[lammps-users] sound in granular

I want to generate sound wave in static granular bed. Can anybody know how to generate sound wave and receive the same?
Thanks in advance,

Wiggle some atoms at a boundary, and a sound wave results?


Thanks a lot. Should I wiggle wall or atoms?


If you just want to produce vibrations at a boundary, fix wall/gran will
work fine. If you want to vibrate a more complex shape, it's best to
create the shape out of atoms and use fix move. That is what I do for
acoustic-to-seismic coupling simulations of irregularly-shaped objects
in sand beds.

Check out the doc pages:

Hi Sam Bateman,
Many thanks for helping me. One more question: how can I create the shape out of atom? For example, I have box filled up of particles; should I group the region(particles in that region) near wall and rigid it ( or freeze?)? Is this right way of doing that?
Thanks in advance,

That's one way to do it. You can input your particles with
a certain type and set the group via type as well.



I wrote some custom scripts to create data files to be used with the
read_data command. I usually have the fixed shape atoms of type 2, and
the rest of the atoms (sand grains) are type 1. Then, create groups:

group shape type 2
group mobile subtract all shape
neigh_modify exclude group shape shape

You can also group by region if that's more convenient. Then use the
various fixes (freeze, move, gravity, etc) to change the motion of these
groups however you want.

It's important to remember that if you use a fix that does its own
motion on the shape (freeze, move), you should not integrate those atoms
with nve/sphere. So:

fix int mobile nve/sphere

Hope that helps.

Hi Sam,
Thanks a lot. Here:

group mobile subtract all shape
neigh_modify exclude group shape shape
is this:
neigh_modify exclude group mobile shape or neigh_modify exclude group shape shape?


"neigh_modify exclude group shape shape" will eliminate unnecessary
pairwise computations between the shape atoms. It's not necessary, but
will probably speed up the simulation. Sorry I wasn't clear earlier, I
just copy-and-pasted those lines.

If you "neigh_modify exclude group mobile shape", there will be no pair
interactions between the two atoms types, which is probably not what you

See http://lammps.sandia.gov/doc/neigh_modify.html