Electrochemical, Aging, and Thermal Investigation of Lithium-ion Cells, Modules, and Packs

Lithium-ion cells are high energy and high-power density electrochemical storage devices that are used throughout the automotive industry. As the climate around the world continues to change, the implementation and use of lithium-ion batteries will continue to increase. Thus, to ensure the best performance is achieved, a thorough understanding of the internal processes of lithium-ion cells and the effects of the arrangement of the systems in which they are incorporated is required. To achieve a deeper understanding of the electrochemical processes within a cell, an electrochemical aging model of a prismatic 131 Ah NCA/Gr lithium-ion cell is developed using COMSOL Multiphysics V6.2 ®. The performance of the base electrochemical model is validated using experimental hybrid pulse power characterization tests at various temperatures. As for the aging model, the diffusion limited growth of the solid electrolyte interphase (SEI) layer is included and validated using experimental calendar aging data. To account for additional aging during the cycling of the cell, the accelerated formation of the SEI layer due to crack formations is incorporated. Additionally, the deposition of lithium metal is considered to account for additional aging at lower temperatures. Throughout the development of the model, it is found that modelling different numbers of electrochemical stacks, and whether a complete or partial stack is modelled, affects the distribution of the solid and liquid potentials throughout the cell. The electrochemical aging model is coupled with a simplified thermal model of a module used in an EV to understand the influence of the surrounding materials and the applied thermal management system. In addition to the model in COMSOL®, a thermal model of a pack used in an EV is developed in Siemens Simcenter Amesim. An equivalent circuit model with predefined parameters is used to capture the electrochemical and thermal behaviour of the cells. To capture the temperature gradient induced by the thermal management system, the thermal mass for the cells in the pack are discretized into a bottom and top fraction. In the modules the cells are grouped into sections to reduce the number of independent variables, while still capturing the temperature gradients about the length of the module. Comparing the performance of the two models, the Amesim model accurately predicts the temperature of a cell in a pack over a fast charge cycle, while the COMSOL® model underpredicts the temperature. It is found that the heat generation of the cells in the COMSOL® model are much lower than in Amesim, due to the lack of material specific experimental data for the development of the COMSOL® model. Each software package has different advantages and disadvantages for use throughout the automotive industry, with Amesim providing fast computations useful for quick development, while COMSOL® provides an opportunity to obtain a deeper understanding of the internal processes that can be used to improve a system's performance.