Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2017-10-04 15:47:42

sb1966
Member
Registered: 2015-03-16
Posts: 48

[SOLVED] Fatal Error while Launching Paravis

Dear All,
I tried to run a simple linear elastic study of a box fixed at the bottom and loaded by a pressure load (PRES_REP) at the top face. Everything runs nicely up to Aster Study. But when I try to launch Paravis, I get error message (the screen-shot attached). I have LGPL-3 package installed. I also tried after installing the latest package (2017.0.1-LGPL), but same proble persists.

It is little awkward, as Paravis was working fine earlier, though I could not visualise the SIEF_ELGA and SIEF_ELNO fields.

I am attaching my .hdf, .comm , .rmed file and the screen shot in an archive for your advice.

Thanks and regards.

Sukumar

Last edited by sb1966 (2017-10-08 16:15:13)


Attachments:
StudyBox.zip, Size: 399.65 KiB, Downloads: 13

Offline

#2 2017-10-05 04:56:15

GXA_William
Member
Registered: 2016-04-30
Posts: 13

Re: [SOLVED] Fatal Error while Launching Paravis

Dear Sukumar,

I can open your HDF and RMED files without issue. 

One suggestion I can make from personal experience is that using Paravis from within the same study is usually a bad idea.  One alternate path to take is to open a second SMECA and just run Paravis, with the first SMECA running the Geom/Mesh/Aster side of things.

What I do is a little different.  My solution to avoid Sigsev errors is to run Paraview separately using using a terminal and runSession (e.g., /opt/appli_V2016/runSession paraview).  In fact I rarely run Aster within the same SMECA and instead typically have a study for Geom/Meshing, then export the mesh as a Med file which I subsequently import into the Aster (or AsterStudy) run, then I run a third study just for Paravis/Paraview.

Not sure why the Sigsev errors occur but best guess is that if anything changes or gets corrupted in the chain between the Geom-Mesh-Aster links the result is a Sigsev error.

Kind regards,

-William

Offline

#3 2017-10-05 08:05:43

sb1966
Member
Registered: 2015-03-16
Posts: 48

Re: [SOLVED] Fatal Error while Launching Paravis

Dear William,

GXA_William wrote:

One suggestion I can make from personal experience is that using Paravis from within the same study is usually a bad idea.  One alternate path to take is to open a second SMECA and just run Paravis, with the first SMECA running the Geom/Mesh/Aster side of things

Thanks a lot for your suggestion and quick response. I will follow your advice and post  my experience here.

With best regards.

Sukumar

Offline

#4 2017-10-05 17:00:01

sb1966
Member
Registered: 2015-03-16
Posts: 48

Re: [SOLVED] Fatal Error while Launching Paravis

Dear William,

Following your advice, I am able to visualize the result file (.rmed) now. Thanks a lot for your support.

One more question. Though my command file contains IMPR_RESU command for printing SIEF_ELGA and SIEM_ELNO fields,:

...
stress = CALC_CHAMP(
    CONTRAINTE=('SIEF_ELGA', 'SIEF_ELNO'),
    EXCIT=_F(CHARGE=load),
    MODELE=model,
    RESULTAT=reslin
)

IMPR_RESU(
    FORMAT='MED',
    RESU=(_F(MAILLAGE=mesh, RESULTAT=reslin), _F(RESULTAT=stress)),
    UNITE=80
)

I cannot see the stress fields in my .rmed file. Would you kindly suggest me how to visualise them? In earlier versions (say SMeca 2016), I used to get them visible.

Thanking you.
With best regards.

Sukumar

Offline

#5 2017-10-08 16:36:36

sb1966
Member
Registered: 2015-03-16
Posts: 48

Re: [SOLVED] Fatal Error while Launching Paravis

Dear All,


I am thankful to dbpatankar for his support in form of a reply to my personal communications . I am posting the following which is based on his reply, hoping that it may be useful for some one like me in future:

1) One can use the Filters > Mechanics > 'ELNO Field to surface' OR 'ELGA Field to surface' (as required).

