Gyrokinetic projection of the divertor heat-flux width from present tokamaks to ITER

Abstract

The XGC1 edge gyrokinetic code is used to study the width of the heat-flux to divertor plates in attached plasma condition. The flux-driven simulation is performed until an approximate power balance is achieved between the heat-flux across the steep pedestal pressure gradient and the heat-flux on the divertor plates. The simulation results compare well against the empirical scaling λ q ∝1/B P γ obtained from present tokamak devices, where λ q is the divertor heat-flux width mapped to the outboard midplane, γ=1.19 as found by T. Eich et al. [Nucl. Fusion 53 (2013) 093031], and B P is the magnitude of the poloidal magnetic field at the outboard midplane separatrix surface. This empirical scaling predicts λ q ≲1mm when extrapolated to ITER, which would require operation with very high separatrix densities (n sep /n Greenwald > 0.6) [Kukushkin, A. et al., Jour. Nucl. Mat. 438 (2013) S203] in the Q=10 scenario to achieve semi-detached plasma operation and high radiative fractions for acceptable divertor power fluxes. Using the same simulation code and technique, however, the projected λ q for ITER's model plasma is 5.9 mm, which could be suggesting that operation in the ITER Q=10 scenario with acceptable divertor power loads may be obtained over a wider range of plasma separatrix densities and radiative fractions. The physics reason behind this difference is, according to the XGC1 results, that while the ion magnetic drift contribution to the divertor heat-flux width is wider in the present tokamaks, the turbulent electron contribution is wider in ITER. Study will continue to verify further this important projection. A high current C-Mod discharge is found to be in a mixed regime: While the heat-flux width by the ion neoclassical magnetic drift is still wider than the turbulent electron heat-flux width, the heatflux magnitude is dominated by the narrower electron heat-flux.

A fast low-to-high confinement mode bifurcation dynamics in the boundary-plasma gyrokinetic code XGC1

“…The scheme allows ambipolar electron and ion loss to the material wall, which results in neutral atomic particle recycling back into the plasma, but still preserving the phasespace volume of the charged particles. The scheme has been parallelized and optimized to reduce data communications in extreme-scale parallel computers, and applied to study the present L-H bifurcation study, 31 the divertor heat-flux footprint study, 62 and other boundary physics. A fast, forced bifurcation of turbulence and transport has been observed for the first time in an electrostatic nonlinear gyrokinetic simulation.…”

“…In the present study, we make the simulation possible by reducing the computational resource requirement as much as possible via a model simplification; i.e., by choosing a fast electrostatic bifurcation case under strong forcing by a high rate of edge heat deposition without prolonging it to the slow, follow-on pedestal build up process. XGC1 simulations and analytic study show that edge turbulence saturation is usually established within 0.1 ms, 33,62 while in the core plasma, the nonlinear turbulence saturation is established in 1 ms or longer.…”

Pedestal and edge electrostatic turbulence characteristics from an XGC1 gyrokinetic simulation

“…The topic of blob generation is of particular importance for predicting the divertor heat flux width on ITER, as recent XGC1 simulations suggest the relative importance of blobby to neoclassical transport leads to a larger (i.e. safer) divertor heat load width on ITER compared to extrapolations using current-day experiments [7].…”

“…The present study uses a realistic DIII-D-like magnetic equilibrium, including lower single-null X-point, with q 95 ≈ 3.5 and B 0 ≈ 2.1 T. Collisions and neutrals were turned off in this particular report, in order to remove the Coloumb collision and neutral particle effects for simplicity, though these are typically used in divertor heat-flux studies in Refs. [7] and [8]). Future reports will study these effects on blobby turbulence one by one.…”

Fast Low-to-High Confinement Mode Bifurcation Dynamics in a Tokamak Edge Plasma Gyrokinetic Simulation

“…In the present study, we make the simulation possible by reducing the computational resource requirement as much as possible via a model simplification; i.e., by choosing a fast electrostatic bifurcation case under strong forcing by a high rate of edge heat deposition without prolonging it to the slow, follow-on pedestal build up process. XGC1 simulations and analytic study show that edge turbulence saturation is usually established in ≲0.1ms [34,31], while in the core plasma nonlinear turbulence saturation is established in ≳1 ms.…”