Nonlinear characterization of gas liquid two-phase flow in complex networks

A new method is proposed to study nonlinear dynamic of gas liquid two-phase flow through a vertical upward pipe with the aim to make clear the two-phase flow behavior, from the complex networks perspective. The experimental test section is a transparent plexiglass pipe with inner diameter is 40 mm and the length is 1.5 m. Five kinds of flow pattern including bubble flow, bubble-slug flow, slug flow, slug-churn flow and churn flow are measured, and then constructed 178 corresponding flow pattern complex networks in order to reveal the dynamical characteristics of two-phase flow. It is found that different flow pattern complex network present different community structure, when the correspondence between the threshold and the mean of similarity matrix meet specific proportion. The results show that the power-law exponent of degree distribution and network information entropy of flow pattern complex network is sensitive to the flow pattern transition, which can be employed to characterize the nonlinear dynamics of two-phase flow. It is demonstrated that complex network approach can be an effective tool to study nonlinear dynamics of air water two-phase flow.