analyze speed

[gorica]

i have made some reinforced concerete structure models . i want to do time history analysis (actually IDA). but there is a problem :

i want to run them at the same time and on the same computer . the CPU ability is goog enough how an analyze just gets 13% of the CPU.

when i run 6 models at the same time they get ( 78%=6*13% ) of the CPU , but the analyze speed decreases .
on the other hand ther is no limitation by the RAM .

i did 6 analyzes in 6 different folders ( in all folders i had opensees.exe file) , but nothing changed . for running i opened opensees.exe and typed " source model.txt "

is there any way to do them in a parallel manner with no decreasing analyze speed for each of them .

thanks for useful guides .

[Ameri]

For parallel processing you can use either OpenSeesSP or OpenSeesMP. Visit the following page for more information: http://opensees.berkeley.edu/OpenSees/p ... rallel.php

[fmk]

13% is abysmal is this any reasonably size model and not some little frame that runs in under 10 sec. .. try cutting down on output (especially puts commands) .. output kills program speed.

[gorica]

my model has 7 stories ( which are rigid diaphram ) and 2 drift recorder for each story . totally i have 7*2=14 recorder .
so the number of recorders is not alot .

i have 3 questions :

1- i did not define torsional and shear stiffness for my RC sections , does it increase analyze time ?

2-why runing similar models in different folders at the same time is not like openseesMP?

3- in following lines , you can see last lines of the model which are for convergence , do you have any suggestion for that to decrease the analyze time ?

#------------------------------------------------------------------------
variable constraintsTypeDynamic Transformation;
constraints $constraintsTypeDynamic ;

variable numbererTypeDynamic RCM
numberer $numbererTypeDynamic

variable systemTypeDynamic BandGeneral;
system $systemTypeDynamic

variable TolDynamic 1.e-6;
variable maxNumIterDynamic 10;
variable printFlagDynamic 0;
variable testTypeDynamic EnergyIncr;
test $testTypeDynamic $TolDynamic $maxNumIterDynamic $printFlagDynamic;

variable maxNumIterConvergeDynamic 2000;
variable printFlagConvergeDynamic 0;

variable algorithmTypeDynamic ModifiedNewton
algorithm $algorithmTypeDynamic;

variable NewmarkGamma 0.5;
variable NewmarkBeta 0.25;
variable integratorTypeDynamic Newmark;
integrator $integratorTypeDynamic $NewmarkGamma $NewmarkBeta

variable analysisTypeDynamic Transient
analysis $analysisTypeDynamic



set Nsteps [expr int($TmaxAnalysis/$DtAnalysis)];
set ok [analyze $Nsteps $DtAnalysis];

if {$ok != 0} { ;
set ok 0;
set controlTime [getTime];
while {$controlTime < $TmaxAnalysis && $ok == 0} {
set controlTime [getTime]
set ok [analyze 1 $DtAnalysis]



if {$ok != 0} {
puts "Trying NewtonWithLineSearch .."
algorithm NewtonLineSearch .8
set ok [analyze 1 $DtAnalysis]
algorithm $algorithmTypeDynamic
}





if {$ok != 0} {
puts "Trying Broyden .."
algorithm Broyden 8
set ok [analyze 1 $DtAnalysis]
algorithm $algorithmTypeDynamic
}





if {$ok != 0} {
puts "Trying Newton with Initial Tangent .."
test NormDispIncr $Tol 1000 0
algorithm Newton -initial
set ok [analyze 1 $DtAnalysis]
test $testTypeDynamic $TolDynamic $maxNumIterDynamic 0
algorithm $algorithmTypeDynamic
}





}
}; # end if ok !0

[fmk]

you do know that not everything has to be a variable:
constraints Transformation
numberer RCM
system BandGeneral

...

anyway
1. no
2. yes, it's just easier to start 1 using 1 input file as opposed to starting may with different input files or arguments
3. if model is big enough, use a different solver.