse have been extracted using Exhaustive Search (ES)[4]. The ES is one of the sparse modeling methods that automatically extract features hidden in high-dimensional data.
The dataset consists of 16 candidate parameters such as $\bar{n}_\mathrm{e}$, $P_\mathrm{in}$, $P_\mathrm{rad}$, $\beta$, impurity line emission intensities, and electron temperature at plasma center $T_\mathrm{e,center}$. $T_\mathrm{e,center}$ is used as typical electron temperature in the present study.
The boundary between the state of plasma where the radiation collapse is likely to occur and the stable state is expressed with the plasma parameters extracted by the ES as$\mathrm{C}\bar{n}_\mathrm{e}^{1.3}\mathrm{OVI}^{2.1}\mathrm{FeXVI}^{0.69}T_\mathrm{e,center}^{-3.9}=1.$
Note that $\mathrm{C}$ is a normalization factor and the intensities of OVI and FeXVI were normalized by $\bar{n}_\mathrm{e}$. $\bar{n}_\mathrm{e}$ and $T_\mathrm{e,center}$ are used in $10^{19} m^{-3}$ and $keV$, respectively. Then the likelihood of the radiation collapse has been evaluated by corresponding the distance from the boundary to the fitting of the frequency of the occurrence by sigmoid function. The likelihood was compared with Sudo scaling in Fig.1. In Fig. 1, the data in the stable state and the state of plasma undergoing the radiation collapse were plotted by dots and crosses, respectively. The distributions of those two states are overlapping in Sudo scaling but separated better in likelihood. By fixing threshold likelihood, eq.1 can be deformed into the density limit expression, e.g., $\bar{n}_\mathrm{e}^{50\%}=0.014\mathrm{OVI}^{-1.6}\mathrm{FeXVI}^{-0.54}T_\mathrm{e,center}^{3.0}$ for 50% likelihood. This is a practical density limit for the control of LHD plasma in the high density regime.
Figure 2 shows a typical discharge with radiation collapse in LHD. In the panel (a), the density exponent and collapse likelihood are shown in blue and red lines, respectively. In this discharge, collapse likelihood increased before density exponent started to change and reached unity in prior to the collapse by about $100\,\mathrm{ms}$. On average, collapse likelihood reaches 80% in prior to the collapse by about $80\,\mathrm{ms}$.
According to equation 1, it was seen that the $T_\mathrm{e,center}^{-3.9}$ is an index defining the boundary between the state of plasma where the radiation collapse is likely to occur and the stable state. Here, it is known that the radiation power due to impurities is a source of radiation loss from the plasma and it depends on impurity amount and cooling rate $L_z$ of impurity ion species[2]. The cooling rate is a function of electron temperature and the dependency differs by species. Figure 3 shows that the cooling rate of oxygen partly has a relationship that is similar to the proportion to $T_\mathrm{e}^{-3.9}$, which agrees with the identified feature in equation 1. Therefore, radiation loss caused by oxygen is expected to be major cause of radiation collapse in this dataset.
References
1. S. Sudo et al. Nuclear Fusion, 30(1):11–21, 1990.
2. K. Itoh, S.-I. Itoh, L. Giannone. Journal of the Physical Society of Japan, 70(11):3274–3284, 2001.
3. B. J. Peterson et al. Plasma and Fusion Research, 1:045–045, 2006.
4. Y. Igarashi et al. Journal of Physics: Conference Series, 1036:012001, 2018.Speaker: Mr Tatsuya Yokoyama (Graduate School of Frontier Sciences, The University of Tokyo) -
In the Large Helical Device (LHD), deuterium (D) plasma experiment has been conducted since 2017. This is the first opportunity in stellarator / heliotron devices to reveal the transport of tritium (T) generated by DD fusion reactions. Dedicated analyses of the remaining T in the LHD vacuum vessel have been conducted, and results of the analyses suggest that T remains densely in divertor tiles and sparsely in the first wall. Furthermore, an asymmetry was observed in remaining amounts of T in divertor tiles located at symmetric positions. This asymmetric distribution is consistent with the distribution of high-energy triton loss positions on the vacuum vessel calculated by a Lorentz orbit following code (LORBIT) [1]. A depth profile of remaining T in a baffle part of divertor tile, on which no divertor plasma struck, shows that the peak of the profile locates several micrometers from the surface, which is much deeper than the range of thermalized T. These results suggest the origin of the remaining T in divertor is mainly high energy triton.
Figure 1 shows the top view of four divertor tile arrays in the helical divertor in a half part of LHD. Yellow dots in Fig. 1 are lost positions of high energy triton calculated by using LORBIT code for the standard operational magnetic configuration with the counter-clockwise (CCW) toroidal magnetic field ($B_{t}$) direction. It is clearly shown in Fig. 1 that high energy tritons are lost to divertor region asymmetrically. In this case, the lost positions are only at red colored tile arrays. In the case of the clockwise (CW) $B_{t}$ case which is not shown here, lost positions are only at blue colored tile arrays. B × grad B drift dominates the orbit loss of high energy triton, and thus the asymmetry depends on the $B_{t}$ direction.
Remaining T in divertor tiles made of graphite was analyzed by using the tritium imaging plate technique (TIPT) [2], which can measure the amount of remaining T within a depth shallower than the escaping depth of $\beta$-rays from T decay (~ 1 $\mu$m for graphite) as the photo-stimulated luminescence (PSL) intensity. Figure 2 shows two divertor tile arrays, which are shown as arrays filled with blue and red in Fig. 1, and positions of retrieved divertor tiles to analyze.
Figures 3(a) - (d) show areal and exposure-time averaged PSL intensities on divertor tiles. These figures clearly indicate that amounts of remaining T near surface is asymmetric, and larger amounts of remaining T are on tiles in R array rather than in L array. In the D experimental campaign in 2017, experimental dates with CCW and with CW $B_{t}$ were 54 days and 12 days, respectively. Therefore, the asymmetry shown in Figs. 3 (a)-(d) is consistent with the above mentioned asymmetry in the high energy triton loss. Relatively large PSL intensities were observed on baffle part tiles at the inboard-side divertor shown as red bars in Fig. 3(b), on which no divertor strike point comes, and the asymmetry is most evident in the baffle part tiles between L (blue bar) and R arrays in Figs. 3(a) and (b). In the cases of divertor tiles struck by divertor plasma (light blue and orange bars), temperature rises and isotope exchanges can cause the smaller retention of T and the asymmetry.
Figure 4 shows a depth profile of the PSL intensity in a baffle part tile (I-3R), which was obtained by using a new method, which uses a sputtering treatment and TIPT. For the short escaping depth of $\beta$-rays from T decay, the profile of remaining T is similar to this profile. Because the depth of the peak of the profile is much deeper than the normal incident hydrogen (H) with energy of 100 keV, the origin of remaining T is considered to be high energy triton with various incident angles.
Analyses of remaining T in the first wall made of stainless steel (SUS316L), and long-term material probes made of SUS316L and tungsten, which had been installed on the first wall, were conducted using TIPT and thermal desorption methods. The averaged PSL intensities of the material probes at deposition dominant (mainly carbon) region near divertor tiles, and erosion dominant region are the order of $10^{-2}$, and $10^{-3}$ [$mm^{2}$/h] or less, respectively. Results of thermal desorption analyses show that remaining T in material probes are up to the order of MBq/$m^{2}$ at the deposition dominant region, and remaining T in divertor tiles are the order of magnitude larger than that in material probes.
A result of calculation using LORBIT code shows that approximately 40 % of generated triton by DD reactions is lost without collisions at divertor [1]. Furthermore, a result of exhausted gas analysis shows approximately 3.9 GBq, which is 60 % of total generated triton, still remained in LHD [3]. From these results, approximately 66 % of remaining T in LHD can be in divertor tiles and the first wall near divertor tiles. The above mentioned experimental and calculated results suggest that T remains densely in divertor tiles and sparsely in the first wall.[1] K. Ogawa et al., Plasma Sci. Technol. 21 (2019) 025102.
[2] T. Tanabe, ”Tritium: Fuel of Fusion Reactors”, (2017) Springer.
[3] M. Tanaka et al., 27th IAEA Fusion Energy Conference (2018) FIP_P1-7.Speaker: Dr Suguru Masuzaki (National Institute for Fusion Science) -
Recent improvements of the EMC3-EIRENE code have allowed to assess for the first time the
detached divertor scenario foreseen for ITER during ELM suppression by resonant magnetic
perturbation (RMP) fields. This is a major breakthrough because ELM suppression is required
for ITER to maintain the integrity of the plasma wall interface. The ITER divertor has been
designed based on extensive 2D (axisymmetric) simulations, but whether the 3D (nonaxisymmetric)
boundary from RMP application remains compatible with divertor operation in a
dissipative, partially detached state remains unknown. New EMC3-EIRENE results show that
detachment transition with RMPs occurs at lower upstream density within the traditional strike
zone of the symmetric configuration (see figure 1). At the same time, however, nonaxisymmetric
strike locations with a magnetic connection into the bulk plasma appear further
outside, and those remain attached - even at higher upstream densities when the symmetric
configuration is already (partially) detached. Neon seeding can mitigate those non-axisymmetric
heat loads by about 30% for an average impurity concentration of 1% at the separatrix.
Fundamental for the extended application range of EMC3-EIRENE has been the numerical
stabilization of the iterative solver by linearization of the electron energy loss term [a].
Furthermore, volumetric electron-ion recombination is now activated in EMC3-EIRENE along
with neutral-neutral collisions (in BGK approximation) and molecular assisted recombination.
Simulations for the ITER Pre-Fusion Plasma Operation (30 MW) show that detachment
transition occurs at lower upstream density within the traditional strike zone of the symmetric
configuration while non-axisymmetric strike locations further outwards remain attached [b].
Figure 1 shows the correlation between the magnetic footprint (a) and the heat load pattern (b) on
the outer divertor target. A reference simulation with comparable upstream density of nup ≈ 1.7 ·
1019 m-3 (evaluated at the midplane position of the separatrix of the symmetric configuration) is
included in 1 (c). The same amount of power as in the symmetric reference configuration is now
distributed over perturbed field lines connecting the bulk plasma to the target (red areas in figure
1(a)). Not only is the upstream heat flux at strike point A reduced (green box in Fig. 1.c), more
energy is lost to neutral gas and dissipated through cross-field diffusion here compared to the
reference (magenta box). Evaluation of the upstream-downstream pressure balance confirms the
earlier onset of detachment here.
