Feature Request: Diffusion coefficient for two step calculation

[khurrumsaleem]

Hello to all the developers.

I am creating this issue for a feature request. I have seen some papers regarding CMM method implementation for determination of diffusion coefficient using OpenMC. Is there any chance of that feature being a part of public branch of OpenMC?

Thank you all and keep up the good work.

[paulromano]

Hi @khurrumsaleem. Pinging @johnnyliu27 on this one since he is the CMM guru. My understanding is that it's not quite ready as a production feature, but my knowledge may very well be outdated. I'll let @johnnyliu27 speak for himself.

[rockfool]

@paulromano Rgiht. I think @johnnyliu27 has trouble in getting access to github currently. As I know, he finished CMM in his local repository based on old OpenMC with Fortran, which is not ready to be merged in the latest OpenMC.

[khurrumsaleem]

Thank you @paulromano and @rockfool.

@rockfool can you inquire that is it possible that @johnnyliu27 will be working on this feature in future? Can @johnnyliu27 share the Fortran based repository with CMM implementation. This might help us to take a crack at the capability...

Thank you and keep up the good work

[johnnyliu27]

Hi @khurrumsaleem, thanks for the interest in CMM and sorry for my late response. CMM is ready for generating diffusion coefficients in lattice calculation (homogenized D_g for infinite medium with all reflective boundary conditions), but it has not been in the main branch since we have been trying to make CMM work for generating heterogeneous diffusion coefficients and transport cross secitons, which is more general like every normal tallies in OpenMC. But this attempt is not straightforward and I spent quite a lot of time in the analytical part.

But for your interest of diffusion coefficients in lattice calculation for two-step calculation I do have a public branch in my own repo here (https://github.com/johnnyliu27/openmc/tree/ima). The branch name ima stands for “incremental migration area”. This branch was half done when I was trying to test its effectiveness on heterogeneous tallies, but it can do the job of computing homogenized D_g for infinite medium with all reflective boundary conditions. I implemented a new score for “incremental migration area” with the score name “ima”, and in the tally results using “ima/flux” will just give you the CMM diffusion coefficients for each group, without any further processing like in the original cumulative tally scheme in the first CMM journal paper. (But the warning here is that the result is only reliable for tallies defined for the whole geometry of infinite medium. Although I implemented the mesh filters for ima, my test results showed that they are not working well.)

For the implementation, I do have a future plan to make it a production feature in OpenMC. But I have lost track of the code update for a long while and I’m tied up with some other work now, so unfortunately I don’t yet have a timeline on this right now. But I’ll be happy to answer your questions about the method or if anyone else wants to give it a try.

[khurrumsaleem]

Thank you very much @johnnyliu27.

[khurrumsaleem]

@johnnyliu27 @paulromano looks like this fell off the radar........ :-(