A gradient-descent-based method for design of performance-scaled rotor for floating wind turbine model testing in wave basins

Can Yang,Zhengshun Cheng,Longfei Xiao,Xinliang Tian,Mingyue Liu,Binrong Wen
DOI: https://doi.org/10.1016/j.renene.2022.01.068
IF: 8.7
2022-01-01
Renewable Energy
Abstract:When performing the model testing of floating wind turbine (FWT) in wave basins, using geometric scaled rotors is unable to achieve the desired thrusts due to the dramatic reduction of Reynolds number experienced by model-scaled blades. Alternatively, a performance-scaled rotor (PSR) is usually utilized. In the present study, a gradient-descent-based method, named as CCB-GD method is proposed to conduct the PSR design for FWT model testing in wave basins. Different from existing methods, the proposed method aims to mimic the radial distribution of normal aerodynamic loads along the blade, rather than the total thrust. This is achieved by combining the blade element momentum theory with gradient descent optimization algorithm. In the present design procedure, a new airfoil with good aerodynamic performance at low Reynolds number is first selected, the optimal radial distribution of twist angles and chord lengths are then determined separately and sequentially based on the gradient descent optimization algorithm. Additionally, high-order Bezier curves are used to smooth the radial distribution of twist angle and chord length. The proposed method is demonstrated by two case studies, in which the PSR design of the NREL 5 MW wind turbine and the DTU 10 MW wind turbine are conducted and compared with existing methods. Results show that the present method can design the PSR with good accuracy. Besides, the present method is robust, generic and also applicable in the PSR design of MW-scale wind turbines for FWT model testing in the wave basin.(c) 2022 Elsevier Ltd. All rights reserved.
What problem does this paper attempt to address?