mass of quad elements for direct soil structure interaction

[sgjsgj]

i wrote two posts later. may be my problem is very funny and simple.
i model the foundation and soil below the structure to consider soil structure interaction in direct approach. when i consider mass density for mass of quad elements, it gives period but results zero data. but when i consider nodal mass , it gives period and results reasonable data.

two simple sample are presented here. in the first case mass density and in the second nodal mass is used.


#sample 1
wipe; # clear memory of all past model definitions
model BasicBuilder -ndm 2 -ndf 2;
set gap -0.0001
set nm 55stg0;
puts "$nm"
set dataDir Data$nm; # set up name of data directory (you can remove this)
file mkdir $dataDir; # create data directory

nDMaterial ElasticIsotropic 3 20e9 0.3
nDMaterial ElasticIsotropic 4 2e9 0.3
set rho1 1000;
set fc 9.806;
set ac cape.acc;
set dts 0.02;
set alphaM 0.315504;
set betaKcurr 0.0;
set betaKinit 0.007868;
set betaKcomm 0.0;
set tol 1.0e-6;
set ite 20;
set dt 0.02;
set st 10;
###########nodes soil
node 1 0 0
node 2 2 0
node 3 2 2
node 4 0 2
node 5 2 4
node 6 0 4
element quad 1 1 2 3 4 1 "PlaneStrain" 4 0 $rho1 0 0
element quad 2 4 3 5 6 1 "PlaneStrain" 4 0 $rho1 0 0

fixY 0 1 1
###

set numModes 2
set lambda [eigen $numModes];

# calculate frequencies and periods of the structure
#---------------------------------------------------
set omega {}
set f {}
set T {}
set pi 3.141593

foreach lam $lambda {
lappend omega [expr sqrt($lam)]
lappend f [expr sqrt($lam)/(2*$pi)]
lappend T [expr (2*$pi)/sqrt($lam)]
}

puts "periods are $T"
set period "Data$nm/Periods.txt"
set Periods [open $period "w"]
foreach t $T {
puts $Periods " $t"
}
close $Periods
loadConst -time 0.0
# DYNAMIC ground-motion analysis -------------------------------------------------------------
set accelSeries "Series -dt $dts -filePath $ac -factor $fc"; # define acceleration vector from file (dt=0.01 is associated with the input file gm)
pattern UniformExcitation 2 1 -accel $accelSeries; # define where and how (pattern tag, dof) acceleration is applied
rayleigh $alphaM $betaKcurr $betaKinit $betaKcomm; # RAYLEIGH damping
#
recorder Node -file $dataDir/dsoil1.out -time -node 5 -dof 1 2 3 disp;
#constraints Transformation
constraints Penalty 1.0e18 1.0e18 ;# Transformation; #
test NormDispIncr 1.e-12 25 0
algorithm Newton
numberer RCM
system ProfileSPD # use Newton's solution algorithm: updates tangent stiffness at every iteration
integrator Newmark 0.5 0.25 ; # determine the next time step for an analysis
analysis VariableTransient; # define type of analysis: time-dependent
analyze $st $dt; # apply 1000 0.02-sec time steps in analysis
puts "Done!"


################
###############
###############
#sample 2
wipe; # clear memory of all past model definitions
model BasicBuilder -ndm 2 -ndf 2;
set gap -0.0001
set nm 55stg0;
puts "$nm"
set dataDir Data$nm; # set up name of data directory (you can remove this)
file mkdir $dataDir; # create data directory

nDMaterial ElasticIsotropic 3 20e9 0.3
nDMaterial ElasticIsotropic 4 2e9 0.3
#set rho1 1000;
set fc 9.806;
set ac cape.acc;
set dts 0.02;
set alphaM 0.315504;
set betaKcurr 0.0;
set betaKinit 0.007868;
set betaKcomm 0.0;
set tol 1.0e-6;
set ite 20;
set dt 0.02;
set st 10;
###########nodes soil
node 1 0 0
node 2 2 0
node 3 2 2
node 4 0 2
node 5 2 4
node 6 0 4
element quad 1 1 2 3 4 1 "PlaneStrain" 4 0 0 0 0
element quad 2 4 3 5 6 1 "PlaneStrain" 4 0 0 0 0
mass 3 2000 2000
mass 4 2000 2000
mass 5 1000 1000
mass 6 1000 1000
fixY 0 1 1
###

set numModes 2
set lambda [eigen $numModes];

# calculate frequencies and periods of the structure
#---------------------------------------------------
set omega {}
set f {}
set T {}
set pi 3.141593

foreach lam $lambda {
lappend omega [expr sqrt($lam)]
lappend f [expr sqrt($lam)/(2*$pi)]
lappend T [expr (2*$pi)/sqrt($lam)]
}

puts "periods are $T"
set period "Data$nm/Periods.txt"
set Periods [open $period "w"]
foreach t $T {
puts $Periods " $t"
}
close $Periods
loadConst -time 0.0
# DYNAMIC ground-motion analysis -------------------------------------------------------------
set accelSeries "Series -dt $dts -filePath $ac -factor $fc"; # define acceleration vector from file (dt=0.01 is associated with the input file gm)
pattern UniformExcitation 2 1 -accel $accelSeries; # define where and how (pattern tag, dof) acceleration is applied
rayleigh $alphaM $betaKcurr $betaKinit $betaKcomm; # RAYLEIGH damping
#
recorder Node -file $dataDir/dsoil1.out -time -node 5 -dof 1 2 3 disp;
#constraints Transformation
constraints Penalty 1.0e18 1.0e18 ;# Transformation; #
test NormDispIncr 1.e-12 25 0
algorithm Newton
numberer RCM
system ProfileSPD # use Newton's solution algorithm: updates tangent stiffness at every iteration
integrator Newmark 0.5 0.25 ; # determine the next time step for an analysis
analysis VariableTransient; # define type of analysis: time-dependent
analyze $st $dt; # apply 1000 0.02-sec time steps in analysis
puts "Done!"



#cape.acc # some data
-.3796879E-03 .4255504E-02 .4052873E-02 .8586178E-03 .3707899E-02
.8218798E-02 .4969572E-02 -.5254377E-02 .1444530E-02 .1197273E-02
.3349923E-03 .2318547E-02 -.2217702E-02 .1109265E-02 .7489674E-03
-.2087855E-02 -.9857360E-02 -.6751592E-02 .9615523E-03 -.4992248E-02
-.2765613E-02 .3343176E-02 -.4722938E-02 -.1023698E-01 -.9769704E-02

[vesna]

Thanks for letting us know about the bug. Until it is resolved I suggest you not to use mass density to define the mass.

[sgjsgj]

thanks a lot Vesna
i have one question else.
i use Hinton-Rock-Zienkiewicz (HRZ) method method to calculate nodal mass of non rectangular quad elements. for a big domain, i saw slithery differences to calculate period between mass density usage of quad elements and nodal mass by HRZ method. what method does opensees use to calculate lumped mass matrix.

best regard

[vesna]

You will have to check the source code to find that info.

[sgjsgj]

thanks a lot Vesna