NameError on re-running already successful simulation

Hello,

Recently I wanted to re-run some of the simulations which have been made by me some time ago and which ended at that time successfuly. However, now - during the simulation I got an error which never happened to me before:

.. __stg1_txt5

  # ------------------------------------------------------------------------------------------

  # ------------------------------------------------------------------------------------------

  POURSUITE(PAR_LOT='NON',

            IGNORE_ALARM=('SUPERVIS_1', 'ALGORITH11_87'),

            LANG='en',

            FORMAT_HDF='NON',

            IMPR_MACRO='NON',

            DEBUG=_F(JXVERI='NON',

                     SDVERI='NON',

                     JEVEUX='NON',

                     HIST_ETAPE='NON',),

            MESURE_TEMPS=_F(NIVE_DETAIL=1,

                            MOYENNE='NON',),

            MEMOIRE=_F(TAILLE_BLOC=800.0,

                       TAILLE_GROUP_ELEM=1000,),

            RESERVE_CPU=_F(BORNE=900,),

            CODE='NON',

            INFO=1,)

  # MĂ©moire (Mo) :     7.58 /     7.58 (Optimum / Minimum)

  # Fin orders No:  0001 user+syst:          0.00S (syst:          0.00S, Elaps:          0.00S)

  # ------------------------------------------------------------------------------------------

ERREUR WITH INTERPRETATION IN ACCAS - INTERRUPTION

>> JDC.py: DEBUT CR phase d'initialisation

   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

   ! erreur de syntaxe,  NameError: name 'MODEL' is not defined ligne 37 !

   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

fin CR phase d'initialisation

 REPORT

 >> JDC.py: FIN REPORT

End of the Code_Aster execution

Code_Aster MPI exits normally

Exited

EXECUTION_CODE_ASTER_EXIT_11=1

What I also noticed for the first time, was a warning message after meshing:

An empty node group has been identified while setting up the simulation. This is most likely the result of having bodies on the structure where the mesh has only one-element thickness on any direction, and has multiple boundary conditions or slave faces of a contact assigned to them. To resolve this problem, you can go to Mesh settings and make a refinement on the problematic bodies of the structure or use custom Element technology to specify a second order mesh. If after trying this the error persists, please contact support.

I also share my project link where you can see the full error log:

Could you please help me to understand what is wrong with my simulations?

Hi there, and thanks for taking the time to reach out to the forum for help.

This is a known problem, that was introduced by a recent update. In short, we are trying to improve the heat transfer between contacting parts of assemblies, but it conflicts with the former contact method. Sorry for the inconveniences caused by this.

A fix for the issue was already released, and I can see that your last run is completed. Please keep testing and report back if the issue persists for you, or any other problem you face.

1 Like

Hi! I have tested the simulations, and indeed it is possible to get the results like volume temperature, however the simulation overall still fails, so it is impossible to see or to download the results. Is this also related to the issue you mentioned?

Hi again!

The error occurs during the final format conversion for postprocessing, after the solution is computed. Actually, the reason is that the result database is too large, because you are storing all the time steps in the results:

image

That means 1780 time steps, which is a lot! My suggestion to fix this is to store just a few time, or even only the last one if you do not need all of them. You can achieve this by using the “Write control definition”, as explained in this page:

Hi again!

Thank you for the detailed explanation! The thing is, I need to observe how the temperature changes on the selected volumes during all the simulation, in best case with 1[s] steps. Temperature plots would be enough. Can I achieve this without storing full results?

Hi there!

In principle yes, you can just remove all the entries under “Solution Fields” and go with your point/surface/volume data.

Please try this if you see fit and let me know how it goes.