Nonlinear Viscous Damper

[kakue]

Now I am trying to analyze a simple 3-story base isolated structure using a stick model.
So, in the OpenSees file, there are 5 nodes. Node 2 is base floor, then Nodes 3-5 are for 1st, 2nd, and 3rd floors.
Below is the code I used.

wipe
model BasicBuilder -ndm 2 -ndf 2
node 1 0.0 0.0
node 2 350.0 0.0
node 3 700.0 0.0
node 4 1050.0 0.0
node 5 1400.0 0.0

mass 2 1.00 1.00
mass 3 1.00 1.00
mass 4 1.00 1.00
mass 5 1.00 1.00

fix 1 1 1
fix 2 0 1
fix 3 0 1
fix 4 0 1
fix 5 0 1

set kb 39.4784176043574
set ks1 2631.89450695716
set ks2 2193.2454224643
set ks3 1315.94725347858
set betab 0.0636619772367581
set betas 0.00190985931710275
set alphab 1
set alphas 1

uniaxialMaterial Elastic 1 $kb
uniaxialMaterial Elastic 2 $ks1
uniaxialMaterial Elastic 3 $ks2
uniaxialMaterial Elastic 4 $ks3
uniaxialMaterial Viscous 5 [expr $betab*$kb] alphab
uniaxialMaterial Viscous 6 [expr $betas*$ks1] alphas
uniaxialMaterial Viscous 7 [expr $betas*$ks2] alphas
uniaxialMaterial Viscous 8 [expr $betas*$ks3] alphas

# Create truss elements - command: element truss trussID node1 node2 A matID
element truss 1 1 2 350.0 1
element truss 2 2 3 350.0 2
element truss 3 3 4 350.0 3
element truss 4 4 5 350.0 4
element truss 5 1 2 350.0 5
element truss 6 2 3 350.0 6
element truss 7 3 4 350.0 7
element truss 8 4 5 350.0 8

set dt 0.020
set GM la21.txt
set accelSeries "Series -dt $dt -filePath $GM -factor 1"
pattern UniformExcitation 2 1 -accel $accelSeries
rayleigh 0.0 0.0 0.0 0.0
constraints Plain
numberer RCM
system UmfPack
test NormDispIncr 1.0e-2 50
algorithm KrylovNewton
integrator Newmark 0.5 0.25
analysis Transient

set RespDisp "NodeDisps.out"
set RespVel "NodeVels.out"
set RespAcc "NodeAccs.out"
set RespForce "ElemForces.out"

recorder Node -file $RespDisp -time -node 2 3 4 5 -dof 1 disp
recorder Node -file $RespVel -time -node 2 3 4 5 -dof 1 vel
recorder Node -file $RespAcc -time -node 2 3 4 5 -dof 1 accel
recorder Element -file $RespForce -time -ele 1 2 3 4 5 6 7 8 localForce

analyze 6000 0.01

For viscous material in the above code, Viscous Material #5 is used for the isolation system.
In the above case, it is normal linear viscous damping type, and it works fine.

However, when I change alphab from 1 to 0.1 to represent nonlinear viscous damping,
I also change betab from 0.0636619772367581 to 6.18874304475616 to have same energy dissipation at expected displacement
It does not work anymore. The damping force and acceleration at base become too high, and it cannot finish the analysis.
I have also tried changing alphab to 0.7. It did work actually, but the results are not what I expected.

Anyone has any ideas what's going on here????
I've been struggling with this for a few days already.

Thank you very much

[vesna]

The viscous material works only for values of alpha greater than 0.5. That is why you could not achieve convergence with alpha of 0.1.

[kakue]

Thank you very much ....

[payamdindar]

Dear kakue
What changes should make the code to work correctly?