Comparison of Different Techniques for Modelling an Ultra-Low Speed Vertical Wind Turbine

Abstract

Two different techniques for simulating the performance of a helical Savonius rotor of 1m high and 0.5 diameter with 180o twist angle are compared in this study. Designed to operate at a wind speed range of between 2 to 5 m/s, this drag type vertical axis wind turbine was evaluated by using commercial Computational Fluid Dynamics code. The first technique is based on the tip speed ratio (TSR) incorporating a sliding mesh interface to estimate its optimal power coefficient. On the other hand, a flow-driven concept is adopted in the second technique where a rigid body dynamic(RBD) parameters of the rotor are defined. Common parameters applied to both models include element size, time step, turbulent models and the boundary conditions. Both techniques predicted the optimal power coefficient of 0.165 to occur at 0.9 TSR with about the same computing time of 50hrs on a four-core machine. As far as the actual turbine output RPM is concerned both techniques have underestimated the value by at least 15% based on the S-A and SST turbulence models. In this case, the S-A model took 30% less computing time than the SST model. The flow-driven model provides a better representation of performance of the actual turbine at varying wind speeds rather than TSR.