hyperelastic

manolo · 2014-03-16 11:10:08

Hi all,
i have to model a rubber pipe with internal pressure and traction along his axis. Above critical pressure pipe shows a bubble on midside.
I try to model the pipe with DKT modelisation and four node plane element in a 3d mesh.
I assign a ELAS_HYPER material tho DKT modelisation.
I run a STAT_NON_LINE problem with raising pressure until critical value but i get following message:
!--------------------------------------------------------------------------------!
! <EXCEPTION> <ELEMENTS_42> !
! !
! Le comportement matériau ELAS_HYPER n'est pas traité. !
! !
! Conseil : !
! Utilisez la commande DEFI_COMPOSITE pour définir une coque monocouche avec ce !
! comportement. !
!--------------------------------------------------------------------------------!
How can I assign ELAS_HYPER material to plane element in a 3d mesh?
Or I have to mesh the pipe with solid elements in a 3D modelisation?
Please see attachment with .comm file.

Thank you.

Attachments:
tubo01.comm, Size: 2.62 KiB, Downloads: 406

keeswouters · 2014-03-16 20:04:35

Hoi manolo,

I think the only shell element capable of hyper elasticity is coque_3d, not DKT.

model=AFFE_MODELE(MAILLAGE=m_tubo,
AFFE=_F(GROUP_MA='tubo',
PHENOMENE='MECANIQUE',
MODELISATION='DKT',),);

I cannot find the overview right now (ie too lazy to dive into the manuals), but I guess changing 'DKT' into 'COQUE_3D' will take you further (and donot forget to change the mesh into tria7 or quad9).

Last edited by keeswouters (2014-03-16 20:04:58)

jeanpierreaubry · 2014-03-17 07:58:09

hello

keeswouters wrote:

I think the only shell element capable of hyper elasticity is coque_3d, not DKT.
.

are you sure
the msg clearly states that one should use DEFI_COMPOSITE with a single layer for ELAS_HYPER law

jean pierre aubry

AsterO'dactyle · 2014-03-17 09:27:36

Quite right, it's a bug.

In fact, this message is for non-isotropic elastic material. In this case, you have to use DEFI_COMPOSITE.
But, in your case, ELAS_HYPER is isotropic material. There is a bug in dxmate.F90 subroutine.

manolo · 2014-03-18 20:38:14

Many thaks to all.
Yours hel is precious.
I will try with coque 3d element and i will not forget to add central node.
Bye.

manolo · 2014-03-20 21:24:14

Hi,
I have tried with coque 3d element but hyperelastic material is not supported.
So I passed to solid element that support hyperelastic non linearities and also large deformation.
Now the problem run and converge but the results do not show the bubble in the middle of the rubber pipe as the experimental tests do.
Next step will be control the pressure by "pilotage" method, monitoring displacement of a middle side pipe.
I will inform you about new calculation results.
Thanks again.
Bye.

Manolo

frotscher · 2014-03-24 10:35:40

Hi manolo,

if you're still interested in using the hyperelastic material on the COQUE_3D model, please read post #8 in this thread: http://www.code-aster.org/forum2/viewtopic.php?id=15686 . Some minor changes to the code do the job.

Regards,

Ralf

#1 2014-03-16 11:10:08

hyperelastic

#2 2014-03-16 20:04:35

Re: hyperelastic

#3 2014-03-17 07:58:09

Re: hyperelastic

#4 2014-03-17 09:27:36

Re: hyperelastic

#5 2014-03-18 20:38:14

Re: hyperelastic

#6 2014-03-20 21:24:14

Re: hyperelastic

#7 2014-03-24 10:35:40

Re: hyperelastic

Board footer