Hello everyone,
I have been working on a microreactor model with control drums positioned around the core. I’d like to model reactivity insertion accident (RIA) by rotating control drums (like a control rod ejection in a PWR).
I have seen some discussions here from past years, but I am still unsure if the latest version of OpenMC can model such transients?
Thanks a lot in advance for any help
Yesim
Hello Yesim!
I am also planning to do some moving control drum simulations in the future. I don’t think there’s any hurdle to implementing a model script that takes an arbitrary drum angle as an argument and creates an XML input for that angle.
I assume that for an accident scenario you will be concerned with the temperatures in the system (which will have reactivity implications as well). You will need some kind of multiphiysics solver to accomplish this. If you are early enough along, I recommend using Cardinal. There will be some tutorials at that link on how to set up a multiphysics simulation that couples to OpenMC, but AFAIK no one has done a drum-rotation transient, coupled multiphysics simulation.
I will be very curious to see your setup if you end up doing this route!
Best,
Lewis
Hello Lewis,
I appreciate your suggestions. I have gone and checked Cardinal and I think it would be even better if I can use it for my project. However, what I am still not sure is that if OpenMC can be used for time-dependent simulations?
Does the version of OpenMC coupled with NekRS in Cardinal is different than the version in the following link? OpenMC 0.15.2
I have also found some papers on Cardinal but it looks like the thermal side is modelled as transient, while OpenMC runs in static mode at each iteration in those studies.
I am puzzled about how to proceed for standalone or coupled neutronics transient model with OpenMC. Do you happen to know if there are such studies with Cardinal ?
Thank you
Yesim
I do believe you can use time bins in OpenMC via the TimeFilter for a standalone simulation, but I’m not sure how this would work properly (if it can work?) with Cardinal. Here are the example tests in OpenMC that use TimeFilters. Keep in mind that having small time bins would require very many particles to get enough contributions in each bin (similar to using a very fine mesh in space).
Doing standalone neutronics, you might be able to define an accurate time-dependent source distributions if you knew enough about your problem, but this all becomes complicated if fission is involved.
I have also found some papers on Cardinal but it looks like the thermal side is modelled as transient, while OpenMC runs in static mode at each iteration in those studies.
You are correct; this is how Cardinal is typically used. To do a time dependent simulation in Cardinal with neutronics, I believe you would do a steady-state OpenMC solve at every point in your transient multiphysics solve. The neutronics becomes “time-dependent” by updating temperature and density as a function of time and then running transport with the updated T and \rho in your system. The neutronics can feedback with the other physics and stay up to date with the physical simulation time.
Does the version of OpenMC coupled with NekRS in Cardinal is different than the version in the following link? OpenMC 0.15.2
Looking at the current Cardinal repo’s devel branch at the /contrib directory (where the OpenMC version is stored), it seems like they are using OpenMC at this commit. Cardinal is very flexible with which physics solvers run, and you may be able to use a simpler fluid solver (as opposed to NekRS, for example THM or SAM). It would be problem dependent and depend on which codes you have access to, but there are Open-Source versions of most of the physics solvers