ERROR:pybamm.logger:Simulation error: Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:1401:

[Akila1993]

Hi @DrSOKane

When I run the Modelling coupled degradation mechanisms in PyBaMM for 500 cycles it gives me the following error.

At t = 292.526 and h = 4.54225e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 294.731 and h = 4.53026e-10, the corrector convergence failed repeatedly or with |h| = hmin.
2023-12-28 12:11:10.017 - [ERROR] callbacks.on_experiment_error(224): Simulation error: Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:1401:
Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:330:
.../casadi/interfaces/sundials/idas_interface.cpp:596: IDASolve returned "IDA_CONV_FAIL". Consult IDAS documentation.
ERROR:pybamm.logger:Simulation error: Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:1401:
Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:330:
.../casadi/interfaces/sundials/idas_interface.cpp:596: IDASolve returned "IDA_CONV_FAIL". Consult IDAS documentation.

This is the same in google collab and locally in my computer.

Thanks !

[DrSOKane]

Happy new year Akila!

I tested this code for 1000 cycles but not for 500, so try again for 1000 cycles. If it still doesn't work, I'll look into it further.

[Akila1993]

Hi Dr @DrSOKane ,

Thank you for your respond and wishing you also a happy new year.

Yes, I have tried it for 1000 cycles and it gives me the same error as below. I hope you will have a look about it.

Thank you for your consideration.

At t = 287.442 and h = 1.18304e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 292.529 and h = 5.72136e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 292.45 and h = 5.75579e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 297.728 and h = 4.76176e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 299.997 and h = 1.51245e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 277.016 and h = 3.56316e-10, the corrector convergence failed repeatedly or with |h| = hmin.
At t = 287.293 and h = 2.05602e-10, the corrector convergence failed repeatedly or with |h| = hmin.
2024-01-03 22:49:55.923 - [ERROR] callbacks.on_experiment_error(224): Simulation error: Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:1401:
Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:330:
.../casadi/interfaces/sundials/idas_interface.cpp:596: IDASolve returned "IDA_CONV_FAIL". Consult IDAS documentation.
ERROR:pybamm.logger:Simulation error: Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:1401:
Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:330:
.../casadi/interfaces/sundials/idas_interface.cpp:596: IDASolve returned "IDA_CONV_FAIL". Consult IDAS documentation.

[ldavezac]

Hi,
Happy new year.
I faced the issue when using the DFN model for simulation:
Error in Function::call for 'F' [IdasInterface] at .../casadi/core/function.cpp:330:
.../casadi/interfaces/sundials/idas_interface.cpp:596: IDASolve returned "IDA_CONV_FAIL". Consult IDAS documentation.
I will be interested if any informations about it are found. Thank you!

[DrSOKane]

After several hours of trial and error, I managed to track down the source of the problem. It's to do with the experiment protocol.

The original experiment was this:

exp = pybamm.Experiment(
    [
        "Hold at 4.2 V until C/100 (5 minute period)",
        "Rest for 4 hours (5 minute period)",
        "Discharge at 0.1C until 2.5 V (5 minute period)",  # initial capacity check
        "Charge at 0.3C until 4.2 V (5 minute period)",
        "Hold at 4.2 V until C/100 (5 minute period)",
    ]
    + [
        (
            "Discharge at 1C until 2.5 V",  # ageing cycles
            "Charge at 0.3C until 4.2 V (5 minute period)",
            "Hold at 4.2 V until C/100 (5 minute period)",
        )
    ]
    * cycle_number
    + ["Discharge at 0.1C until 2.5 V (5 minute period)"],  # final capacity check
)

i.e. everything except for the 1C discharge had a 5 minute period, whereas the 1C discharge itself had the default 1 minue period.

The code was subsequently changed so that everything, including the 1C discharge, had a 5 minute period. When I tried restoring the original experiment protocol, it worked. I am shocked that such a small change would make such a difference, but after years of experience with PyBaMM I am no longer surprised.

I will make a pull request right away to revert the change to the notebook so that future users won't have this problem.

Simon