[lammps-users] Pressure Not Equilibrating in NPT

Steve,

I’m trying to run an NPT simulations of monatomic water with Si.sw pair coefficients (as suggested by a research paper). This model of water ignores the hydrogens and the electrostatics.
The simulation is running; however, the pressure is not equilibrating even after 1000000 steps. The researchers who previously worked on this model used Nose-Hoover thermostat and barostat, which is why i didn’t add a drag value (since it alters the equations). Also, the relaxation time given below is as suggested by the paper. I’m using 32,768 water particles. Any suggestions of how to get my pressure to equilibrate will be extremely helpful.

Thank you.

The following is my input file:

#Lammps input
dimension 3
units metal

boundary p p p

kspace_style none

atom_style full

pair_style sw

newton on off

read_data FilledConfig.data
pair_coeff * * Si.sw Si
neigh_modify delay 4 every 1

fix 1 all npt 350 350 1.0 xyz 1.01325 1.01325 5.0
timestep 0.01
thermo 1000
thermo_style custom step ke pe etotal evdwl press vol temp
compute 1 all temp

restart 100000 Cyclo.restart
dump 10 all custom 1000 bulkmwtr2.dump id type x y z fx fy fz
run 1000000

Following is my the last few runs of my output:

STEP KE PE ETotal E_vdwl Pressure Volume Temp
917000 1486.935 -14006.302 -12519.368 -14006.302 -14.539066 995734.91 351.06752
918000 1488.0142 -14014.755 -12526.741 -14014.755 -5.5419105 995502.16 351.32232
919000 1484.0819 -14027.879 -12543.797 -14027.879 -63.095907 996376.35 350.3939
920000 1485.9079 -14014.155 -12528.247 -14014.155 16.865016 997138.36 350.82504

Hi,

I’m trying to run an NPT simulations of monatomic water with Si.sw pair coefficients (as suggested by a research paper). This model of water ignores the hydrogens and the electrostatics.
The simulation is running; however, the pressure is not equilibrating even after 1000000 steps. The researchers who previously worked on this model used Nose-Hoover thermostat and barostat, which is why i didn’t add a drag value (since it alters the equations). Also, the relaxation time given below is as suggested by the paper. I’m using 32,768 water particles. Any suggestions of how to get my pressure to equilibrate will be extremely helpful.

Thank you.

The following is my input file:

#Lammps input
dimension 3
units metal

boundary p p p

kspace_style none

atom_style full

pair_style sw

newton on off

read_data FilledConfig.data
pair_coeff * * Si.sw Si
neigh_modify delay 4 every 1

fix 1 all npt 350 350 1.0 xyz 1.01325 1.01325 5.0
timestep 0.01
thermo 1000
thermo_style custom step ke pe etotal evdwl press vol temp
compute 1 all temp

restart 100000 Cyclo.restart
dump 10 all custom 1000 bulkmwtr2.dump id type x y z fx fy fz
run 1000000

Following is my the last few runs of my output:

STEP KE PE ETotal E_vdwl Pressure Volume Temp
917000 1486.935 -14006.302 -12519.368 -14006.302 -14.539066 995734.91 351.06752
918000 1488.0142 -14014.755 -12526.741 -14014.755 -5.5419105 995502.16 351.32232
919000 1484.0819 -14027.879 -12543.797 -14027.879 -63.095907 996376.35 350.3939
920000 1485.9079 -14014.155 -12528.247 -14014.155 16.865016 997138.36 350.82504

vidhya,

instantaneous pressure always fluctuates a lot.
you have to look at averages. please check the
manual and the mailing list archives. this topic has
been discussed many, many times.

cheers,
    axel.