RANS Computations of 3D Flow past NASA Common Research Model in High Lift Configuration

Abstract

Accurate prediction of the flow phenomena that occur in high lift flows has been found to be difficult to model with Reynolds-Averaged Navier-Stokes (RANS) equations in conjunction with a turbulence model. The complex interactions among multiple elements of the wing as well as the support hardware can be difficult to predict accurately for the RANS equations with a turbulence model. Three turbulence models are used to evaluate their ability to predict the flow field of NASA Common Research Model in high lift (CRM-HL) configuration. The models are Standard Spalart-Allmaras (SA) model, SA model with Quadratic Constitutive Relation (QCR) SA-QCR2000 and Wray-Agarwal (WA) with QCR WA-QCR2000. Results from the three turbulence models are compared to the experimental data. It is found that the simulations show good agreement with experimental results in the linear range of the lift slope curve. At higher angles of attack above the critical stall angle, the results fail to match the experimental results on multiple meshes from coarse to fine. The computed results on all three grid densities from coarse to medium to fine under-predict the lift and over-predict drag due to separation. In addition, CFD results over-predict the separation region that occurs at high angles of attack, especially towards the tip of the wing. Both SA and SA-QCR2000 models’ predictions are very close to each other, while WA-QCR 2000 model predicts large separation above the critical stall angle.