Comparison of Leading Edge Noise Measurements of an Airfoil in Two Wind Tunnel Facilities

Abstract

This work compares the measured radiated leading edge noise from a NACA 0012 airfoil with a 150 mm chord immersed in isotropic turbulence in two different anechoic wind tunnels. It investigates the cross-tunnel repeatability of leading edge noise measurements in a near-isotropic turbulent flow. Both wind tunnels are open jet anechoic wind tunnel types, with one located in UNSW Sydney, Australia and the other at the University of Poitiers, France. The experimental setups in the two facilities are designed to generate comparable mean flow and turbulence characteristics by ensuring identical turbulence grids in both facilities and equal distances between the grids and the airfoil in both facilities. The airfoil is subjected to two different flow speeds: 20 and 30 m/s corresponding to chord-based Reynolds numbers of Re_c = 2 · 10^5 and 3 · 10^5. The airfoil’s geometric angle of attack is kept at 0 degrees. Two near-isotropic turbulence cases with streamwise turbulence intensities of 5.1 % and 4.8 % (UNSW) and 4.1 % and 3.5 % (Poitiers) are generated by two passive turbulence grids, respectively. Flow measurements are taken using Stereo Particle Image Velocimetry & hot-wire anemometry at UNSW and hot-wire anemometry only at Poitiers. Beamforming arrays are used to localise and quantify the radiated noise from the leading edge of the airfoil in both facilities. A Source Region Integration method is used to extract the leading edge power spectral density. These power spectral densities obtained in the two wind tunnel experiments are then compared between the facilities and between all turbulence- and mean flow speed cases. The noise comparison results reveal that the grids’ self-noise levels have the potential to significantly corrupt the beamforming results of the leading-edge noise across the entire frequency range due to multiple individual factors which are discussed in the present work. The comparison of the turbulence characteristics yields similar integral length scales but identified minor differences in the turbulence intensities. This highlights how sensitive the turbulence characteristics can be to slight differences between the two open jet wind tunnel facilities. Suggestions are given to enable accurate and reproducible leading edge noise measurements in open jet wind tunnels.