hyperelastic simulation with ls-dyna

[shiwencailx]

hi, everyone and Jorgen:
I am new member for polyerfem and am very glad to communicate with members here.

I want to validate one kind of hyperelastic material for future research, and downloaded experimental data of Natural Rubber(55pph CB) in our website.With ls-dyna software, I create one element(unit element, unit system is mm,ms,kn, gpa) to simulate uniaxial tensile phenomena. But I find results are not so good.
With hyperelastic material model in ls-dyna, keyword*mat_hyperelastic_rubber ,*mat_odgen_rubber and *mat_simplified_rubber, result in small strain(less than 0.1) is reasonable with experimental data, after it, stress seems too large.
would you please give me some tips to get right or better result? Thanks.

By the way, for two of three material models above in ls-dyna, another curve of stress relaxation is needed, do you have it for Natural Rubber(55pph CB)?

how to consider strain rate in simulation of hyperelastic model? I have checked some of your simulation results with different strain rate.

let me know if you need more information.
thanks again.
best regards.

[Jorgen]

Can you present a figure that compares your model predictions to the experimental data? No other data is needed for hyperelastic models. LSDYNA, like most other FE programs, allows the hyperelastic model to be combined with a linear viscoelastic model.

You can attempt to use a linear viscoelastic model to capture the strain-rate dependence. If that does not give you accurate enough predictions, then the Bergstrom-Boyce model is likely able to provide better predictions.

[shiwencailx]

I will present the figure later. Thank for your reply.

Would you please show me the range of linear bulk modulus for Natural Rubber(55pph CB)? What is about of specimen length, width and thickness? Thanks again.

[shiwencailx]

Hi, Joegen
Here is the figure.....

Note: Keywords in ls-dyna
Mooney_Rivlin (*Mat_Mooney_Rivlin_Rubber)
Mat77_Odgen (*Mat_Odgen_Rubber without stress relaxation curve)
Mat77_hyper (*Mat_Hyperelastic_Rubber without stress relaxation curve)
Mat_181 (*Mat_Simplified_Rubber)

Thanks.

[shiwencailx]

check attachment file please.

[Jorgen]

The figure seems to have to experimental stress-strain curves. Is that right?

The quality of the model fits is reasonable, and similar to what I would expect for a hyperelastic model.

I personally would not chase after a more accurate hyperelastic model. If the accuracy of the predictions are not satisfactory, then I would switch to a more accurate non-linear viscoelastic model, such as the Bergstrom-Boyce model.

- Jorgen