atoms are far apart

LAMMPS LAMMPS Mailing List Mirror
1

Hi Dear

I simulate polymer chains in the lammps. I think that the initial topology which has been got from Material Studio Soft. is correct. But at the during of simulation, the some atoms are too far apart. The my simple script is following:

units real
boundary p p p
atom_style molecular
bond_style harmonic
angle_style cosine/squared
dihedral_style charmm
read_data vinyl.txt # the polymer chains
pair_style lj/cut 10.5
pair_coeff 1 1 0.1094 3.816 10.5
pair_coeff 2 2 0.0860 3.816 10.5
pair_coeff 3 3 0.2104 3.442 10.5
pair_coeff 4 4 0.2100 3.3224 10.5
pair_coeff 5 5 0.0157 2.974 10.5
pair_coeff 6 6 0.0150 2.918 10.5
pair_coeff 1 2 0.09699 3.816 10.5
pair_coeff 1 3 0.15171 3.629 10.5
pair_coeff 1 4 0.15157 3.5692 10.5
pair_coeff 1 5 0.04144 3.395 10.5
pair_coeff 1 6 0.04051 3.367 10.5
pair_coeff 2 3 0.1345 3.629 10.5
pair_coeff 2 4 0.1344 3.5692 10.5
pair_coeff 2 5 0.03675 3.395 10.5
pair_coeff 2 6 0.03591 3.367 10.5
pair_coeff 3 4 0.2102 3.3822 10.5
pair_coeff 3 5 0.0574 3.208 10.5
pair_coeff 3 6 0.05618 3.18 10.5
pair_coeff 4 5 0.0574 3.1482 10.5
pair_coeff 4 6 0.0561 3.1202 10.5
pair_coeff 5 6 0.01534 2.946 10.5
pair_coeff * 7 0 3.00 10.5

Equilibration ( at 300 K)

velocity all create 300.0 1231

fix equi1 all nvt temp 300 300 10
thermo_style custom step temp
thermo 1
timestep 10
run 10

2

Hi Dear
I simulate polymer chains in the lammps. I think that the initial topology
which has been got from Material Studio Soft. is correct. But at the during

please note that what you think and what LAMMPS does with your data
may be different. whether you have built a structure successfully with
some tool does not guarantee that you have it properly translated data
file.

of simulation, the some atoms are too far apart. The my simple script is
following:

what does "some atoms are too far apart" mean? are they like this from
the first step, or just "fly apart" rapidly?
in any case, there are plenty of possible explanations (bad
parameters, bad force field, bad input structure, incorrect topology),
and 10 MD steps is not exactly "a simulation". a 10fs timestep seems
very large unless you are running a coarse grain model.

axel.