Extraction of the acoustic part of a turbulent boundary layer from wall pressure and vibration measurements

The acoustic pressure in the nearfield of a wall in a flow can be decomposed in different components: a first part is related to the vortices resulting from the perturbation of the flow by the presence of the wall, a second part is due to the acoustic waves generated by the parietal pressure fluctuations diffracted by the wall, and a third part is potentially caused by any other incident acoustic wave. In many situations, one may be interested in the possibility of extracting experimentally the acoustic part of a parietal pressure. It can be achieved by directly measuring the parietal pressure using flush mounted microphones, and by applying post processing like denoising, wavenumber analysis, modal identification. However, the wall pressure fluctuation is generally largely dominated by the non acoustic part, which complicates the identification because of very low SNR. A second possibility is to identify the parietal pressure from the vibration of the wall itself using inverse methods. This approach offers the possibility to use the structure as a wavenumber filter, improving significantly the SNR. The aim of this paper is to compare two approaches based on direct and indirect measurements in the framework of an experiment in a wind tunnel. The direct approach uses several cleaning methods applied to the measured wall pressure fluctuation, and the indirect approach is based on the identification of the pressure from vibration measurements.