Numerical simulation of a slit resonator in a grazing flow under acoustic excitation

摘要

It is known experimentally that a grazing flow has significant influence on the performance of a resonant acoustic liner. As yet, detailed understanding of the effect in fluid dynamics or acoustics terms is not available. One principal reason for this is the small size of the openings of the resonators of present day jet engine acoustic liners. The small size of the holes makes in-depth experimental observation and mapping of the fluid flow field around the opening of a resonator in the presence of a grazing flow extremely difficult. As a result, there is a genuine lack of data leading directly to a lack of understanding. The face sheet of an acoustic liner is entirely covered with holes (the openings of resonators underneath). There is, therefore, a possibility of fluid mechanical interaction between neighboring resonators. However, evidence for such interaction is not available at this time. One of the objectives of the present work is to shed light oil whether this is possible and what is a possible interaction mechanism. In this study, numerical simulations of the flow field around a slit resonator in the presence of a grazing flow under acoustic forcing are carried out. It is observed that at high sound pressure level, vortices are shed from the corners of the resonator opening. Some of these vortices merge together. Others are absorbed by the wall boundary layer or dissipated by viscosity. The simulated results indicate that a strong merged vortex is convected downstream by the grazing flow and persists for a long distance. This suggests that possible fluid mechanical interaction between neighboring resonators of ail acoustic liner could, indeed, be possible because of the interference of this convected vortex with the flow field of the downstream resonator. This interaction, as far as is known, has not been included in any theoretical or semi-empirical model of acoustic liners. Detailed formulation of the computational model, as well as computational algorithm, is provided. The computation code is verified by comparing computed results with an exact linear solution and also validated by comparing with measurements of a companion experiment. (C) 2007 Elsevier Ltd. All rights reserved.