local rotation of RHS-T-joint

[fl0rian]

Hello,

I am investigating the local rotations of rectangular hollow section T-joints under in-plane bending.

I have modelled a chord und a brace which are connected via a tie constraint. The reference point on top is coupled to the top surface of my brace.
Then the RP is loaded with a single moment.
Here you can see my model:

The results look something like that:

My ultimate goal is to get Phi - the local deformation of the joint. I do not want the elastic deformations of the brace Phi_br and the chord Phi_ch.
For now I extract the global rotation - UR1 - of my reference point and substract the values of the elastic rotation of the chord and the brace.
I get those with simple equations (engineering beem teory).

Is there any way to extract just the local joint rotation Phi? If so, could you give me a hint on how to do it.

Thanks!!

 

[FEA way]

Transformation of the results to local coordinate system may be useful here. But I would mesh this structure with shell elements. You might also consider different approach to modeling the connection between both parts than the one with tie constraint. It depends how they are connected in real life.

 

[fl0rian]

thank you for answering! I am going to try the local coordinate system.

In real life the chord and the brace are welded together. Either with a bevel weld or a fillet weld. In the example above it is a full bevel weld. Thats why I went with the tie constraint. Which different approach would you suggest?

At first I also tried modelling it with shell elements, but the results differ to much from real life test data.
All those joints are built and then tested in real life. My job is to build a abaqus reference model to interpolate joint variations, which are not tested. Shell elements result in a approximately 15 percent lower initial stiffness.