How to verify the potential used?

Hi,

In my simulation I have considered the

atom type 1 and 2 as carbon &
atom type 3 as silicon

I would like to consider the following potentials between the atom types.

1 and 2 - only airebo potential
1 and 3 - lj/cut and tersoff potential
2 and 3 - lj/cut and tersoff potential

Here is the part of my input script considering interactions between atoms.

pair_style hybrid/overlay tersoff airebo 3.0 lj/cut 15.0
pair_coeff * * tersoff SiC.tersoff C C Si
pair_coeff 1 1 none
pair_coeff 1 2 none

pair_coeff 2 2 none

pair_coeff * * airebo CH.airebo C C NULL
pair_coeff 1*2 3 lj 0.784727 2.74755 8.24

  • Do the format of my script matches my description about the interaction potential?

** If so, how to verify (for example) that only tersoff potential is considered for atom type 1 and 2. Since I want to make sure that two many potentials are not used for the atom type 1 and 2.

*** Is there any command in LAMMPS that output the interactions between the atoms and potential used for them?

Thanks in advance.
With regards,
Ramki.

Hi,

In my simulation I have considered the

atom type 1 and 2 as carbon &
atom type 3 as silicon

I would like to consider the following potentials between the atom types.

1 and 2 - only airebo potential
1 and 3 - lj/cut and tersoff potential
2 and 3 - lj/cut and tersoff potential

this cannot work. tersoff is not a pairwise additive potential.
neither is airebo. they cannot be selected for individual pairs but
only for entire sets of atoms.
also, overlaying tesoff with lj/cut double counts the non-bonded
interactions since tersoff already includes that.

Here is the part of my input script considering interactions between atoms.

pair_style hybrid/overlay tersoff airebo 3.0 lj/cut 15.0
pair_coeff * * tersoff SiC.tersoff C C Si

this will set up tersoff between *all* types.

pair_coeff 1 1 none
pair_coeff 1 2 none
pair_coeff 2 2 none

these three lines have no effect with hybrid/overlay. and since AIREBO
is not a pairwise additive potential, you cannot "blank out"
individual pairs (there are also interactions between triples and
quadruples, but this list will be incomplete and inconsistent, if you
do something like you appear to be wanting to do.

pair_coeff * * airebo CH.airebo C C NULL
pair_coeff 1*2 3 lj 0.784727 2.74755 8.24

* Do the format of my script matches my description about the interaction
potential?

no.

** If so, how to verify (for example) that only tersoff potential is
considered for atom type 1 and 2. Since I want to make sure that two many
potentials are not used for the atom type 1 and 2.

build a test system where you can compute the interactions manually

*** Is there any command in LAMMPS that output the interactions between the
atoms and potential used for them?

you can use pair_write, but that only works for pairwise additive
potentials. in total, your main problem is that you want to treat a
manybody potential like a pairwise additive potential. secondly, you
cannot simply mix and match different parameterizations and expect
consistent behavior.

axel.

Remove the pair coeff none commands; they aren’t needed.

Otherwise it looks fine. You can “debug” this yourself,

by re-running with different styles turned on or off and monitoring

the energy/forces in the system. You can also use the compute pair

command to extract the energy components of different sub-styles

of the pair hybrid command.

Steve