On the other hand, no pressure and power losses are found at strike point B (blue box) which
remains attached at temperatures well above 10 eV. Simulations with Ne seeding have been
performed to explore a possible mitigation strategy in anticipation of the Fusion Plasma
Operation phase at 100 MW. It can be seen in figure 1 (c) that Ne seeding can indeed mitigate
the non-axisymmetric heat loads, but it is more efficient within the traditional strike zone
(approximately 0 – 15 cm from the separatrix of the symmetric configuration). Nevertheless, a
heat load reduction of about 30 % can still be achieved at strike point B with impurity
concentrations of 1 % averaged along the former separatrix.
Plasma response is key for reliable predictions of
divertor heat and particle fluxes. The present
simulations are based on MARS-F results [c] within a
single fluid, linearized resistive magneto-hydrodynamic
model. The relative phasing between coil rows of the
externally applied perturbation field can be optimized
for ELM control based on the X-point displacement
caused by the edge-peeling component of the plasma
response, but this is found to be correlated with a
relatively large magnetic footprint on the divertor
targets. Despite screening of resonances in the bulk
plasma, field amplification near the separatrix is found
and this determines the magnetic footprint size. It is
possible for the magnetic footprint to extend beyond its
size in the vacuum perturbation field approximation,
and it can be seen in figure 2 that it can even extend
beyond the dedicated high heat flux region on the
divertor targets (dashed line) under certain assumptions
related to plasma rotation. This can bring high heat
loads to those locations, and it is found that Neon seeding is significantly less effective under
these conditions.
Acknowledgments: This work was supported by the U.S. Department of Energy under grants DE-SC0013911, DESC0020357
and DE-SC0020284, by the College of Engineering at the University of Wisconsin - Madison, and by
the ITER Scientist Fellow Network.
[a] H. Frerichs et al., Nuclear Materials and Energy 18 (2019) 62–66
[b] H. Frerichs et al., accepted in Physical Review Letters (2020)
[c] L. Li et al., Nuclear Fusion 59 (2019) 096038Speaker: Heinke Frerichs (University of Wisconsin - Madison) -
This paper first studies the coupling effects between the impurities and the background plasma under BOUT++ framework by extending the six-field two-fluid MHD model 1. The extended model is applied to study the effect of pedestal Li impurity on ELMs during real-time Li powder injection. The extended BOUT++ MHD model is first validated by the good agreement between analytical results and numerical ones over a linear model. The obtained results can help deepen understanding the physical mechanism of ELM suppression resulting from impurity. The validated extended MHD model is then being applied to simulate the ELM evolution in its nonlinear phase.
During the operation in the current tokamak devices, some amount of various types of impurities inevitably penetrates the separatrix and enters the core plasma region, then changes the background plasma, and subsequently affects the MHD instabilities. Therefore, it is necessary and important to understand the physical mechanism behind by constructing reliable models and developing corresponding numerical codes. To couple the effect of impurity equilibrium on the background plasma into the BOUT++ MHD model, we redefine the vorticity by introducing the contribution of impurity to the vertical component of ion velocity and modify the gyro-viscous term by integrating the impurity contribution into the vorticity equation. Using this extended model, we take only diamagnetic drift into account for deriving analytical dispersion relation of strong peeling-ballooning case, and take into account the contributions of diamagnetic drift, resistivity, gyro-viscosity, and the effect of cross term for numerical simulation.
Based on the obtained dispersion relation, we can know that the increment in the plasma mass density due to the Li ions present in the pedestal region not only leads to a stabilizing effect on the ideal peeling-ballooning mode, but also weakens the ion diamagnetic stabilization, an indirectly destabilizing effect on the instability. These two opposite effects compete with each other, as shown in Fig. 1: when the diamagnetic frequency is close to the ideal instability growth rate, the weakening effect of impurity on the ion diamagnetic stabilization becomes dominant; when the diamagnetic frequency is much less than the ideal instability growth rate, the stabilizing effect of impurity on the ideal peeling-ballooning mode is dominant. After further considering the reducing temperature due to the existence of pedestal impurity, the background temperature drops. This cooling effect helps stabilize the peeling-ballooning mode more effectively through reducing the diamagnetic frequency and the ideal instability growth rate by a same percentage.
To understand how the impurity affects the growth rate of individual mode, we present the corresponding numerical results in Figs. 2 and 3. As can be seen, the simulation results have a good agreement with the analytical results, which validates the reliability of this extended MHD model. Note that, in the case of strong peeling-ballooning mode, the effect of resistivity is negligibly small. Based on this finding, we further consider the effect of gyro-viscosity and the cross term in the vorticity definition, respectively, and carry out simulations. The result shows that the pedestal Li ions have a negligible effect on strong peeling-ballooning mode through the additional gyro-viscos term and cross term of impurity in the evolved vorticity equation.
Based on the extended model and the aforementioned findings, we further advance understanding the physical mechanism for ELM suppression by impurities, which originates from the ELM suppression mechanism proposed in 2 for assessing the Li wall conditioning effects on ELMs. The mechanism is shown schematically in Fig. 4. With the Li conditioned wall, some amount of Li species inevitably come into the scrape-off-layer region and some of them finally enters the plasma pedestal. These pedestal Li ions then stabilize peeling-ballooning mode to some degree, as the ion diamagnetic stabilizing effect is not strong enough to stabilize the ideal peeling-ballooning mode completely. In addition, the recycling neutral particles reduces from the PFCs with Li conditioning wall, leading to the decrease in the density at the pedestal base and consequently, the enhancement in the ion diamagnetic stabilizing effect, which can further stabilize peeling-ballooning mode 2. It is found that either the pedestal Li ions or the reduced edge recycling, or the combination of them reduces the ELMs amplitude. In addition, as the ELMs amplitude goes down, the transient heat flux subsequently reduces，The reduction of transient heat flux leads to the decrement of the neutral recycling at the wall and target plates. Thus, a positive feedback loop takes place, and maintains until the density at the pedestal base approaches the threshold value for stabilizing peeling-ballooning mode completely 2.
The remaining work of this paper is to carry out the nonlinear simulation of strong peeling ballooning mode case. This work is on the way now and the new results will be reported on the coming conference.
Reference:
1 T. Y. Xia, et al, “Six-field two-fluid simulations of peeling–ballooning modes using BOUT++”, Nucl. Fusion 53 (2013) 073009
2 Y. Ye, et al, “Experimental study on low recycling no-ELM high confinement mode in EAST”, Nucl. Fusion 59 (2019) 086044.Speaker: Prof. Sun Jizhong (Key Laboratory of Materials Modification by Laser, Ion, and Electron Beams (Ministry of Education), Dalian University of Technology, Dalian 116024, China) -
Tungsten (W)-coated divertor tiles were installed in the wide area of the closed divertor of one inner-toroidal section out of the ten toroidal sections in the Large Helical Device (LHD). The visible and extreme ultraviolet (EUV) spectroscopic measurements show the different dependence of emission intensity of neutral W atoms and highly ionized W ions (W$^{24+}$~W$^{29+}$) on the line-averaged electron density in NBI plasmas. The different dependence suggests that an intrinsic mechanism to reduce core W concentration is observed in helical systems. The suppression of the carbon flakes, which is generated by the exfoliation of a deposition layer formed by eroded carbon divertor tiles is confirmed at the W-coated divertor section. A carbon flake often causes undesirable impact on the particle control in a long pulse discharge in LHD. The results suggest that W-coating is effective for the particle control. The surface analysis shows that the coated W at the strike point was eroded on some divertor tiles. Based on the analysis of divertor plasma parameters and spectroscopy data, the erosion is considered due to carbon bombardment originating from graphite divertor tiles remaining in the torus. The results suggest that understanding of role of low-Z impurity is important to predict a lifetime of W divertor in future devices, where the low to medium Z impurity exist as auxiliary impurity to aid divertor heat load mitigation.
Tungsten (W) has been selected as the divertor material of ITER. It is an important subject to understand the impurity transport of W in fusion plasmas, and consequently to control the W flux into the main plasma. Also, the investigation of the effect on W material on wall retention of fuel particles is important for the establishment of stable particle control. In the previous study, we installed three VPS-W tiles to investigate the effects of W tiles on the LHD plasma and plasma-wall interactions $[1]$. However, the effect was not clear due to the limited installation. In order to investigate the effect more clearly, we have installed W-coated divertor tiles in the wide area of the closed divertor in LHD.
As shown in Figure 1, 10 $\mu$m of W coating on the isotropic graphite tiles (IG-430U, Toyo-tanso) by magnetron sputtering was selected for this study. Molybdenum (Mo) interlayer of 2 $\mu$m is introduced between W and graphite tiles. The W-coated tiles were installed in one inboard side closed divertor section out of the ten toroidal sections, where the relative high heat flux is reached. As shown in Figure 1, 56 closed divertor tiles and 32 sets of dome structure tiles in one toroidal section of the closed divertor system were coated. The high heat flux test facility ACT2 using electron beam experiment shows that W-coating survived without delamination up to 12.5 MW$^2$ for 3 seconds.
Divertor visible spectroscopy was conducted to investigate plasma-surface interaction in the W-divertor region. Also, EUV spectroscopy was conducted to observe behaviours of tungsten impurity ions in the core plasma region. Figure 2 shows density dependence of source and core concentration of W in NBI plasmas. The port-through power of the NBI was 7.8 MW for all discharges shown in the figure. W source deduced from WI intensity by the visible spectroscopy decreases with electron density. On the other hand, core concentration of W deduced from the unresolved transition array (UTA) spectrum of W$^{24+}$~W$^{29+}$ integrated over 50-51 A$^{\rm o}$ measured by EUV spectroscopy increases with the electron density and peaked at $3 \times 10^{19}$ m$^{-3}$. Then the intensity decreases in further high density ($4\times 10^{19}$ m$^{-3}$). Although the UTA intensity strongly depends on the electron temperature, the slight reduction of the electron temperature from 2.3 keV to 2.1 keV accompanying the change of the electron density from $3\times 10^{19}$ m$^{-3}$ and $4\times 10^{19}$ m$^{-3}$ is less likely to be the reason for the significant reduction of the UTA intensity. The results suggest that the W transport is changed by the electron density and the impurity screening effect by the ergodic layer may be one of the candidates for an intrinsic mechanism to reduce the core W concentration. The impurity screening of high Z impurity such as iron in the ergodic layer has been observed in previous study in LHD $[2]$.
