Reflective wall on long molecules!

Hi all,

I’m trying to simulate a box of hexane molecules where on one face I want to prevent the molecules from escaping, something like a reflective wall. However, the sudden reflective force will cause the molecule to rotate and enforce an unwanted rotation on the molecule! Is there any way to apply the reflective force on the center of mass of the molecule? or even use a half harmonic wall with LAMMPS? I know colvars library provides a half harmonic wall but is there such a capability available in LAMMPS also or should I use colvars for that?

I appreciate any suggestions to tackle this problem.

Best,
Kasra.

Hi all,

I'm trying to simulate a box of hexane molecules where on one face I want to
prevent the molecules from escaping, something like a reflective wall.
However, the sudden reflective force will cause the molecule to rotate and
enforce an unwanted rotation on the molecule! Is there any way to apply the
reflective force on the center of mass of the molecule? or even use a half
harmonic wall with LAMMPS? I know colvars library provides a half harmonic
wall but is there such a capability available in LAMMPS also or should I use
colvars for that?

i don't think that there is a ready-to-use feature for this, but i
could imagine that it would be rather straightforward to implement a
variety of fix wall/chunk classes based on the code in the fix wall
variants and fix spring/chunk (position restraint for chunks).

other than that, i'd try a very soft fix wall/harmonic, that should be
much less drastic than fix wall/reflect.

axel.

The fix wall/harmonic

in LAMMPS is already one-sided in that the particle

cannot be inside the wall and is truncated at the harmonic
minimum. So is that what you mean

by half-harmonic?

Steve

Thank you Axel and Steve.
Steve,
Yes, wall/harmonic is actually a half harmonic but it's zone of action is within the domain whereas I was thinking something like colvars wall which the potential is activated whenever the atom falls outside of the wall. But the wall in LAMMPS is defined such that nothing can pass it, in other words, I needed the half harmonic being applied after the center mass of the molecule passes the wall (again as colvars wall defining a dummy atom at the center of mass).
Anyway, I guess wall/harmonic is the closest to what I need or the best would be an implementation of fix wall/chunk as Axel mentioned.

Thank you,
Kasra.

What’s wrong with rotation? Why is it unwanted? That’s what molecules do: they rotate.

Aidan