Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

How to model Rubber properties ( Hyper elasticity in FEM) using Abaqus

Status
Not open for further replies.
Sep 7, 2016
4
0
0
IE
Hi All,

Can Anybody guide me how to model the rubber elastic plastic properties in FEM packages like Abaqus to create the same elasticity effect.

As rubber is viscoelastic or hyperelastic material. How one can define the properties in abaqus to replicate the same effect as material by selecting respective modelling technique in abaqus.

I am familiar with abaqus and I know about different hyperelastic and visco elastic model available in ABAQUS. I just want to know how would you use the uniaxial true stress strain curve in abaqus to define the material. Do you use compression stress-strain part of graph for yielding and use tension stress-strain part for elasticity. Or you define the graph with tension and compression at the same time to define the material property. Any one can give me the right way to model it please.

Kind Regards
Mohamed
 
Replies continue below

Recommended for you

Sorry rubber is a viscoelastic material, and I think that you will not replicate it on Abacus. This is is due the their nature that can be defined as a behaviour of a spring and an absorber on the same time.
Please see details on the Maurice Morton book "Rubber Technology", about this phenomenum.



 
The easiest material model to use is the Marlow model. Feed it your tensile data and you are off and running. If you are in a highly constrained application you will want to run a volumetric test to refine the poisson's value, elastomers are not perfectly incompressible but close. Make sure to run a single element test model into order to compare the model's response versus the test data before moving onto the part model.

But the data you feed it will determine the results. So if you do a simple straight pull, you will get initial elastomer properties that the application will see only once. Therefore your results may be too stiff. If you run a more complex test with built in cycling and pauses to allow the viscoelasticity to come into play then you will have data that more closely represent long term properties of the material. You can also run a viscoelastic test and run a visco FEA model. The typical problem with this is you have to pick what strain you are going to gather data for while the part model will be at various levels of strain.

Good luck.
 
Status
Not open for further replies.
Back
Top