Dear support and users,
I activated the particle tracking module, as well as the post treatment in displacement mode for my simulation. The problem is when I want to open deplacement.CASE, my post processor returns an error. Apparently, it can not find the value of “Time set” and can not open the deplacement.geo**** files.
I have seen on the “Issue tracker” page a bug related to the creation of the ensight file in the RESU directory, so I downloaded the attached patch and put the files in ncs-2.0.0-rc1/src/lagr, but I I got exactly the same problem.
Can someone help me?
I attached the deplacement.CASE file if it can help.
Best Regards,
Stéphane
deplacement.CASE.txt (593 Bytes)
Hello Stéphane,
Could you try with the attached (and slightly modified) deplacement.CASE ? If it doesn’t work, could you post the output of a “ls -l” command in a terminal, so that we can list the different files ?
By the way, which post-processing tool are you using (EnSight, ParaView, …) and which version ?
David
deplacement.CASE.txt (601 Bytes)
Ok… it seems you found a weakness in our post-processing output. Could you test the attached deplacement.CASE (forget the previous one…) ?
I’ve looked further and we have obviously forgotten to specify the time set number. As we never encountered a problem with EnSight, I assume that it uses a default value in case it is not specified. I will fix that for next release (2.0-rc2), probably released early next week.
In the meantime, I’m attaching a possible fix.
David
enswaf.f90 (23 KB)
deplacement.CASE.txt (595 Bytes)
Hello,
I tested the attached deplacement.CASE, it doesn’t work neither. Actually it’s worst, Paraview immediately crashes without returning me any error message, when opening the deplacement.CASE.
And the enswaf.f90 file doesn’t seem to solve the problem, I have still the same problem.
But anyway, thank you for your help.
Stéphane
Which version of ParaView are you using ? I had some crashes when trying to use the animation mode for visualizing displacements (with version 3.6.2 IIRC). But, I was still able to load the case, the crash only happened later on.
Oh sorry, I did not see the first message.
Here is the file list.
I use Paraview 3.6.2 for the post-processing.
Stéphane
No problem! It seems everything is present in the directory…
Could you try latest version of ParaView (namely 3.8) ? It solved my issue with animations.
Ok, but even with the version 3.6.2, I can’t even load the case, it crashes immediately.
Anyway I will try with Paraview 3.8.
I’ve just tried with the newest version of Paraview, I have still the same issue (it crashes when loading the CASE file).
It seems your result files are not too big, could you make a “tar.gz” or similar and post it so that we can test with EnSight on our side? I suspect that the EnSight 6 reader (used for particles output) is not correctly handled by ParaView…
Yes my result files are not too big, I did the simulation only in a few iterations.
Here is the tar.gz file.
Thank you for helping me.
LAGR-06171326-tar.gz (9.19 KB)
Ok… We had several differences with the EnSight 6 format in the way we were writing the particles files (both displacement and trajectories).
Here attached are two files for fixing the bug (hopefully). Could you test them ? You may need to compile a newer version of ParaView (namely the development version 3.9) as they improved the EnSight reader; to be precise, I couldn’t correctly visualize the particles with version 3.8.0…
Thanks again for the reporting and testing!
David
fix-postpro-lagr.tgz (7.28 KB)
I have just tried with Paraview 3.8, and I have still the same problem: I can’t open the file.
I will try with the version 3.9 later on, when I will find it.
Stéphane.
Have you tried to use the fixes I suggest? If so, could you attach the result files for the Lagrangian stuff so that I can check on my computer (with different versions of EnSight and ParaView).
As regards ParaView 3.9, you won’t find any binaries snapshot (AFAIK). Instead, you’ll have to clone the git repository and to compile the code by yourself (a long and disk-memory-consuming task…).
David
Yes I tried with the fixes, here are my results.
Thank you for the information about Paraview 3.9, unfortunately I won’t have time to do that.
I forgot to ask you: is that possible to simulate multiphasic flows with Code_Saturne (for example water + air flow simulation)?
Furthermore, I noticed I can only use 1 core when using “particles + droplets tracking”, is that normal?
Stéphane
LAGR-06230903-tar.gz (39.2 KB)
Hi,
It seems you haven’t properly used the fixes (looking at the results files). The easiest way to use them is to drop the two files in the SRC directory of your test-case, the other way being to drop them in the ncs-2.0.0-rc1/src/lagr directory and to recompile/reinstall the code.
Concerning your two other questions: - Code_Saturne cannot simulate multiphasic “Euler/Euler” flows (though there is a NPHAS variable in the code…). We have a dedicated code for that, sort of “cousin” of Code_Saturne as they share much of the code, but it is not open-source. - Parallel run for “particles tracking” is under development/testing and will available in a 2.1 release of Code_Saturne (probably before the end of the 2010).
David
Hi,
Thank you for your answers. You are right, I did not properly used the fixes, sorry about that. I tried again and it works fine!
Thanks again,
Stéphane
