RDF calculation with different cutoff distances

Hi, I wanted to know if the rdf can be computed using different cutoff distances in the same simulation.

Currently I have atom types 1 and 2, (ions and neutrals) and three possible gr’s (rdf).

However, the cutoff distance between the potentials are different, being 1-1 much larger.

When Lammps calculate the rdf it uses the same cutoff distance for all the interactions, making the rdf 1-1 well resolved but the rdf for 1-2 and 2-2 are too noisy (the features I’m interested for those happen at smaller distances)

Is it possible to compute each rdf with a different cutoff distance, in order to not lose resolution?

This is the way I’m calculting rdf in my input script:

compute rdf all rdf 1000  1 1   2 2   1 2
fix saverdf all ave/time 1 2000 2000 c_rdf[1] c_rdf[2] c_rdf[4] c_rdf[6] file rdf.rdf mode vector

The statistical noise in the RDF is mostly due to the number of pairs that you have at a given distance.
Tweaking the cutoff won’t make a difference since reducing the cutoff will just drop the data that is outside the cutoff.

There is nothing stopping you from defining two compute and fix commands with different settings.

1 Like