Dear Sir,
I want to simulate three types of atom, for example, with 1 2 3 corresponding to Al, Cu and carbon, respectively. When I used the following input commands, I got an error of “Incorrect args for pair coefficients”
pair_coeff * * airebo ./CH.airebo NULL NULL C
pair_coeff 12 12 eam/alloy ./AlCu.eam.alloy Al Cu NULL
When I changed them to:
pair_coeff * * airebo ./CH.airebo NULL NULL C
pair_coeff * * eam/alloy ./AlCu.eam.alloy Al Cu NULL
I could run it without error, though I am not sure if the atom types are assigned as I mean to.
According to the manual, these two kinds of methods seem to be the same, can you help me explain this? Is the second approach right or not?
Thanks you very much!
Best Regards,
Zenan Qi
College of Engineering
Boston University
8 Saint Marry Street
Boston, MA, 02215
U.S.A
Dear Sir,
I want to simulate three types of atom, for example, with 1 2
3 corresponding to Al, Cu and carbon, respectively. When I used the
following input commands, I got an error of "Incorrect args for pair
coefficients"
pair_coeff * * airebo ./CH.airebo NULL NULL C
pair_coeff 1*2 1*2 eam/alloy ./AlCu.eam.alloy Al Cu NULL
this is incorrect. eam/alloy is a manybody potential and _only_ works
with pair_coeff * *, no individual pairs or ranges may be specified.
When I changed them to:
pair_coeff * * airebo ./CH.airebo NULL NULL C
pair_coeff * * eam/alloy ./AlCu.eam.alloy Al Cu NULL
that is how it is supposed to be done.
however, you are still missing the non-bonded potential component
that connects C to both Al and CU.
I could run it without error, though I am not sure if the atom types are
assigned as I mean to.
how about testing this empirically? you could just
run two inputs that have each only one of the two potential types
and then compare the energies to an input with the hybrid definition
but otherwise the same coordinates and the same atom types.
According to the manual, these two kinds of methods seem to be the same, can
you help me explain this? Is the second approach right or not?
the second method of setting pair coefficients (with explicit numbers)
only works for pairwise additive potentials, like LJ.
axel.