Dear Lammps users,
I am encountering some strange behaviour regarding an NPT fix.
I am trying to simulate a box of ~54,000 Fe atoms at high pressure (7.7GPa) and temperature which are described using a MEAM potential (Phys. Rev. B 89, 094102).
I run the simulation for 200k timesteps with periodic boundaries in all directions.
If I use the fix:
fix 2 all npt temp 1860 1860 0.1 iso 77000 77000 1
the pressure stabilises within 2000 timesteps to the correct value.
However, If I try applying the same pressure only in the x direction:
fix 2 all npt temp 1860 1860 0.1 x 77000 77000 1
the pressure consistently stabilizes at ~10GPa throughout the entirety of the run.
A search through the forums brings up a user experiencing exactly the same problem (http://lammps.sandia.gov/threads/msg66409.html) . The problem in their thread
it would seem was the core-shell package.
Could I have a similar problem with the MEAM potential? Or is there something I am misunderstanding.
Any advice would be greatly appreciated.
Cheers,
Brani
Dear Lammps users,
I am encountering some strange behaviour regarding an NPT fix.
I am trying to simulate a box of ~54,000 Fe atoms at high pressure
(7.7GPa) and temperature which are described using a MEAM potential (Phys.
Rev. B 89, 094102).
I run the simulation for 200k timesteps with periodic boundaries in all
directions.
If I use the fix:
fix 2 all npt temp 1860 1860 0.1 iso 77000 77000 1
the pressure stabilises within 2000 timesteps to the correct value.
However, If I try applying the same pressure only in the x direction:
fix 2 all npt temp 1860 1860 0.1 x 77000 77000 1
the pressure consistently stabilizes at ~10GPa throughout the entirety of
the run.
A search through the forums brings up a user experiencing exactly the same
problem (http://lammps.sandia.gov/threads/msg66409.html) . The problem in
their thread
it would seem was the core-shell package.
Could I have a similar problem with the MEAM potential? Or is there
something I am misunderstanding.
Any advice would be greatly appreciated.
are you looking at the total pressure, or at the pxx component?
axel.
I’m looking at pressure as output by :
thermo_style custom press
Brani
I’m looking at pressure as output by :
thermo_style custom press
that is the total averaged pressure over Pxx, Pyy, Pzz.
however with your fix npt command, you are only adjusting the x-dimension
of your box according to Pxx.
so you should output and monitor pxx to determine whether fix npt is
working as expected.
axel.
It seems my mistake was in thinking that feedback for the fix NPT was the averaged pressure, rather than the x component of the pressure tensor. This explains why at much higher temperatures (well past the melt temperature), P_ave is approx. equasl P_xx than at lower (semi-solid) temperatures. Thank you for your reply, I hope this helps anyone experiencing similar issues.
Cheers,
Brani