[PranjalNewton]

Hi there!
I am using Uniaxial Tension data. And fitting for Mooney-Rivlin constants. Will this be enough to model compression and shearing in the material or I need poisson ratio, Shear modulus and bulk modulus. OR I need more experimental data?

Thanks.

[Florin]

Hello Pranjal, if you want to model compression and shearing it would be good to have experimental data for these types of loading; this is because the response is not linear and the uniaxial tension is not enough. The material models in ANSYS will auto-compute the coefficients from the experimental data.

[PranjalNewton]

Originally Posted by Florin

The material models in ANSYS will auto-compute the coefficients from the experimental data.

Does this mean it will also fill in the shear modulus etc. data. (Which of course wouldn't be as accurate as using shear etc. experimental test data)

I don't want to just model everything simultaneously so I wonder if just Uniaxial data will suffice or I still have to define a constant G, poisson ratio and K separately.

Thanks a lot Florin!

[Florin]

Hello, if you do not need to model everything, then the experimental data for just axial load will suffice. The analysis runs with these data. The coefficients that ANSYS determines from experimental data will describe the axial properties; from what I recall no other data is needed.

[PranjalNewton]

Thanks a lot!

[Nabuchadresar]

Hello guys,

I'm trying to help a friend with your master's degree. He needs to compare a experimental data with a FEA data. We have a Uniaxial Tension data of a hyperelastic material. Usually I work with cfd, It's my first simulation with this materials.

So, I searched some videos, forums, papers, but I feel losting yet.

I input the data on ansys: strain and stress

and I choose the method (Yeoh 1st order), solve the curve fit and copy the calculated values:
And I have to compare the exp data with 12 methods [Mooney-Rivlin (3, 5 e 9 parameters); Ogden (1a, 2a e 3a order; Yeoh (1a, 2a e 3a order); Polinomial (1a, 2a e 3a order).]



Then I put this material on my project (Static structural).
I set the mesh:


Then, my boundary conditions: (dispacement = 195mm (like the max displacement on experimental test))


Set my Analysis settings: (most are set as default)



When I run my simulations I have the follow results:

But I don't know if is a correct, because when I compare to experimental data is so different:


I have no idea where the error is, if it exists. Or if it is problems in my engineering interpretations, I am still studying this type of material and its behavior and applications in FEA. Can someone help me? Tips, useful links, things like that?

Sorry for the long post and thanks for any help.

André Moura

[karachun]

1. Look like model curves don't fit experimental data on these pictures, but this is the main purpose of "Curve Fitting ''. if you can not fit experimental data using some models then don't use them.


2. In the experiment we measure deformation of straight parts of the experiment specimen. You should ask a detailed description of the experiment from your friend.

https://www.youtube.com/watch?v=6xrC4huWFfE

3. Hexa!!!! If you are familiar with CFD then you should know that structural hexahedral mesh is preferable than unstructural tetra mesh. Same principle works in FEA.

[Nabuchadresar]

Originally Posted by karachun

1. Look like model curves don't fit experimental data on these pictures, but this is the main purpose of "Curve Fitting ''. if you can not fit experimental data using some models then don't use them.


2. In the experiment we measure deformation of straight parts of the experiment specimen. You should ask a detailed description of the experiment from your friend.

https://www.youtube.com/watch?v=6xrC4huWFfE

3. Hexa!!!! If you are familiar with CFD then you should know that structural hexahedral mesh is preferable than unstructural tetra mesh. Same principle works in FEA.

1. Yes, this model is don't fix well, but I tried more models, some them fixed well (like a MR 9 parameters). But don't work. I'll try split the curve, each part with a method.
2. thank's
3. For while I'm using 2D simulation, If it's work I'll go to 3D (if needed). Thanks for the advice.

[karachun]

1. I can not explain why, but try to run 3D simulation instead of 2D. Some time ago I also ran a 2D axisymmetric simulation with MR 2 model and got odd results. Then I run 3D and everything is OK. It is definitely my error, not Ansys, because there are many tutorials/webinars of hyperelastic in 2D that run OK. But maybe you have similar errors.

3. Nothing actually changed. Quad mesh is better than triangular.