Trouble with assembly

I'm trying to analyse a simple assembly. I've produced the comm file after looking at the press-assembly example.

In case it helps understanding my comm file, I've a thin plate (cover_pl) that is supported by a second piece (back_pla) which has the surfaces const_1 and const_2 that are held rigid. A pressure is applied across cover_pl.


I'm getting this exception

!

---

!
! <EXCEPTION> <FACTOR_20> !
! !
! !
! -> Matrice non factorisable ! !
! Le pivot est presque nul à la ligne 97890 pour le noeud N28764 et !
! la composante DZ. !
! Si le solveur linéaire est LDLT ou MULT_FRONT, on sait en plus que le nombre de décimales !
! perdues est de 13. !
! !
! -> Conseil & Risque : !
! Il s'agit peut etre d'un mouvement de corps rigide mal bloqué. !
! Vérifiez les conditions aux limites. !
! Si vous faites du contact, il ne faut pas que la !
! structure ne "tienne" que par le contact. !
! !

which (with the assistance of google translate) suggests there's problems with a matrix computation.

Then later on in the file I have

JDC.py : ERREUR A L'EXECUTION - INTERRUPTION
>> JDC.py : DEBUT RAPPORT
CR d'execution de JDC en MIXTE
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
! <S> Exception utilisateur levee mais pas interceptee. !
! Les bases sont fermees. !
! error : <EXCEPTION LEVEE> FACTOR_20 !
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
fin CR d'execution de JDC en MIXTE

>> JDC.py : FIN RAPPORT
EXECUTION_CODE_ASTER_EXIT_7739=1
<INFO> Code_Aster run ended, diagnostic : <S>_ERROR

which is where it seems to actually fail.

I'm fairly new to this so expect I've something wrong in the comm file, or is it possible that it's in the mesh?

Any help or pointers will be much appreciated! [file]<br /><br />Post edited by: Joe Dowsett, at: 2010/07/08 23:13

Attachment failed for some reason. Attached is file from flasheur.

The two parts were created as an assembly in an external CAD package, and then exported/imported in step format. For creating the mesh I followed the process in the press frame example - I don't recall there being anything explicit about joining surfaces in the mesh? I need the surfaces to be able to slide so that the cover doesn't add strength to the back.

I've just found another assembly tutorial so shall have a look through that tomorrow.

If you've any suggestions in the meantime then I'll follow them up too.

Many thanks.<br /><br />Post edited by: Joe Dowsett, at: 2010/07/09 23:21

Yes, I think this might be relevant to my problem. No, the const_1 and const_2 are both on the back part. I just experimented removing the sliding nature of the contact and it solved fine.

I guess I over simplified my model and I need to include a couple of pins to locate the cover plate in specific spots (which is how it will be in reality).

I hope to check this later today and will report the result.


Fingers crossed!