Experimental and Numerical Analyses of Sweep-out Phenomenon in Downcomer of Reactor Pressure Vessel

During the core reflooding phase of a loss-of-coolant accident, sweep-out phenomenon may occur in the downcomer region of the reactor pressure vessel (RPV) with the action of steam flowing from the cold-leg. When sweep-out occurs in the downcomer of RPV, liquid is entrained and carried out by the steam. Then, the core reflooding process is affected. To investigate the sweep-out phenomenon in the downcomer, an RPV integral test facility was designed and constructed based on scaling criteria and design features of prototype reactor. The experimental investigations were conducted. Subsequently, the onset point of sweep-out under different operating conditions and its influencing mechanisms were obtained. Based on the experimental data, different prediction models for the onset height of sweep-out in the downcomer were compared and a newly prediction model was proposed. Then, the mechanistic analysis of sweep-out phenomenon was conducted. Experimental results and mechanistic analysis indicate that when the initial liquid level in the downcomer is relatively high, the level gradually decreases under a certain steam flow rate until it reaches the sweep-out initiation point. As the steam flow rate increases, the corresponding dimensionless onset height gradually increases. Within the parameter range of experimental conditions, the maximum observed dimensionless onset height is 2.9. Comparing with the experimentally derived onset height models of sweep-out phenomenon, the existing prediction models show significant deviations. Furthermore, these existing models do not account for the absence of steam flow. Numerical simulation was conducted using the system analysis code with different nodalization schemes by adjusting circumferential nodes or axial node heights. The simulation results indicate that the calculated dimensionless onset heights are higher than the experimental values. The differences between numerical calculations and experimental data are systematically examined. Finally, the applicability of system analysis code in this context was critically discussed and the corresponding analytical methodology was proposed. This approach aims to improve prediction accuracy.