Study On New Flowrate Measurement Method For Air-water Two-Phase Flow

Gas-liquid two-phase flow widely exists and is widely applied in many industries such as power, petroleum and chemical engineering. Measurement of its flowrate is of great scientific and engineering importance. Due to the inherent complexity of two-phase flow, its parameter measurement is quite difficult, and its flowrate measurement is always an unsolved problem around the world. Aimed at. the online measurement of two-phase flowrate, based on a Venturi meter and an ERT conductance sensor, and by using the Least Squares Support Vector Machine (LS-SVM) method, this dissertation carried out the research work of flow pattern identification, void fraction measurement, mass quality measurement and flowrate measurement of air-water two-phase flow.The main work and innovative points are listed as follows:1. Based on the combination of a Venturi meter and a 16-electrode ERT conductance sensor, a new method is proposed for air-water two-phase flowrate measurement with the Least Squares Support Vector Machine (LS-SVM) method and the gas-liquid flowrate measurement correlations for throttling device. The experimental result shows that this method can measure the flowrate of air-water two-phase flow effectively.2. Based on the 16-electrode ERT conductance sensor, a new method is proposed for flow pattern identification of air-water two-phase flow with the Least Squares Support Vector Machine (LS-SVM) classification method. This method uses the conductance data of two-phase flow which is measured by the ERT conductance sensor to develop the flow pattern classifier, and then to identify the flow pattern. The image reconstruction process in conventional ERT flow pattern identification is eliminated. Experimental research indicates that this method is effective for flow pattern identification, and the success rate in flow pattern identification is high for the four typical flow patterns:stratified flow, annular flow, slug flow and bubble flow.3. Based on the 16-electrode ERT conductance sensor, a new method is proposed for void fraction measurement of air-water two-phase flow with the Least Squares Support Vector Machine (LS-SVM) regression method. This method develops different void fraction measurement models for different typical flow patterns of horizontal-pipe air-water flow. In the practical measurement, the appropriate void fraction measurement model is selected according to the flow pattern identification result, and the void fraction is calculated from the conductance data of two-phase flow. Experimental research indicates that this method is effective for void fraction measurement. In this method, the measurement speed is improved because the image reconstruction in conventional ERT void fraction measurement is eliminated, and the measurement accuracy is improved because the flow pattern identification is introduced to minimize the influence of flow pattern on the void fraction measurement. Static experiment results demonstrate that the maximum absolute error of void fraction measurement is 6.0%.4. Referring to the existing void fraction-quality correlation, the void fraction-quality correlation is optimized under the experimental condition and for different flow patterns. In the practical measurement, the appropriate correlation is selected according to the flow pattern identification result, and the mass quality is calculated from the void fraction, so the difficulty to obtain the mass quality in conventional throttling device measurement methods is overcome. Additionally, because of the introduction of flow pattern identification, the influence of flow pattern on the mass quality measurement is minimized. Experimental results show that the root mean squares of relative errors for mass quality measurement are 0.219,0.177,0.097 and 0.150 for stratified flow, annular flow, slug flow and bubble flow, respectively.5. The existing gas-liquid flowrate measurement correlations for throttling device are compared, and then are optimized under the experimental condition and for different flow patterns. According to the comparison results of these correlations, the optimized flowrate measurement correlations based on flow pattern are obtained.6. Combining the flow pattern identifier, the void fraction measurement model, the void fraction-quality correlation and the flowrate measurement correlation, the online flowrate measurement of air-water two-phase flow is implemented. Firstly, the conductance data of two-phase flow obtained by the ERT conductance sensor are used to identify the current flow pattern with the flow pattern classifier. Then, according to the flow pattern identification result. the appropriate void fraction measurement model is selected to calculate the void fraction. Next, the appropriate void fraction-quality correlation is selected to calculate the mass quality from the void fraction. Finally, the flowrate measurement correlation for this flow pattern is used to calculate the flowrate from the differential pressure measured by the Venturi meter and the mass quality, so the online measurement of air-water two-phase flowrate is accomplished. Experimental results demonstrate that the new flowrate measurement method for air-water two-phase flow is effective, and the measurement accuracies are satisfactory for the four typical flow patterns:stratified flow, annular flow, slug flow and bubble flow.