Multiscale modeling and calibration framework for predicting the mechanical response of Li-ion battery cell components

The mechanical response of battery cells plays a vital role in design of electric vehicles e.g., when evaluating their crashworthiness or studying the effects of cell swelling during operation. In this paper, we present a multiscale modeling framework for predicting the mechanical response of battery cell components at different length scales. Two design optimization loops for calibrating material model parameters are established. First, the effective mechanical response of the binder-conductive additive-electrolyte material phase inside the electrodes is estimated by performing homogenization of microscale Representative Volume Elements (RVEs), while utilizing experimental data for the effective electrode layer and data for the electrode particles from literature. Secondly, the effective response of the jellyroll is estimated by creating an RVE of the electrodes-separator stack and perform homogenization, while utilizing experimental data for the individual layers. Close agreement between numerical and experimental results are found for a jellyroll sample and a complete prismatic battery cell exposed to mechanical impact loading. By separating the length scales, and utilizing homogenization and calibration schemes, it is possible to estimate the mechanical response of the various material phases at the different length scales. This enables improved understanding of how battery cells behave under mechanical loads. • Framework to estimate mechanical properties across multiple length scales. • Effective stiffness of binder-electrolyte and jellyroll modeled via homogenization. • Mechanical impact tests performed on jellyroll sample and complete battery cell. • Numerical predictions compared with experimental results.