COMPARISON OF NEAR-FIELD EVENTS AND THEIR FAR-FIELD ACOUSTIC SIGNATURES IN EXPERIMENTAL AND NUMERICAL HIGH SPEED JETS

Abstract

Two different approaches are applied to near-field (NF) velocity field and far-field (FF) pressure signals to gain better understanding of the flow structures that contribute to high speed jet noise. We use laboratory data from a 10kHz TRPIV experiment data of Mach 0.6 jet and numerical data from an 80kHz LES database at Mach 0.9 jet. From the NF, over 20 representative diagnostics are extracted as time traces, of which about half give high correlation with the far-field. Utilizing cross-correlation and wavelet analysis, we locate the frequency band where information is transferred from NF to FF. Furthermore we identify excerpts in time and frequency domain that act as major correlation contributors. The lists of events based on FF only (acoustic footprints) and on NF-FF correlations are compared and show good similarity, which validates both techniques. Finally, the lists of events are separated into categories based on their properties, including magnitude, frequency, and axial and transverse location. INTRODUCTION The noise sources of high speed jet was initially assumed to be random, until the discovery of turbulent coherent structures offered another view to understanding the dynamics. Coherent structures are firstly found to be in part responsible for the occurrence of acoustic spectrum peak (Mollo-Christensen, 1967; Crow and Champagne, 1971). The frequency range associated with the loudest noise was found to be 0.1 < St < 0.7 (Michalke, 1977). Coherent structures are more amenable to flow control toward noise reduction, and related studies dominate the literature. The level of coherence was important for jet noise production since a periodic shear layer would not generate farfield noise. Also, it was shown by Michalke and Fuchs (1975) that while the first few azimuthal modes were associated with far-field noise, the axisymmetric mode is not the most efficient. This was related to the coherence level of the velocity field, defined as the ratio of the size of the source to that of the eddies. From the research of Wei and Freund (2006), the more ordered propagation of near-field structures was related with the far-field noise reduction. Cavalieri et al. (2011b) showed that the far-field pressure reFigure 1. Experimental facilities of jet flow measurement (courtesy K.R. Low). sulted from the near-field wave packets propagating through a modeled flow field had good correspondence with the experimental data. What we observe in this paper adds to the coherent part of analysis. DATA DESCRIPTION In this study, we use two sets of data, one experimental and the other numerical. Our two algorithms, distinct for far-field and near-field processing, are applied to both databases, providing validation of the procedures in spite of the different Mach numbers. Experimental Data The experiment was performed in a large-scale (approximately 8000 f t3) anechoic chamber in Syracuse University. The data we use for this paper is for a cold jet with Ma = 0.6 and its Reynolds number is 700,000. The top view of the experiment facility is shown in fig. 1. The exit of the nozzle has the diameter of 0.0508m. To measure far-field pressure, 6 microphones were