press, energy, density became abnormal after a long time of NPT

Dear LAMMPS users,

I’m trying to run a simulation in NPT( 300 K and 1 atm). Everything is OK (the total energy declined gradually, the average press equilibrated at 1 atm, the temperature was fine and the density increased gradually to a stable and desired value) until the timestep reached 24000000.The total energy and press suddenly increased, the temperature fluctuated more widely and the density decreased from the stable value to a very small and unphysical value.

What caused this phenomena and how to correct it?

Thank you very much!

here is my input file:

units real

atom_style full

bond_style harmonic

angle_style harmonic

dihedral_style hybrid charmm fourier

pair_style hybrid lj/charmm/coul/long 10.0 11

kspace_style pppm 1.0e-5

pair_modify mix arithmetic

special_bonds lj 0.0 0.0 0.0 coul 0.0 0.0 0.8333 dihedral yes

neighbor 2.0 bin #

neigh_modify every 5 delay 5 check yes #

----------------- Atom Definition Section -----------------

read_data 128375537553755.data

#read_restart 128newnpt200000.restart

----------------- Settings Section -----------------

include “5CBs.in.settings”

#----------------- Minimize -----------------

thermo 5

#thermo_style custom step temp pe ke etotal evdwl ecoul elong press density

thermo_style custom step temp pe etotal evdwl ecoul elong press vol density

fix min all box/relax iso 1.0 vmax 0.01

minimize 1.0e-6 1.0e-8 1000 10000

reset_timestep 0

restart 500000 128newnpt*.restart

----------------- Run Section -----------------

timestep 2 #1 fs=10^(-15) s

velocity all create 300 12345678 loop geom

thermo 2000

dump 1 all custom 15000 vmd.lammpstrj id mol type x y z ix iy iz

dump_modify 1 format 1 %10d format 2 %10d format 3 %10d format 4 %12.6f &

format 5 %12.6f format 6 %12.6f format 7 %10d format 8 %10d format 9 %10d

compute 1 all chunk/atom molecule ids once

compute 2 all inertia/chunk 1

fix intertia1 all ave/time 5000 1 5000 c_2[*] file tmp.out mode vector &

title1 “Ixx,Iyy,Izz,Ixy,Iyz,Ixz”

fix 1 all nve

fix tempres all temp/rescale 1 300 300 1 1

run 1000000

unfix 1

unfix tempres

fix 1 all nve

fix tempres all temp/rescale 1 300 300 1 1

fix press1 all press/berendsen iso 1.0 1.0 1500.0 modulus 500000

run 5000000

write_data 128temres300p1.*.data

unfix 1

unfix tempres

unfix press1

fix 1 all nve

fix temp1 all temp/berendsen 300 300 1200.0

fix press1 all press/berendsen iso 1.0 1.0 1500.0 modulus 500000

run 35000000 #35ns

write_data 19128tem300p1atm*.data

density.png

press.png

TotEng.png