Aeroacoustic wind tunnel tests

Abstract

The global demand for renewable energy sources has led to an increase in development and installation of wind turbines. However, noise regulations on wind turbines near populated areas can restrict the installation of new wind turbines and limit the potential energy production. The main noise from wind turbines is aerodynamic and arises from turbulence caused by the air flow around the wind turbine blades in motion. The aerodynamic noise increases with the size of the wind turbine, and with a demand for more productive, and thus larger wind turbines, there is an increased interest in quantifying the aerodynamic noise from wind turbine blades. This thesis investigates experimental methods for estimating the noise of wind turbine blades in a newly-established wind tunnel – the Poul la Cour Tunnel (PLCT). This study has three focus areas; the acoustic design of the wind tunnel airline, the measurement conditions and sound propagation properties in the anechoic, Kevlar-walled test section, and the quantification of trailing edge noise of wind turbine blade sections. The attenuation of background noise (from the wind tunnel airline) is a crucial part of successful acoustic measurements. To investigate this, the wind tunnel design was described and an experimental in-situ method proposed for determining transmission losses of the acoustically treated wind tunnel airline components. The results were compared to a numerical study that was conducted prior to construction, showing good agreement and thereby validating the general acoustic design. In the test section, acoustic measurements are made with a microphone array situated behind a tensioned Kevlar wall inside an anechoic room. The acoustic losses due to sound propagation through the Kevlar fabric was examined both with and without flow in several measurement setups. A comparison of methods and results was discussed and found to agree with the existing literature on the topic. Additionally, methods for quantifying absolute noise levels of an airfoil was examined. Several of these were applied to measurements of a NACA63018 airfoil at flow speeds ranging from about 30 m/s to 70 m/s (Reynolds numbers 2 mio. to 4 mio.) and corrections due to Kevlar losses were applied. The results were compared to a trailing edge noise model, finding good agreement. This dissertation describes the relevant contributions of the PhD study and places it in the context of current aeroacoustic wind tunnel research.