Monday, 10 October 2022: 14:00
Room 315 (The Hilton Atlanta)
J. S. Lopata (University of South Carolina), T. R. Garrick, Y. Zeng (General Motors Holdings LLC), and S. Shimpalee (University of South Carolina)
Conventional porous electrode battery modeling for a variety of applications makes assumptions to reduce the complexity of the simulation where tradeoffs between computational complexity and electrode detail can be realized without significantly impacting the accuracy of the results. The most common porous electrode battery model is the pseudo-two-dimensional (P2D) model as popularized by Newman and colleagues. These porous electrode battery models typically focus on simulating gradients in two dimensions; the primary dimension through the thickness of the battery cell electrodes and separators, and the secondary dimension in the radial direction of spherical active material particles. The particles are distributed throughout what is assumed to be a homogeneous porous electrode domain. Typically, uniformity in particle size, porosity, and tortuosity is assumed.
In this work, the development of a three-dimensional microstructure (3DMS) lithium-ion battery model will be presented. With the increase in computational resources available in academia and industry, simulation of the microstructure in 3D is possible. Active material domains of both the anode and cathode can be generated stochastically to create microstructures representing the electrodes, with voids surrounding the particle representing the electrolyte. A comparison between the P2D and 3DMS models with sensitivity analysis focused on items that can be probed with microstructure simulations will be discussed and similarities and/or differences in charge/discharge characteristics, overpotentials, and concentration profiles will be presented. The findings from this work will provide a path to simulate at the microstructure level, drive electrode design, and translate the effects of heterogeneity into a homogenous porous electrode model.