Adding Angles and Missing Atoms of Graphene

I am currently working on my Master’s Project and have run into a problem with my script. I am using the custom command to create a sheet of graphene and moving one side towards the other. If I only create bonds, the simulation proceeds, but the atoms crumple together. As I add information for the angles and dihedrals, the first iteration generates the desired sheet of graphene, but all of the subsequent iterations “loose” the atoms. I have yet to find a solution by means of the Forum.

log log_test_2.txt

----- Simulation Setup -----

units metal

atom_style molecular

dimension 3

boundary f p p

----- Create Atoms -----

lattice custom 2.461 &

a1 1.0 0.0 0.0 &

a2 0.0 1.732050807569 0.0 &

a3 0.0 0.0 3.34 &

basis 0.0 0.0 0.0 &

basis 0.5 0.166666666667 0.0 &

basis 0.5 0.5 0.0 &

basis 0.0 0.66666666667 0.0

region whole block &

0.0 30.0 &

0.0 10.0 &

-0.5 0.5

— Configuration_1 —

create_box 3 whole &

bond/types 1 &

extra/bond/per/atom 3 &

angle/types 1 &

extra/angle/per/atom 3 &

dihedral/types 1 &

extra/dihedral/per/atom 4

create_atoms 1 region whole

region 1 block INF 1.0 &

INF INF &

INF INF

group left region 1

region 2 block 29.0 INF &

INF INF &

INF INF

group right region 2

group sheet subtract all left right

group contact union left sheet right

set group left type 2

set group right type 3

mass * 12.011

----- Interatomic Potential -----

bond_style harmonic

bond_coeff 1 25.0 1.421

angle_style harmonic

angle_coeff 1 2.0 120.0

dihedral_style harmonic

dihedral_coeff 1 5.0 1 2

pair_style tersoff

pair_coeff * * SiC_1994.tersof C C C

create_bonds all all 1 1.411 1.431

special_bonds angle yes &

dihedral yes

----- Processor Behavior -----

neighbor 2.0 bin

neigh_modify delay 0 &

every 10 &

check yes

----- Simulation Conditions -----

fix 1 all nvt temp 4.0 4.0 0.05

fix 2 left setforce 0.025 0.0 0.0

fix 3 right momentum 10 linear 1 0 0

fix 4 all bond/create 10 1 2 1.431 1

----- Create Output Files -----

dump 1 all image 10 image_*.jpg &

type &

type &

zoom 1.8 &

size 1000 1000 &

view 65 -85

dump_modify 1 pad 4

----- Run the Simulation -----

timestep 1

thermo 10

thermo_style custom &

etotal

thermo_modify lost ignore &

lost/bond ignore

run 100

Stephen Beattie

Graduate Student The University of Akron

Just b/c you added an angle and dihedral style, doesn’t mean

you have added any actual angles and dihedrals to your system. You

have to list those in a data file in sectinos like Angles, Dihedrals

that list which sets of 3 and 4 atoms form an angle/dihedral.

Steve

I am currently working on my Master’s Project and have run into a problem
with my script. I am using the custom command to create a sheet of graphene
and moving one side towards the other. If I only create bonds, the
simulation proceeds, but the atoms crumple together. As I add information
for the angles and dihedrals, the first iteration generates the desired
sheet of graphene, but all of the subsequent iterations “loose” the atoms. I
have yet to find a solution by means of the Forum.

i see at least two better options to do what you are doing.

1. created the entire topology outside of LAMMPS. e.g. use the
nanotube builder in VMD which can also create graphene sheets and you
can select whether you want to get bonds, angles and dihedrals or not.
whether you need impropers depends on the force field. typically not.
together with the topotools plugin, you can output the structure
directly to LAMMPS. or build multiple sheets (and move, rotate, cut
them as needed) and merge them and *then* output to LAMMPS data.

2. use a manybody potential like tersoff or airebo and you don't need
any bonds, angles or dihedrals at all, as they are implicit in the
force field. this is what the majority of LAMMPS users seem to be
doing.

option 1. results in a faster simulation, option 2 is easier to set up.

axel.