Atom topic feed | site map | contact | login | Protection des données personnelles | Powered by FluxBB | réalisation artaban
You are not logged in.
Pages: 1
Hi.
When I run test case ssnv504h, I get the following alarm
!------------------------------------------------------------------------------------------------------------!
! <A> <CONTACT3_88> !
! !
! !
! Contact mthode continue. !
! -> Il y a convergence force sur boucle de gomtrie lors du traitement du !
! contact. !
! -> Risque & Conseil : !
! La convergence force se dclenche lorsque le problme a du mal converger !
! lors de grands glissements relatifs entre deux surfaces de contact. !
! Il y a des risques que le problme soit un peu moins bien trait. !
! Vrifiez bien que vous n'avez pas d'interpntration entre les mailles. !
! S'il y des interpntrations intempestives, tentez de dcouper plus finement en temps votre problme. !
! !
! !
! !
! Ceci est une alarme. Si vous ne comprenez pas le sens de cette !
! alarme, vous pouvez obtenir des rsultats inattendus ! !
!---------------------------------------------------------------------
Can someone explain why this warning occurs and how to fix it? Do I need to worry about it?
Todd.
Offline
Hi,
The contact method 'CONTINUE' in Code_Aster takes contact/friction hypothesis into account in the weak formulation. It is the most precise method available, particularly when dealing with friction. For now the many non-linearities involved in contact are solved by fitting various loops together (that is one for the matching/geometry, one for the friction, one for the contact and finally the general newton one).
In the current formulation some terms are omitted and the convergence rate of the outer loop (geometry) can be slow. The maximum number of iterations of this loop is fixed via ITER_GEOM_MAXI (which defaults to 3). When 3 iterations have been made, convergence is enforced (thus the warning message). You should either increase ITER_GEOM_MAXI or visually check that the gap remains positive (that is no interfering is noticeable).
Ongoing work should enable the removal of this outer loop.
Moreover in version 9.4 the criterion has been fixed so that fewer iterations should be necessary to reach convergence. Speed has also greatly improved with the 'CONTINUE' method between 9.3 and 9.4.
Regards,
TdS
Offline
Hi Thomas
Thanks. I will try increasing ITER_GEOM_MAXI
In the same test case, I am trying to output the strains to gmsh, using the following commands
U=CALC_ELEM(reuse =U,
MODELE=MO,
CHAM_MATER=CM,
RESULTAT=U,
NUME_ORDRE=10,
OPTION=('EPSI_ELGA_DEPL','EPSI_ELNO_DEPL',),);
IMPR_RESU(FORMAT='GMSH',
UNITE=37,
RESU=_F(RESULTAT=U,
GROUP_MA='LOPIN',
TYPE_CHAM='SCALAIRE',),);
but no matter what I choose for CALC_ELEM, I only see stresses in the output.
Is it possible to output the strains to gmsh?
Offline
Ok. To answer my own question. It is possible to get the strain results output in gmesh, by choosing specific field names in the print statement, like so
U=CALC_ELEM(reuse =U,
MODELE=MO,
CHAM_MATER=CM,
RESULTAT=U,
NUME_ORDRE=10,
OPTION=('EPSI_ELGA_DEPL','EPSI_ELNO_DEPL',),);
IMPR_RESU(FORMAT='GMSH',
UNITE=37,
RESU=_F(RESULTAT=U,
NOM_CHAM=('EPSI_ELGA_DEPL','EPSI_ELNO_DEPL',),
GROUP_MA='LOPIN',
TYPE_CHAM='SCALAIRE',),);
Offline
Pages: 1