Point of load application in MPA

[armanattarzadeh]

I have a three-Story one-bay X-braced frame. Here are the assumptions I uesd to model the frame
-Geometric imperfection of brace is parabolic in shape, out of plane with initial imperfection in the middle of brace of Leff/200.
-Beams and columns are modeled with force-based beam-column element .
-Braces are defined with 12 force-based beam-column element
-To diminish convergence error, truss elements with very small stiffness are added to braces.
-Gusset plates are modelled with (i) force-based beam-column element (ii) rotational hinge using article “Improved analytical model for special concentrically braced frames, Po-Chien Hsiao , Dawn E. Lehman, Charles W. Roeder,2012”
-Number of integration points for beam, column and brace is 4 and for rigid link is 3.
To determine periods and modeshapes of the frame in Opensees, all solvers in eigen command(-genBandArpack, -fullGenLapack, -UmfPack, -SuperLU) are used. If i changed the node in which modeshape is read, modeshapes,especially in higher modes seem not to be logical. Periods and modeshapes are also derived from ETABS and observed the periods in opensees and ETABS are almost equal but the modeshapes are not.
I attached my model in the following link:

http://www.mediafire.com/download/c3nxi ... y-SCBF.zip

Run "00-Build-SAC-Generic.tcl "

I want to do modal pushover analysis method (MPA) in multi-X braced frame. I have some questions:
1-When using “nodeEigenvector” command, Which nodes should be read?
2-Where should be the point of load application for MPA analysis(left,middle or right of beam)?


Can anyone check what's going wrong with my model?

[MRG]

i have this problem with modeshapes too that i explain it with simple model and i demontrate its results in the topic ("Difference in ModeShape between Sap & Opensees" in Mon Jul 15, 2013 ) .
please help me too .
thank you