Close this search box.
Clear all

Strain concentrations in elastomers

2 Posts
2 Users
0 Reactions
Posts: 124
Topic starter
Estimable Member
Joined: 17 years ago

Hi there,

I am working on a component made of steel and rubber.

At the inetrface between the two the strain gradient in the rubber is very significant.

The solution detail of interest is exactly the maximum strain at the interface, which I have tried to hunt down refining my mesh and checking the sensitivity.

The fact is that strains keep increasing, up to values in the region of 300%, at a point were my mesh is well too fine.

It made me think about modelling the detail.

In the presence of a singularity in a metallic component would just represent a warning on the appereance of a plastic zone, but with polymers?

Are those strain realistic or simply a product of the mathematical singularity? Do not the viscoelastic dissipation at such point start to represent a feature which cannot be neglected any more?

The same would happen modelling a crack. Using singular elements n ANSYS if needed even the singularity could be represented, but what is the meaning in a material which does not undergo plastic deformation? Is it a call for considering how relaxation will make the crack tip more and more blunt?

Thank you very much for any help


Topic Tags
1 Reply
Posts: 3998
Joined: 5 years ago

Hi Muzialis - Good to hear from you again.

300% strain for a rubber may not be too back, and may not lead to failure.

As you know, rubbers are certainly viscoelastic, and if you use (for example) a hyperelastic representation then the calculated strain may be too high since the material does not dissipate energy. I cannot say if in your specific case the strain magnitude is realistic or an artifact from the material representation.

I agree that for a crack problem the viscoelastic stress relaxation will make the stress and strain values finite, and the crack will blunt as a result.


Topic Tags
1 Reply