Simple hyperlasticity

[blazzzzzz]

Hi,

I am trying to model an intervertebral disc, in compression.
So for now it's just a rod with 2 thin endplates on both edges, one is fixed and one has a pressure load compressing everything.
I am heading towards making it more realistic in terms of geometry and ideally it will porohyperelastic.
But for now Im just getting familiar with ABAQUS and want to make it just hyperelastic. I am using Mooney Rivlin equations and putting the coefficients into ABAQUS. My question is, when I put elastic properties the job is completed, no problem. But when I put hyperelasticity, the job is aborted but gives results anyway (negative eigenvalues, and excessive distortion in the msg file). can I solve this? how much can I trust the data of these aborted jobs? I also noticed that if I put a lower pressure load, then the job is completed.
Can anyone help me? I hope I was clear enough.
Thank you very much

Ben

 

[GBor]

I'm not an Abaqus user, but it sounds like you don't have a small enough time step in your analysis. You probably can't trust the aborted results at all.

Basically, the load is moving the upper plate too quickly and passing through the lower plate before the next time step can be calculated.

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group
Magnitude The Finite Element Analysis Magazine for the Engineering Community

 

[blazzzzzz]

Thank you very much for your reply. I have tried decreasing the time period in the load step, and also change the increment size. But it just wouldn't work.
If I apply a load magnitude of 2 it doesnt work, but 1 works....I dont get it.
Thank you anyway

Ben

 

[blazzzzzz]

Actually with a magnitude of 1, the jobs says its completed but I just checked the msg file, it also says it found negative eigenvalues and excessive distortion.
Actually a job of 2 or 30 or anything more would give the same result. So somewhere between 1 and 2 it just stops calculating....
If anybody has an answer that would be great, thank you

Ben

 

[prost]

Trying ramping up to a much smaller force, say 1/100th of the force you want.

 

[blazzzzzz]

well what I call magnitude is the force. So it works if I put 1, and not 2, actually I would need a lot more than that.
I tried to put 0.05 and it works also, but the deformation is very very small of course...

 

[prost]

From your description this looks like the classical case of flat end poker or indentor bonded to a flat surface. This is a very difficult problem to solve analytically, in fact my research into this problem 10 years ago discovered no analytical model for this special class of material. The reason why you would like an analytical model for a similar situation (for instance, instead of your particular structure, say for the case of a flat end block indenting a flat surface of body composed of this stuff) is that the analytical solution tells you so much about what to expect with the finite element modeling. Because there were no analytical solns, I also did some FE simulations of a flat end block indenting a flat surface of body of Mooney material. If the block is bonded to the flat surface, you see VERY LARGE rotations and deformations in the mesh of the flat body underneath the corner of the block with the flat surface. The mesh gets very distorted there; we tried hundreds of different mesh geometries at the corner; what appeared to work best was mesh that has triangular elements fanning out from the corner; this allowed some rotation and cut down on the distortion to a point.

 

[blazzzzzz]

you seem to know quite a lot about it, I actually thought I was doing something very basic.
I have tried the tet meshing, the job is completed and does not give any error message. Even if I put a force of 50.
BUT the results show almost no deformation. And if I put a bigger force (100), then it doesnt work. But 50 should be enough to see a big swelling....which is by the way what I see with my aborted jobs and the hex meshing, the results show the cylinder really compressed.
I dont understand why meshing changes everything so much.
Thanks for helping....

Ben

 

[dollarbulldog]

Hi

Why don't you test it first with some more simple example? In that case, may be you can grasp its behavior before heading to the actual one.