The pair_style eam in LAMMPS is based on the original formulation of EAM which uses a universal cross mixing interaction: r * phi_ij(r) = 27.2 * 0.529 * z_i(r) * z_j(r). You don't specify the cross interactions because they will always be this. Note that this original eam style is primarily used by potentials before 1990 by Daw, Foiles, or Adams, and most others do not use it. If the cross interaction is different than this, then the potential needs to use one of the eam/alloy or eam/fs pair styles instead.

Good afternoon!

There are potentials for different alloys and elements. For alloy modeling, it is best to choose a potential that takes into account the elements of your alloy. Usually, these are potentials like" eam.alloy".

Each potential of the alloy must reproduce the properties of each simple (single) element of which it is composed.

Potential files here:

https://www.ctcms.nist.gov/potentials/

If you are new to this, then please see below. Here's my favorite example:

# citation: Beland, Lu, Osetskiy, Samolyuk, Caro, Wang, and Stoller (2016), Journal of Applied Physics, 119(8), 085901. DOI: 10.1063/1.4942533. # Example LAMMPS pair_style, pair_coeff lines for Ni-Co potential with zbl hardening for collision cascade simulations # Assumes lengths in Angstroms # pair_style with zbl inner and outer cutoffs of 0.5 and 2.0 Angstroms, respectively pair_style hybrid/overlay zbl 0.5 2.0 eam/alloy # pair_coeff lines for atom types 1=Ni and 2=Co pair_coeff 1 1 zbl 28 28 pair_coeff 1 2 zbl 28 27 pair_coeff 2 2 zbl 27 27 pair_coeff * * eam/alloy NiCo-lammps-2014.alloy Ni Co

https://www.ctcms.nist.gov/potentials/entry/2016--Beland-L-K-Lu-C-Osetskiy-Y-N-et-al--Ni-Co/

Misha Urazaliev Lucas M. Hale

Thank you both. The contents were very useful.

My pleasure