Comparison of supersonic full-scale and laboratory-scale jet data and the similarity spectra for turbulent mixing noise

Abstract

Broadband, partially correlated noise radiated from supersonic jets has characteristics that scale with nozzle size and flow properties. In particular, the spectral content of jet noise and variation with angle in many cases agree with empirically derived similarity spectra for large and fine-scale components of turbulent mixing noise [Tam et al., AIAA paper 96-1716]. In previous studies, measurements made near the F-22A Raptor agreed remarkably well with the similarity spectra, with two exceptions. First, the high-frequency slopes seen in the data were shallower than the similarity spectra at many angles. Second, the data exhibit a double frequency peak, which is absent from the similarity spectra [Neilsen et al., J. Acoust. Soc. Am. 132, 1993 (2012)]. These observations are explored further by examining the spectral characteristics of noise from a different military jet and a laboratory-scale, unheated jet. In both cases, there is evidence that for supersonic cases the measured spectra are shallower than the similarity spectra due to nonlinear propagation effects. In addition, the military data support the observation that the double spectral peak is a feature of full-scale jet noise. Recommendations are made for applying the similarity spectra to predict spectral levels for full-scale jets. © 2013 Acoustical Society of America.