WARNING: Using a manybody potential with bonds/angles/dihedrals and special_bond exclusions

Hi, all

When I used EAM potential, a warning appeared: Using a many body potential with bonds/angles/dihedrals and special_ Bond exclusions. And I did not perform the so-called “exclusion” operation that I actually did not know.

One point I want to clarify is that my system contains water molecules, and the following command is executed:
create_box 5 box bond/types 1 angle/types 1 extra/bond/per/atom 2 extra/angle/per/atom 1 extra/special/per/atom 2

Looking forward to your help, thank you very much!

So what is the problem here?

“exclusions” happen automatically for any pairs of atoms that have bonded interactions. This is part of the force field and done by LAMMPS by default (see the special_bonds command).

Thank you for your explanation. I am no longer worried if this is EAM’s own problem.

However, the explanation of this warning states, "This is like not what you want to do. The exclusion settings will limit neighbors in the neighbor list, which many body potential needs to calculate its terms correctly."

Does this mean that the force field file is not correct for certain calculations?

The explanation assumes the situation where you have only a many-body pair style and also bonds. For example when people take a geometry of a hydrocarbon polymer, load it into VMD and write out a data file without specifying the atom type. Then the data file will contain the bonds that VMD has guessed and the output will have the corresponding topology information. If you then use a many-body force field like AIREBO or ReaxFF, the resulting computations will be bogus, since the atom pairs involved in bonds directly connected to that atom of interest or it next bonded atoms or the two bond apart atoms will all be removed from the neighborlist.

Your situation is special, because you are simulating a hybrid system with a hybrid pair style. For there LAMMPS cannot easily determine whether you have bonds between atoms that are subject to a manybody potential where bonds are implicit and not explicit. All that LAMMPS knows are: 1) there is a many-body interaction in use somewhere, and 2) there are bonds defined somewhere. That is sufficient to trigger the warning.

Whether your model is correct or not is for you to sort out. This is one of the (many) complications of using hybrid potentials.

I see. Thanks for your patient explanation very much!

The same issue i am facing. I have modeled BNNT using VMD. Using topo, bonds and angles are guessed. but dont know how to assign force field. Moreover if i model BNNT using material studio then there is no such force field which can be used for boron. Lammps also has not any .frc file which i can assign while using msi2lmp.

Please check out the various LAMMPS tutorials for learning how to do that.
You didn’t say which force field you want to use.

Sir i have done a lot of research and finally i posted here. I can use Compass or cvff for nitrogen but can not use them for boron. i have checked all the .frc files. None of them support boron.

Force field atom typing is more complex than trying to match chemical elements.
The discussion on which force field to use is something that has nothing to do with LAMMPS but is a crucial choice for your research. Thus you need to discuss this before even considering to start a simulation with your adviser or supervisor or tutor or experienced colleagues. This is not a LAMMPS problem but a problem about how to conduct research and thus you need to discuss it with people that care about your research. I would expect that neither Compass nor cvff are good choices for your compound and that you want to use something else that has been specifically parameterized for it.