Development of a Laboratory Oscillating Surge Wave Energy Converter

Abstract

Ocean wave energy conversion is a promising solution to growing energy demands across the globe, particularly for island and coastal communities. Oscillating Surge Wave Energy Converters (OSWECs) harness energy through a buoyant flap, typically bottom-hinged, which oscillates in response to surging wave forces. With OSWEC design still an active area of research, the technology is not yet mature enough for grid-scale or ``Blue Economy’’ applications. There are numerous experimental studies of OSWECs to-date, with topics ranging from general performance to array behavior. However, further experimental research remains essential to address knowledge gaps in device hydrodynamics and control. While many laboratory-scale experimental OSWECs have been developed, these have been sparsely instrumented and rarely used to understand or optimize control strategies. This thesis focuses on developing the driveline for a laboratory-scale experimental OSWEC intended to study hydrodynamics and control. The device is unique, relative to prior experimental work, given it is designed to emulate power-take-off using a servo-motor and gearbox protected by a submersible housing directly underneath the flap, reducing losses and potential hydrodynamic interactions from a chain drive or hydraulic system. However, the gearbox and rigid shaft coupling included in the driveline may introduce differences in commanded and realized flap position due to backlash or torsion. Characterizing this behavior is a necessary step in readying the driveline for experimental work. Additionally, future experiments may require accurate control of device position or driveline torque, so experimental characterization of motor torque and velocity constants is another step required prior to OSWEC experimentation. To study positional differences between the motor and external shaft as a result of backlash and coupling torsion, a benchtop dynamometer was designed and built. The benchtop fixture secures the following components in-line: a servomotor, gearbox, rigid shaft coupling, encoder, six-axis load cell, bearing, flexible shaft coupling, and a magnetic particle brake. During experiments, the motor commands a sinusoidal velocity profile similar to that seen by the OSWEC during regular operation, and the particle brake commands a resistive torque proportional to the shaft rotational velocity. Position measurements are captured by an encoder internal to the motor as well as an encoder on the shaft outboard of the gearbox and coupling. In order to characterize torsional deflection of the shaft relative to the motor, position of the motor and shaft are compared alongside driveline torque measurements captured by the load cell. Shaft oscillations are found to have a smaller amplitude and are shifted in phase in comparison to motor oscillations. Amplitude attenuation is found to be linearly correlated with root-mean-squared driveline torque, and phase shift linearly correlated with root-mean-squared motor velocity. These relationships are used to create a position control correction which reduces positional differences between the motor and shaft by a factor of two. Derivatives of the benchtop dynamometer are used to experimentally determine motor torque and voltage constants. Hardware setup remains unchanged for the study of motor torque constant during which torque control is employed for the motor while the particle brake is commanded to resist with maximum torque to maintain a stationary driveline. The torque constant is calculated using a linear fit between driveline torque, as measured by the load cell, and command current, as delivered to the motor, and is found to be in agreement with motor specifications. However, significant variability in the calculated torque constant between experiments suggests that precise torque control will require real-time corrections (i.e., closed loop control on measured mechanical torque). During velocity constant testing, the particle brake is disengaged from the dynamometer driveline and the motor is operated under velocity control. The velocity constant is then calculated using a linear fit between the driveline velocity and motor command voltage. The spread of velocity constant values is only 2.8\% of the mean, suggesting accurate velocity control could be achieved with minimal real-time corrections.