A reduced order model for a lithium-ion 3D pouch battery for coupled thermal-electrochemical analysis

Abstract

Newman's Pseudo 2-D or porous electrode theory model, is commonly used in the coupled thermal-electrochemical analysis for lithium-ion batteries. This model consists of a 1D macroscale model that satisfies the conservation of charge and mass in the solid and liquid phases in the superimposed continua considering the reaction rate at the interface between solid and liquid, and a 1D microscale diffusion model. The macroscale model can be expanded to include 3D topology to help describe more complex lithium-ion battery behavior, which is often referred to as Pseudo 4-D. However, given the geometrical complexities of cell construction, a Pseudo 4-D model is computationally expensive when compared to a Pseudo 2-D model. When simulations require multiple cells while maintaining detailed thermo-electrochemistry computations (as in module-level simulations), a model order reduction method is beneficial to efficiently perform coupled thermal-electrochemical analysis on a Pseudo 4-D model.This work presents a new methodology for constructing a Pseudo 4-D reduced order model using common pouch cell geometries as an example. Governing equations for the Pseudo 2-D models are adopted in the Pseudo 4-D framework. The proposed workflow reduces the model size by limiting the number of component layers and keeping a high fidelity to the detailed model by properly scaling relevant material properties. First, coupled thermal-electrochemical analysis is carried on the detailed and reduced order pouch cell models, where the efficiency and accuracy of the reduced order model is evaluated. Second, a series of the reduced order models are assembled to form a battery module model, where the same analysis is carried out and the corresponding results are analyzed. The proposed reduced order model approach can be used to significantly reduce computational cost for coupled thermal-electrochemical analysis while keeping a high fidelity, compared to a detailed Pseudo 4-D model.