EXTERNAL: Re: ZBL funtionality for all pair styles

LAMMPS LAMMPS Mailing List Mirror
#1

please *always* copy the mailing list when replying. thanks.

Thanx Alex. I hadn't thought of it that way. Probably a peaked (average of two ff) function at the cut off. Just modify the LJ pair style code using the peaked function. Cool.

I've be either taking the results with an immense grain of salt and using LJ and very small time steps when I don't have "better" ff. Is there a pair style ZBL for lammps (not contaminated with other potentials) ?

yes, i already mentioned that: http://lammps.sandia.gov/doc/pair_zbl.html

axel.

#2

Thanx again for your time, I found it in the latest distribution it doesn't seem accessible from the main doc html. Confused on the Z must be the same for I and J. Do I and J have to be the same or can I get around it by sqrt(ZI*ZJ) Sorry, I need to play with it a bit before asking intelligent questions.

#3

I think you are misreading the docs.
Axel.

#4

Hi M. Kohlmeyer,
I have written to this group recently about my attempt to combine ZBL with Stillinger-Weber with a spline function between them. Would it be possible to create a potential file ZBL/sw (or hybrid/zbl) that could allow users to combine ZBL with other potentials than Tersoff ? Thank you,

Paule

#5

I found it in the latest distribution it doesn’t seem accessible from the main doc html.

It’s listed here, in the pair style section:
http://lammps.sandia.gov/doc/Section_commands.html#cmd_5

#6

Hi M. Kohlmeyer,
I have written to this group recently about my attempt to combine ZBL with
Stillinger-Weber with a spline function between them. Would it be possible
to create a potential file ZBL/sw (or hybrid/zbl) that could allow users to
combine ZBL with other potentials than Tersoff ? Thank you,i

i don't understand your question. i already recommended a much easier
to implement and cleaner solution to this which does not require
hybrid/overlay to boot. you can implement pretty much any pairwise
potential and it is embarrassingly easy to do. the majority of the
pair styles in LAMMPS differ only in a few lines of their respective
Pair::compute() method and how they store the potential parameters.
replacing the twobody term in the sw style is even easier.

axel.

#7

http://lammps.sandia.gov/doc/pair_style.html

#8

Thanx again for your time, I found it in the latest distribution it doesn't seem accessible from the main doc html. Confused on the Z must be the same for I and J.

no, it doesn't. the docs only say, that you have to provide parameters
for all permutations of types, but only the "self" term is used, since
the potential implicitly derives all other parameters from those.

so you could do something like this

pair_coeff 1 * ...
pair_coeff 2 * ...
pair_coeff 3 * ...

and you would be good.

axel.

Do I and J have to be the same or can I get around it by sqrt(ZI*ZJ)
Sorry, I need to play with it a bit before asking intelligent
questions.

#9

yes - we missed adding it there - just added it

thanks,
Steve