Computational and machine learning analysis on bio-inspired coronary fin designs for enhanced melting in PCM-based next-generation thermal energy storage system

The current analysis objectives to present a novel application of machine learning methodology based on Levenberg Marquardt Technique with Artificial Back Propagated Neural Networks (LMT-ABPNN) and numerical simulation to predict thermal energy storage (TES) efficiency of phase change materials by introducing an innovative horizontal shell and tube unit incorporating bio-mimetic heart coronary-shaped fins for reduce energy storage time. Numerical simulations were used to investigate the thermal behavior of phase change materials (PCMs) during the melting process, specifically focusing on the significance of fin design and orientation. The results show that the melting completion times for the systems with heart coronary fins are significantly reduced: 5360 s for an angle of 43°, 4720 s for an angle of 70°, and 4880 s for an angle of 90°. The optimized fin angles (43°, 70°, and 90°) demonstrated substantial time savings of 30.2 %, 38.54 %, and 36.45 %, respectively, with the 70° configuration achieving the highest performance. Adding fins significantly improves energy storage capacity, with the 70° fin angle (Case III) increasing capacity to 2989.6 kJ from 2960 kJ (Case II). The enhancement ratio (ER) also highlights the benefits of heart coronary fins, with ER peaks between 2000 s and 3500 s due to improved convection currents in Cases II, III, and IV. The novel heart coronary fins, inspired by biomimicry, significantly improve heat storage efficiency, demonstrating the potential of bio-inspired engineering solutions in advancing TES technologies. The outstanding validation of the artificial neural network result against the CFD numerical reference solution for liquid fraction across per unit time was observed.