The OpenSees Community

You can define a uniaxial Viscous material, which will define your C parameter: uniaxialMaterial Viscous $matTag $C $alpha (http://opensees.berkeley.edu/wiki/index ... s_Material)
Additionally, you can define your K parameter using an uniaxial Elastic material: uniaxialMaterial Elastic $matTag $E <$eta> <$Eneg> (http://opensees.berkeley.edu/wiki/index ... l_Material). This is slightly more complex, as you will have to express your E value through your desired sitffness, thus, depending on the complexity of the model, this might be a challenge.

These two materials can be used in element zeroLength as arguments, which will define your springs.

Also, fmk has given an example here: http://opensees.berkeley.edu/community/ ... ot#p103445

Hi,

I am modelling a soil column (ElasticIsotropic and PressureDependMultiYield02 materials) with an attached SDOF to it.
I attach the SDOF to the soil column using the following method:
----------------------------------------------------------------------------------------
set MidTopX [expr {int((($ColumnWidth/$MeshSize)+1)*($MeshSize/2))}]; #coordinate of the middle point at the surface of soil column
set MasterNode $NodeTags([expr {(int($ColumnWidth/$MeshSize)+1)}],1) #tag of this node

set n1 100000001; #bottom of SDOF
set n2 100000002; #free end of SDOF

model BasicBuilder -ndm 2 -ndf 2

node $n1 $MidTopX 0.0
node $n2 $MidTopX 1.0

equalDOF $MasterNode $n1 1 2
#Masternode has 3 DOF, therefore, equalDOF needed

mass $n2 $M 1e-18
#mass assigned to node 2 (top of SDOF) in X direction

uniaxialMaterial Steel01 20000001 0.118623 1.949551 0.03
uniaxialMaterial Viscous 20000002 0.139626 1
uniaxialMaterial Parallel 20000003 20000001 20000002

element truss 80000001 $n1 $n2 1 20000003
----------------------------------------------------------------------

Using this method, I get exactly the same displacements at the top of my SDOF for two completely different models.
What could be the reason for this?

Dear fmk,

could you please explain how this recorder works and what does it actually record? (11, 22, 12)?

I was struggling to find a good explanation on the wiki.

Hello,

I am using a zeroLengthContact2D element in my script to simulate the friction between the base of the SDoF and the soil deposit.
However, after running the script I end up with the following error:

WARNING FullGenLinLapackSolver::solve() - lapack solver failed - 1 returned
WARNING AcceleratedNewton::solveCurrentStep() -the LinearSysOfEqn failed in solve()
StaticAnalysis::analyze() - the Algorithm failed at iteration: 0 with domain at load factor 1
OpenSees > analyze failed, returned: -3 error flag

This is the section of my script, which describes the process:
----------------------------------------------------------------------------------------------------------
---Define SDoF
model BasicBuilder -ndm 2 -ndf 2

set n1 100000001
set n2 100000002
set n3 100000003

node $n1 1.0 0.0
#SDoF base node

node $n2 1.0 0.0

node $n3 1.0 1.0
#SDoF free end node

mass $n3 $M 1e-18
#mass assigned to node 3 (top of SDOF) in X direction

uniaxialMaterial Steel01 20000001 0.118623 1.949551 0.03
uniaxialMaterial Viscous 20000002 $c 1
uniaxialMaterial Parallel 20000003 20000001 20000002

element truss 80000001 $n2 $n3 1 20000003

equalDOF $n1 $n2 1 2
#move the base of SDoF together with the contact slave mode

---Define 2d contact element
set Kn 0.1
set Kt 0.1
#penalties in (m)
set mu [::tcl::mathfunc::tan [expr {$SoilFrictionAngle*($pi/180)}]]
#set friction coefficient
set Nx 0
set Ny 1

element zeroLengthContact2D 987654321 $n1 $MasterNode $Kn $Kt $mu -normal $Nx $Ny

---Analysis
system FullGeneral
test NormDispIncr 1.e-12 25 0
constraints Transformation
integrator LoadControl 1 1 1 1
algorithm KrylovNewton
numberer RCM
analysis Static

analyze 1

----------------------------------------------------------------------------------------------------------

I have read the topic http://opensees.berkeley.edu/community/ ... =2&t=54235, where a person had the same kind of problem and it seems like it is not a convergence issue, but a geometry inconsistency.
If you could point out where is the mistake and explain all the parameters of zeroLengthContact2D in a non-scientific manner I would be extremely grateful to you.

You should read them horizontally and it would be good to provide the script with the file, which has got only one number in a row.

Additionally, quite often you may want to apply the earthquake scaled to a certain peak ground acceleration, which cannot be done using the record handler from the examples.

Here is my handler of PEER records: https://www.dropbox.com/s/en5lwe9hmln5sxf/ParseAT2.tcl

This one creates two files - one is the original unscaled record in the right format and the second one is scaled to the desired PGA.

This script also creates global variables NPTS, DT and PGA so they could be later used in the script.

I assume that the reason is in your analysis bit and truncation/convergence errors.
Try to play around with different types of integrators, tests, etc.
At least, that's what was suggested earlier with a similar type of error:
http://opensees.berkeley.edu/community/ ... =2&t=60632

The thing is that I don't know the theory well as I did not had any dynamics modules yet.
So far I have made the model of my soil column and the model for SDoF.
The next step will be combining these two together.

While as the the soil models works (at least I get something I expected), the SDoF doesn't.

Here is the script of my SDoF model: http://pastebin.com/p3BNftxy (str2.tcl)
The input motion file is available here: http://pastebin.com/gVcwJyVC (STR_inc.out)

The input motion file contains 295 acceleration values with dT=0.01 at the point of interest after the analysis of my soil column.
However, after the SDoF displacement output file Node_disp.out (line 50) contains only 251 lines, while as Node_acc.out contains 289 lines, but still, not 295 as it should.

Could you please tell me what might be the reason for this problem?

Here is also the link to the folder with scripts: https://www.dropbox.com/sh/kti5veryrpfc ... hPx52wmOxa

Hello.

I am currently writing my dissertation on the SSI effect and I have already modelled the soil and SDoF separately.
At the moment, I am applying the earthquake accelerations at the bedrock, getting the accelerations at the ground level and applying them directly at the base of the SDoF.
I understand that this this de-coupled model neglects the SSI effect.

Hence, I am looking for an example, where SSI effect is studied on a SDoF.

If you could give me any recommendations on the modelling of the interface between the soil column and the SDoF or provide some examples (tcl scripts), your help would be much appreciated.

Many thanks!

Up.

All the links now seem to be broken.
It would be good if somebody could upload the latest version of Building Tcl somewhere.

Cheers.