Is there a tool currently that transfers OpenMC cross sections to Griffin? I have looked found the TOGA code (GitHub - lanl/TOGA: Tool for Optimization and Group-structure Analysis of nuclear multi-group cross sections) but I am finding that when calculating removal cross sections from OpenMC, Griffin calculates a negative removal cross section. I am assuming the cause is likely related to scattering and n,xn reactions not being accounted for properly. I was wondering if there might be a similar tool out there or possibly an updated version of TOGA that people are using to transfer the data between codes.
Thanks
Jonathon
1 Like
Hello Jonathon,
I developed TOGA in 2020. Since it was developed for an internship and wasn’t verified against any benchmark, I recommend it primarily as a template for how you can convert OpenMC cross section libraries into isoxml format for Griffin. I am glad it has been useful to you so far.
Removal cross sections can be a bit tricky when you aren’t sure if (n,xn) reactions are being accounted for in the absorption. I would point you to this discussion for further details: Different k-eff in RESULTS and calculation - #13 by paulromano
Note that there is now a ReducedAbsorptionXS class in OpenMC’s mgxs library that might help.
Sincerely,
Miriam
1 Like
Hello,
Thank you for the response! I was able to play around with it quite a bit and eventually get it working properly for Griffin. The conversion to XML part was the most difficult for me but luckily the code in TOGA was very easy to modify once I picked which tallies to use and now it is working very nicely. I am still working on verifying so I am not sure if this is the fully correct setup between the two but I will write what worked for me below in case others have the same questions in the future.
For the scattering matrix, I found that using the scattering probability tally multiplied by the nu-scatter tally worked very well. I also believe that the scattering matrix should then be inverted when inputting into the ISOXML format - atleast that seems to give very good agreement in my tests. The (hopefully correct?) removal xs that I used was then reduced absorption + nu-scatter - the diagonal of the scattering matrix. I am still trying to test this with some various cases against reference solutions however. The setup looks promising so far and using TOGA as the base has also greatly sped up the data transfer process so thank you again for providing this code.
Jonathon
1 Like