In recent years, molten salt reactors (MSRs) have gained new momentum thanks to their potential for innovation in the nuclear industry, and several studies on their compliance with all the expected safety features are currently underway. In terms of passive safety, a strategy currently envisaged in accidental scenarios is to drain by gravity the molten salt, which acts both as fuel and coolant, in an emergency draining tank, ensuring both a subcritical geometry and proper cooling. To activate the draining system, a freeze plug, made of the same salt used in the core, is expected to open when the temperature in the core reaches high values. Up to this point, the freeze valve is still a key concept in the molten salt fast reactor (MSFR), and special attention must be paid to its analysis, given the requirement for passive safety, especially focusing on melting and solidification phenomena related to the molten salt mixture.

This work aims to contribute to the macroscale modeling of melting and solidification phenomena relevant to the analysis of the freeze valve behavior. In particular, the focus is on the identification of the numerical models that can be adopted to achieve the quantitative insights needed for the design of the freeze valve. Among the ones available in the literature, the most appropriate models were selected based on a compromise between accuracy and computational efficiency. A critical look at the models allows for a synthetic and consistent formulation of the numerical models and their implementation in the open-source software OpenFOAM. The code was subsequently verified using analytical and numerical solutions already well established in the literature.

A good agreement between the results produced by the developed solver and the reference solutions was obtained. In the end, the code was applied to simple case studies related to the freeze valve system, focusing on recognizing whether the developed code can model physical phenomena that can occur in a freeze valve. The results of the simulations are encouraging and show that the code can be used to model single-region melting or solidification problems. As such, this work constitutes a starting point for further development of the code, intending to achieve better quantitative predictions for the design of a freeze valve.