Hey, problem with STAT_NON_LINE

Hey guys,

I am trying to implement a soil model with tunnel, through a plate system. It involves 3 steps:

1) Applying gravity to the whole system, and estimating the stresses.

2) Excavation of tunnel, simulated through a hole in the plate. Stresses from Stage1 are applied to the nodes before computation through CREA_CHAMP.

3) Applying circular beam of steel along the circumference of the excavated hole. Stresses from Stage2 are imported before computation through CREA_CHAMP..

However, I'm getting <EXCEPTION> <CALCULEL_55> on the final calculation. Can someone tell me where I'm going wrong?

I'm attaching the .med, .mess and. comm files.

Note: There are 4 groups in the mesh: 1 for the whole plate system "system", 1 for plate with hole as "soil", 1 for the circular lining as "lining", and a 4th group "excavation" for the hole in the plate which is not used.

Thanking in advance,
d_riel

Hello d_riel,
the problem comes from the fact that the modelization you use for solution RESU2 is different from the modelisation of the result RESU1 where you extract the initial stresses and so aster complains that the initial stresses are not found on some part of the model.
In this case the problem comes from the additional group 'lining' which is modeled as POU_D_T.
To solve this problem you have to project the stress field on the new model with the command PROJ_CHAMP (and use option PROL_ZERO='OUI' to fill missing values with zeros) before applying as initial state.

Best regards,
Richard

Hey Richard,

Thanks for the reply. So I was trying with PROJ_CHAMP, but I am not sure if the results I'm getting are what is required. I've attached two command files(no.1 is working, though I'm not sure if its tranferring (only) the stresses as is needed, and no.2 which is not working). Attaching the message files also. Kindly advice.

Regds,
d_riel

Hello d_riel,
unfortunately there are errors in both of your cases.

The first (working) one does not include any initial state for RESU2, so nothing is transferred and all initial values are assumed to be zero.

In the second case you use METHODE='AUTO' in PROJ_CHAMP, this way, at first the nodal values are projected and afterwards the values at gauss points (fields of type SIEF_ELGA_R, for example your initial stresses). But here aster complains that NONE nodal fields are found (which is correct as you only want to project the stresses at gauss points) and so the calculation aborts. I admit the error message is a bit unclear in this instance.
To project only fields that are known at gauss points you have to use METHODE='ECLA_PG'.

Best regards,
Richard

Hey Richard,

Thanks for the advice, but its still not running with METHODE='ECLA_PG'

Attaching the message file.

Regds,
d_riel

Hello d_riel,
I can't see any error, except that you are still using SIGM1 as initial state for RESU2 and not the projected field PROJ.
But this doesn't explain the syntax error....hmm.
I can't test it by myself as I currently have no acces to my notebook.

Which Code-Aster version are u using?

Regards,
Richard