(fix rigid has changed since a year ago.)
I do learn something when I "answer" people's questions.
Thank you for the reference!
Andrew
(fix rigid has changed since a year ago.)
I do learn something when I "answer" people's questions.
you just discovered the greatest value of actively
participating at a public mailing list. to add something
constructive to the discussion, here is a description
of a trick, that is used to simulate linear rigid molecules
in other codes with SHAKE.
http://lists.gromacs.org/pipermail/gmx-users/2009-July/043363.html
you basically avoid the issue of the discontinuity
for solving the constraints for the angle at 180 deg,
by turning the CO2 molecule into a diatomic species
with only one bond constraint. in a way, this would
be similar to the TIP4P water in LAMMPS. i.e.
you have two masses between which you have the
constraint and that you time integrate, but then you
have other sites that have interactions and you
project the forces from them on the two masses,
which have no interactions by themselves.
it is not obvious to me at this point, how we can
easily do this in LAMMPS (perhaps we need a
generic dummy atom facility like gromacs),
but it would be interesting
also for other systems.
cheers,
axel.