[lammps-users] Equilibration problem

Dear friends,
I am unable to get my system to equilibrate at a constant temperature.The atoms seem to fly off from the surface. I am trying to get a 2 simulation of a layer of Copper atoms.Is it because NVE is not used? Could anyone please tell me how to go about equilibrating such a system in LAMMPS? I have done this using a custom C++ code. Trying to get that converted to LAMMPS for efficiency

Code:

dimension 2
units metal
boundary p p p
atom_style atomic
neighbor 3 bin
neigh_modify delay 1

lattice hex 1
region box block 0 142.1 -75 75 -0.25 0.25
create_box 5 box
mass 1 1
mass 2 1
mass 3 56
mass 4 56
mass 5 56

region lof block 0 141.1 0 1.1 INF INF
region lo-slab block 0 141.1 1.1 8.1 INF INF
create_atoms 1 region lo-slab
create_atoms 2 region lof

pair_style eam
pair_coeff * * cuu3.eam
group mobile region lo-slab
group mobile type 2
group boundary region lof

fix 1 boundary rigid single
fix 2 boundary setforce 0 0 0
fix 3 all npt 200.0 200.0 100.0 xyz 0 1 1000.0
velocity all create 200.0 475367 dist gaussian

thermo 50
dump 100 all atom 1 dump.npt2
log log.npt2
run 300000

Thanks in Advance,
Regards,
Nandu Gopan

Nandu,

Have u minimized your system before running NPT… usually minimization is done before running any dynamics…

Navdeep

Using fix rigid and fix npt on the same atoms will integrate
them twice which is bad. You should have got a warning
about this. If all you want to do is freeze the boundary, then
don't use fix rigid. Just don't integrate them, i.e. use the
group option in fix npt. I would start by running
fix nve on the system and see if it does what you expect.

Steve