The importance of renewable and alternative energy is rapidly gaining attention. A national goal of replacing 20% of the United States electricity generation with wind power by 2030 has been proposed but such an ambitious goal is dependent on many parameters. Improved aerodynamic performance of wind turbine blades is one parameter necessary to achieve this goal. Blade testing is traditionally done using 2D airfoils in a laboratory wind tunnel, developing the lift and drag coefficients, and then using this data to predict wind turbine blade performance. Dimensional analysis has been used successfully in design of rotating machinery such as pumps, developing a series of dimensionless pump parameters with which to scale a particular pump design to a larger or small size. These parameters lead to similarity or affinity laws which relate any two homologous states for two pumps that are geometrically and dynamically similar. Affinity laws could be applied to wind turbines however the conditions tested in the wind tunnel do not match what would be expected in a full scale wind machine. As with pumps, the laws would apply only if the model and full scale wind turbine would operate at identical Reynolds numbers and are exactly similar (i.e. relative surface roughness and tip conditions). Reynolds numbers in the model tests are smaller than those achieved by the actual wind turbines while the surface roughness of the model is generally larger. This leads to the need for empirical equations to predict performance. This paper examines current wind tunnel testing and the problems with scaling wind turbine blades. It also outlines a methodology to test 3-D model wind turbine blades in a wind tunnel. Blades are designed and manufactured according to existing criteria, mounted to a generator, and their performance is then tested in the wind tunnel. Challenges with wind tunnel testing as well as extrapolation of the wind tunnel data to actual applications will be addressed.

This content is only available via PDF.
You do not currently have access to this content.