Comparison of SOLPS5.0 and SOLPS‐ITER simulations for ASDEX upgrade L‐mode

Abstract

This paper examines the backward compatibility of SOLPS-ITER with SOLPS5.0 and produces a basic test of the physics/numerics improvements/additions in SOLPS-ITER recommended by developers, taking an ASDEX Upgrade L-mode simulation as an example. SOLPS-ITER, which is emerging as the most advanced tool for edge plasma modelling, can be instructed to mimic SOLPS5.0 physics/numerics. This allows producing a detailed comparison of the two codes, in a framework where they are expected to produce the similar results, thus raising the confidence in using SOLPS-ITER to continue SOLPS5.0 simulations. Under such framework, SOLPS-ITER results match well with those of SOLPS5.0. The remaining differences might be from the ion energy source, and a full benchmark activity is expected to solve this in the future. We then test the effect of the recommended physics/numerics introduced in SOLPS-ITER with respect to the widely used SOLPS5.0. Only deuterium shots are considered as the basic test, where the recommended physics/numerics are simple and expected not to change the simulation results significantly. Electron density and temperature on divertor targets are the key metrics in this study, instead of particle fluxes and power load. Numerical simulations show that the effect of the recommended physics/numerics on the final solution results in only ∼5% differences in the outer mid-plane and target profiles of electron density and temperature. An upstream density scan, covering the full range from attached to detached conditions, also produces closely matching results (∼10% differences). Thus, we believe that recommended physics/numerics do not introduce unwanted spurious effects and are confident about future modelling results of SOLPS-ITER. K E Y W O R D SL-mode, SOLPS5.0, SOLPS-ITER

SOLPS-ITER modeling of ASDEX Upgrade L-mode detachment states

“…The 96 × 36 computational grid of SOLPS-ITER is the same as discussed in [10]. Sub-divertor structures, including turbopump, cryopump and neutral baffles equals to those in [11,12] are considered.…”

“…In order to validate the neutral transport described by SOLPS-ITER, we compare the modeled neutral flux density against measured value at dome and pump positions in table 2. The positions of gauges are the same as in [10]. For the neutral flux density at the dome, it can be seen that the discrepancy between modeled and measured values is within ∼30%.…”