AUTHOR=Komori Chiyuri , Ishikawa Shota , Nunoshita Keita , So Magnus , Kimura Naoki , Inoue Gen , Tsuge Yoshifumi TITLE=Stress Prediction of the Particle Structure of All-Solid-State Batteries by Numerical Simulation and Machine Learning JOURNAL=Frontiers in Chemical Engineering VOLUME=4 YEAR=2022 URL=https://www.frontiersin.org/journals/chemical-engineering/articles/10.3389/fceng.2022.836282 DOI=10.3389/fceng.2022.836282 ISSN=2673-2718 ABSTRACT=

All-Solid-state batteries (ASSBs) are non-flammable and safe and have high capacities. Thus, ASSBs are expected to be commercialized soon for use in electric vehicles. However, because the electrode active material (AM) and solid electrolyte (SE) of ASSBs are both solid particles, the contact between the particles strongly affects the battery characteristics, yet the correlation between the electrode structure and the stress at the contact surface between the solids remains unknown. Therefore, we used the results of numerical simulations as a dataset to build a machine learning model to predict the battery performance of ASSBs. Specifically, the discrete element method (DEM) was used for the numerical simulations. In these simulations, AM and SE particles were used to fill a model of the electrode, and force was applied from one direction. Thus, the stress between the particles was calculated with respect to time. Using the simulations, we obtained a sufficient data set to build a machine learning model to predict the distribution of interparticle stress, which is difficult to measure experimentally. Promisingly, the stress distribution predicted by the constructed machine learning model showed good agreement with the stress distribution calculated by DEM.