about rigid connections in nastran 3

For linear static analysis in NASTRAN, the RBE2 and RBE3 elements are different methods for transfering nodal displacements between grids. The RBE2 element rigidly "ties" the independent and dependent degrees of freedom together in the directions specified. The RBE3 is different. This element enforces a weighted average of motion to the dependent degrees of freedom. There is no rigid stiffness imparted on the finite element model from an RBE3. Hope this helps...

-George

hi george,
i have a dought can u replace RBE3 with MPC ie mutiple pt constrants in a linear analysis.? and is there any different behaviour of RBE3.? u said about linear analysis only what about nonlinear analysis.? what the independent and dependent criteria of nodes in RBE3. ? thnxs

Hi,
If you have defined RBE3 or RB2 then NASTRAN will automatically create an internally MPC card, so you don't need to specify any MPC request in the case control section.
I see that george has said "For linear static analysis in NASTRAN" not "linear analysis only". To my know there is no rule for MPC or R-Type elements to be valid only for linear analysis. In fact, each constrain equation in that elements expresses one dependent DOF as "linear" function of the independent DOF. Any suggestion?

regards

Hi Zuardy,
thanks for clarifying my dought but i had not understood what u said about dependency and independency in ur last line.
for analysis we create spider around the hole so we create it by RBE2 and RBE3 .by RBE2 we use multiple points. for that we select a single node and then we select nodes around the hole. and we say the behaviour of around nodes dependent on first one. right?
so can u explane the same in RBE3 .because while selecting we come across dependency and independency.

-chakri thnxs

Hi c89666,
you are right about choosing nodes for RBE2 and actually it is the same for RBE3. But in DOFs level there is a different between RBE2 & RBE3, In RBE2 the independent DOFs (on the single node in the middle of hole) are all 6 DOFs, but in RBE3 you should specify the apropriate DOF(s) and this DOF(s) is called REFC(reference DOFs component), besides that you must specify the wighting factor. May be your pre-processor refers this DOF also as independent DOF.
Another therms in NASTRAN for those DOFs are m-set for dependent DOFs and n-set for independent DOF.
Hope it can help you

regards

Hello folks,
I like zuardy's post for the explanation of the connection of a node at the center of the hole. In both cases (RBE2 and RBE3) the degrees of freedom at the center node DEPEND on the motion of the surrounding nodes. Also, with both you can specify any DOF that you wish. For example only in-plane motion and no out of plane motion. One important thing to note. The RBE3 element is not a rigid element. Like zuardy's post says, it uses a weighting function. The nodes surrounding the hole are free to flex with the material and the motion of the center hole is averaged based on those nodes. With the RBE2 element all surrounding nodes are rigidly tied together in the specified degrees of freedom. This may or may not be what you want.

-George

Hi all,

Just a point on the use of RB3 elements. Many times I have seen incorrect results when the depedant nodes have components of motion 1-6. We find it is good practice to make the dependant just 1-3 i.e. no rotational. But you still might want to set the reference node to 1-6 if you are applying a moment. If you set the reference node to 1-3 and apply a moment it will not be used.

As freebody said the RBE3 is not rigid. It transmits force without adding stiffness.

Regrads,