The large majority of EAM potentials for Fe suffer mostly from the inability to correctly predict the Peierls barrier shape for dislocation motion (with an exception of Chamati et al. and Proville et al.). However, the best answer for your question would to conduct preliminary research on all of these potentials. The following tests can serve as good starting point to verify with published data (MD, DFT and experiment):

a) The Generalized stacking fault energy curves

b)The differential displacement maps for the dislocations

c) The grain boundary structure and energy (Mendelev et al. may be better for this purpose)

The large majority of EAM potentials for Fe suffer mostly from the inability to correctly predict the Peierls barrier shape for dislocation motion (with an exception of Chamati et al. and Proville et al.). However, the best answer for your question would to conduct preliminary research on all of these potentials. The following tests can serve as good starting point to verify with published data (MD, DFT and experiment):

a) The Generalized stacking fault energy curves

b)The differential displacement maps for the dislocations

c) The grain boundary structure and energy (Mendelev et al. may be better for this purpose)

In this paper they use EAM:

http://dx.doi.org/10.1016/j.msea.2011.03.006

You may also try ReaxFF, used here for iron oxidation:

http://www.sciencedirect.com/science/article/pii/S0167577X11007944

Hope this is of some use,

Fedor

Bsrao Nphy - Thanks. But When I checked the cohesive energy of the system done with Fe_2.eam.fs and Fe_5.eam.fs (Mikhail Mendelev's potentials) it doesn't match with the theoretical and experimental results. I got the following,

Fe_2.eam.fs 

                      Ecoh=-4.1224 eV  a=2.8553 A

Fe_5.eam.fs

                      Ecoh=-4.1341 eV a=2.8553 A

Thanks a lot Ilaksh Adlakha. Do you have the potential files of Chamati et al. and Proville et al. as I couldn't find on internet. It will be so helpful.

I'm working on bcc metals so, SF test is not possible I think. I will carry out other two tests. Is it enough to verify it by comparing the values of 'cohesive energy and elastic constant' from the experiment or literature ?

You can obtain the tabulations of embedding, pariwise andelectron density functions for Chamati et al. from the NIST website (refer the link below). I am not sure if the Proville et al. EAM is openly available, might want to ask the authors. 

 I don't understand your reasoning why generalized stacking fault curves are not possible for BCC metals? There is plenty of literature available for these stacking fault energy in Fe.

In my experience just cohesive energy and elastic constant are not enough to verify especially for BCC metals. In addition to the other suggestions, you could verify crack tip events for Fe single crystal crack simulation against available literature. 

Please let me know, if you need any assistance.

Regards

-Ilaksh

http://www.ctcms.nist.gov/~cbecker/Fe.html