A simplified 3D-resolved microstructure model for high-fidelity lithium-ion battery cell simulations

• Published in: Journal of power sources Vol. 613; p. 234817
• Main Authors: Alberghini, Matteo, Blanco, Giulia, Bertinetti, Andrea, Tommasi, Alessio, Sgroi, Mauro
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2024
• Subjects: 3D-resolved microstructure; Electrochemical modeling; Finite Elements Method; Lithium-ion batteries; Microscale modeling; Lithium-ion batteries; 3D-resolved microstructure; Finite Elements Method; Microscale modeling; Electrochemical modeling
• ISSN: 0378-7753; 1873-2755
• DOI: 10.1016/j.jpowsour.2024.234817

Reliable and representative 3D microstructures of lithium-ion cells are essential for identifying novel and cost-effective optimization strategies for next-generation battery technologies. Numerical simulations are a powerful tool to generate microstructures and evaluate their impact on the cell performance without relying on expensive imaging techniques. This study introduces a simplified porous cathode model using 3D-resolved geometric entities to represent each phase. Specifically, the impact of three levels of detail in modeling additive conductive materials, such as carbon black, is investigated: via (i) sub-micron additive particles, (ii) micron-sized aggregates, or (iii) homogenized within the catholyte, as commonly assumed. Electrode geometries are generated and analyzed via a dedicated script implemented in Octave. Equivalent cathodic properties and cell performance are evaluated via finite elements simulations to assess the impact of the three modeling methods considered. Results demonstrate the effectiveness of the proposed approach with respect to homogenized structures, successfully matching experimental benchmarks of electrode conductivity and tortuosity, and avoiding overestimation of cell performance. These findings pave the wave to ease the simulation of accurate cell microstructures, improving the reliability of 3D simulations and fostering more computationally-efficient models. [Display omitted] •A simple 3D electrode microstructure model for Li-ion cells is proposed.•Rapid and flexible generation of microstructures based on phases volume fractions.•Explicit representation of conductive additives is compared to homogenized phases.•Accurate representation of carbon percolation is achieved with minimal complexity.•Finite element validation against experiments, outperforming reference methods.