creep/plasticity algorithm

[svufer]

Hi
I have a 3d model that consists of 3 parts with 3 different materials. One is elastic the second porous elastic and the last one drucker-prager linear hardening (shear). the element type i use is C3D8.the solution does not converge and the exact warning message is
creep/plasticity algorithm did not converge. i have tried to decrease the min increment size and the element type but the warning still exists. i have no other idea. Please help me.

 

[xerf]

Try to use quadratic elements, in this case C3D20...I think.

 

[PamCrash]

Hi,
please let us know more

1) What kind of creep model are you using ? Abaqus cards ?
2) Does the first increment converge ?
3) Implicit or explicit creep ?
4) Time hardening or strain hardening ?
:
:

Cheers
Pam

 

[svufer]

hi
i have tried to use quadratic elements but still the creep algorithm does not converge. the first increment converges. in fact the first step is completed successfully. i am using strain hardening model in drucker prager card of Abaqus. Thanks for your concern
svufer

 

[PamCrash]

Sorry, can't help on that small data basis.

Some basic tips

1) carefully get rid of all warning messages in pre and dat-file
2) try again with smaller load and/or with simpler material law
3) try stabilize-option with the *viscous card

 

[psantunes]

Hi Svufer!

As you must know the plasticity is a incompressible effect (no change of volume in the material during the deformation). If you are usig the implicit abaqus algorithm (Standard), try to use hybrid formulation in your elements. This kind of formulation includes a independent interpolation of pressure stress preventing non-convergence in almost incompressible or incompressible problems.

Good work!

Paulo

 

[svufer]

Hi,
First of all i want to thank you for your interest. i tried everything you suggested. i would like to inform you about the results.
C3D20 elements achieve better convergence but still the second step of the analysis is not completed. Also i have read that this type of element shouldn't be used with interaction properties.
Also my material isnot viscoelastic. i have a shear hardening of a soil material but also i have tried different material combinations and it seems that this isnot the problem.
Finally the plastic material is Drucker-Prager hardening and it is not incompressible. i tried though hybrid formulation for my elements but it was worse than simple linear elements.
I have made the same model in 2d plane strain conditions. i have still the same warnings about creep/plastcity algorithm but the analysis is completed.
i have no other ideas about the 3d model, so i would be grateful if you helped me again.
svufer