LAMMPS team
I’m a newbie to the MD simulation and get lost with so many MD simulation software. Since the LAMMPS is free opensource and it’s my first choice, now I’m wondering if LAMMPS can simulate ferroelectric thin films, particles, such as Barium Titanate, with external constant electric or stress field? Or simulation of the ferroelectric domain walls? Looking forward and thanks for your kindly reply!
LAMMPS team
I'm a newbie to the MD simulation and get lost with so many MD
simulation software. Since the LAMMPS is free opensource and it's my
first choice, now I'm wondering if LAMMPS can simulate ferroelectric
thin films, particles, such as Barium Titanate, with external constant
electric or stress field? Or simulation of the ferroelectric domain
walls? Looking forward and thanks for your kindly reply!
to the best of my knowledge, these kind of systems usually need
to be treated with quantum mechanics, e.g. at the density functional
theory level. in order to model ferroelectric properties, you need to
include the electrons. LAMMPS does "only" classical interaction models.
cheers,
axel.
Axel
Thanks for your reply 
The problem of first principle lies in that the system could not be too large, and obeys the periodic boundary condition, which requires a very large supercell when dealing with domain walls. Furthermore, symmetries are imposed to the ferroelectric material, which might also fail in the region of domain walls. So, I’m wondering if there’s any remedy or I can do nothing.
Regards, Lin Xie
Axel
Thanks for your reply
that the system could not be too large, and obeys the periodic
boundary condition, which requires a very large supercell when dealing
with domain walls. Furthermore, symmetries are imposed to the
ferroelectric material, which might also fail in the region of domain
walls. So, I'm wondering if there's any remedy or I can do nothing.
get a bigger computer? 
axel.
It does really help for the great information. I think the avoid of periodic boundary condition could help me reduce the total number of atoms. However I’m still annoyed with the artificial symmetries and maybe I’ll have a try of Octopus at last.
Keep it simple and stupid The computing resources are rather limited and I have to figure out the best solution.