Modeling the effect of shock unsteadiness in shock-wave/turbulent boundary layer interactions

Reynolds averaged Navier-Stokes methods often cannot predict shock/turbulence interaction correctly. This may be because RANS models do not account for the unsteady motion of the shock wave that is inherent in these interactions. Sinha et al. [Phys. Fluids, Vol. 15, No. 8 (2003)] propose a shock-unsteadiness correction that significantly improves turbulence prediction across a normal shock in a uniform mean flow. In this paper, we generalize the modification to simulate complex flows using k-ε, k-ω, and Spalart-Allmaras models. In compression-corner flows, the corrected k-ε and k-ω models amplify the turbulent kinetic energy less through the shock compared to the standard models. This results in Improved prediction of the separation shock location, a delayed reattachment, and a slower recovery of the boundary layer on the ramp. In the Spalart-Allmaras model, the modification amplifies eddy viscosity across the shock, moving the separation location closer to the experiment.