Assessment of transition modeling and detached-eddy-simulation of a wind-turbine airfoil near stall

Abstract

Accurate prediction of transition, separation, and wake unsteadiness remains challenging for wind-turbine airfoils. This study investigates the DU 91-W2-250 wind-turbine airfoil at Re c = 3 × 10 6 using fully turbulent and transitional shear-stress transport (SST) k – ω unsteady Reynolds-averaged Navier–Stokes (URANS), delayed detached-eddy simulation (DDES), and improved delayed detached-eddy simulation (IDDES) computations. URANS simulations are first performed over a range of angles of attack, followed by DDES and IDDES analyses at α = 7.08° and α = 10°. The influence of transition modeling is model- and flow-regime-dependent and does not lead to a consistent improvement in the present predictions. In the attached-flow range, the transitional URANS results lie closer to the reported lift, drag, and quarter-chord pitching-moment measurements than the fully turbulent results, with the clearest change occurring in drag because transition treatment reduces the excess skin-friction drag associated with the fully turbulent assumption. At α = 7.08°, where weak trailing-edge separation occurs, the transitional DDES and IDDES results likewise lie closer to the reported measurements than their fully turbulent counterparts. At α = 10° near maximum lift, the fully turbulent URANS and DDES predictions remain closer to the reported lift and drag values than their transitional counterparts. For IDDES, transition treatment reduces the drag overprediction, but shifts the lift prediction from underprediction to overprediction. Transition modeling mainly modifies the upstream boundary-layer state and skin friction, whereas DDES and IDDES primarily affect trailing-edge separation and the associated shear-layer and wake unsteadiness.