[lammps-users] Specifying pair style, and pair coeff for interactions at an interface

Dear Steve and Alex,

Regarding Narumanchi question, would it me sense to use MEAM for both Si and Al?

  • Javier

Dear Steve and Alex,

dear javier,

Regarding Narumanchi question, would it me sense to use MEAM
for both Si and Al?

that is really difficult to say without being an expert
in the respective materials and potentials. i suggest you
search the literature for studies from reputable sources,
that have tried using MEAM for describing a Si-Al interface.
in my _very_ limited experience, i would expect that a Si-Al
embedded atom potential would be parameterized towards
describing a Si-Al alloy, rather than a Si-Al interface.

all i was warning about is that you lose some of the
many-body interactions across boundaries when using pair_hybrid
with two separate/different many-body pair-styles.

you also have to consider what the approximations are, that
go into the respective models, how much properties that you
are interested in depend on the interface area, and what level
of accuracy you need altogether. all of this cannot be
answered in a general statement.

for some systems/problems, it may be required to use a more
sophisticated method, e.g. like density functional theory.
in other cases, a simple mechanical coupling may be fully
sufficient. LAMMPS will allow you to do a lot of things,
even those that are questionable, so you have to convince
yourself that a chosen model is adequate for a given problem.

cheers,
    axel.