atoms run apart while they are relaxing

Dear All,

I tried to relax the system of silicon nitride and PTFE(polymer with the elements of C and F).unfortunately ,while the system is relaxing , the atoms belonging to Si3N4 is running apart as shown below. And the PTFE was supposed to be adsorbed to the Si3N4 surface ,but it didn’t.

it is strange for me because the temperature I set is not high and the potential is fine to my system I guess.

Below is my input file. Hopefully some one can shed light on this problem.

Thank you in advance

Yours sincerely

Tina

units real

dimension 3

boundary p p f

neighbor 0.8 bin

atom_style full

neigh_modify delay 5 one 5000 page 1000000

pair_style hybrid tersoff lj/cut/coul/long 12

bond_style class2

read_data PTFEpackmol.data extra/atom/types 2

read_data si3n4command.data add append offset 2 0 0 0 0 group si3n4 shift -10 -17 20

pair_coeff * * tersoff Si3N4.tersoff NULL NULL N Si

pair_coeff 1 1 lj/cut/coul/long 0.054 4.01

pair_coeff 2 2 lj/cut/coul/long 0.059 3.2

pair_coeff 1 2 lj/cut/coul/long 0.056 3.61

pair_coeff 2 4 lj/cut/coul/long 0.107 3.825

pair_coeff 2 3 lj/cut/coul/long 0.0623 3.635

pair_coeff 1 3 lj/cut/coul/long 0.059 4.4

pair_coeff 1 4 lj/cut/coul/long 0.101 4.23

kspace_style ewald 1.0e-5

kspace_modify slab 3.0

region ptfe block INF INF INF INF INF 20

group ptfe region ptfe

compute tep all temp

velocity all create 300 987098 temp tep

fix 1 all nvt temp 300 300 100

fix 2 all temp/rescale 100 300 300 0.01 1

fix 3 si3n4 setforce 0 0 0

timestep 0.1

thermo 100

dump 1 all custom 100 nvtxyz.lammpstrj id type x y z

run 18000

Have you modified the parameters in the Tersoff potential file according to the real units used in your main script?

Bruce

Huijie Tang <[email protected]…127…> 于 2019年6月23日周日 10:10写道:

there are several problems with this input:

  1. you are choosing real units, but tersoff style potentials in LAMMPS usually use metal units. this is rather irrelevant, since you are using fix setforce to set all resulting forces for the atoms set to interact via tersoff to zero.

  2. however, fix setforce 0.0 0.0 0.0 is not sufficient to immobilize atoms. you also need a velocity of zero for those atoms. but that is not the case here. so those atoms will keep moving at constant velocity (no forces means no change of velocity)

  3. you are using two thermostats on the same atoms. that is a very bad idea. don’t use fix temp/rescale, anyway. it is bad for any serious calculation.

axel.

Dear Axel,

I am so sorry to reply you individually just now cause it was my first time to find help in mail list and i will pay more attention to it in the future. and thanks a million for your detailed and useful reply again !
Acoording to your suggestion, I modified my input but the atoms belonging to Si3N4 still run apart. Would you mind pointing out the mistakes in my current input ? The picture and current input are as follow.

Thank you in advance !

图片.png
units metal
dimension 3
boundary p p f
neighbor 0.5 bin
atom_style full
neigh_modify delay 5 one 5000 page 1000000

pair_style hybrid tersoff lj/cut/coul/long 5
bond_style class2

read_data PTFEpackmol.data extra/atom/types 2
read_data si3n4command.data add append offset 2 0 0 0 0 group si3n4 shift -10 -17 20

pair_coeff * * tersoff SiNCkrollm.tersoff NULL NULL N Si
pair_coeff 1 1 lj/cut/coul/long 0.054 4.01
pair_coeff 2 2 lj/cut/coul/long 0.059 3.2
pair_coeff 1 2 lj/cut/coul/long 0.056 3.61
pair_coeff 2 4 lj/cut/coul/long 0.107 3.825
pair_coeff 2 3 lj/cut/coul/long 0.0623 3.635
pair_coeff 1 3 lj/cut/coul/long 0.059 4.4
pair_coeff 1 4 lj/cut/coul/long 0.101 4.23
kspace_style ewald 1.0e-5
kspace_modify slab 3.0

compute tep all temp
velocity all create 300 987098 temp tep
fix 1 all nvt temp 300 300 100

timestep 0.001
thermo 100

dump 1 all custom 100 nvtxyz.lammpstrj id type x y z
run 1000

you haven’t addressed all the issues i pointed out. issue 2) still applies. axel.

图片.png