Quantitative void fraction measurement method by neutron radiography was developed using a neutron absorber grid based on an umbra method. The cross-sectionally averaged void fraction for various flow patterns in air-water two-phase flow in a tube was measured quantitatively. Uncertainty analyses were carried out for the measured values and the accuracy of the measurement was discussed for each flow pattern. Experimental results agree well with the prediction by drift flux model. The quantitative measurement method was applied to two-phase flow in a concentric annular tube, a tube with a spiral wire, a rod bundle, a capillary tube and a plate fin heat exchanger. The CT reconstruction method was applied to measure the quantitative void fraction distributions in the cross section.
INTRODUCTIONVisualization by neutron radiography is suitable for two-phase flow studies. Measurement of the void fraction is important to analyze the two-phase flow phenomena. However, it is not easy to measure the void fraction quantitatively from the images obtained by neutron radiography. The brightness of the image is not always proportional to the value calculated by linear attenuation equations. Some methods have already been proposed for the quantitative measurement by neutron radiography. Glickstein et al. [1] used Monte-Carlo simulation for quantitative measurement and applied it to void fraction measurement by estimating the neutron scattering as an inverse problem of neutron interaction. No special experimental preparation is required with this method. The profile of the void fraction distribution should be assumed to obtain a unique answer to solve the inverse problem. Hirdlika and Peterka [2], Hibiki et al. [3] and Yoshii and Kobayashi [4] proposed to increase the distance between the converter and the object to reduce the possibility of the scattered neutrons falling on the image in the converter. This method is easy but the spatial resolution is reduced if the L/D of the system is not high.Tamaki [5] inserted a collimator made of neutron absorber between the converter and the object to reduce the neutrons the direction of which was changed after scattering in the object. A honeycomb painted with Gadolium was tested. By changing the L/D of the honeycomb the reduction rate of the scattered neutrons could be estimated. When the L/D of the system was low, the visible field was quite reduced due to the shadow of the collimator on the converter. Kobayashi et al. [6] employed an umbra method to compensate the neutron scattering effects. A cadmium neutron absorber tape was placed between the object and the source to make the umbra of the tape on the converter. The brightness due to the scattered neutron could be measured by the brightness at the umbra. Compensation was made by subtracting the image with the tape from that without it. This method is expected to compensate the effects of the scattered scintillation optical rays. One-dimensional quantitative distribution along the tape is obtainable with this method.Murata et al. [7...
