Flow Characteristics Analysis And Description Of Gas-liquid Two-phase Slug Flow

Slug flow is one of the most common flow patterns in gas-liquid two-phase flow.It’s often encountered in a wide range of industrial applications, such as long oiltransportation pipeline, nuclear power plants, chemical plants and other industrialsituations. The influence of gas-liquid slug changes with different productionprocesses. Therefore research about the mechanism and characteristics of gas-liquidflow is of great importance to the security and stability of production process and ofgreat significance to the optimization of production process as well. With thecomplexity and uncertainty of two-phase flow, it’s quite hard to get detailedinformation of the flow mechanism, and ways to describe flow characteristics alsoneed deep and extensive exploration.This work was based on the research of transportation of multi-phase flow, usinghigh-speed camera to capture details of fluid flow process and extracting motionparameters, to get more information about the mechanism and characteristics ofgas-liquid flow. The main work includes:(1) Beased on reviewing and summarizing the existing works on gas-liquid slugflow, a research programme was selected, in which the motion information of flowprocess was acquired with the high-speed video capture technology, and basicparameters of the flow process was then extracted by the methods of image processingand the flow process was described as well.(2) A high-speed video acquisition system was designed with the multiphase flowexperimental apparatus of Tianjin University Key Laboratory of ProcessMeasurement and Control as slug flow generating devices to capture the flow processof gas-liquid slug flow. Camera calibration experiments were carried out to revise thehigh-speed camera, thus the extraction accuracy of the follow-up parameters wasensured.(3) With the utilization of various image processing methods including imageintensification, background reconstruction, spatial filtering, Otsu algorithm andconnected domain screening, methods to extracted the front and tail velocities of theslug flow were developed. By analyzing the acquired data through linear fitting, theexperienced correlation between mixture velocity and slug transition velocity was obtained.(4) Bubble tracking program based on Mean-shift algorithm was designed andimplemented. The velocity distribution of discrete bubbles in liquid slugs can beacquired and ways to describe the formation and development of discrete bubbles inliquid slugs were explored as well.