Reduction of broadband trailing edge noise by serrations

Abstract

This thesis aims to investigate and reduce the aerodynamic noise source known as trailing edge noise, or airfoil self-noise, by using passive flow control techniques. Airfoil self-noise is produced when a turbulent boundary layer generated on an airfoil surface is scattered by the airfoil’s trailing edge. The investigation is of experimental nature, conducted in the aeroacoustic as well as aerodynamic wind tunnel facilities at Brunel University London and the Institute of Sound and Vibration (ISVR) at the University of Southampton. The research is relevant for any application in which airfoil blades encounter a smooth non-turbulent inflow and hence where trailing edge noise is a dominant noise source. Potential applications can therefore be fan or rotor blades in aero-engines, wind turbine blades or industrial cooling fans. The approach taken for the reduction of trailing edge noise utilises passive flow control techniques through the use of trailing edge serrations and the additional support of porous materials. Both of the aforementioned are inspired by the owl’s silent flight due to its unique wing structure. The research presented here can be divided in three parts: The first part comprises an extensive assessment of the performance of non-flat plate trailing edge serrations for airfoil broadband noise and their aerodynamic performance in terms of lift and drag. It is found that serrations can realistically achieve noteworthy broadband airfoil self-noise reductions, however due to the fact that non-flat plate serrations are directly cut into the airfoil body, the blunt sections in the serration root produce an additional noise source of vortex shedding tonal noise. The second part investigates the two flow mechanisms involved. Regarding the mechanism responsible for broadband noise and the subsequent reductions by the serration geometry, the turbulent boundary layer structures are studied in depth on a serrated trailing edge of a flat plate. Experimental techniques such as hot wire anemometry, liquid crystal flow visualisation, unsteady surface pressure measurements and noise measurements are used. A redistribution of the momentum and turbulent energy near the sawtooth tip and side edges appears to reduce the trailing edge noise scattering-efficiency of the hydrodynamic pressure waves. For the study of the flow mechanism responsible for the vortex shedding tonal noise increase, noise and velocity measurements along with flow visualisation techniques are used for the identification and further understanding of this noise source. A highly three-dimensional wake-flow could be identified in the wake past the serration gap, which differs from the longitudinal vortices shed from a straight blunt serration root. The third part presents the concept of poro-serrated trailing edges as a novel method to substantially improve the overall noise performance of the non-flat plate trailing edge serration type. The use of porous metal foams or thin brush bundles which fill the interstices between adjacent members of the sawtooth can completely suppress the bluntness-induced vortex shedding noise. Most importantly a turbulent broadband noise reduction of up to 7 dB can be achieved without compromising the aerodynamic performances in lift and drag. The new serrated trailing edges do not cause any noise increase throughout the frequency range investigated here. Through noise and velocity measurements near the trailing edge of an airfoil, the reduction of the broadband noise is found to be primarily caused by the sawtooth geometry. The new serrated trailing edges have the potential to improve the industrial worthiness of the serration technology in achieving low noise radiation.