The Prony-series representation of shear modulus G(t) is used for the time -dependent behaviour of viscoelastic materials. These are dissipative materials.

Hyperelastic Materials are, to my knowledge of the definition, non-dissipative (conservative) Materials. The energy you put in you will get 100% out, if you release stress or strain. So, i guess the Prony-series-approach is not appropriate for your hyperelastic material model. But anyway, you need test data for G and K or you have to find some appropriate data in references.

Another question that comes to my mind, since you are talking about base-isolation (earthquake analysis?): In such kind of analysis you need dissipative materials, or? 

best regards

Dear Andreas; thanks before....

In some references, Prony model could describe the rate-dependency of rubber under pure shear load. Base isolation will subjected to cyclic loading (shear) in order to derive the stiffness and dissipation energy through hysteresis loop.

thw rubber is deformed into a larger strain territory for first time, it is very stiff but upon recycling in this same strain, the rubber softens dramatically. This phenomenon is often referred to as the Mullin's Effect. Then The restricted of FE comercial software, they could not present the mullin's effect, so the hysteresis loop will be zero stress at zero strain. That is why we need modeling prony series in order to obtain Mullin's effect on our FE model.

Rubber exhibits a rate-dependent behavior and can modeled as a viscoelastic material, with its properties depending on both temperature and time. When subhected to a constant stress, it creep. When given a precribed strain, the stress decreases with time; this phenomena is called stress relaxation. Collectively, these feature of hysteresis, creep and relaxation-all dependent upon temperature- are often called of "viscoelasticity" (see the texts by Fung-1965, Chritensen-1982, and Ferry-1970).

Dear Ariadne

I had read it. Thanks

it said that we could obtain initial shear (G) and Bulk (K) modulus through constant values (C10,C01 D2 etc).

My questions are :

Does it mean that I do not need input shear and bulk modulus due to the software have calculated them with contant (C10,C01,D2)?

What is mean "initial shear modulus"..? Will it remain constant or change after loading...?

What "initial shear and shear derived by prony series (relaxations test) are same? Because we could not conduct relaxation test (the apparatus is not available). So we need another way to define shear modulus and bulk modulus and model them into FE software.

Hello Yudha:

Remember, the software is allowing you to use one of its predefined equations/functions to define your material property and to represent your stress-strain curve in ANSYS. Therefore, you don't need to calculate G & K.  Now, regarding the viscoelastic(time) part, you need to run relaxation or creep test to simulate the time domain part. Since;  you don't have these tests available. Just run the model without taking the time domain in your simulation and see what happened to the simulation. I mean compare simulation with your experiment. If both correlates very well with reasonable error, then you are okay, otherwise you have got to run relaxation test. Since your material is hyperplastic, it could be sufficient not to include your visco part in the simulation but you have to prove it!.

BR

Mohamed F. Alzoubi

Dear Mohamed F.Alzoubi

In other word, you said the equation of hyperelastic model have been calculate the G and K automaticly? Yes, we will verified our analysis with experimental results, but we can not make a speciemen (base isolation) before make sure our FE Analysis is complete...because it will take a lot of cost.

I thought, I need looking for the reference related to hyperelastic material model which explain those more detail including the values of G & K on their equations.

BR

Yudha Lesmana

Dear Mohamed F.Alzoubi

it means that I MUST carry out relaxation test for modeling shear and bulk modulus in hyperelastic material properties?

The equations of hyperelastic model (such as Neo-Hookean,etc) they only represent initial shear modulus (G) and initial bulk modulus (K). I worry if it will not represent the real behavior of model.

Dear Yudha,

I have done base isolation analysis in Abaqus. Basically my objective was to analyze the isolation characteristic of rubber component with respective to the frequency. My observation is, without viscoelastic property damping effect can not be simulate. So to exactly get the transmissibility and isolation efficiency of rubber, viscoelastic property is needed.

Instead of stress relaxation test or creep test, Dynamic Mechanical Analysis test (DMA) can be also performed to give material property input.