Lattice Boltzmann k-ω SST based hybrid RANS/LES simulations of turbulent flows
High Reynolds number simulations with lattice Boltzmann method (LBM) have been computationally feasible by means of turbulence models for under-resolved physical scales. Hybrid RANS/LES models are efficient alternatives to accurate but computationally expensive LES techniques. This study investigates extensively the hybrid RANS/LES models in the LBM framework for complex wall-bounded turbulent flow simulations on non-body fitting Cartesian grids. The hybridization processes for these models are applied to a widely used two-equation RANS base model, specifically the k-ω SST turbulence model. The functionality of these hybrid models is validated through the simulation of a three-dimensional fully turbulent flow over an airfoil. In addition, they are assessed by performing LBM simulations of flows surrounding a generic vehicle geometry, the Ahmed body, in order to reproduce the solution of a similar simulation by conventional CFD methods and to compare with wind tunnel experimental data. It is found that the near-wall treatments have significant impacts on the local boundary flow structures and possible remedies are also discussed.
Sajad Mozaffari, Shang-Gui Cai, Jérôme Jacob, Pierre Sagaut. Lattice Boltzmann k-ω SST based hybrid RANS/LES simulations of turbulent flows. Journal of Computational Physics, 2024, pp.113269. ⟨10.1016/j.jcp.2024.113269⟩. ⟨hal-04921212⟩
Journal: Journal of Computational Physics
Date de publication: 01-07-2024
Auteurs:
-
Sajad Mozaffari
-
Shang-Gui Cai
- Jérôme Jacob
- Pierre Sagaut