Figures 3(a)-(c) show depth profiles of W, Mo, and C on three positions on a divertor tile obtained by using a glow discharge optical emission spectroscopy (GD-OES). On the surface of the strike point only carbon is seen due to the erosion of coated-W (Fig. 3(a)). Near the strike point, W remains on the surface (Fig. 3(b)). However, the partial erosion of W is observed. On the other hand, in the area far from the strike point, W is not eroded and the carbon is deposited on the W-coated surface (Fig. 3(c)), indicating that the carbon presumably comes from other locations. Figure 4 shows the neutral W intensity as a function of nominal impact energy of carbon ions estimated from divertor probe. The sputtering yield of W from carbon impurity qualitatively coincides with the dependence of the intensity, indicating that carbon impurities can be a cause for the sputtering of W. If the full W-coated tiles are installed in future, the sputtering of W will be significantly reduced. These results demonstrate the plasma-wall interactions of W from low-Z impurities such as carbon and provide the important understanding of the W erosion and transport in the magnetic fusion devices.
$[1]$ M. Tokitani et al., Nucl. Mater. Energy 18 (2019) 23.
$[2]$ S. Morita et al., Nucl. Fusion 53 (2013) 093017.Speaker: Dr Gen Motojima (National Institute for Fusion Science) -
In addition to the existing hybrid electromagnetic (EM) scheme that uses gyrokinetic ions and fluid electrons (Ref. 1), two new EM schemes, which use kinetic electrons, have been implemented in the whole-volume particle-in-cell (PIC) code XGC (Ref. 2). In particular, these new EM schemes in XGC will allow higher-fidelity global simulation of long-wavelength modes, which are important for studying the onset of edge localized modes (ELMs) and their interaction with micro-turbulence and neoclassical dynamics, and also ELM control, via resonant magnetic perturbations (RMPs) and other means. Additionally, EM effects for XGC-focused ITER problems, such as divertor heat-load width, L-H and H-L transitions, pedestal height and shape, and RMP penetration, should include long-wavelength modes self-consistently with other physics effects. Traditionally, the PIC approach exhibited higher numerical sensitivity to the so-called "cancellation problem" (Ref. 3) at long wavelengths than the continuum approach.
The first new kinetic-electron EM scheme that has been implemented in XGC is a fully implicit numerical scheme (Ref. 4) that uses the symplectic formulation of gyrokinetics. The second new scheme that has been implemented is an explicit numerical scheme (Ref. 5) that uses the Hamiltonian formulation of gyrokinetics. Both the implicit and explicit numerical schemes have significant advantages over each other that make both viable for consideration, depending on the particular physics of interest and computational architecture. For example, the implicit scheme completely avoids the numerical cancellation problem that has to be mitigated in the explicit scheme (Ref. 5), whereas, the explicit scheme has a simpler numerical implementation and better performance in general. Our ability to rigorously evaluate these two numerical schemes is, therefore, an important aspect of this work.
The verification of the hybrid EM scheme in XGC was presented in Ref. 1. The hybrid EM scheme, which does not suffer from the cancellation problem, showed that kinetic ballooning modes (KBMs) in NSTX are generally stable in the pedestal region, but could be unstable at the pedestal top, and that the low-$n$ peeling modes generally exist in the steep gradient region of DIII-D and KSTAR pedestals, without a sharp instability boundary, which are contrary to the usual MHD results. We see gyrokinetic peeling modes at low $n<15$ and electromagnetic ITG modes at higher $n>15$, as shown in Figure 1. It appears that the gyrokinetic peeling modes require interaction with mean $E\times B$ shearing and microturbulence to be nonlinearly unstable.
The verification of the new implicit and explicit numerical schemes has been performed for shear-Alfvén modes, as well as the ion temperature gradient (ITG) to KBM transition at finite plasma $\beta$ (Ref. 6). Verification results for shear-Alfvén modes with the implicit scheme are shown in Figure 2. Performing such verification simulations with the explicit scheme would be particularly challenging because of the cancellation problem. Verification results for the ITG-KBM transition for the explicit Hamiltonian scheme are shown in Figure 3, with the cancellation problem successfully mitigated. The stabilization of the ITG instability from 0% to 1.5% plasma $\beta$ is reproduced, and there is good quantitative agreement with the GENE code (Ref. 6). The ITG to KBM transition is also reproduced, and there is again good quantitative agreement with the GENE code for the destabilization of the KBM above 1.5% plasma $\beta$.
Some results from electromagnetic studies of critical ITER edge physics, such as divertor heat-load width, L-H transition, pedestal structure, and ELM onset and RMP control, will be presented.
References:
- R. Hager, J. Lang, C. S. Chang, S. Ku, Y. Chen, S. E. Parker, and M. F. Adams, Phys. Plasmas 24, 054508 (2017);
- S. Ku, C. S. Chang, R. Hager, R. M. Churchill, G. R. Tynan, I. Cziegler, M. Greenwald, J. Hughes, S. E. Parker, M. F. Adams, E. D’Azevedo, and P. Worley, Phys. Plasmas 25, 056107 (2018);
- Y. Chen and S. Parker, Phys. Plasmas 8, 441 (2001);
- G. Chen and L. Chacón, Comput. Phys. Commun. 197 (2015) 73–87;
- A. Mishchenko, R. Hatzky, and A. Könies, Phys. Plasmas 11, 5480 (2004);
- T. Görler, N. Tronko, W. A. Hornsby, A. Bottino, R. Kleiber, C. Norscini, V. Grandgirard, F. Jenko, and E. Sonnendrücker, Phys. Plasmas 23, 072503 (2016).
Speaker: Amil Y. Sharma -
Edge localised mode (ELM) instabilities are a concern for future tokamak operations due to the predicted high heat loads and excessive erosion to the first wall/divertor [1]. A solution to address this focuses on divertor design; the ‘Super-X’ magnetic configuration is designed to alleviate high heat loads and will be tested on MAST-U [2]. Plasma detachment has been predicted in the MAST-U Super-X for L-mode [3] and H-mode [4] plasmas but behaviour during ELMy H-mode is unknown. The work presented here uses simulations to address this by investigating the extent of ELM burn-through and recovery times. Questions of ELM burn-through can only be answered with confidence when MAST-U starts operation, but until then simulations can provide useful guidance. The simulations are first of a kind and are numerically and computationally very challenging.
Simulations are performed using the nonlinear MHD code JOREK [5]. In JOREK the reduced MHD two temperature model has been merged with the neutrals model [6] and developed for divertor recycling due to the increased plasma-neutral interactions in the Super-X. The neutral fluid is described diffusively, the plasma flux incident on the tokamak wall/ divertor targets is reflected away as neutrals, which diffuse into the domain and become ionised at higher temperatures. The flux of plasma at the divertor targets increases target neutral density and an ionization front is established.
This simple model includes only deuterium atoms, the ionization and recombination processes along with radiative processes with energy losses through line and bremsstrahlung radiation. Despite the exclusion of detailed atomic processes like charge exchange, molecular dissociation and impurities required for advanced detachment studies, simulations appear to produce an ionisation front detached from the target, which is sufficient as an initial condition for preliminary studies of ELM burn-through.
It is first important to demonstrate detachment, or to at least obtain a high neutral density in the divertor so an upstream density scan has been performed for a MAST-U Super-X case. As the upstream density is increased (Fig. 1a) the target electron temperature reduces to a few eV, when the target electron Fig. 1a) Target density flux and electron temperature temperature decreases below 5 eV a roll-over in the target density as a function of upstream density, comparison of flux is also observed. The JOREK roll-over is significantly JOREK to SOLPS. b) Ionization in the lower divertor as a function of upstream density, shallower in a comparison to SOLPS for the same Super-X corresponding to each point in a). scenario. Fig. 1b) shows the ionization front, which moves upstream, as the roll-over occurs. A roll-over has been demonstrated for the Super-X case with expanded flux in the divertor, shown here, and for a case without expanded flux, which is used for the ELM burn-through simulations.
The ELM burn-through simulations are initiated from a divertor configuration after the roll-over where single or multiple toroidal mode numbers are included in each simulation. The energy of the modes grows as the equilibrium is unstable to peeling-ballooning modes and an ELM crash is observed. In each simulation the ELM crash results in plasma burn-through in the Super-X and the divertor plasma re-attaches. A peak heat flux of 9.8 MW/m2 to the lower outer divertor is observed for a multiple mode number simulation (n=2,4,6,...,20), shown in Fig. 2a), along with the evolution of various peak outer target values. The peak heat flux to the inner target is 2.2 MW/m2. The filaments that erupt from the plasma edge are observed; the plot of the plasma density in the poloidal plane is shown in Fig. 2b). This is a snapshot of the plasma density when the MHD activity is highest, here n=10 is dominant. When the energy of n=10 has reduced, the thermal energy and particle losses in the pedestal from ψ = 0.9 to ψ = 1.0 are 22% and 20% respectively.
In the divertor and around the x-point regions; lobe structures are observed. These structures are seen even more clearly in the electron temperature (Fig. 2c)). The whole region around the strike point is ergodised, with field lines connecting the pedestal top to the target (Fig. 2d)). A few milliseconds after the ELM crash the divertor heat fluxes and temperatures recover to pre-ELM conditions and the plasma appears to detach again, shown in Fig. 2a) and the ionization evolution in the outer divertors (Fig. 2e)). The recovery times are shorter than the inter-ELM phase, typically 10’s milliseconds for previous MAST experiments. A divertor pumping scan has been performed after the ELM crash. A phase where the ionisation front detaches from the target occurs, after the crash, for all but the highest pumping amplitudes due to the increase in density from the ELM. A pumping threshold is then observed at which the plasma re-attaches for high pumping and at low or no pumping the divertor ionisation front appears to remain detached. The pumping threshold observed is consistent with the pumping threshold before the ELM crash.
