Three-body harmonic interaction potential


I am doing a simulation of the fusion of an oxide mixture in Lammps. The system has three types of atoms (A,B,C). I have tried to reproduce a published work where they use a Born-Mayer-Huggins potential and a three-body harmonic interaction potential.

For this I use, in my simulation, a potential of pairs of type buck/long. On the other hand, since atoms A and C should be presented in structures of type AC3 and AC4 and that each of these groups present angles of type C-A-C with defined angles (110 and 120 depending on the configuration), I am interested in including three- body harmonic interaction potential in the simulation.

What would be the best way to add the potential three-body harmonic interaction ?

I was reading the lammps manual and found that angle_style harmonic potential could be included. However, I have some doubts about how to implement it.

Is it necessary to define the angles prior to their use? That is, indicate for each combination of C-A-C the number of particles

Should I raise previously defined links for use? to define the above, you should define the bonds between atoms C-A

Thank you very much for your help

How about this? pair_style nb3b/harmonic command — LAMMPS documentation


Thanks Axel for the suggestion.
I was reading the documentation and I have a question about its implementation.
In order to consider the two possibilities that angles C-A-C could take depending on whether they are as AC3 or AC4, should I have two types of nb3b.harmonic files specifying in each one the possibility?

I don’t understand what you are asking. You have to provide more details and perhaps a more tangible example.


Sorry if I wasn’t clear.

The study system has 955 type B atoms, 409 type Li atoms and 1636 type O atoms. Type A atoms are accommodated as BO3 and BO4 type units (considering a given cutoff), in a given percentage. For example, 40% of the B’s are in BO4 type units. The O-B-O angles in BO3 units are 120 and the O-B-O angles in BO4 units are 110 degrees approx.

The doubts would be

  • When using the pair_style nb3b/harmonic, how could I specify the coexistence of both types of angles?
  • is there a way to specify/limit the fraction of angles that can take the conformation, for example, declared in pair_style nb3b/harmonic?


With a pair style like this you have to specify which B atoms are part of a BO3 group or part of a BO4 group and then give them different atom types (e.g. B3 and B4). Then in the potential file, you need to list all available permutations of triples for B3, B4, Li, and O and assign parameters. Please have a look at the existing potential file MOH.nb3b.harmonic in the potentials folder.

If you want the B3 atoms change into B4 and vice versa, this would be a different step and require some custom programming.

Dear Axel,

Thanks very much!
I was able to solve it with your recommendations