Discussion Forum

I am trying to simulate the voltage of a primary battery cell by applying a pulsating current. The battery should be functional for more than 5 years with this current.

Because the pulses are small (less than 8 seconds), the timestep should be very small. But having a 1-second time-step for 5 years exceeds COMSOL's capability, so I was wondering what is the best option for this. If I choose Strict Time Stepping and 1 s time-step, it works very nice for a shorter time range, but if I choose Free Time Stepping, the solver jumps over the pulses. However, the Strict option is limited by the maximum number of time steps allowed in COMSOL. My question is...

Is it possible to modify the solver so that it will have a higher density of time-steps around the 8s pulses and coarser otherwise? The pulses are generated using an analytical function

I have created a simplified model to illustrate this issue (see attachment).

If one goes to:

Study 1 > Solver Configurations > Solution 1 > Time-Dependent Solver 1 > Time Stepping Option

and chooses Free in the "Time steps taken by the solver", one can see that the capacity (Q in the results) jumps over the pulses. If one chooses Strict, it will calculate the capacity without any issues, as expected.

However, for simulating a 5-year range, for the second option, the time-steps should be minimum 100 s, which would jump the 8s pulses and will take a long time (sometimes COMSOL freezes if I use 100 s and 5 yr range).