Push down test for shear behavior of adhesives

[als94]

Hello!

I am a young researcher, still learning how to use Abaqus. I would love to model numerically an experimental test made to measure the shear behavior of some particular type of rubber adhesive. The problem is that I am finding some difficulties in the definition of the interaction properties between the different materials.

I have to timber slabs, with a rubber layer on the inner side; between those there is a concrete slab that is pushed down during the test.

I already have tried to define the behavior as General contact with particular definitions on the contact surfaces, but seems that this make the convergence of the analysis impossible.

Do you have some hints or suggestions? It will be really appreciated.

Thanks in advance.

P.S.: In case it will be useful I can also attach the model file.

 

[FEA way]

Yes, please share the .cae and .inp files if possible. Regular contact likely won’t be appropriate here unless you want the parts to slide on each other. You will probably need a tie constraint or a different way to model the adhesive (like cohesive elements or surfaces).

 

[als94]

Unfortunately the slide between the different parts is exactly what I want to investigate. I also would have loved to define and measure the strains on the adhesive layer, that's why a choose a so detailed modelling technique.

I have also thought that maybe I can define a tie constraint and measure the displacement only in the rubber layer (is a 3D deformable element) but I am not sure that the results obtained will be physically valid.

Anyway, I attach here the .cae and the .inp file, hoping you could give me more useful suggestions on this issue.

https://files.engineering.com/getfi...4-4a96-ac16-ce81bc99e696&file=Cracovia2_2.cae

https://files.engineering.com/getfi...3-4f4d-a1e5-c6f5ee49d6cb&file=Cracovia2_2.inp

Thanks.

 

[FEA way]

If you can obtain proper data, using cohesive elements for the rubber adhesive might be the best way. Otherwise, you should just connect all the layers with tie constraints.

 

[als94]

Thank you for the reply. I find out that cohesive is the best solution in this case; unfortunately in this case is a bit difficult to have proper data on the stiffness parameters (Knn, Kss and Ktt) as well as for the Damage Initiation and Evolution.

Do you have some suggestion about some references where I can study better the definition of it? I had difficulties in finding in the manual

 

[FEA way]

Check the documentation section Elements --> Special-Purpose Elements --> Cohesive Elements (especially the chapter "Modeling with Cohesive Elements"). And look for some research papers covering the modeling of similar types of adhesives, maybe you will find examples of input properties and instructions on how to obtain them.