Phase-Locked PIV Measurement in the Wake of Model Wind Turbines Under Various Inflow Conditions

This paper focuses on understanding correlative interactions between boundary layer flow structures and the resultant unsteady wake of a Horizontal Axis Wind Turbine (HAWT) model. Phase-locked Particle Image Velocimetry (PIV) is employed to measure turbulence statistics such as velocity, turbulence intensity, shear stress, vorticity, and to subsequently identify large-scale coherent flow structures. In the first stage, phase-lock experiments were performed under free-stream flow conditions. Ten consecutive downstream locations up to six rotor diameters from the turbine are captured. Ensemble averaged velocity and vorticity fields reveal that while the identity of tip vortices are maintained over five rotor diameters downstream of the turbine, their strength decays exponentially. When the turbine is placed in the wake of other units, the vortical structures exhibit a rapid decay in both coherence and strength and substantially suppress the wake-vortex and vortex-vortex interactions, playing an important role in the wake recovery. These observations inspire the current investigation using low-speed phase-locked PIV Interactions among the near wall flow structures in a turbulent boundary layer, hub and tip vortices will be investigated in this paper. The model turbine has a 0.108 m hub height, rotor diameter of 0.128 m and tip speed ratio of 4. It is located in a wind tunnel under nearly zero-pressure-gradient and thermally neutrally stratified conditions. A tripped turbulent boundary layer generated by a picket fence located at the inlet has a boundary layer thickness, δ, of 0.55∼0.6 m. Measurements are performed at Re = 3×10⁵, 4×10⁵, and 12 × 10⁵.. To achieve sufficient spatial resolution, two measurement fields are taken at each stream-wise location to cover upper and lower half of the turbines. Measurements locations extend ten diameters downstream. Robust turbulence statistics, such as velocity fluctuations, Reynolds stresses, full budget of turbulent kinetic energy, are computed from large dataset, totaling 400 GBytes.