Hi All,
I am very new to LAMMPS and I am having a tough time trying to figure out how to create a patterned substrate. The substrate is supposed to be a series of 1 micron indentations on a flat substrate that are spaced 1 micron apart from each other. Like this (but with the top dashes actually connected to the bottom lines):
| || || || || || |
After searching through the manual and looking at examples, I deduced I had to use lattice and region and I attempted modifying the deposit example and it didn’t create what I wanted. It created a smaller box than the original deposit example. Here are the first few lines of the input script:
units lj
atom_style atomic
boundary p p f
lattice fcc 1.0
region box block 0 5 0 5 0 10
create_box 2 box
region substrate1 block 0 1 INF INF INF 1
create_atoms 1 region substrate1
region substrate2 block 2 3 INF INF INF 1
create_atoms 1 region substrate2
region substrate3 block 4 5 INF INF INF 1
create_atoms 1 region substrate3
It would be wonderful if someone could explain to me what I am doing incorrectly and if I am interpreting the commands of these functions incorrectly also. I just looked at it and I think it could be incorrect due to the units being lj when I actually need them to be nano (for my application of the simulation) but I am assuming there is more that is wrong.
Thanks
Rachel
Hi All,
I am very new to LAMMPS and I am having a tough time trying to figure out
how to create a patterned substrate. The substrate is supposed to be a
series of 1 micron indentations on a flat substrate that are spaced 1
micron apart from each other. Like this (but with the top dashes actually
connected to the bottom lines):
before looking at the geometry, you first need to decide which material
you want to model and what kind of interaction model you want to use for
that. if you want to do this on the atomic scale with each particle
representing an atom, then you may first need to search for a very large
computer and a lot of time on it, as you will need a *crazy lot* of atoms
to represent a micrometer scale system. i suggest you do a little back of
the envelope calculation as to how many atoms you would be needed to
represent your system of interest and then look at the benchmarks page on
the LAMMPS home page to make an estimate (well, you probably need to
extrapolate) as to how much time this would take and how many CPUs you
would need.
axel.
I should have prefaced this in the last email with what the simulation should show. I want to show the deposition of nanoparticles suspended in a solution (like the colloid example) onto the substrate and then heat up the substrate at a certain temperature for a set amount of time to note the change in assembly of the particles. I forgot to change the units from atomic to nano (which I assume I’m supposed to since it is nanoparticles). The substrate is gold throughout and the nanoparticles are PbS. I was going to focus on the creation of the nanoparticles later but for now I just wanted to create the geometry. The gold substrate has little effect on the interaction with the nanoparticles. I am looking more into modeling the effect of temperature. Should I still look into the CPUs and time?
Thanks,
Rachel
I should have prefaced this in the last email with what the simulation
should show. I want to show the deposition of nanoparticles suspended in a
solution (like the colloid example) onto the substrate and then heat up the
substrate at a certain temperature for a set amount of time to note the
change in assembly of the particles. I forgot to change the units from
atomic to nano (which I assume I'm supposed to since it is nanoparticles).
The substrate is gold throughout and the nanoparticles are PbS. I was going
to focus on the creation of the nanoparticles later but for now I just
wanted to create the geometry. The gold substrate has little effect on the
interaction with the nanoparticles. I am looking more into modeling the
effect of temperature. Should I still look into the CPUs and time?
first you need to find a suitable model (potential type and parameters)
that would describe (all!) the interactions and materials you want to
represent.
then you can start estimating the cost of the calculation.
to make this workable, you may need to use some coarse grained
approximation.
but that said, if you are new to MD simulations, you should learn some MD
basics with simpler systems first and you need to find yourself a tutor in
running and analyzing MD simulations properly, because doing correct MD
simulations is not something that you can learn from a software manual and
some e-mails. there is a lot more to it than writing a syntactically
correct input file and making atoms wiggle around.
and in most cases, building the simulation geometry is the last thing to
worry about.
axel.
Rachel,
1 micron is a VERY large size for MD.
That aside …
You have regions which are 1 \AA wide, with a lattice parameter of 1 \AA. This is unlikely to produce what you want.
HTH
Nigel