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

You are not logged in.

#1 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » Today 20:17:29

Hi piyushverma,

When you start the x server, did you check "Disable Access Control" ?

#2 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 21:47:29

Unfortunately it will not be possible for you to use gpu, at least not until you switch to a newer version of wsl2 including wslg.

Microsoft announced gpu compatibility for different cards manufacturers through wsl2, and singularity containers support nvidia gpu acceleration.

So theoretically if you have a nvidia cuda card and update wsl2, you might be able to activate gpu through wsl2 and then through singularity.

Again, this is how it should work on paper, and Microsoft is indeed actively developing and improving the performance of WSL. Running a 3d graphical application was really not imaginable a few years back on WSL so we are already living the revolution.

Best regards

#3 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 21:09:51

Ok it seems hence that your version of wsl2 does not include a native display driver, called wslg.

But there is a solution for your version consisting of installing an x server on windows.

Search for vcxsrv in google, this is the lightest server and it's easy to setup.

Once you install vcxsrv, run Xlaunch in windows and choose the default options except "Disable Access Control" (page 3 I think) which must be checked (access control disabled), otherwise the internal windows firewall might block your wsl.

Now go back to your wsl, and run


(Yes the distro already includes a specific launcher for that).

Finally please note that these steps which may seem complicated are not at all due to salome_meca. Wslg which will be rolled in future updates of wsl2 (I already have it and it seems to be out for a while now) will make user skip the last step of installing xserver in windows.

Best of luck

#4 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 20:50:26

Can you tell me what you get if you launch


In the wsl system, do you see a clock appear ?

#5 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 19:12:21

Maybe you don't have write access to the root of C: ?

Try to create a folder, say C:\Data and use that...

In my case I had a 2nd disk in F:\ where I had write access.

stefano89 wrote:

No problem HB.

Now I receive this:

C:\Windows\system32>wsl --import SalomeMeca2021 C:\ C:\Users\Calcolo\Downloads\smeca-lgpl-wsl2-opensuse.tar
Accesso negato.

then no access. I suppose that is "sudo" user problem...any idea?

All the best

#7 Re: Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 19:02:12

Hi Stefano,

Are you sure that you have WSL2 ?

#8 Salome-Meca installation » Salome Meca 2021 on WSL2 » 2021-11-26 14:32:05

Replies: 15

We have noticed a lot of deception from the community relating to the configuration and run of salome_meca 2021 LGPL based on the singularity (.sif) container.

For that reason, we have prepared the following WSL2 distro (based on OpenSUSE) and which contains the prerequisites for launching any salome meca container file.

To download the WSL2 distro (remove spaces from the link)

https: / / we.tl / t-0MOg4Bz6Zz

To install the distro in windows WSL2 :

wsl --import SalomeMeca2021 <install_path> <path_to_downloaded_tar>

For example

wsl --import SalomeMeca2021 C:\Data  C:\Downloads\smeca-lgpl-wsl2-opensuse.tar

To run (after install)

wsl --distribution SalomeMeca2021 --user user

Login details in the WSL2 linux system:

username : user
pwd: password

To be able to launch salome_meca_2021, first download the singularity .sif file and then run


in the same folder as the sif file (salome_meca-lgpl-2021.0.0-0-20210601-scibian-9.sif).

Other information

To turn off WSL distros

wsl --shutdown

To unregister the SalomeMeca2021 distro from WSL2

wsl --unregister SalomeMeca2021

Let us know how it works for you.

Best regards

HB from the salome_meca EDF Team

#9 Re: Code_Aster usage » Modal Analysis with Fluid - Frequency Output » 2021-11-20 10:26:21

Oh, my bad...

You need to apply REST_GENE_PHYS on your FSI modes so they are projected back to the "physical" mesh. After that they can be printed out normally like the initial modes.


#10 Re: Code_Aster usage » Modal Analysis with Fluid - Frequency Output » 2021-11-20 09:20:10

Hello Stefano,

It is not possible to print out the "fluid modes" using this procedure: CALC_MATR_AJOU followed by CALC_MODES because in reality the fluid is not part of the mechanical model. Only its "effect" (pressure fluctuations generating forces) on the initial structural modes is accounted by calculating added mass (and optionally stiffness).

However, there is 1 specific command that allows to calculate specifically the pressure fields *in the fluid* generated by the modes (by the modal acceleration actually), the command is CALC_CHAM_FLUI. It takes as input basically the same ingredients as CALC_MATR_AJOU but only calculates the fluid pressure fields, one field per mode (optionally the fluid velocity or displacement can be calculated).

It is possible to use CALC_CHAM_FLUI either with the initial (empty) strucutual modal basis, or the final coupled FSI basis for visualisation purposes.

Good luck,

#11 Re: Salome-Meca installation » salome meca 2021 installation problems » 2021-11-13 18:56:43

To be honest the output error is too generic to give you advice, did you follow the steps specific for WSL2 ?

#12 Re: Code_Aster installation » [SOL]Problem in launching ASTER-STUDY Module in Salome-Meca-2021(.sif) » 2021-11-13 09:37:50

Hi Cyprien,

Are you using ubuntu in a virtual machine ?

