Stochastic parameterized modeling for aerodynamic design of wind turbine airfoil

Abstract

Aerodynamic optimization of the airfoil is of great significance to the shape design for wind turbine blade. However, it is tough to calculate the aerodynamic forces of airfoil with random shape variations. In response to this challenge, a stochastic parameterized model based on uncertainty quantification method is proposed. And the influence of non-deterministic airfoil geometry caused by designed parameters on the aerodynamic characteristics is evaluated through the combination of sparse grid-based polynomial chaos (SGPC) and computational fluid dynamic (CFD) taking two airfoils at different Reynolds numbers as examples. OpenFOAM is employed as CFD solver and connected with the in-house stochastic parametric analysis code in a non-intrusive way. Finally, the aerodynamic responses of lift, drag, as well as the surface pressure coefficients are obtained, and the sensitivity of the lift and drag coefficients to the stochastic designed parameters are evaluated. It can be concluded that for the same type airfoil, the sensitive parameters after stall behavior are effected to some extent by increasing Reynolds number, that is, the impact of the maximum curvature index on pressure surface is weakened. As for two kinds of airfoils, S809 is sensitive to parameters near the maximum curvature of the airfoil and the trailing edge of the suction surface under attached flow, while convert to those near the leading edge and maximum curvature of suction face once stalled. FFA-W3-241 airfoil has no sensitivity factors in attached flow and features the same sensitive indicators after stall. These results will provide a reference for coarse-to-fine strategy in optimizing the sample library which is generally the first step of airfoil design process.