SOLPS modelling of ASDEX upgrade H-mode plasma

Abstract

A low density H-mode plasma has been selected for detailed inter-ELM modelling by the SOLPS code package, with the coupled treatment of its plasma (fluid code B2) and neutral (Monte-Carlo code Eirene) parts. Good quality measured midplane density and temperature profiles, covering the pedestal region and stretching far into the SOL, as well as several other parameters and profiles measured in the divertor, have enabled testing the consistency of code solutions with experiment. Once the upstream, midplane profiles have been fitted, and the global parameters (e.g. input power into the computational grid, radiated power) matched, the code reproduces experimental profiles and control parameters in the divertor with an accuracy within a factor of 2. Deviations of modelled parameters from the experiment were found around the strike point position where most of the power was deposited on the target. The deviations are consistent among themselves and all point to one common problem with the modelling: the predicted divertor electron temperature is too low and the density too high, compared with the experiment. The largest inconsistency between the code and experiment was in the magnitude of the peak Hα radiation in the outer divertor, which was larger by a factor of 2 in the code simulations. In addition, the code predicts a somewhat higher sub-divertor neutral flux but lower carbon impurity content in the edge plasma than in the experiment, as well as lower CIII emission. The discrepancy between Hα profiles can to a large degree be attributed to profile effects: the simulated Hα emission profiles are narrower than in the experiment, reflecting the tendency of the neutral–plasma mix to congregate excessively around the strike point in the modelling. At the same time, the integrated Hα emission matches very well with the experiment.Extensive sensitivity studies of the influence of variations in input parameters and assumptions of the code on the modelled divertor conditions have been conducted. They have not resulted in an identification of any SOLPS input/control parameters capable of removing the main disagreement between the code output and experiment. A possibility of parallel transport effects related to low collisionality to increase the effective plasma temperature near the strike point position or of increased perpendicular transport by neutrals (due to some missing reactions in Eirene) to widen the target profiles, will be explored in the future.