[ESPResSo-users] monte-carlo simulation of dipolar lattice

[Fan Zheng]

Dear all,

I am completely new to espresso. But I am wondering if anyone can help me decide if espresso can solve my problem before I dig into it.

My problem is to have monte-carlo simulation of a large dipolar lattice.

Each site is a dipole.
During the MC simulation, the position and magnitude of these dipoles will not change.
Only the direction of the dipole will rotate for each MC step.
But the rotation is limited to only four directions.

The total energy is sum of short range and long range.

Short range energy is the pair-wised interaction between nearest neighbors (six neighbors for each dipole), which I will define by myself.

Long range energy is the dipolar interaction, which needs ewald summation.

The positions of the dipoles are also on a regular grid in this big box.

The system can be very large (number of dipole could be around 1e6).
I want to use the parallel ewald summation (MPI or even GPU) of espresso to save time.

Could anyone illuminate if this problem could be possibly solve by espresso? Will it be very hard to modify the code?

Thanks all,

Fan