Yongliang Feng / Laboratoire de Mécanique, Modélisation et Procédés Propres
Shaolong Guo / Laboratoire de Mécanique, Modélisation et Procédés Propres
Pierre Sagaut / Laboratoire de Mécanique, Modélisation et Procédés Propres
The recent progress in computational fluid dynamics (CFD), and more specifically the lattice Boltzmann method (LBM) along with the tremendous increase of power on high-performance computing enable today to perform high fidelity simulation on real industry applications. Due to the advantages of LBM for massively parallel computing as well as its ability to handle very complex geometries, LBM has gained popularity as a promising approach for CFD.
An efficient lattice Boltzmann (LB) model relying on a hybrid recursive regularization (HRR) collision operator on D3Q19 stencil with an improved thermal equilibrium distribution function is proposed for the simulation of three-dimensional high-speed compressible flows in both subsonic and supersonic regimes. Mass and momentum equations are recovered by an efficient streaming, collision and forcing process on D3Q19 lattice. Then a non-conservative formulation of the entropy evolution equation is used, that is solved using a finite volume method.
The compressible HRR-LB method is assessed considering the simulation of i) 2D supersonic flow over a bump, ii) 2D laminar flows over a flat plate at a range of Ma numbers, iii) unsteady viscous transonic flow over NACA0012 airfoil and iv) 3D transonic flows over airfoil and aircraft.