Fe-Cu-Ni ternary potentials. (Siddharth Sundararaman)

First, How wrong are your results? And what “file” are you using? (By file, I assume you mean a set of tabulated parameters). What potential? EAM?

As for ternary parameterizations: Not too many exist. Most potentials I know of (like EAM) only use simple binary mixing rules i.e. Fe-Cu EAM parameters are simple a geometric average of Fe-Fe and Cu-Cu parameters, etc. As a result they are usually only accurate for dilute solutions. The hetero-pairwise interactions can be fit independently to give better results. Again, even if you found a potential that was fit like this, it might not be fit for what you need it for. When you get into ternary systems life gets more complicated. Then you have to consider the effects of three body interactions Fe-Cu-Ni, Fe-Ni-Cu, Cu-Fe-Ni, etc… These may or may not be a significant source of error. Again dilute solutions may be OK, whereas high alloy concentrations probably are not.

You might look into modifying the eam/cd potential for your purposes (look up papers by Caro). It’s pretty neat in that it includes the heat of mixing in the potential fitting database and as such is more accurate for high alloy concentrations. I imagine if you had the data to fit to, this potential form might do a good job with ternary alloys.

Good luck, fitting potentials is no easy task.


J. David Schall, Assistant Professor
Dept. of Mechanical Engineering, Oakland University
130 DHE, Rochester, MI 48309