Investigation of Flow–Acoustic Coupling in Dual-Stream Coaxial Supersonic Jets Using PIV and Noise Modeling

Abstract

This work presents a numerical investigation into jet noise mechanisms, focusing on two of the three primary components: Turbulent Mixing Noise (TMN) and Broadband Shock-Associated Noise (BBSAN). Reynolds-Averaged Navier-Stokes simulations are used to generate the flow-field data required as input for the ACO-JNS aeroacoustic code, which is based on the Morris and Miller acoustic analogy. The numerical simulations of the flow field are first validated against velocity fields obtained from PIV, to assess the accuracy of the aerodynamic input to the noise modeling framework. The comparison shows good agreement, particularly for the streamwise velocity component of the jet. The acoustic frequency spectra show good agreement with experimental measurements. Maximum deviations for the TMN are limited to approximately 3 dB in the entire frequency spectrum. The BBSAN component, instead, despite following the experimental trend, exhibits larger discrepancies of about 6 dB in the entire frequency spectrum. The study computes and analyzes the equivalent sources and reveals distinct mechanisms for the two noise components: TMN is driven by large-scale turbulent structures downstream of the potential core, resulting in preferential downstream directivity. Conversely, BBSAN sources are confined within the shock-cell region, where their spatial extent contracts toward the nozzle exit at higher frequencies and is governed by local pressure fluctuations and axial wavenumber filtering, leading to a dominant sideline directivity. The results indicate that, while TMN appears to be dominated by the secondary stream alone, BBSAN is clearly dependent on both shear layers.