Different divertor configurations are used for ELM simulations with the single temperature reduced MHD model without neutrals. The ELM simulations show that moving from a conventional case to the Super-X case reduces the peak heat fluxes by an order of magnitude from 8 MW/m2 to 0.8 MW/m2. This can be understood by considering the area the heat flux is incident on in both cases - approximately 0.14 m2 in the conventional case to 1.5 m2 in the Super-X case; even though the peak heat fluxes are in general smaller than expected, in both cases, the comparison between divertor configurations appears to be reasonable.
References:
[1] R. Pitts et al., J. Nucl. Mater. 438, S48 (2013)
[2] W. Morris et al., IEEE Transactions on Plasma Science 46, 1217 (2018)
[3] E. Havlíčková et al., Plasma Phys. Controlled Fusion 57, 115001 (2015)
[4] D. Moulton et al., O5.129, 44th EPS Conference, Belfast, Northern Ireland (UK), Tech. rep.
[5] G. Huysmans and O. Czarny, Nucl. Fusion 47, 659 (2007)
[6] A. Fil et al., Physics of Plasmas 22, 062509 (2015)Speaker: Siobhan Smith (CCFE-UKAEA) -
The Wendelstein 7-X (W7-X) experiment commissioned the first of two neutral beam boxes [1] in the previous divertor campaign, providing 3.6 MW of heating power, achieving of densities above $2\times10^{20}$ $m^{-3}$, and providing the first initial assessment of fast ion confinement in the device. Demonstration of the confinement of fast ions is key to forwarding the stellarator concept as a nuclear fusion reactor. Experiments exploring the interplay between electron-cyclotron resonance heating (ECRH) and neutral beam injection (NBI) were performed through a series of discharges varying the ratio of NBI to ECRH power. It was found that even a small amount of ECRH was enough to arrest a continuous density rise during NBI operation. Discharges solely heated by NBI featured a continuous density rise with strong density peaking in the core of the plasma. In these discharges, densities above $2\times10^{20}~ m^{-3}$ were achieved, opening the possibility to explore OXB ECRH operation. Infrared camera images suggest fast ion wall loads which are consistent with numerical predictions. In general, discharges were free from the presence of Alfvénic activity suggesting future upgrades to assess the triggering these modes. These experiments provide data for future scenario development and initial assessment of fast-ion confinement in a drift optimized stellarator.
The NBI system on W7-X is designed to inject neutral hydrogen at 55 keV providing 1.8 MW of heating per source for up to 5 s (60 keV, 2.5MW, and 10 s for Deuterium), thereby providing particles which mimic fusion alphas (gyro-radius scaling) in a larger Helias reactor. The injection geometry is neither radial, nor tangential, but rather populates both the trapped and passing particles, allowing the assessment of fast ion confinement across the trapped passing boundary. The experiments conducted drove a beam current in a direction which lowers the overall rotational transform consistent with beam line geometry. Discharges solely heated by NBI indicate a drastically different character than those with a combination of ECRH and NBI. A continuous density rise over the discharge was found to be arrested by even a small amount of ECRH consistent with results from Wendelstein 7-AS (figure 1). In addition to acting as a heating/fueling source, the beam-plasma interaction enables spectroscopic measurements of beam attenuation, density and power as well as impurity densities, ion temperature, and rotation measurements, which are key for validation of equilibrium, transport and fast ion codes. The first experiments on W7-X with NBI successfully demonstrated the system and helped to access parameter regimes as yet unaccessible with ECRH operation alone.
Discharges heated solely by NBI demonstrated a continuous density rise achieving densities above $2\times10^{20}~m^{-3}$ [2]. These discharges had density profiles which peaked inside of $r/a\sim0.6$ with little change in density outside this radius (figure 2). Temperatures were relatively modest around 1.0 keV with broad shapes in these discharges. Discharges with similar levels of ECRH and NBI indicated a small density rise, no meaningful peaking of the density profile, and a small change in ion temperature. Even a small amount of ECRH introduced into a NBI discharge was enough to arrest the continuous density rise, with reduction in density peaking in the core of the plasma. Achievement of these high densities suggests the future possibility of operating with OXB ECRH heating above the O-2 cutoff.
The assessment of fast ion confinement in W7-X is key topic in the overall experimental program and NBI is envisioned as the primary method by which to achieve this goal. The particles injected by the NBI system scale, in normalized gyro-radius, to fusion alphas in a larger HELIAS type reactor. It has been predicted that as plasma beta increases, fast-ion confinement improves [3]. These experiments have provided data to help validate our numerical models for fast ion confinement. Predictions of wall overloads [4,5], beam deposition [6], and radial electric fields [7] have already been validated against experimental data provided by the NBI system. These models are now being used in the development of fast-ion diagnostics for future campaigns [8,9].
The first experiments on W7-X with NBI have provided a wealth of information for more detailed studies to come in future campaigns. The ability of the system to sustain plasmas for 5 s, thereby achieving high density operation was demonstrated. In the next campaign, a second beam box with two sources will be brought into operation, doubling the heating power and fueling of the system.
[1] Rust, N. et al. (2011). W7-X neutral-beam-injection: Selection of the NBI source positions for experiment start-up. Fusion Engineering and Design, 86(6-8), 728–731. http://doi.org/10.1016/j.fusengdes.2011.03.054
[2] Wolf, R. C. et al. (2019). Performance of Wendelstein 7-X stellarator plasmas during the first divertor operation phase. Phys. Plasmas, 26, 082504. https://doi.org/10.1063/1.5098761
[3] Drevlak, M. et al. (2014). Fast particle confinement with optimized coil currents in the W7-X stellarator. Nuclear Fusion, 54(7), 073002. http://doi.org/10.1088/0029-5515/54/7/073002
[4] Äkäslompolo, S., et al. (2019). Validating fast-ion wall-load IR analysis-methods against W7-X NBI empty-torus experiment. Journal of Instrumentation, 14(07), P07018–P07018. http://doi.org/10.1088/1748-0221/14/07/P07018
[5] Äkäslompolo, et al. (2018). Modelling of NBI ion wall loads in the W7-X stellarator. Nuclear Fusion, 58(8), 082010–15. http://doi.org/10.1088/1741-4326/aac4e5
[6] Lazerson, S. et al. (2020) Validation of the BEAMS3D neutral beam deposition model on Wendelstein 7-X. (submitted)
[7] Ford, O. et al. (2020) Charge Exchange Recombination Spectroscopy at Wendelstein 7-X. (submitted)
[8] Lazerson, S. et al. (2019). Development of a Faraday cup fast ion loss detector for keV beam ions. Review of Scientific Instruments, 1–6. http://doi.org/10.1063/1.5111714
[9] Ogawa, K. et al. (2019). Energy-and-pitch-angle-resolved escaping beam ion measurements by Faraday-cup-based fast-ion loss detector in Wendelstein 7-X. Journal of Instrumentation, 14(09), C09021–C09021. http://doi.org/10.1088/1748-0221/14/09/C09021Speaker: Dr Samuel Lazerson (Max-Planck-Institut für Plasmaphysik) -
Flux driven pedestal formation in tokamaks:
Turbulence simulations validated against the isotope effectC. Bourdelle1, G. De Dominici1, G. Fuhr2, P. Beyer2, L. Chôné3, F. Cianfrani2, G. L. Falchetto1, X. Garbet1, Y. Sarazin1
1 CEA, IRFM, F-13108 St-Paul-Lez-Durance, France
2 CNRS, Aix-Marseille Univ., PIIM UMR7345, Marseille, France
3 Department of Applied Physics, Aalto University, Espoo, FinlandE-mail: clarisse.bourdelle@cea.fr
Spontaneous pedestal formation above a power threshold at the edge of magnetically confined plasma is modelled for the first time in flux driven three-dimensional fluid simulations of electromagnetic turbulence with the code EMEDGE3D 1. The model implemented in EMEDGE3D is based on nonlinear fluid equations for the charge, energy balance and Ohm's law, the three transported fields being the electrostatic potential, the electron pressure and the magnetic potential 2.
Three key ingredients of the edge turbulent transport are simultaneously included in the flux driven simulations, applied on realistic L mode edge parameters, namely:
- an edge turbulence modelling accounting for resistive ballooning modes as well as drift waves [3,4,5,6]
- the electromagnetic effects on edge turbulence [3,4,5,6]
- a force balance radial electric field accounting for a realistic neoclassical poloidal velocity profile, i.e. with a realistic L mode edge radial variation of collisionality (from banana to Pfirsch-Schlüter regimes) [7,8]
The existence of a threshold on the injected power above which a pedestal forms is recovered. The pedestal formation is shown to be due to the E×B shear of the turbulence, following the BDT criterion [9]. The neoclassical friction and the Reynolds stresses are of the same order, while the Maxwell stress is negligible.
![From [1]. Time averaged profiles of, (a) the equilibrium electron pressure in arbitrary units, (b) the equilibrium radial electric field in arbitrary units, for a heat source lower than the threshold (Q_0=40, dashed line) and a heat source above the threshold (Q_0=55 full line). 1<r/a<1.02 corresponds to the RHS buffer zone.](http://irfm.cea.fr/Pisp/clarisse.bourdelle/Fig2.jpg)
The validity of the physics embedded in the fluid turbulence modelled is further challenged by changing Deuterium for Tritium. A lower threshold value on the power leading to the formation of a pedestal is observed in T versus D, similarly to experimental observations. The E×B quenching is made easier in T due to longer turbulence auto-correlation time.
![From [1]. Radially averaged confinement time over the simulation domain, tau_E in ms with respect to the radially averaged heat flux Q_0 in arbitrary units. For Deuterium (as shown in Fig.1) in black dotted line, for Tritium in yellow dashed line.](http://irfm.cea.fr/Pisp/clarisse.bourdelle/Fig6.jpg)
Even though the fluctuation level of the pressure is around 5 to 10% when approaching the separatrix in L mode, the quasilinear approximation is found to be valid. As in gyrokinetic L mode edge modelling [6], in the fluid EMEDGE3D simulations, the linear and non-linear cross-phases agree with each other [5]. Moreover, the Kubo number (the ratio between the turbulence auto-correlation time and the time-of-flight of the particle, here the eddy turnover time) is estimated and shown to be consistently lower than unity..
