Fusion Science and Technology / Volume 60 / Number 1T / July 2011 / Pages 48-55
Although the most complex currently available fluid-neutral Monte-Carlo plasma boundary code package, SOLPS, has been a major player in the ITER divertor design, it has not yet been systematically used for the study of kinetic phenomena such as ELM transients. This paper investigates the relevance of fluid code results for transients, in particular at the targets where kinetic effects are most manifest, by comparing power and particle fluxes at the targets from SOLPS5 time-dependent simulations of TCV Type III ELMs with those obtained from dedicated Particle-in-Cell (PiC) kinetic transport code (BIT1) simulations. Although reasonable agreement is found in terms of the absolute magnitude of total heat fluxes, the arrival of the ion pulse at the target from upstream is significantly faster in SOLPS than expected on the basis of sonic transit times (as also seen in PiC). Adjustments of kinetic heat flux limiters to render the heat fluxes more convective in SOLPS are necessary in order to correct for this discrepancy. Moreover, because SOLPS does not account for the transfer of heat from electrons to ions inside the sheath, correction terms to the electron and ion power fluxes at the targets are required in SOLPS in order to better match PiC results. However, it does not appear possible within the scope of these sensitivity studies to simultaneously achieve expected delays and ion-electron power sharing in the fluid simulations.