Error messages in ParaView when reading in Transient OpenFoam test case

Hi,

I am reading in a transient openFoam test case using a case.foam file and VTK format on a Windows 10 system. I am editing an earlier post about this issue since I initially believed ParaView 4.4 did not have issues. But, I was mistaken. I have been able to determine that if I write out data from OpenFoam at time steps which involve a scientific format (in this case at time steps of 1e-8 s, 2e-8 s, etc…), ParaView (4.4. and 5.6.x) generate an error message, some of which are attached to this post. This is a reproducible error. For the same test case, if I read in data from OpenFoam written out at time points expressed in decimal format, this error does not appear.

I plan to post it too on the Development forum in the hope that someone on the development team can confirm and/or propose a workaround.
Thank you for your consideration.


Please share your dataset.

Hi,

Thank you for looking into this! I am attaching a few tar archives containing example data from the pitzDaily tutorial of OpenFoam (v1806) exemplifying the issue. The archives contain all the necessary dictionary files to re-run the calculations. I’ve modified controlDict to ensure only two time steps are written out. There are four archives:

  1. pitzdaily.tar.gz
  2. pitzdaily_exp.tar.gz - same as 1, except results are written out at very early time steps such that the data folders are expressed in using scientific notation (5e-06/ and 1e-05/).
  3. pitzdaily_par.tar.gz - same as 1, except the calculation is run in parallel using mpi.
  4. pitzDaily_reconstr.tar.gz - same as 3, but results have been reconstructed using foamToVTK.

Except for archive #3, all examples have a VTK folder too. I have read in data using two approaches:
a) Read in case.foam file in the respective folder. Then Apply.
b) Read in .vtk group file in the VTK folder of the respective example.

I see error pop-up windows appear when I try to read in the data, depending on whether I utilize approach a) or b) listed above using ParaView 5.6.0 on Windows 7, and ParaView 5.6.0 (and 5.6.1) on Windows 10. I understand from a work colleague that no error pop-up windows appeared when he tried loading the data using a Mac with macOS Mojave (10.14.5) and ParaView 5.6.0. The results I obtained are listed below:

Folder Outcome Case.foam Outcome VTK
pitzDaily Error after data Apply It works!
pitzDaily_par Error after data Apply N/A
pitzDaily_reconstr Error after data Apply It works!
pitzDaily_exp Error after data Apply Error after data Apply

Please note how there is also an error with the VTK data input when the results are written out to folders in expoential format (pitzDaily_exp).

Here are the links to the archives.


https://www.dropbox.com/s/p0jsdt2uo9b9290/pitzDaily_reconstr.tar.gz?dl=0
https://www.dropbox.com/s/n724ygf8ll84h9y/pitzdaily.tar.gz?dl=0
https://www.dropbox.com/s/std0ppss971hln6/pitzdaily_exp.tar.gz?dl=0

Many thanks for your help.

Maziar.

I was unable to reproduce your issues with the provided datasets. Could please explain, step by step, how to reproduce the issue ?

Hi,

I am assuming you’ve unpacked the tar archives and the folders are available. I am using a Windows 10 system and ParaView 5.6.1.

Please start an instance of ParaView in the folder containing the “case.foam” file. Load the case.foam file and “Apply” for all test cases except for pitzdaily_par. In the case of pitzdaily_par, which is a parallel run, set the data type to “Decomposed” in place of the default “Reconstructed” before you “Apply”.

The alternative strategy to try is to load the VTK files located in the VTK folder, and then “Apply”. The pitzdaily_par folder does not have a VTK folder.

You should see the errors based on the results I showed in my previous response. For example, where I wrote:

FolderOutcome Case.foam Outcome VTK
pitzDaily Error after data Apply It works!
pitzDaily_exp Error after data Apply Error after data Apply

This means that for the folder “pitzDaily”, if I load the data using the case.foam approach presented above, I get an error. If I load the VTK files, it works. In the case of the folder pitzDaily_exp, I get errors regardless of the approach.

Please let me know if this is clear. I would really like to know what the issue is and how it can be resolved. Thank you!

This is exactly what I did, without any issue.

I’m using Windows 10 with an english language. What about you ?

Yes. I am using an English language version. How to proceed? Can the nature of the error message provide clarification as to what aspect of my setup is the issue? I have attached the error messages for the case of the pitzDaily run.

Please try to reproduce on another computer.

I only have employer-issued hardware running Windows. I have reproduced these problems with hardware running Windows 7 and various versions of ParaView (4.4, 5.2, 5.6, 5.6.1). My colleague, who has an employer-issued Mac, does not have this problem as I reported in this thread. I am not sure what else to test within my means.

Does the error message provide no hint where the problem could reside?

Not really no.

Please try to use our release instead of your own superbuild.
https://www.paraview.org/download/

If by “superbuild” you mean a version I may have compiled from source, I did not.
I have downloaded the “.exe” which I subsequently installed using the install wizard.

sorry, about that. You are right.

You will have to find a way to reproduce your error on another computer.