Dear All,
I am trying to create a cylindrical Cu-Ag core-shell nanoparticle using LAMMPS. I am able to make the core but facing difficulty in making the shell. Everytime I am trying to make the shell it is becoming a solid structure with the shell atoms starting from the core.
Kindly help with the fundamental or algorithm for creating the shell around the core.
Dear All,
I am trying to create a cylindrical Cu-Ag core-shell nanoparticle using
LAMMPS. I am able to make the core but facing difficulty in making the
shell. Everytime I am trying to make the shell it is becoming a solid
structure with the shell atoms starting from the core.Kindly help with the fundamental or algorithm for creating the shell
around the core.
there are two fundamental approaches:
1) create the two kinds of particles separately. for that you need fill a
cylinder and a larger cylinder from which the former is removed
2) create a cylinder filled with atoms of the type for the outer shell and
then define a smaller cylinder inside the first one and change the type of
the atoms there.
then relaxation, equilibration as usual.
axel.
Dear All,
I am able to make the core-shell structure by using the second way suggested by Axel sir, but I am facing two different problems now-
The program is getting terminated with some error message of not being able to run lammps right in the first step of run. The values of potential energy and some other parameters are showing “nan” instead of numbers in that step. But my other lammps programs are running smoothly parallel to it in another command window.
When I am assigning color code using Ovito on the basis of mass, it is showing that the core is of silver and shell is of copper, while I have given the command for the reverse case (Cu core and Ag shell)
I shall be grateful if you kindly provide your valuable suggestions in this regard.
I am also attaching my input script and potential file for your reference.
CuAg.eam.alloy.txt (1.74 MB)
in.cs.txt (1.16 KB)
Dear All,
I am able to make the core-shell structure by using the second way
suggested by Axel sir, but I am facing two different problems now-1) The program is getting terminated with some error message of not being
able to run lammps right in the first step of run. The values of potential
energy and some other parameters are showing "nan" instead of numbers in
that step. But my other lammps programs are running smoothly parallel to it
in another command window.
you get this because you did *not* understand and follow my advice. you
are still creating atoms on top of each other.
let me repeat:
option 1) you have to create two regions without overlap, i.e. the larger
region has to have the smaller region subtracted.
or option 2) you have only one create_atoms command (for the larger region)
and then change the type of the atoms in the smaller region only.
2) When I am assigning color code using Ovito on the basis of mass, it is
showing that the core is of silver and shell is of copper, while I have
given the command for the reverse case (Cu core and Ag shell)
that would be an Ovito problem and not a LAMMPS problem, then.
axel.
Thank you very much Axel sir for helping me again to resolve my first problem.
But I am still facing the second problem which I mentioned earlier and it do not seems to be a Ovito problem. I am using “Color Coding” function of Ovito to assign color on the basis of mass of the atoms. According to that, it is clearly showing from the image (screenshot attached) that my Ag has formed the core and Cu has formed the shell (instead of Cu core and Ag shell). The same information I obtained by putting my cursor on atoms of respective region and it also shows that the mass of the particles of core is of Ag and mass of the particle of shell is of copper.
I am attaching a screenshot of the Ovito for reference.
Thank you very much Axel sir for helping me again to resolve my first
problem.But I am still facing the second problem which I mentioned earlier and it
do not seems to be a Ovito problem. I am using "Color Coding" function of
Ovito to assign color on the basis of mass of the atoms.
if you see it only in ovito, it is an ovito problem.
to make people have a look into whether there is a problem in LAMMPS you
have to provide more convincing proof.
- you have to demonstrate that the dump files already contain incorrect
mass information
- you have to demonstrate this unexpected behavior can be reproduced with
the very latest LAMMPS version (19May2017 at the time of this writing)
until then, we have to assume, that the problem is not in LAMMPS.
here is my proof. with this minimal input:
region box block 0 1 0 1 0 1
create_box 2 box
create_atoms 1 single 0.5 0.5 0.0
create_atoms 2 single 0.0 0.5 0.5
mass 1 4.0
mass 2 2.0
write_dump all cfg two.cfg mass type xs ys zs
i get exactly the masses, that i have assigned in the .cfg file:
Number of particles = 2
A = 1 Angstrom (basic length-scale)
H0(1,1) = 1 A
H0(1,2) = 0 A
H0(1,3) = 0 A
H0(2,1) = 0 A
H0(2,2) = 1 A
H0(2,3) = 0 A
H0(3,1) = 0 A
H0(3,2) = 0 A
H0(3,3) = 1 A
.NO_VELOCITY.
entry_count = 3
4.000000
C
0.5 0.5 0
2.000000
C
0 0.5 0.5
i see no indication that LAMMPS is mixing things up and i would be
extremely surprised of this, considering how many people use the cfg file
format and atomeye to visualize LAMMPS output. if LAMMPS would be creating
incorrect output, it would have long been noticed and reported, since the
code for dumping .cfg style files has been unchanged for quite a while.
According to that, it is clearly showing from the image (screenshot
attached) that my Ag has formed the core and Cu has formed the shell
(instead of Cu core and Ag shell). The same information I obtained by
putting my cursor on atoms of respective region and it also shows that the
mass of the particles of core is of Ag and mass of the particle of shell is
of copper.
please note the following comment from the dump documentation about the cfg
format output:
Note that you will typically want to use the *dump_modify element*
<http://lammps.sandia.gov/doc/dump_modify.html> command with CFG-formatted
files, to associate element names with atom types, so that AtomEye can
render atoms appropriately.
axel.
Hello,
Does it make sense to use tail correction of buck/coul/long potential with ewald/disp summation?
Best regards,
Xavier B.
Hello,
Does it make sense to use tail correction of buck/coul/long potential with
ewald/disp summation?
to some degree.
- Tail correction calculates the long-range correction of the buck part of
buck/coul/long.
- ewald/disp calculates the long-range contribution to coulombic
interactions, but also calculates the long-range contribution of the buck
part in an ewald manner. The latter concerning buck part appears to be
redundant with the tail correction. The difference seems that ewald/disp
doesn't presuppose that g(r) = 1 as it is for tail correction.
Anyone could confirm the redundance?
there is no redundance here. ewald/disp will only compute the long-range
component for the non-coulomb part of the pair style forces, if you use the
pair style buck/long/coul/long with the long setting for the buckingham
part. buck/coul/long will always use a cutoff on the buckingham part of the
interaction and thus tail correction makes sense.
axel.