[lammps-users] Need help implementing mixed (cross) EAM-FS potential for Fe-Cr in LAMMPS


I am trying to write a Dynamo setfl file containing three EAM Finnis-Sinclair (FS) type potentials for an Fe-Cr alloy. The three potentials are for Fe-Fe, Cr-Cr, and the Fe-Cr cross-interactions.

I have a specific potential form for Fe-Cr and as such do Not want LAMMPS to mix the other two potentials for me to create a third Fe-Cr cross-potential.

I can get LAMMPS to work with just the Fe-Fe potential, but I get errors when introducing the cross-potential (for Fe-Cr). Looking through messages in the LAMMPS mail list archives, I believe that I should accomplish this by placing all three potential in the same Dynamo setfl file.

Looking over the “read_file” function in “pair_eam_fs.cpp” file of the LAMMPS source code, I believe I understand how to implement EAM-FS potentials for Fe-Fe and Cr-Cr interactions, but I don’t know how to do it for the Fe-Cr cross potential.

If I’m right, the Dynamo setfl file format should follow this structure (Let i = for element i):

Line 6: [Atomic Number(i)] [Atomic Mass(i)] [Lattice Constant(i)] [Lattice Type(i)]
Line 7 to A: [Embedding Function(i) = F(rho(i))]
Line A+1 to B: [Electron Density(i) = rho(i)]

Line B+1: [Atomic Number(j)] [Atomic Mass(j)] [Lattice Constant(j)] [Lattice Type(j)]
Line B+2 to C: [Embedding Function(j) = F(rho(j))]
Line C+1: [Electron Density(j) = rho(j)]

(Continue on for however many elements you have)

Line D: [Pair Potential between i and i]

Line E: [Pair Potential between i and j]

Line F: [Pair Potential between i and k]

(Continue on for however many elements you have)

Line G: [Pair Potential between n and n]

If I am right, there seems to be nowhere to specify the electron density function for the cross potential. If I am wrong…I’d love to know where I’m wrong.

Any help you can offer me is greatly appreciated!

-Brian Frisbie

If you use a setfl file, LAMMPS does no mixing, since the cross
interactions are in the file. The pair_style eam doc page
describes the format of the setfl file. That and the source
code are all there is.


Thanks for the reference,

I’ve read through that document as well and I believe it agrees with my description of the Dynamo setfl file format in my prior email.

My Problem: I know it is possible to list/specify cross interactions for the Pair Potential, however it does NOT appear possible to list/specify cross interaction for the Density Function, rho®.

I need to specify cross interactions for the Density Function, rho®. Can I do this using LAMMPS? It does not seem possible using the setfl format. I’ve copied the relevant section from the “pair_style eam doc” page to the end of this email for reference.

Thank you again for your suggestions and for getting back to me so quickly.

-Brian Frisbie

I think you want the Finnis-Sinclair format, which is eam/fs in LAMMPS.
It allows a pairwise rho() definition. The eal/alloy (setfl) format
use an element-wise rho().


You’re right! That solves my problem.

Thank you so much for your help!

-Brian Frisbie