CELAS3 vs RBE3

[sych]

I'm modeling a fastener in FEM, which of the following should be used? DOF spring or RBE3. My understanding is, RBE3 gives a more realistic results. Please share this knowledge and if you have any reference, that would be great too. Thanks.

 

[fbdandloadpath]

Sych,

I recommend performing some tests, or at least getting access to some test data, then try modeling it both ways. You'll come up with a defendable position that way. This question has been circulating around Nastran (and other) user discussions for years, but I'm always convinced by analytical correlation to test data.

FBDandLoadPath

 

[SWComposites]

It depends on a lot of things; like a) are you connecting 2 layers of shell or solid elements? b) are you connecting 2 nodes, a line of nodes along the fastener axis, or the boundarys of actual holes in the two plates? c) what results do you want to recover? d) do you want to include the preload clamp-up? etc. You need to be very specific and detailed in what you are trying to do.

 

[SWComposites]

Yes, correlating to test data is the best approach. Though getting good and useful data from a joint test article is very difficult. If you don't have any test data, then at least make a very simple one fastener model and correlate the joint displacement/stiffness to one of the closed form fastener flexibility formulas in the literature (there have been postings about such methods previously).

 

[fbdandloadpath]

Agreed - a search of the Eng-Tips Forums for "fastener flexibility" will result in many approaches to this question as well as links to great references.

 

[sych]

Thank you all for your suggestions. Yes, I have read quite a lot of postings on different forums and I dont get a specific way to model it.

SWComposites: I'm new in using Femap and Nastran for modeling and getting result. Thus, I'm modeling a simple model by connecting two plate elements using nodes.

Thanks