Error while trying to run drive cycle in experiment

[shwetankverma28]

Hey everyone,
I am trying to run a drive cycle in an experiment but it keeps failing. The code is a small modification of this code given herein experiment_drive_cycle.py

import pybamm
import pandas as pd

# import drive cycle from file
drive_cycle_current = pd.read_csv(
    "US06.csv", comment="#", header=None
).to_numpy()

parameters = pybamm.ParameterValues("Chen2020_plating")
experiment = pybamm.Experiment(
    [
        f"Charge at  1C until {parameters['Upper voltage cut-off [V]']} V",
        f"Run US06_A (A) until {parameters['Lower voltage cut-off [V]']} V",
    ]*100,
    drive_cycles={
        "US06_A": drive_cycle_current,
        
    },
)

model = pybamm.lithium_ion.DFN(options={ "particle": "Fickian diffusion", 
                                           "SEI":"solvent-diffusion limited", 
                                           "SEI film resistance":"distributed",})
sim = pybamm.Simulation(model, experiment=experiment,parameter_values=parameters, solver=pybamm.CasadiSolver("safe"))
sim.solve()

# Show all plots
sim.plot()

I want to run drive cycle experiment multiple times but it keeps failing with this error:

At t = 0.00337103, , mxstep steps taken before reaching tout.
At t = 0.000700632, , mxstep steps taken before reaching tout.
At t = 0.000700623, , mxstep steps taken before reaching tout.
At t = 3.30309e-005, , mxstep steps taken before reaching tout.
At t = 3.30324e-005, , mxstep steps taken before reaching tout.
At t = 3.30328e-005, , mxstep steps taken before reaching tout.
At t = 3.30334e-005, , mxstep steps taken before reaching tout.
At t = 3.30292e-005, , mxstep steps taken before reaching tout.

SolverError: Maximum number of decreased steps occurred at t=757.8750000000014. Try solving the model up to this time only or reducing dt_max (currently, dt_max=0.006925422836696886).

Any solutions? I have tried setting dt_max to 1e-6 but that doesn't solve this.

[brosaplanella]

Have you tried plotting the solution for shorter cycles to see how the battery behaves? Could be that the solver struggles due to a very low porosity.