Issue #115: Performance of polarization-curve simulations degrades over time (ESDLab/openfcst)
AslanK
https://bitbucket.org/ESDLab/openfcst/issues/115/performance-of-polarization-curve
Aslan Kosakian:
The time required to calculate one point of a polarization curve increases with every iV point. For instance, if you run examples/PemfcTPPSD/polarization\_curve/1D\_two-phase\_CCL\_polarization\_curve/, the first polarization iteration takes only a second, but the computational time increases with every iteration and reaches unreasonable amounts by the end of such a simple simulation.
Specifically, the bottleneck appears to be somewhere between this output…
`Application->Copy->Newton3ppBase->->AdaptiveRefinement `
...and this output...
`ReactionNames is: ORR `
`KineticsNames is: DoubleTrapKinetics `
`...`
Therefore, the issue is not with the physical simulation itself, it is likely somewhere in the initialization, perhaps of the kinetic class\(es\).
This issue can be reproduced on:
Ubuntu 20.04, gcc 9.4
Ubuntu 22.04, gcc 11.4