Turbulent transport characterization of a large number of stellarator configurations to guide future stellarator optimizations

The use of stellarators is one of the most promising pathway to commercial fusion. In the last decade, large progress has been made to improve their particle confinement and to increase their commercial viability as future fusion power plants. The strongest examples are the improvement of fast particles confinement and neoclassical transport, where metrics and surrogate models have been derived and validated. The next step in stellarator optimization is the improvement of turbulent transport. In recent years, this research has been accelerated by the development of fast GPU native codes, such as GX [1], which allows the evaluation of the non-linear turbulent heat-flux in reasonable times. This opens the door to broader parametric studies of turbulent transport, using a large number of stellarator configurations.

Using a recently created framework including GX, we are able to evaluate the non-linear heat-flux of a large number of optimized stellarators using either adiabatic or kinetic electrons. Results from these studies will be presented. Similarly, we assess the accuracy of existing surrogate models in predicting the saturated turbulent heat-flux.

References:

[1] N. R. Mandell, et al. 2018. J. Plasma Phys. 84, 905840108.