1D beam to 2D plate connection

[Wnuczko]

thread1234-338079

Hi all,

I therad mentioned above the connection 1D beam to 2D plate is done by RBE2 element.

But if end of my beam element is touching my plate element can I force Femap to have coincident node there and no RBE2 connection? Just like when I have 2 beam elements at 90 degrees with coincident node.

Brgds,
Wnuczko

 

[crostolo]

Hi Wnuczko,
yes you can but keep in mind that the nodes of plate elements doesn't have the DOF R3 (the drilling one, rotational degree of freedom along the perprendicular axis of the element), so your beam element would be free to rotate along its axis. Thta is why a 1D element is connected with RBE to a 2D one.

 

[Wnuczko]

Thank you crostolo, but I see my approach is completely wrong.

I see having 1D and 2D elements in connection is not a good idea for checking what is the situation at connection area. Look picture attached.

My RBE2 spider is too small. Would be better to select the nodes that are close to the square tube walls. But then changing the profile enforces to change the spider again - not very efficient.

Do you guys have ani other idea than using only 2D plates and add them to non manifold?

Rgds,

 

[rb1957]

i think the best approach is to model the tube as 2D elements (rather than as a 1D beam).

how does the square tube actually interface with the plate ? welded along perimeter ??

if you must keep 1D beam, then your RBE should come to the footprint of the square tube, and you should have different nodes for the beam on differenst sides of teh plate (so the load goes from the 1D beam thru the RBE across the plate thru the 2nd RBE into the second 1D beam)

another day in paradise, or is paradise one day closer ?