How to obtain rebo potential/force for each atom in single CNT

Dear all,

Since the “compute group/group” only works for pairwise interaction like LJ, I wonder how to obtain the rebo potential and force for each atom in a single CNT, if I use “pair_style rebo”.

Thank you very much in advance.


pe/atom and atomic forces are per-atom properties that are independent of which pair style used. Please see compute pe/atom and dump commands for how to output per-atom properties.


Thanks Ray.

Assume there are two CNTs with rebo for bonded interaction and LJ for non-bonded interaction. How do I get the rebo potential/force for each atom?

Same question -> same answer.

Hi Axel,

Sorry that I could not figure out how come it is the same question.
Since both the rebo and LJ exist, should the potential and force obtained from “compute pe/atom and dump” be the assembly of results of both rebo and LJ? I may only want to see the atomic potential/force from the effect of rebo.




Please note that rebo is an atomic potential so that both rebo and LJ are essentially non-bonded and are considered whenever a pair of atoms fall within the cutoffs.

Extracting pure rebo or LJ from a rebo+LJ pair_style has been discussed quite a few time in the list. Search the list and you should be able to find the answer.



Thanks very much for the reply.
Actually I have been searching for quite a while before posting. Maybe I didn’t search in a correct way. Thanks for the information and I will try to search again to see if I can find anything related.



I’m not clear what you are asking. You have mentioned:

compute group/group, as it you want the force on some atoms due to only some atoms
LJ + rebo, which are both terms within airebo

LAMMPS cannot do compute group/group for manybody potentials
as it doens’t make sense.

LAMMPS does not bookkeep different parts of the AIREBO potential
(LJ vs REBO) so that you can extract different portions of the energy,
either for the entire system or on a per-atom basis.

So you need to clearly explain what energy you want to extract,
and given what I just explained, it may be impossible to extract
on the fly. Another option is to use the rerun command with a different
potential (e.g. airebo with LJ turned off) and do a post-computation
on your dump snapshots to compute some energy term you want.