nuclear reaction fluid modelling

Question

I'm pretty ignorant regarding the dark arts of numerical codes and modelling, but i'm interested in trying to pursue it for a particular pet project. It regards modelling of nuclear reactions like those required for NERVA-like rockets or this one.

I would appreciate any references regarding implementation, models that need to be used (i.e: Navier-Stokes 2D with MonteCarlo neutron diffusion, etc.), or for existing publicly available codes or frameworks that could get me started. I've found that it is very hard to find most of those developed at Livermore because of Export restrictions. I've requested TART and MCNP5 downloads for a couple weeks now and i don't even get a reply

More precisely:

• Are there existing models implemented in publicly available codes for this kind of domain modelling? if not, what is the closest one and what kind of tweaking would be required?

Edit

I was able to find this one called TALYS but i haven't had a time to look at it deeply. Will try to keep this question updated as i learn more about what i need for my project and if/what portion of it is provided by this library

Answer 1

This is a very complex topic, and indeed, for reasons of export control you won't be able to get codes for a 'pet project'. However, there are publicly available documents on results of such computations, and you are advised to look for them in DoE's OSTI and NASA's NTRS systems.

To understand what's going on, would suggest reading up from the basics (like Bell and Glasstone, Nuclear Reactor Theory, Van Nostrand, 1970).

There are commercially available packages that would do at least part of the work for you (COMSOL Multiphysics, Ansys, and others), if you go the 3-D way. You'd still need to code neutron diffusion and fuel behavior yourself (supposing you get cross-section and fuel data somewhere), and tie it together with thermohydrodynamics of liquid hydrogen or methane or whatever.

Answer 2

What a fascinating design for a rocket!

Having -- unfortunately -- had to observe the struggle of some colleagues with the export control issues, I can certainly tell that you won't be able to find any open source codes for this kind of problem in the United States. Everything that could potentially solve such problems would also be able to model nuclear explosions and would, consequently, be export controlled.

I don't know how other countries, say in Europe, treat these issues.

Answer 3

In Europe, nuclear codes are either proprietary or under the responsibility of the OECD Nuclear Energy Agency in Paris (which acts as an intermediate for some codes with RSICC, the US databank). If you live in an OECD country and you are an OECD citizen, you should be able to ask NEA for the country's "liaison officer". She/He is the person who can request codes from the databank.

However, as professor Bangerth mentioned, for anything close to these topics, you will have to have a very good justification (and "pet project" won't fit) and be affiliated to a nuclear research center or university research group to request these codes.