Leveraging large eddy simulations to assess noise source imaging of a controlled supersonic jet

Abstract

Noise source imaging based on phased array measurements is an essential tool in the aeroacoustic analysis of new nozzle designs, especially at full-scale. This investigation aims to assess the capability of a deconvolution-based beamforming technique to accurately estimate the changes in noise sources for model-scale heated military jets when fluid inserts are used for noise control. This goal is achieved by performing well-validated Large Eddy Simulations (LES) to complement the experimental measurements. The LES data is segregated into its hydrodynamic, acoustic and thermal components using Doak’s Momentum Potential Theory (MPT). The near-field MPT-derived components are subjected to Spectral Proper Orthogonal Decomposition (SPOD) to compare with the frequency-dependent noise source maps obtained directly from experiments. It is shown that fluid inserts alter the naturally occurring Kelvin-Helmholtz (K-H) instability in the jet shear layer, which leads to a change in the directivity of the noise radiated in the near-field. The upstream shift in the noise source distribution resulting from the modified K-H instability is accurately captured by the deconvolution-based source imaging technique using just the far-field measurements. These changes in source locations as a function of frequency are documented.