Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Mystical contact problem

Status
Not open for further replies.

flowglow

Mechanical
Apr 30, 2014
14
0
0
FR
Hello everyone!

I am doing a nonlinear simulation of a compression test of a 2D geometry with round inclusions.
Matrix is modeled as hyperelastic material, material of inclusions is chosen to be a lot stiffer then matrix. Geometry is compressed between two analytical rigid horizontal lines. Contact between rigid lines and matrix has Normal Behaviour with Hard Contact pressure-overclosure and Tangential Behaviour with friction of 0.5. Contact between inclusions and the matrix is the same, but friction is equal to 0.9. Compression is imposed by displacement of a top rigid line.
I used Plane Stress triangular elements from standart Abaqus Library (CPS3: A 3-node linear plane stress triangle.)
The task seems to be easy (The geometry and mesh is presented on the first picture), but each time I run calculation, at the first step I have always the same error following by the same warnings:

Excessive distortion at a total of 25 integration points in solid (continuum) elements
Excessive distortion at a total of 3207 integration points in solid (continuum) elements

And after:
Too many attempts made for this increment

So, I started to investigate the problem, and what I found:
1) If I insert only few inclusions in the middle of geometry - calculation runs;
2) If I add any inclusion close to the geometry border - calculation fails on the first step with the above mentioned error;
3a) If I use geometry with all inclusions, but incline the top rigid line (so the contact starts at the corner) - calculation runs
3b) If the slope of inclined rigid line is small - calculation runs, but result is extrusion of the geometry from the area between of two rigid lines (in simple words - geometry flies away).


I am using Abaqus 6.11-2/Standart

Hope someone will help,
Than you in advance

Attachements:

Problem, initial geometry:
Case of a inclined line - resulting geometry:
Case of 3 inclusions - resulting geometry:
 
Replies continue below

Recommended for you

Don't use 3 noded tet elements, they're useless. If you spend a little time and partition the geometry you can used hex elements and get better results.

 
Thank you for your suggestions. I have few additional questions:

to corus: why 3 node elements are bad and useless? Are they less stable in the meaning of possibility of distortion?

to IceBreakerSours: what do you mean under "high quality" element? Fine meshing?

to all:
1) I understand the distortion of the element, but what does the distortion of integration point means?
2) Is it normal to have small penetration of master surface into slave surface?

Additionally I tried to model "compression plates" by deformable geometry: in this case calculation runs. I thought that normally rigid object can simplify the problem, not the opposite.
 
Status
Not open for further replies.
Back
Top