Hi,
I have a DEM code of my own and I would like to use ParaView to postprocess the results. I have been following the procedure for converting LIGGGHTS dump.* files to *.vtk files, and my ultimate goal is to figure out the format of *.vtk files for DEM type results. The problem is I don't have an example dump.* file (e.g. dump.cohesion) to run through the pizza.py process. I tried using a LAMMPS output file that I found on the web, but that did not work.
Could someone please provide me with a simple LIGGGHTS dump.* file so that I can finally create a *.vtk file from DEM data? Equally helpful would be the corresponding *.vtk file as well, so that I can compare my output.
One last question, does the pizza.py process create the old style ASCII *.vtk file or the new XML type (e.g. *.vtp, *.vtu etc.) of output? I have written a writer for my LBM code to create the XML output, which allows compression which is very helpful.
Thanks,
Chris
mahdi_roozbeh | Wed, 11/23/2011 - 05:59
Hi
If I was getting your question right, the answer can be as follows:
There are many examples in liggghts. you can use one of them to create vtk for paraview. The ligggts can be easily downloaded.
What you can do is to run the file with liggght and after that according to input script in examples one file will be created in post folder which is already defined in dump command.
after that you must open your terminal and going to your post folder and run the pizza.py
after that write these commands in order.
d=dump("dump.pour") # pour can be changed according to your example
d.scale()
v=vtk(d)
v.manyGran()
cheers
jc117818 | Wed, 11/23/2011 - 14:56
Difficult for me to run LIGGGHTS
Hi,
Thanks for your reply. All of our computation here is done using the Windows platform so it is difficult for me to get set up with LINUX to run LIGGGHTS.
I have mastered the pizza.py conversion process (I think, I tested it with a LAMMPS dump I found on the internet) but I just need a very simple LIGGGHTS dump file to test it for DEM. A file with only a few particles and some radius, displacement, velocity etc. information is all I need. If anyone can provide a simple file, it would be greatly appreciated.
Cheers,
Chris
richti83 | Fri, 11/25/2011 - 10:10
why double conversion ?
Hi Chris,
In my opinion the easier way is to make a custom reader plugin for paraview for your own dump style.
You can use my code here:
node/620
and adapt the RequestData subroutine in liggghts_reader.cxx.
The ligghts dump file looks for a single dump step like this:
Line0-Line9: unneeded dump information
Line10 and following: ATOMS-id type type x y z ix iy iz vx vy vz fx fy fz omegax omegay omegaz radius
so I splitted every line in double vars and append them to the paraview pipeline in 3xN Array-styles.
In this case every dump-timestep a new dump-file is generated and the existing paraview files-series-reader-class manages loading the right file.
Normaly liggghts generates one file containing every timestep, so pizza has to split the big dump file in many vtk files which is very cost-intensive.
greetings,
Christian
jc117818 | Mon, 11/28/2011 - 16:44
Still unclear on *.vtk format
Hi Christian,
Thanks for your reply. You are right, there is no need for the double conversion. Ultimately I want to code a ParaView writer into my DEM code, but I am still trying to determine exactly what the format for DEM data is in *.vtk files. I thought that I could use pizza.py with an example LIGGGHTS dump file to create a *.vtk file which I could then look at to understand the format. Unfortunately, I have not been able to get my hands on an example LIGGGHTS dump file (and I cannot run LIGGGHTS on my current system) or an example *.vtk file which contains an output of LIGGGHTS data.
Do you have a simple LIGGGHTS dump example file or a simple *.vtk example file that I could look at so I can reverse engineer my own ParaView writer?
Thanks,
Chris