Numerical Study on the Performance of Hybrid Darrieus-Savonius Vertical Axis Wind Turbines Having Twisted Blades

Abstract

Abstract Darrieus and Savonius rotors are the main types of vertical axis wind turbines. Darrieus rotors exhibit high power coefficient but suffer from lower self-starting capabilities. On the other hand, Savonius rotors demonstrate relatively high self-starting capabilities but with an inferior power coefficient. Therefore, the current study proposes three hybrid Darrieus-Savonius designs with different angles of twisting aiming at having a compact turbine with high coefficient of power and easy to self-start. The Savonius turbine is placed inside the Darrieus turbine in the three proposed hybrid designs. Accordingly, the trends of coefficient of power against ratio of rotor tip speed in addition to coefficient of torque over a whole cycle (360 degrees) are computed. The coefficient of static torque during a whole cycle is also calculated. Therefore, three dimensional, incompressible, and unsteady Reynolds-Average Navier-Stokes equations in conjunction with the turbulence model of k-ω shear-stress transport is developed and numerically simulated utilizing ANSYS Fluent. The devolved numerical model is validated using the available measurements from literature. The predicted characteristics of flow field over the three suggested hybrid designs are analyzed and presented. Accordingly, results illustrated that a significant enhancement of the coefficient of power is attained by suggested hybrid designs with twisting blades in comparison with the straight-bladed design. The model with twisted-bladed Darrieus and Savonius turbines with angles of twisting of 90° and 45°, respectively attains the largest peak of coefficient of power of 0.353. This hybrid configuration exhibits lower static torque fluctuations with positive values during the whole cycle. Consequently, using the suggested hybrid turbine with twisting blades improves both coefficient of power and the self-starting capability of the turbine.