Can anyone in the community kindly explain me which is the difference between the contacts and give a simple example when can be used, one instead the other?
1. When is it normal to use a surface to surface contact?
2. When is it normal to use a surface to surface gluing?
Thanks for your time,
Solved! Go to Solution.
Dear Simcenter user,
the surfaces of glue contact never open up during the analysis while any regular ("non-glue") contact can open and close. Glue contact is used if 2 parts are meshed independently but should be treated as one part. Regular surface to surface contact is used if parts can move apart from each other.
Characteristics of glue contact:
Characteristics of regular contact:
Good luck with your analysis! Happy to assist further.
Martin (Siemens PLM employee)
First of all thank you for your quicly answer.
We are tring to simulate an pressfit using contact and is really time cost expensive. So we want to try it using glue contact, but we are wondering about how NX solve the initial penetrations. We can not find any parameter to controle it...
Do you know how it works?
I would add here additional question.
Then what is the difference between glue contact and RBE2 MPC ? What are advantages and disadvantages of using one or another ?
if you do a pressfit analysis with initial penetrations, glue contact is not the right way. Glue contact requires that the nodes are "glued" throughout the analysis, so no initial penetration is possible.
There is - however - a workround with glue contact that might help you. You can work with thermal expansion in such a way that you start with a glue contact but then you "heat up" the interior part in a static analysis in such a way that in an unconstraint situation this interior part would exhibit the required initial penetrations. This of course requires that you know how much temperature difference you have to apply (you can test it with FEM). Also, "temperature" is only used for "expansion" and has no other physical meaning in your problem. No slipping of the contact surfaces is possible even though this might happen if you do a "real" contact analysis. So it is a workaround ... with shortcomings....
A glue contact compares in its behavior to a series of RBE2s where each master node of a contact surface is constraint ("glued") with a separate RBE2 to the opposite element surface defined by the element nodes. By that you would end up with the same number of RBE2s that you have contact nodes - that is not convenient. If an RBE2 spans over several elements, the contact area would suffer an artificial stiffening, and that is not what glue contact does.
I realized that I should have been more precise in my post about comparing RBE2s and glue contact. While a modelling with RBE2 for each node would be one (painful) way to model glue contact, we have done a lot of work in our refinement strategy and connection process to minimize discontinuities. Describing the Nastran way of glue contact as a simple RBE2 connection is therefore not appropriate.
Glue's not going to help. Contact will, and to simulate a press fit, define a surface offset on the contact definition, this is exactly why it's there for.