Error as ProfileSPDLinDirectSolver in eigenvalue analysis

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

Moderators: silvia, selimgunay, Moderators

Post Reply
baikunthasilwal
Posts: 24
Joined: Wed Feb 20, 2013 4:23 pm
Location: University of Virginia

Error as ProfileSPDLinDirectSolver in eigenvalue analysis

Post by baikunthasilwal »

Dear All,
Although I got time period of 0.58 sec for a simple one story model with elastic model, I got following message while solving the same one story model with panel zone.

OpenSees > source onestory.pz.tcl.tcl
ProfileSPDLinDirectSolver::solve() - aii < 0 (i, aii): (6, 0)
ProfileSPDLinDirectSolver::solve() - aii < 0 (i, aii): (24, 0)
ProfileSPDLinDirectSolver::solve() - aii < 0 (i, aii): (33, 0)
T1 = 14.300286564563539 s

Please provide me the feedback or comments to improve the model so that I can get the reasonable time period close to model without panel zone.


The model with panel zone is as below.
# -----------------------------------------------------------------
# Example 1. portal frame in 2D
# all units are in kip, inch, second
# elasticBeamColumn ELEMENT with PANELZONE
# Silvia Mazzoni & Frank McKenna, 2006
#
# ^Y
# |
# 3_________(3)________4 __
# | | |
# | | |
# | | |
# (1) (2) LCol
# | | |
# | | |
# | | |
# =1= =2= _|_ -------->X
# |----------LBeam------------|
#

# SET UP ----------------------------
wipe;
model basic -ndm 2 -ndf 3;

source DisplayModel2D.tcl;
source DisplayPlane.tcl;
source elemPanelZone2D.tcl;

# define GEOMETRY ----------------
# nodal coordinates:
node 11 0.0 0.0;
node 21 504.0 0.0

node 1201 [expr 0.0-20.0/2.0] [expr 432.0+20.0/2.0];
node 1202 [expr 0.0-20.0/2.0] [expr 432.0+20.0/2.0];
node 1203 [expr 0.0+20.0/2.0] [expr 432.0+20.0/2.0];
node 1204 [expr 0.0+20.0/2.0] [expr 432.0+20.0/2.0];
node 1205 [expr 0.0+20.0/2.0] 432.0 -mass 5.18 1e-9 1e-9;
node 1206 [expr 0.0+20.0/2.0] [expr 432.0-20.0/2.0];
node 1207 [expr 0.0+20.0/2.0] [expr 432.0-20.0/2.0];
node 1208 [expr 0.0-20.0/2.0] [expr 432.0-20.0/2.0];
node 1209 [expr 0.0-20.0/2.0] [expr 432.0-20.0/2.0];
node 1210 [expr 0.0-20.0/2.0] 432.0;
node 126 0.0 [expr 432.0-20.0/2.0];
node 127 0.0 [expr 432.0+20.0/2.0];

node 2201 [expr 504.0-20.0/2.0] [expr 432.0+20.0/2.0];
node 2202 [expr 504.0-20.0/2.0] [expr 432.0+20.0/2.0];
node 2203 [expr 504.0+20.0/2.0] [expr 432.0+20.0/2.0];
node 2204 [expr 504.0+20.0/2.0] [expr 432.0+20.0/2.0];
node 2205 [expr 504.0+20.0/2.0] 432.0 -mass 5.18 1e-9 1e-9;
node 2206 [expr 504.0+20.0/2.0] [expr 432.0-20.0/2.0];
node 2207 [expr 504.0+20.0/2.0] [expr 432.0-20.0/2.0];
node 2208 [expr 504.0-20.0/2.0] [expr 432.0-20.0/2.0];
node 2209 [expr 504.0-20.0/2.0] [expr 432.0-20.0/2.0];
node 2210 [expr 504.0-20.0/2.0] 432.0;
node 226 504.0 [expr 432.0-20.0/2.0];
node 227 504.0 [expr 432.0+20.0/2.0];

fix 11 1 1 1;
fix 21 1 1 1;

mass 1205 5.18 0. 0.
mass 2205 5.18 0. 0.

geomTransf Linear 1;

# connectivity: (make A very large, 10e6 times its actual value)
element elasticBeamColumn 1 11 126 3600000000 4227 1080000 1
element elasticBeamColumn 2 21 226 3600000000 4227 1080000 1
element elasticBeamColumn 3 1205 2210 5760000000 4227 4423680 1

# panel zone
elemPanelZone2D 51201 1201 29000.0 1000.0 1.0e5 1;
elemPanelZone2D 52201 2201 29000.0 1000.0 1.0e5 1;

DisplayModel2D NodeNumbers

# define GRAVITY
pattern Plain 1 Linear {
eleLoad -ele 3 -type -beamUniform -7.94 ;
}

# Eigenvalue Analysis
set pi [expr 2.0*asin(1.0)];
set nEigenI 1;
set lambdaN [eigen [expr $nEigenI]];
set lambdaI [lindex $lambdaN [expr $nEigenI-1]];

set w1 [expr pow($lambdaI,0.5)];
set T1 [expr 2.0*$pi/$w1];
puts "T1 = $T1 s";

I am looking forward to receive reason behind these error message and different time period of the frame.
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysi

Post by fmk »

you obviously have problems with the model, in particular that panel zone procedure you are using. the results you are getting are as a result probably gibberish.
baikunthasilwal
Posts: 24
Joined: Wed Feb 20, 2013 4:23 pm
Location: University of Virginia

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysi

Post by baikunthasilwal »

Dear Dr. Frank,

Thank you for your feedback that helps to improve model especially the panel zone modelling and its parameters. Now, I got the reasonable response with some changes in my model.

I have one issue for assigning panel zone shear model properties that make the response changes significantly. Please provide some reference sources or selecting reasonable shear spring parameters. Next, while using Pinching4 uniaxial material model to represent shear spring, I am not sure which load we have to use (Shear or Moment). can we use either shear or moment as a load for the shear spring employing Pinching4? Or we can use other uniaxial material models such as Modified Ibarra-Krawinkler model.
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysi

Post by fmk »

you can use any one of the uniaxial material models .. they all provide a stress-strain/force-displacement relationship .. whether the relationship is for axial, shear, moment, ... the material does not care nor does it know .. the trick is for you to select one who's response curve best suits the behaviour for which you are using it.
baikunthasilwal
Posts: 24
Joined: Wed Feb 20, 2013 4:23 pm
Location: University of Virginia

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysi

Post by baikunthasilwal »

Dear Dr. Frank,

Thank you for your valuable feedback.
smrithi
Posts: 4
Joined: Mon Aug 27, 2018 3:20 pm
Location: Iowa State University

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysi

Post by smrithi »

I am trying to model a 3D steel frame with panel zone elements. I am now facing the same problem with my model using panel zones. Could we have a discussion regarding your modeling strategies. I would appreciate any help I can get.

Thank you,
Smrithi
sriarun
Posts: 3
Joined: Fri Dec 13, 2019 7:48 pm

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysis

Post by sriarun »

I am facing the same issue with modeling panel zone elements. I am unable to figure out the error in the panel zone procedure that is responsible for exciting local panel zone modes, thereby affecting the eigenanalysis and any dynamic analysis. Kindly let me know if this issue is resolved.

Thanks,
Srinivasan
moezajafari
Posts: 1
Joined: Mon Dec 16, 2019 9:29 am

Re: Error as ProfileSPDLinDirectSolver in eigenvalue analysis

Post by moezajafari »

Post Reply