[lammps-users] Periodic Boundary Conditions

LAMMPS LAMMPS Mailing List Mirror
1

Dear Steve,

I have been simulating Couette flow with hexadecane molecules confined between Alumina layers. The initial configuration was realized in Material Studio. Later it was imported into Lammps for dynamics. I had made the hexadecane molecules at ambient temperature and pressure. In lammps, the system was thermostated to ambient temperature. However, the pressure calculted inside the simulation cell (using the stress/atom command), is unphysically very high (of the order of 0.2-0.3 GPa).
Recently I saw a paper by Louwerse and Baerends claming that the periodic boundary conditions induce error in the pressure calculation. Following this I saw your (group) paper, claiming to have rectified this problem. I have two questions

  1. Is this the reason for getting high pressures in my simulation? If so, is there a way to rectify it? Or some patches?
  2. In case you have implemented your ideas into the latest version for lammps, then the reason for me getting high pressures is something else. Do you have any reason about this abnormal high pressures?

Kindly advice on this. Any relavent or related information will be highly appreciated.

Thanks

With Regards
Ajay

2

I don't know, but you can read the paper which came out last year. There
is nothing intrinsically problematic about pressure with PBC. All of
the ideas in the paper are implemented in LAMMPS.

If you have the same system setup in Mat Studio, then I would compare
the pressure between the 2 codes. If they are different for the same
snapshot, then possibly you have translated their FF into LAMMPS incorrectly.
Since this translation can be tricky, this is a good comparison to make,
regardless.

Steve

3

Dear Steve,

Thanks for your reply. I am running the comparisons to Materials Studio right now. However, I have two related questions:

  1. Your paper that came out last year (JCP 131 (2009) 154107) discusses three forms of the stress calculation (group, atom, atom-cell). Which of these is implemented in LAMMPS?

  2. Also, is this incorporated into both the pressure and stress/atom fixes?

Thank you ahead of time for your help.

With Regards
Ajay

4

The paper essentially showed that they are all equivalent. I forget
the terminology, but they are implemented in both global and per-atom
pressure (stress). PBCs make no difference; i.e. they are handled correctly.

Steve