The OpenSees Community

For large structures, since there are only a few large-scale tests that have been done, the validation of the numerical response should be done at a component level. Then you should get back to some engineering fundamentals to check global and local responses.

For component validation, there is a lot of data out there. NEEShub has some data, PEER has many reports, and there are things in the literature for most common cases and failure modes.

I can provide more specifics off this post. Let me know.

Andre
--
http://web.engr.oregonstate.edu/~barbosa/


Luisa wrote:
> Dear Dr. McKenna,
>
> Hi again!
>
> As I understood, I should make my analytical models in an impeccable manner
> and according to accurate prediction of what is going to take place in a
> real physical system. In my M.Sc. thesis, however, I have 3D Reinforced
> Concrete Moment Frame models (6- 9- 12- and 15-story buildings indeed) and
> unfortunately I don't have any idea about their true behavior in real
> circumstances. I know this is not an OpenSees question somehow, but could
> you tell me please which of the experimental works (researches, reports,
> theses, ...) fulfilled on systems such as mine are highly reliable and
> include enough information for verification of my analyses results?
> Actually I need them as a reference to my study in order to gain a better
> understanding of the real behavior of my models, so that I can better
> decide on how to make my analytical models in OpenSees (elements, NIPs,
> ...).
>
> Many thanks for your invaluable help in advance.

For modeling a tapered beam, you can use the forceBeamColumn element and assign different sections to the element using the "-sections $secTag1 $secTag2...". You will need to assign a different section for each integration point along the element.

Please have a look at:
http://opensees.berkeley.edu/wiki/index ... mn_Element

"To change the sections along the element length, the following form of command may be used:
element forceBeamColumn $eleTag $iNode $jNode $numIntgrPts -sections $secTag1 $secTag2 ... $transfTag <-mass $massDens> <-iter $maxIters $tol> <-integration $intType>"

Hi Frank,

There seems to be a very small typo in the ElementOutput header. I think the sixth ResponseType should be Mz_1?


<?xml version="1.0" encoding="UTF-8"?>
<OpenSees
xmlns:xsi = "http://www.w3.org/2001/XMLSchema-instance"
xsi:noNamespaceSchemaLocation = "http://OpenSees.berkeley.edu/xml-schema ... enSees.xsd">
<TimeOutput>
<ResponseType>time</ResponseType>
</TimeOutput>
<ElementOutput eleType="ForceBeamColumn3d" eleTag="110101" node1="101" node2="37">
<ResponseType>N_ 1</ResponseType>
<ResponseType>Vy_1</ResponseType>
<ResponseType>Vz_1</ResponseType>
<ResponseType>T_1</ResponseType>
<ResponseType>My_1</ResponseType>
<ResponseType>Tz_1</ResponseType>
<ResponseType>N_2</ResponseType>
<ResponseType>Py_2</ResponseType>
<ResponseType>Pz_2</ResponseType>
<ResponseType>T_2</ResponseType>
<ResponseType>My_2</ResponseType>
<ResponseType>Mz_2</ResponseType>
</ElementOutput>

Thanks Frank.

Would something like is described in the following link work? Would OpenSeesMP have to be compiled with mpich2?

h***://wiki.mcs.anl.gov/mpich2/index.php/Hydra_Output_Redirection

For now, I tried the following "mpiexec -l -np 3 OpenSeesMP runAll.tcl > logFile.out 2>&1" where the "-l" option prints a prefix output with the process at the beginning of each line. Example of the output is below.

For the stdout it looks fine, but for some lines coming from the stderr, it does not write the prefix, which would be nice to have if I wanted to parse through the logFile for all stderr coming from one process, which in my case corresponds to some stderr from one analysis with one specific eqke record.

