Multi-fluid and dynamic simulation of tungsten impurity in tokamak boundary plasma

Abstract

Accumulation of tungsten (W) in core is a serious challenge for achieving high-performance plasmas in future tokamak reactors, thus W impurity transport is a highly concerned topic in the tokamak physics researches. Multi-fluid model and kinetic model are the numerical tools widely used for investigating and/or predicting impurity behaviors in the boundary of tokamak plasma. Generally, the applicability of multi-fluid model for impurity transport modeling requires that the collision mean-free-path should be smaller than the gradient scale lengths of particles, which may not be always satisfied. It is performed and comparatively investigated to evaluate the applicability of multi-fluid model for W impurity transport modeling, multi-fluid (SOLPS-ITER) modeling and kinetic (DIVIMP) modeling of W impurity transport in the edge of high-confinement plasma in Experimental Advanced Superconducting Tokamak (EAST) during neon impurity seeding. It is found that low-charge-state W ions are mainly located in the divertor region near the target plate where plasma collisionality is relatively high due to the relatively low/high local plasma temperature/density. Hence, the fluid assumption for transport of lowly-charged W ions can be well satisfied. Consequently, the density of lowly-charged W ions predicted by SOLPS-ITER and that calculated by DIVIMP are almost similar. Owing to the fact that the density of highly-charged W ions is relatively low and these particles mainly exist in the upstream (e.g. the main SOL and core) where plasma collisionality is relatively low, the fluid approximation cannot be well satisfied. However, the total W impurity density calculated by the kinetic code DIVIMP and the multi-fluid model SOLPS-ITER are found to be in agreement with each other within a factor of 1.5 for the simulation cases presented in this contribution. Besides, the multi-fluid simulation with bundled charge state model has also been performed, the obtained results are compared with those from the multi-fluid modeling with W ions treated as 74 fluids. It is revealed that in simulation cases with neon impurity seeding and with divertor plasmas in high-recycling or partially detached regimes, the bundling scheme, which is commonly used for saving the computation cost in multi-fluid modeling, tends to overestimate the average charge state of W ions and thus tends to underestimate the radiation power loss, especially in the divertor region. Consequently, under the circumstance that W impurity radiation dominates the radiative power loss in divertor region, plasma temperature/density can be largely overestimated/underestimated, leading to the underestimation of W ion ionization source and W impurity density. Moreover, simulation results demonstrate that W accumulation in core can decrease effectively during divertor detachment promoted by neon seeding.