Novel regenerator design for caloric cycles using artificial neural network — Genetic algorithm method and additive manufacturing

Abstract

The regenerator is a key component to determine the performance of the active regenerative caloric cycle. Although the geometry of the regenerator is an important factor, very limited geometries have been used in prototypes due to the limitations of conventional manufacturing. Therefore, this study proposed a new type of regenerator with a high heat transfer rate for the caloric cycle, which is a packed rod bed, using additive manufacturing. In order to ensure a systematic approach for the new design of the regenerator, a new design optimization using an artificial neural network — genetic algorithm with the help of computational fluid dynamics was introduced. Artificial neural network models were used to predict the j and f factors of the packed rod bed and showed a mean relative error of less than 2.0% for the j factor, and a mean relative error of less than 7.5% for the f factor. The accurate results of artificial neural networks contribute to improving the optimization process. The regenerator optimized through the artificial neural network — genetic algorithm method increased the system efficiency by 4.7% and the cooling capacity by 13.0% compared to the baseline caloric cycle using a parallel plate matrix. Considering that caloric cycles are still in the development stage due to the lack of performance of the magnetocaloric cycle, this eventually may contribute to the commercialization and energy saving of the magnetic refrigeration cycle while the design optimization might also help to improve the performance of other caloric cycles.