Experimental and machine learning-based investigation of additively manufactured PCM encapsulation geometries for enhanced thermal and electrical performance in battery thermal management system

Abstract

In this study, the thermal and electrical performances of phase change material (PCM) based battery thermal management systems (BTMS) with different macro-encapsulation geometries (hexagonal-Hx, square-Sq, circle-Cr) and produced by additive manufacturing method were experimentally evaluated. The developed BTMS structures provided a safe and stable structure by preventing PCM leakage and direct contact with the battery. Hexagonal (Hx) geometry showed superior thermal performance compared to other geometries by providing the lowest battery temperatures at all C rates. Hx BTMS limited the increase in internal resistance by increasing the heat transfer from the battery to the PCM, thus maintaining voltage stability and increasing the energy density by up to 6.90 %. At high discharge rates, the latent heat storage feature of PCM was activated only in the Hx structure, and this made active heat management possible. It was shown in the analyses performed with an artificial neural network (ANN) that the experimental data could be predicted with high accuracy, and it was determined by SHAP analysis that Hx geometry had the highest positive effect on the voltage. These results demonstrate the critical importance of macro-encapsulation geometry in BTMS design and demonstrate that Hx geometry offers a superior solution in terms of thermal safety, voltage stability, and long cycle life.