Development of a Generalizable Understanding of Cross-Flow Turbine Design: Intracycle Velocity Control and Parametric Changes in Rotor Geometry

Abstract

Cross-flow turbines, often referred to as vertical-axis wind turbines (VAWTs), offer potential for wind and hydrokinetic energy generation. However, their complex, fluctuating angle-of-attack, dynamic stall, and flow recovery cause periodic and unsteady dynamics that are difficult to model, and an optimal design process remains undetermined. This dissertation investigates two independent pathways to improve turbine performance by regulating blade-flow interactions: (1) intracycle variable-velocity control to adjust the relative flow experienced by the blades, and (2) modifying rotor geometry---including camber, blade pitch, and chord-to-radius ratio (c/R)---to influence the blade aerodynamics. These approaches are explored through a combination of laboratory performance measurements and theoretical analyses rooted in classical steady-state aerodynamic principles to inform generalizable design rules. By focusing on optimizing both the control scheme and blade geometry, significant gains can be achieved without adding substantial increases in complexity. All experiments were conducted in the Alice C. Tyler flume at the University of Washington, where temperature and flow conditions were held constant, to enable broader non-dimensional interpretation of results. To modify and control cyclic fluctuations in angle-of-attack, several works have evaluated active blade pitch control and found benefits; however, this introduces design complexity and additional failure pathways. A simpler fixed-geometry approach is sinusoidal variable-velocity (intracycle) control, where velocity varies with blade position to regulate relative velocity and improve efficiency. A preliminary study by Strom et al. (2017) demonstrated a 53% increase in efficiency over constant-speed control, but the broader implications remain poorly understood. For example, while increased performance has been generally shown to increase loading, the extent of this effect has not previously been studied for this type of control. This work experimentally explores a systematic sweep of intracycle kinematics by varying the phase and amplitude of sinusoidal changes in rotation rate, and examining their impact on efficiency and loading. Performance was found to be most sensitive to changes in the sinusoid's phase shift, which regulates synchronization between angular velocity, torque production, and near-blade hydrodynamics. Optimal kinematics increased efficiency by up to 13.2% over constant-speed control or reduced maximum forces by 20%. Performance gains depended strongly on the Reynolds number, which explains the smaller improvements observed here (on a percentage basis) relative to prior work, despite higher peak efficiency. It was found that triggering an earlier stall, followed by blade acceleration, can reduce maximum loading by 12% while simultaneously increasing efficiency by 3% by broadening the performance peak. Particle image velocimetry connected near-blade flow structures with key intracycle kinematics and identified beneficial hydrodynamic structures for future control strategies. To understand how fixed rotor geometry can impact turbine performance, loading, and near-blade hydrodynamics of cross-flow turbines, the influence of blade camber, pitch, and c/R, were examined simultaneously. These geometric parameters have a coupled effect on performance, adding to design complexity. The circular path of cross-flow turbine blades causes the incident flow to appear curved, inducing a virtual camber and pitch, which alter the lift, drag, and pitching moment of the blades. While virtual camber and pitch are largely determined by c/R, they are expected to interact directly with geometric pitch and camber, complicating turbine performance optimization. While cambered blades may be beneficial, there is little consensus in the literature on which camber direction might be most favorable. By considering both virtual and geometric effects, this work clarifies how pitch, camber, and c/R are coupled, enabling a better understanding that can help reduce design complexity. Optimum geometric camber is seen to be somewhat rotation-rate-dependent, with concave-in (negative) camber showing potential to reduce loading while enhancing performance at higher rotation rates. Peak efficiency was not found to be particularly sensitive to percent camber, though a -2% (NACA-2418) cambered blade shows relatively consistent benefits across the tested rotors. The overall optimum net camber, defined as the combination of geometric and virtual components, was found to vary somewhat across c/R. For the higher tested c/R values, 0.49 ≤ c/R ≤ 0.74, performance is quite similar, and the optimal net camber appears to lie between 6% and 8%. This range of optimum net camber is likely due to the non-linearity of geometric camber's influence on dynamic stall and recovery. Departures in performance at low c/R deserve further exploration, as they may be related to either Reynolds number effects or a change in the sign of geometric camber to achieve the same net camber. Importantly, by utilizing net pitch, defined similarly to net camber, optimal performance collapses to a greater extent across c/R. The optimum net pitch for all foils (independent of camber) is approximately zero, explaining previous observations for symmetric foils that demonstrated the optimal geometric blade pitch to be related to c/R. Finally, thin airfoil theory is used to draw parallels between pitch and camber using the zero-lift angle-of-attack, and to define a new “total” pitch parameter that encapsulates the effects of camber, pitch, c/R, and their flow-curvature effects into a single variable that is correlated with the range of performance and loading within a single rotation. This new parameter provides a physical representation of the lift generation at infinite tip-speed ratio and demonstrates a linear collapse of both the maximum force and the variation in efficiency over a cycle. Together, the research contained in the chapters of this dissertation establishes experimental links between turbine kinematics, blade geometry, and near-blade hydrodynamics, demonstrating how control and aerodynamic design can enhance efficiency while mitigating structural loads. The results provide new physical insight and practical frameworks for the future optimization of cross-flow turbines for marine and wind energy applications.