Thanks

[0] -----------------------------------------
[1] -----------------------------------------
[0]Gravity load analysis: starting
[1]Gravity load analysis: starting
[2] -----------------------------------------
[2]Gravity load analysis: starting
[2]Entering Model.Build.tcl
[0]Entering Model.Build.tcl
[1]Entering Model.Build.tcl
[1]Gravity Analysis: starting
[2]Gravity Analysis: starting
[0]Gravity Analysis: starting
[2]CTestEnergyIncr::test() - iteration: 1 current EnergyIncr: 2973.19 (max: 1e-012)
[1]CTestEnergyIncr::test() - iteration:
1 current EnergyIncr: 2973.19 (max: 1e-012)
[0] OpenSees -- Open System For Earthquake Engineering Simulation
[0] Pacific Earthquake Engineering Research Center -- 2.2.2
[0] (c) Copyright 1999,2000 The Regents of the University of California
[0] All Rights Reserved
[0] (Copyright and Disclaimer @ http://www.berkeley.edu/OpenSees/copyright.html)
[0]CTestEnergyIncr::test() - iteration: 1 current EnergyIncr: 2973.19 (max: 1e-012)
[2]CTestEnergyIncr::test() - iteration: 2 current EnergyIncr: 4.50689 (max: 1e-012)
[1]CTestEnergyIncr::test() - iteration: 2 current EnergyIncr: 4.50689 (max: 1e-012)
[0]CTestEnergyIncr::test() - iteration: 2 current EnergyIncr: 4.50689 (max: 1e-012)
[2]CTestEnergyIncr::test() - iteration: 3 current EnergyIncr: 0.0677937 (max: 1e-012)
[1]CTestEnergyIncr::test() - iteration: 3 current EnergyIncr: 0.0677937 (max: 1e-012)
[0]CTestEnergyIncr::test() - iteration:
3 current EnergyIncr: 0.0677937 (max: 1e-012)
[2]CTestEnergyIncr::test() - iteration: 4 current EnergyIncr: 3.85188e-006 (max: 1e-012)
[1]CTestEnergyIncr::test() - iteration:
4 current EnergyIncr: 3.85188e-006 (max: 1e-012)
[0]CTestEnergyIncr::test() - iteration: 4 current EnergyIncr: 3.85188e-006 (max: 1e-012)
[2]CTestEnergyIncr::test() - iteration: 5 current EnergyIncr: 2.5565e-013 (max: 1e-012)
[1]CTestEnergyIncr::test() - iteration: 5 current EnergyIncr: 2.5565e-013 (max: 1e-012)
[2]Gravity analysis: done - Model.Build.tcl
[2] -----------------------------------------
[2]Sylmar_OV - Gravity load analysis: 10 seconds
[2] -----------------------------------------
[2]Sylmar_OV - AnalysisType analysis: starting...
[2]Sylmar_OV - Dynamic.EQ.bidirect analysis: complete...
[2]$Sylmar_OV - AnalysisType analysis: 0 seconds
[2] --------------- End Analysis -----------------
[1]Gravity analysis: done - Model.Build.tcl
[1] -----------------------------------------
[1]LGPC - Gravity load analysis: 10 seconds
[1] -----------------------------------------
[1]LGPC - AnalysisType analysis: starting...
[1]LGPC - Dynamic.EQ.bidirect analysis: complete...
[1]$LGPC - AnalysisType analysis: 0 seconds
[1] --------------- End Analysis -----------------
[0]CTestEnergyIncr::test() - iteration:
5 current EnergyIncr: 2.5565e-013 (max: 1e-012)
[0]Gravity analysis: done - Model.Build.tcl
[0] -----------------------------------------
[0]Tabas - Gravity load analysis: 11 seconds
[0] -----------------------------------------
[0]Tabas - AnalysisType analysis: starting...
[0]Tabas - Dynamic.EQ.bidirect analysis: complete...
[0]$Tabas - AnalysisType analysis: 0 seconds
[0] --------------- End Analysis -----------------
[1]Process Terminating 1
[0]Process Terminating 0
[2]Process Terminating 2

Maybe what may be useful is something like the following command to output absolute accelerations (velocities and displacements, also) in X, Y and Z, for example:

recorder Node -file output.out -timeSeries $IDloadTagX $IDloadTagY $IDloadTagZ -time -node 1 -dof 1 2 3 accel

Thanks

Hi,

How can I redirect stdout and stderr for the process running in a $pid to a different file (see example below)?

I tried the following command that outputs all stdout and stderr from all $pid to the same logFile: mpiexec -np 3 OpenSeesMP runAll.tcl > logFile.out 2>&1

Thank you
Andre
*****************************************************************
set pid [getPID]
set numP [getNP]
set count 0

foreach {EqName } {
Tabas
LGPC
Sylmar_OV } {

if {[expr $count % $numP] == $pid} {
puts "$EqName do something"
}

incr count 1
}

I agree. Why don't you add it to future requests?

Thank you for the reply. I had also reached that conclusion that for now it does not seem to be possible, but by trying out on a cantilever SDOF. In an effort to reduce the number of recorders for a big structure, it might be worth adding an option that for a 3D case could look something like "-timeseries $accelX $accelY $accelZ", in which the three component rigid base accelerations ($accelX $accelY $accelZ) would be added to the relative acceleration response obtained by the node recorder "-dof 1 2 3 accel". Just an idea. Thanks

Thank you for the reply. It worked well.

One more question. Would the format be the one I have in the next line if I want to use one recorder to output the absolute acceleration in more than one direction, for example X and Y?

recorder Node -file output.out -timeSeries $IDloadTagX $IDloadTagY -time -node 1 -dof 1 2 accel

By the way, it might be worth changing the very nice and clear example of the eigen analysis shown available in: http://opensees.berkeley.edu/wiki/index ... hear_frame

Thanks again for the good job.

Thank you very much again.

Hello,

If I wanted to define the UniformExcitation using timeSeries so that I could get the absolute nodal acceleration directly using a recorder, can my commands look like the following?

timeSeries Path $IDloadTag -dt $dt -filePath $outFile -factor $GMfact
pattern UniformExcitation $IDloadTag $GMdirection -accel $IDloadTag; # create Uniform excitation


For now, I have been doing my earthquake analysis by setting the commands below, which seems to be working well, but the nodal acceleration output is the relative acceleration and it can be cumbersome sometimes to add the ground acceleration to the recorder output after the analysis is complete, especially when an adaptive time stepping scheme is used.

set AccelSeries "Series -dt $dt -filePath $outFile -factor $GMfact"; # time series information
pattern UniformExcitation $IDloadTag $GMdirection -accel $AccelSeries; # create Uniform excitation

Thank you

That makes perfect sense now. Thank you for that very clear response.

Dear All,

I realize the eigenvalues are computed when we set the [eigen $numModes] command without using the analyze command. Is it also possible to output the eigenvectors without invoking the analyze command? If not, I would like to understand why not?

Thank you

Thank you. Have a good weekend.