Quantitative Simulation of Battery Aging on a Microstructural Level
The quantitative simulation of battery aging on a microstructural level
The quantitative simulation of battery aging on a microstructural level is jointly pursued by Math2Market and the MaDE group of Prof. Vanessa Wood (ETH Zürich) in the framework of the EU-project “SOLVED!”.
The experimental approach is to analyze the microstructure of NMC cathode and graphite anode through 3D in-operando x-ray tomography and electrochemical characterization.
The digital approach is to use GeoDict to simulate the local stresses and expansion of the electrode microstructures due to the local lithium-ion concentration in the active material during charging and discharging of the battery. The simulation monitors and predicts the influence of microstructural changes on the battery performance.
We proceeded in the following steps:
- Simulating battery cycling by combining battery charging simulations and mechanical expansion simulations in GeoDict
- Analyzing the dynamic behavior of the microstructure of the battery during cycling through in-operando x-ray tomography by the MaDe group of ETH Zurich
- Validating simulation results by comparison with the experimental data
- Analyzing the influence of material and structural modifications on strain-induced battery aging
What was the result?
This unique digital workflow represents an efficient, state-of-the-art methodology for digital R&D material design of energy materials for e-mobility and energy storage.
What does this mean for our customers?
Reliable quantitative simulations are the way to a time-saving and streamlined approach to develop new battery prototype materials with superior lifetime and performance.
Research & Development of Battery Materials with GeoDict
Battery degradation is in part caused by changes in the microstructure of the electrode during battery cycling.
Lithium that enters an electrode, during charging or operating a lithium-ion battery, leads to expansion of the active material. The local lithium concentration determines the extent of this expansion that causes mechanical stresses on the micro- and macro-scale.
These stresses during cycling of the battery lead to degradation of the material due to cracks, delamination, and deformation of the inactive layers and the casing . The altered microstructure is said to ‘age’ and it suffers from capacity loss and damaging effects, like lithium-plating .
Simulation of dynamic processes during cycling to detect structural changes.
We use GeoDict to simulate the local stresses and expansion of the electrode microstructures during charging or discharging of the battery. Experimental data is used to validate these degradation simulations. In these simulations, local deformations and damage due to micro-scale lithium intercalation are linked to the local lithium-ion concentration in the active material and its local volumetric changes. The 3D simulations of the electro-chemical battery charging and the mechanical deformation are coupled to simulate the electrode expansion.
The altered microstructure is considered dynamically for each charging and discharging step. Digitally, we predict and monitor the influence of structural changes on the electronic and ionic transport processes and on the macroscopic performance of the cell.
In-operando measurements of electrodes for validation.
We analyze the dynamic behavior of the microstructure of the battery during cycling through in-operando x-ray tomography .
In the SOLVED! project the dynamic behavior of the microstructures of a lithium ion battery during cycling are studied with coupled electrochemical and mechanical simulations. Verification of the simulation is done through comparison with in-operando x-ray tomography measurements of an LiNiMnCoO2 cathode and a graphite anode. Electrochemical characterization of the same materials are used to parametrize the simulation.
The datasets and accompanying 3D renderings are provided by the MaDE group of ETH Zurich.
Shown are the validation results for the half cell simulation of a graphite anode in GeoDict.
To speed-up the computation, the LIR solver in GeoDict is adapted for battery charging simulations.
Cycling protocols are defined for the digital battery, to control the charging and discharging process in the automated coupling of electro-chemical charging and mechanical deformation simulations.
The well established GeoDict workflow of importing, processing, and analyzing 3D image data of battery materials is expanded by using powerful structure generators to create realistic statistical Digital Twins of the battery materials and a digital battery. Additionally, the damage due to mechanical deformation is simulated.
Finally, with automated evaluation and data post-processing, the changes in capacity and performance during battery cycling, can be simulated for whole electrodes.
The simulated expansion of a cut-out of the graphite anode is compared with the corresponding in-operando measurement to validate the simulation. The total expansion of the particles is 10% (measurement) and 9.8% (simulation) with free boundary conditions for the simulation.
A comparison of a slice of the structure for 60% state-of-charge is shown in Fig. 8. Voxels with the same material in the measurement and the simulation are shown in white for electrolyte and in gray for graphite. Voxels that are labeled differently in the simulation and the measurement are shown in blue and red respectively.
These differences are caused mainly by the fact, that the simulation results shown are for isotropic expansion. Due to the orientation of the graphite flakes, the real expansion is anisotropic, observed in the measurement. This anisotropic expansion of graphite will be simulated as a next step in the project.
 Quantification and modeling of mechanical degradation in lithium-ion batteries based on nanoscale imaging. S. Müller, P. Pietsch, B.-E. Brandt, P. Baade, V. De Andrade, F. De Carlo, V. Wood, 2018, Nature Communications, DOI: 10.1038/s41467-018-04477-1
 Investigation on the temperature dependence of lithium plating in commercial Li-ion batteries. M. A. Cabañero et al., Electrochimica Acta 171, 1217–1228 (2017).
 X-Ray Tomography for Lithium-ion Battery Research: A Practical Guide. P. Pietsch and V. Wood, Annu. Rev. Mater. Res., 2017, 47, 451–479
We thank our partners at ETH Zurich, MaDe Group, Vanessa Wood and her team, for their excellent collaboration and for providing the experimental data.