Monte Carlo simulations of pseudo hard spheres

This code can run Monte Carlo simulations using a pseudo hard sphere potential.

It is written from the ground up to be very fast and efficient.

Build

To build and use this code, first create a build directory

mkdir build

then, inside this directory run the following

cmake -GNinja ..

This command expects that you have installed the ninja build system.

Then, just run

ninja

In case you only have make just run

cmake ..

and then run

make

Either way, an executable mcfort will be created inside the build directory. This is the main executable.

Usage

The mcfort executable expects a file called input.in with four values in it:

• Packing fraction, a value between strictly larger than 0 and strictly less than 0.49. Any value in between is acceptable. This value will determine the density of the system.
• Number of particles, an integer. The larger this value, the longer it will take to run a simulation.
• Number of q-vectors, the vectors to be used in order to estimate the structure factor. The larger this value, the longer it will take to run a simulation.
• The number of bins to use for the observables. The larger the value, the more precise the observables will be. This means that more data will be computed.
• Number of Monte Carlo cycles, this is the value to use in order to compute the observables.

Observables

By default, this code will compute two main observables:

• The radial distribution function, also known as g(r);
• the structure factor, from the Statistical Mechanics definition, also known as S(q).

These will appear as files named gr.dat and sq.dat.

Caveats

There is no acceleration on this code. This is plain, old O(n^2) complexity. Although, it is very robust, and works as expected. It might take a long time, but it will give you the correct, physically-relevant result.

The only potential supported is the pseudo hard sphere potential. Maybe in the future other potentials might be added.