Name of output file names - where do they come about? [Solved]

[teepeeaye]

Hi When I start ParaView after running a case, I get the following error from Paraview:

Code: Select all

cannot open data file "case_t0002.vtu"

The actual output files are named:

Code: Select all

case0001.pvtu
case0002par0001.vtu
case0001par0001.vtu

This leads me to believing that there is "something" messed up in the handling of file naming.

• I would expect the naming to have the case numer (0001) following the text "case" and the partition numbers (0001 and 0002) following the text "par" but naming is different.

• The name used when calling Paraview seems to be what the name would be if it was not partitioned - but it is.

• The appearance of "_t" in the file name as generated by ElmerGUI differs from the filenames generated by the output module. but nether of them seems to be correct.

I am on Ubuntu Xenial (16) using the ppa ElmerGUI and the ElmerSolver of my own 8.4 build. If I were to try to align the file names - which sourcefiles should I look in?

[raback]

Hi

I think the problem is in inconsistency of the versions for ElmerSolver and ElmerGUI. Any particular reason you have so old build for ElmerSolver? This convention has been there for several years.

We always recommend using the current "devel" branch.

-Peter

[teepeeaye]

Thank you for the response. 8.4 is the latest build from the packet repository for my system. I have tried to build the GUI from scratch but no success so far.
The resulting ElmerGUI is missing the object browser on the left (Below screenshot is from my working ElmerGUI):

Screenshot at 2023-06-18 12-13-46.png (41.41 KiB) Viewed 700 times

This unfortunately inhibits using the GUI to compose sif-files.

My workaround for the naming is to open the sif editor, add _t to the output file basename and then start the calculations from the sif editor. I realise this is not of common interest and the workaround is OK.