The Influence of Surface Gravity Waves on the Performance and Near-Wake of an Axial-Flow Marine Hydrokinetic Turbine

Thumbnail Image

Publication or External Link





Surface gravity waves can significantly impact operating conditions for axial-flow marine hydrokinetic turbines, imparting unsteady velocities several orders of magnitude larger than the ambient turbulence. The complex interactions between the turbine and the wake, particularly in the presence of waves, are not well understood. Furthermore, detailed experimental data are needed for numerical model validation. Thus, the influence of surface waves on the performance and wake of a two-bladed axial-flow hydrokinetic turbine was investigated experimentally using an in-house designed and built towed Particle Image Velocimetry (PIV) system in the large towing tank facility at the U.S. Naval Academy. The turbine model has a 0.8 m diameter (D) rotor with a NACA 633-618 cross section that is Reynolds number independent with respect to lift coefficient in the operating range of Rec ≈ 4x10^5. Performance measures (i.e. power and thrust) were taken for both the steady (no wave) and unsteady (wave) cases. Average performance parameter values for the unsteady case were found to closely match those of the steady case, regardless of selected wave parameters. However, instantaneous values were found to depart substantially from the mean value having significant implications for power quality and fatigue loading. A wake survey was conducted under steady conditions to a downstream distance of 2D. Wake characteristics such as a decrease in the inflow velocity upstream of the turbine, wake expansion well-described by a 1/3 power law expression, a maximum velocity deficit of 2/3 the free stream velocity, and prominent turbine tip vortices were all observed. Methods developed for helicopter rotor analysis were applied to identify and characterize turbine tip vortices. Adjacent vortex filament interaction, thought to be the initial mechanism of wake break down and re-energization, was observed. A recently-developed vortex center averaging methodology was employed with new implications for the interpretation of turbulence statistics. A wake survey was also conducted under unsteady conditions over a similar downstream range. Blade-phase averaging was shown to be a poor descriptor of wake characteristics. Blade and wave phase averaging offers a clearer picture of wake dynamics in the presence of waves. The unsteady velocities induced by the waves were shown to change the spatial characteristics of the tip vortex helix. The strong helical structure that characterizes the steady case to a distance of approximately 1D persists in the steady case, however, the unsteady vertical wave velocity appears to convect vortex filaments into the wake region, potentially enhancing kinetic energy transport and wake re-energization. A simple, potential flow-based model was proposed to simulate the behavior of the wake influenced by waves, and a parametric study employing the model provided insight as to what factors most significantly affect vortex filament position and the characteristic length scales of the wake. An additional length scale was proposed to describe the shear layer in the unsteady case and is shown to agree well with experimental observations.