ANSYS rubber modelling

[bijoyraj]

Kindly help me out in analysis of the following problem in ANSYS.

Model consists of alternate layers of steel ( total -9nos) and rubber pad(8nos) in between the steel plates.I am using uniaxial tensile test data from ASTM specimens (Hyper elastic Mooney Rivlin model c1=1.2ksc and C2=0.8ksc)

Loading condition1: Top steel plate is fixed and pressure is applied on the bottom steel plate which gives an axial compression.I am able to match the axial compression with actual test data.

Loading condition 2: alongwith pressure; load is applied that causes shear deformation of one plate w.r.t other.(Problem:There is a mismatch observed between the load actually required to move the top steel plate and the load computed from analysis.)

I understand that load required depends on the shear modulus of the rubber and ANSYS takes intial shear modulous as 2(C1+C2) from the tensile test data which gives the wrong result.Also If we increase the pressure applied on the plates then the load required decreases.(How to simulate the variable shear modulus with pressure in ANSYS?Is it possible?)

I have both tensile and shear specimen data for the rubber used.When tensile data is used to compute the C1=1.2ksc and C2 =0.8ksc the axial compression matches.But when shear properties are included c1 changes from 1.2ksc to 3.3ksc and c2 changes from 0.8ksc to -3.2 ksc.Convergence is not attained with this set.

Do we need to use biaxial data for this model?will the load match in that case?

[Jorgen]

Here are a few commends and questions:
- What does ksc stand for?
- How poor are the "alongwith pressure" results?
- I think it is possible that you do not have either the correct parameters for the Mooney Rivlin model, or the Mooney Rivlin model is a poor model for your rubber.
- You cannot easily make the shear modulus depend on pressure. It might not be necessary if you use an appropriately calibrated material model.
- How did you calibrate your material model?

[bijoyraj]

Here are a few commends and questions:
- What does ksc stand for?
- How poor are the "alongwith pressure" results?
- I think it is possible that you do not have either the correct parameters for the Mooney Rivlin model, or the Mooney Rivlin model is a poor model for your rubber.
- You cannot easily make the shear modulus depend on pressure. It might not be necessary if you use an appropriately calibrated material model.
- How did you calibrate your material model?
************************************
I am sorry that I confused you with the units.
ksc-stands for kg/cm2 (1kg/cm2 = 0.0981N/mm2).

Alongwith pressure results are matching using Mooney Rivlin constants.Sir can you suggest any other models which I can use?

I didnt callibrate the model.I had fit in the stress strain values obtained from tensile test specimens and derived the C1 and C2 constants using ANSYS.However this model was callibrated by one of my friends by modelling tensile test specimen in ANSYS.

I am new to polymer analysis and hope to get your support.Based on your advice I can also post the actual model used for analysis to give further clarity.

[Jorgen]

The Mooney-Rivlin model is notoriously sensitive incomplete experimental data sets for the purpose of calibration. In your case it might be that the C2 term is incorrect.

I would start by trying a simple neo-Hookean (NH) model. The NH model is more robust and might help with the predictions.

[msna]

I would suggest using shell elements for the steel layers and using solid elements to model the rubber between them. If you use a layered shell element (like SHELL181) you can do the axial load simulation but as the layers cannot slip, you can not use it for the sheer analysis.