Research On The Two-phase Flow Behavior And Its Predication In The Water Pipe

In the long-distance pipeline water delivery process,the liquid-gas two-phase flow in the pipeline is a common phenomenon.Understanding this phenomenon is very important for pipeline protection.The key to understanding this phenomenon is to describe the two-phase flow patterns in the pipeline.Two-phase flow pattern is formed by phases interaction,pressure drop and heat and mass transfer effects.However,due to the complex geometry of the pipe,the two-phase flow undergoes significant characteristic changes in the flow pattern.The focus of this paper is to investigate the effect of a sudden expanded region in a closed flow loop with a transparent conduit on the flow characteristics of the twophase liquid-gas flow.The study of the dynamics of the slug flow in the sudden expansion pipe is the focus in this paper.The experimental setup was used to test the effects of two key parameters,liquid flow and gas flow.Based on these two parameters,the flow characteristics of single-phase flow and two-phase flow near the sudden expansion zone were studied.Using high-speed photography technology,the characteristics of water jets in single-phase flow and two-phase flow in horizontal tubes were studied,and the velocity of bubbles in the expansion region was quantified.The observed bubble collapse phenomenon was quantitatively described by image analysis and the liquid film thickness of the bubble was estimated.At the same time,this paper studies the influence of the sudden expansion zone on the water jet trajectory profile from both experimental and theoretical aspects,and on the basis of the existing experimental data,the pipeline water jet model was modified to improve the prediction accuracy.Finally,the Bayesian network model was used to discretize the flow patterns obtained by the experiment,and the information of the flow patterns was used as the data sample to finish the parameter learning of the Bayesian network.Finally,the Bayesian network model was obtained and the flow pattern was predicted.