Photon Heating

[philipJFpfahl]

Hello,

I am trying to calculate the heat deposited in the moderator using Cardinal and OpenMC.
I observed a difference between the results. The heating for Cardinal in the moderator is around 2.1% of the power, while it is 7% for OpenMC. Presumably, the photon heating is causing the issue. I can reproduce the 2.1% heating in OpenMC by setting the photon transport flag to false (settings.photon_transport = False).

Can Cardinal capture the photon heating, or am I missing something?

All the best,
Philip

[aprilnovak]

Hi Philip, are you running your Cardinal simulation with the same settings for photon transport, and same tally score, as in OpenMC? You'll want to make sure that you have photon_transport set to True in your XML files for Cardinal, and to use the same tally score for comparisons.

Here's the list of tally_score options we have in Cardinal: link

[aprilnovak]

Thanks for sending! Taking a look now

[aprilnovak]

Hi Philip,

Good news! This is not actually anything related to photon transport. In your standalone OpenMC runs, you are using a mesh tally (which in OpenMC uses a collision estimator). However, in Cardinal you're using cell tallies (which by default use tracklength estimators). The difference in the tally results is only coming from the different estimators. If you were to increase the number of particles, you'd see that the tallies agree within statistics.

Alternatively, you can set the tally estimator used for Cardinal's tallies to a collision estimator (to match what you had in OpenMC), and you can confirm the results are identical.

[Problem]
  type = OpenMCCellAverageProblem
  tally_estimator = collision
  ...
[]

[philipJFpfahl]

Hey April,

Yes good news indeed.
That also answers the question. Thank you.

Does that actually mean that I do 2 tally calculations when adding a tally.xml file?

Philip

[aprilnovak]

Correct, Cardinal will create all the tallies it needs for multiphysics in-memory automatically (users don't need to do anything except indicate tally_score in the cardinal.i), so they won't appear in the tallies.xml.

You can still add additional tallies in a tallies.xml if you wanted to, but they won't be passed into Cardinal (so you'll have to explore them using OpenMC's tally visualization/manipulation features).

[philipJFpfahl]

Ohh okay that is good to know :)

Thank you again for your help.