[lammps-users] Questions about using LAMMPS to simulate nanoparticle aggregate

Hi, Steve:

I would like to try using LAMMPS to study nanoparticle aggregation in liquid. Here is the case: nanoparticles (80nm) are uniformly dispersed in a liquid fuel droplet (1mm). Fuel droplet will be burned so the content of particles will increase. We want to know the aggregate properties (aggregate size distribution, volume fraction and so on) over time till all the liquids are consumed. It would also be interesting if we can get the structure of the particles agglomerate at the end.

I will be very grateful if you can tell me if it is possible to achieve this using LAMMPS. And any advise on this simulation work will be appreciated.

Thanks quite a lot.


Yanan Gan
PhD Candidate, Research Assistant,
Mechanical Engineering,
Purdue University

Dear Yanan,

A particle with a 80nm diameter contains around 10^7 atoms (meaning
*extremely* long computation times), and I don't speak of a 1 mm
diameter drop (10^20 atoms). Typically classical Molecular Dynamics is
able to tackle up to 10^5 atoms in reasonable computation times, so
you won't be able to simulate *exactly* your experimental system.

However MD can be an relevant tool to get some insight into the
problem your interested in, but clearly you'll have to think of the
size of your simulated system. If you're not familiar with MD, be
aware that you will have first to learn the hard way (read basic books
like the ones of Allen and Tildesley, or Frenkel and Smit; learn on
much simpler systems), because only by understanding the theory behind
lammps you can hope to produce physical meaningful results. And this
can take a while (the most efficient way is probably to find a local
expert in MD and learn with him/her).


2010/10/6 Yanan Gan <[email protected]...>:

Laurent is correct, especially if you wanted to do this
as an all-atom simulation. LAMMPS also has many
coarse-graining potentials and methods. E.g. the pair_style
colloid treats each big particle as a single particle. There
are various coarse-grain solvent models in LAMMPS as
well, including the new fix srd command. So the general
answer to your question, is that you need to figure out how
to model your problem before asking if LAMMPS can do it.

E.g. find a paper where someone
has done this and what kind of model and potentials
they used, and start from there. Or formulate your own
model. Until you do that it can't be said whether LAMMPS
or any other tool can help you.

I would have to go thru the same steps if I wanted to
model this phenomenon.


You might start with this.


It was done with the LAMMPS features Steve mentioned. I think the nanoparticle size is large enough that you could get away with a coarse grained solvent. I guess this depends on the solvent you are trying to model, though. You should know the answer to that before you ask more questions on the list. Or at least have an idea of how you are going to find out. It will be hard for people to give you helpful answers otherwise.

First, thank you all for your reply. I really appreciate your help here.

I will go ahead and find any related paper doing the similar work. It seems like coarse-grain solvent models might be appreciate since I have read one paper using LAMMPS to simulate nanoparticle aggregation. I will let you guys know if I make any progress.

Thank Matt for your paper.

Thanks you all.