Tersoff Potential Regarding

Hi Steve,

I have been making use of Tersoff Potential file in my simulation to simulate machining. I have six different type of atoms as the first 3 belongs to workpiece which is essentially single crystal silicon & the other 3 types of atoms are diamond which is tool. I used to specify the following:

pair_style tersoff

pair_coeff * * SiCGe.tersoff Si(D) Si(D) Si(D) C C C

I want to now model SiC-Silicon carbide as workpiece instead of silicon and is in this regard referred LAMMPS manual. Couldn’t get a hint how to specify it. Could you help in this regard?

Hi Steve,

I have been making use of Tersoff Potential file in my simulation to
simulate machining. I have six different type of atoms as the first 3
belongs to workpiece which is essentially single crystal silicon & the other
3 types of atoms are diamond which is tool. I used to specify the following:

pair_style tersoff

pair_coeff * * SiCGe.tersoff Si(D) Si(D) Si(D) C C C

I want to now model SiC-Silicon carbide as workpiece instead of silicon and
is in this regard referred LAMMPS manual. Couldn't get a hint how to specify
it. Could you help in this regard?

so you input a SiC crystal as input structure. where exactly is the
problem with that?

axel.

If you're asking how to change the pair coeff command to use different elements
from the potential file, then you need a valid potential file for that system
and to use the descriptors (e.g. Si, C, etc) given in that potential file. That
is simply input to LAMMPS.

Steve

Hi Axel,
Thanks for your response.

What i meant to ask is that how should i specify the potential function to distinguish between mono crystal Si & SiC as workpiece. Is the following correct in this context:

for silicon carbide as workpiece

pair_style tersoff
pair_coeff * * SiCGe.tersoff Si(D) Si(D) Si(D) C C C

& for single crystal silicon as Workpiece

pair_style tersoff
pair_coeff * * SiC.tersoff Si Si Si C C C

Regards,
Saurav

Hi Axel,
Thanks for your response.

What i meant to ask is that how should i specify the potential function to
distinguish between mono crystal Si & SiC as workpiece. Is the following
correct in this context:

you cannot have two different tersoff potentials, so you'll
have to validate that the SiC parametrization is valid for
both SiC and plain Si. this is normally the case.

note that the Si.tersoff file in the lammps distribution refers
to a specific tersoff parameterization for Si only, the original
one, referenced a Si(B) in the combined files.

for silicon carbide as workpiece

pair_style tersoff
pair_coeff * * SiCGe.tersoff Si(D) Si(D) Si(D) C C C

& for single crystal silicon as Workpiece

pair_style tersoff
pair_coeff * * SiC.tersoff Si Si Si C C C

have you compared the actual parameters?
particularly for the Si Si Si lines?

axel.

Hi Axel,
Thanks very much for your response. This was confusing but understood the following:

  1. For SiC : Sic.tersoff & SicGe.tersoff with Si(D) holds good.
  2. For Si : SiCGe.tersoff Si(B) will hold good.

Have following questions:
1.You have said that “SiC parametrization is valid for both SiC and plain Si”. Does that mean that if i specify potential function for Si & C using Arhart-Albe Potential than it is valid for both SiC & Si as workpiece. However, they specified different potentials for Si & SiC.

  1. How to specify interaction between Si(B) & C ?

Regards,
Saurav

Hi Axel,
Thanks very much for your response. This was confusing but understood the
following:

1. For SiC : Sic.tersoff & SicGe.tersoff with Si(D) holds good.
2. For Si : SiCGe.tersoff Si(B) will hold good.

that is not what i said. read the comments in the files, read the
papers they are referring to and then look at papers that cite
those papers and that will help you to determine, what has
proven to be good enough for what you have in mind. there is
no way to avoid using up some (virtual) shoeleather to get this
sorted out.

Have following questions:
1.You have said that "SiC parametrization is valid for both SiC and plain
Si". Does that mean that if i specify potential function for Si & C using

i didn't say that either. i said that it is often the case.
it is your job to validate that. not mine. you cannot give
in a paper as a reference for what you were doing "some
random dude on the internet and that i don't know said so".

Arhart-Albe Potential than it is valid for both SiC & Si as workpiece.
However, they specified different potentials for Si & SiC.

in lammps you have to have only one potential file,
but that file can have multiple parameters sets.
like SiGeC.tersoff does. the problem is: there are no
combination rules to derive mixed parameter sets.
so those mixed terms need to be parameterized, too.

2. How to specify interaction between Si(B) & C ?

you cannot, and you should not.

axel.