A Time-Domain Linear Method for Jet Noise Prediction and Control Trend Analysis

Abstract

Large-scale turbulent structures in the form of coherent wavepackets play a significant role in the generation of prominent shallow angle noise radiation of jets. Economical prediction tools often model these wavepackets in the frequency-domain using stability modes of the mean flow. The use of simplifying choices, such as parabolized equations and azimuthal decomposition, provide efficient methods but can impose constraints on rate of streamwise variation of the mean state or geometric complexity. The current investigation develops a time-domain linearized Navier-Stokes-based approach predicated on the mean basic state for two goals: i) to obtain the radiated shallow-angle noise field, including that from imperfectly expanded jets containing shock trains, and ii) to estimate noise control trends with actuator frequency. A previously developed implicit linearization technique repurposing native non-linear Navier-Stokes code capabilities avoids any additional constraints on nozzle geometry, while its time-domain nature facilitates control analysis through transient pulse response. Two other integral components of the method are the sifting of linearized perturbations to isolate the acoustic component with Doak's Momentum Potential Theory, and subsequently Dynamic Mode Decomposition to analyze the response in different spectral ranges. Comparisons with well-validated LES databases show accurate model predictions for super-radiative shallow angle noise, including for hot jets from military-style nozzles. For a specific jet with extensive published experimental data using plasma actuators, it is shown that the method correctly predicts noise amplification at lower frequencies and reduction at higher values, at the observed crossover location. Considerations on the costs associated with the approach, which exploits linearity to extract multiple frequencies with each simulation, are outlined.