[lammps-users] Spatially-Resolved Bond Angle Histograms


I would like to generate histograms of bond angle sets calculated for different layers within a block of atoms. I thought I might accomplish this through the "compute angle/local" and "fix ave/spatial" commands.

I get an error that states "compute angle/local used when angles are not allowed". I am using "atomic" atom style which I understand does not support angles. Is there some way I can command LAMMPS to compute bond angle distributions while using "atomic" style?


Nicholas Jabari Lee

Compute angle/local only knows about angles you have defined.
It doesn't find angles by doing triple loops over atoms and pairs.
You'd have to write a compute to do that yourself, kind of like
the Tersoff of SW potential does. You could make the compute output
a list of "local" info, like compute angle/local does, which could
then be output by other commands. Fix ave/spatial only
works with per-atom inputs, not local, however.