Catia V5 connection properties

[PerKr]

I'm doing a series of FE analysis on a plastic shelf with steel reinforcement bars and just today noticed an annoying error I made: I have been using a "rigid connection" between the reinforcement bars and the plastic. As I understand it, the rigid connection results in an infinitely thin surface which is infinitely rigid acting between the connected parts. So now I'm running a couple of them using "smooth connection". I suppose "rigid connection" would be used for things such as welded connections in a spaceframe? The Catia documentation really doesn't explain when to use which connection type, so if anyone could explain this in relatively easy-to-understand terms that would be great.

 

[fl11]

PerKr,

Basically, CATIA creates a null–length rigid bar at the middle point between the centroids of the two surfaces to connect. Then, each extremity of this null-length bar is connected via a rigid spider (equivalent to MSC/Nastran RBE2 element) to the nodes of the surfaces. Being a rigid spider means that the nodes on the surface will not deform releative to each other. A round hole will remain round. (displacement of independant node = displacement of independant nodes)

For the smooth connection property, the nodes on the surface will be able to deform relatively to each other. Instead of being a rigid spider, it is a smooth spider (equivalent of RBE3 in Nastran). Therefore, the round hole can now deform and becomes flat for example. (displacement of dependant node is a weighted average of the displacement of all the independant nodes)

The question now is what is more representative of your structure. Smooth or rigid could be a good assumption depending on what you think is closer to "reality"

Changing from smooth to rigid can change the load distribution as well as stress distribution.

You could also use a user-defined connection properties which allow you to define separate condition on both surfaces and how they are connected to each other like Rigid - spring - Smooth. (meaning that there is a ridig spider on one surface, a smooth spider on another surface and they are both connected using a spring element with a defined stiffness.)

I hope it is not too much confusing and that it answer your question. And yes, the Catia Documentation is not very clear, so sometimes, you need to make a lot of test case just to understand what exaclty does the different icons.

 

[PerKr]

in this case, I have 2 cylindrical steel rods completely encapsulated in a plastic structure, simply supported at both ends and carrying a load of 200N. the rigid connection results in virtually no deformation at all for neither the plastic nor the steel rod. the smooth connection seems to result in an excessive deformation of the plastic but the deformations of the steel rods seem somewhat more correct.

I suppose I could iterate my way to the most suiting connection property. All that's needed is time, patience and a suitably simple model to work with. Guess that will have to be on my to-do list

 

[fl11]

That is right. The reality is somewhere in between. The only thing that you can do to have a better notion of what is really going on is to use a non-linear solution with contacts. (Not possible in Catia) but this is anoher game. But is it really worth the cost of it?