Numerical and experimental investigation of low Reynolds number wind turbine airfoils under stall and post-stall conditions

This work concerns a numerical and experimental investigation of low Reynolds number wind turbine airfoils under stall and post-stall conditions. A verification of high angle of attack semi empirical models, namely Viterna-Corrigan and Aerodas from Spera, is needed to assess the input data reliability for Blade Element Momentum (BEM) based Vertical Axis Wind Turbine (VAWT) performance code and also the performance predictions for Horizontal Axis Wind Turbine (HAWT) where passive stall control must be guaranteed by a reliable post-stall aerodynamic coefficients distribution. A great challenge is inherited by the low Reynolds number related to the phenomenon (considering VAWT for urban installation and small size generation) and the presence of possible laminar bubble burst induced stall. In case for which no laminar bubble promotion is observed on the stall behavior, still a strong dependency on turbulence intensity is present for the maximum lift coefficient attainable. Attention is also paid to not clean VAWT operating conditions as those occurring far from maintenance period in which a transition from laminar to turbulent flow regime could be supposed to be promoted by dirty material accumulation on the blade leading edge region. Computational Fluid Dynamics (CFD) simulations using OpenFOAM library and a transitional turbulence model are used as a comparison for the experimental results derived from the test campaign conducted in the closed test section wind tunnel facility of the Department of Industrial Engineering of University of Naples, for both free and fixed transition condition. A further comparison with Delft University experiments is performed for similar Reynolds number, and moreover, an analysis of blockage effect by means of another test campaign in the open test section wind tunnel facility is conducted to understand the influence of wind tunnel walls on secondary lift coefficient peak and maximum drag coefficient value.