[gmx-users] why the pulling force between Graphite and CNT is repulsive force?

[Cun Zhang]

hi,all
I'm doing a  simulation about the vdw interaction of Graphite and
Carbon Nanotube.

In my model is a trilayer graphite and a CNT which is perpendicular to
the graphite plane,like
          | |
          | |
          | |                 --
                              |      the distance between CNT and Graphite
-------------------------    --
-------------------------
-------------------------
.
In the initial configuration,the distance between them is about 1.8nm.
Then I pull CNT closer to the graphite,
and obtain the file pullf.xvg.

In this plot, I found the pulling force is a constant (a horizontal
line), but not zero. and the peak is less than zero.
Say, the interaction force between CNT and graphite from 1.4nm to
0.2nm is repulsive force.

The curves can be seen from the following links:
http://edwardpku.com/cun/wp-content/uploads/2010/07/time-force.jpg
http://edwardpku.com/cun/wp-content/uploads/2010/07/Distance-Force.jpg

Absolutely this result is wrong. Like the vdw force between two atoms,
there should be a equilibrium position
between the CNT and Graphite where the vdw force is zero.When the
distance is larger than it, the vdw force
between them is attractive force; when the distance is smaller than
it, it becomes repulsive force.

I put the main mdp parameters in the end.  Can you tell me  what's
wrong with my simulation?
Furthermore, if possible, I hope know how the pulling is operated,
and the pulling force is calculated and outputed in GMX?

Thank you!

Yours,
Zhang Cun


The main mdp parameters is as follow:
dt                           = 0.0005
energygrps              = CNT GRA
; CNT is the group of Carbon Nanotube,GRA is the group of graphite

vdw-type                 = Cut-off
rvdw                        = 1.0

Tcoupl                     = v-rescale
tc-grps                    = CNT GRA ;
tau_t                       = 0.1 0.1
ref_t                        = 300 300
constraints              = all-bonds

freezegrps               = GRA
freezedim                = Y Y Y

;pulling
pull = constraint
pull_geometry = distance
pull_init1 =
pull_start = yes
pull_dim = N N Y
pull_constr_tol = 1e-6
pull_nstxout = 100
pull_nstfout = 100
pull_ngroups = 1
pull_group0 = ref ;the toplevel graphite layer
pull_group1 = pull; the toplevel CNT layer
pull_vec1 = 0.0 0.0 1.0
pull_rate1 = -0.001


--
Blog: http://www.edwardpku.com/cun