Modelling and localizing low frequency noise of a wind turbine using an array of acoustic vector sensors
The large size and low rotational speed of modern wind turbines are often linked to the generation of low frequency noise. This paper proposes a simplified approach to model the sound produced by a wind turbine based on moving monopole sources. Time-dependent Green functions are used to account for the Doppler effect introduced by the relative changes in position between the moving elements and the fixed sensors. The proposed model can be used for understanding how different mechanical defects have an impact on the perceived sound. The sound field is hereby studied through an array of acoustic vector sensors (AVSs) since it enables locating low frequency sound sources with a relatively small aperture. A beamforming method is applied upon the synthetic data for locating the noise emission points along the moving blades. An experimental investigation is also presented introducing a novel in-situ calibration procedure for adjusting the AVS orientation. Both numerical and experimental results show that the proposed approach is suitable for modelling and localizing the sources of noise emission with a low number of acoustic vector sensors.
Fernandez Comesana, D., Nambur Ramamohan, K., Perez Cabo, D., Carrillo Pousa, G., (2017). Modelling and localizing low frequency noise of a wind turbine using an array of acoustic vector sensors. In Proceedings of 7th International Conference on Wind Turbine Noise.