[lammps-users] Meam potential

Hi

I have a couple of questions regarding MEAM potential.

  1. MEAM potential is said to consider angular terms in atom-atom interactions to account for orientation of electron cloud in the direction of covalent bonds. So, for simulating SiC, do I have to define bonds, angles, and bond energies if I am using meam potential with the library.meam and SiC.meam files included with the LAMMPS package? Or does the definition of SiC interactions in meam package consider all these required features of the covalent bond?

  2. If I have to consider a 3rd element (X) interaction with SiC, is there a simple way to do so using the meam files for SiC and for the third element or do I have to use something more complicated? In other words, can the meam package use some kind of rule of mixing to evaluate the interactions?

Any help will be deeply appreciated.

Rohit

Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
boundary="----_=_NextPart_001_01CA2741.9108990C"

Hi

I have a couple of questions regarding MEAM potential.

  1. MEAM potential is said to consider angular terms in atom-atom interactions to account for orientation of electron cloud in the direction of covalent bonds.

Aside: angular terms in MEAM are as much associated crystal symmetry as with bond directionality.

So, for simulating SiC, do I have to define bonds, angles, and bond energies if I am using meam potential with the library.meam and SiC.meam files included with the LAMMPS package? Or does the definition of SiC interactions in meam package consider all these required features of the covalent bond?

the MEAM force routine uses cartesian representations of the angular interactions. you do not have to express them explicitly. The optimum geometry of SiC is implicit in the parameters.

  1. If I have to consider a 3rd element (X) interaction with SiC, is there a simple way to do so using the meam files for SiC and for the third element or do I have to use something more complicated? In other words, can the meam package use some kind of rule of mixing to evaluate the interactions?

You only need interactions of each element with itself and interactions between each pair of elements. You should be able to construct all of the input files for the ternary in analogy with SiC.

Steve

Steven M. Valone
MST-8, MS G755
Los Alamos National Laboratory
Los Alamos, NM 87545
smv@…795…
office: (505)667-2067; fax: (505)667-8021

License Exception TSPA/Correspondence

The bond, angle, etc potentials defined by the bond_style, angle_style, etc
commands in LAMMPS are for permanent bonds, angles, etc. Potentials
like MEAM, REBO, Tersoff, that create/destroy "bonds" dynamically are
all treated as pair styles by LAMMPS, even though they may compute 3-, 4-body,
etc interactions. So no, you should not define a bond_style if you are using
MEAM. Nor would you list such bonds in your input data file.

Steve