Warnings from Lammps

_Huang_Haoxiang · July 5, 2015, 8:26pm

Hi, all, I found two warnings in my simulation; the information is below:

Finding 1-2 1-3 1-4 neighbors ...
Special bond factors lj: 0 0 0
Special bond factors coul: 0 0 0
  4 = max # of 1-2 neighbors
  8 = max # of 1-3 neighbors
  22 = max # of 1-4 neighbors
  23 = max # of special neighbors
Finding 1-2 1-3 1-4 neighbors ...
Special bond factors lj: 0 0 1
Special bond factors coul: 0 0 1
  4 = max # of 1-2 neighbors
  8 = max # of 1-3 neighbors
  12 = max # of special neighbors
1280 atoms in group CNT
23 atoms in group ket11
24 atoms in group ket12
30623 atoms in group mket
1327 atoms in group CNTket-1
PPPM initialization ...

WARNING: System is not charge neutral, net charge = 0.03 (kspace.cpp:297)

  G vector (1/distance) = 0.241562
  grid = 54 54 27
  stencil order = 5
  estimated absolute RMS force accuracy = 0.000364241
  estimated relative force accuracy = 1.0969e-06
  using double precision FFTs
  3d grid and FFT values/proc = 5200 1512
Neighbor list info ...
  1 neighbor list requests
  update every 1 steps, delay 10 steps, check yes
  master list distance cutoff = 14
Setting up run ...

WARNING: Inconsistent image flags (domain.cpp:682)

Memory usage per processor = 14.4138 Mbytes
Step Atoms Tcnt Tket Press Volume PotEng KinEng TotEng E_long Lx Ly Lz Xlo Xhi Ylo Yhi Zlo Zhi
0 31903 388.15863 209.31861 -2368.8103 369333.35 2129901.6 20586.098 2150487.7 -29443.064
ERROR on proc 39: Bond atoms 8746 8747 missing on proc 39 at step 44 (neigh_bond.cpp:65)

what I guess is the bond atoms may be the reason for the nonzero net charge. At the same time, it causes the inconsistent image flags. If the problem is caused by this bonding, how can I fix that?
(bring the atoms closer or use fix_bond)

Thanks,
Haoxiang Huang

akohlmey · July 5, 2015, 8:42pm

Hi, all, I found two warnings in my simulation; the information is below:

Finding 1-2 1-3 1-4 neighbors ...
Special bond factors lj: 0 0 0
Special bond factors coul: 0 0 0
  4 = max # of 1-2 neighbors
  8 = max # of 1-3 neighbors
  22 = max # of 1-4 neighbors
  23 = max # of special neighbors
Finding 1-2 1-3 1-4 neighbors ...
Special bond factors lj: 0 0 1
Special bond factors coul: 0 0 1
  4 = max # of 1-2 neighbors
  8 = max # of 1-3 neighbors
  12 = max # of special neighbors
1280 atoms in group CNT
23 atoms in group ket11
24 atoms in group ket12
30623 atoms in group mket
1327 atoms in group CNTket-1
PPPM initialization ...

WARNING: System is not charge neutral, net charge = 0.03 (kspace.cpp:297)

0.03 is not a lot.

  G vector (1/distance) = 0.241562
  grid = 54 54 27
  stencil order = 5
  estimated absolute RMS force accuracy = 0.000364241
  estimated relative force accuracy = 1.0969e-06
  using double precision FFTs
  3d grid and FFT values/proc = 5200 1512
Neighbor list info ...
  1 neighbor list requests
  update every 1 steps, delay 10 steps, check yes
  master list distance cutoff = 14
Setting up run ...

WARNING: Inconsistent image flags (domain.cpp:682)

this is worrisome. unless you have deliberately set up bonds that wrap
around periodic boundaries, this is an indication of problems with the
data file, most likely wrong PBC handling or a bad bond topology.

Memory usage per processor = 14.4138 Mbytes
Step Atoms Tcnt Tket Press Volume PotEng KinEng TotEng E_long Lx Ly Lz Xlo Xhi Ylo Yhi Zlo Zhi
0 31903 388.15863 209.31861 -2368.8103 369333.35 2129901.6 20586.098 2150487.7 -29443.064
ERROR on proc 39: Bond atoms 8746 8747 missing on proc 39 at step 44 (neigh_bond.cpp:65)

what I guess is the bond atoms may be the reason for the nonzero net charge. At the same time, it causes the inconsistent image flags. If the problem is caused by this bonding, how can I fix that?

your guess is wrong. charges and bonds are not related. a total charge
error can be accrued through accumulating rounding errors. e.g. when
you have groups of atoms with a partial charges that don't add up to
zero. example: 3x -0.0833 and 1x 0.25.

the inconsistent image flags warning and also the pretty high positive
potential energy are indicators that there is something wrong with
your bond topology, like a bond that is stretched across a very long
distance. then your atoms will speed up quickly, and if they move too
fast, you "lose" them.

as usual, your should do a trajectory dump for every step and watch
closely which atoms/bonds are causing problems. also, close
inspection/visualization of the data file can help.

(bring the atoms closer or use fix_bond)

fix bond what?

axel.

_Huang_Haoxiang · July 5, 2015, 9:08pm

Thank you for your suggestion.
1. I may try to fix the PBC first to see what's going on ( for inconsistent image )
2. The coefficients I will check that again.

the last thing is, when I define the special_bond coul 0.0 0.0 1.0, the bonding will be missing at the beginning of simulation. (ERROR on proc 39: Bond atoms 8746 8747 missing on proc 39 at step 44 (neigh_bond.cpp:65))
Also, as I define the special_bond coul, It is hard to minimize the energy: It will be still really high after 40 steps, but it suddenly drops to high negative number like -6.400 e+15 after step 57.

If I exclude the special_bond coul, even the warning of inconsistent image still exists, the system works OK under energy minimization, NVE, NPT and NVT.

I am not sure why special_bond coul command will change my system seriously.

Best,
Haoxiang Huang

akohlmey · July 5, 2015, 9:19pm

Thank you for your suggestion.

I may try to fix the PBC first to see what’s going on ( for inconsistent image )

The coefficients I will check that again.

the last thing is, when I define the special_bond coul 0.0 0.0 1.0, the bonding will be missing at the beginning of simulation. (ERROR on proc 39: Bond atoms 8746 8747 missing on proc 39 at step 44 (neigh_bond.cpp:65))
Also, as I define the special_bond coul, It is hard to minimize the energy: It will be still really high after 40 steps, but it suddenly drops to high negative number like -6.400 e+15 after step 57.

If I exclude the special_bond coul, even the warning of inconsistent image still exists, the system works OK under energy minimization, NVE, NPT and NVT.

I am not sure why special_bond coul command will change my system seriously.

That would happen, if your non-bonded parameters were not correct, or your bond topology. Do you have the correct units setting? Some charges get very close and are not properly balanced by the LJ parameters. Then the forces get large and ka-boom!

Axel

_Huang_Haoxiang · July 5, 2015, 9:45pm

the units I am using is “full” all the time. I think it is correct.

Thanks.
Haoxiang