[lammps-users] Simulation of the silicon melt

Sorry if I seemed abrupt. Your question seemed to be: someone
did a model with a complex geometry in a paper, will someone read it,
understand it, and show me how to do it in LAMMPS.

Now you are asking a specific question, about the lattice command.
The lattice command is tricky for me as well, and I wrote it. The only
way to use it for complex geometries and be sure you are getting
what you want is to viz the results, starting with a small system.
You can always generate the atomic coords yourself and put them
in a data file to read in if you know exactly what you want.

This paragraph from the doc page is pretty precise - what part is
unclear? And I don't know
what you mean by "step direction".

The orient option specifies how the unit cell will be rotated when mapping it into the simulation box. The >dim argument is one of the 3 coordinate axes in the simulation box. The other 3 arguments are the >crystallographic direction in the lattice that you want to orient along that axis, specified as integers. E.g. >"orient x 2 1 0" means the x-axis in the simulation box will be the [210] lattice direction. The 3 lattice >directions you specify must be mutually orthogonal and obey the right-hand rule, i.e. (X cross Y) points in >the Z direction.