Cross-Flow Turbine Fluid Mechanics: Experimental Optimization and Analysis

摘要

Cross-flow turbines are devices for converting the kinetic energy of wind or water currents to rotational mechanical energy. The dual objectives of this work are to increase our understanding of cross-flow turbine flow physics and to improve the energy conversion performance of individual rotors and arrays. Experiments are conducted using scale models in flumes or tow-tanks. The first group of studies examines cross-flow turbine rotor geometry. To make the parameter space tractable, this work is restricted to straight-bladed turbines with NACA0018 blade profiles. The shape and location of structures used to mount the blades to the center shaft are found to have substantial impact on turbine power output. Likewise, performance is found to be sensitive to the blade mounting angle. Finally, results are presented from a large multi-parameter study on how optimal mounting angle, number of blades, and chord length change with the scale of the turbine. Optimal geometry parameters are found to be strongly co-dependent. The most efficient turbine geometry for small and large-scale rotors is found to differ significantly due to blade boundary layer effects. In the following chapter, measurements of the wake of a cross-flow turbine using particle image velocimetry are presented. A new fluid analysis approach for extracting oscillatory flows is introduced and used to describe the wake features. The form and trajectory of Lagrangian coherent structures in the wake are described. Strong span-wise (axial) flow is observed in the core of shed vortices for the first time in a cross-flow turbine wake. The last chapter focuses on advanced control of cross-flow turbine rotors. A rotor control scheme that optimizes the local flow conditions on the blade by varying the rotor angular velocity is presented, and shown to increase turbine performance by 59% over standard control methods. Control and geometric optimization of an array of two turbine rotors is performed. This includes the introduction of a new array controller that seeks to optimize interactions with the coherent structures observed in the wake analysis. Beneficial interactions between rotors are shown to increase the array performance by 1.3 times that of isolated turbines.