Paraview/Elmer communication

Post processing utility for Elmer
Post Reply
dams2one
Posts: 4
Joined: 02 Feb 2017, 20:18
Antispam: Yes

Paraview/Elmer communication

Post by dams2one » 22 Feb 2017, 19:28

Hello dear gents,

I'm facing an issue as I can only get the geometry but no value as VM, using Elmer 8.2 (Win) + Paraview 5.3 (tested with 5.1.2 64bits; 5.2 32 bits).

In the [Information] tag I can see :
in [Statistics]
Type: Unstructured Grid
Number of Cells: 58269
Number of Points: 11247
Memory: 5.4 MB

and

in [Data Arrays],
"vonmises double [0,0]"

I've got float64 in my ptvu.

Do I missed some parameters,switches...

Thank you in advance for your help

Damien

mzenker
Posts: 1807
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Paraview/Elmer communication

Post by mzenker » 23 Feb 2017, 10:14

Hi,

difficult to say what is the problem without more information.
Could you post your sif file, and the output file (or a test file) for ParaView exposing the problem?

Matthias

dams2one
Posts: 4
Joined: 02 Feb 2017, 20:18
Antispam: Yes

Re: Paraview/Elmer communication

Post by dams2one » 23 Feb 2017, 11:39

Hi Mattias,

Thank you for such a fast answer. I'm not sure :? what you meant with test file for Paraview, by the way I tried to attached the Test.out file that comes from ELMER without any success as .out file aren't allowed.
Let me know if you need some more data.

Best regards

Damien
case.sif
(2.42 KiB) Downloaded 92 times

mzenker
Posts: 1807
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Paraview/Elmer communication

Post by mzenker » 23 Feb 2017, 12:13

Hi,

I don't see an obvious problem in your sif file.
Could you post your output file myfile.vtu? Or, if it is confidential or too large (even zipped), could you build a small test case reproducing the problem and post the vtu file?

Matthias

raback
Site Admin
Posts: 3328
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Paraview/Elmer communication

Post by raback » 23 Feb 2017, 13:42

Hi

You seem to save the results prior to computing the solution. That might be the problem. Control the solver numbering or just set "Exec Solver = After All", for example.

-Peter

mzenker
Posts: 1807
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Paraview/Elmer communication

Post by mzenker » 23 Feb 2017, 15:28

Oops, I didn't pay attention to that one.
However, in the sif file it says:

Code: Select all

Solver 2
  Equation = Linear elasticity
  Exec Solver = Always
  ...
End

Solver 1
  Equation = Result Output
  Exec Solver = Always
  ...
End

Equation 1
  ...
  Active Solvers(2) = 2 1
End
Doesn't that mean that the Result Output would be called after Linear Elasticity?
In any case, Exec Solver = After All is appropriate for Steady State, of course.
Exec Solver = Always would mean that each iteration step is saved.
At least as far as I know...

Matthias

dams2one
Posts: 4
Joined: 02 Feb 2017, 20:18
Antispam: Yes

Re: Paraview/Elmer communication

Post by dams2one » 23 Feb 2017, 20:18

Hi guys,

Once again, thank you for your support. As I'm newbe with Elmer, sometime I miss some parameters.
Regarding GUI, the option "After All" doesn't exist. Obviously, I can change it through .sif file.
Have got options : Always Before Simulation, After Simulation, Before timestep, After timestep & Never.
Guess no so obvious. :)
Then, by checking "After timestep", I suppose to define "After All", don't I?

Damien

dams2one
Posts: 4
Joined: 02 Feb 2017, 20:18
Antispam: Yes

Re: Paraview/Elmer communication

Post by dams2one » 23 Feb 2017, 20:22

Seems to be solved.
I changed Result Output => Exec Solver = After Timestep and it does work well.

Thank you for your help. I'll try to be a bit more carefull now.

Regards

Damien

mzenker
Posts: 1807
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Paraview/Elmer communication

Post by mzenker » 23 Feb 2017, 21:02

Hi,

I would spontaneuosly say "After Simulation" for Steady State. But "After Timestep" seems to do the same since it works for you. :)

Matthias

raback
Site Admin
Posts: 3328
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Paraview/Elmer communication

Post by raback » 27 Feb 2017, 00:49

Hi

"Always" refers to to the main iteration loop wherein the solvers are executed according to their number. Both "after timestep" and "after all" are performed after the main iteration loop. Their order again is competing with other solvers in the same "exec solver" slot.

-Peter

Post Reply