Problem running a CS2 case

[Alexandre Lebouvier]

Hello,

I have just installed code saturne V2 and I tried it with a case of mine working on V1.3.3.

I get this error that I don't understand. Is that a installation problem ? I have added some output file to this message to help.

Thank you


Single processor Code_Saturne simulation

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

Working directory (to be periodically cleaned) :
/b1/alexa/tmp_Saturne/TEST_CS2.CAS1.01071628

Kernel version: /usr/local/saturne/cs-2.0-beta2
Preprocessor version: /usr/local/saturne/cs-2.0-beta2

***************************************************************
Compilation of user subroutines and linking of Code_Saturne
***************************************************************

********************************************
Preparing calculation
********************************************


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

/b1/alexa/tmp_Saturne/TEST_CS2.CAS1.01071628/localexec: line 3: 25021 Instruction illégale /b1/alexa/tmp_Saturne/TEST_CS2.CAS1.01071628/cs_solver $@
Error running the calculation.

Check Kernel log (listing) and error* files for details


********************************************
Error in calculation stage.
********************************************

[Brennan Sharp]

Hello Alexandre,

A simple test you could make, if you haven't already,  is just create a new, simple study with v2 and see if that runs. If it does then you should have some confidence in your installation.

Next, it is reasonable that the v2 solver would expect the study to have been prepared using the 'code_saturne create .....' command of v2, rather than of v.1.3.3, so I suggest re-creating your v1.3.3 case using the v2 study preparer in case case of other differences. The case preparer sets up some scripts that might be dependent on how Code_Saturne was compiled.

Lastly, if I remember correctly, the choice of fortran compiler changed between v1.3.3 and v2 and I think it is necessary to convert any user fortran source you may have to the newer fortran used by v2.

Regards,

Brennan

[Alexandre Lebouvier]

Hi,
 
Thanks for the answer.
 
I used "code_saturne create" to create that new study with the CS v2 preparer. I  used the created f90 files that I modified to correspond to my Fortran CS1.3.3 files. The compilation of these new f90 fortran went right. 
 
You're right I will test a simple case like a tutorial and I will let you know if it's work.
 
Alexandre

[Alexandre Lebouvier]

Finally, it was a boundary conditions problem. CS2.0 seems to not say when boundary conditions are not applied to certain free faces as in CS1.3.
 
Alexandre

[Yvan Fournier]

Hello,

Actually, Code_Saturne 2.0 should also output an error message, and even a visualizable error output (ensight.error) case when boundary conditions are not defined correctly. It is possible that some specific physics modules such as the electric arcs module do not do this, as they have some specific input routines. We'll have to check this.

Best regards,

  Yvan