2D Concrete frame with shear wall

salman · Post by **salman** » Thu Nov 26, 2009 5:06 pm

Dear Dr. Mazzoni:
i`m trying to model 5bay 3 story concrete frame with shear wall.but after running my model(static gravity analysis) an error occurs .

WARNING BandGenLinLapackSolver::solve<> -LAPAK routine returned 377
WARNING NewtonRaphson::solveCurrentStep<> -the Linear SysOfEgn failed in solve<> StaticAnalysis::analyze<> - the algorithm failed at iteration:0 With domain at load factor 0.1
Opensees > analyze failed,returned: -3 error flag

my model are as follows.
I couldn't solve this problem.
would you please help me?

thanks

# number of storeys = 3 & number of beys = 5
# units KN-m-s
wipe
# start of model generation
model BasicBuilder -ndm 2 -ndf 3
set displayMode "displayON"
# node command
node 1 0 0
node 6 25 0
node 7 0 3
node 12 25 3
node 13 0 6
node 18 25 6
node 19 0 9
node 24 25 9
# Fix Supports At Base Of Columns & Shear Walls
# tag Dx Dy Rz
#colums
fix 1 1 1 1
fix 6 1 1 1
#walls
fix 25 1 1 1
fix 26 1 1 1
fix 27 1 1 1
fix 28 1 1 1
fix 29 1 1 1
fix 125 1 1 1
fix 126 1 1 1
fix 127 1 1 1
fix 128 1 1 1
fix 129 1 1 1

#Define Nodal Masses
set mrfext 5.188
set mstrext 5.813
set mrfint 7.032
set mstrint 8.251
# external columns
# storeys
mass 7 $mstrext 0 0
mass 12 $mstrext 0 0

mass 13 $mstrext 0 0

mass 18 $mstrext 0 0
#roof
mass 19 $mrfext 0 0
mass 24 $mrfext 0 0
uniaxialMaterial Concrete02 1 -43422.5 -0.0052 -37845.5 -0.0133 0.1 4342.25 2780000
set fy 450000.0 ; # Yield stress
set E 200000000.0 ; # Young's modulus
uniaxialMaterial Steel01 2 $fy $E 0.02

# defining sections
source RCsection.tcl
source RCsectionBeam2.tcl
source wallsection1.tcl

proc wallsection1 {id midcore midcover n1 n2 steelid n a} {
section Fiber $id {
patch quad $midcore $n1 $n2 -2.35 -0.025 2.35 -0.025 2.35 0.025 -2.35 0.025
patch quad $midcover $n2 $n2 -2.5 -0.05 -2.35 -0.025 -2.35 0.025 -2.5 0.05
patch quad $midcover $n1 $n2 -2.5 -0.05 2.5 -0.05 2.35 -0.025 -2.35 -0.025

patch quad $midcover $n2 $n2 2.35 -0.025 2.5 -0.05 2.5 0.05 2.35 0.025
patch quad $midcover $n1 $n2 -2.35 0.025 2.35 0.025 2.5 0.05 -2.5 0.05

layer straight $steelid $n $a -2.35 0.025 2.35 0.025
layer straight $steelid $n $a -2.35 -0.025 2.35 -0.025

}
}
# column section
RCsection 1 .3 .3 .04 1 1 2 3 0.00020096 5 5 2 1
# beam section
RCsectionBeam2 2 0.30 0.25 0.04 1 1 2 3 2 2 0.00025434 5 5 2 1
#wall section
wallsection1 3 1 1 10 1 2 15 0.00011304
# -------------------------define columns
geomTransf PDelta 1
set np 3
element nonlinearBeamColumn 1 1 7 $np 1 1 -iter 1000 1.0e-1
element nonlinearBeamColumn 2 6 12 $np 1 1 -iter 1000 1.0e-1
element nonlinearBeamColumn 3 7 13 $np 1 1 -iter 1000 1.0e-1
element nonlinearBeamColumn 4 12 18 $np 1 1 -iter 1000 1.0e-1
element nonlinearBeamColumn 5 13 19 $np 1 1 -iter 1000 1.0e-1
element nonlinearBeamColumn 6 18 24 $np 1 1 -iter 1000 1.0e-1
# ----------------------define walls
set nx 4
set ny 4

set Plate ShellMITC4
set eleArgs "3"
set n 0
set l 0
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 5 0
2 10 0
3 10 3
4 5 3
}

set n 1
set l 1
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 5 3
2 10 3
3 10 6
4 5 6
}

set n 2
set l 2
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 5 6
2 10 6
3 10 9
4 5 9
}

