Test data required to quantify hyperelasticity

[enda]

Hello,

I am new to the polymer FEM so my first question is a basic one. I am attempting to quantify the material properties of a hyperelastic material. I have performed a uniaxial tensile test on the material. What are the minimum number of other tests that need to be performed to give a good estimation of the material properties.

Abaqus 6.8 is the software I am using, I have already tried just inputting the tensile test data and using the ogden constitutive equation to represent the stress strain curve. When I then evaluated the coefficients it seems the stress values for the material in the planar results are too high. Are planar and biaxial tests required or can I get away with one or the other.

Enda.

[Jorgen]

Many people recommend that you use at least 3 loading modes, e.g. uniax, biaxial, shear.

The truth, however, is if you stick with a "simple" hyperelastic material model (e.g. Neo-Hookean) it is sufficient to use one loading mode!

-Jorgen

[Fracture]

Hi guy,
I am modelling epoxy under simple tension test using LS-Dyana, I fix one end of a say cylindrical model and I exert uniaxial force in the other end but my results are not correct, My freinds say me to exert displacement instead of force on the second end,
1-As I am new to this software I don't kknow how to exert displacement on some nodes?
2-I am not sure if this makes any difference
tanx

[Fracture]

I should add that I am modelling a cylinder of redius=1mm and length=100mm,
The epoxy is assumed to have isotropic elastic-plastic behaviour. Does anyone know if LS-Dyna is a good software to model this behaviour?

[Jorgen]

LS-DYNA should be a fine software for your application.

You can apply a the displacement boundary conditons by prescribing the nodal displacements at a node set.

Using Forces instead of displacements should make not significant difference. If you get strange results, then I suspect that you don't have an accurate model for the epoxy.

-Jorgen

[Fracture]

Thank you,
Do you think isotropic elastic-plastic behaviour is not suitable for epoxy, I use the following as entering mechanical properties( as in my deck file):
ro =1.1500E-6 g=1 sigy=0.5 etan=0.01 bulk=15

These values are not fix but I found them from different sites and papers.
If I exert an axial force in one direction (say z direction) the resultant sigma(z) is 1/10 of the calculated value I mean sigma=E*epsilon where epsilon=displacement/initial length

[Jorgen]

Isotropic elastic-plastic modeling is a good starting point. I recommend that you first try a single element under tenson to make sure you get the stress-strain response that like.

-Jorgen

[Fracture]

Thanks,
Sorry for silly question by single element do you mean that my model should have several elements but all should have the same type?

[Jorgen]

No, I meant a single element. That will run faster and will give the stress-strain response that you can check.

-Jorgen

[Fracture]

Hi
Thanks Jorgen, I used a single element but I don't know why my results are so silly!! A cube under tensile load(say in z direction), expands and contracts several times in x and y direc. in the first 0.02 ms (time=0.02) and then its behaviour become more reasonable. I have set endtime=1 and dt=0.001, As I change endtime to 0.1 or 0.01 this fluctuation in both x and y direction exists.

[Jorgen]

Based on your description I guess that you are running an explicit simulation with displacement-based boundary conditions. The effects that you are mentioned are sometimes caused by inertial effects at time 0.

-Jorgen

[Fracture]

Thank you, I thought exerting a force in fraction of a time may be the cause of this violation so I changed the end time to 1000ms, which means that the force will be exerted in one second so that we won't have this fluctuations and now the problem is when I ran a time demanding problem after normal termination of run, when I go to LS-PREPOST and open d3plot file it indicates this error message:

realloc failed in:state array
requested memory size=275515 kbytes

I am sure that my computer has enough memory, so I will be thankful if you show how I can solve the new problem

[Jorgen]

Hmm, I would start by rebooting the computer and then trying again. If that does not work I would contact LSTC support. It sounds like a bug.

-Jorgen

[Fracture]

Thank you,
I also contacted them and they suggested me to install version 2.4 of LS prepost and it did not work out I installed version 2.3 and I also installed the new version of LS-Dyna manager, but it had no effect. So I tried to reduce the number of out put slides in LS-prepost. I think a large number of slides like 100,000 slides are unnecessary.

[Fracture]

Hi,
Another problem,I try to define time interval using *Dbase=>Binary-D3plot
and I set DT. But As I run my model What I see in the running time is that dt has no relation with what I set (e.g. I usually define DT=0.01 but when I run the program it solves in increments of 8E-09!!) so the running take alot of time. How can I control time increment

[Jorgen]

The *Dbase command is for specifying the output frequency, isn't it?

There is another command for the time incrementation. I don't recall at the moment what it is. Check the manual.

-Jorgen

[Fracture]

Thank you,
I could decrease the run time by reducing the density value that has no effect on the final responses.
I have a new question ther are alot of models material in LS-dyna for plastic, foams, adhesives and cohesives and I like to model a PMC with epoxy matrix so can I use those materials recomended for foam, adhesive and cohesive for this purpose?

[Jorgen]

Yes, that should be OK.

-Jorgen

[Fracture]

Thanks,
But I had another question, I copy it here:
I have a new question:there are alot of models material in LS-dyna for plastic, foams, adhesives and cohesives and I like to model a PMC with epoxy matrix so can I use those materials recomended for foam, adhesive and cohesive for this purpose?

[Jorgen]

I cannot recommend a model without more information about the specific materials and the the applied loading conditions, etc :(

-Jorgen