Closing the gap between simulations and experiment with GRILLIX

Understanding and predicting the scrape-off layer and divertor plasma is a crucial step for building confidence in fusion reactor designs. Building a predictive model is challenging, however, since the scrape-off layer and divertor plasma is challenging to model. The dynamics in these regions is governed by a complex interaction of effects, including strong magnetic shaping and the “sheath” plasma-surface interaction.In this talk, I will discuss how the GRILLIX fluid turbulence code was extended to treat realistic experimental magnetic geometries. I will also discuss two validations which were made using the extended version of GRILLIX — on the TORPEX and TCV devices at the Swiss Plasma Centre. The agreement found with TCV shows that the GRILLIX code can closely predict the absolute magnitude and shape of many outboard midplane, divertor volume and target measurements. These results are particularly remarkable because they come from “hands-off” GRILLIX simulations — the only free “knob” is the density source rate since the neutral ionisation is not yet included. These simulations therefore represent a significant step towards building predictive capability. Finally, I will discuss the potential applications of the validated code — including a promising ongoing project regarding predicting the turbulence levels within advanced divertor concepts.