Hi. I’m attempting to model a cylindrical sample of amorphous silica. I’m using the wall/region command with LJ93 to create a cylindrical wall bounding the atoms. The animation on Ovito shows that the atoms remain in the cylindrical region. However, the bounding box is of course a prism. Would this correctly model a cylindrical sample? Thank you.
Hi. I’m attempting to model a cylindrical sample of amorphous silica. I’m using the wall/region command with LJ93 to create a cylindrical wall bounding the atoms. The animation on Ovito shows that the atoms remain in the cylindrical region.
However, the bounding box is of course a prism.
Bounding box of what? and how do you determine its shape and dimensions?
Would this correctly model a cylindrical sample? Thank you.
as noted in the documentation, the wall/region fix applies forces to atoms for as long as the region itself is inside the simulation cell (regardless of boundary conditions).
this is a property of the region feature in LAMMPS.
Axel.
As I already mentioned, only the parts of the region that are inside the simulation box will create confinement forces with fix wall/region, even with periodic boundary conditions. The LAMMPS manual is pretty clear about this.
A simulation cell of 80X80X160 Angstroms was created (maybe “bounding box” was not the right term). Inside that simulation cell, atoms filled a cylindrical region of diameter 80 Angstroms and length 160 Angstroms. A cylindrical wall with a diameter two Angstroms greater than the cylindrical region was created to confine the atoms in the cylindrical region. (The two angstrom gap was to ensure that the atoms would start in the cylindrical region).
Would you therefore recommend that I create a simulation box which is larger so as to confine the cylindrical wall entirely in the simulation box? The atoms do seem perfectly confined in the cylindrical region as is however.
Yes a larger simulation box that completely includes the wall region is a must for your case.
while it may look as if things work as expected, there is no guarantee that this is not by accident or for other reasons than the confinement potential.