2) Alternatively one can also calculate the equivalent field(Von-Mises, Tresca etc) from available stress fields, by using CALC_CHAMP and CRITERES as SIEQ_ELNO and/or SIEQ_NOEU.

...
stress = CALC_CHAMP(
    reuse = stress,
    RESULTAT=stress,
    CRITERES=('SIEQ_ELNO', 'SIEQ_NOEU'),
    )
...

These fields contain the Von-Mises, Tresca and Principal stresses and the components of the Stress tensor for their  individual viewing. This is a better option in  design problems for checking permissible/yield stresses etc.


Thanks to dbpatankar and GXA_William for their supports. I find this a really great community promptly helping each other.

With regards

Sukumar

Last edited by sb1966 (2017-10-08 16:46:27)

Offline

#6 2017-10-10 22:26:48

Thomas DE SOZA
Guru
From: EDF
Registered: 2007-11-23
Posts: 2,851

Re: [SOLVED] Fatal Error while Launching Paravis

Hi,

sb1966 wrote:

Everything runs nicely up to Aster Study. But when I try to launch Paravis, I get error message (the screen-shot attached). I have LGPL-3 package installed. I also tried after installing the latest package (2017.0.1-LGPL), but same proble persists.

Be advised that there is a known bug that may explain the difficulties you had. It will be fixed in a subsequent 2017.0.x release.

It involves the use of the AsterStudy feature "Open in ParaViS" to directly open a MED result file in the ParaViS module. In some cases, using that once may corrupt the current study (HDF file) and make it impossible to activate ParaViS again or even some panels of AsterStudy using the graphical view.

TdS

Offline

#7 2017-10-11 18:21:59

sb1966
Member
Registered: 2015-03-16
Posts: 48

Re: [SOLVED] Fatal Error while Launching Paravis

Dear Thomas,
Thanks a lot for your clarification and advice.

Tds wrote:

It involves the use of the AsterStudy feature "Open in ParaViS" to directly open a MED result file in the ParaViS module. In some cases, using that once may corrupt the current study (HDF file) and make it impossible to activate ParaViS again or even some panels of AsterStudy using the graphical view.

In this regard, I request you to kindly clarify if the advice given by GXA_William:

GXA_William wrote:

One suggestion I can make from personal experience is that using Paravis from within the same study is usually a bad idea.  One alternate path to take is to open a second SMECA and just run Paravis, with the first SMECA running the Geom/Mesh/Aster side of things./

is safe enough to prevent corruption of the hdf file and/or Aster Study GUI. If this is not advisable, what should we do to prevent such problems.

Thanks a lot.
Best regards.

Sukumar

Offline

#8 2017-10-13 14:16:38

Thomas DE SOZA
Guru
From: EDF
Registered: 2007-11-23
Posts: 2,851

Re: [SOLVED] Fatal Error while Launching Paravis

sb1966 wrote:

In this regard, I request you to kindly clarify if the advice given by GXA_William:

is safe enough to prevent corruption of the hdf file and/or Aster Study GUI. If this is not advisable, what should we do to prevent such problems.

Short answer: yes.

Long answer: as long as you don't use "Open in ParaViS" feature and that you're using 2017.0.1, corruption of the HDF should not occur. There are still some major bugs related to SALOME to be fixed in a subsequent release though and you should regularly save your study, especially after having run an analysis.

TdS

Offline

#9 2017-10-16 20:11:02

sb1966
Member
Registered: 2015-03-16
Posts: 48

Re: [SOLVED] Fatal Error while Launching Paravis

Dear Thomas,

TdS wrote:

Long answer: as long as you don't use "Open in ParaViS" feature and that you're using 2017.0.1, corruption of the HDF should not occur. There are still some major bugs related to SALOME to be fixed in a subsequent release though and you should regularly save your study, especially after having run an analysis.

Thanks a lot for your clarifications. Sorry for my delayed response.

Best Regards.

Sukumar

Offline

Board footer