In the LAMMPS documentation for pair_eam (https://lammps.sandia.gov/doc/pair_eam.html) it states that "for the alloy case LAMMPS mixes the single-element potentials to produce alloy potentials..."
Does this mean there is a mixing technique specifically for EAM that can generate alloy EAM parameters from multiple single element (e.g. A-A, B-B) EAM parameter sets? Somehow avoiding the need to parameterize EAM from experiments/DFT for the alloy (A-B) interactions?
If this is true:
- Is there any documentation/literature I can read on how it accomplishes this? Is it just simple averaging?
- Is there a similar mixing technique for MEAM?
Regards,
Luis Costa
In the LAMMPS documentation for pair_eam (https://lammps.sandia.gov/doc/pair_eam.html) it states that “for the alloy case LAMMPS mixes the single-element potentials to produce alloy potentials…”
Does this mean there is a mixing technique specifically for EAM that can generate alloy EAM parameters from multiple single element (e.g. A-A, B-B) EAM parameter sets? Somehow avoiding the need to parameterize EAM from experiments/DFT for the alloy (A-B) interactions?
as explained in the documentation for eam and eam/alloy:
while you may use some form of mixing to determine A-B interactions for the simple EAM potentials, the eam/alloy format allows for more flexible and thus accurate fitting of potential data.
i don’t know much about the exact implementation in pair style eam, but as for mixing per-element data, to generate eam/alloy potentials, you may want to look at the program in the tools/eam_database folder and the associated publication mentioned in the README file.
If this is true:
- Is there any documentation/literature I can read on how it accomplishes this? Is it just simple averaging?
since the LAMMPS documentation explicitly specifies that the eam pair style behaves like the DYNAMO software as far as inferring the mixed terms is concerned, i would look up the corresponding publications and documentation.
- Is there a similar mixing technique for MEAM?
again, i am not an expert on that, but from the documentation it appears, that there is some general database of “generic” potential parameters provided, but that - same as in the EAM case - those have limited accuracy without additional modifications, which are applied through the second input file and modify the parameters for specific pairs of atom types to improve the accuracy for alloyed compounds.
axel.