Acoustic Sensing for Regime Identification and Local Heat Transfer Characterization of Flow Condensation

Accurate characterization of flow condensation is essential for optimizing phase-change processes but remains a significant challenge. This study introduces a non-destructive framework using Acoustic Emission (AE) sensing to probe flow regimes and local heat transfer behavior during condensation. A series of tube-in-tube internal and external flow condensation experiments are performed under varying working fluids, flow rates, and vapor qualities with AE sensors recording high-frequency signals at 1 MHz. The results show that two-phase condensation significantly amplifies acoustic power compared to pure vapor and liquid flows. Analysis of AE power spectra reveals strong correlations with local heat transfer characteristics during condensation. A hybrid machine learning approach is developed, where unsupervised clustering distinguishes superheated vapor from two-phase flows, while supervised regression models predict local equilibrium vapor quality, heat flux, and heat transfer coefficient. Additionally, an acoustic indicator based on Gaussian Mixture Models of AE absolute energy is proposed to detect slug flow regime. This work demonstrates the potential of AE sensing as a robust and non-destructive measurement for regime identification and heat transfer characterization in flow condensation.