Rigid

[AndyLake]

Results (displacements) are not reliable when I use rigid elements and I have fixed only T1, T2 and T3. Why?! When I use equations (T1, T2 and T3) results are correct.

 

[AndyLake]

I use MSC/NASTRAN for windows 4.0.3

 

[Caronte]

Wich kind of rigid Nastran elements are you using? You must be using any type of MPC, but RBE2, RBE3 or wich kind?

 

[SWComposites]

1) How do you know the results are not reliable?

2) Rigid elements can have rigid body motions (translations and rotations), so they can act very different from MPC equations. Both give correct results; the user just has to understand how each one behaves and select the proper modeling technique.

Rigid elements and MPC are often abused and misunderstood. Build simple models first to understand their behavior (this should be done whenever you are using an FE feature that you have not used before.

FEA = A (SOMETIMES CLOSE) APPROXIMATION TO THE VERSION OF REALITY INPUT BY THE USER

 

[GregLocock]

"FEA = A (SOMETIMES CLOSE) APPROXIMATION TO THE VERSION OF REALITY INPUT BY THE USER"

I prefer

ALL MODELS ARE WRONG. SOME MODELS ARE USEFUL.

Cheers

Greg Locock

 

[AndyLake]

I built simple model. I used rigid element and fixed all DOF (translations & rotations) between two nodes. Everything was all right: translations & rotations were same between above-mentioned nodes.

Then I modifyed model. I fixed only translations. Now translations between two nodes weren't same!