[lammps-users] I've a problem when I run nvt after energy minimization

Dear All:

I’m trying to balance a nickel particle. I use minimize first, then nvt-nve. It’s said if the temperature is stable during nve, it means the system is in equilibrium status. However, it always gose run when I run this input file. Sometimes, it goes wrong at the very beginning. Sometimes, it goes wrong after minimization but before nve. The error is “segmentation fault” all the time. I’ve searched some information on the internet, it seems there’s a bug in the old version of Lammps. But it’s fixed already. So I really don’t know what’s the problem of this input file. Besides, I’ve tried to run minimization and nvt-nve separately, and it goes well. In addition, is nve the best choice to do the balance in this case? How about npt? How should I choose the right one?

So, is anybody can help me figure out this problem? I really appreciate it ! Thanks !

#-------------------- 3d simulation of nickle particle—balance------------

units metal
atom_style atomic
boundary p p s
neighbor 2.0 bin
neigh_modify delay 5
timestep 0.001

#-------------------- create geometry---------------------

lattice fcc 3.52 orient x 1 0 0 orient y 0 1 0 orient z 0 0 1
region sphere_region sphere 0 0 0 5 units lattice
region box_region block -10 10 -10 10 -10 10 units lattice

create_box 1 box_region
create_atoms 1 region sphere_region

mass 1 58.710

EAM potential

pair_style eam/opt
pair_coeff 1 1 /home/lyang/lammps-20Feb10/potentials/Ni_u3.eam

define groups

group all type 1

#------------------- minimize-------------

min_style cg

minimize 1.0e-8 1.0e-8 10000 10000

#------------- compute and thermo---------------------

compute 1 all ackland/atom
compute 2 all msd com yes
compute 3 all gyration
variable msd equal c_2[4]
variable gyration equal c_3
thermo 10
thermo_modify flush yes
thermo_style custom step temp vol press pe ke etotal epair v_msd v_gyration

------------- Equilibration and thermalisation ----------------

reset_timestep 0
dump 4 all custom 20000 dump2.atom.gz id type x y z vx vy vz
dump 5 all xyz 1000 dump2.xyz.gz
dump 6 all custom 1000 ackland2.gz c_1

fix NVT all nvt 300.0 300.0 0.1 drag 0.2
run 20000
unfix NVT

fix NVE all nve
run 20000

I see no errors running your script on 1 proc (for less timesteps
in NVT and NVE). No memory errors either. So I can't reproduce
your problem. You probably do want to set your initial temp
(after minimization) to 300K, via the velocity command. NVT with
NH chains does not equilibrate from 0 up to 300K very quickly without
starting close to your target T.

Steve

2010/3/20 Lingqi Yang <[email protected]>: