Mixing EAM potentilas to create alloy potential in LAMMPS


I am facing a problem in making a alloy EAM potential for Ni-Co by using the individual element EAM potentials from NIST database, and merging them using averaging of phi interactions

I have attached the 2 individual EAM potential files. For combining them, I copy pasted them one below the another. Another thing that I kept in mind was that the Nr phi terms were removed from their place in individual potentials and added at last along with phi 2-1 calculated by averaging in excel.

One of the problems is that in this setfl format, we are allowed to specify Npho, dpho, Nr and dr only once in the beginning. They will be same for all the elements present.

But, the individual files for Ni and Co that are available on NIST have different values of these distances dr and dpho. In these case how to mix these potentials?

For time being I just used the values of Co in the file, but I am getting errors ( lost atoms) while relaxing the alloy structure. I think this is a bad potential that I have created.

Can somebody please guide me on mixing EAM potentials to create alloy potential?

Co_PurjaPun_2012.eam.alloy (670 KB)

Ni1_Mendelev_2012.eam.fs (745 KB)

merging potentials is not a kind of a black magic… I’ll be surprise that a “simple” copy-paste + averaging will ever give you a decent physical result! most probably Do you have a paper describing such a procedure? Then follow it. Btw, generating new potentials is a (very) complex task, that can take years! This is not something that has to be on a mailing list. Discuss with your supervisor(s), teacher or colleagues who have already an experience in this task. Julien.