RandomVariablePositioner

A place to ask questions on how to use OpenSees to perform Finite Element Reliability Analysis

Moderators: silvia, selimgunay, mhscott, Moderators

Post Reply
lorcancon
Posts: 25
Joined: Thu May 22, 2014 5:22 am
Location: Roughan & O Donovan

RandomVariablePositioner

Post by lorcancon » Mon Jun 16, 2014 5:52 am

Hello,

I noticed this command in the paper entitled "Finite Element Reliability and Sensitivity Methods for Performance-Based Earthquake Engineering (2004)". It is used to add random variables to a finite element model. However, when I tried to use this command, the opensees command window stated that it didn't exist. Is this command now obsolete? If so, what is the new command to add a random variable to a FE model? For example, how would you add a nodal load, as a random variable to an FE model? I was also wondering if there was any manual available for reliability analysis in opensees?

I would really appreciate any assistance,
Kind Regards,
Lorcan Connolly

fmk
Site Admin
Posts: 5883
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: RandomVariablePositioner

Post by fmk » Wed Jun 18, 2014 8:00 am

Lorcan,

while we await Mikes reply you can have a look at some simple examples here:

http://opensees.berkeley.edu/WebSVN/lis ... ability%2F

i also suggest reviewing Mikes seminar on the subject:

http://opensees.berkeley.edu/wiki/index ... g_OpenSees

lorcancon
Posts: 25
Joined: Thu May 22, 2014 5:22 am
Location: Roughan & O Donovan

Re: RandomVariablePositioner

Post by lorcancon » Thu Jun 19, 2014 6:51 am

Frank/Mike,

I'll look through those examples. In the meantime, would it be possible for you to run my code and see if you get the same error? The code is aimed at calculating the probability of the displacement of a simple beam exceeding 60mm with a central point load as a random variable ~N(460,46):
# Reliability analysis of Simplebeam

# OpenSees reliability analysis -- Lorcan Connolly (lorcan.connolly@rod.ie)

wipe
wipeReliability

model basic -ndm 2 -ndf 3
reliability

set m 1.0
set mm [expr 0.001*$m]
set kN 1.0
set N [expr 0.001*$kN]
set MPa [expr 1000*$kN/($m*$m)]

node 1 [expr 0.0*$m] [expr 0.0*$m]; fix 1 1 1 0
node 2 [expr 5.0*$m] [expr 0.0*$m];
node 3 [expr 10.0*$m] [expr 0.0*$m]; fix 3 1 1 0

# Define all structural properties
set E [expr 200000.0*$MPa]
set Ecov 0.05

set fy [expr 300.0*$MPa]

set b 0.1
set d 0.5
set I [expr $b*$d*$d*$d/12]


geomTransf Linear 1

element elasticBeamColumn 1 1 2 [expr $d*$b] $E $I 1
element elasticBeamColumn 2 2 3 [expr $d*$b] $E $I 1

parameter 1 loadPattern 1 loadAtNode 2 2
randomVariable 1 normal 460 46; set param(1) 1
randomVariablePositioner 1 -rvNum 1 -parameter 1

performanceFunction 1 "0.06-\[nodeDisp 2 2\]"

sensitivityIntegrator -static ;# or -transient
sensitivityAlgorithm -computeByCommand -parameters

set Nsteps 100
randomNumberGenerator CStdLib
probabilityTransformation Nataf -print 3
reliabilityConvergenceCheck Standard -e1 1.0e-4 -e2 1.0e-4 -print 1
functionEvaluator Tcl -file "analyze $Nsteps"
gradientEvaluator Implicit; #FiniteDifference
searchDirection iHLRF; #Improved Hasofer and Lind, Rackwitz and Fiessler
meritFunctionCheck AdkZhang -multi 2.0 -add 50 -factor 0.5; # determine the suitability of a step size
stepSizeRule Armijo -maxNum 10 -base 0.5 -initial 0.3 5; #Max No. step size reductions allowed before accept, base number for stepsize, initial value in Eq
stepSizeRule Fixed -stepSize 1.0; #This seems to be overuling the previous line and setting constant step size.
startPoint Mean; #Where to start searching
findDesignPoint StepSearch -maxNumIter 20;# -printDesignPointX designPointX.out

runFORMAnalysis simple.out;

Thanks again,
Lorcan.

Post Reply