Abstract.A European project was undertaken to improve the available SOL ICRF physics simulation tools and confront them with measurements. This paper first reviews code upgrades within the project. Using the multi-physics finite element solver COMSOL, the SSWICH code couples RF full-wave propagation with DC plasma biasing over "antenna-scale" 2D (toroidal/radial) domains, via non-linear RF and DC sheath boundary conditions (SBCs) applied at shaped plasma-facing boundaries. For the different modules and associated SBCs, more elaborate basic research in RF-sheath physics, SOL turbulent transport and applied mathematics, generally over smaller spatial scales, guides code improvement. The available simulation tools were applied to interpret experimental observations on various tokamaks. We focus on robust qualitative results common to several devices: the spatial distribution of RF-induced DC bias; left-right asymmetries over strap power unbalance; parametric dependence and antenna electrical tuning; DC SOL biasing far from the antennas, and RF-induced density modifications. From these results we try to identify the relevant physical ingredients necessary to reproduce the measurements, e.g. accurate radiated field maps from 3D antenna codes, spatial proximity effects from wave evanescence in the near RF field, or DC current transport. Pending issues towards quantitative predictions are also outlined.
ICRF antennas as active PlasmaFacing ComponentsThe phased strap arrays used to launch Ion Cyclotron Range of Frequency (ICRF, 30-80MHz) waves into magnetic fusion devices poorly radiate in vacuum: efficient coupling of the fast wave to the main plasma necessitates minimizing the distance from the straps to a critical peripheral density: the R-cutoff layer. In the Scrape-Off Layer (SOL) the antennas then behave as Plasma-facing Components (PFCs), subject to PlasmaSurface Interactions (PSI). As RF-field-emitting structures, ICRF antennas are active PFCs able to create RF-specific PSI and to modify locally their environment: enhanced ion energies, heat loads, erosion and density modifications have been observed locally. These phenomena are critical in the prospect of long-pulse machines with high-Z plasma-facing materials. Predicting the magnitude and spatial location of these processes, in relation with plasma parameters, launcher design and electrical settings, remains challenging. So far, realistic tokamak predictions relied on very simple models of oscillating double probes. A European project was undertaken to improve the available SOL RF physics simulation tools and confront them with measurements. This paper first reviews code improvements and associated basic research within the project. The available simulation tools are then applied to interpret experimental observations on various machines. From these results we exhibit qualitative properties common to the various devices, identify relevant physical ingredients necessary to reproduce the measurements, as well as pending issues towards quantitative predictions.
I...