Coupled multi-physics simulation and experimental study of dynamic seawater freezing under volumetric driving forces

Off-grid islands face energy and freshwater shortages, and integrating seawater desalination with ice thermal storage can efficiently utilize seawater resources, optimize energy storage, and improve energy efficiency. However, current seawater freezing models often neglect the coupled effects of natural convection on the phase, concentration, and temperature fields, leading to discrepancies in ice crystal morphology and growth rates. This study integrates the phase field model with computational fluid dynamics (CFD), accounting for the impact of temperature and concentration gradients on the flow field. An improved high-precision numerical model for seawater freezing is developed and validated using dynamic freezing experiments on a micro-scale platform under varying salinities and supercooling. Results show that the improved model accurately predicts seawater dendrite morphology under different conditions, with simulation results closely matching experimental observations. Predictions of the effects of supercooling and initial salinity on water volume fraction align with experimental data, with deviations of 10 % for supercooling and 25 % for initial salinities of 3.5 % and 5.0 %. Compared to existing models, the improved model achieves a 70 % increase in accuracy and captures both longitudinal growth and transverse branching of dendrites, providing a more realistic representation of dendrite growth. The study also reveals that higher initial salinity increases the number of dendrite branches, narrows their widths, and decreases their growth height. Ice growth rates increase by 10–14 % for every 1 °C decrease in initial temperature, while each 1 % increase in salinity reduces ice growth by about 6 %.