Dear Gerolf,
thank you for your explanation. I will try to use GULP later for checking the EAM potential.
In the case of LAMMPS, I understood that this code is 'only' for dynamics simulation but not static simulation, in which we can check the reliability of the EAM potential. Does "write some scripts" mean that we must make script (input command file) to be able to use LAMMPS code? I am very grateful if you explain in a little more detail how to do this.
Thanks again,
Chol-Jun
Dear Chol-Jun
I think what Gerolf means by "write some scripts" is to write your on code
or program to get your desired property after the MD simulation has
finished. This is also called post processing. I don't know anything about
your system, but as an example, I can extract the radius of gyration of a
polymer by writing my own C code or "write some scripts" from an xyz
coordinate dump file from an MD simulation of a polymer.
Jan-Michael
Deal Jan-Michael,
I mentioned the check of reproducibility (or reliability) of the EAM potential, which is generally appropriate for metal or alloys. The EAM potential can be constructed from the ab initio data or experimental data such as lattice constant, elastic properties and some energetics of the single crystal.
Therefore it would be necessary to check the reproducibility of the EAM potential after construction. To do so, we have to calculate the energetics of the crystal by means of static simulation rather than dynamics simulation. My question is, can LAMMPS code provide such capability? If so, how can we do this (in a little more detail)? If we have to make C code or Fortran code to do this, it is not direct way using LAMMPS. Again, how can we use LAMMPS code to calculate the ZERO-temperature properties of the single crystal in the meaning of statics?
Best regards,
Chol-Jun
Jan-Michael Y. Carrillo wrote:
dear chol-yun,
Therefore it would be necessary to check the reproducibility of the EAM
potential after construction. To do so, we have to calculate the
energetics of the crystal by means of static simulation rather than
dynamics simulation. My question is, can LAMMPS code provide such
capability? If so, how can we do this (in a little more detail)? If we
the conventional way to check for lattice constant (and bulk
modulus) is to run a series of single point calculations with
the atoms on lattice positions (= scaled coordinates) and then
collect the total energy vs. the lattice constant and fit the
result against an equation of state. pick your favorite text book
on solid state electron structure calculations and check it out.
have to make C code or Fortran code to do this, it is not direct way
using LAMMPS. Again, how can we use LAMMPS code to calculate the
ZERO-temperature properties of the single crystal in the meaning of statics?
there is no simple way to integrate this into a package like LAMMPS
in an automated fashion, but then again, it is not very laborious
to do the calculation and several EOS programs are available for
download on the web.
cheers,
axel.
There is a minimize command in LAMMPS to relax
a system. You could also write a loop in your input
script (see jump, label, next commands) and re-run
LAMMPS many times with a variable for lattice constant,
thus evaluating the energy of a crystal at different
geometries.
Steve