Contact analysis- a general question

[nastranuser123]

Hi,

I have asked this question Siemens UG NX community as well.

But, since this question is a general question related to contact analysis, I thought of asking in thsi community as well.

My mesh is of master and slave contact edges is not a confirmal mesh (i.e. the nodes of master and slave edges / region do not coincide).

In that case, is it a good practice to define two contact pairs ? That is : one pair slecting region 1 as source and region 2 as target and then another pair with source and target sapped (i.e. region 1 as target and region 2 as source).

Is this a good practise?

Or, will it make the problem over-constrained? I guess it would suit the purpose as it would create a symmetrical contact.

Thanks.

 

[Penguineer]

You shouldn't need to define two contacts for a single location, that would be redundant. Simply define what you initially stated, with the smaller face as the source and the region it will come into contact with as the target. Under Definition -> Behavior you can set the contact to behave symmetrically, which prevents the contacting surface from penetrating the target and same for the reverse. I recommend just defining your single contact condition and observing the results with the Behavior set to "Program Controlled" first and see how it turns out.

 

[rickfischer51]

There is a good discussion of this in section 3.5 in the Contact Technology Guide in the Mechanical APDL Manual

Rick Fischer
Principal Engineer
Argonne National Laboratory