Correlation-enhanced short-time analysis of coherent structures for acoustic bursts in subsonic jets

A correlation-enhanced event-driven decomposition (CEED) method is developed to analyze the coherent structure during the acoustic burst in subsonic jets. To sample the acoustic burst signal for the CEED, a conditional filter is proposed. The correlation between the acoustic burst signal and the signal from the flow field is introduced into the decomposition, to deal with the low signal-to-noise ratio issue between these two signals. Suitability for the subsonic jets is demonstrated for the CEED, based on validated large eddy simulation results. Then, the short-time evolution of the coherent structure is analyzed by the CEED. The results show that the sampled acoustic burst signal is intermittent. Its spectrum appears the same shape as that of the original signal. Thus, the acoustic burst is dominant in the sound radiation. The correlation is found to accelerate the convergence of the CEED. The leading mode of the CEED is shown able to represent the short-time evolution of the coherent structure clearly. During the acoustic burst, the leading mode appears as a train of puffs, and it presents three stages: (1) Two in-phase puffs enclose an anti-phase puff. These puffs move downstream with their size growing. (2) The anti-phase puff vanishes, while the two in-phase puffs merge into a large puff. The large puff forms a wavefront. (3) Three puffs form waves in sequence. Together, these waves compose a train of waves. The merging accelerates the puff into a supersonic state intermittently; therefore, the merging is responsible for the acoustic burst.