Examination of the GEKO Turbulence Model for Hypersonic Turbulent Boundary Layer Applications

Abstract

This study investigates the performance of Reynolds-Averaged Navier-Stokes (RANS) with the Generalized K-Omega turbulence model (GEKO) turbulence model in comparison with Menter's Shear Stress Transport (SST) model for predicting hypersonic turbulent boundary layer flows. Two configurations are considered: a Mach 4.9 zero-pressure-gradient (ZPG) flat-plate boundary layer and a Mach 4.9 curved-wall turbulent boundary layer. Both configurations were simulated using GEKO and SST models in ANSYS Fluent, and the relative accuracy of each model in predicting wall quantities, mean flow profiles, Reynolds Stresses, and turbulent heat fluxes was determined by comparing against the high-fidelity Direct Numerical Simulation (DNS). For the Mach 4.9 flat-plate case, the GEKO model demonstrated superior agreement with DNS for most flow quantities, particularly in capturing wall shear stress, thermal gradients, and Reynolds stresses. The SST model tended to overpredict wall heat transfer and skin friction values, offering more conservative estimates but less accuracy in matching DNS trends. In contrast, for the curved wall configuration, the SST model exhibited better consistency with DNS in regions influenced by pressure gradients, especially for wall-normal and shear stresses, though both models GEKO and SST produced nearly identical trends in mean profiles and turbulent fluxes. The analysis shows that GEKO offers flexibility through tunable parameters, it occasionally overestimates turbulent fluctuations and wall heat fluxes, particularly in regions of pressure gradient. Overall, both models capture the essential physics of hypersonic boundary layer development, though differences in quantitative accuracy emphasize the importance of turbulence model selection based on application needs. These findings support the applicability of GEKO for high-speed modeling and highlight the validation role of DNS.