Characterization of the horizontal gas–liquid two-phase flow patterns using acoustic measurement techniques

The characterization of gas–liquid two-phase flow patterns is crucial for monitoring stability in industrial applications. However, the impact of these flow patterns on gas–liquid two-phase flow-induced sound (GTFIS) emissions remains inadequately understood. In this paper, the GTFIS signals at high liquid velocities within a horizontal pipe are captured using precision hydrophone. A novel approach for the analysis of acoustic signals that synergistically combines Variational Mode Decomposition, Singular Value Decomposition, and nonlinear signal processing methods is proposed to assess the flow dynamics. The results show that the GTFIS signals exhibit chaotic characteristics. Two distinct Hurst exponents are observed for each acoustic signal: one is greater than 0.5 and the other is less than 0.5. The coexistence of randomly moving small bubbles and intermittent bubbles contributes to an increase in the complexity of the attractor phase trajectory of acoustic signals, resulting in a maximum value in the correlation dimension. The expansion radius and correlation dimension of mesoscale acoustic signals can serve as early warning indicators for the transition from dispersed bubble flow to slug flow. In conjunction with the characteristic parameters of multi-scale entropy, the flow patterns can be effectively characterized.