Experimental Investigation of Nano-encapsulated Molten Salt for Medium-Temperature Thermal Storage Systems and Modeling of Neural Networks

Molten salts were chosen as a thermal storage medium because they were best suited for medium-temperature thermal energy storage applications. Their nano-sized capsules allow for a more efficient design of thermal energy storage systems. This research work utilized the emulsification sol–gel method to synthesize two different types of nano-encapsulated phase change material (NEPCM) salts (i.e., SiO 2 shell-based PCM and TiO 2 shell-based PCM). The chemical structure of NEPCM salts was investigated using X-ray diffraction and Fourier transformation infrared analysis. A scanning electron microscope and a particle distribution analyzer were used to examine nanocapsules' surface morphology and size distribution. The phase change characteristics and thermal stability of the NEPCM samples were determined using simultaneous differential scanning calorimetry and thermogravimetric analyzer equipment. The activation energy (AE) of the pure PCM and NEPCM samples were calculated by the Kissinger, Ozawa, and Starink models. The artificial neural network models were developed to predict the thermophysical properties of nano-encapsulated PCM samples at different heating rates. The experimental differential scanning calorimetry outcomes are taken to train the neural networks. The optimum neural architecture is obtained at a 3-36-1 structure. This optimum neural network effectively predicts nanocapsules' thermophysical properties with higher accuracy (i.e., R 2 = 0.99). Graphical Abstract