CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Running, Solving & CFD

Moving unstructured mesh with changing topology

Register Blogs Community New Posts Updated Threads Search

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   April 26, 2008, 07:23
Default Heya! I am an OpenFOAM newbie
  #1
New Member
 
Mark Eaton
Join Date: Mar 2009
Posts: 4
Rep Power: 17
meaton is on a distinguished road
Heya!
I am an OpenFOAM newbie and currently I want to evaluate whether I can use OpenFOAM to simulate the flow of a melting material, the initial structure given as an unstructured tetrahedral mesh. The flow results in strong mesh deformation, for example pinching off drops or recombination of drops falling onto other material.
Because the initial structure is very nested, and it fills only a small fraction of a circumscribed box, I want to use a surface tracking solver rather than the volume-of-fluid approach.

I'd appreciate if you help me get started by answering these two questions or of course suggesting anything that you think could ease the solution:

1) This is the simple one. Are there any (documented!) examples how to have OpenFOAM operate on unstructured meshes? For example, if I don't blockmesh the box of the cavity tutorial case, but have it meshed by some CAD program, is there anything necessary to know besides how to convert the mesh into the OpenFOAM format?
Do I need to tell OpenFOAM that it is a tetrahedral (or general polygonal) mesh rather than a hex mesh as generated my blockMesh?

2) This is probably complicated, and I don't expect an answer that enables a newbie to achieve everything I mentioned above. But please give an overview what is available so that I know where to put my effort.
I have read http://openfoamwiki.net/index.php/Ho...mic_mesh_cases. Firstly, the classes mentioned there don't exist in recent versions. Does an update of this Howto exist?
Secondly, the Howto doesn't describe how to detect and repair regions of low mesh quality, but instead assumes that you know a priori where to add or remove layers. Do you have any example (documented or not) which does surface tracking of strongly deformed meshes and includes *robust* mesh motion and automatic topology correction?

Regards
Mark
meaton is offline   Reply With Quote

Old   April 26, 2008, 14:03
Default Mark, 1. You have several o
  #2
Senior Member
 
Sandeep Menon
Join Date: Mar 2009
Location: Amherst, MA
Posts: 403
Rep Power: 25
deepsterblue will become famous soon enough
Mark,

1. You have several options to convert from generic meshes to the native Foam format. (Available in: OpenFOAM/OpenFOAM-1.4.1-dev/applications/utilities/mesh/conversion)
Do take a look. FOAM will handle polyhedral meshes without problems, and you needn't specify anything.

2. I'm working with the dev version, and I know that the necessary classes for dynamic meshes with topology changes exist there (I haven't checked the OpenCFD version, though).
By detecting and repairing low mesh quality, I'm guessing that you're talking about general tetrahedral meshes. Layer addition and removal applies specifically to hexahedral zones in your mesh that have orthogonal motions prescribed on them. FOAM currently does not have the capability to handle run-time manipulation on tet-meshes (Although I'll be working on that pretty soon).
Some of my work on FOAM currently involves topological changes to 2D triangular meshes with strong deformations, surface tracking and mesh-motion. If this applies to you, I can probably help you out.

Hope this helps.
__________________
Sandeep Menon
University of Massachusetts Amherst
https://github.com/smenon
deepsterblue is offline   Reply With Quote

Old   April 26, 2008, 16:36
Default Dear Sandeep, thank you for th
  #3
New Member
 
Mark Eaton
Join Date: Mar 2009
Posts: 4
Rep Power: 17
meaton is on a distinguished road
Dear Sandeep, thank you for the help.

Could you tell me which classes for changing the topology are available? I need something to get started, because there is hardly any documentation except for the very few examples in the user guide.

As you say that OpenFOAM has no tet-mesh manipulation capabilities, I'm interested to work on this after understanding the general concepts a bit better. Do any non-commercial tools (independent from OpenFOAM) exist for this kind of mesh manipulation that could be integrated in FOAM?

Mark
meaton is offline   Reply With Quote

Old   April 26, 2008, 17:11
Default Mark, If you can get a copy
  #4
Senior Member
 
Sandeep Menon
Join Date: Mar 2009
Location: Amherst, MA
Posts: 403
Rep Power: 25
deepsterblue will become famous soon enough
Mark,

If you can get a copy of the dev version from the SVN ( http://openfoam-extend.svn.sourceforge.net/ ):
and compile it successfully, you'll find the classes under:

/OpenFOAM/OpenFOAM-1.4.1-dev/src/topoChangerFvMesh

These classes are derived from classes available in:

/OpenFOAM/OpenFOAM-1.4.1-dev/src/dynamicMesh

Take a look at the tutorials (under icoDyMFoam) for a better idea on setting up dynamic mesh cases.

I'm not aware of any non-commercial tools that perform what you're looking for. If you need run-time manipulation of tet meshes, it's time to roll-up your sleeves and start writing some code.

Hope this helps.
__________________
Sandeep Menon
University of Massachusetts Amherst
https://github.com/smenon
deepsterblue is offline   Reply With Quote

Old   April 26, 2008, 17:43
Default Thank you, I'll study the movi
  #5
New Member
 
Mark Eaton
Join Date: Mar 2009
Posts: 4
Rep Power: 17
meaton is on a distinguished road
Thank you, I'll study the movingCone tutorial in detail.

>If you need run-time manipulation of tet meshes,
>it's time to roll-up your sleeves and start writing
>some code.

Why can for example the routines within gmsh not be used to do run-time mesh updates? Can you mention a few keywords or references that I can search for to learn about state of the art algorithms? Then I will dive into implementing these for OpenFOAM.

Mark
meaton is offline   Reply With Quote

Old   April 26, 2008, 19:15
Default Because those routines are int
  #6
Senior Member
 
Sandeep Menon
Join Date: Mar 2009
Location: Amherst, MA
Posts: 403
Rep Power: 25
deepsterblue will become famous soon enough
Because those routines are intended for mesh generation. Regenerating the entire mesh is not only hard to automate, but a huge waste of computational time. Local re-meshing is a much better alternative.

As for algorithms, here's a popular one:
Two Discrete Optimization Algorithms for the Topological Improvement of Tetrahedral Meshes.
by Jonathan Richard Shewchuk
__________________
Sandeep Menon
University of Massachusetts Amherst
https://github.com/smenon
deepsterblue is offline   Reply With Quote

Old   April 27, 2008, 07:56
Default >As for algorithms, here's a p
  #7
New Member
 
Mark Eaton
Join Date: Mar 2009
Posts: 4
Rep Power: 17
meaton is on a distinguished road
>As for algorithms, here's a popular one:
>Two Discrete Optimization Algorithms for the Topological Improvement of Tetrahedral Meshes. by Jonathan Richard Shewchuk

So you suggest these algorithms to be promising candidates for implementation in FOAM? Do you know of any existing implementations (for validation or reusing code)?

If someone else could confirm (or if noone contradicts) this paper to be a good starting point, I'll try my best after getting familiar with OpenFOAM.

Regards
Mark
meaton is offline   Reply With Quote

Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Change mesh topology during transient run Alex77 CFX 2 October 31, 2007 02:35
Changing Mesh Velocity Anil CFX 0 October 29, 2007 09:47
changes in mesh topology Neil Main CFD Forum 0 March 27, 2003 10:44
Unstructured mesh moving boundary Ivano Di Domenico Main CFD Forum 3 February 20, 2001 11:29
Unstructured moving grid M.A.Leschziner Main CFD Forum 0 July 15, 1999 07:25


All times are GMT -4. The time now is 09:56.