Semi-Rigid Roof Diaphragm

[CBEngi]

I have modeled a sloped roof with plates. I have noticed that when looking at the stresses, the plates appear to be removing compression force from the top chord of the trusses below. I have selected the "plane Stress" option so the plates shouldn't do this (or that's how I understand it).

I want the plates to act as a diaphragm only and not interfere with compression force in the top chord of the member below. I was wondering if anyone else has encountered this problem and how they fixed it.

Thanks in advance

 

[JoshPlumSE]

Plane stress plates don't remove flexural moments from the beam members they are attached to because they only have in-plan / diaphragm type stiffness. But, when they connect to the chord of a truss, their in-plane / axial stiffness will take axial force away from that truss.

Somme ideas about working around the issue.
1) Don't, accept this within the context of your lateral model. But, do the final truss design for gravity loads on it's own without the diaphragm interfering (i.e in a separate mode).

2) Keep it in the same model, but make sure the trusses don't connect to the plates, or carefully control how they connect (with plate corner releases, or nodal slaving or such). Potentially tricky, but it should work in theory.

3) Don't model the truss, at least not in your lateral model. Replace it instead with an "equivalent beam". The beam would be a single member with a moment of inertia similar to the overall truss.