This study investigates the spatiotemporal coupling between wall heat flux mechanisms, buoyancy-driven flow and melting interface dynamics during phase change material (PCM) melting in a confined cylindrical enclosure. A high fidelity computational fluid dynamics model based on the enthalpy-porosity approach is developed for a representative low-temperature PCM heated radially through a sidewall and cooled axially through a backing layer. The simulations cover Stefan numbers of approximately ~ 0.05 – 0.35 and Rayleigh numbers of ~ 10⁴ – 10⁸ over a full melting duration of about 5000 seconds. By synchronising the moving solid-liquid interface with local wall heat flux signals, four successive wall heat transfer regimes are identified: pure conduction, boundary-layer instability, developed natural convection with plume formation, and post-melting thermal equilibration. The transition from conduction to convection occurs when the near wall thermal boundary layer becomes unstable, whereas local completion of melting produces an exponential decay in wall heat flux. Temporal autocorrelation and spatial two-point correlation of the heat-flux signals reveal characteristic time and length scales associated with plume generation and convective organisation. Regime-wise energy analysis shows that the conduction- and instability-dominated stages contribute approximately 24% of the total heat transferred while occupying less than 20% of the melting duration. The results demonstrate that global melting is governed by strong two-way coupling among wall induced thermal gradients, interface morphology, and bulk convective circulation. The proposed non-dimensional framework provides physical insight into PCM melting dynamics and supports future modelling, correlation development, and design-oriented optimisation of latent heat thermal storage systems. • Four melting regimes are identified in cylindrical PCM melting. • Wall heat flux is strongly coupled with melting-front evolution. • Regime transitions are characterized using Stefan and Rayleigh numbers. • Correlation analysis reveals integral time and length scales of instabilities. • Early melting regimes contribute nearly 24% of total heat transfer.
Khaliliparam et al. (Wed,) studied this question.