Reax/c and nve temperature keep increasing

I searched all the problem case of the mail list, but there is no situation like my problem. I really need you guys help.

I used the latest lammps version, which is August 7, 2019. I’m trying to test the B13C2 stability at 300K. I first minimize the system, and then give a 10K stage, after that, I use NVE and temp/rescale to increase the temperature from 10K to 300K. Everything here seems ok, but when I use NVE to run 100000 steps (10ps) to test the stability, the problem happened. The total energy fluctuant just 2eV, seems reasonable, but the temperature slowly keeps increasing from 300K to 340K.

I tried to change timestep to 0.01 and run 10ps, the temperature still keeps increasing.

My input file and part of the log file are shown below. Attachment is my total running package.

Any help is really appreciated.
Peng

Heat B13C2 crystal, ReaxFF test

Initialization

units real

dimension 3

processors 2 2 2

boundary p p p

atom_style charge

box tilt large

Atom definition

read_data data.B13C2

replicate 5 5 5 bbox

Settings

pair_style reax/c NULL

pair_coeff * * ffield.reax B C

fix 1 all qeq/reax 1 0.0 10.0 1.0e-6 reax/c

thermo 1

thermo_style custom step temp press pe ke etotal vol density epair

dump 99 all custom 1 dump.reax.atom id type mass xs ys zs

timestep 0.1

Minimization

min_style cg

minimize 1.0e-15 1.0e-15 10000 10000

Heating Stage

velocity all create 10.0 12345 rot yes dist gaussian

fix 2 all nve

run 10000

unfix 2

Heat up

fix 3 all nve

fix 4 all temp/rescale 1 10.0 300.0 1.0 1.0

run 10000

unfix 3

unfix 4

fix 5 all nve

run 100000

LAMMPS (7 Aug 2019)

Reading data file …

triclinic box = (0 0 0) to (5.1868 4.73801 4.53553) with tilt (2.11049 2.11049 1.37031)

2 by 2 by 2 MPI processor grid

reading atoms …

15 atoms

read_data CPU = 0.187428 secs

Replicating atoms …

triclinic box = (0 0 0) to (25.934 23.69 22.6776) with tilt (10.5524 10.5524 6.85153)

2 by 2 by 2 MPI processor grid

bounding box image = (0 0 0) to (0 0 0)

bounding box extra memory = 0.00 MB

average # of replicas added to proc = 38.00 out of 125 (30.40 %)

1875 atoms

replicate CPU = 0.000590121 secs

WARNING: Using ‘neigh_modify every 1 delay 0 check yes’ setting during minimization (…/min.cpp:168)

Neighbor list info …

update every 1 steps, delay 0 steps, check yes

max neighbors/atom: 2000, page size: 100000

master list distance cutoff = 12

ghost atom cutoff = 12

binsize = 6, bins = 8 6 4

2 neighbor lists, perpetual/occasional/extra = 2 0 0

(1) pair reax/c, perpetual

attributes: half, newton off, ghost

pair build: half/bin/newtoff/ghost

stencil: half/ghost/bin/3d/newtoff

bin: standard

(2) fix qeq/reax, perpetual, copy from (1)

attributes: half, newton off, ghost

pair build: copy

stencil: none

bin: none

Setting up cg style minimization …

Unit style : real

Current step : 0

Per MPI rank memory allocation (min/avg/max) = 147.9 | 148 | 148.1 Mbytes

Step Temp Press PotEng KinEng TotEng Volume Density E_pair v_eqeq

0 0 -83658.22 -285267.37 0 -285267.37 13932.629 2.4516693 -285267.37

45 0 -172668.82 -290227.34 0 -290227.34 13932.629 2.4516693 -290227.34

Loop time of 89.6945 on 8 procs for 45 steps with 1875 atoms

99.1% CPU use with 8 MPI tasks x no OpenMP threads

Minimization stats:

Stopping criterion = linesearch alpha is zero

Energy initial, next-to-last, final =

-285267.36735 -290227.34287 -290227.342859

Force two-norm initial, final = 4843.69 12357.2

Force max component initial, final = 143.549 7263.64

Final line search alpha, max atom move = 5.00848e-17 3.63798e-13

Iterations, force evaluations = 45 719

MPI task timing breakdown:

Section | min time | avg time | max time |%varavg| %total

data.B13C2 (888 Bytes)

ffield.reax (17.6 KB)

in.B13C2 (1.04 KB)