Methods

Protocol

Constant-current discharges at 0.05C, 0.1C, 0.2C, 0.5C, 1C, 2C to the lower cut-off, from the file's initial state and ambient temperature, with the model its header names (DFN if none). Thermal model only if the file has thermal properties.

Measured data

Only from the file's own Validation section, replayed with the measured current and compared over the time both cover. Rates above the highest measured one are marked.

Not done

Parameter sets are not ranked. Failures are shown with their reason.

Numerics

PyBaMM default mesh (x_n 20, x_s 20, x_p 20, r_n 20, r_p 20) and IDAKLU solver. Doubling the mesh changes capacity by under 0.01% and voltage by at most 3 mV RMS.

Benchmark

Latest patch of each release line from the last 12 months, plus the pinned release, on one machine. Releases that read none of the files are left out. One warm-up, then the median of 3 runs including model build.

Code

simulation_protocol.py, simulate-parameter-sets.py, benchmark-simulators.py. Results are written once per file checksum, PyBaMM version and generator version.