Problem with pair_style lj/cut/dipole/cut

Dear LAMMPS users,

I am trying to obtain the pair distribution function for several Au3 clusters, charged, constrained to move on a 2D plane, assuming repulsive interactions between them, following this article: PRL 106, 156101 (2011). For this, I used pair_style lj/cut/dipole/cut (like in the dipole Example) and I obtain this:
ERROR: Pair dipole/cut requires atom attributes q, mu, torque (…/pair_lj_cut_dipole_cut.cpp:362).

I do not understand as it has well created my box with its 800 atoms, with their charge and dipole moment (“800 settings made for dipole”).

Could you please help me and show me my mistakes?

Thank you very very much! (By the way, I am an experimentalist, and no one around me knows how to use LAMMPS…). I feel quite lost…

Here is my program (using LAMMPS 32-bit 2015-12-15 from LAMMPS site):

Initialization

units metal
atom_style dipole
dimension 2
newton on off
boundary p p p

Atom definition

lattice hex 22.5
region mybox block -10 10 -10 10 -0.1 0.1 units lattice
create_box 1 mybox
create_atoms 1 box

Force fields

mass 1 590.91
set group all charge 0.1599
set group all dipole 0.0 0.0 0.6636
velocity all create 1000.0 87287
pair_style lj/cut/dipole/cut 100.0
pair_coeff * * 0.001691 2.934

Settings and fixes

neighbor 2.0 bin
neigh_modify delay 0
compute g6 all hexorder/atom nnn 6
compute PDF all rdf 500
fix 1 all ave/time 500 1 500 c_PDF file tmp.rdf mode vector
fix 2 all planeforce 0.0 0.0 1.0
fix 3 all enforce2d
timestep 0.001
run_style verlet

Output

thermo_style custom step atoms temp epair evdwl ecoul nbuild
thermo 500

dump STMim all image 500 image.*.jpg type type
dump_modify STMim element Au pad 5 acolor 1 black adiam 1 2.7 backcolor white

Actions

run 50000
echo screen

Dear LAMMPS users,

I am trying to obtain the pair distribution function for several Au3
clusters, charged, constrained to move on a 2D plane, assuming repulsive
interactions between them, following this article: PRL 106, 156101 (2011).
For this, I used pair_style lj/cut/dipole/cut (like in the dipole Example)
and I obtain this:
ERROR: Pair dipole/cut requires atom attributes q, mu, torque
(../pair_lj_cut_dipole_cut.cpp:362).
I do not understand as it has well created my box with its 800 atoms, with
their charge and dipole moment ("800 settings made for dipole").

Could you please help me and show me my mistakes?

you are missing the torque property, which is not included in atom_style dipole.
you need to use: atom_style hybrid dipole sphere

please note that you have no time integration below, thus your
simulation will not have any progress.

axel.

And with hybrid sphere dipole you will need to

use fix nve/sphere update dipole (or similar for nvt/npt) to do the integration.

I added a few lines to the atom_style command to explain

that atom_style dipole by itself is not sufficient

for Stockmayer-like models.

Steve

Thank you very much for your help!!

It is working now!

I wish you all a good day and a nice week-end!

Marion