A modified partially averaged Navier-stokes model for the turbulent flows over a backward facing step

In current study, a partially averaged Navier-stokes model based on a modified SST k-? model is proposed (MSST PANS) to predict the turbulent flows with flow separations, recirculation and reattachment. A benchmark case, turbulence flows over the backward facing step (Re=50000), is treated to evaluate the capacity of the MSST PANS model. Several other turbulence models, i.e., the standard k-? model, SST k-? model, the standard k-? PANS (SKE PANS) model and SST k-? PANS (SST PANS) model are also implemented for comparisons with some available experimental data. Results show that the MSST PANS model performs the closest results in predicting the reattachment length as well as the corner vortex. Furthermore, the MSST PANS model yields improved statistics on the skin frictions, pressures, velocity profiles together with Reynolds stresses. In contrast to the SST PANS models, the modifications on the eddy viscosity and ? equation with consideration of streamline curvature show promising in capturing the small-scale flow separations, recirculation and reattachments in some industrial applications.