Non-spherical particles by clumping multiple spheres together

Submitted by marc on Thu, 09/01/2011 - 16:52

Hi everyone,

I'm trying to set up a simulation with 3 spheres clumped together, exactly as seen in LIGGGHTS Features > Non-spherical particles > Example - pouring a continuous stream of non-spherical particles, where can be seen particles made of 3 spheres glued together.

I'm exploring some possibilities to do a modification of this study http://www.nt.ntnu.no/users/skoge/prost/proceedings/ecce6_sep07/upload/1... by simulating a packed bed with non-spherical particle as my PhD study. My idea is starting with 3 spheres glued and then move on with different particle geometries.

As I'm still very new in LIGGGHTS, I'm having dificulties in doing this simulation.

Has someone did something similar to that example and could share some of the structure to set up this simulation?

Thank you in advance for your help.

All the best

Marc

Forums:

willsmithumich | Tue, 11/13/2012 - 22:13

I too am interested in using non-spherical particles in my simulations. The example of non-spherical-particles under "LIGGGHTS Features" has been available for some time, yet I have not seen an update to the code.

I could try to create many rigid bodies, or perhaps use a hybrid interaction to provide bonds, but these seem like poor work-arounds. I would greatly appreciate any help or hints you can provide to make this feature available.

dbxmcf | Mon, 04/08/2013 - 21:05

Hi, Christoph:

I checked out the branch above and successfully complied this branch using the command "make fedora", however when I entered the bondsexample directory and typed the command "liggghts < in.template", it prompts an error:

[dbxmcf@dbxmcf-1 bondsexample]$ liggghts_bond < in.template
LIGGGHTS (Version LIGGGHTS-PUBLIC 2.2, compiled 2013-04-08-13:32:22 by ccava based on LAMMPS 20 Apr 2012)
ERROR: Invalid atom style (atom.cpp:343)

Because there is currently no document on this branch, I am wondering if I can actually use this branch for bonded particles?

Thanks a lot!

Roy

dbxmcf | Mon, 04/08/2013 - 22:54

Hi, Christoph:

I have figured that out by finding one README file in the src folder, I copied all bond-related files in to the src folder and re-compiled LIGGGHTS, the in.template seems to be working now, thank you for providing the git branch, I am also wondering is there any major bugs/problems with this branch?

Thanks!

Roy

ckloss's picture

ckloss | Wed, 04/10/2013 - 08:11

Hi Roy,

>>I am also wondering is there any major bugs/problems with this branch?
this was a student's project and is really in beta stadium. At the moment it's not actively supported by us. If it would, it would have been merged into LIGGGHTS-PUBLIC already. You can try emailing Patrick [polyun.p (AT) gmail] if you have questions

Hope I could help

Cheers
Christoph

dbxmcf | Thu, 04/11/2013 - 20:49

Hi, Christoph:

Could you please tell the major differences between the branch and the bonded model you mentioned here:

node/525

Thanks a lot!

Roy

ckloss's picture

ckloss | Fri, 04/19/2013 - 15:08

Hi Roy,

the bonded particle model mentioned at node/525 is for LIGGGHTS 1.5.3 and thus outdated

Cheers
Christoph

Madadi | Mon, 05/06/2013 - 13:14

Hi Christoph
I tried to compile your example of "Bonded Particle" before and there were an error : ERROR: Invalid atom style (atom.cpp:343)
I saw the video files in branch and one of them may help me in my PhD project : (bondedregionfallsoncone.avi)
Unfortunately I encountered to the same error (ERROR: Invalid atom style (atom.cpp:343)) again!
I also sent mail to Patrick [polyun.p (AT) gmail] you suggested, but didn't get the answer.
please guide about this example (bondedregionfallsoncone.avi).
Regards
Hossein

dbxmcf | Tue, 05/07/2013 - 16:58

Hi, Hossein:

Can you show your steps to repeat this problem? What did you do when re-compiling LIGGGHTS? Did you copied all the source files in bondpackage/ folder and then type "make fedora"?

RL

Madadi | Wed, 05/08/2013 - 08:19

I copied all the source files in src folder, but what do you mean about "make fedora"? I typed that but didn't work.
Thanks
Hossein

dbxmcf | Sun, 05/12/2013 - 06:55

Hi, Hossein,

Can you copy and paste all your screen output so others can know what exactly happened on your computer? thank you!

Roy

Madadi | Mon, 05/13/2013 - 07:41

Hi Roy

hossein@hossein:~/Desktop/LIGGGHTS-PUBLIC/examples/LIGGGHTS/Tutorials_public/Bond1$ ~/Desktop/LIGGGHTS-PUBLIC/src/lmp_openmpi < in.template
LIGGGHTS (Version LIGGGHTS-PUBLIC 2.2.4, compiled 2013-03-16-16:31:45 by hossein based on LAMMPS 20 Apr 2012)
ERROR: Invalid atom style (atom.cpp:343)
hossein@hossein:~/Desktop/LIGGGHTS-PUBLIC/examples/LIGGGHTS/Tutorials_public/Bond1$

When I change atom style to granular, an error of pair style appears and etc.
Hossein

dbxmcf | Mon, 05/13/2013 - 18:57

Hi, Hossein:

You might need to re-compile LIGGGHTS after you have copied all bonds related files to your src folder, i.e. put all files in src/bondspackage to src folder and the re-compile LIGGGHTS.

Roy

paulbrumby | Wed, 09/14/2011 - 09:53

Hi Marc,

I have not done anything like this yet but (and I could be wrong) I think the developers are working on including multisphere and/or particle bonds in the future. These sound like they could be useful for your research.

All the best for your PhD.

Best wishes,

Paul

anandmds's picture

anandmds | Wed, 07/01/2015 - 13:07

I am new to Liggghts so can u tell me which src folder do we have to paste the bond files ? /usr/bin or /usr/local or where ? and what is the procedure after that

Anand.M
M.S. Mechanical Engineering,
I.I.T. Madras