Turbulence optimization of the HSX stellarator via small coil-current modifications

Turbulent transport commonly limits confinement in fusion devices. In stellarators, the 3D structure of the confining magnetic field provides the potential for reducing transport by optimizing the field geometry. To date, turbulence-optimized configurations have been obtained primarily for idealized or synthetic geometries, and no such configurations have been validated against experiments. In order to make stellarators competitive, turbulence optimization needs to be demonstrated in existing devices. To that end, it is shown numerically (M. J. Gerard et al. 2023 Nucl. Fusion 63 056004) that small changes in the magnetic field of a stellarator equilibrium can result in a significant reduction in trapped-electron-mode (TEM) activity in the Helically Symmetric eXperiment (HSX). This is accomplished by numerically generating > 10⁶ stellarator equilibria by modifying individual coil currents. A hierarchy of models is used to reduce this set of configurations to ~ 10² promising experimental-candidate configurations. In so doing, it is found that TEM growth rates can be regulated by modifying trapped-electron resonances via their precessional drifts across helically-linked magnetic trapping wells. Moreover, it is found that by elongating the plasma cross-section, TEM growth rates can be reduced without sacrificing helical symmetry. Considering configurations of varying elongation, nonlinear simulations reveal that the reduction in growth rates is correlated with an increase in the density gradient at which a sharp increase in heat flux is observed. Three configurations are identified in which the highest heat flux is a factor of two larger than the lowest heat flux, with a 30 % reduction from the standard configuration achieved by a moderate increase in elongation. These results are compared against quasilinear models to determine what impact the linear stabilization has on transport. This provides a set of candidate configurations that will be explored in future experiments on HSX.