Measurement Of Local Flow Characteristics In Gas-Liquid Two-phase Flows Using Planer Laser-induced Fluorescence Method

Gas-liquid two phase flows are widely existed in chemical engineering,pharmaceutical engineering,petroleum industry,energy industry,power engineering,ect.The flow presents various distribution and structure due to complex interaction and relative motion in two-phase flows,which lead to varied flow patterns.It notes that declined films are obviously observed in two-phase slug and churn patterns,and the film structure presents significant unstable wavy feature;with large turbulent energy,obvious structure revolution will be generated in the declined films,and the films are easily entrained with multiscale dispersed bubbles.Study on local structure features in two-phase flow has great significance on the analysis of flow pattern transform mechanism and establishment of flow parameter models.In this paper a planer laser induced fluorescence measurement system has been proposed to obtain local flow structure of gas-liquid two-phase flow in vertical upward small pipe,and the flow features and transform regular of slug,churn and bubble flow patterns have been studied.We extracted the edges of two-phase flow based on two image processing methods and meanly studied the time and space distribution of slug flow and churn flow,as well as the micro structure feature of the entrained bubbles.The innovated results of this study are stated as follows:1.A planer laser induced fluorescence measurement system for the gas-water two-phase flow local structure measurement was designed,and the measurement system from the perspectives of experiment device and the light path model were optimized,considering the light path distortion caused by refraction and reflect in the two-phase intermittence.The vertical small pipes gas-water flow experiment was conducted,and clear gas water 2D flow structure images were obtained.According this,the feature and transform regular of gas-water two-phase slug,churn and bubble flow were analyzed.2.Automatic threshold binarization and digital mask methods were used to process PLIF images of gas-water two phase flow,and film thickness transient transform process,space distribution characteristics and probability density distribution features were studied,and then the relationship between film structure time and space evolution and gas-water two-phase flow parameters were explored,and thus we draw the conclusion that the film thickness parameter can be used as an effective parameter for the gas-water two-phase flow patterns identification and transform mechanism study.3.The Random Structure Forests method and Reacher Convolution Feature neural network method were used to extract the edges based on the PLIF gas-water two-phase flow images.Using the image edges results,the bubble amount and size information entrained in the liquid films were obtained,and the multiscale bubbles distribution characteristics entrained in the films were obtained,and then the bubble generative mechanism in the films was discussed.This part can provide valuable clues to research the flow patterns transform regular in gas-water two-flow.