In order to decommission nuclear reactors and to improve the safety of BWR, it is important to estimate the falling behavior of molten core jet in the reactor vessel of BWR when an accident occurred as can be seen from Fukushima Daiichi nuclear power plant accident. Since the BWR lower plenum is consisted with various complicated structures, it is suggested that the jet falling behavior is affected by these structures. Thus we are developing the numerical simulation method to estimate the molten core falling behavior in BWR. To verify the code for the case of the BWR core melt accident, it is necessary to obtain the experimental data and validate the code by comparing the numerical results with the experimental results. The purpose of this study is to investigate the influence of these structures on behavior of jet breakup and fragmentation, and to construct the benchmarks of the numerical simulation experimentally. We used molten core simulant material and simulate the molten core falling behavior, focusing on the hydrodynamic behavior. The 1/10 planar type test section simulated the arrangement of complicated structures in the BWR lower plenum is used. Jet injection experiments were conducted under some conditions that experimental parameters were flow rate and nozzle diameter. To clarify the influence of complicated structures on the jet behavior, experiments were performed in the conditions with and without structures. Jet falling behaviors were recorded by a high speed video camera. The fragment diameters were measured from image by means of image processing techniques. Visual measurement is usually used to measure fragment diameter, but it will contain the arbitrariness and the amount of fragments are small. Since the outline of fragment is easy to recognize by the difference of refractive index between gas and liquid, image processing for measuring the diameters is used in gas liquid flow. On the other hand, it is difficult to recognize the interface in liquid-liquid flow. We developed the new image processing filter for detecting the outline of fragments precisely and established the image processing method including this filter. We measured about ten thousand fragments precisely and automatically. The measurement of fragment diameter was implemented by the image processing method mentioned above. The histogram of fragment diameter distribution shows that it can be fitted by the lognormal distribution in condition with and without structures. We calculated the volume median diameters in all conditions. The diameters were smaller that depended on the increasing injection velocity. Comparing between condition with and without structures, the fragment diameters became small in condition with structures than without structures. Since the velocity of tip of the jet was larger in condition with structures (Saito et al., J. Nucl. Sci. Tech, 2015), the velocity gradient between the jet and ambient fluid also would be larger. The shear force strongly acting on the interface made the diameter small.

This content is only available via PDF.
You do not currently have access to this content.