[gmx-developers] external potential ?

[Pim Schravendijk]

Now the easiest way to implement a wall would be by conserving perdiodic 
boundaries, adding 2 layers of frozen atoms and leaving an empty space the 
size of the cutoff:

http://www.gromacs.org/pipermail/gmx-users/2003-January/030522.html

Point is, I really want to have implicit walls, say at z=0 and z=L. One
way to do this would be to remove periodic boundary conditions in the z
direction by removing all interactions between atoms and their minimum
images at the other side of z=0 or L.

I figure the most logical place to do this is at the point where the
neigbour list is made. Now do_force calls a function ns in force.c, which
on its turn calls search_neigbours in ns.c. There I'd expect an algorithm
testing if any charge group or atom falls within a cutoff value and then
add that to the list. However the code there is a bit too cryptic for me,
I can't find anything which resembles something like that. I would be glad 
to get any hints on this..

Greetings, Pim

--
Pim Schravendijk - PhD Student
Max Planck Institute for Polymer Research
http://www.mpip-mainz.mpg.de/~schraven/