LAMMPS_inputs

LAMMPS Documentation

A LAMMPS simulations is all build from a single file or script. This is known as the input script and often has the extension .in (although this is not required). The simplest way to run a LAMMPS input is to simply call it from the command line:

lmp -in script.in

Where lmp is the compiled LAMMPS executable.

The input script can have a multitude of command within it but some of the standard things you will need are:

• units
• atom_style
• read_data
• bond/angle/dihedral/improper_style
• timestep
• dump
• fix
• thermo
• run
• write_data

Let us first talk about read_data (I discuss the other commands in input.md); to start we need a data file. We can use LipParGen to create a molecule and OPLS force field for it. Let's use ethylene glycol as an example.

Navigate to the LigParGen server and input the SMILES string for ethylene glycol, OCCO, as the input structure. Choose the force field you want (you can choose from two OPLS forcefields), and hit submit molecule. After a short wait (with fingers crossed) you should get a series of options including LAMMPS. Download the LAMMPS file, this is your data file. I have put the output in this repo as OCCO.data. This data file holds the information about the simulation cell, the molecule and the forcefield that governs it, as well as some other stuff. I have written more about this in data_file.md.

Taking OCCO.in and OCCO.data we can run a simulation of a single molecule of ethylene glycol with the command (as above): lmp -in OCCO.in.

This should output several things:

• log.lammps
• traj.atom
• msd.msd
• 20000.data These are mostly self explainatory or I have explained them elsewhere in this repo.