Messages

Hello everyone,

A colleague and I are currently simulating a solid-state battery cathode consisting of NCM + LPSCl in BatteryDict (GeoDict 2024). Unfortunately, the charging simulation fails immediately with the following error:
Failed to solve the InitialPotential-system!
BESTmeso returned with error code 1
The battery report additionally states:
Charge rate: 0.1 C
Applied current: 0 A
Transferred charge: 0 Ah
Maximum cell capacity: 0.00078 Ah
However, the effective transport properties of the microstructure seem to be available:
-Electronic conductivity of the cathode is non-zero.
-Ionic conductivity through the LPSCl phase is also non-zero.
Therefore, I would have expected the simulation to start.
Some additional information:
-Cathode: NCM + LPSCl composite
-Anode: Lithium
-The structure was successfully imported and BatteryDict identifies a finite cell capacity.

Has anyone encountered this error before?

This is what the error file says:
15:58:27.787: [BEST:er] Starting discretization...
15:58:27.840: [BEST:er] Function Norm ||F(x)||_2: 0.0264661
15:58:27.841: [BEST:er] Convergence criterion NewtonIncrementMaxComponentChange: error: 0.016342 relativeTolerance: 1e-05
15:58:27.841: [BEST:er] NewtonSolver Error: Maximum number of iterations reached!
15:58:27.841: [BEST:er] Failed to solve the InitialPotential-system!
15:58:27.841: [BEST:er] Writing output-file of the last successful timestep...
15:58:27.842: [Error] [BEST:er : Error] BesterStoppingException occurred in main() while initializing or solving: Failed to solve the InitialPotential-system!


Any suggestions would be greatly appreciated.
Thank you very much!
Finn and Caroline

I understand. Thank you very much for the quick replies.

Best regards,
Caroline

Hello Jürgen,

thank you very much, for this quick response. I understand what you mean, and this is exactly what I want. The only thing is, that I don't have this option to check "Has Tortuosity", because I don't have the tab "Porosity/Tortuosity" when I want to compute conductivity. Do I need to active something before? I only have the tabs "Electrical conductivity" and "Specific Electrical Contact Resistivity" when I open the Edit-window for setting up the simulation, even when I use porous materials. I have GeoDict 2024, is that too old for this feature? Or do I have to change something in the material database? I already set the material to porous and gave it a porosity of 50 % and a certain tortuosity. Everything looks exactly as on the website you sent me, but I am missing this one tab.

Best,
Caroline

[solid]

Hello everyone,

I am running simulations in ConductoDict using a solid material for the active material of my battery, and porous material for the carbon-binder-domain. How does the porosity of this porous material influence the conductivity results? Because so far, I get the same overall conductivity results no matter if I use a solid or porous CBD, when I manually insert the same intrinstic conductivity value. I thought I would get lower conductivities if my CBD is porous (e.g. 50 percent), as my overall structure thus contains more parts with low-conductive air. In other words, I though that the "solid" part of my porous CBD would adopt the conductivity that I manually enter and the air inside the porous material would adopt the same intrinsic conductivity as the air or electrlyte surrounding the CBD and active material. It seems like I am wrong. What difference does it make for conductivity simulations to have a porous material or a solid material, if the results are always the same?

Best,
Caroline

Dear Math2Market,

I heard that it is possible to rescale the simulation domain and the voxel size not only by integer values, but also by decimal numbers. Could you tell me how this can be done?
I would really like to run simulations faster this way, without losing too much accuracy.

All the best,
Caroline

Dear Janine,

Thank you very much for this detailed reply! I received the email and will try some simulations with it.
Reducing the data points also makes a lot of sense. Thank you! I look forward to the results.

Best,
Caroline

Hello Janine,
Yes, I run a homogenized simulation. I only have GeoDit 2024. Is there also any way to speed that up?

Thank you for the link to the video! I will try that. Does GeoDict contain ready-to-use text files with concentration-dependend values, of does the user always write their own text files? I am simulating NMC622 as active material for my cathode.

Best,
Caroline

Hello everyone,
I am running charge simulations for my electrode structures of 500*500*544 voxels à 100 nm. When using constant parameters for the electrolyte, the simulations take about 3-5 hours. When using concentration dependend lithium ion diffusivity and ionic conduction, the simulations repeatedly take 95 hours. Is this difference in time normal? I know simulation time depends on the computer hardware, but maybe you can tell me, if a factor 20-30 increase is normal.

On that note: is it also possible to use concentration depended values for lithium ion diffusivity for the active material? Just out of curiosity.
Thank you!