Dear Sir/Madam,
I've been trying to run a 2D simulation on a simple hexagonal system
(resembling a graphene sheet).
The potential that we are using is very simple and consists of only harmonic
bonds and angles; there are no non-bonding interactions or dihedrals etc...I've generated the structure on material studio v6.0 and have converted the
structure into a readable format via msi2lmp and have made some slight
modifications to the file to suit my needs.
I'm running a simple minimization simulation however the program is
returning a number of warnings which are:- WARNING: Inconsistent image flags (../domain.cpp:594)
- WARNING: Bond/angle/dihedral extent > half of periodic box length
(../domain.cpp:667)I've checked the system using VMD and it appears that the atoms at the
boundary are not bonding with their image atoms but with the atoms in the
same cell (as can be seen in this image:
http://tinypic.com/view.php?pic=2j29euc&s=5#.UrgQ0vRDtl0).
i don't see any bonds at all in this image.
please note that a) vmd's periodic display is just taking the regular
rendered cell and replicating it, so it cannot directly render bonds
between periodic images and b) that LAMMPS will set up bonds to the
"closest image", thus the warning about a bond longer than half the
periodic box is harmless in your case. the warning is printed because
LAMMPS cannot distinguish between your case and the case where a
molecule is "broken" due to inconsistent image flags (which can spell
problems down the road, e.g. when trying to use the replicate command.
I've defined all of the bonds in the structure file (even bonds crossing the
image plane) however the problem still persist. From the diagnostic tests
that I've carried out it appears that the bonds crossing the periodic
boundaries are being bonded with atoms within the same cell. I've tried
searching for a possible solution and have tried tampering about with the
script as well as structure file however I have not managed to address this
problem.For reference, here's a copy of the script that I'm using:
http://pastebin.com/raw.php?i=9LjwJmSR
this is useless without the data file.
axel.