[lammps-users] Grouping atoms on region boundaries

LAMMPS LAMMPS Mailing List Mirror
1

I am trying to group atoms in an orderly fashion using spatial
dimensions. I am finding that if I split my area of interest into
subregions and then group, if there are atoms sitting exactly on the
subregion boundary the atoms will be grouped in both of the groups on
the two sides of the boundary. Is there away to control in a
predetermined way which group an atom would belong to (i.e. group only
if <= a boundary and > another boundary). I see in the code that the
"match" in regions tests for the inclusion in groups, but I can not find
where "match" is actually defined.

Thanks,

Rob

2

The match function is defined in the region*cpp files. The regions
are all defined to include atoms exactly on their boundaries. So
for your problem, I suggest you input the upper boundary to the region
on the left as x and the lower boundary of the region on the right
as x + epsilon, where epsilon is something like 1.0e-8. As long
as no atoms fall in the epsilon crack, you should be fine.

Steve