enforcing periodic boundary conditions

Hello all,

I have a datafile of a copper polycrystal for which I known there were periodic boundary conditions imposed in the z-direction during previous simulations. The problem is I only have atom coordinates and not the simulation box size. Is there a way in LAMMPS to automatically set the box boundary to exactly encompass the extent of atoms? If I do it manually, I seem to move the boundary too close or too far, which will create a layer of incorrectly positioned atoms on the periodic boundary.

Sincere thank you for your replies,

Kristian Kuppart

Hello all,

I have a datafile of a copper polycrystal for which I known there were
periodic boundary conditions imposed in the z-direction during previous
simulations. The problem is I only have atom coordinates and not the
simulation box size. Is there a way in LAMMPS to automatically set the box
boundary to exactly encompass the extent of atoms?

Why should Lammps know better than you what the exact size of the box is?
Perhaps once AIs dominate the planet.

If I do it manually, I seem to move the boundary too close or too far,

which will create a layer of incorrectly positioned atoms on the periodic
boundary.

You could start with a guess that closely bounds all atoms and then run
some box/positions energy minimization (or damped dynamics) on the system
to get an equilibrated structure. Perhaps some atomic reshuffling will take
place especially at the GBs but I would not see that as a problem (have a
polycrystal remember?). Make sure to always carefully monitor the system
evolution to avoid the creation of undesired interfaces, etc.
Carlos

Hello all,

I have a datafile of a copper polycrystal for which I known there were
periodic boundary conditions imposed in the z-direction during previous
simulations. The problem is I only have atom coordinates and not the
simulation box size. Is there a way in LAMMPS to automatically set the box
boundary to exactly encompass the extent of atoms? If I do it manually, I

no. neither can any other software recover lost information.

why do you need to do this inside LAMMPS? by “manually” you mean you don’t know how to find the right box size?