Geometric design optimization of a Savonius wind turbine

Abstract

Savonius vertical-axis wind turbines are becoming an outstanding alternative for small-scale power generation mainly due to their simple design, high self-starting performance at low wind speeds, omnidirectional capability, and low cost. Over the past few decades, extensive research has suggested turbine configurations that maximize their aerodynamic performance, evaluating different influencing geometric parameters. Nevertheless, few studies have considered running large-scale experimentation, simultaneously assessing different influencing parameters, and implementing mathematical optimization techniques. Consequently, there is a need to establish a set of procedures and techniques that address these gaps. The current research proposes a new design method for a Savonius wind turbine to maximize its aerodynamic performance. The proposed method defines a numerical model for a geometry with four influencing parameters, which automatically achieves solutions to the equations that govern the motion of fluids for different turbines. The model is successfully validated with experimental results available in the literature. The numerical solutions of 340 different turbines build a response surface with excellent quality metrics based on the Kriging technique. Two mathematical optimization techniques are implemented on the response surface to determine the configuration that maximizes its aerodynamic performance. This optimal configuration is characterized under different operating conditions. The results indicate that the optimal turbine has an aspect ratio of 8.38, an overlap ratio of 0.08, a twist angle of 174.05°, and two blades. The behavior shown by the optimal turbine (maximum power coefficient of 0.21 at a wind speed of 12 m/s) is remarkable since its aerodynamic performance is superior to multiple reported turbines. The instantaneous torque that it develops is always positive during its operation, which indicates a greater capacity for self-starting and continuity in its rotation. Likewise, it supplies a maximum rotor power of approximately 160 W at a wind speed of 12 m/s. These results make the optimal turbine an attractive design for electric power generation. The proposed method provides novel insights and powerful tools for researchers to explore various geometries and operating conditions that, together with the established optimization process, accomplish the turbine that maximizes its aerodynamic performance.