Parallel Adaptive Mesh Refinement Simulation of the Flow Around a Square Cylinder at Re = 22000

In this paper a parallel adaptive mesh refinement for LES simulation of turbulent flows is presented. The AMR scheme applies a cell- based refinement technique to get enough grid-resolution to solve the small scales structures, adapting the mesh according to physics- based refinement criteria. A flexible tree data structure is used to keeping track of the mesh adaptation and an edge-based data structure to save and search cell connectivity. The AMR framework is combined with parallel algorithms for partitioning and balancing of the computational mesh. Numerical results for turbulent flow around a square cylinder at Re = 22000 are compared to experimental data.