Pair_style hybrid problem

LAMMPS LAMMPS General Discussion
1

Hi, all
l’m going to carry out a simulation containing 6 elements which are Co, Cr, Fe, Ni, Al, Ti.
But there is no potential involve all 6 elements.
I only could find “CoCrFeNiAl.eam.alloy”.
I also found “CoTi.meam”、“VTiCr_IFS.eam.alloy”、“FeTi.meam”、“NiAlTi.meam”
I think maybe I can use “pair_style hybrid” command.
Here are two questions:

  1. Do I still need library.meam for hybrid?
  2. I have looked up the “pair_coeff” command, but I still don’t know how to write the commands to hybridize them successfully. Can anyone give me a hint ?
2

You cannot use pair style hybrid to combine multiple embedded atom potentials. It cannot compute the embedding energy across the different potentials and it cannot generate mixed pairwise potential terms.

This has been discussed many times before, you can find these discussions in the archives.

4

So…it could be “pair_style hybrid/overlay” ?

5

No. Same problem.

7

So the answer is there is no any way I can hybridize them.

8

I already explained why.

1 Like
9

@akohlmey I have also a similar doubt. I have been working with a system of four metals W, Fe, Ni and Co.
I have used the following comands to define the pair style and it was running. Will it end up inaccurate ?
pair_style hybrid meam eam/alloy
pair_coeff * * meam library_NiW.meam Ni W Fe Co NiW.meam Ni W Fe Co
pair_coeff * * meam library_FeNi.meam Ni Fe W Co NiFe.meam Ni Fe W Co
pair_coeff * * meam library_CoW.meam Co W Ni Fe CoW.meam Co W Ni Fe
pair_coeff * * meam library_Co_Fe.meam Co Fe Ni W CoFe.meam Co Fe Ni W
pair_coeff * * meam library_NiCo.meam Ni Co W Fe NiCo.meam Ni Co W Fe
pair_coeff * * eam/alloy FeW.eam.alloy Fe W NULL NULL

10

Yes.

But the problem is worse than a simple “hybridation issue”, there is no logic at all behind these 6 commands if you used them like that. For instance, the first command

pair_coeff * * meam library_NiW.meam Ni W Fe Co NiW.meam Ni W Fe Co

assumes that atom of type 1 is Ni, atom of type 2 is W, 3 is Fe, and 4 is Co. Since there are four atom types in total in the simulation, that should be enough to fully define your system. All the other commands attribute a different element to your atom types, so atom of type 1 is a Ni, and a Co, and a Fe, all at the same time.

Simon

1 Like
11

… and yet you do the same thing that I have been saying multiple times in this thread that should not be done and that cannot work correctly.

It will be complete and utter garbage.

1 Like