Using rigid link elements with zero length elements!

[amkopensees]

I have a very simple model as follows. The models is two rigid link beam elements which are connected to each other with a zero length element. I want to pull one end of one of the rigid elements and evaluate the deformation and force of the rigid link element. However it returns wrong answer.
Would you please help me with this?

# Testing a high type fixed bearing in longitudinal direction
wipe

# Model builder
model BasicBuilder -ndm 2 -ndf 3

set npile 1.2
set keff [expr $npile*7.0]
set k1pile [expr 2.333*$keff]
set d1pile 7.62
set f1pile [expr $d1pile*$k1pile]
puts "f1pile= $f1pile"
set k2pile [expr 0.428*$keff]
set d2pile 25.4
set f2pile [expr ($f1pile+$k2pile*($d2pile-$d1pile))]
puts "f2pile= $f2pile"
set f3pile [expr 0.999999999*$f2pile]; # It does not accept to have f2pile and f3pile exactly the same!!!!!
set d3pile [expr 1.000000001*$d2pile]
puts "f3pile= $f3pile"
puts "d3pile= $d3pile"

# uniaxialMaterial Hysteretic $matTag $s1p $e1p $s2p $e2p <$s3p $e3p> $s1n $e1n $s2n $e2n <$s3n $e3n> $pinchX $pinchY $damage1 $damage2 <$beta>
uniaxialMaterial Hysteretic 1 $f1pile $d1pile $f2pile $d2pile $f3pile $d3pile [expr -$f1pile] [expr -$d1pile] [expr -$f2pile] [expr -$d2pile] [expr -$f3pile] [expr -$d3pile] 0.85 0.45 0.0 0.0 0.15


# Defining nodes
node 1 0. 0.
node 2 1000. 0.
node 3 1000. 0.
node 4 2000. 0.

#Boundary conditions
fix 1 1 1 1;
fix 2 0 1 1
fix 3 0 1 1
fix 4 0 1 1

# transformation:
geomTransf Linear 1

# Defining element
rigidLink beam 1 2
element zeroLength 1 2 3 -mat 1 -dir 1
rigidLink beam 3 4

recorder Node -file disp.out -node 4 -dof 1 disp
recorder Node -file reaction.out -node 1 -dof 1 reaction

recorder Element -file force.out -ele 1 force
recorder Element -file deformation.out -ele 1 deformation

pattern Plain 1 "Linear" {
load 4 1.0 0.0 0.0
}

integrator DisplacementControl 4 1 0.10

# system SparseGeneral -piv;
system BandGeneral;
constraints Penalty 1.0e-12 1.0e-12
test NormUnbalance 1.0e-5 100000
numberer Plain;
algorithm Newton

analysis Static

analyze 100

[gor]

Do you really need the rigid links? It looks like you could just fix node 2 and apply the load to node 3 to get the same result

[amkopensees]

Yes, you are right. But actually I need to make this work for another model. I want to know why this simple model does not work.

[gor]

Ah ok, that makes more sense. What do you mean by wring answer? What does it say for the unit load?

[amkopensees]

Its just meaningless random numbers. If you run the model you will see what I mean.

[gor]

You need to put a wipe command at the end of the script to print the last line to the .out files

I presume you mean the 1.0e-14 numbers for the forces are the meaningless ones? If you change your constraints from Penalty to Transformation, it works fine and the forces are around 150(units?) at the end of the analysis. For the reaction forces, these are zero because you are recording the nodes that you fixed and have rigid links attached to them, so the reaction forces are actually at the other end of the rigid links (nodes 2 and 3, instead of 1 and 4) because 2 and 3 have now been 'fixed' because of the rigid links

Hope this helps