Aerodynamics and aeroacoustics of diffuser-augmented wind turbines

As global energy demand continues to rise, particularly in urban centers, so does the environmental cost of using fossil fuels as energy sources. This growing demand, coupled with the urgent need to reduce greenhouse gas emissions, highlights the importance of increasing renewable energy production. While solar energy has become a dominant solution in urban areas, urban wind energy offers a complementary alternative which presents the challenges of low wind speeds, changes of direction, high turbulence and limited available space.

The diffuser-augmented wind turbine (DAWT) is a promising solution that enhances performance through a diffuser-shaped duct that surrounds the rotor and accelerates airflow. This configuration increases energy output and improves resilience to varying and misaligned wind directions, making it well-suited for rooftop or building-integrated installations in urban environments. Like all wind turbines, DAWTs generate aerodynamic noise from the rotor blades and, additionally the diffuser emits noise which dominates in most directions. Because DAWTs are designed for urban installations, the acoustic impact must be predicted and controlled to ensure public acceptance and regulatory compliance.

The rotor generates aerodynamic noise primarily through trailing-edge noise, a broadband mechanism resulting from the scattering of turbulent boundary layer fluctuations as they advected over the blade’s trailing edge. In this thesis, rotor broadband noise was modeled using Amiet’s theory through a strip approach using inputs from either a two-dimensional RANS or a three-dimensional RANS simulation. Additionally, tonal noise originating from periodic blade-passing effects is predicted using a steady-loading analytical model, with its scattering by the diffuser captured using a high-order finite element method.

In addition to scattering rotor noise, the diffuser introduces significant broadband noise through its trailing edge. This mechanism becomes dominant when the rotor-induced flow enhances turbulence near the diffuser surface and accelerates the flow, which amplifies trailing-edge noise. The broadband noise contribution was modelled with two analytical methods, derived based on Amiet’s theory, extending it to cylindrical geometries. The models for predicting leading- and trailing-edge broadband noise were validated by predicting the noise of a isolated duct and the diffuser of the DAWT.

The predictions revealed how the tonal noise generated by the rotor is scattered by the diffuser geometry, with the scattering strongly influenced by the rotor’s axial position. The results also indicate that the rotor and diffuser broadband noise radiate differently: diffuser trailing-edge noise dominates most directions, whereas rotor noise remains especially significant downstream, where diffuser emissions are minimal. These findings suggest that acoustic modeling and strategic placement could minimize the overall acoustic disturbance for nearby communities.