failed to converge

For posts concerning the documentation, errors, ommissions, general comments, etc.

Moderators: silvia, selimgunay, Moderators

Post Reply
jndlsmt
Posts: 18
Joined: Thu Nov 25, 2010 2:37 am
Location: IIT Kanpur

failed to converge

Post by jndlsmt »

hi friends,
I am getting this warning, please help me to sort out it.
WARNING:-CTestNormDispIncr:: test <failed to converge>
after 40 iteration
NewtnRaphson:: solveCurrentStep<> the convergence test object failed in test<>
Static analysis :: analysis <> the Algorithm failed at iteration: 3 vwith domainat load factor 0.4
opensees > analysis failed
please reply....
Sumit Jindal
IIT Kanpur (India)
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Post by fmk »

i suggest you search the user forum where there are many posts that are similar to this
jndlsmt
Posts: 18
Joined: Thu Nov 25, 2010 2:37 am
Location: IIT Kanpur

Post by jndlsmt »

Dear fmk,
thanks for suggestion but the problem is something different .
in my analysis i am applying cyclic load at a particular node.
out of this two analysis (gravity or cyclic) one is going to fail.
here i am writing my code..... please look at this..
1st input file...Input1.txt

## INPUT FILE FOR SHALLOWFOUNDATIONGEN
#Input data for foundation 1
#SoilProp $SoilType $c $Phi $Gamma $G $Nu $Crad $Tp
SoilProp 2 0.0 40.0 16200.0 16000000.0 0.4 0.05 0.05
#CapSoil $Qult $Pult $Tult $Kv $Kx
#CapSoil 1800.0 1000.0 1000.0 2000.00 3000.0
#FootProp $Lf $Bf $Hf $Df $Ef $Wg $beta # Ef (aluminium)
FootProp 2.80 0.65 0.66 0.0001 68900000000.0 283312.8 0.0
#MeshProp $Rk $Re $le/L
MeshProp 5.0 0.09 0.02
#End of input dat
############################
Main code file....
###########################

##"A shear wall supported by a strip footing"
#--Written by Sumit Jindal
#--units: N,m
wipe
wipeAnalysis
file mkdir OUTPUT; ##### Create output directory
# Create ModelBuilder
model BasicBuilder -ndm 2 -ndf 3
# Set wall and footing dimensions
set footL 2.8

set LengthWall 2.5;
set WidthWall 0.38;
set HeightWall 10.1;

set AspRatio 0.41; ; #---->CH; #--aspect ratio (M/H/L)
set pod [expr $AspRatio*$footL] ; #----point of lateral load application
# Set structural nodes
node 1 0. 0.
node 2 0. $pod
node 3 0. $HeightWall
# set wall properties
set AWall [expr $WidthWall*$LengthWall]
set EWall [expr 68.9*pow(10,9)]; #----[N/m^2] aluminium
set IWall [expr $WidthWall*pow($LengthWall,3)/12.]
# set wall material
uniaxialMaterial Elastic 1 $EWall
# set geometric transformation
geomTransf Linear 1
# set wall element
#element elasticBeamColumn $eleTag $iNode $jNode $A $E $Iz $transfTag
element elasticBeamColumn 1 1 2 $AWall $EWall $IWall 1
element elasticBeamColumn 2 2 3 $AWall $EWall $IWall 1

# set wall mass
set g 9.81; #--m/s^2
set FSv 5.2
set MWall 28880.0; #---mass of structure in prototype scale (kg), provided by Gajan
mass 2 [expr $MWall/2.] [expr $MWall/2.] 1
mass 3 [expr $MWall/2.] [expr $MWall/2.] 1
set Wg [expr $MWall*$g]
puts "Wg=$Wg"
#-------------------------------------------------
# Use ShallowFoundationGen command to
# attach shallow foundation with Foundation tag=1
# at node 1
#--------------------------------------------------
##
set FoundationTag 1
#ShallowFoundationGen $FoundationTag $ConectNode $InputFileName $FootCondition
ShallowFoundationGen $FoundationTag 1 "Input1.txt" 5
source Foundation_$FoundationTag.tcl
set MassFooting 28880.0
mass 1 $MassFooting $MassFooting 1
#-------------------------------
# Eigen Value Analysis
#-------------------------------
set PI 3.1415926
set lambdax [eigen 1]
set lambda [lindex $lambdax 0]
set omega [expr pow($lambda,0.5)]
set Tn [expr 2*$PI/$omega]
set fn [expr 1/$Tn]
puts "1st mode, Tn=$Tn sec, fn=$fn Hz"
#-------------------------------
# Recorder
#-------------------------------
###-wall
recorder Node -time -file OUTPUT/WallRoofdisp.dat -node 1 2 3 -dof 1 2 3 disp
recorder Element -file OUTPUT/WallElementforce.dat -time -ele 1 2 localForce
###--------recorder for footings------------------------######
###-end Spring left-------
recorder Node -time -file OUTPUT/EndSprLdisp.dat -node $endFootNodeL_1 -dof 1 2 3 disp
recorder Element -file OUTPUT/EndSpringLforce.dat -time -ele $endSprEleL_1 force
###-end Spring right-------
recorder Node -time -file OUTPUT/EndSprRdisp.dat -node $endFootNodeR_1 -dof 1 2 3 disp
recorder Element -file OUTPUT/EndSpringRforce.dat -time -ele $endSprEleR_1 force
#-----------------------
# Gravity LOAD PATTERNS
#-----------------------
set deadLoad [expr ($MassFooting+$MWall)*$g*$FSv]; #---total gravity load on footing (footing+wall)
#puts $deadLoad
pattern Plain 1 "Linear" {
load 1 0. [expr -$deadLoad] 0.
}
#--------------------
# gravity analysis
#--------------------
system UmfPack
constraints Plain
test NormDispIncr 1.0e-8 25 0
algorithm Newton
numberer RCM
integrator LoadControl 0.1
analysis Static
analyze 10; #use 10 analysis steps
#--------------------

