These slides were presented at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

The fabrication process for lithium-ion electrodes typically involves four main steps: mixing, coating, drying, and calendaring. Despite extensive research, the impact of each step and its associated parameters on the final electrode microstructure and performance has not been thoroughly investigated. This study presents a novel multi-scale multi-physics computational framework to predict process-to-property relationships in lithium-ion cathode manufacturing. This work is motivated by two critical challenges in battery production: 1) the difficulty of implementing lab-scale research findings in mass production without substantial modifications, and 2) the low yield of acceptable batteries, where achieving failure rates below 10% remains challenging. In this research, first, using discrete element method (DEM), a slurry containing active materials, a carbon-binder domain, and solvent was generated. Second, a macro scale computational fluid dynamics (CFD) model was constructed to replicate the coating machine and drying step. Third, DEM was coupled with the CFD and thermo-mechanical finite element models to simulate the evaporation phase and predict the temperature distribution and fluid flow regime inside the coater as well as across the cathode electrode, thickness of the electrode, and the overall shrinkage of the sample. Finally, the electrode was further compressed to mimic the calendaring step. By incorporating the damage criterial, the numerical simulation could also predict the onset of crack formation and delamination phenomena which are considered the primary failure mechanisms in cathode electrode manufacturing. To validate the numerical simulation results, experimental test was conducted to fabricate NMC 111 cathode electrode. Then, materials characterizations including scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), and micro-CT were performed at various sections to analyze the microstructure formation. The experimental findings were then compared to simulation results at corresponding spatiotemporal points within the computational domain. Good agreement was established between the results for the process parameters investigated. The numerical simulation results demonstrated how variations in manufacturing process parameters affect the microstructure, performance, and aging characteristics of lithium-ion cells. Additionally, the simulation results confirmed that by employing high-fidelity physics-based numerical simulation at different scales, the optimum process parameters, in particular, drying rate and temperature, can be identified for producing defect-free electrodes with tailored microstructures.