Gap simualtion

[Lei Lei]

Dear STKOteam:

I'm simulating a four-point bending experiment just like Figure 1 shows.

figure 1

37eef323283d9d37cc1e1a9d4f9e1f0.png (35.15 KiB) Viewed 2828 times

However, there are three gaps in the middle joint of the separated curved beam.

figure2

90399de4d1cd98f15307509dc3e6f58.png (11.54 KiB) Viewed 2828 times

Is there any way to simulate the behavior of gaps? With the gaps, the two parts of the curved beam will rotate and then touch each other.

Looking forward to your reply!

[kesavapraba]

Yes, it is possible to achieve that task using zeroLengthContactASDImplex. Please refer to the webinar in the link below.

https://www.youtube.com/watch?v=tXA5guozrFU

[Lei Lei]

Thanks for your reply!

To use a 2D beam with a fiber section to replace the solid curve beam, what should I do to simulate the gaps' behavior?

Does the method you recommend still work?

Looking forward to your reply!

[kesavapraba]

Yes, you can still use the element by creating two auxiliary nodes at the vertex, where gap is simulated. It's because the beam element has 3 DOF at a node in 2D, but the zeroLengthContact element works with 2DOF in case of 2D problem. Hence, we need those auxiliary nodes. You can do it like:
1) create two auxiliary nodes (say, A and B) at the same location where the gap is present
2) connect left beam with A using equal dof interaction (only displacements)
3) similarly, connect right beam with B
4) use zeroLengthContact element between A and B

[Lei Lei]

Thanks for your reply!

1) create two auxiliary nodes (say, A and B) at the same location where the gap is present

Can I select two nodes on the ends of the beams instead of creating two auxiliary nodes?

2) connect left beam with A using equal dof interaction (only displacements)

Since the gap will rotate, Why don't you couple all DOFs of the auxiliary node and beam?

4) use zeroLengthContact element between A and B

What's the difference between zeroLengthContact element and zeroLengthSection element?

I guess I also need to define a fiber section for the zeroLengthContact element to simulate the construction of the gap.

Looking forward to your reply!

[kesavapraba]

Can I select two nodes on the ends of the beams instead of creating two auxiliary nodes?

No, it is due to the limitation of the number of degree of freedom imposed by the zeroLengthContact element. As already mentioned, beam element has 3 DOFs, whereas the zeroLengthContact can handle only 2DOF in 2D problems.

Since the gap will rotate, Why don't you couple all DOFs of the auxiliary node and beam?

Gap cannot rotate, it is the structural elements that can deform and close the gap and make contact with each other. Coupling of the DOFS between the auxiliary nodes and the beam ends has nothing to do with gap rotation, it is because zeroLengthContact element can deal only with 2 DOFs at a node. For this reason, the auxiliary nodes need to be created. Had it been some quad/triangle element in 2D, you don't need the auxiliary nodes.

What's the difference between zeroLengthContact element and zeroLengthSection element?

while both elements involve zero-length characteristics, the ZeroLengthContact Element is focused on modeling contact behavior, whereas the ZeroLengthSection Element is used for connecting points with a specific force-deformation relationship representation

I guess I also need to define a fiber section for the zeroLengthContact element to simulate the construction of the gap.

Contact element doesn't require such definition. Please watch the webinar by Dr. Massimo which I shared in the previous answer.

[Lei Lei]

Thanks for your detailed explanation！

Best wishes！

[kesavapraba]

You are welcome

[Lei Lei]

Dear STKOteam,

I'm trying to simulate this curved beam with a complex gap in the middle using a 2D beam element and a zero-length section element.


However, It always comes up with a converge problem and my monitor doesn't show the wanted curve. Where is the problem?

I defined the section property and material for both beams and the gap. Is my modeling strategy correct?

Because I wanted to get the joint opening-moment curve at the gap, I outputted the rotation of one node coupled with another and the moment of the gap element. But the results were so strange.


Looking forward to your reply!

[kesavapraba]

Hi, It is challenging to understand your exact requirement. However, I can suggest you to look at zeroLengthContact incase if it helps.