Recent numerical developments in scrape-off-layer global fluid simulations using the GBS code
Turbulence in the scrape-off-layer (SOL) of magnetic fusion devices is one of the most outstanding issues in magnetic fusion. This open fied lines region determines the boundary condition of the core plasma and controls the plasma refueling, heat losses and impurity dynamics, therefore governing the fusion power output of the entire device. In this work, we present the global fluid code GBS [Ricci et al., Plasma Phys. Control. Fusion 54, 124047, 2012]. It employs a 5 field drift-reduced Braginskii model for both electrostatic and electromagnetic turbulence in a limited configuration. The model simulates a turbulent steady state resulting from plasma sources mimicking the plasma outflow from the core, turbulent perpendicular transport and parallel losses at the limiter sheaths. Unique features of the code are that gradients are a-priori unknown and there is no separation between the background gradient and the fluctuations. We will focus on recent advances to extend GBS from an infinite aspect ratio model to a general geometry model. One of the main features of SOL turbulence is its strong anisotropy characterized by k// /k⊥ << 1, It is therefore crucial to correctly describe the parallel gradient derivative, in particular at the limiter plates where the plasma is lost. In view of more complicated situations such as a diverted geometry, the GBS code does not employ field- aligned coordinates. The fluid fields are discretized on a toroidal and poloidal grid. Using alternative schemes that will be presented. These schemes are tested in a simplified model describing the propagation of shear-Alfven waves, which is the fastest wave propagating for this simulation model. Then, GBS nonlinear simulations using these new schemes will be presented and compared. Among those, we will also discuss some ot the simulation results, focusing on circular geometry with finite aspect ratio. In particular, it is shown that the characteristic pressure length can be well described by the gradient removal theory [Ricci et al., Phys. Plasmas 20, 010702, 2013] that uses the flattening of the gradient by the perturbation as a saturation mechanism.