Weight windows in OpenMC 0.13.0

Hey! I have been intrigued by openmc for some time now, so after years of using MCNP I started with some simple examples comparing MCNP/openmc results. I have seen that in the latest version of openmc the weight windows have been introduced, but I could not find some simple examples (or workflow) of how to use this option. Any suggestions where to look for the information?

I’ve been working on some simple examples, I don’t think they are fully finished but perhaps they are still helpful.

Ultimately I’m hoping to make an example with a 3d mesh and another one with a neutron spectra tally.

Part of the reason I’ve been working on other things and not finished these examples is because the generation of the weight window values is difficult to do manually. I plan to come back to this once automatic weight window generation is possible, so I think that is the area to focus on for now.

1 Like

Yes, unfortunately we don’t have a good example of it at present. You can take a look at the documentation of the new WeightWindows class, but I realize that’s not super useful without a real example. @pshriwise has been working on a capability for automated weight window generation and we’re hoping to get that in the main branch of the code in the coming months.

1 Like

Thank you both for replying. I have quickly checked the provided files (and the WeightWindows class) - if my understanding is correct, using the global variance reduction technique I would be able to produce the upper/lower ww bounds (e.g., weight_window_values.py) and use them in my subsequent run as in @Shimwell examples.

@akolsek yes that sounds correct to me. I will of course update those examples when possible but it might be best to keep an eye out on the openmc repo pull requests. That is the place where the docs will first appear and be reviewed by the developers and community (including yourself if you are keen)

Thanks for the help - I have implemented a simple global variance reduction into the “build_xml.py” file and did some very simple tests.

2 Likes