warning in eigen value

[soroushali]

Hi
I build a steel moment frame unsing Ibarra model with constrained plasticity. Then I did eigen value analysis on it.
It give me following warning:
ProfileSPDLinDirectSolver::solve<> - aii <0<i,aii>: <<25, 0>
I check the constrains and EqualDof.
Could you please help me.

[fmk]

do a "print node" after it fails and find a node with an ID of 25 .. either node is free, not enough conatrstints or ele framing into it are producing a negative value for kii .. in which case either incorrect definition, or material is yielding and you have chosen the wrong type of solver as matrix no longer symmetric POSITIVE definite

[soroushali]

Thanks
I download Dynamic_example and MRF_PanelZone_example from openSEES site. They have the same warning when performing the eigen command.
Is this a common warning?
Can I Ignore it for further IDA analysis?

[fmk]

it is typically not something that can be ignored as it typically means that the analysis has stopped .. if only occuring in the eigen command, do you get eigenvalues?

[soroushali]

Yes, I got eigenvalues.
I changed the stiffness of zero length element in RotLeaningCol.tcl file that written by Laura Eads, to a very large value. This time, The warning didn't appear.

[AngusHunter]

Node: 2101
Coordinates : 0.381 4.572
Disps: 0 0 0
Velocities : 0 0 0
commitAccels: 0 0 0
Mass :
2.5 0 0
0 1e-016 0
0 0 1e-016

Rayleigh Factor: alphaM: 0
Rayleigh Forces: 0 0 0
Eigenvectors:
-2.16529e-017 6.77502e-008 -1.23543e-009
-0.447214 -4.26859e-008 7.6036e-010
0.053354 -5.90529e-008 1.05151e-009

Hi I get some Negative eigenvectors when I use command “print node”. Is it right? I am not sure, PLEASE help me !

[selimgunay]

Negative values of eigenvectors at some nodes are fine. It all depends on the mode shape

[robertomartinez]

can eigenvectors be negative even for the first mode shape ?