Search for a command to run...
The present work entails investigation of the unsteady aerodynamic characteristics of a wind turbine airfoil operating in four stall-induced vibration (SIV) regimes, including normal (SIV-1) and reverse (SIV-2) flow conditions. We conducted computational fluid dynamics simulations using OpenFOAM (OF) over the full angle-of-attack range (180° to +180°) and compared the results with those of semi-empirical dynamic stall models obtained using BLADED software, including the Øye, Beddoes–Leishman (BL), and IAG models. Numerical verifications confirmed that both conformal and non-conformal meshes could reproduce the key aerodynamic trends with limited sensitivity, except near the stall onset. In the SIV-1 regime, the OF simulations accurately captured hysteresis loops compared to wind tunnel data; the IAG model demonstrated superior performance among the engineering models, while the Øye and BL models significantly underestimated the hysteresis characteristics. In the SIV-2 regime representing reverse flow conditions relevant to parked or idling wind turbines, pronounced asymmetries were observed between the positive and negative ranges owing to directional vortex evolution and geometric effects of the airfoil. The OF results revealed early separation, irregular vortex shedding, and wide hysteresis loops that could not be reproduced by the existing semi-empirical models without additional treatments. Flow-field analysis confirmed that the reverse flow aerodynamics is dominated by coherent and directionally dependent vortical structures, suggesting that conventional definitions of the stall onset are inadequate in the SIV-2 regime. We observed that setting the critical stall normal force to zero improved the prediction of dynamic stall for the IAG model, although the strong amplitudes of the unsteady loads remained underrepresented. Overall, our findings highlight the need for advanced modeling frameworks capable of capturing the asymmetric vortex-driven phenomena in reverse flow regimes that are critical to standstill load assessments.