The effect of tip-speed ratio and free-stream turbulence on the coupled wind turbine blade/wake dynamics

Abstract

Wind turbines operating within wind farms experience complex aerodynamic loading arising from the interplay between wake-induced velocity deficits, enhanced turbulence, and varying operational conditions. Understanding the relationship between the blade's structural response to the different operating regimes and flow structures generated in the turbine's wake is critical for predicting fatigue damage and optimizing turbine performance. In this work, we implement a novel technique, allowing us to simultaneously measure spatially distributed blade strain and wake dynamics for a model wind turbine under controlled free-stream turbulence (FST) and tip-speed ratio ($λ$) conditions. A $1$ $\mathrm{m}$ diameter three-bladed rotor was instrumented with distributed Rayleigh backscattering fibre-optic sensors, while synchronised hot-wire anemometry captured wake evolution up to $4$ rotor diameters downstream. Experiments were conducted covering a wide $\{\mathrm{FST}, λ\}$ parameter space -- $21$ cases in total. Results reveal that aerodynamic-induced strain fluctuations peak at $λ\approx 3.5$, close to the design tip -speed ratio ($λ_d = 4$), with the blade's tip experiencing a contribution from the aerodynamically-driven strain fluctuations of up to $75\%$ of the total fluctuating strain at design conditions. Spectral analysis shows frequency-selective coupling between wake flow structures and the blade response, dominated by flow structures dynamically related to the rotor's rotating frequency (\textit{eg.} tip vortex structure). The novel experimental methodology and results establish a data-driven foundation for future aeroelastic models' validation, and fatigue-informed control strategies.

Figures (18)

• Figure 1: a): Characterisation of the free-stream turbulence (FST) conditions used in the experiment. Each case corresponds to a different grid-turbine spacing $x/M$, generating distinct turbulence intensity ($TI$) and integral length scale $\mathcal{L}/R$. Increasing turbulence levels are denoted by cases $A$, $B$, and $C$ with decreasing $x/M$ ratios. Integral length scale is defined both as a ratio of the turbine radius ($R$), and length of the chord at midspan ($MC$). b): picture of the experimental apparatus set at the centre of the $10'\times 5'$ wind tunnel top section.
• Figure 2: ($a$) Schematic of the Rayleigh backscattering sensor layout along the pressure side of the wind turbine blade. The sinusoidal fiber path, defined on a reference grid with spacing $a = 40,\mathrm{mm}$, allows the retrieval of both the edgewise and flapwise strain components. ($b$) Schematic of the nacelle, where the optical slip ring (MFO100) is mounted on the nacelle, at the base of the rotor shaft, allowing the continuous optical signal transmission from the rotating blade, to the stationary optical interrogator (ODiSI-B).
• Figure 3: Schematic of the hot-wire array set-up locations, along the spanwise direction of the turbine. The wind turbine is mounted at the centre of the test section of the wind tunnel, and the six hot-wire probe array is mounted on a traverse system that moves along the spanwise, and streamwise direction of the wind tunnel allowing the characterisation of the generated wake dynamics by the wind turbine.
• Figure 4: a): Phase averaged strain measured across the blade under a quiescent background, averaged over $s/R=[0.15,0.95]$, at different phases $\phi$, for each of the rotational conditions respective of the tested operating tip speed ratios. b): Ensemble average of strain measurement across the blade's span, for the different tested rotating velocities and expected acceleration induced by centrifugal force $R \Omega^2$.
• Figure 5: Power coefficient ($C_P$) of the wind turbine model as a function of tip speed ratio ($\lambda$).