[philippose]

Hello,

I was testing out OpenFOAM ver.1.4 a few days ago, and found that when I try to visualise a decomposed mesh with paraFoam, it throws up an error saying that the controlDict file could not be found.

In the earlier versions of OpenFOAM, I could go into the case folder of a decomposed case, and use (for example):

paraFoam . processor0

or

paraFoam . processor1

and it would open ParaView and display that part of the mesh which belongs to the particular processor.

In the paraFoam file of version 1.4, the script looks for a valid "controlDict" or "fvSolution" or "fvSchemes" file before moving ahead... since the "processor<x>" folder has no "system" folder within, paraFoam exits with an error.

Was this modification made with the idea of introducing (in the near future) the capability of being able to visualise decomposed cases in ParaView without first having to reconstruct them, or was it just an oversight?

Currently I use the ParaFoam script from one of the earlier versions, and things work fine....

Have a nice day!

Philippose

[mattijs]

> or was it just an oversight

It was just an oversight. Remove the check or use the old script.

[olesen]

Attached is a version that does the correct checking paraFoam for serial and decomposed cases.
BTW: you can simply issue 'paraFoam' to view the results in the current working directory or 'paraFoam DIR' to specify a particular directory (see 'paraFoam -h')

[philippose]

Hi Mark,

Thanks a lot for the modified paraFoam script :-)! Works like a charm!

Now all that is remaining, is for someone to take the time out to modify the vtkFoam reader to enable a decomposed mesh to be visualised (in its entirety) without having to reconstruct it.

Have a nice day!

Philippose

[olesen]

Hi Philippose,

From what I've gathered, paraview 3 might make it easier to write a parallelized reader.