[shimwell]

Hi all

Perhaps this is not quite ready to merge, I wanted to share this code for downloading and converting JENDL 4.0 nuclear data and ask a question about metastables elements.

The other convert scripts have a section where the script opens the ACE file and changes the ZAID number to ensure it is identified as metastable.

I wanted to know if this is necessary when converting an NJOY file ? It appears that the metastables are identified when the h5 files are produced so I guess not. I just wanted to double check because this is the main difference between this script and the others.

The other difference is the use of openmc.data.IncidentNeutron.from_njoy(filename) which is slower than openmc.data.IncidentNeutron.from_ace(filename)

Best,

Jon

[paulromano]

Thanks for contributing this @shimwell. I have a few requests for changes if you don't mind. Also, I'd prefer the name of this script be generate_jendl.py. My naming convention is:

• convert_*.py = Convert existing ACE files
• generate_*.py = Generate ACE files using NJOY and then convert

[paulromano]

One more comment I forgot to put before. I know you fusion guys don't care, but it would be good to include the JENDL thermal scattering sublibrary too. I might make some updates to this script though once merged, so feel free to ignore that and I can take care of it.

[shimwell]

I also spotted a few convert scripts where we are downloading files without using the handy download function in the openmc _utils.py script. So I was able to remove a decent amount of lines in these scripts.

[shimwell]

At the moment I am running all the download scripts to check that they work but it takes a while with the njoy parts. I guess this repo is not well suited to CI testing due to the large amounts of data downloading and time required for processing. Perhaps a topic for another PR but in the future we could carry out some automated testing with the different libraries. Perhaps something like this ...

# Run all download scripts within a Circle CI container with OpenMC installed
python3 data/convert_fendl.py
python3 data/convert_jeff32.py
python3 data/convert_nndc71.py
python3 data/convert_tendl.py
python3 data/generate_endf71.py
python3 data/generate_jendl.py

# set OPENMC_CROSS_SECTIONS path to each library 
# run simple sphere model for each isotope in the library with a neutron spectra tally
# flag up inconsistencies with the tally results

[paulromano]

Thanks for cleaning up some of the other scripts -- I had been meaning to get rid of the manual downloading!