High Reynolds Number Horizontal and Vertical Axis Wind Turbine Experiments

Abstract

The large scale of modern wind turbines has created a disconnect between laboratory experiments and field-scale operation. Due to the added complexity of a non-dimensional frequency, based on the turbine angular velocity, conventional wind tunnel facilities are not able to match all non-dimensional parameters between laboratory and field-scale. The ever-increasing size of wind turbines continues to highlight this discrepancy, yet more is demanded of new models and simulations in the form of accuracy with actual turbine operation. To address this issue a unique wind tunnel facility has been employed to study the Reynolds number scaling of both horizontal and vertical axis wind turbine geometries, or HAWTs and VAWTs, respectively. With the additional control provided over the fluid density, Reynolds numbers and non-dimensional rotational rates are matched between model and full-scale. To operate successfully in the pressurized environment, an entire suite of models, sensors, and test platforms were developed. This allowed for highly controlled and accurate measurements to be made of wind turbine performance with respect to both power and thrust coefficients.

The results indicate that a Reynolds number based on the flow at the rotor chord is the preferred way to scale Reynolds number behavior for both horizontal and vertical axis wind turbine models. For the first time, a wide enough range of Re was explored to confirm the value for which the models achieved Reynolds number invariance. In addition, the performance of the tripped HAWT rotor was explored with a well-characterized trip geometry and it was observed that the threshold to achieve invariance in the power coefficient was reduced from the un-tripped case. Vertical axis experiments included a study of several solidity values across a range of Reynolds numbers with the goal of adding to the understanding of scaling effects for a wide range of VAWT designs. The conclusions drawn, and reference data gathered during the course of these experiments is now available for use as a validation case for further development of models, comparison to numerical simulations, and for increasing the knowledge base of canonical HAWT and VAWT operation.