Study of using simulants for the assessment of LWR meltdown models

In the past few years, several analytical models and computer codes have been developed for the analysis of light water reactor (LWR) severe accident phenomena (e.g., SCADAP, MAAP, MMELRPI, new versions of MARCH, and others). In these codes, several assumptions are usually required concerning the modeling of various complicated physical mechanisms associated with core degradation and melting. Such assumptions can be verified only by comparing the results of calculations against experimental data. Unfortunately, the existing experimental evidence is very limited in general, and, in particular, in the areas of molten fuel relocation, melt interaction with various core and reactor vessel structures, and failure mechanisms of these structures. Moreover, many experiments performed so far are basically integral rather than separate-effects-type experiments, and thus they are of a limited help to model developers. Because of the high cost of experiments using actual reactor materials, the only practical approach to detailed parametric experiments oriented toward studying specific phenomena is through the use of simulants. The main question in this case concerns the similarity between the phenomena observed in the experiments and those occurring during hypothetical reactor accidents. The purpose of this paper is to present data and the results of analysis concerning the use of eutectic metal and paraffin wax to simulate molten reactor fuel (UO2) and solid core structures (steel), respectively