Thermal and Electrochemical Modelling of Lithium-Ion Batteries Applied in Immersed Battery Management Systems.

Lithium-ion batteries are widely adopted in the automotive field for their high energy density, long cycle life and efficiency, but they tend to generate heat during operations, which can degrade their performance and lifespan. High operating temperatures specifically impose the need for effective thermal management strategies. While the working principles behind batteries are straightforward, their internal processes are highly complex, with several factors contributing to potential overheating. Accordingly, designers are constantly invested in developing tools capable of understanding, predicting and mitigating thermal risks. This thesis addresses these challenges by developing thermal models to accurately predict the thermal and electrochemical behaviour of lithium-ion batteries. The following paper considers various models of a 21700 lithium-ion battery using COMSOL Multiphysics, a tool capable of integrating a variety of interacting physical phenomena. A simplified 2D lumped semi-empirical model is employed to study the thermal and electrochemical characteristics of the battery. The procedure involves selecting a set of reduced inputs, tuning fitting model parameters with parameter estimation analyses, and validating the simulations against experimental data obtained from laboratory. The study focuses on simulating charging and discharging cycles, though other load cycles can be performed. In terms of results, the electrical model, corresponding to the cell potential, correlates well with experimental results, whereas the thermal model, characterized by the cell core temperature, shows some deviations. This model provides a simplified yet accurate representation of the battery dynamics, resulting in better understanding and prediction of battery performance under different conditions. Furthermore, the research extends to a 3D model of a battery pack consisting of nine cells in a 3s3p configuration, providing insights into the temperature distribution within the pack. Overall, the findings of this paper could serve as a benchmark for future enhancements for the optimization of immersed battery management systems.