Hello lammps users, I have a physical question, anyone know how you could
calculate the radial distribution function for droplets in the vacuum, the
routine is the same for the bulk systems, always the radial distribution
function (RDF), g(r), is defined as the number of atoms a distance r from
a given atom inside the spheric shell compared with the number of atoms at
the same distance in an ideal gas, but the atoms in the surface only have
neighboring in the radial direction to the center of the drop, there is
some consideration being taken or is the same for the bulk system and
don't matter if they are atoms in the surface or inside. Thanks, I hope
you soon replay.
The RDF is just a statistic, and you are free to calculate a conditioned/redefined version anyway you wish, whether meaningful or not. The RDF is only typically meaningful for statistically homogeneous systems, where the center of mass position of any pair can be trivially integrated out. If its a fully periodic system with no location specific forcing or means of distributing particles, your case is statistically homogeneous and you can use the traditional g®.
Hello lammps users, I have a physical question, anyone know how you could
calculate the radial distribution function for droplets in the vacuum, the
routine is the same for the bulk systems, always the radial distribution
function (RDF), g(r), is defined as the number of atoms a distance r from
a given atom inside the spheric shell compared with the number of atoms at
the same distance in an ideal gas, but the atoms in the surface only have
neighboring in the radial direction to the center of the drop, there is
some consideration being taken or is the same for the bulk system and
don't matter if they are atoms in the surface or inside. Thanks, I hope
you soon replay.
please see above for a "replay" of your inquiry and find my "reply" below.
a radial distribution function is a statistical mechanical concept
that is derived from treating a simple atomic liquid as an ideal gas
with only pairwise additive interactions added. as such it is a
measure of (short-range) structure and can thus be related to
diffraction data and the potential function. for droplets or clusters,
this will have a bit of an odd shape since you have a convolution of
the short range structure with the long(er) range structure of the
clusters.
what you seem to be asking for, however, is something different, a
spherical density distribution. that will require a rather different
kind of calculation. for starters, the reference point would not be a
specific atom, but rather the center of mass of each cluster.
i think you need to think a bit more about what you really want to
compute and learn from that property. a look in an advanced stat mech.
book might help, e.g. "theory of simple liquids" by hansen and
mcdonald.
Adding to Axel’s comments, in the case that you do have a “nice” problem.
You might look into the structure factor as well, i.e. the Fourier space counter part to the RDF. If you’re lucky enough to have a separation of scales, namely the mean droplet size being much smaller than the distance between droplets, the mixing of intracluster and intercluster separation data should not be as severe. Both scales will show their signature.
