Characterization of Gas–Liquid Two-Phase Flow by Correlation Dimension of Vortex-Induced Pressure Fluctuation

Characterization of flow behaviors is one of the most challenging problems in a gas-liquid flow system. In this paper, correlation dimension, a chaotic characteristic indicator, was introduced to characterize the gas-liquid two-phase flow behaviors by using the fluctuating pressure induced by a bluff body. An artificial neural network was trained to help select suitable flow parameters that were combined with correlation dimension to construct a novel gas-liquid flow pattern map, which was able to distinguish between the bubble, bubble/plug transitional, plug, slug, and annular flows with reasonable accuracy. Furthermore, a quantitative correlation with the form of u _g = AD ₂ ^B u ^C was established by the universal fitting and the pattern-specific fitting with the coefficients of determination R ² approaching to 1. In view of the simplicity and the convenience of vortex generation and pressure measurement, the correlation dimension-based method provides an effective and practical idea to gas-liquid two-phase flows study.