Dear all,
I'd like to run a simulation of Gay-Berne mesogens with charges on
them. I've adapted one of the ellipses examples from the lammps
distribution and defined a new
pair_style hybrid ellipsoid charge
However, it seems that charge and ellipsoid are incompatible somehow.
I've set all charges to zero, so only the Gay-Berne potential should
be effective.
The output of this simulation is vastly different from a similar
simulation with only pair_style ellipsoids. The problem is that the
temperature and kinetic energy show too much flcutuations; i.e.
integration is not stable, even with damping.
Could anyone confirm this for me / suggest a fix?
This is the input stream:
units lj
atom_style hybrid ellipsoid charge
dimension 2
lattice sq2 0.05
region box block 0 5 0 5 -0.5 0.5
create_box 2 box
create_atoms 1 box basis 1 2
set group all quat/random 18238
mass 1 1.0
mass 2 1.0
shape 1 3 1 1
shape 2 3 1 1
group cations type 1
group anions type 2
set group cations charge 0
set group anions charge 0
compute ast all temp/asphere
compute ave all pe
velocity all create 1.0 87287 loop geom
pair_style hybrid/overlay gayberne 1.0 3.0 1.0 5.0 coul/cut 10.0
pair_coeff 1 1 gayberne 1.0 1.0 1 1 0.2 0 0 0
pair_coeff 1 2 gayberne 1.0 1.0 1 1 0.2 0 0 0
pair_coeff 2 2 gayberne 1.0 1.0 1 1 0.2 0 0 0
pair_coeff 1 1 coul/cut 10.0
pair_coeff 1 2 coul/cut 10.0
pair_coeff 2 2 coul/cut 10.0
neighbor 0.8 bin
neigh_modify every 1 delay 0 check no
thermo_style custom step epair etotal press vol c_ast c_ave
thermo 1000
timestep 0.000166
fix 1 all nvt/asphere 1.0 1.0 0.1 drag 1.0
fix 2 all enforce2d
dump 1 all custom 1 dump.resquared &
tag type x y z quatw quati quatj quatk q
run 0
undump 1
run 20000
# run dynamics
unfix 1
fix 1 all nve/asphere
fix 3 all ave/time 1 5000 5000 c_ave file ave.out
run 100000