Dynamically change geometrical parameters

[lefile]

Hi guys,

rather than using the sub-model from pybamm, I have a custom mechanical model coupled to a pybamm simulation. I run pybamm sim step-by-step, updating electrode porosity as particles breathe.

I would also like to update particle radius as well, as it affects solid-state diffusion. Ideally, I'd like to update every particle in the DFN electrode, but a single value would also be fine. However, pybamm does not allow for geometrical parameters to be set as inputs.

The additional question would be if it is needed to update electrode thickness as well? Obviously, this would affect electrode capacity (which ofc is not being changed), so in that case active material volume fraction should be decreased.

Can I maybe bypass the need to change particle radius and electrode thickness parameters by respectively scaling the solid-state diffusivity and electrolyte conductivity/diffusivity?

[lefile]

Regarding my first problem - influence of an increased particle radius - I took a look on the model equations and it seems that the following changes are needed in the case if radius increases by factor a:

• Solid state diffusivity of the electrode needs to be decreased by factor a^2
• Exchange-current density, since it is scaled by a specific surface, needs to be decreased only by a

I took some dummy parameter set that I'm currently working on and simulated 1C charge-discharge with following parameter changes (a=1.5):

from Custom_parameter_set import get_parameter_values

param1 = pybamm.ParameterValues(get_parameter_values())
param2 = pybamm.ParameterValues(get_parameter_values())
param1.update(
     {
          "Positive particle radius [m]": 1.4e-06 * 1.5,
          "Positive electrode diffusivity [m2.s-1]": 1e-15,
          "Positive electrode exchange-current density [A.m-2]": 2.12e-2
    }
)
param2.update(
     {
          "Positive particle radius [m]": 1.4e-06,
          "Positive electrode diffusivity [m2.s-1]": 1e-15 / 1.5 / 1.5,
          "Positive electrode exchange-current density [A.m-2]": 2.12e-2 / 1.5
     }
)

Here are the results (it's a Na-ion cell, that's why low voltage):

So, voltage difference is not zero, but is in micro-volts. Any idea why models are so close, but not exactly equivalent?

As of my second problem - increase of the electrode thickness - it is harder as I would need to scale electrolyte diffusivity and conductivity based on domains. I.e. anode and cathode would have different scaling factors while separator would not be scaled. Is it possible to define domain-specific electrolyte properties?

[lefile]

I'm bumping this thread since I saw that the new PyBaMM version includes the option to set geometrical parameters as inputs!

However, the release post states that "This mesh does not work straight out of the box with default models, but it can be used in custom-built models."

So, is there any example on how to set-up model that can take geometrical parameters as inputs?