The Analysis Of Gas-liquid Two-ghase Flow Pattern Based On DFA

Gas-liquid two-phase flow is widespread in nature and in industrial processes.The two-phase flow characteristics has a very important role in the study of two-phaseflow. It can largely affect on the research and development, design and normaloperation of equipment, and also affect the measurement accuracy of flow processparameters and other relevant characteristics. Therefore, the analysis of two-phaseflow characteristics will continue to be an important direction for two-phases.Detrended fluctuation analysis (DFA) method is used to analyze gas-liquidtwo-phase flow patterns of three typical (bubble flow, slug flow, churn flow) in thispaper. In view of the non-linearity and non-stability of two-phase flow pattern signals,Hurst exponent which is the characterization volume of the nonlinear complex systemwas applied to study flow pattern signals. First, multi-scale wavelet decompositionbased on wavelet analysis was used for analyzing the collected gas-liquid two-phaseflow conductance fluctuation signals. We use DFA method to calculate the Hurstexponent for each scale. By comparing the results we found that two methods coulddescribe the fractal characteristics of gas-liquid two-phase flow at different scales, andreveal the two-phase flow motion characteristics in the details of local and global.From a practicalpoint of view, the use of a single fractal analysis of different flowcharacteristics is not enough, so we use multi-fractal detrended fluctuation analysis(MF-DFA) to analyze conductance wave signals multi-fractal characteristics of thevertical rising pipe of gas-liquid two-phase. By calculating wave function, generalizedHurst exponent, scaling function, multi-fractal spectrum parameters, the fluctuationsingularity of the conductance fluctuating signals in partial and different levels wasquantified in the paper. The results of the study show different conductance fluctuatingsignals have long-range correlation and is multi-fractal process, multi-fractal spectrum parameters can reflect dynamic characteristics of different flow patterns.Finally, thepaper studied the causes of the two-phase multi-fractal by comparing the Hurstexponent of the front and rear signals after the random rearrangement and found it iscaused by both long-range correlation and thick-tailed probability density distribution,and it is mainly caused by the correlation. The results show the multi-fractal detrendedfluctuation analysis of conductance signals can have better characterizing changes inflow patterns, and it is a useful tool to understand the flow pattern transitionmechanism and dynamic evolution characteristics.