error with gas combustion - specific physic - dp_C3P

[sylvain_serra]

Hello everyone,

I'm trying to run a case on saturne 2,1,4.

This case use a particular physics « gas combustion ».
I use a xml file created with SaturneGUI

so I add in the:
DATA directory, the dp_C3P file

SRC direcory, the usppmo.f90 file with ippmod(icoebu) = 2
the usebu1.f90 usebui.f90 and usebuc.f90 files


Here is the error I obtain:


--------------------------------


@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@

@ @@ ATTENTION : ARRET A L'ENTREE DES DONNEES (COLECD)

@ =========

@ PHYSIQUE PARTICULIERE (COMBUSTION GAZ)

@

@ Erreur a l'ouverture du fichier parametrique.

@

@ Le calcul ne sera pas execute.

@

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@

Pile d'appels :

1: 0x2af026227ddb <colecd_+0x201b> (libsaturne.so.0)

2: 0x2af026109755 <pplecd_+0x25> (libsaturne.so.0)

3: 0x2af0260db6d5 <varpos_+0x1e5> (libsaturne.so.0)

4: 0x2af02607ebd2 <iniusi_+0x1f2> (libsaturne.so.0)

5: 0x2af02607e6db <initi1_+0x1b> (libsaturne.so.0)

6: 0x2af025fde117 <cs_run+0xc7> (libsaturne.so.0)

7: 0x2af025fde985 <main+0x1e5> (libsaturne.so.0)

8: 0x39aac1d994 <__libc_start_main+0xf4> (libc.so.6)

9: 0x4024e9 <main+0x129> (cs_solver)

fin de la pile
--------------------------------

It's seems that the dp_C3P file is not found...

Is there something special to do in the runcase?

I tried to add in the DATA directory the cs_user_scripts.py
and modify the line domain.thermochemistry_data = 'dp_C3P'


but the error is the same.

Can someone help me?

Thank you very much

Sylvain

[Yvan Fournier]

Hello,

This might be either us user error or a bug related to to the name of the dp_C3P file (as different combustion variants use similar names).

In your temporary execution directory, do you have a file with a name similar to dp_C3P ?
Otherwise, could you post your study here ?

Best regards,

Yvan

[sylvain_serra]

Hello Yvan,

The dp_CP file is not in the temporary execution repertory.
I add the case and the mesh in this post.

I tried with colecd to write ficfpp and the result is not dp_C3P but dp_thcp

Thank you for your help

Sylvain

[Yvan Fournier]

Hello,

In the cs_user_scripts.py, you commented the "if domain.param == None:" line before setting domain.thermochemistry_data, but you did not change the indentation for the chemistry data. As Python bases its scoping on indentation, this means that the domain.thermochemistry_data is set base on the previous test, relative to domain.meteo_data.

If you simply replace:

Code: Select all

#if domain.param == None:
    domain.thermochemistry_data = 'dp_C3P'

by:

Code: Select all

#if domain.param == None:
domain.thermochemistry_data = 'dp_C3P'

then things improve, and the file is read.

Running your case, I get another error, which is:

The kernel variables number 8 and the calculated one by the GUI 6 are not the same.


Pile d'appels :
1: 0x7f732adae259 <bft_error+0xde> (libbft.so.0)
2: 0x7f732bf62873 <csvnum_+0x1a9b> (libsaturne.so.0)
3: 0x7f732b7f32c9 <iniusi_+0x63d> (libsaturne.so.0)
4: 0x7f732b7f2880 <initi1_+0x30> (libsaturne.so.0)
5: 0x7f732b6a8723 <cs_run+0xc7> (libsaturne.so.0)
6: 0x7f732b6a93b2 <main+0x204> (libsaturne.so.0)
7: 0x7f7328742c8d <__libc_start_main+0xfd> (libc.so.6)
8: 0x401c19 <> (cs_solver)
Fin de la pile

Meaning there is a conflict between the numberof variables determined by the GUI and that determined by the user subroutines and specific physics.

As gas combustion is not handled by the GUI yet, I guess the GUI does not read the dp_C3P file, the number of variables computed is false.
Solutions would be either:
- add variables in the GUI matching the ones that should be added by gas combustion so that checks may proceed (this might work, but I am not sure it is safe)
- assume that gas combustion is not handled by the GUI yet, and do the data setup using only user subroutines.

I checked the gas combustion is not handled by the soon-to-be released version 2.2, and I do not know whether it will be in 2.3 (around June/July) or in 3.0 (I can check with collegues), so for the short term, using user subroutines might be a safe bet.

Best regards,

Yvan

[sylvain_serra]

Thank you very much Yvan,

I will try without Gui.

Sylvain