forces between two atoms using reax

Hi John,

The energies you’ve outputted from LAMMPS are not only from vdW interactions, probably because of a bug in energy tallying. The manual explains how to output various energy terms from reax/c and reax properly. I’ve checked the values and the minimum is indeed about -0.15 kcal/mol and is exactly the same for both cases.



Thanks for setting John straight about the difference between LAMMPS evdwl and ReaxFF van der Waal energy. Just to be clear, this behavior is a feature, not a bug, because it has been carefully documented for both pair_style reax and reax/c:

“The thermo variable evdwl stores the sum of all the ReaxFF potential energy contributions, with the exception of the Coulombic and charge equilibration contributions which are stored in the thermo variable ecoul.”


I am not surprised that the values of evdw calculated by the two pair styles do not exactly match. The two codes use slightly different default values for some of the more obscure bits of ReaxFF. Again, this is a feature, not a bug. In the case of reax, some of these defaults can be changed using optional keywords to the pair_style command. In the case of reax/c, you can change many default values using the control file. In your case, it may be that adding the ‘tripflag 1’ will change the C-N bond energy, match it to reax/c. This is why it is helpful to compare individual energy sub-categories. Read the doc pages for more information.


Thank you for your replies. I understand now.