using the Odgen material model , you need three differnt types of load vs stretch data for 3D models and only two types of load vs stretch data for 2D models.
1. tension-compression
2. simple shear
3. pure shear
The odgen model is quite complex but characterizes the rubber material better than Moodey Rivlin.
Suggest the Moodey Rivlin for near linear load-stretch curve.
If the rubber is really stretched into the S curve the Odgen is required. Use the minimum number of terms in the odgen series to chaacterize the material, since the solution stability decreases with increase in the terms used to characterize the material.

The Arruda Boyce material model is superior to Odgen but Algor does not have this feature. The new Gent material model is also less complicated than Arruda Boyce and just a good.

Thank you for your reply to my post.  What I have available for this material is the result of a tension test.  What I can do with this?  Mooney-Rivlin?  The deformations involve are quite large, so I probably are going to be away from the linear range.  Is there any check that I can do to find out when Mooney-Rivlin is no longer valid?  Thank you for you insight.

Regards,

Hector

Here's a slide presentation on the different methods:

http://www.hks.com/reference/epaper/elastomer_FEA.pdf

It states in this presentation to use Ogden if you have three different test data.

I performed this exercise with an Algor Application Engineer. I received test data from Supont on the rubber I was analyzing. They had three different test curves (stress-strain) for the material as sterndrive mentioned above. The curves were small, and the info had to be inputed manually into Algor. I called Dupont to see if they had more accurate curves for me to use. They were not so pleasant though.

Find a supplier of the material you will be using, and call to see if they had already performed the testing for this material. There are companies out there that will perform tests for you for a price.

oharag