Viscoelasticity in MSC Marc

[Roham]

Greetings all,

I am trying to use gerenalized maxwell viscoelastic model for polyimide polymer films (Kapton) in MSC Marc . I have never used the prony series representation of the maxwell model (hereditary integral approach). Has anyone ever used MSC Marc to incorporate the prony series relaxation to the material? Here is my game plan for doing this:

1. I have the uniaxial relaxation curve for the polymer film, I obtain elastic modulus as a function of time for this model by curvefitting the relaxation curve and then obtain the prony series coefficients (I don't know how yet).

2. By assuming a constant bulk mudulus (time-independent dilatational response) I calculate the bulk modulus using equation K=E(0)/(3*(1-2*nu)).

3. I obtain prony series representation of shear modulus by using the equation G(t)=E(t)/(2*(1+nu)).

4. I input K and G(t) to the VISELPROP command of Marc.

My questions are:

- Does this procedure make sense?

- Can I still use the elastic-plastic scheme for an isotropic material using ISOTROPIC command or it has to be just elastic?

- Is there any easy way to get the prony series coefficient from uniaxial tensile test (Instron)?

Thanks a lot.

Roham.

[Jorgen]

I'm not an expert on MSC Marc, but here are a few comments:

- Your procedure sounds fine.

- I suspect that you cannot combine elastic-plastic with a Prony series. You will likely have to stick to a elastic (or hyperelastic) model.

- You can get the Prony series coefficients from a set of uniaxial tension test at different strain rates.

/Jorgen

[Roham]

Thanks a lot Jorgen,

I don't understand exactly how I can get the Prony series coefficients from a set of uniaxial tension tests at different strain rates. I'm a little confused. Could you please explain it more. I thought Prony series coefficients can just be obtained from relaxation data.

Roham.

[Jorgen]

The procedure would be as follows:

(1) guess a set of prony series parameters, and a set of hyperelastic parameters

(2) use the material parameters to predict the stress-stress response in uniaxial tension at the strain rates that was used in the experiments

(3) compare the experimental data and the predicted data, and based on this comparison modify the material parameters in order to improve the predictions. Goto (2)


This approach can be made completely automatic using a program like Matlab.

-Jorgen