Skip to content

This is the setup files for a Kob-Andersen liquid at specified reduced T and density.

Notifications You must be signed in to change notification settings

mathemaphysics/koband

Repository files navigation

Generate a Kob-Andersen System

Introduction

This is a set of scripts and compiled code for generating a input files for a Kob-Andersen system in three dimensions in a number of different molecular dynamics codes. The primary MD codes are GROMACS and Tinker.

The Kob-Andersen liquid is a simple atomic liquid which is commonly used in the field of condensed matter physics for studying glassy liquid behavior. It is a liquid which displays a structural glass transition around T = 0.45 (in reduced temperature units).

There are two atom types within the model, type A and type B. In the 80:20 KA liquid, there are 4 times as many A as B. Type A is larger in Vanderwaals radius than B. However, the A-B pair Vanderwaals energy of interaction, \epsilon_{AB}, is greater than that of either A-A or B-B. It is partially for this reason that the system is glassy.

Details

The following are the input parameters used for this simulation of the Kob-Andersen liquid.

N_A = 4000
N_B = 1000
T_r = 0.55

Variables N_A and N_B are the numbers of Type A and Type B atoms of the Kob-Andersen model, and T_r is the target temperature specified in reduced units (to be described below.

Reduced Units

In situations like this one, it turns out that not all parameters used to specify the state of the system are independent. For example, \sigma_A, \sigma_B, and \Delta x are related to one another as a result of the fact that each of these depends on available space.

Notes on Files

The file names are intended to serve as a guide for what they do, I hope. The MDP files are for different purposes each. The file step-min.mdp is generated for minimization. So use it to do a minimization after the initial packmol system is generated. It is important to note that you still need to do this minimization with the target software since packmol doesn't use the same potential. You can end up blowing up your simulation if you dive right in with the file straight from packmol, i.e. confin-box.gro. The MDP file step-pre.mdp will set up a run for a small fraction of the total time of the full simulation in the constant energy ensemble in order to smooth things out before runnin at the target temperature. Finally, there's the full run MDP file step-run.mdp, which will set generate random velocities initially, as needed by the isoconfigurational (IC) ensemble, and will proceed to run for the allotted time using the velocity rescaling method for running canonical ensemble simulation.

At this point there are also two different files to feed to packmol. The originally named file, packmol.inp will center the system's origin at the positional coordinates (0, 0, 0) nm. In addition, there's a file named packmol-pos.inp which places the bottom lower lefthand corner of the system cube at (0, 0, 0) nm instead. You can choose which of these is used by editing the Makefile to set PKG_INPFILE. The default at this point is packmol-pos.inp, since it is what is expected in general by GROMACS. However the downside of this is that TINKER is generally unhappy with this setup. So I recommend using the original packmol.inp file for simulations for which you want to use TINKER.

About

This is the setup files for a Kob-Andersen liquid at specified reduced T and density.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published