Parallel Adaptive Mesh Refinement Scheme for LES of Turbulent Premixed Flames

Abstract

A block-based adaptive mesh refinement (AMR) algorithm is proposed for performing large-eddy simulations (LES) of turbulent premixed combustion using body-fitted multi-block hexahedral computational mesh. The AMR framework is combined with a high-resolution finite-volume scheme with limited linear reconstruction and a parallel implicit time-marching scheme to solve the Favre-filtered Navier-Stokes equations for a thermally perfect compressible reactive mixture. A flexible block-based octree data structure is used to facilitate automatic solution-directed mesh adaptation according to physics-based refinement criteria. The data structure also enables an efficient and scalable parallel implementation via domain decomposition. The use of body-fitted mesh permits the use of anisotropic grids and resolution of thin boundary, mixing, and shear layers. The implicit formulation makes use of a dual-time-stepping-like approach, Jacobian-free inexact Newton method, and preconditioned generalized minimal residual (GMRES) algorithm to solve the system of nonlinear algebraic equations arising from the temporal discretization procedure. An additive Schwarz global preconditioner is used in conjunction with block incomplete LU type preconditioners for each sub-domain. Both the thickened-flame and flame-surface-density subfilter-scale models are considered for representing turbulence-chemistry interactions. Numerical results for a methane-air slot burner demonstrated the potential of the approach for performing LES of turbulent premixed flames.