Vertical Axis Wind Turbines: Analysis of Experimental Data at Full Dynamic Similarity

Abstract

As the world’s energy demand increases and the negative impacts of traditional energy sources become increasingly evident, the need for new, clean and sustainable energy technologies such as wind energy has become most clear. Due to the diversity of needs and resources around the world, different sustainable energy technologies are being developed, and Vertical Axis Wind Turbines (VAWT) are one of those technologies currently under study that will soon enter the market in large numbers. Thus, research on this technology and how its design can be best perfected for commercial applications can have important effects on the energy transition. This thesis uses experimental data collected at Princeton's High Reynolds number Test Facility (HRTF) to analyze VAWT performance. The experiments from which this data was collected were carried out on a geometrically scaled vertical axis wind turbine model over an unprecedented range of Reynolds numbers, including and exceeding those of the full-scale turbine, while maintaining dynamic similarity, which has outputted large amounts of valuable data only available at this facility. This thesis has analyzed that data in a variety of forms: fist studying time-averaged and phase-averaged data in the time domain and then moving to the frequency domain. The results allow us to validate the assertion of dynamic similarity, as well as gain important insights into the aerodynamic behavior of VAWTs. The optimal conditions for maximum power coefficient of VAWTs was found to be at tip-speed ratios close to unity, and turbine configurations with 5 blades were seen to perform best at those conditions. However, configurations with higher solidity were found more suitable at higher tip-speed ratios, which might be found at off-design conditions. Phase-averaged data of torque coefficient showed worse collapse at tip-speed ratio values higher than 1 but lower than 2 which is attributed to the effect of aerodynamic lift caused by the turbine blades and large variations in forces due the large range of possible angle of attack values. Fourier transforms were employed to study the data in the frequency-domain and, as expected, the highest peaking frequency was the rotational frequency of the turbine and that related to the number of blades. However, other frequency peaks were also found which were related to the rotational speed of the turbine. These results serve to better understand VAWT performance and make design recommendations for future turbines so that they can soon start replacing, or supplementing, older forms of energy generation.