Fixed Source Energy Spectrum Reverting to Watt Default

User Support
1

Hello,

I am attempting to define a custom energy spectrum for a fixed source simulation as part of a neutron spectrum shaping project. I am using a custom 1353-group energy spectrum that spans 1E-9 to 20 MeV and is known to work with MCNP. The bin edges have been converted to eV for use with OpenMC (version 0.12.0 for python 3.7). The associated probabilities (numbering 1353 values) have been normalized to sum to 1. The first and last values have been set to 0. I am using an openmc.stats.Tabular distribution with histogram interpolation.

No errors are raised when this energy spectrum is loaded or exported to xml. The xml file appears to contain all of the data and inputs in the proper format. However, after running the fixed source simulation, a plot of the results of a flux tally on the inner-most cell reveals a Watt fission spectrum. Note that all of the cells in this geometry are voids and should not be affecting the neutron energies. No errors are raised nor are any abnormal notifications printed. I have tested the same setup with simple (e.g. 4-group) custom energy spectra and found that the resulting plots to show tally results matching the input distributions. Does anyone have suggestions for a solution to this issue or any errors I have made in the input?

Also, a previous post https://openmc.discourse.group/t/source-energy-distribution-for-neutron-of-different-energy-group/477 indicated that the probabilities should be divided by the associated energy bin width. Is that still the case?

I would greatly appreciate any assistance on this. Thanks!

Google Drive folder with xml files

Edit: added links to xml files on Google Drive

2

Hi @imwilkinson and welcome to the community! The probabilities for the distribution should be given per eV. That is, if you integrate the probability distribution (in your case represented as a histogram function) over energy, you should get one. In your case, they do not appear to be normalized per eV, which will result in the bins for lower energies receiving much less probability (since they are multiplied by the bin width). It’s not surprising then that you only see the higher end of the spectrum, which might look like a Watt spectrum.