Experimental and theoretical characterization of acoustic noise from a 7.6 m diameter teetered rotor wind turbine

Abstract

An experimental investigation into the acoustic noise from a small (7.6 m diameter) teetered rotor wind turbine, set at various yaw angles up to 90 degrees of yaw, was conducted. The results revealed 1/3-octave spectra dominated by a broad peak in the higher frequency range, at all yaw angles and wind speeds investigated. This prompted a theoretical investigation to reveal the mechanisms producing the dominant feature in the experimentally obtained noise spectra and resulted in the development of a wind turbine aerodynamic noise prediction code, WTNOISE. The location near busy roads and the relatively rough terrain of the wind test site caused difficulties in obtaining useful noise spectral information below 500 Hz. However, sufficiently good data was obtained above 500 Hz to clearly show a dominant hump in the spectrum, centered between 3000 and 4000 Hz. Although the local Reynolds number for the blade elements was around 500,000 and one might expect Laminar flow over a significant portion of the blade, the data did not match the noise spectra predicted when Laminar flow was assumed. Given the relatively poor surface quality of the rotor blades and the high turbulence of the test site it was Copyright © 1998 by the American Institute of Aeronautics and Astronautics, Inc. and the American Society of Mechanical Engineers. All rights reserved. therefore assumed that the boundary layer on the blade may have tripped relatively early and that the turbulent flow setting should be used. This assumption led to a much better correlation between experiment and predictions. The WTNOISE code indicated that the broad peak in the spectrum was most likely caused by trailing edge bluntness noise. Field data confirmed this hypothesis and also showed the beneficial effect of sharpening the trailing edge.