Optimization of a thermal energy storage system enhanced with fins using generative adversarial networks method

Abstract

Optimizing fin configurations is an effective method to improve the performance of a thermal energy storage system. The present research aims to determine the optimal fin spacing and inclination angle to minimize the melting process of a phase change material (PCM) in a rectangular enclosure, which is exposed to a constant temperate finned wall while other walls are adiabatic. Oleic acid (OA) is used as a PCM with a melting temperature of 287–288 K. Computational fluid dynamics (CFD) simulations are employed to comprehensively investigate the thermal behavior of the system by solving continuity, momentum, and energy equations simultaneously. Initially, fifteen different datasets are simulated using CFD methods to explore various fin configurations. After that, fifteen additional datasets are generated to verify the reliability of the generative adversarial networks (GANs) method, which supplements the CFD approach. An ensemble learning approach is implemented to create a hybrid dataset consisting of thirty datasets, and then the optimum dataset is determined. To validate this approach, a single numerical simulation is conducted at the optimum dataset, and the obtained optimum value, minimum melting time of PCM, is compared with the result obtained from CFD simulation. It is identified as the optimal fin configuration, resulting in a fin spacing of 15.75 mm and an inclination angle of −11.6°, highlights the efficacy of the combined CFD and GANs approach. The melting time captured from the numerical simulation and the one produced with GANs differed by about 4.7 %. The current study concludes that the GANs method might be effectively worthwhile in many energy storage applications.