Extended source models for long range wind turbine noise propagation

Abstract

Wind turbine noise can be perceived at distances greater than one kilometer and is characterized by amplitude modulations at the receiver. In order to predict this noise, it is necessary to model the dominant aeroacoustic noise sources as well as the main outdoor propagation effects. In most studies from the literature, the wind turbines are modeled as point sources to simplify the coupling between source and propagation models, but this assumption is not always justified. In this study, two original methods are proposed to couple an aeroacoustic source model based on Amiet’s theory and a Split-Step Padé parabolic equation code for acoustic propagation in an inhomogeneous atmosphere. In the first method, an initial starter is obtained for each segment of the blade using the backpropagation approach. This method enables us to accurately model the directivity of the noise sources but is very computationally intensive. In the second method, the blade segments are viewed as moving monopole sources, and only a limited number of parabolic equation simulations are needed which strongly reduces the computation time. These two methods are validated using an analytical reference solution in a homogeneous medium and compared in various inhomogeneous atmospheres.