meam and reax

I have installed the last version of LAMMPS (20apr12)
standard version + meam/reax packages.
I would like to simulate a gold bulk (111) with
six layers in c-direction using the meam potential
plus the charge equilibration (qeq).
The prepared crystal is correct.
The simulations are working well if I use only meam.

If I create just one layer of the gold crystal the
meam+qeq functions very well and I get
the right charges during the dynamics.
However, if I have more than one layer I get a polarization
catastrophy so the charges are growing very fast, then they become
‘nan’ and after that the system acts like there are only
meam interactions. This happens even if I use smaller timesteps.

I used always the meam and qeq parameters which are given
in the LAMMPS-examples.

This is the input:

units metal
boundary p p p
atom_style charge
lattice fcc 4.07 orient x 1 1 1 orient y 1 -1 0 orient z 1 1 -2
region box block 0 2.0 0 2.0 0 1.5
create_box 1 box
create_atoms 1 box units lattice
group gold type 1
pair_style meam
pair_coeff * * library.meam Au Au.meam Au
neighbor 0.3 bin
neigh_modify delay 5
set group gold charge 0.0
thermo 100
fix qeq all qeq/reax 1 0.0 10 0.000001 param.qeq
fix thermostat all temp/berendsen 298 298 0.1
fix movement all nve
dump 1 all custom 10 gold_a_meam.lammpstrj x y z q
timestep 0.001
run 1000

I don’t know what I can do now? Could you please help me?

I have installed the last version of LAMMPS (20apr12)
standard version + meam/reax packages.
I would like to simulate a gold bulk (111) with
six layers in c-direction using the meam potential
plus the charge equilibration (qeq).
The prepared crystal is correct.
The simulations are working well if I use only meam.

If I create just one layer of the gold crystal the
meam+qeq functions very well and I get
the right charges during the dynamics.
However, if I have more than one layer I get a polarization
catastrophy so the charges are growing very fast, then they become
'nan' and after that the system acts like there are only
meam interactions. This happens even if I use smaller timesteps.

I used always the meam and qeq parameters which are given
in the LAMMPS-examples.

This is the input:

units metal
boundary p p p
atom_style charge
lattice fcc 4.07 orient x 1 1 1 orient y 1 -1 0 orient z 1 1 -2
region box block 0 2.0 0 2.0 0 1.5
create_box 1 box
create_atoms 1 box units lattice
group gold type 1
pair_style meam
pair_coeff * * library.meam Au Au.meam Au
neighbor 0.3 bin
neigh_modify delay 5
set group gold charge 0.0
thermo 100
fix qeq all qeq/reax 1 0.0 10 0.000001 param.qeq
fix thermostat all temp/berendsen 298 298 0.1
fix movement all nve
dump 1 all custom 10 gold_a_meam.lammpstrj x y z q
timestep 0.001
run 1000

I don't know what I can do now? Could you please help me?

you are disregarding the fact, that the MEAM potential
is completely oblivious to any charges that you set.

so using fix qeq/reax is meaningless.

axel.