A universal multi-factor coupled model for close-contact melting

Latent heat storage (LHS) dominated by gravity-driven close-contact melting (G-CCM) is limited by the late‑stage drop in driving force, resulting in low utilization efficiency in medium-to-low temperature waste heat recovery. To overcome this bottleneck, this study proposes a novel LHS system architecture based on efficient pressure-driven CCM (P-CCM). Using an experimentally validated P-CCM numerical model, this work investigates the effect mechanisms of applied pressure, geometric dimension, and wall temperature on the hydrodynamic and heat transfer characteristics within the molten liquid film and the thermal storage performance. The universal calculation models for the liquid fraction ( β ) are derived using the nonlinear least-squares method. The results reveal that applying pressure enhances both conductive and convective heat transfer, whereas increasing the width primarily lengthens the drainage path with minimal effect on average velocity and pressure drop. Increasing the wall temperature accelerates the melting rate by increasing the temperature gradient. The complete melting time ( t c ) is positively correlated with the modified Archimedes number ( Ar *) and the Stephen number ( Ste ), and negatively correlated with the aspect ratio ( AR ). The specific power ( P m ) reaches its maximum at 100 Pa, first increasing and then decreasing as Ar * increases. Computational models for G-CCM ( β = 21.6293( Fo·Ste 0.75 ·AR −0.65 ) 0.7653 , R ² = 0.9983) and P-CCM ( β = 1.5677( Fo·Ste 0.7 ·Ar * 0.5 ·AR −0.6 ) 0.8907 , R ² = 0.9992) show excellent agreement with the β curves of the 80 cases. These findings provide practical guidance for elucidating the performance enhancement mechanisms of P-CCM and predicting β curves in multidimensional parameter spaces.