Thanks a lot for your answer, axel.

Inclusion and exclusion of connectiviy information with notation of non-zero number of bonds gave different results with enormous enegy difference for AIREBO pontential and Tersoff potential.

Please provide examples demonstrating your claim and state which

version of lammps you are using.

Axel

Dear axel:

The calculation was done for relxation of graphene.

The result for the case that bonds information is deleted is as follows.

LAMMPS (30 Sep 2011)

Scanning data file …

Reading data file …

orthogonal box = (-0.612951 -0.707952 -8e-06) to (48.4271 41.762 30)

17 by 1 by 1 processor grid

800 atoms

Finding 1-2 1-3 1-4 neighbors …

0 = max # of 1-2 neighbors

0 = max # of 1-3 neighbors

0 = max # of 1-4 neighbors

1 = max # of special neighbors

WARNING: Resetting reneighboring criteria during minimization (min.cpp:160)

Setting up minimization …

Memory usage per processor = 4.06111 Mbytes

Step Temp E_pair E_mol TotEng Press

0 0 -5864.3473 0 -5864.3473 28204.605

3 0 -5864.3482 0 -5864.3482 28204.573

Loop time of 0.11466 on 17 procs for 3 steps with 800 atoms

Minimization stats:

Stopping criterion = energy tolerance

Energy initial, next-to-last, final =

-5864.34730996 -5864.34815715 -5864.34815813

Force two-norm initial, final = 0.605478 0.00140365

Force max component initial, final = 0.0213557 8.33994e-05

Final line search alpha, max atom move = 1 8.33994e-05

Iterations, force evaluations = 3 6

Pair time () = 0.00229779 (2.00401)
Bond time () = 0 (0)

Neigh time () = 0 (0)
Comm time () = 0.0289522 (25.2505)

Outpt time () = 0 (0)
Other time () = 0.0834099 (72.7455)

Nlocal: 47.0588 ave 60 max 40 min

Histogram: 11 0 0 0 0 0 0 0 0 6

Nghost: 170.353 ave 200 max 156 min

Histogram: 6 0 0 0 10 0 0 0 0 1

Neighs: 0 ave 0 max 0 min

Histogram: 17 0 0 0 0 0 0 0 0 0

FullNghs: 847.059 ave 1080 max 720 min

Histogram: 11 0 0 0 0 0 0 0 0 6

Total # of neighbors = 14400

Ave neighs/atom = 18

Ave special neighs/atom = 0

Neighbor list builds = 0

Dangerous builds = 0

System init for write_restart …

The result for the case that connectivity is inlcuded is as follows.

LAMMPS (30 Sep 2011)

Scanning data file …

2 = max bonds/atom

Reading data file …

orthogonal box = (-0.612951 -0.707952 -8e-06) to (48.4271 41.762 30)

17 by 1 by 1 processor grid

800 atoms

1160 bonds

Finding 1-2 1-3 1-4 neighbors …

3 = max # of 1-2 neighbors

6 = max # of 1-3 neighbors

18 = max # of 1-4 neighbors

18 = max # of special neighbors

WARNING: Resetting reneighboring criteria during minimization (min.cpp:160)

Setting up minimization …

Memory usage per processor = 4.86823 Mbytes

Step Temp E_pair E_mol TotEng Press

0 0 -205.43316 0 -205.43316 671.49155

20 0 -206.58163 0 -206.58163 0.066929141

Loop time of 0.78102 on 17 procs for 20 steps with 800 atoms

Minimization stats:

Stopping criterion = energy tolerance

Energy initial, next-to-last, final =

-205.433156537 -206.581627833 -206.581627903

Force two-norm initial, final = 12.3536 0.00519482

Force max component initial, final = 2.4606 0.00364947

Final line search alpha, max atom move = 0.03125 0.000114046

Iterations, force evaluations = 20 49

Pair time () = 0.000459559 (0.0588408)
Bond time () = 0 (0)

Neigh time () = 0 (0)
Comm time () = 0.190028 (24.3307)

Outpt time () = 0 (0)
Other time () = 0.590533 (75.6105)

Nlocal: 47.0588 ave 60 max 40 min

Histogram: 11 0 0 0 0 0 0 0 0 6

Nghost: 170.353 ave 200 max 156 min

Histogram: 6 0 0 0 10 0 0 0 0 1

Neighs: 0 ave 0 max 0 min

Histogram: 17 0 0 0 0 0 0 0 0 0

FullNghs: 62.5882 ave 260 max 28 min

Histogram: 13 2 0 0 0 0 0 0 0 2

Total # of neighbors = 1064

Ave neighs/atom = 1.33

Ave special neighs/atom = 16.67

Neighbor list builds = 0

Dangerous builds = 0

System init for write_restart …

The version of LAMMPS program was 30 Sep 2011.

There is not any other difference between two cases except for the change in data file.

I will send the files for it to you.

Best regards,

Young

2011/10/27 JhonY. I. <yijhon@…8…>

Dear axel:

The calculation was done for relxation of graphene.

The result for the case that bonds information is deleted is as follows.

LAMMPS (30 Sep 2011)

Scanning data file …

Reading data file …

orthogonal box = (-0.612951 -0.707952 -8e-06) to (48.4271 41.762 30)

17 by 1 by 1 processor grid

800 atoms

Finding 1-2 1-3 1-4 neighbors …

0 = max # of 1-2 neighbors

0 = max # of 1-3 neighbors

0 = max # of 1-4 neighbors

1 = max # of special neighbors

[…]

Ave special neighs/atom = 0

Neighbor list builds = 0

Dangerous builds = 0

System init for write_restart …

The result for the case that connectivity is inlcuded is as follows.

LAMMPS (30 Sep 2011)

Scanning data file …

2 = max bonds/atom

Reading data file …

orthogonal box = (-0.612951 -0.707952 -8e-06) to (48.4271 41.762 30)

17 by 1 by 1 processor grid

800 atoms

1160 bonds

Finding 1-2 1-3 1-4 neighbors …

3 = max # of 1-2 neighbors

6 = max # of 1-3 neighbors

18 = max # of 1-4 neighbors

18 = max # of special neighbors

[…]

Ave neighs/atom = 1.33

Ave special neighs/atom = 16.67

Neighbor list builds = 0

Dangerous builds = 0

System init for write_restart …

The version of LAMMPS program was 30 Sep 2011.

There is not any other difference between two cases except for the change in data file.

there is one additional difference: the “special neighbors”

just compare the outputs. by default atom pairs that have

an exclusion (e.g. share a bond) are removed from the

neighbor lists. this is due to the default of

special_bonds lj/coul 0.0 0.0 0.0

if you add

special_bonds lj/coul 1.0 1.0 1.0

to your input with bonds, you will get the correct answer.

LAMMPS is doing what is reasonable, i.e. assume that

if you include bond definitions, that you want to do exclusions.

it cannot easily know without checking for each manybody

potential explicitly by name, that you may not want that.

and then again, if you do a hybrid calculation with a pairwise

force field with bonds, you do need those exclusions.

that would be the case, where you *have* to remove the

bonds from the atoms that are treated with a manybody

potential to avoid them being excluded.

cheers,

axel.

Many thanks for your answer, axel.

It helps a lot for my understanding the bonds in LAMMPS.

Best regards,

Young