I am using the LAMMPS version 29Oct2020. I have a CoCrNi94-Al3Ti3 alloy which was created using a hybrid MD/MC simulation. I learned from a couple of other posts on this forum that for alloys, an eam/alloy interatomic potential that covers the interactions between all the metal atoms in the alloy should be used by assigning that single potential file as “pair_coeff ** file …”. I haven’t been able to find any potential file that covers the 5 metal atoms in the alloy that I am using. I have checked the online repositories but the best I could find was a ‘FeNiCrCoAl-heaweight.setfl’ file that covers 4 of the 5 metal atoms involved. Is there any other way to implement an interatomic potential that covers all the 5 metal atoms? Any help would be appreciated. Thank you.
If you cannot find a ready-to-use potential file in any of the databases, you need to search the literature if a parameterization exists by somebody and then teach yourself how to generate a potential file from the information provided. There is some example software (in fortran, I think) in the LAMMPS tools folder, but there should be others available online somewhere.
As usual in these matters, common sense suggests that you write the software to first reproduce an already existing potential and then verify it by comparing energies and properties computed with LAMMPS.
If no exact parameterization exists, you have to do the next step and teach yourself how to add missing parameterizations. For that you can follow the example of (many) publications describing the same process. It is not easy if you don’t have much experience in parameterization and/or programming.
The final option, if all of the above is not feasible or fails, would be to find a collaborator with experience in the field of parameterizing EAM alloy potentials and then do that step together.
Dear Rajesh
In addition to Axel’s comment, and in a first attempt, you can read in the documentation, it exists a tool for generating EAM aloys (https://docs.lammps.org/Tools.html#eamdb) : 10.4.9. eam database tool
The tools/eam_database directory contains a Fortran program that will generate EAM alloy setfl potential files for any combination of 16 elements: Cu, Ag, Au, Ni, Pd, Pt, Al, Pb, Fe, Mo, Ta, W, Mg, Co, Ti, Zr. The files can then be used with the pair_style eam/alloy command.
The tool is authored by Xiaowang Zhou (Sandia), xzhou at sandia.gov, and is based on his paper:
X. W. Zhou, R. A. Johnson, and H. N. G. Wadley, Phys. Rev. B, 69, 144113 (2004).
a coorected version is available on the NIST interatomic potential website : https://www.ctcms.nist.gov/potentials/entry/2004–Zhou-X-W-Johnson-R-A-Wadley-H-N-G–Al/
Notes: The file Zhou04_create_v2.f is an updated version of create.f modified by L.M. Hale (NIST) following advice from X.W. Zhou (Sandia National Laboratory). This version removes spurious fluctuations in the tabulated functions of the original potential files caused by single/double precision floating point number conflicts.
File(s): README_v2 Zhou04_create_v2.f EAM.input.Al EAM_code
Since Cr data are missing in Zhou EAM database tool, you can add them using the following reference :
Lin Z B, Johnson R A and Zhigilei L V 2008 Computational study of the generation of crystal defects in a bcc metal target
irradiated by short laser pulses Phys. Rev. B 77 214108
It does not mean the resulting potentials are adapted to your problem. And thus you should parametrize by ourself or with experienced collaborators a dedicated potential, as Axel suggests.
thanks for the note on the updated version of the EAM database tool. I will incorporate the changes into the current LAMMPS distribution.
Please note that the version on the NIST website uses some very old fortran constructs that are not supported with very new fortran compilers.
The version in the LAMMPS distribution has been “modernized” (somewhat), so that this is not an issue. So I’ll have to do a 3-way-merge procedure.
Should be released with the next patch release.
If there is a person here with good knowledge in both Fortran and Python, it would be extremely helpful to have a python version of this kind of code to generate EAM potentials.
It would make it much easier to a lot of younger researchers that have little to no experience in programming and modifying Fortran code, and especially very old style fortran code.
Dear Axel,
I confirm this will be very useful to have a Python version for younger scientists.
I tried to learn but being above 60 year old, I had some difficulties ;-).
Nevertheless, I will try in a couple of months with an internship.
For user interested including Rajesh, I have attached the EAM input file that contains the original 16 element parameters + Cr ones. This can be ran with the 2 versions of the Zhou’s eam tool.
You can add it too in the tools/eam_database directory of the next LAMMPS version.
Best
Pascal
It is important to note that the EAM database contains models trained for the pure metals only, not for the alloys. The alloy models are generated by applying simple mixing rules to generate function parameters for every pair of elements. The mixing rules work well for alloys composed of elements that are chemically very similar, but they do not work well for elements that are chemically different. In the case of CoCrNi94-Al3Ti3, the ground state structures of the pure elements are:
FCC: Ni, Al, Co*
BCC: Cr
HCP: Ti,Co*
*Co groundstate is temperature-dependent
It is unlikely that the mixing rules will work well for such a diverse set of elements.