Two-way coupling code_saturne / code_aster

[Oliveira77]

Hello Yvan,

I'll make the changes and I'll test a case in the next days. I'll post any progress!

Thank you!

Flávio

[Yvan Fournier]

Hello,

Great. I may have limited network connection next week, but at worse, I'll check on my return. If none of this works, I'll see with the colleagues who last ran this model (but the person who did most of that work may not be easily available).

Regards,

Yvan

[Oliveira77]

Hello Yvan,

I changed the cs_case_domain.py file, as indicates the FSI.patch.txt, however, the same error occurred again:

Code_Saturne is running
***********************

Version: 4.0
Path: /home/flavio/saturne/4.0.4

Result directory:
/home/flavio/FSI_Auto/TESTE01/RESU_COUPLING/20160606-1745


Single processor Code_Saturne simulation.
Single processor Code_Aster simulation.

****************************************
Compiling user subroutines and linking
****************************************


****************************
Preparing calculation data
****************************


***************************
Preprocessing calculation
***************************

Creating YACS coupling scheme.
Configure parser: processing /home/flavio/saturne/4.0.4/etc/salome/fsi_appli_config.xml ...
--- add module KERNEL /home/flavio/salome/V2015_1/modules/KERNEL_V7_5_1
--- add module GUI /home/flavio/salome/V2015_1/modules/GUI_V7_5_1
--- add module YACS /home/flavio/salome/V2015_1/modules/YACS_V7_5_1
--- add module FSI /home/flavio/saturne/4.0.4
WARNING: context file does not exist

**********************
Starting calculation
**********************

Terminating SALOME on port 2810...
th. 139689970734848 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/NamingService/SALOME_NamingService.cxx [1488] : CORBA::SystemException: unable to contact the naming service
runSalome running on GENIUS
Searching for a free port for naming service: 2810 - OK
Searching Naming Service + found in 0.1 seconds
Searching /Registry in Naming Service + found in 0.5 seconds
Searching /Kernel/ModulCatalog in Naming Service +th. 139644581013312 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/ModuleCatalog/SALOME_ModuleCatalog_Server.cxx [101] : Module Catalog Server: Naming Service was found
found in 0.5 seconds
RunStudy
Searching /myStudyManager in Naming Service + found in 0.5 seconds
Searching /Containers/GENIUS/FactoryServer in Naming Service + found in 0.5 seconds
Start SALOME, elapsed time : 2.1 seconds
graph = fsi_yacs_scheme.xml options: display=0 verbose=0 stop-on-error=1 shutdown=1 reset=0 dumpErrorFile=dumpErrorState.xml
Starting 0 ms
Elapsed time before load: 223 ms
Elapsed time after load: 228 ms
Elapsed time after validation: 228 ms
Elapsed time after check consistency: 229 ms
+++++++++++++++++++ start calculation +++++++++++++++++++
th. 140080115181312 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [628] : [GiveContainer] waiting 60 second steps container /Containers/GENIUS/7107_FSI_SATURNE_0xedd3c0
th. 140080123574016 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [628] : [GiveContainer] waiting 60 second steps container /Containers/GENIUS/7107_FSI_MILIEU_0xecc630
th. 140080106788608 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [628] : [GiveContainer] waiting 60 second steps container /Containers/GENIUS/7107_FSI_ASTER_0xec9f70
th. 140080115181312 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [426] : [GiveContainer] container /Containers/GENIUS/7107_FSI_SATURNE_0xedd3c0 launched
SalomeContainer launched : /Containers/GENIUS/7107_FSI_SATURNE_0xedd3c0 GENIUS 7134
th. 140080123574016 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [426] : [GiveContainer] container /Containers/GENIUS/7107_FSI_MILIEU_0xecc630 launched
SalomeContainer launched : /Containers/GENIUS/7107_FSI_MILIEU_0xecc630 GENIUS 7136
th. 140080106788608 - Trace /local00/home/I27518/TRAV/B4/V2015_1/modules/src/KERNEL/src/Container/SALOME_ContainerManager.cxx [426] : [GiveContainer] container /Containers/GENIUS/7107_FSI_ASTER_0xec9f70 launched
SalomeContainer launched : /Containers/GENIUS/7107_FSI_ASTER_0xec9f70 GENIUS 7138
Error while trying to create a new component: component 'FSI_SATURNE' is not installed or it's a wrong name
SALOME_Container::create_component_instance failed
Error while trying to create a new component: component 'FSI_ASTER' is not installed or it's a wrong name
SALOME_Container::create_component_instance failed
Problem in connectService: fluid_service Component is probably not launched. Modify your YACS file
+++++++++++++++++++ end calculation +++++++++++++++++++
Proc state : FAILED
Elapsed time after execution: 21310 ms
Execution has ended in error
<error node= proc state= FAILED>
<error node= inter_service state= ERROR>
Problem in connectService: fluid_service Component is probably not launched. Modify your YACS file
</error>
<error node= fluid_service state= ERROR>
Error while trying to create a new component: component 'FSI_SATURNE' is not installed or it's a wrong name
SALOME_Container::create_component_instance failed
</error>
<error node= solid_service state= ERROR>
Error while trying to create a new component: component 'FSI_ASTER' is not installed or it's a wrong name
SALOME_Container::create_component_instance failed
</error>
</error>
shutdown SALOME container: /Containers/GENIUS/7107_FSI_MILIEU_0xecc630
shutdown SALOME container: /Containers/GENIUS/7107_FSI_SATURNE_0xedd3c0
shutdown SALOME container: /Containers/GENIUS/7107_FSI_ASTER_0xec9f70
/home/flavio/FSI_Auto/TESTE01/RESU_COUPLING/20160606-1745/run_solver: linha 15: $localexec: redirecionamento ambíguo
solver script exited with status 1.

