Surfacic cube for deformation, displacement and constraint

MarcoB

Hello!
I just join the forum so I can understand a whole lot about code aster.
So I created a simple cube of 111 with 6 faces thanks to vortexes and segments.
from there create 2 groups.
"sol" which is 1 face which will be my locked boundary condition.
then "cube" which is all the faces.

I mesh everything with 2D quadrangles and 1D local length 0.1.
I create once again the group sol and cube from the geometry.
!

I go to asterstudy.

mesh = LIRE_MAILLAGE(
UNITE=20
)
model = AFFE_MODELE(
AFFE=F(
GROUP_MA=('cube', ),
MODELISATION='DKT',
PHENOMENE='MECANIQUE'
),
MAILLAGE=mesh
)
elemprop = AFFE_CARA_ELEM(
COQUE=F(
A_CIS=0.8333333,
COEF_RIGI_DRZ=1e-05,
COQUE_NCOU=1,
EPAIS=0.005,
GROUP_MA=('cube', )
),
MODELE=model
)
mater = DEFI_MATERIAU(
ELAS=F(
E=210000000000.0,
NU=0.3,
RHO=7850.0
)
)
fieldmat = AFFE_MATERIAU(
AFFE=F(
MATER=(mater, ),
TOUT='OUI'
),
MAILLAGE=mesh
)
load = AFFE_CHAR_MECA(
DDL_IMPO=F(
DRX=0.0,
DRY=0.0,
DRZ=0.0,
DX=0.0,
DY=0.0,
DZ=0.0,
GROUP_NO=('sol', )
),
MODELE=model,
PESANTEUR=F(
DIRECTION=(0.0, 0.0, -1.0),
GRAVITE=9.81
)
)
reslin = MECA_STATIQUE(
CARA_ELEM=elemprop,
CHAM_MATER=fieldmat,
EXCIT=F(
CHARGE=load
),
MODELE=model
)
reslin = CALC_CHAMP(
CONTRAINTE=('SIEF_ELGA', 'SIEF_NOEU'),
DEFORMATION=('EPSI_ELGA', 'EPSI_NOEU'),
RESULTAT=reslin
)
IMPR_RESU(
FORMAT='MED',
RESU=F(
NOM_CHAM=('DEPL', 'SIGM_ELNO', 'EPSG_ELNO'),
RESULTAT=reslin
),
UNITE=2
)

I got a fatal stop telling me !
<EXCEPTION> <PLATE1_40> ║
║ ║
║ L'élément de plaque ne peut pas être orienté. Par défaut, pour orienter l'élément, on y ║
║ projette l'axe global X. ║
║ L'axe de référence pour le calcul du repère local est ici normal à ║
║ l'élément. ║
║ Il faut donc modifier l'axe de référence en utilisant ANGL_REP ou VECTEUR dans ║
║ AFFE_CARA_ELEM. ║
║ -------------------------------------------- ║
║ Contexte du message : ║
║ Option ║
║ : SIEF_ELGA ║
║ Type d'élément : MEDKQU4 ║
║ Maillage : 00000001 ║
║ Maille : ║
║ 321 ║
║ Type de maille : QUAD4 ║
║ Cette maille appartient aux groupes de mailles suivants : ║
║ ║
║ cube ║
║ Position du centre de gravité de la maille : ║
║ x=0.000000 y=0.950000 ║
║ z=0.950000

tpa

MarcoB Il faut donc modifier l'axe de référence en utilisant ANGL_REP ou VECTEUR dans ║
║ AFFE_CARA_ELEM

Here code_aster reports the issue and explains what to do. Did you try that ?

MarcoB

I want to add I tried each bloc one after the other and once I use impr_resu I get that problem. with a red dot.

Nemax

Can you use this 3D solid cube as discrete elements at all???

MarcoB

Nemax hello! if you ask me why I don't make it a volume and work in 3D. it's because I want to learn how to truly use aster study step by step. I though a cube made of surface would be "easier".

Nemax

MarcoB
I meant, can you use MODELISATION='DKT' in this case? Is that possible?

