Tersoff vs Vashishta potential

LAMMPS LAMMPS Mailing List Mirror
1

Dear Anders,

Many thanks for your reply.

I did the simulation for a smaller box. I have attached the potential functions. You can find the script in the following.

You can see that while using vashishta potential, atoms start flying, however, for the Erhart potential, the simulation is fine.

units metal
boundary p p p
dimension 3
newton on
atom_style atomic

lattice diamond 4.3613 origin 0 0 0 orient x 0 1 0 orient y 0 0 1 orient z 1 0 0
region box1 block -0.2 20.2 -0.2 20.2 -0.2 20.2
region box2 block 0 20 0 20 0 20
create_box 2 box1
create_atoms 2 region box2 basis 1 1 basis 2 1 basis 3 1 basis 4 1 basis 5 2 basis 6 2 basis 7 2 basis 8 2
group box2 region box2

pair_style vashishta
pair_coeff * * SiC.vashishta Si C

#pair_style tersoff
#pair_coeff * * SiC_Erhart-Albe.tersoff Si C

neighbor 5.0 nsq
neigh_modify every 2 delay 10 check yes cluster yes binsize 5
mass 1 28.0855
mass 2 12.0107

min_style sd
minimize 2.5e-6 1.0e-6 100000 100000
velocity box2 create 300 900000
fix 1 all nve
fix 2 box2 temp/rescale 100 300 300 2 1
fix_modify 2 energy yes

thermo 1000
thermo_style custom step temp pe ke etotal
timestep 0.001
run 5000
unfix 2

thermo 100
thermo_style custom step temp
dump 1 all atom 100 dump.Equilib*.lammpstrj
dump_modify 1 scale no
run 200

SiC.vashishta (1.54 KB)

SiC_Erhart-Albe.tersoff (1.6 KB)

2

Thanks,

If you open your system in VMD or Ovito, you will see a cubic box, all covered with Silicon atoms. Since you are using periodic boundary conditions, this means Si atoms are next to Si atoms in a way that has high potential energy. You can easily count the number of atoms: 34461 silicon, 32000 carbon. Why this system seems stable with tersoff I can’t explain, but vashishta does not like this which we see. The total net charge in the system is not zero and silicon atoms are next to each other.

To solve this, you should create a slightly smaller box to remove the extra layer of silicon atoms. For instance, this gave correct number of atoms:

region box1 block 0 20 0 20 0 20

and was also stable with both potentials.

Btw, you should not use neighbor 5.0 nsq with such a large system!

Anders