Numerical Investigation of Prediction on the NASA High-Lift Common Research Model

Abstract

An assessment of a hybrid Reynolds-averaged Navier–Stokes/large-eddy simulation (HRLES) approach for prediction is presented for the NASA High-Lift Common Research Model (CRM-HL). Both free-air and wind tunnel configurations of the CRM-HL are investigated, and the results are compared to the QinetiQ wind tunnel experiments and two other numerical approaches: RANS and wall-modeled LES (WMLES). For the free-air configuration, HRLES is shown to address some of the known shortcomings with RANS and prevent inboard and outboard flow separation particularly in the region of and poststall. To achieve these improvements over RANS, LES-appropriate grids and numerical discretizations are required. It was also found that, when applying HRLES to a RANS best practice grid and numerics, the HRLES method significantly underperformed RANS. For the in-tunnel configuration, HRLES showed good agreement with the loads, surface pressure, and oil-flow photographs obtained in the experiment. HRLES was able to improve upon the RANS simulations, which showed a sharp loss of lift at the two highest angles of attack due to large-scale inboard and outboard separation on the wing, by correctly predicting the corner flow separation and showing remarkably close agreement in the flow topologies with the experiment.