High-Throughput Process Space Exploration Using Mesoscale Model of Microstructure Evolution During Battery Material Drying Process

Electrode drying is one of the most time and energy consuming processes in Li-ion battery cell manufacturing. As an electric vehicle OEM and Ultium battery cell manufacturer, General Motors seeks to enhance the understanding of the drying mechanisms towards producing high quality battery electrodes with reduced cost and energy usage. In this presentation, we will present an integrated modeling framework to build computational databases for exploring material process space. Coarse-Grained Molecular Dynamics (CGMD) is employed to describe the sedimentation of solid particles and solvent evaporation in the slurry including active materials (AM) particles, conductive carbon solubilized binder and solvent; Smoothed particle hydrodynamics (SPH) is used to describe the multiphase fluid dynamics in porous structures formed by active particles; and Phase-field approach is utilized to describe the species diffusion, convection and pore evolution. With the integrated model, computational database about the effect of initial and operation conditions on the evolution of temperature, drying kinetics, binder distribution and pore structure are built with high-throughput simulations. Then data-driven time-dependent deep learning is applied for the exploration of process space. The results demonstrate the capability of modeling framework for improving the understanding of drying mechanisms and optimizing the drying process parameters to achieve desired microstructures and minimize energy assumption.