High-Order Accurate Numerical Simulation of Supersonic Flow Using RANS and LES Guided by Turbulence Anisotropy

This paper discusses accuracy improvements to Reynolds-Averaged Navier–Stokes (RANS) modeling of supersonic flow by assessing a wide range factors for physics capture. Numerical simulations reveal complex behavior resulting from shock and expansion waves so, jet emanating rectangular nozzle is considered. PIV based experimental data the available literature used validation purposes. Effect various boundary conditions turbulence approaches assessed qualitatively quantitatively. Of particular interest are inlet considering intensity effect upstream air supply duct, wall surface roughness on internal downstream, y+ sensitivity layer resolution laminar turbulent transition modeling. In addition mesh sensitivity, domain dependency conducted evaluate appropriate size capture kinetic energy dissipation downstream nozzle. To further improve characteristics, accounting anisotropy Reynolds stresses also one focuses. Therefore, non-linear eddy viscosity-based two-equation model stress transport investigated. Additionally, results baseline linear (Boussinesq) RANS compared. Corresponding comparisons with high-fidelity LES presented. Jet self-similar all simulation fidelities it appears that in becomes self-similar, but not RANS. Finally, such as geometry numerical choices influencing discussed.