LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Simulation of a LJ liquid with solid walls on top and bottom
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Simulation of a LJ liquid with solid walls on top and bottom

From: Axel Kohlmeyer <akohlmey@...24...>
Date: Thu, 3 Aug 2017 14:46:56 -0400

On Thu, Aug 3, 2017 at 2:20 PM, m n via lammps-users <> wrote:
Dear all,

I want to simulate a 2D liquid (square domain) with two copper solid walls on the top and bottom of the domain and apply two different high and low constant temperatures on the copper walls in order to calculate the temperature distribution as a result of the temperature gradient. I would really appreciate if somebody could answer my following questions: 

1. How can I create my geometry? Should I define two different lattices for liquid and copper walls or I need to use other softwares such as VMD or Packmol to create the data file for geometry?

​the lattice command has no impact whatsoever on ​the simulation. you can change it as many times as you want. the only purpose for the lattice command is to have a simple template (or pattern) that can be used to conveniently specify region or box dimensions that are commensurate with that lattice, which applies to all commands that support the "units lattice" or "units box" keywords (e.g. region).

when you specify the lattice command multiple times, the current command wipes out the previous lattice settings, but does not modify any position/length information specified before. when building a periodic system with multiple lattice commands, you have to make certain, that the lattices are reasonably well compatible, since the simulation box can only be defined once and thus with an incompatible later lattice command, you may fill a region with atoms that overlap at the boundary.

for simplicity, it is recommended to define all regions with the same lattice active when using "units lattice", i.e. the default, or always use "units box" to have a consistent reference for lengths.

for your specific system, you have to first figure out how to build a single thermostatted wall layer, as that will require multiple regions (an immobile region, a strongly thermostatted region and a non-thermostatted region), then add a second wall layer on the opposite side of your non-periodic box dimension, and check whether the layers become properly equilibrated and thermalized. only then would i fill the remaining void with your liquid atoms. you will have to experiment at this stage, what would be a suitable lattice constant for it to have a proper fully equilibrated system. 

of course, VMD scripting or packmol or moltemplate or writing some custom script/program that all would lead to a data file are all suitable alternatives. what works best depends a lot on your personal preferences and skills. i myself would probably try using create_atoms in LAMMPS first.

2. Is it appropriate to use LJ potential for LJ liquid particles, "eam" potential for copper particles and again LJ potential for the interaction between LJ liquid and copper particles? 

​this is frequently done. whether that is appropriate depends on what kind of system it is that you want to model and what level of accuracy you expect. since the metal wall part doesn't seem to be very adding much to the system outside of a "somewhat wiggly​, slightly corrugated surface", it may even be sufficient to use LJ interactions throughout (provided the equilibrated and thermalized solid LJ lattice geometry would be close enough than what you get with EAM).

overall, my recommendation is to first figure out step-by-step how you can build the system in general without paying too much attention to the kind of potential and then worry about improving your model later. switching potential parameters or pair styles is very easy to do in LAMMPS and independent from how to construct or input a system.


Many thanks in advance,


Check out the vibrant tech community on one of the world's most
engaging tech sites,!
lammps-users mailing list

Dr. Axel Kohlmeyer  akohlmey@...12...24...
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.