set n 4
set l 4
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 15 0
2 20 0
3 20 3
4 15 3
}

set n 5
set l 5
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 15 3
2 20 3
3 20 6
4 15 6
}

set n 6
set l 6
set n1 [expr (25+$n*25)]
set l1 [expr (28+$l*16)]

block2D $nx $ny $n1 $l1 $Plate $eleArgs {
1 15 6
2 20 6
3 20 9
4 15 9
}

for {set j 0} {$j < 2} {incr j 1} {
for {set i 0} {$i < 5} {incr i 1} {
equalDOF [expr 50+(25*$j)+$i] [expr 45+(25*$j)+$i] 1 2 3
}
}
for {set j 0} {$j < 2} {incr j 1} {
for {set i 0} {$i < 5} {incr i 1} {
equalDOF [expr 150+(25*$j)+$i] [expr 145+(25*$j)+$i] 1 2 3
}
}
# nodal masses on walls
# storeys
mass 45 $mstrint 0 0
mass 49 $mstrint 0 0

mass 145 $mstrint 0 0
mass 149 $mstrint 0 0

mass 70 $mstrint 0 0
mass 74 $mstrint 0 0

mass 170 $mstrint 0 0
mass 174 $mstrint 0 0

# roof
mass 95 $mrfint 0 0
mass 99 $mrfint 0 0
mass 195 $mrfint 0 0

mass 199 $mrfint 0 0
##beams
geomTransf Linear 2
element nonlinearBeamColumn 19 7 45 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 20 49 145 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 21 149 12 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 22 13 70 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 23 74 170 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 24 174 18 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 25 19 95 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 26 99 195 $np 2 2 -iter 1000 1.0e-1

element nonlinearBeamColumn 27 199 24 $np 2 2 -iter 1000 1.0e-1
#------------------

pattern Plain 1 Linear {
eleLoad -ele 19 -type -beamUniform -21.82
eleLoad -ele 20 -type -beamUniform -21.82
eleLoad -ele 21 -type -beamUniform -21.82
eleLoad -ele 22 -type -beamUniform -21.82
eleLoad -ele 23 -type -beamUniform -21.82
eleLoad -ele 24 -type -beamUniform -21.82
eleLoad -ele 25 -type -beamUniform -19.37
eleLoad -ele 26 -type -beamUniform -19.37
eleLoad -ele 27 -type -beamUniform -19.37
}
constraints Plain
numberer RCM
system BandGeneral
test NormUnbalance 1.0e-5 200 3
algorithm Newton
integrator LoadControl 0.1

analysis Static
analyze 10[img][/img][img][/img][img][/img]

silvia · Post by **silvia** » Thu Nov 26, 2009 7:13 pm

it is a convergence problem.
please look at the examples manual for parts of the script that deal with non-convergence

salman · Post by **salman** » Fri Nov 27, 2009 1:43 am

as i understood,you confirm that ]
my frame and shear wall modeling is correct,but only the problem is related to convergence part.

silvia · Post by **silvia** » Fri Nov 27, 2009 8:24 am

no, this does not conirm it, of course,
you have to check it.

salman · Post by **salman** » Fri Nov 27, 2009 9:52 am

Dear Dr. Mazzoni:
i appreciate you very much
i think that my problem is in shear wall modeling.so after modeling a simple shear wall with elasticmembraneplatesection without any meshes and 4 nodes around the wall , i fixed nodes 1&2 at base and put nodal loads equal to 3KN on each 3&4 nodes.
but after running, the reactions of nodes 1&2 are very greater than (3+3)KN.
also when i changed my nemberer from plain to RCM or changing Algorithm from linear to Newton,,,, the analysis failed.
please advice me in this regard

best regards
salman

salman · Post by **salman** » Fri Nov 27, 2009 11:46 am

Dear Dr. Mazzoni:
i found another thing.
i model the above mentioned r.c.wall in 3D space but fixed in for 2D analysis.the reactions are true and system,algorithm,numberer.......are without problem.
but still i don`t know why?

i model the wall in3D space with elastic section & shellMITC4 without problem ,but when i use fiber section the nalysis failed

silvia · Post by **silvia** » Fri Nov 27, 2009 7:09 pm

it is not converging, try the scripts to see if you can make it past that point.

salman · Post by **salman** » Sat Nov 28, 2009 1:49 am

thanks for your helpful advices.
i can`t understand your purpose. would you please explain your advice a little more
thank you very much

silvia · Post by **silvia** » Sat Nov 28, 2009 8:45 am

please look at the examples manual