One-way fluid–structure interaction modeling and multi-objective optimization of horizontal-axis wind turbine blades equipped with winglets

This study aims to investigate the effects of backward-directed winglets applied on the NREL Phase-II reference wind turbine geometry on structural and aerodynamic characteristics and to reveal optimized configurations. The aerodynamic pressure distributions on the blades were imposed as boundary conditions for the FEM analyses since the structural characteristics of the turbine were analyzed by adopting a one-way Fluid Structure Interaction (FSI) approach. The Taguchi analysis with an L27 orthogonal array revealed the descending order of importance of design variables on the coefficient of power (CP): height, cant angle, toe angle, curvature radius, and finally twist angle. For von-Mises stress, the descending order of importance is height, cant angle, toe angle, twist angle, and curvature radius. A multi-objective genetic algorithm using artificial neural network (ANN) models was used to optimize the objectives CP and von-Mises stress. Pareto-optimal solutions with off-design points were obtained, and their performances were compared to the reference NREL Phase-II geometry. As a result, it was observed that there was an increase in CP from 8.07% to 15.56% and the von-Mises stress showed a considerable increase up to 231.13%. The study demonstrates the aerodynamic benefits and structural drawbacks of various winglet designs in horizontal axis wind turbines.