Strut-Frame connection with Rigid Diaphragm in a 3D frame

[AhmedHisham]

Hello STKO team,
I'm modelling a 3D infilled frame (6-storey, 5-bay) using an infill macro-model. The macro-model is composed of 2 diagonal struts (4 in total) with a lumped mass at the intersection to simulate the OOP behaviour. The frame has a few interactions, namely;
1. Strut-Frame, where the frame nodes are the master while the strut nodes are the slaves, (equalDOF is applied in all DOFs)
2. Strut vertical Interaction, where the OOP nodes at the top are master to those lower
3. Strut Master node, where one of the top OOP nodes is the master, and the other node is a slave.
(equalDOF is applied in Uz and Ux or Uy, depending on the orientation) for both 2 and 3.

Picture3.png (94.03 KiB) Viewed 1884 times

4. Rigid diaphragm (RD) constraint applied floor-wise to the frame nodes.

I analyse the frame under gravity loads and perform eigenanalysis afterwards. I unfortunately faced a few issues with the solutions I found.

At the beginning, I used the transformation constraint handler with "Newton with Line Search" algorithm and "NormDispIncTest". The analysis converges, but the results I found seem off. The strut nodes don't follow the equalDOF condition in both the gravity analysis and the eigenanalysis. I thought it might be due to the fact that I didn't include the strut nodes in the RD interaction, but the problem persists even after I did so.

Picture1.png (1.25 MiB) Viewed 1884 times

Picture2.png (145.32 KiB) Viewed 1884 times

I then changed the constraint handler from transformation to penalty, which only works when I reduce the frame to a 1-storey frame. I have a similar observation when I use the Auto handler. The model behaves correctly kinematically, and mode shapes appear physically consistent. However:

My concern is that:
Transformation converges but may lose part of the kinematic constraint topology, producing physically incorrect behaviour.
Penalty/Auto preserve correct kinematics, but lead to severe convergence problems in the full-scale model.

This seems to be caused by the complex interaction between rigid diaphragm constraints and multiple equalDOF MP constraints.

Would you recommend a different modelling or constraint strategy for this type of system?
Any guidance or best-practice recommendations would be greatly appreciated.
I attached a demo version of the frame (not the actual struts I use; I substituted them with truss elements instead), and everything else should be the same

Thank you very much for your time and support.

Best regards,
Ahmed

IF_1.zip

(1.18 MiB) Downloaded 18 times

[STKO Team]

Your trusses are split in the middle, thus they have no stiffness in the OOP direction. Are you sure you didn't want to model them as beams?