A penalization technique to model plasma facing components in a tokamak ă with temperature variations
To properly address turbulent transport in the edge plasma region of a ă tokamak, it is mandatory to describe the particle and heat outflow on ă wall components, using an accurate representation of the wall geometry. ă This is challenging for many plasma transport codes, which use a ă structured mesh with one coordinate aligned with magnetic surfaces. We ă propose here a penalization technique that allows modeling of particle ă and heat transport using such structured mesh, while also accounting for ă geometrically complex plasma-facing components. Solid obstacles are ă considered as particle and momentum sinks whereas ionic and electronic ă temperature gradients are imposed on both sides of the obstacles along ă the magnetic field direction using delta functions (Dirac). Solutions ă exhibit plasma velocities (M = 1) and temperatures fluxes at the ă plasma-wall boundaries that match with boundary conditions usually ă implemented in fluid codes. Grid convergence and error estimates are ă found to be in agreement with theoretical results obtained for neutral ă fluid conservation equations. The capability of the penalization ă technique is illustrated by introducing the non-collisional plasma ă region expected by the kinetic theory in the immediate vicinity of the ă interface, that is impossible when considering fluid boundary ă conditions. Axisymmetric numerical simulations show the efficiency of ă the method to investigate the large-scale transport at the plasma edge ă including the separatrix and in realistic complex geometries while ă keeping a simple structured grid. (C) 2014 Elsevier Inc. All rights ă reserved.
Alejandro Paredes, Hugo Bufferand, Guido Ciraolo, Frédéric Schwander, Eric Serre, et al.. A penalization technique to model plasma facing components in a tokamak ă with temperature variations. Journal of Computational Physics, 2014, 274, pp.283-298. ⟨10.1016/j.jcp.2014.05.025⟩. ⟨hal-01464707⟩
Journal: Journal of Computational Physics
Date de publication: 01-10-2014
Auteurs:
-
Alejandro Paredes
-
Hugo Bufferand
- Guido Ciraolo
- Frédéric Schwander
- Eric Serre
-
Philippe Ghendrih
-
Patrick Tamain