Divertor plasma conditions and their effect on carbon migration in the ASDEX Upgrade tokamak

摘要

Divertors play a critical role in power and particle exhaust and plasma-material interaction in tokamaks. Of particular concern is the release of impurities from the material components, which reduces the lifetime of the vessel wall and affects the fusion performance. Furthermore, impurities with low charge number can migrate and re-deposit forming potentially tritium-rich layers in the tokamak. Sophisticated numerical simulations are required to understand how the various complex and often nonlinear processes in the plasma boundary affect the divertor performance. This thesis investigates the plasma conditions and the migration of impurities in the divertor region, using numerical code packages with the most complete available description of the relevant physical processes. It is known from earlier studies that the codes have both weaknesses and uncertainties in their models, which limits our current predictive capabilities. In this thesis work, carbon injection experiments at the ASDEX Upgrade tokamak are modelled. The work assesses the present-day boundary plasma models and elucidates impurity migration processes by comparing the simulations with experimental data. The results show that both the local migration and the net erosion of carbon are sensitive to the divertor plasma conditions. Good agreement between the modelled and measured divertor plasma parameters is obtained for a limited operational regime at low density. In this regime, the measured carbon migration is also reproduced in the simulations, and it is shown to be significantly affected by the electric field and the collisionality of the plasma in the divertor region. A significant improvement in code-experiment agreement is obtained after drifts arising from electric and magnetic fields are included in both plasma and impurity simulations. At higher plasma densities, collisionless hot electrons are a potential explanation for the deficiencies observed when benchmarking the plasma models against experiments.