2
$\begingroup$

I have a nuclear reactor design I would like to model. I would like to show the individual atoms and how they interact with each other in the reactor (specifically, I would like to model decay modes, interactions with photons). I was wondering if there was any software which would help me model this in 3D.

This reactor design is one of my own, so I know all about what happens inside of the reactor. I have used Geant4 and similar software before, but I would just like a simple graphical interface in which I could model nuclear systems.

$\endgroup$
  • $\begingroup$ Of course there are several such codes, but the leading examples are frameworks for simulation that assume the user to be both (a) a pretty competent programmer and (b) pretty familiar with reactor design and the nitty-gritty details of what goes on inside a running reactor. Your profile suggests that your current accomplishment is more like an undergraduate physics major than an active researcher, so they may not be for you. Perhaps you could add some information about your preparation? $\endgroup$ – dmckee --- ex-moderator kitten Apr 29 '18 at 17:41
  • $\begingroup$ @dmckee This reactor design is one of my own, so I know all about what happens inside of the reactor. I have used Geant4 and similar software before, but I would just like a simple graphical interface in which I could model nuclear systems. $\endgroup$ – Aakash Sunkari Apr 29 '18 at 18:09
  • $\begingroup$ I'm afraid that at the level of the codes I'm aware or the user has to provide their own interface. $\endgroup$ – dmckee --- ex-moderator kitten Apr 29 '18 at 18:41
1
$\begingroup$

If you would like to reliably model the inner workings of a nuclear reactor, you're looking for MCNP.

Originally the acronym stood for "Monte Carlo Neutron Photon," but in the past decade more physics has been added and the official interpretation is now "Monte Carlo N-Particle." Features include a language for specifying arbitrarily complicated geometries and associated material compositions, an extensive library of energy-dependent cross section data, and, most importantly, a history of accurate benchmarks between simulations and functioning reactors that goes back for decades. There are several associated visualization tools that have been improving in quality recently.

If what you want is "a simple graphical interface," then MCNP is not for you. But at dmckee says elsewhere, that's par for the course in this business. The costs associated with the independently-developed MCNP Visual Editor and its worldwide training sessions suggest to me that there is not some magical superior free alternative floating around on the internet.

MCNP is a product of the US government and its export is restricted. To obtain MCNP, you have to open a correspondence with the Radiation Safety Information Computational Center at Oak Ridge National Laboratory. The RSICC also distributes software with less restrictive (and less expensive) licenses; you might explore there to see if anything meets your needs.

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

Two codes used by the neutrino community to generate predictions of the anti-neutrino flux expected for our detector systems are DRAGON and MURE.

I'm not an expert on either code, but I know they produce slightly different level of detail. DRAGON is a parameterized code while MURE is a full particle-level Monte Carlo. This means that DRAGON will generally be faster but MURE will generally capture more detailed behavior.

Like Geant these are frameworks with which the user builds a model of a particular device. So the user has to supply, geometry, fuel composition and distributions, as well as other physical parameters. They are not for the faint of heart.

To give you a sense of both the detail available from the codes and the work required to get there consider these preprints:

(Preprints select because I was part of those collaborations and knew where to look for them).

$\endgroup$
0
$\begingroup$

A "general" reactor solver and GUI doesn't really exist. Especially one where you want to show individual particles. (I'm not sure how you would show individual particles, there are approximately $10^{10}$--$10^{20}$ particles in a reactor system.)

If you have a new reactor concept that you want to model, you would normally break it down into the following components: geometry, materials, cross sections, particle transport, depletion, thermal-hydraulic modeling and feedback. You would then apply applicable physics to each part. Different reactor concepts require different physics. For example, you wouldn't model a high-power molten salt reactor with the same tools you model a barely critical graphite pile. The physics are different.

If you can provide more information about your concept (specifically the geometry, power, coolant, and materials), we can try to give a better answer.

$\endgroup$
0
$\begingroup$

You can use following :

  1. MCNP - Export controlled by US
  2. OpenMC - free
  3. MONK and RCBEND - Can be obtained from UK (costs approx 3,500 euros)
| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.