Comparative study on thermal performance of two graphite fin thermal energy storages based on experiment, simulation, and artificial neural network

The poor thermal conductivity of phase change material prolongs the charging time of the latent heat storage (LHS) from residue renewable energy. To overcome this challenge, the finned multi-tube configuration is applied to expand the heat transfer area. Moreover, natural graphite sheet is an innovative material for fins owing to its high thermal physical properties, ultra-lightweight, and non-corrosivity. In light of this, a comparative investigation of the thermal performance of two graphite fin shell-and-tube LHSs with different configurations, a circular fin in a cylindrical shell (CF-CS) and a square fin in a rectangular shell (SF-RS), was conducted using experimental and numerical methods. The melting time of the CF-CS LHS was 14% shorter and power was 23% higher than that of the SF-RS LHS. An ANN was employed to predict the natural convection heat transfer characteristics of the two LHSs with a high level of accuracy. A new correlation was established based on the predicted values to calculate Nusselt number in the melting phase with a low disparity of approximately 10.67% for the SF-RS LHS and 13.58% for the CF-CS LHS. This study provides essential guidelines for designing and optimizing an efficient thermal storage unit.