Re: [ESPResSo] LJ Clusters

[Olaf Lenz]

Hi!

On Thu, September 6, 2007 7:15 pm, address@hidden said:
> is there a way to measure the cluster size distribution in Espresso?
> (2 Particles belong to the same cluster if they have less than a certain
> distance.)

As far as I know, there is no such function in ES yet.
I think, that a function to measure the cluster size distribution would be
useful. I've implemented it on the Tcl-level so far, using "measure
nbhood".

However, there are a few issues to it. First, I'm not sure whether a pure
distance criterium is useful, because it would classify particles to
belong to the same cluster, even if two particles just pass by each other
with high velocities. Therefore, it might be useful to employ an energy
criterium. I do not currently have the reference at hand, but the general
idea is to compare the relative kinetic energies to the negative relative
potential energies: when the negated potential energy is higher than the
kinetic energy, the particle belongs to the cluster.

Also, think of a way to specify the particle types/particle groups of
which clusters are to be counted.

Best regards
Olaf