rigid, fix ave/atom

Hi, all

I was wondering if the fix ave/atom command deals with rigid molecules properly? I did a simulation with a number of water molecules.

I used the following commands:

compute 1 water property/atom xu yu zu

fix 1 water ave/atom 50 100 _1[1] c_1[2] c_1[3]

fix 2 water rigid/npt/small molecule temp 280 280 100 iso 10 10 1000

dump 1 water custom .1 id type f_1[1] f_1[2] f_1[3]

visualizing the dump file with VMD, several water molecules connected each other (as image attached). what is wrong? how can we output the time averaged xu yu zu calculated by command fix ave/atom for rigid molecules?

a.png

Hi, all

   I was wondering if the fix ave/atom command deals with rigid molecules
properly? I did a simulation with a number of water molecules.

if you are using version 30 Jan 2015 of LAMMPS or later, fix rigid
should have no effect. before that, please look at the version of the
manual that was bundled with it.

please also note that averaging absolute positions only makes sense
for crystals and otherwise immobilized objects. particularly for rigid
water molecules, there is little meaning to it, since it will almost
always result in unphysical configurations. specifically the rotation
of the molecules will result in a shrinkage of the structure (as the
average position of a circle is its center.

axel.