Bio-Inspired Trailing Edge Noise Control

Abstract

Trailing edge noise remains a primary limiting factor in the widespread implementation of wind turbines, particularly near populated areas. Noise regulations commonly require acoustic de-rating of existing turbines, leading to reduced output and revenue. This presentation will describe an experimental study aimed at trailing edge noise control inspired by the unique features found on the wings of owls that use acoustic stealth while hunting prey. One of these features is a thin layer of fine hairs which grow from the exposed surfaces of the flight feathers. These hairs have been investigated and found to form a sort of canopy suspended above the surface of the owl's feathers. Previous wall-jet tunnel measurements have shown that high open-area canopies of similar characteristics can reduce surface pressure fluctuations on the underlying surface by as much as 30dB, and significantly attenuate roughness noise generated by that surface. In the present work, treatments designed to replicate the effects of the canopy in a form suitable for application to an airfoil have been designed and tested in the Virginia Tech Stability Wind Tunnel. Over 20 variants of these designs have been tested by performing aeroacoustic wind tunnel measurements on a tripped DU96-W180 airfoil at chord Reynolds numbers up to 3 million. Exact details of the treatments are not given here since they are the subject of a current patent application, but the treatments will be described during the presentation. Variations include treatment thickness, density, length, position relative to the trailing edge and the effectiveness of treating only one side of the trailing edge. The treatments were placed over the center-half span of the airfoil in the trailing edge region. Measurements included far-field acoustic data from a 117-microphone phased array and mean surface pressure data from 80 pressure taps distributed over the airfoil profile. For some conditions a rake of Pitot and static probes was used to measure profiles through the airfoil wakes and infer the drag using a momentum balance approach. Compared to the unmodified airfoil the treatments were found to be quite effective. Acoustic beamform maps and integrated spectra show up to 10dB of broadband attenuation of trailing edge noise in the vicinity of the treatment. The majority of the noise attenuation was observed in the frequency range above 1500Hz, but measurements below this frequency are inconclusive because of the large spot size of the phased array at these frequencies. The treatment remains effective throughout a wide parameter range and is not highly dependent on a particular geometry, but there appears to be strong potential for optimization. Treatments were found to be effective over an angle of attack range that extends over 10 degrees from zero lift. Compared to the unmodified airfoil, no additional noise was measured from the treated airfoil past this 10 degree range. The mean surface pressure data revealed that the presence of the treatment had little impact on the lift characteristics of the airfoil model. Drag rake results showed a small increase in drag proportional to the increase in wetted area resulting from the addition of the treatment to the unmodified airfoil.