Jinghua Zhao, Zhouzun Xie, Yansong Shen
The solid–liquid phase change in multiphase flow is widely encountered in various engineering applications, such as metal melting. However, the understanding remains limited due to the complexity of the process and the lack of reliable numerical models that can explicitly account for solid–liquid interactions. This study develops a new fully coupled computational fluid dynamics–discrete element method–volume of fluid (CFD–DEM–VOF) model. Particularly, for the first time, this model can simulate the phase change process based on the lever rule in the phase diagram, while also explicitly capturing both the heat transfer and motion of solid particles during the transition. The model is validated in two cases: 1) single droplet spreading over a spherical particle, and 2) liquid flow dripping through a packed bed. Then, the model is applied to simulate the melting of iron ore in a coke bed with heated gas injection, with the molten metal fraction transiently determined. Further, the effects of carbon content variations, gas inlet velocity, and heat transfer coefficient of gas are comprehensively studied. The results demonstrate that the lever rule significantly improves thermodynamic accuracy compared to the Gaussian error function or linear approximations previously used as a compromise. As a result, this method effectively captures key phenomena, for the first time, such as molten metal softening, melting processes involving solid-to-liquid phase transitions, and temperature distributions within the solid bed. Additionally, for dripping behaviour and melting degree, the higher inlet velocity slows molten metal trickling, while an increased heat transfer coefficient leads to a higher melting degree in the lower coke bed from 27.74% to 35.11%, indicating the critical role of conductive heat transfer. This work provides a generic framework for understanding multiphase flow processes involving phase change and improving industrial process optimisation and design.