Application and Evaluation of the Wray-Agarwal Turbulence Model with Compressibility Corrections in SU2 for RANS Hypersonic Flow Prediction

Accurate and efficient prediction of aerodynamic heating remains a critical challenge in the design and analysis of hypersonic vehicles, where both computational efficiency and precision are essential. This study evaluates the performance of modified one-equation Reynolds-Averaged Navier-Stokes (RANS) turbulence models, specifically focusing on the Spalart-Allmaras (SA) and Wray-Agarwal (WA) models, including compressibility corrections developed by Catris and Aupoix. Additionally, a newly proposed inner-layer semi-local scaling variable-property correction to the WA model is derived and analyzed, showing promising results across many turbulent boundary layer cases. Cases were selected from a supersonic and hypersonic turbulent boundary layer database of Direct Numerical Simulations (DNS) results, as well as from the NASA hypersonic shock-wave/turbulent-boundary-layer interactions experimental database, which contains a range of 2D and axisymmetric high-speed configurations. Simulations were conducted using the open-source CFD code Stanford University Unstructured (SU2). Turbulent boundary layer results reveal that the velocity profile predictions of the SA model are the most accurate; however each model showed comparable performance for the thermal-velocity coupling profiles when compared to DNS. Comparisons of compression corner model predictions for wall shear stress, pressure, and heat flux with experimental data highlight notable differences in model performance between the SA and WA models with minimal impact from the Catris and Aupoix compressibility modifications. This work underscores the potential for refinement in turbulence modeling to enhance prediction accuracy for complex hypersonic flows and establishes a foundation for future model adjustments in high-speed aerodynamic applications.