This paper introduces a novel hybrid source model into an existing acoustic analogy approach to obtain improved predictions of the turbulent mixing noise from cold, round, subsonic, and supersonic jets. The model incorporates new features of the Reynolds stress autocovariance tensor components found in recent experiments. The model parameters are determined from a Reynolds-averaged Navier-Stokes flow solution and experimental data. It is shown that this model significantly improves the predictions relative to previous results, particularly at observer polar angles between 90 degrees to the jet axis and the peak noise direction, indicating the importance of properly modelling relatively subtle characteristics of the autocovariance functions. The results are used to infer the relative importance of individual terms that make up the formula for the acoustic spectrum as a function of jet Mach number, frequency, and observer location.
展开▼
