[gmx-users] g_dipoles -slab option
wsh.gromacs at gmail.com
Wed Jul 13 22:40:03 CEST 2011
I am testing g_dipoles tool in GROMACS with a simple system containing only
tip3p water molecules in a cubic box (6X6X6nm) with periodic boundary
conditions in xyz. After the simulation is finished, using g_dipoles, I get
the average epsilon=99.776 and the epsilon vs times also indicate that
epsilon stabilizes around 100, which are expected as in FIG 1 in ref.: JCP
124, 024503 (2006).
Then, when I looked at the g_dipoles -slab option, I found that it gives
only the average dipole moment (over time) per slab and not the epsilon per
slab. So my *first question* is that is there any particular reason for not
giving the epsilon per slab, such as due to the dipole correlation of water
molecules between slabs?
I have done a little bit more into this problem, I modified the
gmx_dipoles.c code in src/tools and output the following parameters:
1) the sum of dipole moment for all the molecules in xyz (in code M_av[XX],
M_av[YY], M_av[ZZ]) vs. time and correspondingly
2) the sum of molecular dipoles per slab (i) in xyz direction vs time, i.e.
slap_dipole_av[i][XX], slap_dipole_av[i][YY], slap_dipole_av[i][ZZ]
If I use g_dipoles -slab -axis Z, I find that sum of molecular dipoles per
slab fluctuates about 1/sqrt(slab number) of the M_av both in XX and YY
correspondingly, which is presumably correct. However, the slab dipoles in
ZZ direction is much smaller (than that in XX or YY). My original thought
was that because the system is homogenous, the slab dipole fluctuation
should be the same/similar in all XYZ direction, as well as the following
values in gmx_dipoles.c i.e. M, M2. But apparently, I was wrong. So my *second
question* is that is it because of the dipole correlation? I am not very
clear. Because I think it is related to the -slab option currently
implemented in GROMACS, I would appreciate greatly if anyone could help me
on this problem. Thank you very much.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the gromacs.org_gmx-users