Non-linear Analysis of Two-Phase Flow Fluctuation by Means of SVD Entropy Surrogate Data Analysis

Two-phase flow in pipes displays a diverse range of patterns, influenced by factors such as pipe inclination, liquid superficial velocity and gas superficial velocity. Accurate classification of these flow patterns is essential for controlling fluid transport in various industries. This study investigates the non-linear dynamics of two-phase flow in inclined pipes using voltage fluctuations as a proxy signal. Singular value decomposition (SVD) entropy and surrogate data analysis were employed to quantify the non-linearity content in the voltage signals. The results, obtained from examining two-phase flow patterns under varying pipe inclination and gas superficial velocity, reveal that both entropy and non-linearity are dependent on flow conditions. The primary non-linear contribution was found to be linked to discontinuities in the proxy signal, reflecting the heterogeneity of the flow pattern. This analysis demonstrates the effectiveness of the entropy-non-linearity approach in classifying flow patterns based on pattern diversity and non-linearity contribution.