TY - JOUR
T1 - Physics-based modelling and data-driven optimisation of a latent heat thermal energy storage system with corrugated fins
AU - Tavakoli, Ali
AU - Hashemi, Javad
AU - Najafian, Mahyar
AU - Ebrahimi, Amin
PY - 2023
Y1 - 2023
N2 - Solid-liquid phase transformation of a phase change material in a rectangular enclosure with corrugated fins is studied. Employing a physics-based model, the influence of fin length, thickness, and wave amplitude on the thermal and fluid flow fields is explored. Incorporating fins into thermal energy storage systems enhances the heat transfer surface area and thermal penetration depth, accelerating the melting process. Corrugated fins generate more flow perturbations than straight fins, improving the melting performance. Longer and thicker fins increase the melting rate, average temperature, and thermal energy storage capacity. However, the effect of fin thickness on the thermal characteristics seems insignificant. Larger fin wave amplitudes increase the heat transfer surface area but disrupt natural convection currents, slowing the melting front progress. A surrogate model based on an artificial neural network in conjunction with the particle swarm optimisation is developed to optimise the fin geometry. The optimised geometry demonstrates a 43% enhancement in thermal energy storage per unit mass compared to the case with planar fins. The data-driven model predicts the liquid fraction with less than 1% difference from the physics-based model. The proposed approach provides a comprehensive understanding of the system behaviour and facilitates the design of thermal energy storage systems.
AB - Solid-liquid phase transformation of a phase change material in a rectangular enclosure with corrugated fins is studied. Employing a physics-based model, the influence of fin length, thickness, and wave amplitude on the thermal and fluid flow fields is explored. Incorporating fins into thermal energy storage systems enhances the heat transfer surface area and thermal penetration depth, accelerating the melting process. Corrugated fins generate more flow perturbations than straight fins, improving the melting performance. Longer and thicker fins increase the melting rate, average temperature, and thermal energy storage capacity. However, the effect of fin thickness on the thermal characteristics seems insignificant. Larger fin wave amplitudes increase the heat transfer surface area but disrupt natural convection currents, slowing the melting front progress. A surrogate model based on an artificial neural network in conjunction with the particle swarm optimisation is developed to optimise the fin geometry. The optimised geometry demonstrates a 43% enhancement in thermal energy storage per unit mass compared to the case with planar fins. The data-driven model predicts the liquid fraction with less than 1% difference from the physics-based model. The proposed approach provides a comprehensive understanding of the system behaviour and facilitates the design of thermal energy storage systems.
KW - Deep neural networks
KW - Machine learning
KW - Optimisation
KW - Phase change material
KW - Thermal and fluid flow modelling
KW - Thermal energy storage system
UR - http://www.scopus.com/inward/record.url?scp=85168783785&partnerID=8YFLogxK
U2 - 10.1016/j.renene.2023.119200
DO - 10.1016/j.renene.2023.119200
M3 - Article
AN - SCOPUS:85168783785
SN - 0960-1481
VL - 217
JO - Renewable Energy
JF - Renewable Energy
M1 - 119200
ER -