Level-educed Wavepacket Representation of Mach 1.8 Laboratory-Scale Jet Noise

The search for an equivalent acoustic source model for high-speed jet noise has recently focused on wavepacket representations. A wavepacket is defined as a spatially extended source with an axial amplitude distribution that grows, saturates and decays, an axial phase relationship that produces directional noise, and correlation lengths longer than the integral length scales of the turbulent structures. This definition of a wavepacket has the same characteristics as the large-scale turbulent mixing noise; if the turbulent mixing noise can be isolated, the associate equivalent acoustic wavepacket—defined as a pressure fluctuation on a cylinder around the jet nozzle—can be found. An estimate of the frequencydependent, spatial variation in the large-scale turbulent mixing noise comes from a similarity spectra decomposition of the measured autospectral density, which in turn leads to data-educed wavenumber axial spectra associated with the frequency-dependent equivalent wavepackets. This wavepacket eduction technique has been applied to acoustical measurements of an unheated, Mach 1.8 jet in the near and far fields. At both locations, the resulting frequency-dependent, data-educed wavenumber spectra exhibit different types of self-similarity for low and high frequency regimes that become apparent when the axial wavenumber is scaled by the acoustic wavenumber, with a transition band between the two regimes. As expected, the data-educed wavenumber spectra can be used to predict field levels in the dominant radiation lobe. Addition of an uncorrelated source distribution, derived from the similarity spectra decomposition associated with the fine-scale turbulent mixing noise, creates a model that accounts for the sideline levels. This field-prediction ability of the wavepacket-plus-uncorrelated-distribution model is tested using the near and far field measurements. When predicting the field at the other location, the model’s average error is less than 2 dB for St = 0.04-0.25 but increases for larger St because the apparent directivity changes from near to far field, likely due to the frequency dependence of the extended source region.