eigen values, fiber sections, nonlinear time-history anlysis

[horizon]

Dear All,
I am working on modeling of a four-storey RC "moment resisting frames+shear
walls" building in OpenSees and going to run lots of nonlinear time-history
analyses.On the other hand, I have the model of the building in SAP2000.In
OpenSees, first time I used elastic frame sections to define the beams,
columns
and shear walls.I connected the frames which represent shear walls to the
beams
through their widths by rigid fictitious elastic elements. All of these are
same in SAP2000 model.Secondly, I defined the columns as fiber sections and
used elastic sections again for the beams and shear walls.
I created three different tcl files:
1) elasticlinear: "elasticBeamColumn" elements used
2) elasticnonlinear: "Elastic" sections and "nonlinearBeamColumn" elements
used.
3) fibernonlinear: "Fiber" sections (just for columns) and
"nonlinearBeamColumn" elements used.
I got the same results for eigenvalues from the first and second models but
they
differ from SAP2000 results.I almost catch the same value both in OpenSees
and
SAP2000 for the 1st mode and this means there is no problem with the
stiffness
representations and mass assginments between two programs, but I got
SAP2000's
2nd mode in the 3rd mode of OpenSees.Is it possible to get modal
participating
mass ratios for each mode from OpenSees? Do you think whether it is related
to
numerical solution in OpenSees or modeling approach? On the other hand, I
get
negative eigenvalues from the third model in OpenSees and it is not
reasonable!
Eigen elasticlinear elasticnonlinear SAP2000
1 2.45E+02 2.45E+02 2.35E+02
2 2.51E+02 2.51E+02 3.44E+02
3 3.51E+02 3.51E+02 4.65E+02
4 2.59E+03 2.59E+03 2.76E+03
5 2.80E+03 2.80E+03 3.21E+03
This is my first problem and the second one:
Assuming that I have no problem with modeling, I run a time-history analysis
under tabasFN and tabasFP records but it failed after 3rd step with the
error
message "failed to get compatible element for deformations for element:..".
Is it because of the fiber sections?I tried to reduce the number of fibers
but
it failed again.
On the other hand, in that model I used fiber sections just for columns but
I
have to describe shear walls with fiber sections too.I mean there will be
lots
of fiber sections in my model...
Any idea?Anyone had similar problems?
Thanks and regards
Ufuk

[seddighi]

test dispBeamColumn element instead of nonlinearBeamColumn

[vesna]

OpenSees does not output modal participating mass ratios.

Negative eigenvalues are most probably caused by modeling errors.

After attempting to run time history analysis the error message indicates that your columns are failing. Check on your fiber section. You may have an error there. Make sure to define your fibers relative to local coordinate system. Check your geometric transformation to assure that your local axes are correctly defined.