Dear all:
I am trying to build hetero structure using lammps, Here I am including the script for that,
units metal
dimension 3
boundary p p p
atom_style atomic
neigh_modify exclude type 1 3
neigh_modify exclude type 2 4
neigh_modify exclude type 3 4
neigh_modify exclude type 1 4
lattice wurtzite 3.189
region whole block 0 33 0 20 0 75 # units box
create_box 4 whole
lattice wurtzite 3.189
region middle block 1 32 1 19 1 48 #units box
create_atoms 4 region middle basis 1 1 basis 2 1 basis 3 2 basis 4 2
lattice wurtzite 3.112
region lower block 1 32 1 19 49 69 #units box
create_atoms 4 region lower basis 1 3 basis 2 3 basis 3 2 basis 4 2
lattice fcc 4.045
region upper block 1 32 1 19 87 95 #units box
create_atoms 4 region upper
set region upper type 4
mass 1 69.45 # Gallium
mass 2 14.01 # Nitrogen
mass 3 26.98 # Alluminum
mass 4 26.98
pair_style hybrid/overlay sw tersoff eam/alloy
pair_coeff * * sw GaN.sw Ga N Ga N
pair_coeff * * tersoff AlN.tersoff_vamsi Al N Al N
pair_coeff * * eam/alloy AtCu.eam.alloy At At At At
velocity all create 300 123456 dist gaussian
#fix 1 all nve
fix 1 all nvt temp 300.0 300.0 100.0
timestep 0.0001
thermo 100
dump 1 all xyz 200 dump.expInGaN_300K_xyz_m
run 100000
My doubt here is that,
Can I include two different lattices in the same simulation box or Is it a mistake to do that ?
Because I want to include wurtzite lattice and fcc lattice in the same simulation box. I tried to simulate the
structure but it is melting down at 300K itself. Could some body tell me what could be the reason.
And If there is any mistake in the input script please let me know.
Thanks in advance 
vamsi
Dear all:
I am trying to build hetero structure using lammps, Here I am including the script for that,
units metal
dimension 3
boundary p p p
atom_style atomic
neigh_modify exclude type 1 3
neigh_modify exclude type 2 4
neigh_modify exclude type 3 4
neigh_modify exclude type 1 4
lattice wurtzite 3.189
region whole block 0 33 0 20 0 75 # units box
create_box 4 whole
lattice wurtzite 3.189
region middle block 1 32 1 19 1 48 #units box
create_atoms 4 region middle basis 1 1 basis 2 1 basis 3 2 basis 4 2
lattice wurtzite 3.112
region lower block 1 32 1 19 49 69 #units box
create_atoms 4 region lower basis 1 3 basis 2 3 basis 3 2 basis 4 2
lattice fcc 4.045
region upper block 1 32 1 19 87 95 #units box
create_atoms 4 region upper
set region upper type 4
mass 1 69.45 # Gallium
mass 2 14.01 # Nitrogen
mass 3 26.98 # Alluminum
mass 4 26.98
pair_style hybrid/overlay sw tersoff eam/alloy
pair_coeff * * sw GaN.sw Ga N Ga N
pair_coeff * * tersoff AlN.tersoff_vamsi Al N Al N
pair_coeff * * eam/alloy AtCu.eam.alloy At At At At
velocity all create 300 123456 dist gaussian
#fix 1 all nve
fix 1 all nvt temp 300.0 300.0 100.0
timestep 0.0001
thermo 100
dump 1 all xyz 200 dump.expInGaN_300K_xyz_m
run 100000
My doubt here is that,
Can I include two different lattices in the same simulation box or Is it a mistake to do that ?
You can use the lattice command multiple times, but only one is active
at a time.
Because I want to include wurtzite lattice and fcc lattice in the same simulation box. I tried to simulate the
structure but it is melting down at 300K itself. Could some body tell me what could be the reason.
You assign 3 different potentials to each atom. That doesn't seem right.
Axel
Dear Axel/All users:
So, Is there no other way that I can implement the two different lattice in the same
simulation box using lammps. If I want to change the code where should I look at, please suggest me.
Thank you,
vamsi
Dear Axel/All users:
So, Is there no other way that I can implement
the two different lattice in the same
simulation box using lammps. If I want to change the code where should I
look at, please suggest me.
i don't understand the question. the lattice command is just
a way to place a grid across the simulation box in order to
easily define atom positions.
you define the first lattice, then generate the relevant atom
positions using this lattice, next you define the second lattice
and then generate the atom positions using this lattice.
this works fine. *however* in this scheme
only one lattice at a time is "active".
that all being said, i doubt that the lattice
command is your problem.
there are two possible issues with your input
that stand out. the first is that it is *your*
responsibility to pick lattices and thus the
simulation cell dimensions, that are suitably
commensurate. since you have periodic boundaries
this matters a lot. the second is your very
unusual potential definition. i had to bet
money in something causing the problems,
that would be my first candidate. this looks
very non-sensical to me.
axel.
Dear Axel/All Users:
So, About the potential file:
I want to use four atoms in my simulation i.e., Ga, N, Al, At
1)where one structure is with Ga and N with sw potential
2) another structure with Al and N with tersoff potential
3) last structure with At on the whole with eam/alloy potential
So Is there a way to include all these potentials in a single input script. Is pair hybrid/overlay
cannot work for these kind of simulations. Please let me know how can I place the potential files correct in the
input script.
Thanks,
vamsi
Dear Axel/All Users:
So, About the potential file:
I want to use four atoms in my simulation i\.e\.,
Ga, N, Al, At
1\)where one structure is with Ga and N with sw
potential
2) another structure with Al and N with tersoff
potential
3) last structure with At on the whole with
eam/alloy potential
So Is there a way to include all these potentials
in a single input script. Is pair hybrid/overlay
cannot work for these kind of simulations. Please let me know how can I
place the potential files correct in the
input script.
first of all, you should read the documentation more carefully.
second, you should create a *separate* input for each
of the structures and make it work *by itself*.
lumping all of these together in one file without
knowing whether even parts of it work is the
best recipe for instant doom.
third, multi potential type calculations have to be
set up very carefully, especially when they contain
manybody potentials. this usually works well, if
you have separate areas that are each described
by one potential type (i.e. you use pair style hybrid
and not hybrid/overlay) and where the interaction
between the different segments can be described
by a simple pairwise additive potential.
in your specific case, that means, that you will
have to have two types of N atoms. check out
some of the examples of using pair style hybrid
that were discussed on this mailing list in the
mailing list archives.
forth, you have to resolve how you'll make your
different components with different lattice constants
commensurate. a quick glance at your input shows,
that your input structure is a mess and most certainly
not what you expect.
in general, it is almost always a mistake
to use too many things that you don't fully
understand at the same time. build it in
steps and solve one problem at a time
and your chances to be successful will
increase massively. what you are trying
right now, is far to complicated and
effectively broken beyond repair via e-mail.
axel.