Numerical and Experimental Results of a Passive Free Yawing Downwind Wind Turbine

Abstract

The background of this PhD study concerns a medium sized 3 bladed wind turbine in a free yawing and downwind configuration. Largely funded by an EU Marie-Curie IAPP grant, the project was jointly setup by the renewable energy consultant 3E, DTU Wind Energy and the TU Delft. The 3 bladed free yawing downwind concept is pursued in an attempt to increase the robustness of a wind turbine by eliminating the traditionally actively controlled, and sometimes failure prone yawing mechanism. Under certain conditions, such as for remote (off shore) and off grid applications, a decreased failure rate can increase the economical competitiveness significantly compared to more traditional power supplies. This work presents aeroelastic analysis and results of a wind tunnel test campaign for the 3 bladed free yawing downwind concept. The investigated topics concern free yawing stability and how it is affected by coning angle, blade sweep, and blade flexibility using both numerical and experimental methods. The wind tunnel tests were organized in the Open Jet Facility of the TU Delft, and the thesis discusses the experiment’s design, construction, operation, and gives an analysis of the results. It provides a dataset to compare aeroelastic simulations with experimental results for varying yawed inflow conditions and free yawing stability. The experimental data scope is limited to rotor speed, tower base bending moment, yaw angle, and blade root flapwise bending moments measurements. Other work covered during the PhD study is a parametric blade sweep investigation for the NREL 5MW turbine, and a detailed study on load extrapolation methods based on aeroelastic simulations. It is concluded that the 3 bladed, free yawing, and downwind wind turbine can operate in a stable manner. However, numerical studies indicate a less stable operating region when the rotor flow is about to stall. The experiments confirmed the free yawing stability, but the unstable region indicated in the simulations could not be tested due to limitations with respect to the generator torque control. It is also ob-served that coning angle and blade sweep can help to further increase the stability of the free yawing concept. Finally, aeroelastic loads of the tower and blades obtained from HAWC2 simulations are validated against experimental results. It is noted, however, that a number of uncertainties remain unresolved regarding the blade load measurements.