Anisotropic adaptive body-fitted meshes for CFD

Abstract

Mesh adaptation for immersed solid is one of the most challenging topics in computational mechanics. In this contribution, we propose a novel anisotropic adaptive body-fitted mesh method in particular for Computational Fluid Dynamics (CFD) where the boundary layers and the high gradient of a solution at a fluid–solid interface play a major role. It is characterized by its simplicity to implement, and by its generality to tackle complex geometries without additional efforts. It combines two successive iterations, knowing that the order of each one is important. First, a gradient based metric construction uses the gradients of the level-set of any immersed object to generate an anisotropic well adapted mesh. Since the elements are well stretched along the interface, and therefore the interface passes through these elements, a second step is implemented and applied. It is based on R-adaptation and swapping, to provide a sharp anisotropic fitted mesh. The proposed test cases, with and without a CFD solution, show the flexibility and the accuracy of the proposed automated h- and r- adaption framework designed for immersed methods.