Suitable shell element for double contact in Abaqus/Standard

[Sia27]

Can someone please guide me?

What shell element is suitable for double contact modeling in Abaqus? I am trying to model contact between 3 surfaces using contact pair, surface-to-surface. The element in the middle will therefore be modeled as having double sided contact. I have been trying it with S4R but the analysis doesn't seem right.

Thanks in advance.

 

[prz235]

Hi!
What version of Abaqus do you use? And what kind of contact you used with your shell elements?

 

[Sia27]

Hi,

I'm using Abaqus 6.12-2

It is a surface-to-surface contact between 3 steel surfaces. I modeled all 3 surfaces using deformable S4R shell element. The surface in the middle acts as a double sided contact. For instance, the middle surface (assigned as Master1) and the surface on its right side (assigned as Slave1) forms Contact1. And then, the same middle surface (assigned now as Slave2) with the surface on its left side (assigned as Master2)forms Contact2.

I ran a buckling analysis and looking at dat. file & msg.file, it seems that my problem is with Contact2. I think Contact1 works fine and the surfaces interact correctly but then following that interaction, the same middle surface (used in Contact1 and 2) may have deformed a bit too much such that the slave nodes on it finds no intersection with the corresponding Master 2 surface on the left side. I know this because I have the following warning and the said node (Node 177008)is located in the middle.


***WARNING: SLAVE NODE 177008 FOUND NO INTERSECTION WITH THE MASTER SURFACE.
THE INITIAL CLEARANCE PROVIDED FOR THAT NODE IS IGNORED

Below is the code for Contact1:
**
*CONTACT PAIR, INTERACTION=CONTACT1, SMALL SLIDING, TYPE=SURFACE TO SURFACE
S_CONTACT1, M_CONTACT1
*SURFACE INTERACTION, NAME=CONTACT1
*SURFACE BEHAVIOR, DIRECT, PRESSURE-OVERCLOSURE=HARD
*FRICTION
0.0
**
*CLEARANCE, SLAVE=S_CONTACT1, MASTER=M_CONTACT1, VALUE=0.0000
**

And below is the code for contact2:
**
*CONTACT PAIR, INTERACTION=CONTACT2, SMALL SLIDING, TYPE=SURFACE TO SURFACE
S_CONTACT2, M_CONTACT2
*SURFACE INTERACTION, NAME=CONTACT2
*SURFACE BEHAVIOR, DIRECT, PRESSURE-OVERCLOSURE=HARD
*FRICTION
0.0
**
*CLEARANCE, SLAVE=S_CONTACT2, MASTER=M_CONTACT2, VALUE=0.0000
[highlight #EF2929]*CONTACT INITIALIZATION DATA, INTERFERENCE FIT, NAME=CONTACT2,
SEARCH ABOVE=10.0, SEARCH BELOW=10.0
**[/highlight]

The difference between the two codes is that I just added the *CONTACT INITIALIZATION DATA command to fix the problem of slave2 finding no intersection with master2. But it seems to make no difference. Did i miss something? How can i fix this?

Thank you so much.

 

[prz235]

Hmm, I'm not an experienced engineer, but I did some analysis with double sided contacts on shell elems, on 6.13-4.
Firstly, I was always told to be extreamly careful with SMALL SLIDING contact forumulation, especially on curved surfaces, so I'd adivise trying finite sliding.
Secondly, last time I did 2sided contact on shells, I used general contact (which in fact uses surf2surf, finite sliding approach), which, according to documentation, "uses sophisticated tracking algorithms to ensure that proper contact conditions are enforced efficiently".
That's all I have in mind by now,
hope it helps.
BR!