CONTACT Problem

[hacinilyes]

HI,
I have some problem when modeling contact with TARGE69 and CONTA175. Target elements are penetrating too much the contact elements when using keyopt(2)=1 (penalty algorithm).
Any suggestion to solve the problem?
Thanks.
Lyes

 

[prost]

Try to back off on the loads--do the bodies penetrate for really small loads?

Also, I don't know your particular contact algorithm, but I can guess a similar approach is used with this one--many contact algorithms allow the Target to penetrate the Contact Body, but won't allow the Contact Body to penetrate the Target (or vice versa).

 

[salmon2]

The commercial FEA code will re-mesh according to the default criteria or the overlapping limits set by users. You also should be asked before which one is rigid and which one is deformable. Tighter criteria will limit penetration too much, but on the other hand will increase iteration/solution time should be set very carefully.

 

[kemechial]

Hi,

Maybe if you try Augmented Lagrange algorithm for the contact it can work.

 

[meshparts]

Hello,

I am with kemechial, Lagrange could work better. But if the Penalty-Method is mandatory, I would increase the contact stiffness and decrease the step time.

Regards
Alex

 

[Stringmaker]

I agree w/ the two previous posters. Also, try creating symmetric contact if those two suggestions don't work. The Augmented Lagrange algorithm is very good however and should solve your issue.