[gmx-users] Problem with Lennard Jones cross terms

[xrysostomos sfhsdfghsdfhsdgh]

I am trying
to simulate the following system: 5-layer FCC substrate consisting of Si particles
which interact only with Lennard-Jones potential and vertical to the substrate
I put polymer chains of (CH2-OA), starting from a distance of 0,4nm. 

What I want
is to be able to manage completely independently the LJ interactions between
the substrate and the polymers and not according to the combination rules, i.e.
to have very very strong LJ interaction only between the Si particles but these
to interact normally with the polymer chains.

I use the
Gromos-96 force-field. The initial values for the c6 and c12 parameters are the
following:

[ atomtypes ]

name  at.num  
mass      charge  ptype      
     c6                      c12

OA          8          0.000     
0.000     A         0.0022619536 
  1.505529e-06

CH2        6          0.000     
0.000     A         0.0074684164 
  3.3965584e-05

SI            14        0.000      0.000    
A         0.01473796         2.2193521e-05

 

[ nonbond_params ]

i           j      func         
c6                       c12

SI      
OA      1      0.0000000      0.00000000

SI     
CH2     1      0.0000000      0.00000000

SI     
CH3     1      0.0000000
     0.00000000

 

What I did is to run 2 different simulations in NVT ensemble. In the first, I multiplied by 10 the c6, c12
values of Si in the [atomtypes ]  and in
the second I multiplied (the initial
values again) by 100. In both cases I didn’t change the values on the [nonbond_params
].

I would expect that in both cases there should be no adsorption of the
polymers in the substrate, since I  change the interaction between the substrate and
the polymers in the [nonbond_params ] equal to zero.

What I observed is that after 50000 steps with 2 fs time step the
polymers in the second case adsorbed to the surface.

Can anybody help me or explain me why is this happening?

 

Thanks in advance

 

Chrysostomos



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100701/57887171/attachment.html>