Pushover analysis on 2d frames

Forum for OpenSees users to post questions, comments, etc. on the use of the OpenSees interpreter, OpenSees.exe

Moderators: silvia, selimgunay, Moderators

Post Reply
foadEng
Posts: 28
Joined: Sun Jan 24, 2010 12:51 pm
Location: Iran

Pushover analysis on 2d frames

Post by foadEng »

Hi
I want to do a pushover analysis on a 2D steel frame. I have done some pushover analyses on 3D models before. In 3D models, we could use reigidDiaphragm command to make every floor rigid but in 2D models, this command is not usable any more. So two problems is released here :
1.While performing eigen value analysis, the modal shapes of the nodes in a same floor would be different;
2.There is no master node, like we had in 3D models, so lateral loads should be applied at which node? every node would have different results.
I really appreciate if anyone could help me to solve this problem.
Thank you
Foad Mohajeri Nav
Iran University of Science and Technology
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Post by vesna »

Here is an example of eigen analysis of a shear frame:

http://opensees.berkeley.edu/wiki/index ... hear_frame

This should resolve the modeling issues you have.
foadEng
Posts: 28
Joined: Sun Jan 24, 2010 12:51 pm
Location: Iran

Post by foadEng »

Thanks! Eigen analysis works properly but the problem arises after that. I'm using "sp" load command to impose predefined displacement at some nodes. In 3D models, we choose master node as a representative in each story, but in 2D model "rigidDiaphragm" command is not usable any more, so there is no master node in each story. I have chosen some nodes to impose displacement loads but after that, the displacements of nodes in the same story are different. But I want to have a rigid story which I don't know how to do that!
Foad Mohajeri Nav
Iran University of Science and Technology
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Post by vesna »

The best way to model a 2D frame is to model the beams appropriately taking into account the effect of the slab.

If you want to idealize your model to have a rigid floors you can use MP constraints (look at the equalDOF command). Using this command you can equate nodal displacements at one storey. However, a node that is constrained to some other is not going to be in eqilibrium for those dofs that are constrained. If you decide to use MP constraints avoid using force based beam-column element to model the beams since the element gets messed up if axial deformation is 0 (which happens if you constrain horizontal dofs).
Post Reply