I want to model a rail as a flexbody in ADAMS. However to do this I need to create rbe2 on the top surface of the rail shown in red and create a translational joint between the rail and another part. This will make the rail rigid.
I think I did not pose my question properly. What I am trying to ask is:
Why do you need to create the rbe2 elements on the "whole" top surface? Why do you not create the rbe2 elements at, for example, where you have connections?
Well, as JSlatten mentioned, you can use RBE3 elements. For that AUTOMSET parameter needs to be set in the SOL 103 deck.
General note: when you define an ideal joint such as a tranlational joint on a surface, it is as if the whole top surface is glued t a single point in reality. Is this the case in the actual physical test rig?