Application of an automated grid deformation tool for divertor shape optimization in SOLPS-ITER

Abstract

An optimized divertor design is crucial to maximize the lifetime of plasma-facing components and reduce costs of future fusion power plants. Numerical shape optimization could be a powerful tool to obtain improved designs in an automated way. However, it is not trivial to apply due to the need of a field-aligned and boundary-fitted grid for simulating the plasma and quantifying the heat load. This paper shows how a grid deformation tool can automate the gridding process while safeguarding the grid quality. Additionally, sensitivities of shape parameters are computed using finite differences and compared to those obtained by remeshing using standard tools such as CARRE2. The plasma and neutral behavior is simulated using the unstructured SOLPS-ITER code with the latest advanced fluid neutrals model for an ASDEX Upgrade test case. The comparison shows that, contrary to the remeshing strategy, the grid deformation approach yields smoother sensitivities. Furthermore, it is shown that the deformed grids have better mesh quality in terms of poloidal cell size ratio compared to the grid generated with CARRE2, which improves the accuracy of the simulation. This supports the use of the grid deformation tool for automated shape design in future work.