Memory errors. Particles getting lost? (or some other problem)

User Support
1

Hi,

I have a problem running a model and I’m not really sure what the issue is, sorry about the subject line. I run the code and get

`
…
…
Loading ACE cross section table: 73181.09c
WARNING: Negative value(s) found on probability table for nuclide 73181.09c
Creating unionized energy grid…
Initializing source particles…

geometry.xml (5.07 KB)

materials.xml (916 Bytes)

settings.xml (529 Bytes)

2

Hi Kaur,

What cross sections are you using for your simulations? When I switch cross sections to use the NNDC-distributed cross sections, it seems to work fine. I’ve confirmed you don’t have any syntax errors in the XML files.

FYI- hexagonal lattices will likely be in OpenMC version 0.7. See this pull request.

Best,
Paul

3

Hi,

I’m using ENDF/B-VII ace cross section data distributed with an older version of Serpent. Here’s the header of one of the files:
1001.12c 0.999167 1.0341E-07 09/03/09
ENDFB-VII Data for Serpent 1.1.7 (1001 at 1200.00K)

I’m going to try the ones from NNDC later.

Cheers,

Kaur

4

Thanks. It’s enough to take the Ta-181 cross sections from the NNDC library to make it run. I suppose OpenMC is having some trouble with that ace file for Ta-181 at 900K. The same model was also run in Serpent so I don’t think it’s the data itself.

Kaur

5

Thanks for the observation Kaur. I think I have a copy of that Serpent data sitting around somewhere so I’ll see if I can dig to the bottom of it.

Best,
Paul

6

It looks like the root cause of the problem is the fact that the unresolved resonance probability tables for Ta-181 have values that can result in a negative total cross section. This is not really OpenMC’s problem, but it’s reasonable to expect OpenMC to still run. I’m going to try to get a fix in the next release which should be out within a few weeks.

Thanks,
Paul