Feature of acoustic sound signals involved in vapor bubble condensation and its application in identification of condensation regimes

Experiments were carried out on the sound pressure oscillations condensation regime map, and bubble collapse frequency during the direct contact condensation of vapor with the aid of an acoustic hydrophone and a high-speed video camera. The injection rate of vapor and liquid subcooling in the experiments were 0.19-3.73 m3/h and of 10-70 K, respectively. Four different condensation regimes were identified according to features presented on the bubble surface and whether and when the bubble collapse occurred. The state-of-the-art signal processing methods (statistical, spectral, fractal and discrete wavelet transform analyses) were applied to processing the detected acoustic signals. The results showed that only the kurtosis and DWT in the four methods could distinguish the different regimes well. Furthermore, the spectral and fractal analyses showed that strongly persistent behavior in the signals corresponded to the dominant frequency in the range of 120-400 Hz might be arisen from the periodic variation in the vapor bubble volume. While that corresponded to peaks with frequency higher than 7000 Hz in transition and capillary wave regimes were probably the high-frequency oscillation in pressure induced by sudden bubble collapse. Contrarily, the first peak in 0-200 Hz caused by the periodic bubble break-up or split-up was high anti-persistent. DWT analysis showed that the sound pressure oscillation introduced by bubble collapse was similar to that by bubble split-up for all condensation regimes, whereas at very high frequency two different types of oscillations arose. Furthermore, the bubble collapse frequency increased with increase in liquid subcooling and vapor injection rate, and could be obtained from the spectral and Hurst analyses of the signals indirectly.