Inelastic Scattering and data reading bugs/issues

[Andrew Davis]

I've been doing some cross code validation of several MC codes and included OpenMC, I have found a number of issues I think you need to be aware of and/or should be addressed. What is the best form to provide the group with these results?

[Paul Romano]

Hi Andrew,

Thanks for getting in touch! Emailing this list is fine. Other users might be interested in your findings as well.

Look forward to hearing more.

Best regards,

Paul

On Tue, Nov 19, 2013 at 11:18 PM, Andrew Davis <andrewda...@gmail.com> wrote:

I've been doing some cross code validation of several MC codes and included OpenMC, I have found a number of issues I think you need to be aware of and/or should be addressed. What is the best form to provide the group with these results?

--
You received this message because you are subscribed to the Google Groups "OpenMC Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openmc-users...@googlegroups.com.
To post to this group, send email to openmc...@googlegroups.com.
Visit this group at http://groups.google.com/group/openmc-users.
For more options, visit https://groups.google.com/groups/opt_out.

[Andrew Davis]

Hi I've just realised that I didn't post anything!

Please find attached the PDF https://www.dropbox.com/s/guuplo39nf240qq/validation_report.pdf, the first one to be aware of is N-15, very few codes can get this one right and there are a few others. Ar-38 is an issue, Ca-40 is also a problem. 

One thing that is very good with regards to OpenMC is that a lot of data checking is done, which is not done by MCNP for example. If you have any comments or questions please feel free to contact me.

Thanks

Andy

[Adam Nelson]

Hi Andy,
Thanks for posting this. Let me get the context straight here: it looks like you set up an infinite media, single nuclide fixed source problem with a monoenergetic source at 20MeV and tallied the flux spectra in ultrafine groups?  Is that right? It looks like you did 1E7 particles total?  Finally, this was the ENDF/B VII.1 release which is shipped with MCNP6, or is it a custom NJOY run?

Next, that's actually pretty good to see we get most of the reactions right.  Since most nuclides are just fine, we probably do have the energy/angle sampling right, but just need to better handle some NJOY & ACE 'eccentricities.'  In fact, just this week I was struggling with a Ca-40 issue; looking back at my notes it seems that (in ENDF VII.0 at least) there are two incoming energies for angular distributions (i.e., file 4 data) which are the same value. 

[Paul Romano]

Thanks for the excellent study Andy. It appears we do have some issues to look into.

Adam -- is that you volunteering to figure out the discrepancies? :)

Best regards,

Paul

[Adam Nelson]

I'm definitely interested, I think progress will be slow for the next months (limited time to graduate and a new baby coming in 6 weeks), but I can definitely volunteer for it.