Temporal And Spatial Evolution Characteristics Of Gas-Liquid Two Phase Flow Pattern Based On Image Texture Spectrum Descriptor

Two phase flow pattern is an important research aspect and the new investigation method of flow pattern dynamics is one of the significant research direction.In this study, firstly the applicability of local binary pattern algorithms to extract image texture feature of the gas-liquid two phase flow is discussed, then the existed method is employed to effectively analyze the gas-liquid two phase flow dynamic image. In vertical upward gas-liquid two phase flow pipe,the study results show that the amplitude fluctuation of texture characteristic parameter indexes of bubble flow is lowest and shows very random complex dynamic behavior; the amplitude fluctuation of slug flow is higher and shows intermittent motion behavior between gas slug and liquid slug, and the amplitude fluctuation of churn flow is highest and shows better periodicity; the amplitude fluctuation of bubble-slug flow is from low to high and oscillating frequence is higher than that of slug flow,and includes the features both of slug flow and bubble flow; the slug-churn flow loses the periodicity of slug flow and churn flow,and the amplitude fluctuation is high. In inclined upward gas-liquid two phase flow pipe, the amplitude fluctuation of texture characteristic parameter indexes of bubble flow is lower than that of slug flow and shows very random, and the amplitude fluctuation of slug flow is relatively higher with a certain periodicity.The study indicates that the image texture characteristic parameter indexes of different flow pattern can reflect the flow characteristics of gas-liquid two phase flow, and provides a new approach to understand the temporal and spatial evolution of flow pattern dynamics.