LIGGGHTS® - User Forum

I am studying the LIGGGHTS (lammps-13Jun13) for granular simulation under windows platform but I encountered a number of questions. It is a great appreciation if anyone could give me some clues.

Hello all,

I want to simulate a large number of rigid bodies in a granular simulation. Currently I am using fix rigid, which does not scale very well in parallel. Would I get an appreciable speed up if I use POEMS to handle the rigid bodies? Additionally, I was thinking about using very stiff lj/coul bonds to simulate the rigid bodies, would this increase my speed in parallel?

Thank you!
Zamir

HI all
I see doc but I dont understand what is the particlerate!or how particlerate insert particles?
in one example:
nparticles 1000 particlerate 5000 insert_every 1000
how particlerate can be biger than nparticle?
please help me

thank you

Hallo everyone,

I recently found out that the characteristic impact velocity (used for hookeian contact; described in pair_gran) have a high influence to the outcome of my simulation.

I'm not sure which is the meaning of this impact velocity and how to calculate it, maybe someone is able to explain it to me. Since im simulating an almost steady pebble bed, I assume I should set the impact velocity to zero or near zero. Is this correct?

Kind regards,
Philipp

Hi all

I'm simulating a ball mill and I would like to know when the particles have filled 15% of the ball mill volume in order to stop de insertion. Is there any way to do this adding a line in the script?

I'm filling the ball mill with insert/stream fix but I don't think this is the best way. is there any better way to insert the particles in this case? I would like to make start the spin of the ball mill sooner.

Best

Hello,

I was trying to run LIGGGHTS examples using liggghts executable. But it shows the following error -
ERROR: Illegal pair_style command

I am using newest version of LIGGGHTS. But when i write 'liggghts' on terminal shows-
LIGGGHTS 1.5 based on lammps-10Mar10

What might be the reason for that error?

Please help me out.

Thanks
Manjura

UPDATE 7/14 - I've attached a file containing the CAD files for the tutorial cases.

*******************************************************************

I've recently worked up something of a tutorial for running some basic LIGGGHTS simulations that I hope others may find of some use. The attached document contains some sample input decks with some comments as to why certain commands are included. There is also a section at the end of 'best practices' compiled from some user experiences.

Comments, additions, or suggestions for improvement are welcome!

Is there a way to combine a rotary and linear motion so that the axis of rotation moves with the mesh?

When I apply more than one fix move/mesh/gran I do not get the motion I am expecting. I think I need a way to update the rotation point so that after the linear motion the rotation axis origin is updated. Can this be done with a variable that is updated in a for loop with inputs from the linear motion parameters?

Any ideas about how I can set something like this up?

Hi,

new versions are avaibale and some things changed.

Hello, I was wondering if there's any way to group atom types together.

I'm clumping multiple spheres together to act as one particle. Each aggregate particle requires a "fix rigid" on a group, which is works fine. The pain is that the "fix property/global" requires vectors (peratomtype) and matrices (peratomtypepair) of parameters, even if they are all the same material (have the same parameters). For a few particles, it's no big deal. More than a few, and it's very tedious.

In my mind, I'd like the following to work (in pseudo-code):