Leading-Edge separation behaviors in SA RANS and SA-based DDES: Simple modifications for improved prediction

Abstract

In this study, delayed detached-eddy simulations (DDESs) based on the Spalart–Allmaras turbulence model are investigated for separation flows. Three simple modifications are considered: the Reynolds-averaged Navier–Stokes (RANS) turbulence model coefficient, Crot, is calibrated to achieve a better leading-edge separation prediction in DDES; the large-eddy simulation (LES) coefficient, CDES, is assessed to obtain better dissipation control in DDES; and dirty-cell treatments in three-dimensional unstructured grids are conducted for a smooth LES/RANS transition. Numerical results confirm the effects of the three aforementioned steps, such as the reproducibility of the measured pressure distribution over the main wing in unsteady turbulence simulations of low-speed buffet around the NASA Common Research Model. Thus, these modifications will potentially serve as good alternatives, without major programming efforts, to the conventional approaches for practitioners.