The role of neutral gas in validated global edge turbulence simulations

Abstract

To make predictions for and design fusion reactors, a multitude of physical processes must be considered. In the edge and scrape-off layer (SOL), turbulent fluctuations intertwine with the plasma background, which is largely determined by neutral gas, and magnetic geometry plays an important role. A diffusive neutrals model has now been implemented in the global Braginskii edge turbulence code GRILLIX. The code is based on the flux-coordinate independent (FCI) approach, which allows efficient turbulence simulations in diverted equilibria. We validate simulations across the ASDEX Upgrade edge and SOL against measurements in discharge #36190, and find much better agreement thanks to the neutrals. Disentangling the effects of the neutral gas, we find that it affects the plasma in several ways. Firstly, the ionization of neutrals modifies the radial profiles of plasma density and temperature, leading to a transition of the turbulence drive from the general ballooning type to the ion temperature gradient type. Secondly, strong poloidal asymmetries can be induced due to divertor recycling, depending on the ionization pattern. As ballooned perpendicular plasma transport is stronger at the low-field side, neutrals penetrate deeper into the plasma at the high-field side, leading to significant ionization and radiation there. With increasing divertor neutrals density, the targets cool down while plasma density increases, more strongly at the high-field side. Much of the dynamics take place directly around the X-point and along the separatrix, which can be resolved by the FCI approach. Potential remains in extending the model and the code, but our results build confidence that predictive capability is within reach for major design questions for fusion reactors, such as the near SOL fall-off length.

Physics-informed machine learning techniques for edge plasma turbulence modelling in computational theory and experiment

“…But the underlying model approximations tend to only be marginally satisfied in fusion plasmas and there is a scarcity of clear-cut validation tests of the theory against higher-fidelity simulations or experimental measurements. Even with modern drift-reduced Braginskii codes, diagnostic comparisons tend to be qualitative with considerable uncertainties in quantitative accuracy [181,180,161,151,184,159,185,46,61], although these efforts are improving [234]. This challenge in precisely ascertaining the reduced tur-bulence theory's predictive reliability is in part due to difficulties in aligning nonlinear calculations with numerous free parameters for chaotic systems [162,128,238].…”

“…Various adaptations of these equations have been recently taken to investigate several important edge phenomena including pedestal physics [56], blob dynamics [179], neutral effects [210], and heat fluxes impinging plasma-facing components [157]. While experimental trends are at times reproduced in these works [234], direct quantitative agreement between the two-fluid turbulence theory and observations is generally lacking on a wide scale due to difficulty in aligning global simulations with plasma experiments where relevant measurements may be sparse or missing altogether. Fusion plasma diagnostic measurements are inherently noisy and limited in their spatiotemporal scope (e.g.…”

Validation of edge turbulence codes against the TCV-X21 diverted L-mode reference case

“…By adding neutral dynamics, we expect that the parallel advection in the divertor will be reduced. The neutrals will also introduce plasma cooling in the vicinity of the X-point and an asymmetry in the density source, which can change the profiles and, thus, the turbulence drive [75].…”

“…This version does not use the Boussinesq approximation, includes electromagnetic effects and couples to a kinetic neutrals model [57]. For GRILLIX, a refactored version will be used for testing boundary conditions, including neutrals and extending the resolution scan [75].…”

Kinetic modeling of neutral transport for a continuum gyrokinetic code

“…These experimental studies have generated interest in coupling models of neutral transport to first-principles plasma turbulence codes. Couplings with fluid neutral models have effectively modeled some scrape-off layer (SOL) conditions, 6,7 but are rigorous only where the neutral mean free paths are shorter than the characteristic length scale. Comparisons between a fluid neutral code and a kinetic model were improved recently with a more accurate numerical treatment of geometry.…”