We present an attempt to improve the predictive accuracy of RANS turbulence models for low-speed flow separation on airfoils and wings near stall. It is based on a proposal for a new wall-law for the mean velocity in the inner layer of a turbulent boundary layer at adverse pressure gradient The wall-law is a composite law with a log-law in the inner part and a half-power (or sqrt-) law in the outer part of the inner layer. The wall-law is calibrated using a large experimental database. The slope coefficient for log-law and sqrt-law and the wall distance for transition between them depends on the pressure gradient parameter. We describe the corresponding wall-law for the turbulent viscosity and for it and ω in k-ω RANS turbulence models. We propose a modification for the k- and ω-equation such that the wall-law for the inner layer is satisfied. We apply the modified SST model to a new turbulent boundary layer experiment and to airfoil flows near maximum lift at different Reynolds numbers. The modification leads to a reduction of the skin-friction coefficient in the adverse pressure gradient region and causes an upstream shift of the separation line.
展开▼