Skip to content

mazeau/surface_w_sensitivity

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This is the master branch

All things happen in the testing folder.

  1. Run whatever in linear-scaling-tests on discovery
  2. Run scratchfiles.py to copy all species dictionaries and cantera files from the most recent run of linear-scaling-tests
  3. Run runmefirst.py to make sure runparallel.sh and copyme-parallel.py are up to date as well as remove old outputs from previous trials
  4. Run ./start_all.sh to start all simulations

This file level is the "base" case. The figures folder on this level shows the output of the "base" copyme-parallel.py The linearscaling folder is for each C and O binding energy. Inside each linearscaling folder, there exists: - dict_conversions_selectivities where the output of simulations are stored before any sensitivity stuff takes place (like a base case) - an identical copy of copyme-parallel.py found up 2 file levels - the outputs from the mechanism generation run by RMG - another figures folder where the figures are all saved - a sensitivities folder where the sensitivities are all saved

The fixed-temp folder is for the imposed temperature profile simulation. The rxnpath folder is where reaction paths would be stored (but stopped working randomly on discovery)

While things are running/queued, run findsens.py to see if any sensitivity simulations failed (so I can rerun before everything has finished)

After all simulations have been rerun successfully, run plots.py or plots.ipynb to generate lsr plots. All plots will be saved in the lsr folder.

The test folder is left over for when I was trying to figure out how to get CPOX to run because the cat_area_per_vol was too small, and oxygen would all attach to the surface and just stay there. This will survive the purge for now.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 87.4%
  • Python 12.5%
  • Shell 0.1%