Comparison of plasma response models for RMP effects on the divertor and scrape-off layer in KSTAR

TL;DR

This work addresses how resonant magnetic perturbations create helical footprints in the scrape-off layer and divertor of a KSTAR H-mode plasma, and how different plasma-response models alter predictions. Using FLARE to compute footprints based on GPEC, MARS-F, M3D-C1, and JOREK, coupled to EMC3-EIRENE for heat-load predictions, the authors find footprint sizes $S$ spanning from about $2\,$cm to $14\,$cm across models. The heat-load predictions generally overestimate measured striations, though the smallest footprints combined with lower SOL power or higher upstream density and radiative losses improve agreement, with M3D-C1 giving the closest match in several scans. The results underscore the importance of boundary treatment near the separatrix and motivate self-consistent free-boundary benchmarking (e.g., JOREK-STARWALL) and better constraints on separatrix density and impurity radiation for improved predictive capability.

Abstract

Resonant magnetic perturbations (RMPs) are beneficial for control of edge localized modes (ELMs) in tokamaks. Nevertheless, a side effect is the appearance of a helical striations in the particle and heat loads onto divertor targets. The extent and field line connection of these striations is significantly altered by the plasma response to external perturbations. For an ELM suppressed H-mode plasma at KSTAR, magnetic footprints are computed by FLARE based on plasma response from GPEC, MARS-F, M3D-C1 and JOREK with substantial differences in the resulting footprints (from 2 cm to 14 cm). This is reflected in EMC3-EIRENE simulations of the resulting heat loads: it is found that either the peak value or the extent of the striations appear to be overestimated compared to IRTV measurements. Reasonable agreement can only be achieved for the smallest footprint for lower input power and lower cross-field transport, or for higher upstream density and radiative power losses.

Figures (10)

• Figure 1: Perturbed magnetic separatrix (dark green, brown) for different plasma response models. The puncture points of a high density Poincaré plot are colored based on the minimum of $\psi_{N}$ taken along the corresponding field line. The core boundary for the EMC3-EIRENE simulations is shown in black.
• Figure 2: Magnetic footprints on the outer lower divertor target. The distance along the target is measured from the strike point of the reference configuration without RMPs. The traditional SOL with one poloidal turn manifests as an extension of the new helical SOL. The far SOL connects with 1/2 poloidal turn from the lower outer target to the upper outer target.
• Figure 3: Fourier harmonics of the external perturbation field (a) and the total perturbation field including plasma response from different models (b-d).
• Figure 4: Cross-section of computational mesh (lower resolution for visualization purposes) with block-structured layout in double null configuration. Plasma cells (EMC3) are shown in black. The mesh for neutral particles (EIRENE) is extended into the core and to the vacuum vessel.
• Figure 5: (a) IRTV measurement of the heat load by slow rotation of the external perturbation with corrections for a drifting equilibrium VanBlarcum2025. (b-d) Heat loads on the outer lower divertor target for $n_{\mathrm{sepx}}\xspace = 0.5 \cdot 10^{19} \, \meter^{-3}$ and $P_{\mathrm{SOL}}\xspace = 3 \, \mega\watt$ for different plasma response models.