To genarate molecular crystal

I think packmol should be what you want:
http://www.ime.unicamp.br/~martinez/packmol/

Lingti

That's a good suggestion. The only LAMMPS
command that could help is "replicate", which would
require you to have one copy of the molecule in
a small box, then replicate it into a crystal. But that
assumes you won't generate overlaps by doing that.

Steve