MarcoB

tpa Oh yes I did! but the problem is the vector cann't be used. see my cube has a face on each plane. so each normal get coincidence with a vector. And when I put a 1;1;1 vector it just doesn't recognize any element following that vector.

tpa

You could try to divide the domain in element groups each with its own orientation vector.

MarcoB

tpa I see what you mean I will try that!
Now I'm trying with a Volumic cube. with a small cube I have no problem but with a cube 8 times biggers. the mesh is causing me trouble that's weird. I didn't change anything except the size.

MarcoB

tpa tpa Hello, thanks TPA. I got other error and use AI to help me understand then I just stop using CALC_CHAMP to see if my model would run with only DEPL in IMPR_RESU. it works. so I had a problem to understand what was messing it with CALC_CHAMP. With AI again I finally got my answer. I can't get result on any sieq or epsg_noeu directly. I have to get elno then use post_champ and transform my sieg_elno to get sieg_noeu with the next calc_champ which will be the result file of post_champ.

For now I can't get any deformation value as post_champ don't have ESPG_NOEU in its list.
I will keep updating this post if I get more problem.

MarcoB

Hello thanks for the solution! I have tried with a cube of 8meter. I had some serious deformation (oh a long plate that thin 0.005m deform no kidding!). So I tried to add beams inside it on z and y.
I also added another form against the cube. (see the picture).

I tried with that code :
DEBUT(LANG='EN')

mesh = LIRE_MAILLAGE(identifier='0:1',
UNITE=2)

model = AFFE_MODELE(identifier='1:1',
AFFE=(_F(GROUP_MA=('FaceNormalY_Z', 'FaceNormal_X'),
MODELISATION='DKT',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('renfortY', 'RenfortZ'),
MODELISATION='POU_D_T',
PHENOMENE='MECANIQUE')),
MAILLAGE=mesh)

elemprop = AFFE_CARA_ELEM(identifier='2:1',
COQUE=(_F(EPAIS=0.005,
EXCENTREMENT=-0.00025,
GROUP_MA=('FaceNormalY_Z', ),
INER_ROTA='OUI'),
F(EPAIS=0.005,
EXCENTREMENT=-0.00025,
GROUP_MA=('FaceNormal_X', ),
INER_ROTA='OUI')),
MODELE=model,
ORIENTATION=(F(CARA='VECT_Y',
GROUP_MA=('renfortY', ),
VALE=(0.0, 0.0, 1.0)),
F(CARA='VECT_Y',
GROUP_MA=('RenfortZ', ),
VALE=(1.0, 0.0, 0.0))),
POUTRE=(F(CARA=('HY', 'HZ'),
GROUP_MA=('renfortY', ),
SECTION='RECTANGLE',
VALE=(0.005, 0.15)),
_F(CARA=('HY', 'HZ'),
GROUP_MA=('RenfortZ', ),
SECTION='RECTANGLE',
VALE=(0.005, 0.15))))

mater = DEFI_MATERIAU(identifier='3:1',
ELAS=_F(E=210000000000.0,
NU=0.3,
RHO=7850.0))

fieldmat = AFFE_MATERIAU(identifier='4:1',
AFFE=_F(GROUP_MA=('renfortY', 'RenfortZ', 'FaceNormalY_Z', 'FaceNormal_X'),
MATER=(mater, )),
MAILLAGE=mesh)

load = AFFE_CHAR_MECA(identifier='5:1',
DDL_IMPO=F(DX=0.0,
DY=0.0,
DZ=0.0,
GROUP_MA=('sol', )),
MODELE=model,
PESANTEUR=F(DIRECTION=(0.0, 0.0, -1.0),
GRAVITE=9.81))

reslin = MECA_STATIQUE(identifier='6:1',
CARA_ELEM=elemprop,
CHAM_MATER=fieldmat,
MODELE=model)

#comment: reslin = CALC_CHAMP(identifier='7:1',
#comment: reuse=reslin,
#comment: CONTRAINTE=('SIEF_ELNO', ),
#comment: DEFORMATION=('EPSI_ELNO', ),
#comment: RESULTAT=reslin)

#comment: resu_sup = POST_CHAMP(identifier='8:1',
#comment: EXTR_COQUE=_F(NIVE_COUCHE='SUP',
#comment: NOM_CHAM=('SIEF_ELNO', 'EPSI_ELNO'),
#comment: NUME_COUCHE=1),
#comment: RESULTAT=reslin)

