Speaker
Description
Due to complex non-linear interactions of various phenomena in the tokamak plasmas, integrated modelling is a proper tool to understand physics behind them. TRIASSIC (Tokamak Reactor Integrated Automated Suite for SImulation and Computation), which is a flexible integrated suite of codes for interpretive/predictive analyses, is under development for this purpose by exploiting merits of the standardized data model, IMAS [1]. Its structure is depicted in figure 1.
For easy maintenance, plasma analysis codes contained in this integrated modelling tool are designed to be fully independent and modularized, so that the integration of a new code or model is simple. Based on the Python framework, modularized analysis code can dynamically be called without losing fast computation speed and unnecessary file IO. The Python based graphical user interface (GUI) was also developed, which lowers the entry barrier to integrated modelling and enables users to easily conduct integrated simulations without loss of consideration on delicate modelling options.
For the validation of this integrated suite of codes, a database of ~50 stationary discharges of various KSTAR scenarios was established. First, we conducted interpretive simulations to compare plasma stored energy which can be calculated by using plasma density, temperature, and fast ion energy by TRIASSIC or by EFIT [2] relying on magnetic diagnostics. As shown in figure 2, comparison of energy calculated from TRIASSIC with EFIT yields significant overestimation of the plasma stored energy from TRIASSIC in high β_P discharges with TAE, which indicates that the additional fast ion transport model is needed in addition to the classical fast ion model [3]. The calculation on remnant majority of discharges, however, showed good agreement with experiments.
To test predictive capabilities of the integrated suite of codes, comparison of electron density and plasma energy was conducted. Using anomalous transport model, TGLF [4], significant amount of electron density and stored energy reduction was observed. It was found that this underestimation was due to the absence of neutral gas puff modelling [5]. Systematic scans on absolute amount of cold neutral inward flux could resolve experimental electron density and stored energy level. The electron density and stored energy was able to be predicted accurately, mostly within 10% deviation from its experimental values as shown in figure 3.
References:
[1] F. Imbeaux et al. 2015 Nucl. Fusion 55 123006
[2] L.L. Lao et al., 1985 Nucl. Fusion 25 1611
[3] R J Goldston et al., 1981 J. Comput. Phys. 43 61
[4] G.M. Staebler et al., 2005 Phys. Plasmas 14 055909
[5] S. Tamor, 1981, J. Comput. Phys. 40 104
| Affiliation | Seoul National University |
|---|---|
| Country or International Organization | Korea, Republic of |
