[Dune] Boundary IDs of the faces of a cube within the DGF file

Dedner, Andreas A.S.Dedner at warwick.ac.uk
Thu May 16 11:56:13 CEST 2013


Okay, if one boundary of your cube is a=(0,0,0) to b=(10,10,0) then you could
use that in the boundarydomain and it would probably work but it would have to
involve a double comparison of the point generated form the interval block to see
if they lie inside the interval [a,b]. So the domainblock does not generate any boundary
one just provides an interval [a,b] and for each boundary segement generated by the
interval block (or any other process for that matter) a test is performed to see if all
the vertices of the segment fall within this interval. Using the exact boundaries to define
this domain is a bit dangerous since the double comparison can go wrong.
That's why in the example I extend the interval - thats why you have the -1 and 11 (-10, 100 or -0.1 and 10.1 would work just as well). In fact one should probably use
(-1, -1, -1) to (11, 11, 0.0001)...
Does that explain the process?
Best
Andreas

________________________________
From: Sacconi, Andrea [a.sacconi11 at imperial.ac.uk]
Sent: 16 May 2013 10:45
To: Dedner, Andreas; dune at dune-project.org
Subject: RE: Boundary IDs of the faces of a cube within the DGF file

Hi Andreas, hi all,

yes, I had a look at the documentation, and since I didn't find what I was looking for I posted my question.

Let's have a look at the example you pointed: examplegrid6.dgf
The grid is a cube, with dimension 10, precisely from the origin (0, 0, 0) to (10, 10, 10).
The comment says that you want the bottom boundary to have ID 2, and 1 for all the others.
The bottom boundary is a square with South-West corner (0, 0, 0) and North-East corner (10, 10, 0).

Why do you use the interval from (-1, -1, -1) to (11, 11, 0) ?? Why -1 and 11?
Does this command work exactly as the block "Interval" used to generate the grid? I.e., we generate a cuboid, not a segment with extremes (A, B, C) and (D, E, F). If so, we would have a cuboid, with dimensions 12, 12, 1.
Would it be the same if I used an interval from (0, 0, 0) to (10, 10, 0)?

Hope that now my questions are more clear.
Thanks again for your help!

Andrea
__________________________________________________________

Andrea Sacconi
PhD student, Applied Mathematics
AMMP Section, Department of Mathematics, Imperial College London,
London SW7 2AZ, UK
a.sacconi11 at imperial.ac.uk
________________________________
From: Dedner, Andreas [A.S.Dedner at warwick.ac.uk]
Sent: 16 May 2013 09:25
To: Sacconi, Andrea; dune at dune-project.org
Subject: RE: Boundary IDs of the faces of a cube within the DGF file

Hi Andrea.
Have you had a look at the documentation (e.g. examplegrid6.dgf)
http://www.dune-project.org/doc/doxygen/dune-grid-html/group___dune_grid_format_parser.html
If you find that unclear then please describe in more detail what is unclear to you
so I can adjust the examples and the documentation.
Best
Andreas
________________________________
From: Sacconi, Andrea [a.sacconi11 at imperial.ac.uk]
Sent: 14 May 2013 13:50
To: Dedner, Andreas; dune at dune-project.org
Subject: RE: Boundary IDs of the faces of a cube within the DGF file

Hi all,

thanks Andreas for your answer.
I had a look at the boundarydomain, and the documentation says: "Each line consists of an integer greater than zero and two vectors describing an interval in dimworld space."
So, and I am not sure if it is directly possible; if you have a 3D cuboid tessellated with tetrahedra, and you want all the triangles lying on one of the faces of the cuboid to be identified by the same boundary ID, how do you do that, without enumerating all the boundary faces of the elements involved?

Thanks again!
Andrea
__________________________________________________________

Andrea Sacconi
PhD student, Applied Mathematics
AMMP Section, Department of Mathematics, Imperial College London,
London SW7 2AZ, UK
a.sacconi11 at imperial.ac.uk
________________________________
From: Dedner, Andreas [A.S.Dedner at warwick.ac.uk]
Sent: 13 May 2013 14:41
To: Sacconi, Andrea; dune at dune-project.org
Subject: RE: Boundary IDs of the faces of a cube within the DGF file

Hi.
You can use the interval block to define your domain (like in 2d) but that has nothing to do with the
boundary ids. Have a look at the boundarydomain block which should do exactly what you need
(it also uses an "interval" to define a region with the same boundary id).
Best
Andreas
________________________________
From: dune-bounces+a.s.dedner=warwick.ac.uk at dune-project.org [dune-bounces+a.s.dedner=warwick.ac.uk at dune-project.org] on behalf of Sacconi, Andrea [a.sacconi11 at imperial.ac.uk]
Sent: 13 May 2013 14:37
To: dune at dune-project.org
Subject: [Dune] Boundary IDs of the faces of a cube within the DGF file

Hi everybody,

I would like to ask you a question about how to prescribe boundary IDs within a DGF file.
I just had a closer look at the DGF format, but I couldn't find how to do what I need.

Let's say that the domain for your computations is a cube (or a rectangular cuboid, it doesn't matter), and you want to impose different boundary conditions on the six faces.
In order to do that, I would introduce different boundary IDs to be read in the code, to guarantee that the faces are treated accordingly.

If you want to assign the same ID to all the faces of elements lying on a certain big face of the cube, is it possible to do it directly, without enumerating all those faces?
I mean, in 2D there is the excellent command "Interval".
What happens if I use the keyword "Interval" (in 3D), passing the four segments forming the perimeter of the (big) face?

Thanks in advance!

Andrea
__________________________________________________________

Andrea Sacconi
PhD student, Applied Mathematics
AMMP Section, Department of Mathematics, Imperial College London,
London SW7 2AZ, UK
a.sacconi11 at imperial.ac.uk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.dune-project.org/pipermail/dune/attachments/20130516/e4f05281/attachment.htm>


More information about the Dune mailing list