Error running the calculation.

Check Code_Saturne log (listing) and error* files for details.


****************************
Saving calculation results
****************************

Error in calculation stage.

Regards.

Flávio

[Yvan Fournier]

Hello Flàvio,

Your error messages are slightly different from the ones I had (I had a similar message for the Code_Aster component, not for the others), so I still suspect some installation issue, but am not too sure where/how.

I don't know in which files additional error details may be logged, or how to activate more detailed tracing for SALOME, so I can ask some colleagues from the SALOME team, but June is very busy, so maybe checking for debugging recommendations on the www.salome-platform.org forum might help also...

I'll keep you informed if I find more info before you do....

Regards,

Yvan

[Oliveira77]

Hello Yvan,

I got a new advance in coupling problem. Checking the log files (in the /tmp directory), I observed that the same problem ("GFORTRAN_1.4' not found") appears in the log of the FSI_SATURNE. Then I added the command "export LD_PRELOAD=$LD_PRELOAD:/usr/lib/x86_64-linux-gnu/libgfortran.so.3" in the file 'runcase' and the FSI_SATURNE component works without problem.

In the FSI_ASTER log, I observed that the problem was with the memory of the computer. The problem required more that 4 Gb of the ram memory. I added more memory in my computer (8Gb) and the FSI_ASTER component works too.

However, a new problem occurs. I think that the problem is with the command file of ASTER. I'll make new tests for to check this!

Regards,

Flávio

[Yvan Fournier]

Hello Flávio,

This is great news. I am not familiar with Code_Aster, but if you are, the last issue should not be the most difficult.

If needed, I also have a test case or two, which I have not run, but can provide.

Regards,

Yvan

[Oliveira77]

Hi Yvan,

I would like to have these cases. Any contribution is welcome!

Thank you!

Flávio

[Yvan Fournier]

Hello,

Here is the smallest of the cases I have.

Disclaimer: I did not test it yet on my side (2 concentric cylinders)...

Regards,

Yvan

[Oliveira77]

Hello Yvan,

I finally solved a fluid-structure interaction problem using the code_saturne/code_aster coupling (I am very satisfied with this progress), but the results obtained show a relative error of at least 33% in relation to the results of the literature. I'm trying to implement a variable mesh viscosity as function of the deformation of the structure and for this I'm modifying the usvima subroutine. Could you let me know if this modification is enough?

In my example, the viscosity is considered to be high in a radius near the end that exhibits the greatest displacement (avoiding a large mesh distortion) and this value decreases with increasing this distance (propagating the movement of the structure along the mesh of the fluid). In this case, some problem is occurring in the solution of the problem and the process is being interrupted in the first interaction.

Regards,

Flávio

[Yvan Fournier]

Hello,

Good to know you made progress. Did you include the patches on the code_aster side ? They have been integrated into the main Code_Aster development branch since last year, but I am not sure exactly at which point (you would need to check the repository.

On the CFD side, I believe usvima is enough, but will need to check with Martin, who knows that part better than me.

Best regards,

Yvan