Two-phase flow pattern transition behaviors on experimental established ordinal pattern networks

Identifying the flow pattern transition dynamics is a fundamental challenge in modeling a two-phase flow system. In this paper, we investigate the gas-liquid two-phase flow pattern transition behaviors with analyzing the topology structures of the experimental established gas-liquid two-phase flow complex networks. First, we carry out a series of gas-liquid two-phase flow experiments in a vertical 50 mm inner diameter pipe. During the experiments, the two-phase flow fluctuation signals are collected and used to establish the ordinal pattern complex networks, which represent different flow patterns. Then, we employ a K-core decomposition method to identify the hierarchical structures of our established flow pattern networks. We find that the decay rate of the K-core size is sensitive to the flow conditions and can be a potential metric for identifying the flow pattern transitions. Additionally, we analyze the network homology persistence, which indicates the loop structures in the flow pattern networks. The persistence indexes-maximum persistence and persistence entropy-are used to investigate the flow pattern oscillatory behaviors along with the flow pattern transitions. This research provides a novel way for investigating the flow pattern transition behaviors of a gas-liquid two-phase flow system, which are expected to be applicable in other complex fluid systems.