"Core-shell" composite active material

[lefile]

Hello!

There are electrode designs where one active material is coated on the surface of a second active material particle. It is effectively a composite electrode, but not in a way implemented right now, where composite active materials are independent. This could be implemented as a one particle with radially-dependent solid diffusion properties and similar.

[TomTranter]

Spatially dependent diffusion would be simple enough but I'd worry that the continuum representation wouldn't be enough to model things like grain boundary and stresses. Any FE models out there or data looking into this type of thing?

[rtimms]

There is this core-shell model for positive electrode degradation. The authors implemented it in their own fork of PyBaMM (link in paper), so you could probably take a similar approach to handle a coated particle (thinking carefully about what happens at the interface).

[mzzhuo]

Hi Tom and Rob, since Pybamm has changed so much, updating my previous branch did not work well, so I just build the submodel upon the latest Pybamm release, now it is almost done and I wrote a script tutorial that can reproduce the results in my paper. I am now testing the codes (nox -s unit), but it will take some time and may not pass as I defined some domains, expression tree etc. At the same time, would the pybamm team like to review the source codes to speed up the whole process?

[rtimms]

Great! Yeah, if you like some feedback you can open a pull request and mark it as draft until it’s ready for a proper review.

[mzzhuo]

Hi Rob, I have created a PR, but it may not look proper like others, as it is my first time to contribute. At the moment, I think it would good if the pybamm team can seriously check/review the physics including equations, boundary conditions, etc in the submodel. Also, I kind of found two pybamm bugs (one leads to crash one to computation inaccuracy) which need attention. Here is a link to my handwritten formula of scaling equations and bcs for your reference.

[rtimms]

Thanks, will take a look. Can you open separate issues for the bugs please?

[mzzhuo]

Sure, I will raise the issues after the pe submodel is merged as it may be invoked for demonstration