Performance of turbulence models for transonic flows in a diffuser

Eight turbulence models frequently used in aerodynamics have been employed in the detailed numerical investigations for transonic flows in the Sajben diffuser, to assess the predictive capabilities of the turbulence models for shock wave/turbulent boundary layer interactions (SWTBLI) in internal flows. The eight turbulence models include: The Spalart-Allmaras model, the standard κ ϵ-model, the RNG k-model, the realizable κ-model, the standard κ ϵ model, the SST κ ω model, the u²-f model and the Reynolds stress model. The performance of the different turbulence models adopted has been systematically assessed by comparing the numerical results with the available experimental data. The comparisons show that the predictive performance becomes worse as the shock wave becomes stronger. The v2-f model and the SST κ-model perform much better than other models, and the SST κ-model predicts a little better than the v2-f model for pressure on walls and velocity profile, whereas the v2-f model predicts a little better than the SST κ model for separation location, reattachment location and separation length for strong shock case.