Dear all,
Currently I am familiarizing myself with lammps.
What I am trying to do is to simulate Lennard-Jones chains where the lennard jones monomers are connected
by a spring potential. It resembles a united atom approach for alkanes.
Also I want to create binary mixtures of such chains where each chain consists of a different number of monomers.
Finally I want to see how the transport properties of these these „molecules“ behave (thermal conductivity, viscosity)
Currently I am stuck at how to tell lammps what molecule I want. Also I am not sure if lammps can provide what I would like to do with it.
I am looking forward to any pointers.
Best regards!
Philipp Crusius
Dear all,
Currently I am familiarizing myself with lammps.
What I am trying to do is to simulate Lennard-Jones chains where the lennard jones monomers are connected
by a spring potential. It resembles a united atom approach for alkanes.
Also I want to create binary mixtures of such chains where each chain consists of a different number of monomers.
Finally I want to see how the transport properties of these these „molecules“ behave (thermal conductivity, viscosity)
Currently I am stuck at how to tell lammps what molecule I want. Also I am not sure if lammps can provide what I would like to do with it.
LAMMPS is very well suited to model such bead-spring polymers. people
have been doing it for ages. 
with the current version of LAMMPS, there are two strategies to set up
such a system.
a) create an external tool, that will generate a suitable data file.
the chain.f fortran code in the tools directory is such a tool. and
the def.chain.ab example input is a possible starting point.
b) define molecule template files suitable for use with the molecule
command and then use the create_atoms to create an initial system.
in both cases, you will generate highly randomized structures with
very high potential energy due to overlaps. so you'll have to follow
some "unoverlapping" procedure. see the micelle example for a possible
approach to do this.
axel.