#comment: resu_sup = CALC_CHAMP(identifier='9:1',
#comment: reuse=resu_sup,
#comment: CONTRAINTE=('SIEF_NOEU', ),
#comment: DEFORMATION=('EPSI_NOEU', ),
#comment: RESULTAT=resu_sup)

#comment: IMPR_RESU(identifier='10:1',
#comment: FORMAT='MED',
#comment: RESU=(_F(NOM_CHAM=('DEPL', 'SIEF_NOEU', 'EPSI_NOEU'),
#comment: RESULTAT=resu_sup),
#comment: _F(RESULTAT=reslin)),
#comment: UNITE=80)

IMPR_RESU(identifier='11:1',
RESU=_F(MAILLAGE=mesh,
NOM_CHAM=('DEPL', ),
RESULTAT=reslin),
UNITE=3)

FIN()

I also tried to use the point of the face against the ground and also just the point of its edges
I also tried with Liaison_elem or even Liaison_mail but those doesn't seem to fit what I'm doing.
With Elem it seem I have to create one for each beam... and mail doesn't seem to fit the bill too
then ended up with that error...
<EXCEPTION> <FACTOR_11> ║
║ ║
║ Problème : la matrice est singulière ou presque singulière : ║
║ Lors de la factorisation de la ║
║ matrice, on a rencontré un problème ║
║ (pivot nul ou presque nul) à la ligne 601 qui correspond ║
║ au degré de liberté donné ci-dessus. ║
║ ║
║ Risques et conseils : ║
║ * Si la ligne correspond a un ║
║ degré de liberté physique, il s'agit probablement 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. ║
║ Vérifiez également les ║
║ caractéristiques matériaux (module d'Young, ...). ║
║ ║
║ * Si la ligne correspond a un degré de ║
║ liberté de Lagrange, il s'agit sans doute d'une condition ║
║ limite redondante. ║
║ En ║
║ particulier, il se peut que la relation linéaire surabondante provienne des conditions de ║
║ contact. ║
║ Peut-être devriez vous exclure certains noeuds des conditions de contact ║
║ ║
║ (mots clés SANS_NOEUD et SANS_GROUP_NO). ║
║ ║
║ * Si le solveur utilisé est LDLT ou MULT_FRONT, ║
║ vous pouvez utiliser le solveur MUMPS ║
║ car celui-ci est le seul à pouvoir factoriser les ║
║ matrices qui ne sont pas définies positives. ║
║ ║
║ * Parfois, en parallèle, le critère de ║
║ détection de singularité de MUMPS est trop pessimiste ! Il reste néanmoins souvent ║
║ ║
║ possible de faire passer le calcul complet en relaxant ce critère (augmenter de 1 ou 2 la ║
║ valeur du mot-clé NPREC) ou ║
║ en le débranchant (valeur du mot-clé NPREC=-1) ou en ║
║ relançant le calcul sur moins de processeurs. ║
║ ║
║ * Il se peut aussi que ce phénomène soit ║
║ tout à fait normal avec X-FEM si la fissure passe ║
║ très près d'un noeud. ║
║ Si le nombre ║
║ de décimales perdues n'est pas trop grand (max 10 décimales), ║
║ vous pouvez relancer le ║
║ calcul en augmentant le nombre de décimales perdues autorisé : ║
║ mot-clé NPREC du mot clé ║
║ facteur SOLVEUR. ║
║ Sinon, contactez l'équipe de développement.

And lastly I'm thinking maybe I shouldn't use beam element but just coq and turn those edge into face with the dimension I want them to have before adding their thickness in AFFE_CARA_ELEM.

StanislavPavlica

Hello MarcoB,

add

EXCIT=F(CHARGE=load),

to MECA_STATIQUE

Stano

MarcoB

StanislavPavlica
Hello! Thanks for the proposition, sadly it gives me the same error. I think I was in the middle of trying each option one by one to check what was wrong, that's why it wasn't in the scrit.

MarcoB

Nemax I think so! But I will surely only use COQUE. Even for my beam