Digital Twin of Battery Electrode Manufacturing: Discrete Element Simulation of the Calendering Process

The development of so-called digital twins can harness the optimization of components for battery manufacturing, reaching important time-to-market reduction. These digital twins are virtual replicas of the manufacturing process, which help the analysis and definition of specific parameters without relying on costly and long experimental testing campaigns.

One crucial stage in battery production is the manufacturing of electrodes, which include different steps such as slurry preparation, coating, drying, calendering, cutting, or stacking. Among these, calendering is of utmost importance as it will provide the electrode, and eventually the cell, with the specific electrochemical properties and performance. 

Calendering is a process in which the coated electrode, resulting from the drying process, passes through a pressing machine (typically a roller). Until this moment, the electrode mesostructure consists of a blend of active particles and additives. The compression promotes particle rearrangement resulting in a more compact electrode. Therefore, the thickness of the electrode decreases while the density increases. This results in an increase of the electrical conductivity and storage capacity. Also, the compression enhances the adhesion of the active material to the current collector. However, this process is critical and the electrode may be harmed or become useless if the compaction parameters (e.g., pressure, gap between rollers) are not correctly chosen. 

Within this context, the development of simulation tools become relevant, as they can provide the design team with tools to analyze the effect of the compaction parameters for given material mesostructures. Moreover, they can be used to optimize the composition of the electrodes with targeted properties such as the tortuosity, electrical conductivity or storage capacity. Due to the discrete nature of the material mesostructure, particle methods (e.g., lattice model, lattice-particle model, discrete element method) are suitable for the simulation of the calendering process. Unlike continuum-based methods such as the finite element method, these methods account for the individual behavior of the particles and, most important, their interactions (e.g., friction, cohesion, collision). This can help gain insight into the local aspects to better understand particle redistribution and compaction under pressure. Moreover, these models allow to represent the heterogeneities of the mesostructure such as the particle size distribution, shapes or orientations. Finally, they can also provide a final mesostructure that can be used in the simulations of further manufacturing steps, regardless of the type of models to be used (i.e., continuum- or discrete-based).

The main goal of the Battery Cell Assembly Twin (BatCAT) project is the development of digital twins of the manufacturing processes to comprehensively optimize battery components by integrating data-driven and physics-based methods. Particle models are a lever to gain insight into the link between the electrochemical performance of the electrodes and the manufacturing process. Therefore, these can be used to narrow manufacturing parameters like roller pressure or speed without the need for large experimental campaigns, improving the efficiency of the component and the manufacturing process, while increasing the quality of the final product.

Francisco Montero Chacón,
Universidad Loyola Andalucía
October 2024