The validity of the quasilinear approximation up to the separatrix gives perspectives towards a reduced quasilinear model adapted to the L-mode edge region, as long as such a reduced model accounts for the two important effects of electromagnetic destabilization and stabilizing diamagnetic coupling.This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.
References
1 G. De Dominici, G. Fuhr, P. Beyer, C. Bourdelle, L. Chôné, F. Cianfrani, G. L. Falchetto, X. Garbet, Y. Sarazin, submitted to Physical Review Letters, available on ArXiv https://arxiv.org/abs/1912.09792v1, 2019.
2 G. Fuhr, P. Beyer, S. Benkadda, and X. Garbet. Physical Review Letters,101(19), 2008.
3 B. N. Rogers, J. F. Drake, and A. Zeiler. Physical Review Letters, 81(20):4396_4399, 1998.
[4] B. D. Scott. Physics of Plasmas, 12(6), 2005.
[5] G. De Dominici, G. Fuhr, C. Bourdelle et al, Nuclear Fusion, 59(12):126019 2019.
[6] N. Bonanomi, C. Angioni, P.C. Crandall et al, Nuclear Fusion, 59(12):126025, 2019.
[7] L. Chôné, P. Beyer, Y. Sarazin, G. Fuhr, C. Bourdelle, S. Benkadda. Physics of Plasmas, 21(7):070702, 2014.
[8] G. Y. Park, S. S. Kim, Hogun Jhang, P. H. Diamond, T. Rhee, X. Q. Xu., Physics of Plasmas, 22(3), 2015.
[9] H. Biglari, P. H. Diamond, P. W. Terry. Physics of Fluids B: Plasma Physics, 2(1):1_4, 1990.Speaker: clarisse bourdelle (CEA, IRFM, F-13108 Saint-Paul-lez-Durance, France.) -
W7-X is the first large fusion device where stationary reactor relevant densities have been achieved with electron heating only as it will be also the case for alpha particle heated fusion reactors. Employing the multi-pass ECRH scenario in the second harmonic O-mode (O2-ECRH), stationary (12 s) densities of up to 1.5 1020 m-3 had been achieved with hydrogen gas fueling. This scenario also made the stationary divertor detachment possible [1]. With a strongly reduced power load at the divertor tiles, detachment is a candidate for particle and power exhaust in fusion reactors. Even though electrons have been heated primarily, the collisional electron-ion coupling at that high density brought the ion temperature up as well. In particular in combination with pellet injection transiently ion temperatures Ti above 3 keV and close to the neoclassical limit of the W7-X magnetic configurations could be achieved. Thus enabling to test the W7-X neoclassical optimization [2]. Also the beneficial effect on impurity control by electron heating could be demonstrated.
The 140 GHz EC-wave with ordinary (O2) polarization give access to densities beyond the cut-off limit of the commonly used extra-ordinary (X2) polarized waves, but for the expense of incomplete single pass absorption for the W7-X plasma parameter. This disadvantage could be compensated by a special the multi-pass scenario, where the partially (60-70%) absorbed ECRH-beams have been reflected by specially shaped tiles and passed three times through the plasma core with an overall absorption of up to 90%. A detailed analyses of the absorption and losses for all 10 ECRH beams in the experiments enabled further optimization of the reflector tiles, thus in the next campaign an overall absorption of 95% is envisaged. This on the first glance small improvement will reduce the none-absorbed ECRH stray radiation by 50%, which is a remarkable step forward for prospective future steady state operation, since the microwave stray radiation gives an additional load to all W7-X components even if they are outside a line of sight to the plasma [3]. The O2-opereration scenario has been also routinely used for plasma operation below the X2 cutoff at densities above 0.8 1020 m-3 prohibiting the otherwise high risk of uncontrolled beam deflection in the X2-mode ECRH case. In order to keep the absorption high, the density feed-back controls for a central Te above 2 keV. In this scenario stable and stationary detachment for 24 s could demonstrated, which was only limited by the energy limit given by the uncooled plasma facing W7-X components in OP1.2. A further benefit of the O2-ECRH scenario was the compatibility with high neutral pressure at the plasma edge. Even though after the boronization of the W7-X wall no glow discharge cleaning has been performed any more the density control was never lost and the high density operation was very robust. But on the other side in the presence of the high neutral fluxes, plasma radiation and charge exchange losses pushed down the edge temperatures as shown in Fig 1.
The possible large density range enable a combined operation with pellet injection without the risk of approaching the cut-off condition for the here pellet-induced peaked density profiles. After the pellet injection phase, the transport properties are being improved for the ion and thus high plasma performance with high tripple product values have been achieved [4]. Here the ion power flux approaches the neoclassical value enabling to test the neoclassical transport optimization of W7-X.
The neoclassical impurity transport in stellarators predicts an inward pinch and thus a impurity accumulation. Electron temperature gradient driven turbulence is counteracting here and thus in ECRH-plasmas with gas fueling no accumulation has been found [5]. In particular the laser blow off experiments estimated impurity confinement time of the order of 70-80 ms which is far below the neoclassical confinement times of 2-10s [6]. Even more in cases, where impurity accumulation has been found, like in the exclusively NBI-heated high density plasmas, additional O2-ECRH significantly flattened the otherwise peaked Impurity profiles and pushed the impurities toward the plasma edge.
The high density operation with the multi-pass O2-ECRH scenario showed an excellent plasma performance. However the maximal achieved temperature were limited by the amount of available heating power (6MW for 20s) and the respective transport parameters of the plasma scenarios. In particular the gas fed ECRH-plasmas suffered from low plasma edge temperatures, which prohibit the efficient use of magnetic confinement region. For the next operation campaign an upgrade of the available ECRH power to 10 MW and more effective multi-pass reflector tiles are planned. In particular a new more powerful (1.5 MW) gyrotron is being developed now [7] and number of gyrotrons and beamlines will be increased from 10 to 12. In addition the maximal power capability of the in-air quasi-optical transmission line is being enhanced with a strong air drying system. Plasma performance enhancement is also expected by an improved edge neutral density control with cryo pumps in the divertor pumping gap and gas valves in the divertor region. In addition a steady state pellet injector will enable continuous core fueling at low neutral edge density.
References:
[1] M. Jakubowski et al, submitted to Phys. Rev. Lett. (2020)
[2] C. Beidler et al. in preparation (2020)
[3] H.P. Laqua et al. In Proceedings of the 28th EPS Conf. Control. Fusion and Plasma Phys.,Funchal 2001, ECA 25A, European Physical Society, Geneva 2001, 1277-1280.
[4] T.Sunn Pederson et al. Plasma Physics and Controlled Fusion, Volume 61, Number 1
[5] A. Langenberg, F. Warmer, G. Fuchert et al. PPCF 61 014030 (2019)
[6] T. Wegner et al. RSI 89, 073505 (2018); https://doi.org/10.1063/1.5037543
[7] G.Gantenbein et al., this conference (2020)
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training program 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European CommissionSpeaker: Heinrich Laqua (Max-Planck-Institute for Plasma Physics, Greifswald, Germany) -
To improve the performance of ECRH, a real-time control system for the deposition location of ECRH was newly developed to compensate the effect of refraction in standard oblique injection of ECRH on LHD. By the control absorption power higher than that without control was maintained under line-averaged electron density $n_{\rm e,avg}\simeq 3\times 10^{19}$ m$^{-3}$. In addition to that, a method of perpendicular injection was developed in order to be more insensitive to the effect of refraction. The achieved central electron temperature in the case of perpendicular injection was about 2 keV higher than that in the case of standard oblique injection for $n_{\rm e0}\sim 1\times 10^{19}$ m$^{-3}$ by 1 MW injection. With such improved performance of ECRH, high density ECRH plasma of $n_{\rm e0}\sim 8\times 10^{19}$ m$^{-3}$ was successfully sustained after multiple pellet injection.
In tokamak and helical fusion plasmas, adjustments of launcher settings of ECRH are necessary to produce high-performance plasma, to realize desired power deposition profiles, to decrease the stray radiation level in the vessel, and to prevent damages of in-vessel components from unabsorbed power during high-power long-pulse injection. The precise evaluation of deposition profiles is also essential for transport studies [1]. For such purposes, two heating methods have been developed to improve performance of ECRH in LHD: (i) real-time deposition location control and (ii) perpendicular injection.
The real-time control system for the deposition location was newly developed with a fast field programmable gate array (FPGA) [2]. Appropriate settings for a steerable launcher were obtained by evaluating deposition profiles for various $n_{\rm e}$ profiles using the ray-tracing code LHDGauss. Figure 1 shows a typical demonstrated result of the real-time deposition location control in order to obtain and to sustain high absorption power under the EC wave refracted by time-varying $n_{\rm e}$. The vertical target position $Z_{\rm f}$ was controlled according to a gradual increase in $n_{\rm e}$ profiles. Since a deposition location can be controlled mainly by changing $Z_{\rm f}$ with the launcher on an equatorial outer port of LHD, the toroidal target position was fixed almost at $-0.4$ m, which is a standard toroidally-oblique injection setting. Under $n_{\rm e,avg}\simeq 2\times 10^{19}$ m$^{-3}$ at 5 s, the deposition location in the case with the control was maintained at $r_{\rm eff}/a_{99}\sim 0.3$, while the deposition location in the case without the control shifted outward, which indicates that heating in the plasma core region was maintained longer due to the control. After 5 s, the absorption power in the case without the control gradually decreases due to refraction of the EC wave by a gradual increase of $n_{\rm e}$, while higher absorption power was maintained longer due to the control under $n_{\rm e,avg}\simeq 3\times 10^{19}$ m$^{-3}$. At higher densities, a decrease in absorption power was observed even in the case with the control. The absorption power level was almost the same as that in the case without the control, which suggests that the effect of multi-pass absorption is expected to be dominant in the experimentally-evaluated absorption power.
