Modeling Stiffened Plates Using DKT/DST and POU_D_E/POU_D_T using LIAISON_MAIL

Nidish

I've been trying to create stiffened plates for a class I'm teaching on aircraft structures. I'm trying to use the beam elements for the stringers and the DST elements for the skins. The end goal is to create a stiffened skin wing with ribs, spars, and stringers that I want to use for structural design.

I'm trying to enforce the constraints using LIAISON_MAIL but I've not been able to get this to work yet. I'll appreciate any help!

Here's a screenshot of the mesh :

I basically have a 2d plane and a beam running through the middle lying a constant distance from the plate (I'm modeling the beam through its centroidal axis).

Here's my stage - I'm simply interested in conducting a bending analysis :
`mesh = LIRE_MAILLAGE(UNITE=3)

model = AFFE_MODELE(AFFE=(_F(GROUP_MA=('MidLine', 'Face'),
MODELISATION='DST',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('Beam', ),
MODELISATION='POU_D_T',
PHENOMENE='MECANIQUE')),
MAILLAGE=mesh)

elemprop = AFFE_CARA_ELEM(COQUE=F(EPAIS=0.005,
GROUP_MA=('Face', ),
VECTEUR=(0.0, 1.0, 0.0)),
MODELE=model,
POUTRE=F(CARA=('HY', 'HZ'),
GROUP_MA=('Beam', ),
SECTION='RECTANGLE',
VALE=(0.005, 0.25)))

mater = DEFI_MATERIAU(ELAS=_F(E=210000000000.0,
NU=0.3))

fieldmat = AFFE_MATERIAU(AFFE=_F(MATER=(mater, ),
TOUT='OUI'),
MODELE=model)

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_MA=('bot', )),
FORCE_ARETE=F(FY=1.0,
GROUP_MA=('top', )),
LIAISON_ELEM=_F(GROUP_MA_1=('Face', ),
GROUP_MA_2=('Beam', ),
OPTION='2D_POU'),
MODELE=model)

reslin = MECA_STATIQUE(CARA_ELEM=elemprop,
CHAM_MATER=fieldmat,
EXCIT=_F(CHARGE=load),
MODELE=model)

IMPR_RESU(RESU=_F(CARA_ELEM=elemprop,
RESULTAT=reslin),
UNITE=80)
`

When I run it, it keeps throwing the following error :
`# Commande #0007 de /local00/tmp/run_aster_q6rt_n14/./RunCase_1_Stage_1.comm.changed.py, ligne 46
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_MA='bot'),
DOUBLE_LAGRANGE='OUI',
FORCE_ARETE=F(FY=1.0,
GROUP_MA='top'),
INFO=1,
LIAISON_ELEM=_F(GROUP_MA_1='Face',
GROUP_MA_2='Beam',
OPTION='2D_POU'),
MODELE=model,
VERI_AFFE='OUI',
VERI_NORM='OUI')

MAILLE : M49
MAILLE : M50
MAILLE : M51
MAILLE : M52
MAILLE : M53
MAILLE : M54
MAILLE : M55
MAILLE : M56
MAILLE : M105
MAILLE : M106
MAILLE : M107
MAILLE : M108
MAILLE : M109
MAILLE : M110
MAILLE : M111
MAILLE : M112

╔════════════════════════════════════════════════════════════════════════════════════════════════╗
║ <EXCEPTION> <MODELISA6_96> ║
║ ║
║ les 16 mailles imprimées ci-dessus n'appartiennent pas au modèle ║
║ et pourtant elles ont été affectées dans le mot-clé facteur : FORCE_ARETE ║
╚════════════════════════════════════════════════════════════════════════════════════════════════╝
`

The same error was also encountered here : https://forum.code-aster.org/public/d/25467-solved-force-arete-in-a-group-ma-ca-saying-the-group-doesn-t-exist
But that relates to joining plates together, not a plate and a beam.

I'm attaching a .hdf file that can be run with salome_meca here : https://drive.google.com/file/d/1UlW8UbxlbrMj4Kx1xI0oFBN_EuUmkBiH/view?usp=sharing

Nidish

I've managed to get this working with LIAISON_GROUP since it looks like LIAISON_ELEM can't handle beams embedded on shells yet. Here's the comm file for anyone that's interested :

mesh = LIRE_MAILLAGE(UNITE=3)

model = AFFE_MODELE(AFFE=(_F(GROUP_MA=('MidLine', 'Face'),
                             MODELISATION='DST',
                             PHENOMENE='MECANIQUE'),
                          _F(GROUP_MA=('Beam', ),
                             MODELISATION='POU_D_T',
                             PHENOMENE='MECANIQUE')),
                    MAILLAGE=mesh)

elemprop = AFFE_CARA_ELEM(COQUE=_F(EPAIS=0.005,
                                   GROUP_MA=('Face', ),
                                   VECTEUR=(0.0, 1.0, 0.0)),
                          MODELE=model,
                          POUTRE=_F(CARA=('HY', 'HZ'),
                                    GROUP_MA=('Beam', ),
                                    SECTION='RECTANGLE',
                                    VALE=(0.005, 0.25)))

mater = DEFI_MATERIAU(ELAS=_F(E=210000000000.0,
                              NU=0.3))

fieldmat = AFFE_MATERIAU(AFFE=_F(MATER=(mater, ),
                                 TOUT='OUI'),
                         MODELE=model)

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_MA=('bot', ),
                                  GROUP_NO=('bot_vn', )),
                      DOUBLE_LAGRANGE='NON',
                      FORCE_NODALE=_F(FY=1000.0,
                                      GROUP_NO=('LoadPt', )),
                      LIAISON_GROUP=_F(COEF_IMPO=0.0,
                                       COEF_MULT_1=(1.0, ),
                                       COEF_MULT_2=(-1.0, ),
                                       DDL_1=('DY', ),
                                       DDL_2=('DY', ),
                                       GROUP_NO_1=('Beam_n', ),
                                       GROUP_NO_2=('MidLine_n', )),
                      MODELE=model)

reslin = MECA_STATIQUE(CARA_ELEM=elemprop,
                       CHAM_MATER=fieldmat,
                       EXCIT=_F(CHARGE=load),
                       MODELE=model,
                       SOLVEUR=_F(ELIM_LAGR='OUI',
                                  METHODE='MUMPS',
                                  STOP_SINGULIER='NON'))

IMPR_RESU(RESU=_F(CARA_ELEM=elemprop,
                  RESULTAT=reslin),
          UNITE=80)

PS: Of course, to be perfectly correct, I must also account for the offset with the rotational degrees of freedom, I agree. But I figured, for the future, it might be easier to model unsymmetric sections for the beam (still need to figure that out) and make the beam elements on top of the plates directly...