Search for: All records

The accuracy of single particle (SP) models for lithium-ion batteries at high C-rates is constrained by lithium concentration gradients in the electrolyte, which affect ionic conductivity, overpotential, and reaction rates. This study addresses these limitations using extreme gradient boosting machine learning (ML). By training our ML model with data from a comprehensive electrochemical (P2D) model and performing sensitivity analysis on key battery parameters, we enhance predictive accuracy. Compared to conventional SP and P2D models under constant current loading, our ML-based SP model achieves similar predictive accuracy to P2D, with significant improvements in computational efficiency. Additionally, the ML-based SP model demonstrates improved predictive accuracy under dynamic loading conditions, providing a practical framework for improving battery management and safety. 

Abstract Thickening electrodes is one effective approach to increase active material content for higher energy and low‐cost lithium‐ion batteries, but limits in charge transport and huge mechanical stress generation result in poor performance and eventual cell failure. This paper reports a new electrode fabrication process, referred to as µ‐casting, enabling ultrathick electrodes that address the trade‐off between specific capacity and areal/volumetric capacity. The proposed µ‐casting is based on a patterned blade, enabling facile fabrication of 3D electrode structures. The study reveals the governing properties of µ‐casted ultrathick electrodes and how this simultaneously improves battery energy/power performance. The process facilitates a short diffusion path structure that minimizes intercalation‐induced stress, improving energy density and cell stability. This work also investigates the issues with structural integrity, porosity, and paste rheology, and also analyzes mechanical properties due to external force. The µ‐casting enables an ultrathick electrode (≈280 µm) that more effectively utilizes NMC‐811 (LiNi_0.8Mn_0.1Co_0.1O₂) cathode and mesocarbon microbeads anode active materials compared to conventional thick electrodes, allowing high‐mass loading (35.7 mg cm⁻²), 40% higher specific capacity, and 30% higher areal capacity after 200 cycles, high C‐rate performance, and longer cycle life.