Tracking and depletion with continuous densities

Suggest new ideas here
Post Reply
keckler
Posts: 47
Joined: Fri Jun 30, 2017 11:48 pm
Security question 1: No
Security question 2: 43
Location: Berkeley, CA

Tracking and depletion with continuous densities

Post by keckler » Fri Aug 31, 2018 7:40 pm

I understand there is now the capability to do tracking through materials with continuously varying densities, as used with the multiphysics interface. Has there been any consideration within the Serpent team for allowing continuously varying materials in depletion calculations as well?

It seems to me that the code now has most of the pieces in place to allow for this (Functional Expansion Tallies, continuous tracking), and all that would be needed is the capability to express material isotopics in terms of functional expansions as well. Once something like this is done, the flux could be tallied using the same order expansions as are used to express the isotopics, and during the depletion calculations the expansion coefficients, instead of the actual number densities, could be depleted. This ability was explored using OpenMC in the thesis by Matt Ellis, but as far as I know it has never been implemented into a widely distributed code version.

I can imagine that the way Serpent allows varying densities is actually on an overall density basis for a given material. From this perspective I can imagine it might take some changes to allow for the relative ratios between isotope number densities within a given material to change with time, but this would be required to allow for spatially-continuous depletion.

I could see this being interesting for a number of reasons, but most obviously would be the ability to remove the sharp discontinuities in material definitions at cell boundaries. Obviously the trade-off would be that, in the previous situation you had to track n isotope densities corresponding to n discrete cells representing a particular isotope within a given component. In the continuous case you would have to track m expansion coefficients corresponding to an m'th order expansion for an isotope in the same component. In order to achieve the same "fidelity" I would guess that n = m, but this may not be the case. Actually one thing that could be very interesting would be to allow the expansion order to vary for different isotopes within a given material, to dial in the fidelity where it is really needed. Regardless, the continuous representation would get rid of the sharp discontinuities which we have become so accustomed to, and I think that would be worthwhile in and of itself.

Has this type of ability been given any consideration for Serpent? Might be quite interesting!

User avatar
Jaakko Leppänen
Site Admin
Posts: 2179
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Tracking and depletion with continuous densities

Post by Jaakko Leppänen » Sun Sep 02, 2018 10:10 am

We have thought about this, but it would require such drastic modifications in the depletion routine that the practical implementation would become pretty complicated. So this is something that is being considered, but at least for the time being, not actively pursued.
- Jaakko

Post Reply