Hi!
I was wondering if there is a way to create composite objects in LIGGGHTS made out of other objects. For example, building a tetrahedron out of 4 spherical atoms.
If this is not possible, are there other ways to create objects with more complex geometry that can move and interact like atoms? I would like to create these objects but still have the "smoothed" spring-like model for interaction between each other.
Stefan
ckloss | Mon, 02/28/2011 - 19:42
There is a fix rigid that
There is a fix rigid that does integration for spheres clumped together as a rigid body (as long as they do not overlap).
And we are currently working on a "multisphere" feature that is not in the public release - you can email me for more info.
Christoph
schteppe | Tue, 03/01/2011 - 11:17
Okay, I've tried using the
Okay, I've tried using the rigid fix. For me, it doesn't work though... I make a tetrahedron and want the particles to "stick together" like a rigid body but they don't. They are behaving like separate spheres.
The fix command is a bit ambigous about the groups in this case, maybe I just got it wrong? I've tried the following two commands:
fix tetrafix0 all rigid group 1 tetra0
fix tetrafix0 tetra0 rigid group 1 tetra0
Below is the full output for creating the particles and joining them in the rigid.
# Create tetra 0
# Create particle 1
create_atoms 1 single 1.000000 0.000000 2.000000 units box
Created 1 atoms
set atom 1 diameter 1.000000
Setting atom values ...
1 settings made for diameter
# Create particle 2
create_atoms 1 single 0.000000 1.000000 2.000000 units box
Created 1 atoms
set atom 2 diameter 1.000000
Setting atom values ...
1 settings made for diameter
# Create particle 3
create_atoms 1 single 0.000000 0.000000 3.000000 units box
Created 1 atoms
set atom 3 diameter 1.000000
Setting atom values ...
1 settings made for diameter
# Create particle 4
create_atoms 1 single 1.000000 1.000000 3.000000 units box
Created 1 atoms
set atom 4 diameter 1.000000
Setting atom values ...
1 settings made for diameter
group tetra0 id 1 2 3 4
4 atoms in group tetra0
fix tetrafix0 all rigid group 1 tetra0
1 rigid bodies with 4 atoms
The last line of output is comforting but sadly it does not affect my simulation :(
Maybe it has something to do with the integration? At the moment, I run this line to apply an integrator to the system. It affects all atoms.
fix integr nve_group nve/sphere
Stefan
ckloss | Tue, 03/01/2011 - 19:16
Fix rigid performs the
Fix rigid performs the integration in this case, and using fix nve/sphere messes this up, and each atom is integrated twice
You should have got a warning about this. Remove the nix nve/sphere and it should work
Christoph
schteppe | Wed, 03/02/2011 - 09:12
Thanks, you are right about
Thanks, you are right about the errors. How to apply gravity now? Neither of the below commands seem to work... :(
fix gravi tetra0 gravity 9.82 vector 0.0 0.0 -1.0
fix gravi all gravity 9.82 vector 0.0 0.0 -1.0
How do I make the rigid integrator aware of gravity?
ckloss_ | Thu, 03/03/2011 - 02:18
That's weird... are you using
That's weird... are you using the latest version (from the liggghtsdev git repository)? if yes, can you post the script?
Christoph
schteppe | Fri, 03/04/2011 - 09:07
I use a version that I
I use a version that I checked out from the git repos in january. Maybe I should update.
Anyway, here's the code:
http://upload.steffe.se/files/20110304/in.temp
And I also got a mesh in the simulation:
http://upload.steffe.se/files/20110304/box0.stl
ckloss_ | Fri, 03/04/2011 - 16:10
yes you should update, there
yes you should update, there was a bug in fix rigid... I would recommend 1.2.6 from the liggghtsdev repo.
If the error persists, let me know and I will have a look
Christoph
schteppe | Mon, 03/07/2011 - 15:58
Thanks! That did it! Just one
Thanks! That did it!
Just one small thing.. Starting the program outputs "LIGGGHTS 1.2.5-beta based on lammps-10Mar10". Did I get the wrong version or is this text not modified yet? :)
Btw, I don't know if you tested my code or not but the STL file contained some skewed normals. Can be good to know for future readers of this thread.
ckloss | Mon, 03/07/2011 - 16:00
yes, should be 1.2.6, but
yes, should be 1.2.6, but sometimes the counter does not increment correctly...
Christoph
htm | Tue, 12/13/2011 - 06:09
multisphere feature/pouring example
Hi Christoph,
Is the multisphere feature that is used for the break example of the git repository the same one used for pouring a continuous stream of non-spherical particles video shown in the LIGGGHTS features? Can you share the in file for this example?
thanks
Horacio