time history analysis doesn't complete

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

Moderators: silvia, selimgunay, Moderators

Post Reply
Eli
Posts: 44
Joined: Sat Dec 22, 2012 5:16 am

time history analysis doesn't complete

Post by Eli »

Hi everybody
In time history analysis, what does the following meaning?

UmfpackGenLinSOE::setsize – n 60 nnz 684 IVal 13680
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: time history analysis doesn't complete

Post by fmk »

nothing .. that us just a debug statement that never got removed form the code telling me the number of equations, the number of non-zero entries in the matrix and the size of the array neeed for storage.
Eli
Posts: 44
Joined: Sat Dec 22, 2012 5:16 am

Re: time history analysis doesn't complete

Post by Eli »

When I multiply g by 5 in GMfatt (set GMfatt [expr $g*5.0])
Time history analysis doesn’t converge. While before that everything is ok.
How can I solve it?

wipeAnalysis
loadConst -time 0.0
set outFile tabas.txt
set Nsteps 1642
set dt 0.02;
set g 9.81

set GMfatt $g;

set gamma 0.5
set beta 0.25;
set alphaM 0.446;
set betaK 0.00426;
set betaKcomm 0
set betaKinit 0
system UmfPack
constraints Plain
test EnergyIncr 1.0e-4 1000;
algorithm NewtonLineSearch;
numberer RCM
integrator Newmark $gamma $beta
analysis Transient

set Gaccel "Series -dt $dt -filePath $outFile -factor $GMfatt";
pattern UniformExcitation 2 1 -accel $Gaccel;
analyze $Nsteps $dt;
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: time history analysis doesn't complete

Post by fmk »

do something in the analysis script in case the analyze command fails,

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

i have seen cases where there are 20 if {} blocks within the while loop!
Eli
Posts: 44
Joined: Sat Dec 22, 2012 5:16 am

Re: time history analysis doesn't complete

Post by Eli »

you mean i should use while loop?
what do you mean by: i have seen cases where there are 20 if {} blocks within the while loop!?


and another problem:
in UniformExcitation Pattern, when i change factor ( or ground motion records), the result of analysis e.g. displacement, does not change. what is the reason?
i set the mass at nodes.

wipeAnalysis
loadConst -time 0.0
set outFile tabas.txt
set Nsteps 1642
set dt 0.02;
set GMfatt 9.81;
set Gaccel "Series -dt $dt -filePath $outFile -factor $GMfatt";
pattern UniformExcitation 2 1 -accel $Gaccel;
set gamma 0.5
set beta 0.25;
system UmfPack
constraints Plain
test EnergyIncr 1.0e-4 1000;
algorithm NewtonLineSearch;
numberer RCM
integrator Newmark $gamma $beta
analysis Transient
analyze $Nsteps $dt;
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: time history analysis doesn't complete

Post by fmk »

you need to use a while loop.
while { ...} {
set ok [analyze 1 $dt]
if {$ok != 0} {

}
if {$ok != 0} {

}
..
..
}

2. if you change the factor it should change the response.
civil
Posts: 54
Joined: Sat May 25, 2013 3:44 am

Re: time history analysis doesn't complete

Post by civil »

Dear fmk

you wrote: "i have seen cases where there are 20 if {} blocks within the while loop!"

could you please write it here? It will be very useful.

Thanks in advance
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: time history analysis doesn't complete

Post by fmk »

no. sorry i don't have it and the student has long left stanford.
civil
Posts: 54
Joined: Sat May 25, 2013 3:44 am

Re: time history analysis doesn't complete

Post by civil »

Thanks for your attention.
is there any way to find it?
Post Reply