[lammps-users] Energy between two groups of atoms

Hello everyone,

I have three groups of atoms in my system.

group water type 1 2
group cation type 3
group anion type 4

I want to calculate energy between group1(water) and group2 (cation) (coulombic+ vdw). The manual says that " compute group/group " command can do this. But there is a restriction that this compute doesn’t include long-range interactions defined by kspace_style.

Is there any other way to do this?

Thanks in advance

srinivas

LAMMPS doesn't try to include long-range interactions
for group-group interactions, because I don't think it's
clearly defined what it even means. Does it mean
an infinite set of images of just those two groups,
excluding the A-A and B-B interactions in that
infinite array of images? How would you compute
such a thing via Ewald?

Steve

LAMMPS doesn't try to include long-range interactions
for group-group interactions, because I don't think it's
clearly defined what it even means. Does it mean

hi,

we just recently had a discussion on that subject in
our group. in my opinion group-group interactions can be
well defined, but only if you are using _explicit_ coulomb
throughout and no cutoff on it, and that in turn requires
that the groups are not too scattered, so that you can
compute the whole set of interactions within a distance
that is maximally the box length.

when using ewald or some other long range interaction method,
you would thus have to compute those interactions twice.
the long range part does not allow the separation of different
local contributions (which is exactly why it is so efficient
for the long range part).

for the given example of an electrolyte, i'd make some experiments
with systematically changing the concentration of electrolyte
and see if there is a convergence towards low concentration.
also you can still compute the _changes_ in the local contributions
to group/group interactions and rely on error cancellation
for the long range part.

cheers,
   axel.