Developing All-Solid-State Batteries (ASSBs) presents both challenges and opportunities for battery manufacturers worldwide. Digital material design and numerical simulation have proven to be an invaluable asset in research and development of conventional lithium-ion batteries. [1] This approach creates a digital twin of a battery cell based on segmented 3D image data. Charging and discharging cycles, as well as other physical properties such as thermal and electrical conductivity, are predicted using numerical simulations. The geometric properties of the digital twin can be further varied to develop optimized structures.
In this work, we will show how established tools, like the simulation platform GeoDict, the digital material laboratory, and the BEST solver, are applicable to ASSBs. In addition, the simulation results will be compared with experimental data sets to establish a validated ASSB battery model. [2, 3]

This work is being carried out in the framework of DELFIN, a research project funded by the German Ministry of Education and Research (BMBF in German). In this project, a validated simulation model for ASSBs is being developed through the collaborative efforts of industry partners and leading academic institutions. The project covers the entire process from the acquisition of 3D image data using µ-CT, through image processing and segmentation, to the generation of a digital twin, and the creation of a validated battery model.
The first step is the acquisition of the microstructure of the battery electrodes by means of µ-CT scans. ASSBs present challenges for segmentation due to the lack of contrast between active materials and solid sulfuric electrolytes. We use multi-channel segmentation and AI algorithms to enhance contrast and resolution. The result is a detailed, digitized 3D microstructure. A structure generator based on a statistical parametric model is being developed to model the microstructure of ASSB cathodes and to conveniently generate digital twins of typical cathode structures in large numbers.
For accurate electrochemical simulations and analysis, GeoDict includes adjustments for the absence of a concentration gradient, as well as for the determination of Butler-Volmer rates based on experimental exchange rate constants. Data to validate the material parameters used in GeoDict simulations are provided by the collaboration with Dr Anja Bielefeld's group at JLU and will culminate in a validated ASSB model. With the validated model and ability to conveniently generate ASSB cathode structures, an optimization loop will be created to enhance ASSB research and development capabilities and design future ASSB materials.


References
[1] S., Möller et al.; Journal of Power Sources, 2024, 610, 234681.
[2] GeoDict simulation software Release 2024, by Math2Market GmbH, Germany, doi.org/10.30423/release.geodict2024
[3] Fraunhofer ITWM, Battery and Electrochemistry Simulation Tool (BEST), Available at: https://www.itwm.fraunhofer.de/de/abteilungen/sms/produkte-und-leistungen/best-battery-and-electrochemistry-simulation-tool.html, 2024, accessed:  13 September 2024.