Development of a horizontal two-dimensional melt spread analysis code, THERMOS-MSPREAD Part-2: Special models and validations based on dry spreading experiments using molten oxide mixtures and prototype corium.

• MSPREAD was validated based on the three dry spreading experiments. • Influences of generated gases on the friction force and heat transfer were modeled. • The cell size convergence was confirmed for the highly viscous melt spreading. • The gas enhanced heat removal at the upper surface decelerated spreading. • The weir anchoring model was developed to interpret the anisotropic spreading. MSPREAD has been developed to evaluate the multi-dimensional melt spread on both dry and wet floors while considering a number of influential mechanisms. Part-1 of this two-part series described models, numerical methods and verification results. Part-2 focused on the validation of the models related to dry spreading behaviors. The special models necessary for in-depth studies of the experimental results were introduced, then the target phenomena included in the dry spreading were related to the three dry spreading experiments employing molten oxide mixtures (KATS-12 and ECOKATS-1) and prototypic UO 2 -ZrO 2 corium (VULCANO-VEU7). To interpret data with significant gas generation, the bubble suspension viscosity model learned from the rheology field was implemented. This enabled to represent the wide-ranging viscosity change by the Ramacciotti model based on the solid fraction coefficient = 6.3 common to all three experiments. In KATS-12, the predicted spread distance was not sensitive to the cell size in the range of 10 to 50 mm. In VULCANO-VEU7, the highly viscous corium spread a short distance on the ceramic and concrete sectors and showed abrupt stop particularly on the concrete sector. These observations were explained by a gas sparging enhancement of the melt upper surface heat transfer that is in consistent with the heat flux reconstructed based on the measured temperature histories. In ECOKATS-1, the weir anchoring model was proposed to simulate the anisotropic spreading on the 2D plane. Although the present prototype model requires a user to specify locations and timings of the weir formation and break, the successful interpretation of complex spreading behaviors encouraged the authors to improve its predictability by mechanically modeling accumulation of floating crusts at the melt fronts, the transformation into the weir that can obstruct the melt flow, and the break due to ablation caused by the melt convection heat transfer. [ABSTRACT FROM AUTHOR]