SEI disappear when active material loss

[Yuxiao-Lan]

Hello pybamm!
I am running coupled-degradation example and modified the following parameters:

param["Outer SEI solvent diffusivity [m2.s-1]"] = 2.5000000000000002e-22 * 20
param["Negative electrode LAM constant proportional term [s-1]"] = 1.2e-6 * 20

After running 365 cycles, I found that the trend of porosity versus throughput capacity declines initially and then increases, as shown in the Figure. I am curious about which factors causes the effect.
In the reaction_driven_porosity.py file, line 60, the code illustrates how PyBaMM calculates the solid volume fraction, $\epsilon_s$:

$$ \epsilon_s = \epsilon_{init} - a_kL_{tot} $$

According to this, porosity will increase as $a_k$ decreases due to the loss of active material. This is somewhat counterintuitive to me because the SEI mass generated before is disappearing. How can I modify the equation to represent the change in porosity over time as follows?

$$ \frac{d\epsilon_s}{dt} = - a_kdL_{tot} $$

I would greatly appreciate any insights or assistance in clarifying this matter!

[DrSOKane]

Hi @Yuxiao-Lan, I am aware of this issue and have proposed a fix. However, the fix caused an unexpected problem for models such as SPM where the option"x-average side reactions" is set to "true". We have been trying to track down the source of the bug for months now, with no success.

[Yuxiao-Lan]

Thank you for the rapid response and explanation.