An experimental investigation of wind turbine wake evolution and flow over complex terrain

2017-02-17T01:57:22Z (GMT) by Sherry, Michael John
This thesis investigates two flow fields that a wind turbine in a wind farm might experience. The first is the near-wake of a wind turbine where the focus was on the helical tip and root vortices. Two scale model wind turbines were investigated using PIV. A geometrically scaled rotor was observed to generate a chaotic wake due to poor aerodynamic performance at the experimental Reynolds number. Flow visualisations on a static 3D wing confirmed laminar separation is likely on the geometrically scaled model. A research orientated model was designed based on the optimum Glauert rotor to investigate the stability of the tip and root vortices. A pairing instability was observed in the tip vortices. The onset of this instability was found to be dependent on the tip speed ratio. The root vortices become unstable due to their proximity to the turbine support structures. The influence of freestream turbulence was studied by varying the turbulence intensity using passive turbulence grids including a novel tethered sphere grid. Increased turbulence intensity was observed to hasten the breakdown of the vortices, via turbulent diffusion rather than the pairing instability which was absent in the phase-locked average velocity fields. Tip and root vortices of both turbine models were characterised using Galilean invariant vortex identification schemes. The meander of the vortices was observed to be Gaussian at early vortex ages when interaction between vortices is minimal. Meander magnitude was shown to increase with distance absolutely. However, the magnitude of meander was found to be dependent on tip speed ratio and freestream turbulence intensity. A secondary focus was the flow fields above complex terrain features, a common location for wind farms. The recirculation region which formed downstream of various escarpment geometries was characterised using PIV. The size of the recirculation region was found to be dependent on the escarpment angle, the boundary layer to step height thickness ratio, the Reynolds number and the freestream turbulence intensity. An application of POD phase-averaging revealed the dynamic nature of the recirculation region. The wind speed-up above the escarpment beneficial in a wind energy sense was observed to be coupled to a vertical velocity component. Further, significant turbulence generation in the separated shear was observed which questions the appropriateness of complex terrain as a wind farm location.