Generalized Acoustic Analogy Modeling of Hot Jet Noise

Abstract

A generalized acoustic analogy model is implemented for the hot and cold static high-speed jet cases corresponding to conditions of the Strategic Investment in Low-carbon Engine Technology (SILOET) experiment performed by QinetiQ. The model is statistical and based on the covariance of fluctuating Reynolds stresses and enthalpy source terms in accordance with Goldstein’s theory. These quantities are obtained from a solution of the validated large-eddy simulation accelerated on graphics processing units. The covariance of fluctuating enthalpy terms is represented by the Khavaran and Bridges model, which involves the velocity autocorrelation function and temperature gradient in the jet. The velocity autocorrelation function and the covariance of fluctuating Reynolds stresses are approximated by an analytical Gaussian-exponential function. The input parameters of the acoustic model include single-point quantities such as the time-averaged local streamwise velocity, sound speed, temperature, vorticity magnitude, turbulent kinetic energy, and dissipation rate. Modeling choices for the acoustic length and time scales based on either the turbulent kinetic energy dissipation rate or mean flow vorticity are discussed. In each case, the model involves two dimensionless parameters associated with the acoustic correlation length and time scale. These two parameters are calibrated based on the far-field noise data at 90° observer angle. The sensitivity of the calibration coefficients to the jet type, cold versus hot, is investigated. Furthermore, the same acoustic model without recalibration is applied for far-field noise predictions of two NASA Small Hot Jet Acoustic Rig jets for acoustic Mach numbers 0.5 and 0.9. To probe the sensitivity of the acoustic model to the input flow data, Reynolds-averaged Navier–Stokes solutions of the same jet cases are considered along with the large-eddy-simulation solutions. The comparison of the predicted far-field noise spectra with the experiment is discussed in each case.