#loadConst
loadConst -time 0.0
####################################
# Define cyclic loads and loadNode
####################################
set inFile input-disp-SSG02_02.txt######### [url]link for this
file.www.4shared.com/document/95_-I_kY/input ... 02_02.html
set np 16555
set dt 100.
set dir 1
set factor 0.001
set gdisp1 "Path -filePath $inFile -dt $dt -factor $factor"
puts "input file= $inFile"
#----------------------------------------------------
pattern MultipleSupport 4 {
groundMotion 4 Series -disp $gdisp1
##imposedMotion nodeTag? dirn? gMotionTag?
imposedSupportMotion 2 $dir 4
}
#Create analysis
#---------------------------------------------

system UmfPack
constraints Transformation
test NormDispIncr 1.0e-3 40 0
algorithm KrylovNewton
numberer RCM
#Create the integration scheme Newmark with gamma = 0.5 and beta = 0.25
integrator Newmark 0.5 0.25
analysis Transient

#------------------------------------------------
#Perform analysis
#------------------------------------------------
puts "np=$np, dt=$dt"
analyze $np $dt
Sumit Jindal
IIT Kanpur (India)
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Post by fmk »

you need to put the analyze in a while loop and try different things if one step fails .. look at example3.3 here:

[url]
http://opensees.berkeley.edu/WebSVN/lis ... cd9592fe53
[/url]
jndlsmt
Posts: 18
Joined: Thu Nov 25, 2010 2:37 am
Location: IIT Kanpur

Post by jndlsmt »

thanks fmk for reply
but i tried that also , i am getting same problem,
please run my program and give me a solution for that.
i fed up with this, i have to submit this analysis to my thesis supervisor.
please help me to sot out it.
Sumit Jindal
IIT Kanpur (India)
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Post by fmk »

your soil cannot take even the gravity load! .. i suggest you look at the soil properties you have defined or the loading you are putting on it.
jndlsmt
Posts: 18
Joined: Thu Nov 25, 2010 2:37 am
Location: IIT Kanpur

Post by jndlsmt »

thanks fmk
there was prob of load whatever i was applying that was to much,facotor of safety was too high, now analysis is safe in gravity analysis but cyclic load analysis is going to fail again...
plz check it again
use "" set FSv 1.5""
Sumit Jindal
IIT Kanpur (India)
andy0071
Posts: 18
Joined: Sat Feb 26, 2011 11:22 pm
Location: SYjianzhu university

Re: failed to converge

Post by andy0071 »

to Sumit Jindal
I copy your model and put the analysis "while loop" module and run it. It failes in all kinds of analysis i used. Although i didn't make it run but there are some suggestions to you.
1.all the units used in OpenSees is British units just like "inch" and "kip" , but i saw your model units writed by N and M . Is that the reason for your soil model failure to use input file in metric units?
2.If you want your model run for a helpful but not accuracy try to find where is the errors come from, maybe you can amplify your Tolerance and the No. of the test. I change it to Tol 1 No.2000 but still get the error. So i gess the wrong is from the model build but not the analyze. Perhaps it from the units and the parameters you used.

wish your model run. any new evolution please let me know :D
jroyce1
Posts: 1
Joined: Wed Nov 09, 2011 6:15 pm
Contact:

Re: failed to converge

Post by jroyce1 »

Thank you for the info!
Hello, I'm Jake. I am looking forward to interacting with this online community. direct payday lenders no third party
emanueletrainini
Posts: 5
Joined: Wed Nov 30, 2011 8:12 am

Re: failed to converge

Post by emanueletrainini »

Hei Sumit Jindal. Are you using the "white loop" module? Can you give us some more infos abou your soil model? I will try to compare it with my experience and give you some feedback. Bye from Italy
Emanuele Trainini, architetto :: Brescia
Progettazione | Certificazione Energetica
Denunce di Successione | Pratiche catastali
Post Reply