I am new to lammps and have been trying to improve. I wrote a script
intending to use both meam and tersoff: meam for Si-Si and Si-C while
tersoff for C-C interactions respectively. My script goes thus:-
you are new to LAMMPS and yet, you are trying to set up a very complex and
tricky input? how does this fit together? what happened to first learning
how to do simple set ups properly and correctly before moving on.
even more so, since the kind of force field combination you want to is is a
*bad* idea. both mean and tersoff and manybody potentials, which are _not_
pairwise additive, and thus cannot be decomposed the way you intend to do.
in fact, they require atom types to be "wholly contained", i.e. if you
include atom type 1 and atom type 2, you also have to include interactions
between type 1 and 2. furthermore, this is not limited to immediate pairs,
but also extends to atom triples and beyond (depending on the pair styles),
which are inferred from pairs.
so you have to make a choice: either you use MEAM for everything or
Tersoff. even if you sort out the syntax issues, any mixing the way you
proposed is going to result in inconsistent interactions and bogus
mind you, this topic is repeatedly and extensively discussed on this very
mailing list, so you are advised to look up old conversations on the same
topic and learn from them.
pair_style hybrid meam tersoff
pair_coeff * * library.meam Si C SiC.meam Si Si Si C
this is not following the correct syntax for pair_coeff with pair style
hybrid; it is missing the designation of the sub-style.
pair_coeff * * tersoff SiC.tersoff NULL C NULL NULL
on top of being not meaningful, as explained above, this is also
inconsistent with the previous line.
but kept getting error message (ERROR: Pair coeff for hybrid has invalid
style (../pair_hybrid.cpp:396)). Please tell me what am missing.
what you are missing is, that you haven't been reading the documentation
carefully enough and are not following it.