Long term/Instataneous

[Bugz]

Dear all,

Several questions regarding rubbers and their characterisation.


How do you choose strain speed rate when you make tests to characterize your material ? How can determine if your data are long term before relaxation test are made ?
I mainly refer here to the definition used in Abaqus for Long term and instantaneous modulus .

To my knowledege, the relaxation never stops, rubber will always relaxed so how deal with it ...

In Abaqus , you can fit your data with all different types of Strain Energy Potential ? Are you selecting the model for which the best fitting is obtained or is there others criteria that should be taken into consideration ?

Thanks for your answers ,

Bugz

[Jorgen]

Hello Bugz,

The most appropriate experimental strain rate will depend on the strain rate that is used in your intended final application. It is typically a good idea to use a similar strain rate in your experiments to the strain rate that is actully occuring in the real product.

The long term and instantaneous moduli can be approximated from experiments performed at very fast and very slow rates. In practice, it is often better to use data at different intermediate strain rates and then find the necessary material parameters from the experimental data that you got.

There are certainly many different hyperelastic model available to chose between. It does not matter too much which model you use, as long as you use appropriate material parameters. One thing to keep in mind is that some of the hyperelastic models are not unconditinally stable. Hence if you are not careful you might end up with really crazy results. A good model to start with is the Arruda-Boyce 8-chain model since it is always stable.

In thing that you should think about, and you did not mention this, is what overall constitive model are you planning on using? Are you going to be satisfied with a hyperelastic model? are you considering a viscoelastic model or something even better?

Jorgen

[Bugz]

Dear Jorgen,

Thanks for your reply.

Yes I plan to use a more sophisticated material model. Nevertheless, there are quite unknown parameters that make my characterisation of materials difficult. In fact my aim was to use FEM more or less with an iterative approach. The application works with friction, but the value of friction is not known with a strong confidence.

It is therefore very difficul to know at which strain rate the material is deformed, and what is the cylic loading during service.

For sure, the Mullins effect will be taken into consideration because we know that during the first loading the friction is quasi negligible. After the friction becomes extremely random, and stick-slip phenomena occured.

I should observe a softening of the material in general. so it makes sense to my mind to incorporate Mullins effect. For the hysteresis, if I refer to your article might also be a huge contribution in my application.

Do you know what material properties are predominant to extract for study application with friction ?

I would also include relaxation and damping:
Does relaxation in pure shear in time domain gives an indication of the damping properties or only frequency data do ? What exact information give the huge drop observed in rubber when relaxation test are performed?

Thanks again for your time and attention,

Cheers,

Bugz

[Jorgen]

Typically, you want to separate out the friction from the experimental data, and study the inherent material properties separately. Note that friction is not an inherent material property, but an interaction property between two material surfaces.

Yes, relaxation in pure shear will give a good indication of the damping properties of the material -- you do not have to perform frequency tests.

About the "huge drop in relaxation tests", I am assuming that you are refering to the stress relaxation that occur during interrupted loading. The reduction in stress is caused by the delayed response of the molecular microstructure of the elastomer, and can be rather accurately modeled.

Jorgen