Prediction of battery thermal behaviour in the presence of a constructal theory-based heat pipe (CBHP): A multiphysics model and pattern-based machine learning approach

Abstract

This study investigates the thermal conductivity of a constructal theory-based heat pipe and presents the predction of a lithium-ion battery's thermal behaviour during charge and discharge by combining a special form of machine learning with a multiphysics numerical simulation. A series of multiple physical processes such as boiling, evaporation, and condensation were assumed to find the variable thermal conductivity of heat pipes. We used a combination of physics-informed machine learning and visual tracking method (pattern-based) to find the pattern of each feature, including temperature, for the first time. The findings reveal that a heat pipe design based on constructal theory can reduce the average and maximum temperatures of the battery by up to 13.43% and 27%, respectively, during the charge/discharge cycle. An approach based on constructal theory to the geometry of the heat pipe could reduce length (by up to 12%) without compromising efficiency. Additionally, by employing pattern-based machine learning (PBML), training time and transfer data were reduced significantly. Also, thermal conductivity could be predicted for heat pipes during charge/discharge cycles. The results of this study provide insight into adaptable thermal management systems for developing a new generation of compact battery packs