Modeling the Impact of Microstructure Variation on Charging Capability in Lithium-Ion Batteries

Abstract

Battery models can be used to predict conditions ripe for lithium metal plating in the anode during rapid charging. One common procedure is to simulate a rapid charge using a pseudo-two-dimensional (P2D) model and determine if and when the minimum local potential reaches 0.0 V with respect to metallic lithium in the anode. This normally occurs at higher states of charge during a DC fast charging or rapid regenerative charging event. However, we have recently determined that three-dimensional (3D) models, while broadly agreeing with P2D results, demonstrate much variation in the in-plane and through plane direction, which P2D models cannot resolve. Because P2D models employ effective medium approximations, any “local” value in the P2D space along the through-plane coordinate represents an average across a slice of the domain. Therefore, it may be insufficient to use minimum anode potentials from P2D models as indicators of local plating and may only be representative of global lithium plating onset.The results of rapid-charge 3D microstructure simulations are compared to P2D results. This allows us to verify whether rapid charging induces negative local potentials at the anode active material surface that are left undetected in a P2D model. The expected result implies that the 3D microstructure model is necessary to predict the onset of lithium plating. We then extend our previous study to look at the impact of design changes on the microstructure performance, and then link the microstructure performance changes to performance metrics at the battery cell, module, pack, and vehicle level. Finally, the impact of variation in the microstructure domain considering particle size variation is quantified to understand tolerances that may be necessary when developing charging schemes.