The working conditions of airfoils along modern wind turbine blades are putting new focus on the importance of properly characterizing the aerodynamic performance of different airfoil families also at high angles of attack (AoAs) beyond stall and at Reynolds numbers much lower (from few thousands to one million) than those commonly analyzed before. Several test cases are showing that even higher-order computational methods (like RANS/URANS CFD) are unable to properly capture the complex flow physics taking place past the blades, when deep stall occurs or when the AoA changes so rapidly to provoke the onset of dynamic stall. To fill this gap, the use of high-fidelity methods, like the Large Eddy Simulation (LES) is proposed, even though it implies a massive increase of the calculation cost. In order to analyze the prospects of using LES in comparison to RANS for low Reynolds, high AoAs, this work presents an in-depth study of the NACA 0021 aerodynamics at the Reynolds number of 80,000, by means of both traditional RANS approaches and high-fidelity (LES) simulations using the OpenFOAM suite. The selected airfoil has been showing in fact several issues in the correct characterization of its performance in similar conditions in many recent wind energy applications. The LES approach showed the ability to overcome the limitations of traditional RANS simulations, improving the accuracy of the results and reducing their dispersion thanks to the fact that the flow structures in the separated-flow regions are properly captured. Overall, this work underlines that accurate investigations of the aerodynamic performance of the NACA 0021 at low Reynolds require multiple sensitivity studies when RANS approaches are used, and suggests the use of LES simulations in order to increase the accuracy of estimations, especially when studying the stalledflow operating conditions of the airfoil.
On the prospects of improving the numerical analysis of symmetric airfoils at low Reynolds numbers and high angles of attack by a LES approach: the case of NACA0021 profile