CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   IcoTopoFoam update (http://www.cfd-online.com/Forums/openfoam-solving/60631-icotopofoam-update.html)

hjasak March 28, 2005 21:32

Hi Henry, Attached is a pre
 
Hi Henry,

Attached is a pretty major update for icoTopoFoam + another application (moveTopoMesh) which moves and morphs my meshes without updating the topology.

There are 3 test cases for the stuff which should allow you to test the installation - all are set up and ready to run. Here are the most important changes:

topoFvMeshes library
- there is a virtual base class for topoFvMeshes, with a bunch of meshes derived from it. All derived meshes are fully self-contained, i.e. there is no need to run the set utility before you run the case.
- in order to run the case, run blockMesh first and then just start the case. The mesh will check whether the mesh modifiers and zones have been added and if not it will add them automatically. Moving and morphing then happens as expected
- the mesh setup is controlled from the meshMotionDict, with mesh selection + parameters for each mesh type. I have also changed from motionProperties to meshMotionDict for all other apps in FOAM to have a consistent setup + changed all FoamX files - you will get those in the next merge.

apps:
- both apps now talk to the topoFvMeshes through the virtual interface with run-time selection
- moveTopoMesh just moves the mesh
- icoTopoFoam solves icoFoam on it

other stuff:
- I am not too happy with topoFvMesh being derived from fvMesh rather than polyMesh but our class hierarchy is a bit screwed up here to favour the FV discretisation. Haven't got a good solution though...
- I would like icoTopoFoam to replace the one in the release - it has removed the nasty compilation switches and the slider case is of no interest.
- moveTopoMesh should go together with moveMesh into utilities.
- I am not sure where the library should live as some of it is case-related (you write a derived topoMesh for a class of cases) - any ideas are welcome.

There may be other changes required, but this should compile out of box. There is also a set of library bug fixes a bit too complex for me to keep track, so they will arrive after the next merge.

Shout if there's trouble,

Hrv


All times are GMT -4. The time now is 13:08.