The above-mentioned situation led to the necessity of perpendicular injection (i.e., perpendicular to the ECR layer) because it is expected to be more insensitive to the effect of refraction. It is a similar heating method in tokamaks and W7-X, and will be also used in a helical DEMO reactor FFHR-c1. When 77-GHz EC waves are injected from the outside of the horizontally-elongated cross section, on-axis ECRH with perpendicular injection is available only in the magnetic field increased with sub-cooled helical coils. Figure 2 shows comparisons of the $T_{\rm e}$ responses between perpendicular injection and oblique injection in power-modulated ECRH. The perturbation amplitude of $T_{\rm e0}$ in the case of perpendicular injection was higher than that in the case of oblique injection. Perpendicular injection showed better central heating than oblique injection, although the absorption power were almost the same in both cases. Refraction and Doppler-shifted absorption in oblique propagation of the EC wave caused broadening of the deposition profiles, in particular at high density, even with the real-time deposition location control. This result clearly demonstrated that perpendicular injection is more insensitive to refraction and Doppler effects than oblique injection as expected.
Achieved $T_{\rm e}$ profiles were compared between the two kinds of injection of 1 MW without modulation. Plasmas were sustained by other two 154-GHz gyrotrons with 1 MW injection power each. $T_{\rm e0}$ increased from 4 keV to over 6 keV by perpendicular injection for $n_{\rm e,avg}\simeq 1\times 10^{19}$ m$^{-3}$. The achieved $T_{\rm e0}$ is about 2 keV higher than that in the case of oblique injection.
Such an improved ECRH performance has opened up a new operational region in ECRH plasmas. As shown in Fig. 3, high density plasma with $n_{\rm e0}\sim 8\times 10^{19}$ m$^{-3}$ was successfully sustained after injection of three consecutive deuterium pellets for the first time in ECRH plasmas of LHD. Hollow $n_{\rm e}$ profiles by gas puffing changed to peaked profiles after the pellet injection. Equipartition heating was significant in the high-$n_{\rm e}$ region: $T_{\rm i0}\sim T_{\rm e0}\sim 1$ keV were obtained. The new high-$n_{\rm e}$ region in ECRH plasmas will contribute to comparative studies in transport between W7-X and LHD and to helical reactor designs.
These two techniques for efficient first-pass absorption in the plasma core region are beneficial not only for preventing damages of in-vessel components during long-pulse operations but also for extending high-$T_{\rm e}$ and high-$n_{\rm e}$ operational regimes and precise transport studies.
[1] K. Tanaka et al., Nucl. Fusion 59 (2019) 126040.
[2] T. Ii Tsujimura et al., Fusion Eng. Des. 131 (2018) 130.
Speaker: Dr Toru Tsujimura (National Institute for Fusion Science) -
In non axis-symmetric, magnetic confinement fusion devices like the optimized stellarator Wendelstein 7-W (W7-X), recent theoretically predicted aspects on impurity transport, as the existence of a mixed collisionality regime $^1$ or the build up of a radial electric field $E_r$ $^2$, have been addressed in several initial experimental studies. Based on measurements of impurity transport times $^{3,4}$ or radial impurity diffusivity profiles $^{5,6}$, these studies are hinting for a strong anomalous impurity transport mechanism in W7-X.
In this work, a possible suppression of the anomalous impurity transport in a so-called ion-root plasma scenario is investigated, as in those scenarios already a significant improve of the energy confinement has been observed $^7$. Using X-ray imaging spectrometer (XICS) data, impurity density profiles of neighboring Ar charge states, namely $n_{Ar}^{15+}$, $n_{Ar}^{16+}$, and $n_{Ar}^{17+}$ (see Fig.1), are used to derive impurity fluxes of $Ar^{16+}$ as described in detail elsewhere $^5$.
In Fig.2, experimentally derived, radial fluxes of $Ar^{16+}$ are shown comparatively for an experiment program during a pure ion-root confinement time interval (Fig.3 a) and a central electron root confinement (CERC) phase (Fig.3 b).
As evident from Fig.2, in the CERC scenario a positive, radially outwards directed Ar flux from the plasma center up to half of the plasma radius $\rho$ = 0-0.5 (see dashed line) is observed, being dominant over a negative, radially inwards directed Ar flux from half of the plasma radius to the plasma edge. In the ion-root scenario, this positive Ar flux is restricted to the plasma center, now with a dominant negative radially inward directed Ar flux in the entire bulk plasma region of $\rho$ > 0.2.
Finally, Fig.3 c)+d) show measured diffusion and velocity profiles, derived from the above shown Ar flux measurements for the ion-root plasma scenario. Compared to typical diffusivities of D ~ 1.5-3 m$^2$/s measured in CERC plasmas (see shaded area in Fig.3c) $^{5,6}$, one finds a significantly reduced impurity diffusivity D in the plasma bulk region $\rho$ = 0-0.6, accompanied by a strong negative convection velocity v for the ion-root confinement scenario.
As the the ion- and electron-root plasma scenarios exist at different values of $n_e$, $T_e$, and $T_i$, additional neoclassical STRAHL simulations are performed to disentangle possible $n_e$, $T_e$, and $T_i$ contributions to the observed changes in the Ar fluxes, by comparing STRAHL simulated and measured Ar density profiles, given the measured diffusion and velocity profiles.References:
$^1$ P. Helander, S.L. Newton, A. Mollén et al. Phys. Rev. Lett. 118, 155002 (2017)
$^2$ N.A. Pablant, A. Langenberg, A. Alonso et al. Physics of Plasmas 25, 022508 (2018)
$^3$ A. Langenberg, F. Warmer, G. Fuchert et al. Plasma Physics and Controlled Fusion 61, 014030 (2019)
$^4$ Th. Wegner, B. Geiger, F. Kunkel et al. Review of Scientific Instruments 89 073505 (2018)
$^5$ A. Langenberg, N.A. Pablant, O. Marchuk et al. Nuclear Fusion 57 086013 (2017)
$^6$ B. Geiger, Th. Wegner, C.D. Beidler et al. Nuclear Fusion 59 046009 (2019)
$^7$ R. Wolf, A. Alonso, S. Akkäslompolo et al. Physics of Plasmas 26 082504 (2019)Speaker: Andreas Langenberg (Max-Planck-Institut für Plasmaphysik,17491 Greifswald, Germany) -
This work reports on advances in the validation of QH-mode modeling and the integration of momentum transport effects from ion-orbit loss (IOL), neutral interaction and neoclassical poloidal ion-flow damping with the NIMROD code [Ref 1]. Present understanding of steady-state tokamak edge-plasma solutions without edge-localized modes (ELMs) is interpretive. While MHD modeling is successful in simulating the low-n dynamics associated with quiescent H-mode (QH-mode [Ref 2]), predictive modeling requires integration with transport effects outside the MHD model. Next generation tokamaks are challenged by the need to accommodate associated high heat fluxes on the divertor and innovative divertor solutions are being developed. It is critical that these solutions are compatible with edge solutions without ELMs and, given the cost of such devices, predictive simulation is required.
Regarding DIII-D QH-mode dynamics, present simulations focus on the local validation through comparisons of beam-emission spectroscopy (BES) measurements and a synthetic diagnostic with simulated fluctuations. Prior MHD simulation demonstrates the sensitivity to shear flow and produces the preferential density transport as compared to thermal transport [Ref 3] that is observed in experiment. Analysis of DIII-D QH-mode (shot 163518) during a phase with edge harmonic oscillations (EHO) shows that the MHD state is a nonlinear phenomenon which can not be explained by a linear-growth-rate analysis. Simulation of a reconstruction based on the best fit to experimental measurements is stable and the instability drive is increased by narrowing the pedestal width in the measured state. Nonlinear 3D MHD fluctuations broaden the pedestal by relaxing the profiles back toward the initial state. The correct underlying drive is not known a priori; thus a scan of profile gradients similar to that performed in gyrokinetic validation studies is needed. With a weak drive a coherent mode develops while at strong drive a turbulent nonlinear state with large amplitude fluctuations is found as shown in Fig. 1. The line of sight of the BES measurements is not aligned with the magnetic field line pitch in these reversed-current discharges making a full synthetic diagnostic critical for comparison. The BES synthetic diagnostic shows MHD simulations contain low-frequency coherent dynamics similar to those measured during EHO. Examination of the cross-power spectrum from poloidally separated measurement points shows a double-peaked spectrum near the last closed flux surface (LCFS) and a single-peaked spectrum further towards the core in both simulation and experiment.
In terms of momentum transport, the effects from IOL, neutral interaction, neoclassical-ion-poloidal-flow and fast-viscous-parallel-flow damping are critical to determine the pedestal flow profiles. IOL occurs when passing particles inside the LCFS drift outwards on to trajectories that hit the divertor as illustrated in Fig. 2. The resulting co-current torque from IOL is incorporated through a drift-kinetic calculation from Ref. 4 that captures the impacts of collisional filling of the phase-space loss cone, the radial dependence of the force relative to the LCFS and orbit-squeezing by the electric field. Analysis from an inter-ELM period without large applied 3D perturbations of DIII-D shot 164988 shows a large ion-orbit loss torque near the LCFS. To balance this torque a dynamic neutral model based on the equations of Ref. 5 that includes ionization, recombination and charge exchange is used. Figure 3 shows the 1D neutral distribution from core to wall for an initial state set by a local equilibration of ionization and recombination (red line) and after equilibration including the effect of ballistic expansion (blue line). The resultant drag from neutrals can balance the IOL torque leading to pedestal steady-state-flow profiles.
Planned work focuses on validation of the fluctuations observed during the broadband QH-mode where local diagnostics observe both ion- and electron-flow-directed dynamics that indicate the potential importance of two-fluid effects. Ultimately validation and incorporation of two-fluid effects and transport is required to provide understanding of the compatibility of ELM-free regimes and divertor
This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Awards DE-SC0019070, DE-SC0018311, DE-SC0018313 and DE-FC02-04ER54698.
Ref 1: Sovinec et al., JCP 195, 355 (2004); Sovinec and King, JCP 229, 5803 (2010)
Ref 2: Chen et al., Nucl. Fusion 57 086008 (2017); Burrell et al., Phys. Plasmas 23 056103 (2016); Garofalo et al, Phys. Plasmas 22 056116 (2015) and refs. Within.
