I have been trying to calculate the Radial distribution function of amorphous silicon using LAMMPS. Initially I used the compute rdf command which gave the RDF for smaller distances. However for my research purpose I need the RDF for longer distances. I searched through the mail archive and learnt about the command rerun. The method I have used is to 1) get the atomic coordinates using script 1.txt (attached) in the Dump.atom file , and then 2) use script2. txt to get the complete RDF using ‘rerun’ command. But, still obtain RDF for small distances.
Can someone suggest any remedy
script1.txt (618 Bytes)
script2.txt (618 Bytes)
You might have an easier time using the RDF feature in VMD. Alternatively, you can increase the neighbor distance beyond the force cutoff with the skin parameter of the “neighbor” command. I’m assuming the rdf function just goes through the neighbor lists, which are, without this command, built based on the force cutoff.
As per your advice I tried to use VMD for calculating the RDF. But, when I load the dump file and use the Radial Pair distribution Function tool I get the following message saying 0 frames processed and a blank plot. Can someone suggest whats going wrong?
script1.txt (617 Bytes)
Dump.atom (163 KB)
Looks like you’ve entered inconsistent values in the frames text-boxes. Since you have three frames in your trajectory file - and if you want to compute the g® from all three - you might want to put ‘0’ for the first frame, and ‘2’ for the last frame.
By the way, if you have specific questions, there is a mailing list for VMD.
Your 2nd script is asking for the RDF out to 100 Angs.
I presume when you did that it gave you an error
that you can’t compute the RDF beyond the cutoff.
The cutoff in your 2nd script is set by the pair style
which in your case is SW. That has a cutoff hard-wired
in the potential file. Since you don’t care about the
pairwise interactions in your 2nd script to compute RDF,
just use pair lj/cut. And set the cutoff to whatever distance
you want, so long as it is >= the RDF distance.
If you want to use 100 Angs, you should be able to,
though that will be take a while to calculate and take
a lot of memory for a big system.