analysis error

  • Leonardo
  • Topic Author
  • Offline
  • New Member
  • New Member
More
8 years 3 weeks ago - 8 years 3 weeks ago #8586 by Leonardo
analysis error was created by Leonardo
Hi,
I'm running a linear static analysis with Salome Meca, but I got just this error in the .mess file.

!
!
! <EXCEPTION> <FACTOR_20> !
! !
! (Solveur linéaire LDLT ou MULT_FRONT) Matrice non factorisable ! !
! On sait en plus que: !
! - pivot est presque nul à la ligne 6729 pour le noeud N2243 et !
! la composante DZ, !
! - nombre de décimales perdues : 15. !
! !
! -> Conseil & Risque : !
! Il s'agit peut être 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. !
!
!

What I have found out is that it is a problem about internal constraint and that I should use the LIAISON_MAIL command to resolve it. I'm not sure of this. There is someone who knows more about it?

That's the part of .comm file on the boundary condition

CHAR=AFFE_CHAR_MECA(MODELE=MODE,
DDL_IMPO=(
_F(GROUP_MA='fixed_1',
DX=0.0,
DY=0.0,
DZ=0.0,),
_F(GROUP_MA='fixed_2',
DX=0.0,
DY=0.0,
DZ=0.0,),
_F(GROUP_MA='fixed_3',
DX=0.0,
DY=0.0,
DZ=0.0,),
_F(GROUP_MA='fixed_4',
DX=0.0,
DY=0.0,
DZ=0.0,),
_F(GROUP_MA='fixed_5',
DX=0.0,
DY=0.0,
DZ=0.0,),
_F(GROUP_MA='fixed_6',
DX=0.0,
DY=0.0,
DZ=0.0,),
),
PRES_REP=(
_F(GROUP_MA='press',
PRES=1.0,),
),
);

Thanks in advance
Last edit: 8 years 3 weeks ago by Leonardo.

Please Log in or Create an account to join the conversation.

More
8 years 3 weeks ago #8587 by Claus
Replied by Claus on topic Re: analysis error
This error usually means, that you have rigid body movements, meaning that the object is free to move in one om more axes.
Some of the of the objects probably aren't connected (if this is an assembly) correctly, allowing them to move.
LIAISON_MAIL eg. can 'glue' those parts together.

/C

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

  • Leonardo
  • Topic Author
  • Offline
  • New Member
  • New Member
More
8 years 2 weeks ago #8588 by Leonardo
Replied by Leonardo on topic Re: analysis error
Hi,
thanks for the answer Claws. I used LIASON_MAIL, running EFICAS, but now I got a new error (a lot of this one). Here it is

!
!
! <A> <CALCULEL5_7> !
! !
! Alarme utilisateur : !
! Vous utilisez la commande PROJ_CHAMP ou un mot clé nécessitant de "projeter" !
! des noeuds sur des mailles (par exemple LIAISON_MAIL). !
! L'un des noeuds (N292) du maillage (MAIL) a été projeté à une distance D non nulle significative. !
! (D>10% de la taille de la maille la plus proche (M1290) du maillage (MAIL)). !
! !
! !
! Ceci est une alarme. Si vous ne comprenez pas le sens de cette !
! alarme, vous pouvez obtenir des résultats inattendus ! !
!
!

I searched on the web but I didn't find anything helpful. I have a question (I'll try to explain myself). I choose as GROUP_MA_MAIT a solid, and as GROUP_MA_ESCL a surface to connect. The slave surface must be completely coincident with the surface of the master solid?

Thanks again for help

Please Log in or Create an account to join the conversation.

More
8 years 2 weeks ago #8589 by Claus
Replied by Claus on topic Re: analysis error
I think the error means, that the two mesh are too irregular in regards to each other. When I use LIAISON, I try to set a identical submesh on the mating surfaces, so they have approx. same number and equidistant nodes.

The calculation should run with those alarms (not errors :) ), but the results might be incorrect, as the message states.

/C

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version

Please Log in or Create an account to join the conversation.

  • Leonardo
  • Topic Author
  • Offline
  • New Member
  • New Member
More
8 years 2 weeks ago - 8 years 2 weeks ago #8590 by Leonardo
Replied by Leonardo on topic Re: analysis error
I solved it. The problem was that too many LIAISON_MAIL messed the analysis. So my advice is to use less possible of these, fusing the solids that has the same material.
Last edit: 8 years 2 weeks ago by Leonardo.

Please Log in or Create an account to join the conversation.

Moderators: catux
Time to create page: 0.132 seconds
Powered by Kunena Forum