Lost atoms

LAMMPS LAMMPS Mailing List Mirror
1

I first tried equilibrating alumina atoms using NPT at 300 K and 1 bar. Then after that I an heating alumina from 300 K to 3000 K using NPT in restart input script. But around 2200 K I loose atoms . Melting point of alumna is around 2200 K

A part of my input script is here.

read_restart aluminaequil.in

#Define Potential
pair_style vashishta
pair_coeff * * AlO.vashishta Al O

timestep 0.005 #picoseconds
thermo 1
thermo_style custom step temp press vol etotal ke pe density

fix 3 all npt temp 300 3000 1.0 iso 1.0 1.0 1.0
run 200000

2

I first tried equilibrating alumina atoms using NPT at 300 K and 1 bar. Then
after that I an heating alumina from 300 K to 3000 K using NPT in restart
input script. But around 2200 K I loose atoms . Melting point of alumna is
around 2200 K
A part of my input script is here.

there is not enough information here.
the manual describes causes that can cause lost atoms. see if any
apply to your simulation and resolve them.

axel.

3
4

Yes sir I tried using 0.5 femtosecond and my problem is resolved.

However I am not able to capture the melting point. I am trying to look potential energy/atom and mean square displacement variation with respect to temperature . i expect sudden change in slope of these parameters near the melting point.

Should i further reduce timestep?

Thanks

Nikhil

5

Please open a new thread with an appropriate subject as the original problem has been resolved, thanks.

Melting point is a delicate property to capture using MD, and the approach you have taken is a rather crude one. It depends more on the heating rate than the timestep size.

The most reliable way of pin-pointing the melting point is to construct a bi-layer structure with co-existing liquid and solid phases and run it at different (constant) temperatures. Please consult the literature or your more experienced colleagues.

Ray