creep behaviour of adhesives in bonded joints

[miguelmiravalles]

Hi everyone!
I´m studying the creep behaviour of bonded joints, working with epoxy adhesives. In order to know their properties and behaviour some short-term lap shear tests will be carried out and then a FEM analysis should be made.
I´ve checked in Abaqus if I could use the command *creep to model the long-term behaviour, but the problem is that it is only useful for metals, and some reports I´ve read have not been able to create a subroutine that could represent the long-term behaviour accurately. Is there any other program or any written code to represent what I´m looking for?

I suppose that there is not so much work done about that, so I´ll probably focus on the tests and try to represent the stress and strain distribution of the bonded joint. If I just represent these things, which program would you reccomend me to use?

Thanks very much for your attention

[Jorgen]

Depending of the strain, you might be able to use a linear viscoelastic material model to capture the creep. If that is not accurately enough, then you can always use a specialized user-material model. I have modeled creep of thermosets rather accurately using a new model that I am developing.

No matter what model you select, you still need to run some experiments on your epoxy adhesives. What experimental data do you have?

- Jorgen

[miguelmiravalles]

we are planning to run some tensile tests in the bulk adhesive, under different loads, and then, if it´s possible, try to model the behaviour of a joint

[miguelmiravalles]

I´ve been thinking about the experimental data and I have realized that if tensile tests are carried out at different percentages of the ultimate load, it would take a lot of time because the testing machine can only carry one experiment at a time.
If bending tests were carried out, I think it would be faster.
Would we get the same experimental data from bending tests in order to get the parameters that will be used in the model?
Thanks a lot

[Jorgen]

Bend testing is not used as often as uniaxial testing when it comes to finding material parameters. The reason for this is that in bending you have a non-homogeneous stress field. That is, the stress is different different through out the thickness, which makes it difficult to calibrate a material model.

If you can, I recommend that you perform uniaxial or shear experiments.

- Jorgen