ansys simulation of silicone rubber

[Indigo456]

Hello altogether,

I am new here in this forum. My probleme is a thin membrane (circle) made of silicone rubber. A pressure is applied from one side and I am interested in the deformation of the membran.
Well, the difficulty for me is the characterization of the right model for silicone rubber.
I am going to use the model data for silicone rubber that is provided here.
I am working with Ansys workbench, so I tried to get the parameters for the Neo-Hookean Model by fitting the uniaxial data.
I get for the constants: C1 = 0,18 MPa and - of course - the deformation is zero.
Then I tried the Mooney-Rivlin Model. The results of the fitting are: C1=0,1437 MPa and C2=-2,03e-2 MPa.
Actually, which part of the curve (uniaxial compression) do I have to fit? The lower or upper part?
And anyway, are these constants of the right range?

As you can probably see, I am very new to fem-simulations. Is there anything I have to consider when I am doing non-linear simulations with Ansys workbench?

Thanks for your help in advance,
best regards,
Indigo

[Jorgen]

Sounds like you are on the right track! I would also start with the Neo-Hookean model, since that is the simplest hyperelastic model. I don't understand, however, why you say that the deformations are zero.

Ideally, the material model that you select should be able to fit the experimental data both at small and large strain. If it does not do a good job at all strains then I would consider switching to another more accurate model.

The material parameters that you quoted seem reasonable.

- Jorgen

[Indigo456]

Thanks for your response, Jorgen!

There are just a few things which are not really clear to me.

The data you provide on polymerfem.com for silicone rubber is unaxial compression.
In my problem I am going to strech the silicone membrane. Do I need data for tension or can I used the compression data to calculate the material parameters?

Or does it even matter if I use data of uniaxial compression or tension? Do I get different material parameters or is it all the same?

Concerning the parameters for the Mooney Rivlin Model (2 parameters). Do they have to satisfy certain conditions? Are negative values ok?


Actually I tried fitting the uniaxial data to a Mooney-Rivlin Model. I got the attaced curve fit (M_R_2_both.jpg), and - as far as I can see it - the fit is not very good.
Or do I only have to use one part of the uniaxial compression data, e.g. the upper part?

Best regards,
Indigo

[Jorgen]

You can use either compression and/or tension data for calibrating the material model. Typically one gets very similar parameters for tension and compression (at least for rubbers).

With improper values of the Mooney Rivlin parameters the model can become unstable at large strains. Most FE programs will print a warning message if the model is unstable.

I agree that the fit is not very good. To get more accurate predictions you can either use another hyperelastic model, or a more advanced material model such as the Bergstrom-Boyce (http://polymerfem.com/index.php?pageid=umats) model.

-Jorgen

[Dan]

Hey,

I saw your post on the silicone rubber. Do you have a product name associated with your data? I have a project I am working on where your material would fit quite nicely.

Thanks,

Dan

[Jorgen]

I am afraid that I don't have a trade name for that material.
The data is a few years old and I don't recall the name.

- Jorgen