I am simulating a simple Lennard-Jones fluid confined between two fixed walls and I am analyzing the autocorrelation function of the velocity along the direction of the confinement (normal to the walls).
I observe an exponential like decay which is in line with the rough expectation, but in addition I observe periodic peaks. These peaks turn out to be the sound modes reflected back and forth between the two walls, as the time between the peaks increases when I increase the walls distance. And when I calculate the sound velocity for my system, it turns out to be equal to the walls distance divided by the time between the peaks.
Now I know about the sound mode, I am not satisfied with just having the velocity autocorrelation for looking at the sound modes. I would like to visualize the mode with more geometric features such as the velocity field, as I expect rich structures.
Of course the problem is that in such equilibrium simulations a snapshot of the particle velocities is just a random pattern, so some kind of averaging over the particle trajectories and velocities should involved to make the sound mode visible. Any suggestion ?
