Reynolds-Averaged Navier-Stokes/Large-Eddy Simulations of Supersonic Base Flow

TLDR

The study discusses the capabilities of various hybrid RANS/LES methods to predict high‑Reynolds‑number supersonic base flows. The authors assessed numerous numerical parameters relevant to hybrid RANS/LES methods. Hybrid RANS/LES approaches were evaluated on 5 × 10⁶ and 13.5 × 10⁶‑point grids, showing that accurate prediction of the highly compressible separated shear layer requires proper boundary‑layer thickness, minimal numerical dissipation, and careful tuning of the C DES constant, while subgrid‑scale models have limited impact when small‑scale structures are resolved.

Abstract

Several zonal and nonzonal hybrid Reynolds-averaged Navier-Stokes/large-eddy simulation approaches have been assessed to handle a high Reynolds number supersonic base flow. The results obtained on 5 and 13.5 x 10 6 points grids are compared to the available experimental data (Herrin, J. L., and Dutton, J. C., Supersonic Base Flow Experiments in the Near Wake of a Cylindrical Afterbody,AIAA Journal, Vol.32, No. 77,1994), and the capabilities of these different methodologies to predict supersonic flows are discussed. The highly compressible separated shear layer proved to be a challenging issue for hybrid methods due to an alteration in the instability process (as compared to the incompressible case), leading to three-dimensional coherent structures. Numerous numerical parameters relevant to hybrid methods have been assessed. The incoming boundary layer thickness needs to be properly set, while a weak influence of the subgrid scale model is observed when small-scale structures in the separating mixing layer are resolved. Another finding of the present study is the dramatic influence displayed by the numerical dissipation on the flowfield. Sensitivity to the C DES model constant is observed even with the finest grid, leading to a delay in the generation of instabilities in the shear layer.