Impose mesh displacement

Questions and remarks about Code_Saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Djeen
Posts: 8
Joined: Fri Aug 09, 2019 1:53 pm

Impose mesh displacement

Post by Djeen » Fri Aug 09, 2019 2:25 pm

Hello,

I am dealing with a case consisting in simulating a flow between two moving plates. The movement of these plates is imposed. The problem is that the internal mesh don't follow the plate displacement, even when I use a big mesh viscosity (see joined picture). I'll like to know how to impose mesh displacement with Code_Saturne 5.0.

Thank you.

Best regards,

Djeen.
Attachments
Capture d'écran de 2019-08-09 14-59-11.png

Yvan Fournier
Posts: 2689
Joined: Mon Feb 20, 2012 3:25 pm

Re: Impose mesh displacement

Post by Yvan Fournier » Fri Aug 09, 2019 8:37 pm

Hello,

What does the mesh displacement visualization look like ? Your figure does not show the deformation field, only the viscosity. I recommend also using the extract block filter (or separate meshes writer option in Code_Saturne) to separate boundary and volume meshes in ParaView.

Regards,

Yvan

Djeen
Posts: 8
Joined: Fri Aug 09, 2019 1:53 pm

Re: Impose mesh displacement

Post by Djeen » Mon Aug 19, 2019 4:16 pm

Hello
Please find attach the mesh displacement visualization and the runcase. The separate meshes writer in Code_Saturne doesn't anything.
Best regards.
Attachments
ALE_Plaques.xml
run case
(8.22 KiB) Downloaded 8 times
capture.png
Mesh displacement

Yvan Fournier
Posts: 2689
Joined: Mon Feb 20, 2012 3:25 pm

Re: Impose mesh displacement

Post by Yvan Fournier » Tue Aug 20, 2019 12:29 am

Hello,

I probably will not have the time to check your setup in the next 2 weeks.

The fact that the boundary cells are much wider than the following cells probably leads to very small deformation inside the mesh. Uding a more regular mesh will provide better résults in most cases.

Best regards,

Yvan

Luciano Garelli
Posts: 217
Joined: Fri Dec 04, 2015 1:42 pm

Re: Impose mesh displacement

Post by Luciano Garelli » Tue Aug 20, 2019 8:31 pm

Hello Djeen,

In your xml file (case) you set fixed_displacements for two boundaries (upper and lower I guess), and fixed_boundary for the rest. So, the boundary nodes con't move. You have to set sliding boundary. I think that this is the problem.

Regards,
Luciano

Djeen
Posts: 8
Joined: Fri Aug 09, 2019 1:53 pm

Re: Impose mesh displacement

Post by Djeen » Thu Aug 22, 2019 12:35 pm

Hello Luciano,
With the sliding boundary, I have a good mesh motion. Thanks.
I have another problem. I received the following error. I don't know if it depend of my parameters or if it's an internal thing of Code_Saturne. Can you or somebody explain it to me.
Best regards.

$$
Signal SIGTERM (terminaison) reçu.
--> calcul interrompu par l'environnement.

Pile d'appels :
1: 0x7f0ed2711ce6 <+0x12ce6> (mca_btl_openib.so)
2: 0x7f0ed6837eda <opal_progress+0x4a> (libopen-pal.so.13)
3: 0x7f0ed7cac565 <ompi_request_default_wait_all+0x235> (libmpi.so.12)
4: 0x7f0ed7ce626f <PMPI_Waitall+0x9f> (libmpi.so.12)
5: 0x7f0eda6d6fe9 <cs_halo_sync_var_strided+0x459> (libsaturne.so.5)
6: 0x7f0eda8e3f37 <cs_gradient_vector+0xa7> (libsaturne.so.5)
7: 0x7f0eda6d0bb2 <cs_field_gradient_vector+0xb2> (libsaturne.so.5)
8: 0x7f0eda6d0c6f <cs_f_field_gradient_vector+0x5f> (libsaturne.so.5)
9: 0x7f0eda6501af <alemav_+0x2cf> (libsaturne.so.5)
10: 0x7f0eda656fa4 <caltri_+0x1ff4> (libsaturne.so.5)
11: 0x7f0eda627c2f <cs_run+0x56f> (libsaturne.so.5)
12: 0x7f0eda62757f <main+0x12f> (libsaturne.so.5)
13: 0x7f0ed7495b15 <__libc_start_main+0xf5> (libc.so.6)
14: 0x400989 <> (cs_solver)
Fin de la pile
$$

$$
Signal SIGSEGV (accès à une zone mémoire interdite) intercepté !

Pile d'appels :
1: 0x7f6f38d07104 <clipsa_+0x1d0> (libsaturne.so.5)
2: 0x7f6f38d378b5 <turbsa_+0x2315> (libsaturne.so.5)
3: 0x7f6f38ac2d5d <tridim_+0x61ad> (libsaturne.so.5)
4: 0x7f6f3895ae12 <caltri_+0x1e62> (libsaturne.so.5)
5: 0x7f6f3892bc2f <cs_run+0x56f> (libsaturne.so.5)
6: 0x7f6f3892b57f <main+0x12f> (libsaturne.so.5)
7: 0x7f6f35799b15 <__libc_start_main+0xf5> (libc.so.6)
8: 0x400989 <> (cs_solver)
Fin de la pile
$$

Yvan Fournier
Posts: 2689
Joined: Mon Feb 20, 2012 3:25 pm

Re: Impose mesh displacement

Post by Yvan Fournier » Thu Aug 22, 2019 2:08 pm

Hello,

Which exact version (5.0.x release) of the code are you using, and on which type of system ?

Do you have user C or Fortran files, or only the XML definition ?

Regards,

Yvan

Djeen
Posts: 8
Joined: Fri Aug 09, 2019 1:53 pm

Re: Impose mesh displacement

Post by Djeen » Thu Aug 22, 2019 4:01 pm

Hello,
I'm using the version 5.0.9 of Code_Saturne. The OS is Mageia version 6.
I'm only using the XML definition.
I forgot to say that simulation are launched on a cluster.
Regards,
Djeen

Yvan Fournier
Posts: 2689
Joined: Mon Feb 20, 2012 3:25 pm

Re: Impose mesh displacement

Post by Yvan Fournier » Thu Aug 22, 2019 11:03 pm

Hello,

The error looks like a missing parallel sync or inconsistent check issue.

Could you post your mesh also (directly if it is not too big, using a link otherwise).
And if possible, the partition_output directory from the execution directory, so as to reproduce in the same partitioning.

Regards,

Yvan

Djeen
Posts: 8
Joined: Fri Aug 09, 2019 1:53 pm

Re: Impose mesh displacement

Post by Djeen » Fri Aug 23, 2019 9:23 am

Hello,
The following link is for the mesh.
https://1drv.ms/u/s!Ahb7Y7V_16BngRA21BtvmBtSUBVe
Find attached the execution directory.
Regards,
Attachments
Fichiers.zip
All files
(1.98 MiB) Downloaded 6 times

Post Reply