Validation and improvement of a numerical model for freezing and blockage formation of solid-liquid flow of molten fuel in the core disruptive accident of FBR

In order to validate and improve a numerical model of the SIMMER code, which analyzes event progression of core disruptive accidents (CDAs) in sodium-cooled fast reactors, a numerical study was conducted. We focused on the numerical model for freezing and blockage formation of solid-liquid multiphase flow in this study. This model has not yet validated thoroughly in the SIMMER code, while a freezing and blockage formation model for flow of simple molten fuel had already been established mechanistically and introduced into the SIMMER code. Since the solid components may inhibit the molten-fuel discharge during CDAs, it is important to simulate the solid-liquid flow appropriately. Therefore, we performed validation and improvement of the freezing and blockage formation model of solid-liquid flow in the SIMMER code through comparison with a series of THEFIS experiments performed at KIT in 1980s. This experimental study could simulate well the freezing and blockage formation of solid-liquid flow by injecting molten alumina as a molten-fuel simulant into a quartz tube where an alumina particle bed was installed. The SIMMER analysis reproduced well penetration behavior of the alumina melt observed in the THEFIS experiment. However, in a few cases, the melt did not penetrate over the particle bed differently from the experimental result. This was caused by the inconsistency between the physical phenomena in the THEFIS experiment and numerical modeling in the SIMMER code. The numerical model would overestimate the effect of the particle bed on the blockage formation of solid-liquid flow. Therefore, we improved the numerical model based on understanding of the physical phenomena involved in freezing and blockage formation. As the result, the experimental penetration behavior was reproduced by the improved SIMMER code better than the existing SIMMER code. In conclusion, it was confirmed that the SIMMER code could analyze appropriately the freezing and blockage formation of solid-liquid flow by improvement of the numerical model through the comparison with the experimental study. (author)