Ref 3: King et al., Phys. Plasmas 24 055902 (2017); King et al., Nucl. Fusion 57 022002 (2017)
Ref 4: Shaing, Phys. Fluids B 4 3310 (1992)
Ref 5: Meier and Shumlak, Phys. Plasmas 19, 072508 (2012)heat-flux solutions in the next generation of tokamaks.Speaker: Jacob King (Tech-X Corporation) -
Isotope effect on energy confinement time and thermal transport has been investigated for L-mode plasmas in LHD. Detailed and comprehensive assessment in stellarator-heliotron plasmas has progressed since the last IAEA-FEC [1]. Plasmas of hydrogen(H), deuterium(D) and their mixture including helium (He) have exhibited no significant dependence on the isotope mass $M$ in thermal energy confinement time. This is not consistent with simple gyro-Bohm model. Comparison of thermal diffusivity for dimensionally similar H and D plasmas in terms of gyro radius $\rho_*$, collisionality $\nu_*$ and thermal pressure $\beta$ has clearly shown robust confinement improvement in D to compensate for unfavorable mass dependence predicted by the gyro-Bohm model. Also, the transient electron heat transport does not show a substantial difference in H, D and their mixture plasmas. These observations have shown distinction to tokamaks and rouse challenge to common physics picture of isotope effect.
It has been recognized that thermal transport in toroidal plasma is dominated by the turbulence with the characteristic scale of ion gyro-radius. This model is referred to as gyro-Bohm. However, while plasma of heavier hydrogenic isotope would have larger thermal diffusivity according to this gyro-Bohm model ($\chi \propto \rho_* \propto M^{1/2}$), major experimental observations in tokamak have shown better confinement in D plasma than in H plasma. This isotope effect remains a long-standing mystery in fusion plasma research. Clarification of the origin of the isotope effect is a key issue to project a DT burning.
Extensive and elaborated comparison of H and D, and also mixture plasmas including He have been done on LHD. The effective mass $M_{eff}$ and the effective charge $Z_{eff}$ range from 1.0 to 3.41 and from 1.12 to 3.0, respectively. The surveyed ranges of magnetic field, line averaged density and absorbed heating power of NBI and ECH are summarized as 1.38T $\le$ $B$ $\le$ 2.75T, 0.5 $\times 10^{19}m^{-3}$ $\le$ $\bar{n}_e$ $\le$ 5.7 $\times 10^{19} m^{-3}$, 1.5MW $\le$ $P_{abs}^{NBI}$ $\le$ 12.5MW, and 1.8MW $\le$ $P_{abs}^{ECH}$ $\le$ 3.5MW. Here it should be noted that electron heating is dominant for all data as $P_{abs}^{e}/P_{abs}$ = 0.81 $\pm$ 0.11.
Figure 1 shows the comparison of thermal energy confinement time in the experiment with the scaling expression derived from statistical regression analysis; $\tau_{E,th}^{scl} \propto M^{-0.02} B^{0.89} \bar{n}_e^{0.75}P_{abs}^{-0.90}$. Thermal stored energy has been evaluated by profile documentation and dilution of ions due to major impurities of helium and carbon is also taken into account. Here no significant dependence on $M$ is identified and this expression is inconsistent with the gyro-Bohm model suggesting $M^{-0.5}$ dependence. When the energy confinement time normalized by ion gyro-frequency $\Omega_i$ is assumed to be expressed by 4 major dimensionless parameters, this scaling expression is rephrased into the dimensionless expression of $\tau_{E,th}^{scl}\Omega_{i} \propto M^{0.98} \rho_*^{-2.99} \nu_*^{0.21} \beta^{-0.35}$. It should be noted that the gyro-Bohm nature is described as $\tau_{E}\Omega_i \propto M^{0} \rho_*^{-3}$. Therefore, additional clear mass dependence is identified here, which compensates unfavorable negative dependence on mass in the gyro-Bohm model. At the same time, it should be emphasized that gyro-Bohm dependence of $\rho_*^{-3}$ persists.
Then thermal diffusivity in dimensionally similar plasmas has been compared in order to clarify the peculiarity of isotope effect seen in the energy confinement time. Since the three operational parameters $B$, $\bar{n}_e$, and $P_{abs}$ are controllable in the experiment, dimensionally identical condition in terms of $\rho_*$, $\nu_*$, and $\beta$ can be fulfilled for plasmas with different $M$. With using the confinement improvement factor of $\alpha=\tau_E^D / \tau_E^H$, which is 1/$\sqrt{2}$ if gyro-Bohm, the operational conditions for dimensionally similar H and D plasmas are given by the following relation, $B^{D}=2^{3/4}B^{H}$, $\bar{n}_e^{D}/\bar{n}_e^{H}$ = 2, $P_{abs}^{D}=2^{3/4}\alpha^{-5/2}P_{abs}^{H}$ [2,3]. In results, dimensionally similar pairs have been obtained around $\alpha$=1 as predicted by the discussion about the energy confinement time. Since these pairs of H and D plasmas have the same $\rho_*$, $\nu_*$, and $\beta$, normalized thermal diffusivity $\chi/\Omega_i$ would be the same based upon whichever neoclassical, Bohm or gyro-Bohm transport plays an essential role. However, Fig. 2 (a) shows that $\chi_{e} /(S_{2/3} \Omega_i)$ is improved in D from H by a factor of around 2 over the surveyed range of $\nu_*$, which may implicate 1/$M$. Here it should be noted again that electron heating is dominant under the present condition. The ratio of ion thermal diffusivity shows a different trend. One important element in comparison of thermal transport in H and D plasma is difference in collisional electron-ion energy exchange [4]. Heat transfer between electrons to ions $P_{ei}$ is proportional to $n^{2}(T_{e}-T_{i})/(MT_e^{3/2}$). Since $T_e$>$T_i$ in plasmas studied here, it is expected that enhancement of $P_{ei}$ in H plasmas leads to the increase of ion heat flux. Figure 2(b) shows the ratio of the electron heat flux $q_e$ to the ion heat flux $q_i$ corresponding to the data plotted in Fig.2(a). While electron heat flux decreases with the increase of $\nu_*$ ($\propto n/T^2$), this trend is less pronounced in H plasmas than in D plasmas. This is because the density is set at the double for D plasma in this comparison and the effect of mass on electron-ion energy transfer is cancelled out. Electron loss channel stays dominant, in particular in H plasmas, and ion loss channel does not become dominant. Therefore, improvement of electron thermal diffusivity in D shown in Fig.2(a) is consistent with the significant mass dependence ($\propto M^{0.98}$) seen in the non-dimensional expression of the energy confinement time. In this regard, global ($\tau_E$) and local characteristics ($\chi$) matches each other in L-mode plasmas on LHD.
In addition, emergence of a loop-type bifurcation during the modulated ECH (MECH) in the flux-gradient relation has been attracting interests, which corresponds to rapid and nonlocal responses of turbulence intensity and transport to heating power [5,6]. The transient response to MECH has been investigated for L-mode plasmas of H, D, and their mixture. While the trajectories on the flux-gradient diagram (see Fig.3) leap when the MECH is turned on and off, their hysteresis widths are the same for plasmas of H, D and their mixture. This observation is consistent with the fact that there is no significant mass dependence of energy confinement and thermal diffusivity.
-
H.Yamada et al., https://conferences.iaea.org/event/151/papers/5760/files/4346-Yamada2.pdf
-
H.Yamada et al., Phys. Rev. Lett. 123 (2019) 185001.
-
C.F.Maggi et al., Nucl. Fusion 59 (2019) 076028.
-
P.A.Schneider et al., Nucl.Fusion 57 (2017) 066003.
-
S.-I.Itoh, K.Itoh et al., Nucl.Fusion 57 (2017) 022003.
-
T.Kobayashi et al., Nucl. Fusion 58 (2018) 126031.
Speaker: Prof. Hiroshi Yamada (The University of Tokyo) -
-
The influence of plasma scenarios on the amplitude and radial structure of Zonal Flows has been investigated in the plasma edge of the TJ-II stellarator. The main results reported in this paper include: i) first experimental observation of the influence of the ion mass in the radial width of Long Range Correlations (LRC) as proxy of zonal flows; ii) the investigation of the influence of operational limits on Zonal Flows showing a decreasing in the amplitude of LRC as approaching the plasma density limit.
Plasma scenarios and experimental set-up. Plasma scenarios sustained by electron cyclotron resonance heating (ECRH) (two gyrotrons, 53.2 GHz, P ¬≈ 300 kW each, suitable for X2 heating in B = 1 T) and neutral beam injection (NBI) heating (two H0 injectors, E ¬ 30 kV, P ≈¬ 600 kW each) have been investigated in Hydrogen (H) and Deuterium (D) dominated plasmas in TJ-II. Direct generation of NBI plasmas in TJ-II with lithium-coated walls has allowed to investigate the influence of magnetic fields [B = 0.7 – 1 T] in the density limit. A dual system of multi-probes arrays, placed at two different toroidal and poloidal locations, has been used to characterize the properties of LRC in floating potential fluctuations in the TJ-II plasma edge [ ≈ 0.85 – 1].
Influence of isotope mass and plasma heating on the radial width of LRC. The mechanism governing the influence of the ion mass on plasma transport is still one of the main scientific conundrums facing the magnetic fusion community after more than twenty years of intense research.
In stellarators, the ambipolarity condition determining the radial neoclassical electric field (Er) has two stable roots: the ion root with typically negative Er, usually achieved in high density plasmas heated by NBI, and the electron root with positive Er, that is typically realized when electrons are subject to strong heating in ECRH scenarios. Therefore, plasmas produced by different heating schemes (ECRH and NBI) are characterized by different mean radial electric field (Er) in TJ-II in agreement with neoclassical predictions 1.
LRC in floating potential signals are dominated by frequencies below 20 kHz. As the frequency decreases, the oscillating Er structure asymptotically approaches to the mean potential profile. Mean Er profiles are comparable in H and D plasmas both in L-mode and H mode scenarios 2. The width of the oscillating Er structures depends on its frequency as well as the heating scheme, being the radial size of the LRC profile larger in the case of the ECRH phase than in the NBI phase both in H and D plasmas [Fig. 1]. The level of edge broadband turbulence [in the range 20 – 500 kHz] is significantly reduced in the transition from ECRH to NBI scenarios. Normalized values of density fluctuations and rms in floating potential are in the range of 10 - 20% and 5 - 10 V in D and H plasmas respectively [ ≈ 0.9].
