simple concrete frame (please help)

[Ahm]

Hi all,
I’m working on a simple concrete frame with 1 floor and 1 bay.
I used Fiber "T section" to model the beam, and I get the error message

{WARNING: CTestNormDispIncr::test() - failed to converge
after: 10 iterations
NewtnRaphson::solveCurrentStep() -the ConvergenceTest object failed in test()
StaticAnalysis::analyze() - the Algorithm failed at iteration: 6 with domain at
load factor 0.7
OpenSees > analyze failed, returned: -3 error flag
-3}
when I use rectangle section the analysis completed.

What is the problem with that " T section"?
thanks alot


# Units: KN,m
# ------------

model basic -ndm 2 -ndf 3

set width 4
set height 3.65

node 1 0.0 0.0
node 2 $width 0.0
node 3 0.0 $height
node 4 $width $height

fix 1 1 1 1
fix 2 1 1 1

uniaxialMaterial Concrete01 1 -35e3 -0.003 -25e3 -0.014

uniaxialMaterial Concrete01 2 -20e3 -0.002 0.0 -0.006

uniaxialMaterial Steel01 3 3.6e5 2.0e8 0.01

set colWidth 0.3
set colDepth 0.7

set cover .025

set As .000201

set y1 [expr $colDepth/2.0]

set z1 [expr $colWidth/2.0]

section Fiber 1 {

patch rect 1 10 1 [expr $cover-$y1] [expr $cover-$z1] [expr $y1-$cover] [expr $z1-$cover]

patch rect 2 10 1 [expr -$y1] [expr $z1-$cover] $y1 $z1
patch rect 2 10 1 [expr -$y1] [expr -$z1] $y1 [expr $cover-$z1]
patch rect 2 2 1 [expr -$y1] [expr $cover-$z1] [expr $cover-$y1] [expr $z1-$cover]
patch rect 2 2 1 [expr $y1-$cover] [expr $cover-$z1] $y1 [expr $z1-$cover]


layer straight 3 6 $As [expr $y1-$cover] [expr $z1-$cover] [expr $cover-$y1] [expr $z1-$cover]
layer straight 3 6 $As [expr $y1-$cover] [expr $cover-$z1] [expr $cover-$y1] [expr $cover-$z1]
layer straight 3 2 $As [expr $y1-$cover] .06666 [expr $cover-$y1] .06666
layer straight 3 2 $As [expr $y1-$cover] -0.06666 [expr $cover-$y1] -0.06666
}

geomTransf Linear 1

set np 5

element nonlinearBeamColumn 1 1 3 $np 1 1
element nonlinearBeamColumn 2 2 4 $np 1 1





section Fiber 2 {

patch rect 1 5 1 -0.4394 0.1 0.2106 -0.1
patch rect 1 1 5 0.1406 0.1 0.2106 0.575
patch rect 1 1 5 0.1406 -0.575 0.2106 -0.1
patch rect 2 1 2 0.2106 -0.575 0.2356 0.575
patch rect 2 2 1 0.1156 0.575 0.2356 0.6
patch rect 2 2 1 0.1156 -0.6 0.2356 -0.575
patch rect 2 1 2 0.1156 -0.575 0.1406 -0.1
patch rect 2 1 2 0.1156 0.1 0.1406 0.575
patch rect 2 2 1 -0.4394 0.1 0.1156 0.125
patch rect 2 2 1 -0.4394 -0.125 0.1156 -0.1
patch rect 2 1 2 -0.4644 -0.125 -0.4394 0.125


layer straight 3 3 0.000201 0.2106 0.1 0.2106 -0.1
layer straight 3 2 0.000113 -0.1144 0.1 -0.1144 -0.1
layer straight 3 4 .000201 -0.4394 0.1 -0.4394 -0.1
layer straight 3 3 0.000113 0.1406 0.575 0.1406 0.1
layer straight 3 3 0.000113 0.2106 0.575 0.2106 0.1
layer straight 3 3 0.000113 0.1406 -0.1 0.1406 -0.575
layer straight 3 3 0.000113 0.2106 -0.1 0.2106 -0.575
}
geomTransf Linear 2
element nonlinearBeamColumn 3 3 4 $np 2 2


pattern Plain 1 Linear {

eleLoad -ele 3 -type -beamUniform -35
}




system BandGeneral

constraints Transformation

numberer RCM

test NormDispIncr 1.0e-4 10 3

algorithm Newton

integrator LoadControl .1

analysis Static


recorder Node -file 3-5.xls -node 1 2 -dof 1 2 reaction


analyze 10
#-----------------------------------------------------------
##----------------------------------------------------------

[jwaugh]

It could just be that it is taking longer to converge because it is a more complicated section.

Mkae this line:
test NormDispIncr 1.0e-4 10 1

and then see if it is actually converging. If it is just increase the number of iterations it can do at each step.

[silvia]

or try to reduce the step increment, too.

[ocean2000]

the ConvergenceTest failed,you can try other Algorithm,but it need so much time.

[Ahm]

dear all.....
I think that the problem is the two values of the width of the beam
so the program failed in working with 2 values of
width and need one only.........