Modell for a thin walled pc/abs plastic

[Fredrik]

Hello

I am working on my Thesis work, concerning modelling of a pc/abs polymer. The specimen thickness is about 1 mm and I have large deformations with high strain rates. I am going to use existing models in Abaqus and thinking about following:

I plan to combine a hyperelasic-viscoelastic model with a plastic model, using an overlay method. Do you have any comments upon my choice?Which hyperelastic model is suitable for my type of plastic? The model should capture different strain rates as well as the Bauschinger-like unloading.

Thanks for a great home page!

/Fredrik

[sq]

Good luck with your thesis, it sounds like interesting work!

I'm not sure why you're going with a hyperelastic model, it sounds like your polymer would be glassy. Do you expect large elastic strains?

To answer your second question, in general, there are few rate-dependent hyperelastic models. Jorgen Bergstrom wrote what is probably the best one in existence. There is also Johnson, et. al., Drozdov-Dorfmann (which I find wicked confusing), and many applications of BKZ theory (of which I have found none suitable to high strain rates). Of course, there are lots and lots of examples of people adding linear viscoelasticity to some hyperelastic model (i.e., Prony series), which is in some cases a good-enough engineering approximation.

[Fredrik]

I do expect large elastic strains so hyperelastic part is necessary.

I am planing to add a linear viscoelastic model to a hyperelastic model, my question is which one is preferable for a plastic like PC/ABS. (i.e. Neo-Hooke, Yeoh etc.) What are the behavior differencies in these models?

[sq]

All of these models exist to try and predict the same physical behavior, so it's hard to say what the "behavior" differences are, except in terms of stability, number of parameters needed, things like that.

My best advice is to look up the excellent review paper that Drs. Boyce and Arruda wrote for Rubber Chemistry & Technology (I forget exactly when, maybe back in 2000 or so).

[Jorgen]

Here's the reference to the Boyce-Arruda paper: Boyce, M.C.Arruda, E.M.: Constitutive Models of Rubber Elasticity: A Review. Rubber Chemistry and Technology, 73. (2000) 504-523.

I am somewhat confused by your approach. What temperature range are you interested in modeling? If the temperature is less than Tg then a combined hyperelastic + linear viscoelastic will only be accurate for strains less than about 5 - 10%.

For temperatures above Tg then your modeling approach can be accurate to larger strains, but then you will need to watch out for inaccuracies cause by the small range of strain-rates for which linear viscoelasticity is accurate.

In any case, I agree with sq that the choice of hyperelastic model will not make much difference for your predictions. Like many others, I tend to use the Arruda-Boyce 8-chain model as a basic foundation when modeling these kinds of materials.

- Jorgen