Hi all,
I am running a simple case with 2 particles bouncing in a closed box.
At certain time, I can see there is a gap between particles and the wall but still it detects as a contact which seems to be wrong. (visualization in paraview )
I am writing the history of contact-area in the dump files called "localwall.contact" . You might be able to view this file when you run it.
In my case, it happens at 9579, 28851,48123 time-steps and so on i.e it happens for every 19272 time-steps after the first wrong detection.
Can anyone please have a look into this?
I attached my data-files and mesh file to this post.
Thank you.
| Attachment | Size |
|---|---|
 box-2particles.tar_.gz | 10.83 KB |
richti83 | Thu, 02/12/2015 - 19:36
investigation
Hi clarence, I checked your script but can not confirm that there is any problem. But as your dumpstep is 50 iterations I wonder how you can see step 9579 ?
I changed to dumpstep=1 and let it run again, I've seen that the particles are at the bottom of your box between 9579 and 9694.
http://www.richtisoft.de/transfer/9579.png
http://www.richtisoft.de/transfer/9694.png
I made this little animation for you (note i scaled radius to 0.5 to make the contacts visible !):
https://www.youtube.com/watch?v=Ye97H9-7svc&feature=youtu.be
Dumpstep=50. Runtime 50.000 iterations
Script is here:
http://www.richtisoft.de/transfer/in.box
Maybe your problem is related to this bug:
http://www.cfdem.com/forums/sign-error-fixwallgran-relating-cwl-addwall1
try changing line 864 in fix_wall_gran.cpp to vectorSubtract3D and recompile liggghts.
https://github.com/richti83/LIGGGHTS-WITH-BONDS/commit/60ae395908c1b8d5c...
Do you use Glyph filter or PointSprites in PV to visualize Particles ? It's known that pointsprites does not scale well when you zoom to close. When you use glyphs check your settings for scalefactor=1, radius=1, scalars=radius)
Did you solve your plan of bookkeeping walls ? What do you simulate ? Some kind of fuel cell, battery ? Where are you located ?
Good Luck,
Christian
clarence91 | Thu, 02/12/2015 - 22:29
I really apprieciate for your help.
Hi richti,
I just realized that mistake was from my side.
I was visualizing in other direction and I thought 2 particles are just in the middle of the box. I forgot about gravity which is acting in z-direction(-z) and even about particles which were hitting the bottom wall.
Thanks alot.
If you dont mind can you tell me what option did you use to show the contacts between wall-particle or particle-partilcle using paraview ? (your video)
I couldnt solve the bookkeeping of wall .
I am working on my master's thesis (particle charging due to contact). I am from USA.
Thank you.
richti83 | Fri, 02/13/2015 - 09:01
visualisation of contacts
Hi,
I used my forcechainreader for ParaView, you can download a precompiled for PV 4.3.1 64 bit linux here:
https://github.com/richti83/ParaView_Reader_for_LIGGGHTS/tree/master/pre... (use raw-> save as, in PV Tools->manage plugins->load new load the downloaded libliggghts_forcechainreader.so)
Than use exactly the dump commands from my example script for dmp_pc and dmp_wc. To make the zylinder I appended a TUBE filter to the contact chains.
Best,
Christian
clarence91 | Fri, 02/13/2015 - 17:00
Thanks alot.
Thanks alot.