User login

You are here

Modelling Hyper-elastic Plate Using ABAQUS

J Mahmud's picture

Dear All,

For months, I have tried to model a plate with circular hole which has the properties of hyper-elastic material (Ogden) but the results are still not promising compared to literature (Y Basar 1998).  The main problems are:

1. The maximum load I succeeded to apply was q=26, while literature had applied load up to 90. For q > 26, the analysis will terminate with errors such as too many attempt, excessive distortion etc.

2. The results I get for q<26 are not the same as literature. 

I attach my best result for your viewing.

I wonder if any one has successfully solved this problem using any FE software such as Nastran, Ansys, FEBio etc... If not, your comment, guidance or any advice on this matter will be very much appreciated.

Thank you so much.

Kind regards,

Jamal

PhD Student, Cardiff University.

 

Comments

Hi, Jamal:

I think ABAQUS is already a good choice for this problem. I wonder which element you use for your simulation. And for the ogden model what is your n?

 

 

J Mahmud's picture

Hi Li:

Thank you for your comment and showing your interest in this matter. For your information, I have tried several element types (solid and shell) to simulate this particular problem. My main intention is to re-produce any published problem related to hyper-elastic plate using either Abaqus or Nastran. In this case, I chose Basar (1998) because it's the basis of my study. If I can validate my results, then I am confident to apply this skill in solving other types of hyper-elastic plate problem. Previously, I have never used Abaqus or Nastran to solve hyper-elastic problem.

I attach 2 new files for public view. One is the results using CPE8RH elements. For the Ogden, n=2 (Please refer to hyper elastic 2.doc). The other file is its *.inp file (Refer 2D Large deformation.inp_txt). You can run the *.inp file provided you rename it to *.inp.

Kind regards.

ruogang zhao's picture

Hi Jamal,

Are you sure their result is not based on Mooney-rivlin material? I tried your inp with M-R material, the deformation is quite similar to theirs. 

J Mahmud's picture

Hi Zhao,

I appreciate your effort in trying the inp. The paper I am referring to (Basar & Itskov, 1998) is about Ogden Material. A paper reporting their work using M-R material for the same problem was published earlier. (Basar & Itskov, 1997: Int. J. Solids & Structures, Vol 34, page 1687-1708). So, yes I am sure that their results were based on Ogden (1998) & M-R (1997).

Other than the material input, do you have any comment on the boundary condition, element type, step input, etc. All inputs were based on my own judgment. I couldn't confirm whether they were the most appropriate.

I updated the attachment (Hyper elastic 2.doc) which now shows the numerical results. I also included more references related to this problem.

 

Note:

I pay my great respect to the late Prof. Yavuz Basar (1935-2002). http://www.cimne.upc.es/iacm/News/memoriam.pdf

At first, I agree with your simulatioin results. I did run your model. It is like what you said. But that doesn't mean ABAQUS cannot handle this hyperelastic material model. For your problem if you want to simulation with ABAQUS I think you really need a good material model which should based on the material testing data, especially biaxial testing. So why we both got this result which is different from other people's data, I think it might be due to the ways each code use to interpret the material model during the simulaiton and the behavior of the elements we use in our model which are different from those used in the published results. In my opinion, it is better validate our simulation results with expirement result.

Besides, the following things we could do to improve our simulation:

1. Check the physical meaning of our simulation results. Is the strain and stress too high? Because I didn't use Ogden model a lot in our simulation, so I cannot tell you this strain or stress for this material is large or not. 

2. For BC, I think it is better to use symmetry, because we use 1/4 model. 

3. Check our load: is that too high for this material?

4. We still could run a whole model simulation instead of 1/4 model.

5. The way we use to apply load, (surface traction), is that reasonable physically? Could we use pressure or concentrated force in points ? 

 

 

Subscribe to Comments for "Modelling Hyper-elastic Plate Using ABAQUS"

Recent comments

More comments

Syndicate

Subscribe to Syndicate