The Mesa warnings and errors are not harmful and can be safely ignored. They basically relate to hardware unavailable features for rendering which are explicitly used by some salome modules and viewers, but everything should work fine in software fallback mode. The only effect can be performance for big models and for that a good nvidia GPU should fix everything.

Mesa and other graphical libs are already packaged from with in the singularity container and are carefully configured for salome, there is no need whatsoever to install anything on the host system, the lightest linux system should do just fine, only prerequisite is Singularity.

If you really need to hide the errors, you can try to append


To your terminal command with singularity, but that may also hide errors that you would otherwise still get to monitor.


#13 Re: Code_Aster installation » error in downloading unstable version » 2021-11-10 12:41:47

Here is the contents of my $HOME/.hgrc file (mercurial main configuration file)

username = MY NAME <my.email@edf.fr>
merge = meld


cmd.vdiff = meld

Cloning the sourceforge dépôt seems to work just fine from my side.

Here is a snapshot

#14 Re: Code_Aster installation » error in downloading unstable version » 2021-11-09 22:20:41

A syntax error maybe in the .hgrc mercurial configuration file ?
Is it possible to clone any other repository?

#15 Re: Code_Aster usage » Comparison SM2019 vs SM2021 - odd contact results » 2021-11-09 21:41:26


Regarding the naming, this is normal in v15 which is fully "pythonic" which means that concepts are real python variable and thus no more limit is present for the naming.

The code generates a concept name automatically and this is what you see in the messages file.

As for the real problem encountered, this requires indeed further investigation.

#16 Re: Salome-Meca installation » salome meca 2021 installation problems » 2021-11-09 21:06:18

At one point during our tests on WSL2, we found it necessary to do the following

sudo rm -r /dev/shm
sudo mkdir /dev/shm
sudo chmod 777 /dev/shm

Hope it helps, let us know please.


#17 Re: Code_Aster usage » [Solved] Calculate additional fields based on existing result-file » 2021-11-08 14:42:19

Hi Jonas,

The MED file is not sufficient in order to complete the post processing since it does not contain information about the finite element model.

A good approach is to save a "base" (basis) of your calculation, a restart file that saves all context and data from the previous calculation. From a "base" you can continue your calculation (POURSUITE command) and/or do other post processing commands, print tables and other fields, etc.

Best regards,

#18 Re: Salome-Meca usage » Paravis / AsterStudy Post-process results visualization » 2021-11-08 13:50:46

Javier_Podetti wrote:

I couldn't fit the file in the Forum's max size. I'm pasting a link to a WeTransfer site.
h ttps://we.tl/t-tpltqlzU5K

Thanks but the rmed file in the RAR file seems to be corrupt, I can not read it in Paravis or Mesh.

#19 Re: Code_Aster installation » Problem with salome_meca-lgpl-2021.0.0-1-20210811-scibian-9 -> SIGSEG » 2021-11-05 22:11:58


This issue will be fixed in version 2021.0.1

It is mainly encountered with code_aster v15 and due to the new concept naming in the pythonic version of code_aster.

If you can share the med result file I can verify it will be all good with the patched upcoming version.


#21 Re: Salome-Meca usage » Using Filter ELGa Field to Surface (cell averaged) » 2021-11-05 21:49:48

Hello and welcome to the forum.

ELGA field is a special field which contains result data for all integration point (GAUSS points) in each element.

For that reason, in each element the field can have multiple values, this is totally normal but is pretty hard to visualize by Paravis.

The filter creates a simple element field, thus containing a single value per element, which is either the average on all GAUSS point, the min or the max.

That's the significance of these 3 fields created by the filter.

Best regards,


#22 Re: Salome-Meca usage » Paravis / AsterStudy Post-process results visualization » 2021-11-05 21:44:58

Hello Javier,

ELNO and ELGA fields require using filters before any results can be visualized in Paravis.

ELNOtoSurface for example creates a point field by "shrinking" the element cells.

The usage of ExtractGroup filter is more tricky, I recommend you do the IMPR_RESU on parts of the mesh instead of using Paravis for thar, it is more robust in my opinion.

Finally, I'm interested to have a look at your med result to figure out why the integrated post processor fails for your case.


#23 Re: Salome-Meca installation » Questions about singularity version of salome_meca 2021 » 2021-10-05 19:51:09

ing.nicola wrote:

i'm continuing to try to run "./as_run" of code aster fronted but it doesn't work.... it continue to give me "can't import numpy"

singularity run -B /tmp:/local00/tmp ~/containers/xxxxx.sif shell -- as_run --run case.export

Works like a charm

#24 Re: Salome-Meca usage » How to remove perspective view from post-processor of Aster Study » 2021-10-01 22:01:53

Unfortunately this is currently not possible using the interface, it is only possible using python.

It is a good idea to add this possibility, we'll add it to the backlog.

#25 Re: Salome-Meca installation » Salome-Meca 2020 in Singularity Containers: Present Status? » 2021-10-01 21:58:11

https hmm/ code-aster.org/V2/spip.php?article295

The singularity container for v2021 is out!