[lammps-users] Concerns in AIREBO potential for solid Carbon Structure

dear all,

     I have a concern on the AIREBO potential if used for solid state
carbon structure such as Carbon Nanotubes or Graphene. As claimed in
Brenner's 2002 J. Phys.: Condens. Matter paper: for pure carbon
structure the bicubic spline Pij should be zero.

     However, because the file "pair_airebo.cpp" follows Stuart 2000
paper, the Pij term was not turned off at the pure Carbon structure.
Instead Pij contribute to the bond-order (potential) and force,
therefore leads a discrepancy. My tests on (8,8) pristine Carbon
Nanotubes suggested around 5% discrepancy in total energy calculation
which is significant. My impression is that the Pij term should be
effectively turn off when compute pure Carbon system. Can anyone
confirm me if I am right or not.



Haibin Chen Ph.D.
Mechanical Engineering Dept
Carnegie Mellon University
Pittsburgh, Pa, 15213

Can anyone comment on the following AIREBO question?


Please email. Ase Henry <ase@…212…>. Ase wrote the code for this potential and will provide first hand information to your questions.

Hi Haibin,

I checked the two original papers about AIREBO and REBOII potentials.
Actually there is deviation about the determination of P_ij term. The
present lammps software should use the Stuart's (2000) determination
which is shown in Table VIII in this paper.

In your pure carbon study, does the total energy increase or decrease
after the P_ij term turned off? After effectively turning off the
P_ij, are the total energy or other calculated properties more
approaching the experimental results?

Recently I used lammps to calculate the young's modulus of graphene
and found the result is always 5-10% smaller than the published
result. So I'm glad to hear your finding. I also want to know who can
make clear this problem.