I am trying to solve a modeling problem. I have a door that is connected to a frame with hinges on one side, and on the other side of the door, it gets latched to the frame. I have both parts individually meshed with NASTRAN tet10 elements. I am trying to simulate contact between the two by the usage of RBE2s. For some reason, I cannot get it working properly.
Could anyone provide some reference documents that explains the concept of utilizing RBE2s to connect two bodies together? Basically, I have a pressure force acting on the door, which I would like to be transferred into the frame. My frame is grounded using RBE2's, CBUSHES, and SPCs.
Thanks in advance.
EDIT: I should add - I currently have a node floating in space that is RBE2 to the door. This node floating in space is then connected to another node using a CBUSH element with stiffness defined in only one direction (since the constraint is only in on translational DOF). I am trying to then tie this node onto the frame elements using RBE2. But right now, it doesn't seem to be attaching properly.
Ok I'll add some pictures soon.
Maybe I'll ask the question in a different manner.
If I have two different Tet meshed parts. How do I interlink them together? Basically, what I have is a door supported by a frame. The Frame is seeing pressurized loads. I am trying to connect the door to the frame since the frame supports the door. The frame itself is grounded with RBE2's / CBUSHES/ SPC. I am trying to get the load from the door into the frame into the grounded constraints. So what is the best way of interlinking them together to get the loads/deflections transferred appropriately?
As you say use RBE2 elements and also springs with zero or near zero rotational stiffness to model hinge. As I understand the door is simple supported.
Could anyone provide some reference documents that explains the concept of utilizing RBE2s to connect two bodies together?.
Basically , the dependent nodes "follow" the independent node. RBE2 "builts" a rigid body between independent node and dependent node. In the solver the dependent nodes are get rid of the system. In fact, you could join the independent with the dependant nodes through beams with very hihg stiffnees but this colud lead to an ill-condtioning matrix causing numerical problems
I guess this simple explanation helps.
The following link explains RBE2 RBE32 concepts. It is in spanish but google translator may help you