Search for: All records

Flow condensation offers efficient heat transfer but may suffer from performance degradation due to flow regime transitions. These transitions are difficult to characterize in real-world applications, particularly in internal flow condensation due to limited thermofluidic measurements and lack of direct visualization. This study introduces a non-destructive approach for slug flow condensation regime identification using acoustic emission (AE) signals. Experiments were conducted using a vertical downflow condensation setup with normal perfluorohexane (npfh) as working fluid and water as cooling fluid. The AE sensor is mounted on the outer enclosure. A parametric study was conducted over target inlet vapor qualities of 0.6–1.15 and vapor mass velocities of 116–419 kg/m²s, and the resulting acoustic signals were analyzed in both frequency and time domains. Results show a strong positive correlation between acoustic power and vapor mass velocity, with Pearson correlation coefficients of 0.71–0.91. There also exists a weak positive correlation between acoustic power and inlet vapor quality, with the coefficients of 0.02–0.22 across the frequency range. In addition, a novel method for identifying slug flow condensation is developed based on the probability density function (PDF) of acoustic absolute energy (EAbs) using a two-component Gaussian Mixture Model (GMM). Slug flow was found to occur when there is bimodal distribution in the PDF of EAbs, due to the strong periodicity in the EAbs. The significant difference in the Bayesian Information Criterion (BIC) between the one- and two-component GMMs confirms that the model is not overfitted, demonstrating the robustness of the identified bimodal distribution. This method was verified by the pressure drop and acoustic wave propagation analysis based on two accelerometers. The results show that the PDF of EAbs provides a robust, non-destructive, non-optical approach for identifying slug flow condensation in acoustic-based flow regime characterization.