[gmx-users] use of Walls in Martini simulations

[Ramon Reigada]

Hi,

 I would like to post a question about simulating with "Walls" in gromacs
and using the Martini force field. 
The question is very simple. I am trying to simulate a simple box of water
molecules with pbc=xy and walls at z=0 and z=z_box. I use Berendsen
T-coupling and no pressure bath.
I am trying:

pbc=xy
nwall =2
wall_atomtype = C1 C1   (C1 are the alkane carbons defined in the martini
force field file)
wall_type = 9-3 
wall_density = 110 110
wall_r_linpot = 1

and all what I have is that the initial box of water molecules breaks in two
pieces that separate each other in the z-direction.
The same result (or a segmentation fault) is obtained by using 10-4 o 12-6
LJ potentials, large wall_r_linpot values, other values of wall_densities,
so I think that the problem is with the wall_atomtype. I tried the water
bead (W) instead, and segmentation fault is obtained. 


  thank you very much in advance,

  Ramon