Reynolds Averaged Navier Stokes Models Compared to Direct Numerical Simulations in an Adverse Pressure Gradient Boundary Layer Over a Flat Plate

Abstract

In Reynolds-Averaged Navier Stokes (RANS) models simplifying assumptions breakdown in near wall regions. Wall functions/treatments become inaccurate and the homogeneity and isotropy models may not hold. To see the effect that these assumptions have on the validity of boundary layer results in a commercially available RANS code, key boundary layer parameters are compared from laminar, transitional, and fully turbulent RANS results to an existing direct numerical simulation (DNS) simulation for flow over a flat plate with an adverse pressure gradient (APG). Parameters compared include velocity profiles in the free stream, boundary layer thicknesses, skin friction coefficient and the pressure gradient parameter. Results show comparable momentum thickness and pressure gradient parameters between the transition RANS model and the DNS simulation. Differences in the onset of transition between the RANS transition model and DNS are compared as well. These simulations help evaluate the models used in the RANS code. Of most interest is the transition model, a transition shear-stress transport (SST) k–omega model. The RANS code is being used in conjunction with an APG boundary layer experiment being undertaken at the Idaho National Laboratory (INL).