Research On A New Method For Void Fraction Measurement Of Gas-Liquid Two-Phase Flow

Void fraction is one of the most important parameters in gas-liquid two-phase flow, which is indispensable to process control and other parameter measurement of gas-liquid two-phase flow system. Due to the complexity of gas-liquid two-phase flow, void fraction measurement is a difficult problem which has not been solved both in scientific research field and industrial application for a long time. Conductivity detection technique is a classical method for void fraction measurement of gas-liquid two-phase flow. However, this technique is mostly based on contact conductivity detection, which limits its long-term application in industrial field.Capacitively Coupled Contactless Conductivity Detection (C4D) is a new conductivity detection technique, which has the advantages of non-contact, simple construction, low cost, etc.. However, few research works concerning the application of C4D technique to the parameter measurement of two-phase flow in conventional pipes have been reported. In this thesis, a new method for void fraction measurement of gas-liquid two-phase flow in conventional pipes is presented combining C4D technique and data mining techniques.Main innovation points and contributions of this thesis are listed as follows:(1) Combining C4D technique and data mining techniques, a new method for void fraction measurement of gas-liquid two-phase flow is presented.(2) In order to overcome the influence of flow pattern on the void fraction measurement, the void fraction measurement models for three typical flow patterns (bubble flow, annular flow, and stratified flow) are established by a developed six-electrode array C4D sensor. Three kinds of modeling methods are studied and compared:1) Modeling method based on PLSR technique,2) Modeling method based on the combination of PCA technique and LS-SVM technique,3) Modeling method based on the combination of PCA technique and WLS-SVM technique.(3) Static experiments are carried out for the void fraction measurement with the outer diameters of 25 mm,32 mm,40 mm and 50 mm, respectively. Experimental results show that the three kinds of void fraction measurement models based on data mining techniques are successful. The maximum absolute differences between the measured void fraction and the reference one for three typical flow patterns in four pipes are all less than 10%. The void fraction measurement model based on the combination of PCA technique and WLS-SVM technique has a higher accuracy. Dynamic experiments are carried out with the outer diameter of 25 mm. Experimental results show that the maximum absolute differences for three typical flow patterns are less than 15%. Static and dynamic experimental results show that the presented method for void fraction measurement is feasible and effective.