Prediction of Sound Generated by Complex Flows at Low Mach Numbers

In many practical applications, sound is generated by interaction of turbulent flow with solid objects. In such cases, sound waves experience multiple reflections from solid objects before they propagate to a far-field point. In these situations the sound spec-trum exhibits a rich frequency content consisting of broadband and tonal components. To predict the acoustic field in these situations, a general aero-acoustics framework is required to operate in complex environments. More importantly, the employed method must avoid many simplifying assumptions often made about the geometry, compactness or frequency content of sound sources. The objective of the present work is to develop, validate and demonstrate the functionality of such a computational framework. The prediction of flow-generated sound requires accounting for the physics of both unsteady flow and sound waves simultaneously. Since these two phenomena exhibit very different energy and length scales, the prediction of flow-generated sound is challenging especially from the numerical perspective. Sound waves carry only a minuscule fraction of flow energy and high-order numerical schemes are required to keep the sound waves intact. Additionally, at low Mach number regimes, the acoustics CFL number imposes extremely