Prediction of diffuser noise footprint of a ducted wind turbine

The aerodynamic broadband noise generated by ducted configurations, such as ducted wind turbines or propellers, is of research interest due to its implications for noise pollution. Accurate prediction methods for such configurations remain unavailable for low-fidelity approaches. This paper applies the thin annulus and segmentation models, an Amiet’s theory based analytical framework for predicting broadband noise from hollow cylinders, to two ducted configurations. The segmentation approach is further extended to account for the angle of attack in duct geometries. First, the thin annulus model is validated using an isolated duct, showing good agreement between predictions and experimental data. Then, the models are applied to predict the broadband trailing-edge noise of a diffuser-augmented wind turbine. The results highlight the dominance of the diffuser trailing-edge noise in this configuration and demonstrate the improved prediction accuracy of the segmentation model due to its ability to account for the angle of attack. This study provides a comprehensive framework for broadband noise prediction of ducted turbine applications, using inputs derived from RANS simulations, and demonstrates how low-fidelity approaches can be effectively utilized to address the noise prediction of a ducted turbine.