Predicting supercooling of phase change materials in arbitrarily varying conditions

Phase change materials are promising for thermal energy storage; however, one major bottleneck for their practical implementation is their unclear supercooling behaviors. In this work, we introduce a framework to predict the degree of supercooling for a phase change material subject to arbitrary geometrical and thermal conditions by analyzing the phase change material’s intrinsic nucleation characteristics with a statistical model. The prediction capability of our framework is successfully validated with experiments using magnesium chloride hexahydrate as a phase change material. For a system with a uniform temperature distribution, our framework can predict the average degree of supercooling. For a general case such as phase change materials embedded in a heat sink, the framework can accurately predict the expected time, with less than 8% deviation, for nucleation under given conditions. This work provides important insights in understanding and predicting supercooling behavior, thereby providing guidelines for the optimal design of phase change material-based thermal energy storage applications.