[lammps-users] Rigid spherical, fiber and plate packing


I need to look at the packing of spherical and plate-like rigid particles. I’m confused by this statement on the body particle how to page (https://docs.lammps.org/Howto_body.html)

You should not use body particles to model a rigid body made of simpler particles (e.g. point, sphere, ellipsoid, line segment, triangular particles), if the interaction between pairs of rigid bodies is just the summation of pairwise interactions between the simpler particles. LAMMPS already supports this kind of model via the fix rigid command.

What would be the LAMMPS best practice for modeling a mixture of spherical, short fiber and plate packing?

I suspect fix rigid would be the most flexible in defining the overall particle shape?

Thanks for the help,

Atom style body only makes sense, if you then also write a suitable pair style that can handle the interactions of the bodies. In your case you’d also need to devise your own body style as well.

Thus fix rigid or its variants is most likely the only way to move forward. but you can probably replace the different shapes with just a collection of point particles.

When you say a collection of point particles, do mean something like a collection of LJ particles with a bonding potential to keep them tied together? If so, it was my first consideration but figured the rigid body capabilities were probably the “right thing to do”.

For rigid bodies you don’t need bonds between the constituent particles.

You probably want to look up the work of the group of Sharon Glotzer to get some inspiration.