Universes definition using the "solid" card

Questions and discussion about applications, input, output and general user topics
Post Reply
alex.aimetta
Posts: 6
Joined: Mon Nov 02, 2020 2:37 pm
Security question 1: No
Security question 2: 7

Universes definition using the "solid" card

Post by alex.aimetta » Mon Jun 14, 2021 11:30 am

Dear all,
I'm dealing with coupled neutron-photon transport simulations of breeding blanket of the fusion reactor ARC. I have defined the geometry using a CAD software (SolidWorks) and then imported the geometry in Serpent with the STL format (one STL file per component of the reactor). Initially, I have defined the geometry of the whole reactor in one universe using the "solid" card.
Now, I am interested in defining one universe per each component of the reactor, in order to perform the group constant generation using "set gcu". I have some problems in this respect, mainly due to the overlap of cells. The only solution that I have found to this overlap issue is to define series of concentric boxes, each one filled by a different component (i.e universe). Unfortunately, in this way, parts of the components are eliminated, due to the complexity of the geometry (the geometry of the breeding blanket is not cylindrical).
Is there any simple way to assign each STL to a different universe using the "solid" card, even when the STL geometries have arbitrarily complex shapes?
Alex Aimetta

Ana Jambrina
Posts: 551
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by Ana Jambrina » Mon Jun 14, 2021 11:53 am

Decompose the geometry into multiple solids, obtaining multiple STL universes (and associated search mesh). Also, you might want to divide the geometry into multiple levels, nesting into each other the STL components previously defined as separated universes.
- Ana

alex.aimetta
Posts: 6
Joined: Mon Nov 02, 2020 2:37 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by alex.aimetta » Mon Jun 14, 2021 2:57 pm

Thanks for your answer. Actually, I have already tried to define multiple solids, each one assigned to a different universe. I would like to underline that the STL files are associated with concentric geometries. The input is the following:

Code: Select all

surf box cyl 0 0 800 -800 800    % Box containing the reactor

cell 100 0 fill BLANKET   -box  % Fill box with mesh based universe BLANKET
cell 101 0 fill VACUUM2   -box  % Fill box with mesh based universe VACUUM2

cell 2 0 outside   box % Outside of the box 

% --- Create the BACKGORUND universe (filled with void)

surf sINF inf
cell 4 BACKGROUND void -sINF

% --- Create the reactor domain importing the .stl files with the card "solid"

solid 2 BLANKET BACKGROUND
10 4 5 4 3 2
2 1E-5
body Blanket Blanket_cell FLB
file Blanket "Blanket.stl" 1.00 0 0 0

solid 2 VACUUM2 BACKGROUND
10 4 5 4 3 2
2 1E-5
body VV2 Vacuum_cell inc718
file VV2 "Vacuum2.stl" 1.00 0 0 0
The output gives me a geometry error (overlap). Should I define a different containing box and a different background universe for each solid?

Ana Jambrina
Posts: 551
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by Ana Jambrina » Mon Jun 14, 2021 3:41 pm

It should not be necessary to define a different background universe. A similar case is presented with the ‘pbed’ card, explicit stochastic (pebble bed) geometry, with multiple files containing the particle/pebble data and a single background universe definition.
What exactly do you mean by “concentric boxes”? Each of the solids should represent each of the components (shielding, vacuum vessel, coils, etc.), so in principle should not be any kind of overlapping. Have you checked the geometry plots?
- Ana

alex.aimetta
Posts: 6
Joined: Mon Nov 02, 2020 2:37 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by alex.aimetta » Mon Jun 14, 2021 4:41 pm

I mean that the reactor ARC consists of a vacuum vessel made of concentric layers (an external layer of Inconel 718, a FLiBe channel inside of it, a Beryllium layer further inside and so on). Therefore I have built the STL files with this structure in mind.
If I use only one universe (as I have initially done) and one single solid composed of all the STL files, everything works fine (no overlap). When I try to define multiple universes I have overlap. I think that this issue arises because cell 100 and cell 101 in the input are defined inside the same surface (surf box), but I don't know how to work it out.

Ana Jambrina
Posts: 551
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by Ana Jambrina » Mon Jun 14, 2021 5:51 pm

Have you tried replacing the material entries (solid) by the previously defined universes using the ‘fill’ option (nested universes)?
- Ana

alex.aimetta
Posts: 6
Joined: Mon Nov 02, 2020 2:37 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by alex.aimetta » Wed Jul 21, 2021 3:13 pm

Hi Ana, at Politecnico di Torino we have made a request for version 2.1.32 of Serpent, in order to use the 'fill' option in the 'solid' card. Unfortunately, we have been told that currently there are issues in the distribution of new Serpent updates outside Finland, at least for the next few months. Meanwhile, do you think that there are alternative solutions? Thanks

Ana Jambrina
Posts: 551
Joined: Tue May 26, 2020 5:32 pm
Security question 1: No
Security question 2: 7

Re: Universes definition using the "solid" card

Post by Ana Jambrina » Wed Jul 21, 2021 3:58 pm

The ‘fill’ option is available within the ‘body’ entry in the 'solid' definition already in previous versions of Serpent, e.g. 2.1.30 or 2.1.31. Nonetheless, there is always more than one way to define the geometry/model.
- Ana

Post Reply