In ECRH plasmas the LRC radial structure could not be fully explored due to the limited edge plasma region accessible to the dual probe system. Within experimental uncertainties, both the amplitude and radial width of LRC are comparable in H and D plasmas with a slight but systematic increases in the amplitude of LRC in D as compared to H scenarios. The case of NBI plasmas is different. Although the maximum amplitude of LRC is similar in H and D scenarios, its radial position is shifted radially inwards (in the range of 0.5 cm) in the case of D plasmas with respect to pure H plasmas. Furthermore, the radial size of the LRC is of about 1.5 times larger in D than in H plasmas, which is comparable to the ratio of ion Larmor radius (D vs H) in NBI plasmas [Fig. 1].
TJ-II findings show that whereas the neoclassical transport determines the radial electric field on long length scales (in the range of few tens of gyro-radius) zonal flows can control short radial length scales (in the range of few gyro-radius) that increases with Larmor radius. Furthermore, TJ-II results have shown that both Er scales are strongly intertwined and that turbulent and neoclassical mechanisms are involved in the dynamics of edge Zonal Flows [3]. These results emphasize the role of multi-scale mechanisms to unravel the physics of the isotope effect and pave the way for the validation of the influence of ion mass in GK simulations [4].
Influence of plasma density limit on LRC. Density limit is manifested in tokamaks, stellarators and RFPs. In stellarator the density limit is related to radiation collapse mechanisms [5] whereas in tokamaks edge transport can play an important role [6].
LRC in floating potential fluctuations has been investigated in the proximity of the density-limit in the NBI plasma scenarios in TJ-II [Fig. 2]. The density limit decreases with the magnetic field (B = 0.7 – 1 T) in qualitative agreement with the empirical stellarator scaling law for density limit [5]. At low densities the LRCs, measured at rho ≈ 0.9 for frequencies below 20 kHz, are quite large (≈ 0.8) and vary only slightly with increasing density. When approaching the density-limit the amplitude of LRC reduces rapidly with increasing plasma density, in agreement with previous results [7]. The reduction in the LRC as approaching the density limit is accompanied by a reduction in edge mean radial electric field and the level of plasma turbulence. Results point to the role of collisionality, mean E × B flows and level of turbulence on the amplitude of zonal flows in the proximity of the density limit.
TJ-II findings suggest that at a threshold radiation value for the density limit the degradation of confinement would be partially due to the damping of Zonal Flows as recently pointed out [8]. If this is the case, density limit in stellarators would depend on the transport optimization criteria, a prediction that could be validated experimentally.
1 J L Velasco et al., Phys Rev Lett 109 (2012) 135003
2 U. Losada et al. Plasma Phys. Control. Fusion 60 (2018) 074002
[3] R. Gerrú et al Nucl. Fusion 59 (2019) 106054
[4] M. Nakata et al. Phys. Rev. Lett. 118 (2017) 165002
[5] M. Hirsch et al., Plasma Phys. Control. Fusion 50 (2008) 053001
[6] M. Greenwald Plasma Phys. Control. Fusion 44 (2002) R27
[7] Y. Xu et al Nucl. Fusion 51 (2011) 063020
[8] R. J. Hajjar et al., Physics of Plasmas 25 (2018) 062306
Speaker: Dr Carlos Hidalgo (Laboratorio Nacional de Fusión - Ciemat) -
Strength of the internal transport barrier (ITB) is quantified in deuterium (D) and hydrogen (H) plasmas using a unique criterion based on the transport nonlinearity on the temperature. Two distinct isotope effects are found: (i) Stronger ITBs in D plasmas and (ii) edge confinement degradation accompanied by the ITB formation emerging in H plasmas. Principal component analysis reveals the important role of the density peaking for the strong ITB formation.
The isotope effect in magnetically confined plasma physics has been a long-standing mystery unsolved. In general, there are a number of experimental case studies on the isotope effect in tokamaks, but less in stellarators/heliotrons. The isotope effect is particularly prominent in transport barrier formation in tokamaks. For systematic understanding of the background physics, the isotope effect in the transport barrier property in stellarators/heliotrons need be assessed. Here, we report a recent progress on isotope effect studies for the ITB in Large Helical Device (LHD).
Unlike the case of tokamak plasmas, there is no generally accepted criterion for the ITB strength in stellarators/heliotrons. Here, a new criterion for the ITB strength is proposed by defining a unique scalar coefficient. The typical L-mode plasmas in LHD are characterized by the dome-shaped temperature profile with the diffusion coefficient being proportional to the temperature to the power of a factor $\alpha$, i.e., $\chi \propto T^\alpha$, where $\alpha=1$ is widely applicable. The reference L-mode profile $T_{\rm L}^{\rm ref}(r)$ is defined by extrapolating the edge temperature profile to the core according to the solution of the thermal diffusion equation with $\chi \propto T^\alpha$. By comparing $T_{\rm L}^{\rm ref}(r)$ with the entire temperature profile, the profile gain factor as the ITB strength is defined as $G_{1.0} = \int n (r) T (r) dV/\int n (r) T_{\rm L}^{\rm ref} (r) dV$, where the subscript 1.0 represents the $\alpha$ value used [A].
The proposed method is applied to both D and H discharges with the line averaged density $\bar{n}_{\rm e}$ scan in the shot-to-shot basis. In LHD, the ITB in the ion temperature profile is typically formed when an intense neutral beam heating is applied to low $\bar{n}_{\rm e}$ plasmas. Figures 1 (a) and (b) compare the high and low $\bar{n}_{\rm e}$ discharges for corresponding pairs of D and H plasmas, where the neoclassical transport is optimized by shifting the magnetic axis inward. Symbols and curves correspond to the measured data and $T_{\rm L}^{\rm ref}(r)$, respectively. Inserts present the electron density $n_{\rm e}$ profiles in the low $\bar{n}_{\rm e}$ discharges, showing that $n_{\rm e}$ tends to peak in D plasmas. However, the peaking of $n_{\rm e}$ is very weak and no ITB in $n_{\rm e}$ profile is present. Stronger ion temperature ITB is formed in the low $\bar{n}_{\rm e}$ D plasma with $G_{1.0} \sim 1.5$ compared to the H plasma with $G_{1.0} \sim 1.3$. The ITB foot-points in D and H plasmas are similar. Unlike the L-H transition, it is hard to define the threshold power for the ITB in LHD. In stead, we discuss $\bar{n}_{\rm e}$ dependence of the ITB strength $G_{1.0}$ in Fig. 1 (c), where the systematic comparison between D and H plasmas is performed. When $\bar{n}_{\rm e}$ is high, $G_{1.0} \sim 1$ that corresponds to the L-mode, and decreasing $\bar{n}_{\rm e}$ leads to a non-monotonic increase in $G_{1.0}$. Larger $G_{1.0}$ is routinely observed in D plasmas in $\bar{n}_{\rm e} < 2 \times 10^{19} {\rm m}^{-3}$. In addition, the edge temperature decreases when the ITB is formed in the H case as shown in Fig. 1 (b). The insert in Fig. 1 (c) shows the $G_{1.0}$ dependence of the edge ion temperature $T_{\rm i,edge}$ averaged in $0.68 < r_{\rm eff}/a_{99} < 0.83$. As $G_{1.0}$ increases, $T_{\rm i,edge}$ clearly decays in H plasmas. This limits the core temperature increment even in the presence of the ITB in H plasmas [B].
It is essential to identify parameters that play a role for determining the ITB strength. To resolve the relation among parameters that change with $\bar{n}_{\rm e}$, the principal component analysis (PCA) is performed. As appeared in literature, the inverse electron density gradient length $L_{n_{\rm e}}^{-1}$, the carbon density $n_{\rm c}$ and its inverse gradient length $L_{n_{\rm c}}^{-1}$ are predicted to impact on the ITB formation. The PCA is applied for the database in the five-dimensional parameter space ($\bar{n}_{\rm e},G_{1.0},L_{n_{\rm e}}^{-1},n_{\rm c},L_{n_{\rm c}}^{-1}$), where data with different magnetic configurations and the isotope mass are taken all at once. Here, the local parameters are obtained at the ITB foot-point of $r_{\rm eff}/a_{99} = 0.6$. Note that the density profiles apparently having no ITBs are treated as the background elements that possibly affect the ITB property. It is found that the primary and secondary principal components are almost equivalent to $\bar{n}_{\rm e}$ and $L_{n_{\rm e}}^{-1}$, respectively, and $\bar{n}_{\rm e}$ and $L_{n_{\rm e}}^{-1}$ have high correlations with $G_{1.0}$. According to the PCA, $G_{1.0}$ is plotted as a function of $\bar{n}_{\rm e}$ and $L_{n_{\rm e}}^{-1}$ for D and H plasmas in Fig. 2. Stronger ITBs are formed when $L_{n_{\rm e}}^{-1}$ is large and $\bar{n}_{\rm e}$ is small. A working hypothesis is drawn from the observation: In D plasmas $n_{\rm e}$ profile tends to peak due to the particle transport or the beam fueling. The peaked $n_{\rm e}$ profile enhances the ITB formation, possibly due to the ion temperature gradient mode stabilization [C]. The radial electric field shear remains approximately unchanged with $\bar{n}_{\rm e}$, showing less impact on the ITB formation.
The isotope effect in the transport barrier formation property is stronger than that in the L-mode plasma confinement in LHD [D-F]. This is qualitatively similar to the tokamak case, where the isotope effect in the transport barrier threshold power is much clearer with respect to that in the confinement scaling exponent. A detailed comparison between LHD and tokamaks stimulates the phenomenological understanding of the isotope effect.
[A] T. Kobayashi et al., Plasma Phys. Control. Fusion 61, 085005 (2019)
[B] T. Kobayashi et al., Sci. Rep. 9, 15913 (2019)
[C] M. Z. Tokar et al., Phys. Rev. Lett. 84 895 (2000)
[D] H. Yamada et al., Phys. Rev. Lett. 123, 185001 (2019)
[E] T. Kobayashi et al., submitted to Nucl. Fusion (2020)
[F] H. Yamada et al., this conference (2020)Speaker: Tatsuya Kobayashi (National Institute for Fusion Science)
-
-
