Alloys

LAMMPS LAMMPS Mailing List Mirror
1

Hi my friends!
I want to simulate an alloy (base is Cu and then different percent of Al will be added, for example: (%100 Cu_%0Al ),(%90Cu_%10Al),…,(%0Cu_%100Al)).

I want to use “Mendelev_Cu2_2012.eam.fs” for Cu & “Al99.eam.alloy” for Al as force field,too.

How can I use set command for my target?

pair_style eam/fs
pair_coeff * * Mendelev_Cu2_2012.eam.fs Cu

pair_style eam/alloy
pair_coeff * * Al99.eam.alloy Al

Please help and guide me…

So thanks,
Nili

2

Dear Nili,

I believe you can use pair_style hybrid for this purpose. http://lammps.sandia.gov/doc/pair_hybrid.html
I imagine something like (suppose atom type 1 = CU, type 2 =Al)

pair_style hybrid eam/fs eam/alloy SOMETHING
pair_coeff * * eam/fs Mendelev_Cu2_2012.eam.fs Cu NULL # Cu-Cu interaction

pair_coeff * * eam/alloy Al99.eam.alloy NULL Al # Al-Al interaction

pair_coeff 1 2 SOMETHING ? # what is Cu-Al interaction ?

FYI, there is also a tool in lammps package called eam database (http://lammps.sandia.gov/doc/Section_tools.html#eam-database-tool) which create eam/alloy out of eam for single element, if it is what you need.

Cheers,
Trung

3

Dear Nili,

I believe you can use pair_style hybrid for this purpose.
LAMMPS Molecular Dynamics Simulator
I imagine something like (suppose atom type 1 = CU, type 2 =Al)

pair_style hybrid eam/fs eam/alloy SOMETHING
pair_coeff * * eam/fs Mendelev_Cu2_2012.eam.fs Cu NULL
# Cu-Cu interaction
pair_coeff * * eam/alloy Al99.eam.alloy NULL Al
    # Al-Al interaction
pair_coeff 1 2 SOMETHING ?
                # what is Cu-Al interaction ?

​this is not a good idea. this way the embedding term​ of EAM is not
computed for Cu-Al interactions. it would be somewhat acceptable to use a
hybrid pair styles, if there as an Al object and a separate Cu object, but
for an alloy, you need to have the mixed terms. There are several such
potential files provided even in the LAMMPS potentials folder.

axel.