Question on using rigid diaphragm for a RC frame with fiber sections

[baozaizhang]

Dear All:

I have a question about how to use rigid diaphragm (eqDOF) in a 2D RC frame model.
I have a 5-story RC frame, I modeled beams and columns using forcebasedbeamcolumn elements. In order to apply the rigid diaphragm assumption, I used eqDOF to force all beam-column joints at each story move in the same magnitude in horizontal direction. Then I did a pushover analysis. I expected to see zero beam axial forces since all beam-column joints are coupled together in the horizontal direction. However, I saw large axial forces in the beam elements and I see zero axial deformation in the beam elements.

For a RC beam section, the neutral axis moves up (higher than the beam gross section centroid) as concrete cracks. However, if a eqDOF is applied to the two ends of a beam, the beam should have zero deformation and both ends move in the same magnitude. In this way, the section neutral axis has to be stayed at the cross section centroid. I think this is the reason why I saw axial forces in the beam elements but not sure if I am right or not. In addition, if it is right, then how opensees calculate the beam section axial forces and how are these beam axial forces treated in force equilibrium? Because the summed column shear forces at each story equals to the applied lateral loads, and the existence of the beam axial forces broke the force equilibrium at each story.

I deleted the eqDOF commands and ran the pushover analysis again. I still see large axial forces in beam elements.

Here are my two questions:
1. If I want to use forcebasedbeamcolumn elements to model RC frame beams and want to have super small beam axial forces in pushover analysis, may I ask if you have some suggestions?

2. May I ask when I use eqDOF to a RC frame using forcebasedbeamcolumn elements, where is the axial forces in the beams from? And how opensees achieve the force equilibrium?

Thank you very much for your help and best wishes.

Sincerely,
Zhi

[tthdl]

Dear Zhi
I meet the same problem as you do. The use of equalDOF would lead to inrealistic axial force and moment in beam when using the nonlinear element (here is the ref: https://opensees.berkeley.edu/wiki/inde ... nsequences). I also wonder have you manage to solve question 1?

[pattibeck]

I encounter the same issue that you do.

geometry dash

[selimgunay]

There can be two solutions to overcome this issue:

1) Define the beam cross sections using force-resultants (i.e. moment-curvature) instead of fiber sections
2) Use zerolength element at the ends of the beams with very small axial stiffness along the beam longitudinal direction