How to use PVDRecorder (Version 2.5.0. Release Tag 6236

[sb1966]

Hi fmk and all,
The change log of version 2.5.0 (R Tag 6236) talks about PVD recorder for viewing in Paraview. This is a welcome development and I thank the developer for the same.
My questions are:
1. Where can I find documentation for the same?
2.What is the the OpenSees cammand for using the same?
3. What are the valid queries for PVDRecorder?
Are the queries same as the ordinary node or element recorders?

Thanks.

[sb1966]

Dear All,
May I request you, colleagues who have used or have the knowledge on PVDRecorder to kindly give me the OS command? It is important as some of my graduate students need to use the new recorder in order to save their postprocessing efforts/time.

Hope I get some response from this otherwise responsive forum.

Regards and Thanks in advance.

[sb1966]

Dear All,
1. Would any one please enlighten me on the command syntax of PVD recorder available in version 2.5.0?
2. There is a command available for a (related?) software at http://web.engr.oregonstate.edu/~zhum/command.html. Is the same syntax applicable for OpenSees as well?

Hope to listen from you. I think it is important for many users.

Thanks.

[sb1966]

Dear All,
1. After waiting long for the required commands I tried the PFEM commands as described in the web page (my last post).
1a. The command syntax shown is as follows:

recorder pvd $filename <disp> <vel> <accel> <reaction> <mass> <eigen> <-precision $precision> <eleResponse >

Note: A directory named $fileName must pre-exist in the pwd for pvdRecorder to work. The eleResponse must be the last argument in the pvdRecorder
command.

2. Using these commands I am getting the node responses such as displacement, velocity acceleration to use in Paraview.
3. However, I have not been able to get the element response. I tried both "stresses" and stress" and also "strain" and "strains" for each of the Gauss points, one by one.
4. Would any one please guide me whether it is a bug or I have wrongly used any command ?

5. I have used the command as follows:

....
recorder pvd pvdRes disp vel accel eleResponse 1 material 1 stresses
....

Hope to listen from the community, especially from the developers.

Regards and Thanks

[fmk]

do those ele args work in a recorder

[sb1966]

Hi fmk
Thanks for your reply. Yes. They work in normal element recorder.

[sb1966]

Dear fmk,

1. I think your question was related to eleResponse command and not regarding Element recorder command.
2. I have tested it as follows (a single quad element in 2D domain):
--------------------------------------------------------------
set fileId [open OnlyStresses.out w]
set d [eleResponse 1 material 1 stresses]
puts $fileId $d
close $fileId
-------------------------------------------------------------
3. It has given me the following output for a particular time step:
---------------------------------------------------------------------------------
-3.345461 -6.237660 -3.833249 -0.333165 0.000000
------------------------------------------------------------------------------
4. The first three data are OK and matching with the output of corresponding element recorder

5. The last two data, I suppose are spurious.
6. I believe there should only be sigma_x, Sigma_y and sigma_xy in a 2D plane-strain problem in a gauss point.

6. So can I assume the pvdRecorder to be buggy at this moment?

Hope to listen from you soon.

[sb1966]

Dear fmk and all,
Waiting eagerly for your support on the issue.
Thanks in advance for your time.

[zhum]

In the command:

recorder pvd pvdRes disp vel accel eleResponse 1 material 1 stresses

Is the first "1" the element tag?

for the pvd recorder, the element tag is not needed, since it records for all elements, for example,

the eleResponse commands are like:

eleResponse 1 forces
eleResponse 1 material stress

the corresponding pvd recorder commands will be

pvd $filename disp vel eleResponse forces
pvd $filename disp vel eleResponse material stress

can you try again? Thank you!

[sb1966]

Dear Zhum,

Thanks a lot for your reply.
Now the pvdRecorder command is clear to me from your reply and I will try them now.

Infact, since I was unable to get any eleResponse recorded in pvdRecorder so I tried it along with element tag later.

Also would you please shed some light on my question 5 and 6 in my earlier post?

Thanks for your time and support. Have nice time ahead

[zhum]

For the eleResponse command outputs, I'm really not sure about the output formats and meaning. Sorry can't help with that.

[sb1966]

Dear Zhum,
Thanks a lot for your reply. I am sorry for my delayed response. I will concentrate on the issues raised in my questions 5 and 6 later. If i find some reasonable answer, i will let you know.

Thanks for your help. I am not yet able to test those commands, being away from my campus. If needed, i may further bother you. Hope you won't mind.

With best wishes and warm regards
Sukumar.

[sb1966]

Dear Zhum,
Sorry for my delayed response.
1. I was trying to use the pvdRecorder command as suggested by you (earlier post).
2. I have a single quad element excited by a sinusoidal base acceleration.
3. Used the following command to get pvd file for displacement, velocity acceleration and element stresses:
3a) recorder pvd pvdRes disp vel accel eleResponse material stress
3b) alternatively, recorder pvd pvdRes disp vel accel eleResponse material stresses
4. While opening the pvd file in Paraview, I am getting the displacement, velocity and accelerations as expected.
5. But could not find the stress results in any of the Gauss points.
6. Instead I get a field named compositevtkIndex which I could not interpret.
7. Opening vtu part files in text editor reveals that there is no field or values named as stress or stresses
8. I am pasting the last part (after the acceleration response) of a vtu file below:
====================================
</DataArray>
<DataArray type="Float32" Name="Acceleration" NumberOfComponents="3" format="ascii">
1.1419348469e-12 3.7229866373e-12 0.0000000000e+00
-9.0470044718e-13 3.7326133627e-12 0.0000000000e+00
-2.9949812432e-01 9.0649433984e-13 0.0000000000e+00
-2.9949812432e-01 -1.3268020652e-14 0.0000000000e+00
</DataArray>
</PointData>
<CellData>
<DataArray type="Int32" Name="ElementTag" format="ascii">
1
</DataArray>
</CellData>
</Piece>
</UnstructuredGrid>
</VTKFile>
======================================================
9. Kindly let me know what is wrong in my command or is it a bug?
10. I have compiled version 2.5.0 from source in Ubuntu 14.04, 64 bit.

Thanks and regards

[sb1966]

Hi all,

Finally I could locate the problem. I was using the unncessary input argument 'material' in the command.

An example of correct command would be:

recorder pvd $fileName <disp> <vel> <accel> <eleResponse stress >
recorder pvd $fileName1 <disp> <vel> <accel> <eleResponse strain>

To repeat, note that (i) the eleResponse must be the last argument,
(ii) directories named as $fileName and $fileName1 must preexist in the working directory.

However, I could not get desired results by combining both stress and strain in the same command.

Hope this is helpful for some one.

[clpgeo]

Dear All,

can someone advise on the possibility of undersampling the outputs of the pvd recorder according to a desired time step?

In this way, similarly to classical recorder, one may avoid to write on file huge amount of data.

Thank you in advance

Claudio