I 'backported' the Foam2.2 foamToEnsight converter to Foam2.1. It solves a few problems with complex cell shapes. Contact me if you are interested.

Sorry to post here, I couldn't create a new thread (??). Has anyone written a openFOAM to Tecplot converter?

[hani]

Does version 1.2 support postprocessing in Ensight76?

I had problems when converting from version 1.0.2 to 1.2. The OpenFoam option disappeared in Ensight76.

When I added the following to my .cshrc file it worked again:
source $HOME/.OpenFOAM-1.0.2/apps/ensightFoam/cshrc
This is however from the 1.0.2-version. Should I have to do this or do you think I did something else wrong?

[niklas]

Hola senore,

Should work just as good as before.

Check the CEI_HOME variable in cshrc
and make sure you have source .../ensightFoam/cshrc in your ~/.OpenFOAM-1.2/cshrc.

[hani]

CEI_HOME is not set in either of the following:
~/.cshrc
$HOME/OpenFOAM/OpenFOAM-1.2/.OpenFOAM-1.2/cshrc

I got no ~/.OpenFOAM-1.2 directory when installing 1.2

[niklas]

You have forgotten to
cp -r OpenFOAM/OpenFOAM-1.2/.OpenFOAM-1.2 ~/.

dont forget to also change you .cshrc accordingly

N

[hani]

I can't find that 'cp -r ...' in the installation instruction. The README file just tells me where to find the .OpenFOAM-1.2 directory, not that it should be copied. I am also instructed to use the original path to the cshrc file i.e.
$HOME/OpenFOAM/OpenFOAM-1.2/.OpenFOAM-1.2/cshrc
when sourcing in my .cshrc file. I'm a bit confused.

I guess that I also have to un-comment the line containing
SOURCE $HOME/$FOAM_DOT_DIR/apps/ensightFoam/cshrc
in
$HOME/OpenFOAM/OpenFOAM-1.2/.OpenFOAM-1.2/cshrc

[niklas]

Hmmm, Havent read the readme.
seems like it has changed

anyway, it doesnt matter where you keep it, in $HOME or $HOME/OpenFOAM/OpenFOAM-1.2 as long as you source it and
yes, uncomment the source-line.

N

[hani]

Then I think that the problem is likely to be in the commented line
SOURCE $HOME/$FOAM_DOT_DIR/apps/ensightFoam/cshrc
If it had been something like
$HOME/OpenFOAM/OpenFOAM-1.2/.OpenFOAM-1.2/cshrc
the file would have been found and I would just have to uncomment the line, without copying the whole $HOME/OpenFOAM/OpenFOAM-1.2/ directory

[stefanke]

I have a strange problem with the foamToEnsight converter.
When I use the converter for a specific timestep all is fine. Performing the converter over all timesteps for a transient problem the following error occurs trying to open the case file:

[read casefile]
Error> verify_increasing_times() in timesets.c:559
of EnSight Server 8.0.7(j):
Error sequencing data.
Descending/duplicate ordered time values: t[19:18]={0,60}
Error> readensight_casefile() in resu.c:934
of EnSight Server 8.0.7(j):
Error sequencing data.
initializing time-set spans using case data format

OpenFoam 1.2, Ensight 8.0.7 Gold

any ideas how to get rid of this problem?

[mattijs]

This is with any OpenFOAM case? Non-parallel run?

Can you have a look in the generated case file and compare it to one which does work? Can you figure out the fix (source is in postProcessing/dataConversion/foamToEnsight) and post it? We do not have Ensight here so cannot test.

[stefanke]

Hi Mattijs,

I have fixed the sequencing data error. But there are another problems:

no1: moving meshes
mesh.moving() always returns 0 but it is a moving mesh (engineFoam). This results in missing ensight mesh files for the different time steps!

no2: lagragian
If there are no particles at the begining, the converter aborts because in the lagrangian directroy there is only a points file. The problem is that the converter expect (in case of lagrangian=true) that all lagrangian data are available.


I hope I can fix the bugs in the foamToEnsight converter especially for the problem no1!

[mattijs]

1: the mesh movement is/should be detected in moveMesh.H.

If it detects a "points" file it does a mesh.movePoints(...) which sets the 'moving()' flag. Maybe it doesn't find the points file?


2: Guess that never was foreseen, to have lagrangian without data.

[stefanke]

1. the points dict was not found!
fixed with mesh.Name() -> polyMesh::meshSubDir

2. The problem occurs i.e. in engine applications because at the begining there are no particles but after SOI there are!

There are some minor problems with moving meshes but I will try to fix them.

[stefanke]

Mattijes,

I am on the way ....

one problem:
------------
I have to change the geometry file name (in the ensight case file) in case of moving meshes. But "a priori" I don't know if there is a moving mesh (this information is first available at the second time step) so how to write the right filename to the file?

without moving mesh: xxxxxx.000.mesh
with moving mesh: xxxxxx.***.mesh

The only solution coming to my mind is to manipulate the file at the end of foamToEnsight.C, after the case file was written (stream manipluation: substitute xxxxxx.000.mesh -> xxxxx.***.mesh if mesh.moving() == true). But I do not know how to this?

[mattijs]

Neither do I. Sounds a bit scary what you're trying to do.

In my opinion a better way is to do a scan first of the time directories to see if it is a moving mesh case.

[stefanke]

the fixed version of foamToEnsight:
http://n.ethz.ch/student/stefanke/do...foamToEnsight/

fixed bugs:
-----------
- some minor bugs when dealing with moving meshes
- wrong time step ordering in the case file
- wrong position of "measured ...***.lagrangian"

[mattijs]

Thanks. Will put it into next release.

[stefanke]

Mattijs,

there is one file which was missing!
http://n.ethz.ch/student/stefanke/download/OpenFoam/src/foamToEnsight/ensightCas eTail.H

[stefanke]

Hi Mattijs,

there still one bug (perhaps introduced by myself) in the foamToEnsight (OF v.1.3) converter.

Bug info: wrong position of the lagrangian file definition in the case file. Ensight complains about a unkown keyword in the case file.

To fix this bug please correct the ensightCaseHeader.H to:

if (Pstream::master())
{
ensightCaseFile << "FORMAT" << nl;
ensightCaseFile << "type: ensight gold" << nl << nl;
}

I hope the converter will works now without any problems!