Orateur: Clinton Groth, Institute for Aerospace Studies, University of Toronto
Titre: Parallel High-Order Finite-Volume Method with Anisotropic Block-Based AMR for Multi-Scale Physically-Complex Flows
A high-order central essentially non-oscillatory (CENO) finite-volume scheme combined with a block-based anisotropic adaptive mesh refinement (AMR) algorithm is proposed for the solution of multi-scale physically-complex flows on three-dimensional multi-block body-fitted mesh consisting of hexahedral elements. The cell-centered CENO method uses a hybrid reconstruction approach based on a fixed central stencil. Smooth and fully resolved solution data is represented using an unlimited high-order k-exact reconstruction. In cells deemed to have under-resolved/discontinuous solution content based on a smoothness indicator, the high-order reconstruction reverts to a lower-order limited linear scheme. The high-order CENO finite-volume scheme is implemented within a flexible multi-block hexahedral grid framework. Highly efficient parallel implementation and local anisotropic grid adaptivity are achieved by using a hierarchical block-based domain decomposition strategy in which the connectivity and refinement history of grid blocks are tracked using a binary tree data structure. Physics-based refinement criteria as well as criteria based on the smoothness indicator are used for directing the mesh refinement. Numerical results for several compressible inviscid and viscous flows, ideal magnetohydrodynamics flows, as well as reactive flows are all described in order to demonstrate the accuracy and efficiency of the proposed high-order AMR finite-volume method for a range of problems.