Parallel Multigrid Acceleration of 3D Compressible Flows on Unstructured Meshes

摘要

the itnerest in using unstructured meshes for Computational Fluid Dynamics (CFD) applications appears to be increasingly important in the industrial community. Concerning flow solvers, the main question appears to be the lack of efficiency demonstrated by unstructured mesh solvers compared to structured ones. Part of this loss of efficiency has been recovered by developping parallel versions of unstructured mesh solvers taking benefit of the potentially high performance offered by current generation distributed memory MIMD architectures. Clearly, resorting to parallel computing greatly contributes to reduce the overall simulation time however this has to be done in close relation with improvements on the numerical efficiency of the underlying solution methods. This parper is concerned with the construction and evaluation of parallel linear multigrid algorithms for the acceleration of 3D compressible flow calculations on unstructured finite element meshes. A parallel multigrid by agglomeration solution strategy is implemented in a dedicated CFD software and applied to the numerical simulation of complex flows of industrial relevance.