OpenMC Tutorial | Converting the World's Most Detailed MSRE CAD Model to Simulation

[@NAJDAR]

How and where do you assign the materials to the geometry? I see you are filling the regions with cad models, but how do you tell the code to put materials in specifically assigned places?

[@bobbooker1948]

Which nuclear data set should I use for this tutorial?

[@saschaapel7628]

iam working for my degree on a GUI with own created, examples of openmc. with comments and explanations to every step. to show the workflow and some important essentials. as good as it is to couple a CAD with openmc. You have to read and understand the code. especially for the geometry file of opnemc. it's a stiff climb into understanding openmc. besides the physics you als need to have a understanding.

[@erikbergbackknudsen6346]

Of course - understanding is key. Part of our intentions with this project is give people a simple way of gaining knowledge and understanding of the MSRE - and allow them to tinker with the reactor.

[@saschaapel7628]

@@erikbergbackknudsen6346 question: what kind of people you want to understand about msre? Ian asking because it took me a complete studies to get a picture of reactor physics. It took me many hours to get an idea of how openMC is working and how I would make an attempt to train people, who have a knowledge of the physical basics of a reactor, in this program. Programs like scale came to the same conclusion and have an integrated gui. It’s still complex to work with. In my opinion on the didactic level you started in the middle and not the start. I would like to know, at whom you aim with your intentions. It’s not meant as an offence!

[@erikbergbackknudsen7695]

@@saschaapel7628 The way I think of this, is that several sets of people may be interested in this: - some interested in trying out reactor computations, may have read about the MSRE, who may then later back up to understand the physics behind. - people who know about reactors and reactor sims, can have easy access to the tools to run an extremely detailed MSRE-model. As you know, building a model of something as complicated as the MSRE is not trivial. - already advanced simulators who may interested in migrating to OpenMC and what it has to offer, can get a leg up. Last, personally I think it is a pity that MSRE is talked about as a failure. It was an experiment pushing the envelope, and was IMHO very successful in doing so. The time was not perhaps right for commercial deployment of MSR-reactors, but that does not detract from the accomplishment, and I want to address that. Didactically speaking you are of course right. It is not sensible to try teach a course in reactor physics in 15 minutes :-). OTOH - to give anyone practical access to a detailed MSRE-model does make sense to me.

[@erikbergbackknudsen6346]

@@saschaapel7628 I sort of agree with you on the didactic level. If we were aiming at creating a course in nuclear reactor physics and design this is not a sensible way to go about it - but that could be said for any 15-minute video. It would have to be a whole set of videos. Wrt. target audience: actually a set of different people: - Tinkerers who would like to get started doing something with OpenMC - and here MSRE is one example among many. - More experienced users who are now able to easily use this very detailed model of MSRE. Building a model MSRE from scratch is a huge task. - Advanced users have a vehicle for applying OpenMC to an MSR. MSRE remains the only MSR to have run for a long time - and thus it is an important example. Also an important point was to showcase some software packages than allow anyone to build geometries with easily accessible open-source tools and transport particles around in them. All in all - what we aim for is to get more people interested in nuclear design - who may then take the step back and really learn the physics. I hope that at least succeeded on some level 🙂

[@saschaapel7628]

@@erikbergbackknudsen6346 Kopenhagen atomics was the reason for my theme of my thesis. that's why I aim asking. your company had an announcement for ppl. who wanted to quote: Soften the learning curve for students new to OpenMC. so I follow your company for a while.

[@BriannaGouteyron]

I am trying to use cad_to_openmc but I need to associate materials to geometries in my step file, however, I cannot do that with salome directly, did you incorporate the materials in the cad_to_openmc code or did you associate the materials in open mc before a simulation ? also what mesh parameters did you use because the documentation of cad_to_openmc entions nothing about submeshes and if/how we can create them.

[@madsenism]

This is great! Thanks Erik

[@erikbergbackknudsen6346]

You are most welcome. Glad you enjoyed it.

[@mr_happygolucky7095]

Did you get it to run? I don't know how to get the .xml files.

[@zeinabmohammadi1379]

Thanks for your great tutorial. I have a question regarding the activate demosim, when I do that, I can't see demosim on the left like what you have? I was wondering does it make any problem?

[@erikbergbackknudsen6346]

Back from vacation: I am glad you liked it - thanks for getting back to me. What is on the left of the prompt is purely cosmetic. It's merely a reminder that you have an environment activated. I'm afraid I don't know precisely why you don't see it like that - it really depends on how your system environment is set up. Specifically, what is written on the left of the prompt is governed by the $PS1 environment variable. In my case it is normally: " \[\033[01;32m\][\u@\h\[\033[01;37m\] \W\[\033[01;32m\]]\$\[\033[00m\]" which gets altered to "(demosim) \[\033[01;32m\][\u@\h\[\033[01;37m\] \W\[\033[01;32m\]]\$\[\033[00m\]" when the environment is activated. Roughly speaking, what matters is what is in the $PATH-variable. This one should have a first entry that points to a bin-directory inside the environment location. If it does then that installation will take precedence over system locations.

[@jooch_exe]

Nice work! What's the error margin of the models used?

[@CopenhagenAtomics]

That is unfortunately a very difficult question to answer - seeing that the experiment was carried out 70+ years ago, and to our knowledge none of the equipment remains. The model is built from drawings, reports, and measurements that we have access to. More data exists, on its way to long term storage, but it is (we are told) possibly intermingled with classified things, meaning access is non-trivial. That said, in terms of criticality, the CAD-model gets significantly closer to experiment, than the CSG-models that exists, e.g. in terms of the IRPhE-benchmark." An upcoming article discusses some of these models in detail: www.frontiersin.org/articles/10.3389/fnuen.2024.1385478/abstract

[@gavinridley5727]

Epic. But to clarify, does that say 40 particles per second in the final moments of the video?

[@erikbergbackknudsen7695]

It does, but it is an artefact of the video recording process. Just reran the same thing (on that machine) at >5000 /s (because I couldn't remember the number)

[@gavinridley5727]

@@erikbergbackknudsen7695 Ah nice, thank you!