# Born-potential

Hi，

Recently, I am doing some calculation with pair_style born-coul-long.
In LAMMPS, the form of born is , in front of D, the sign is plus. And in the examples of LAMMPS manual, “pair_coeff 1 1 6.08 0.317 2.340 24.18 11.51”, the D coefficient is given as 11.51, a positive number. I have analyzed that the parameters in this example are for NaCl systems with “real” units.
So, in LAMMPS, the sign in front of D is plus, and D is set as positive for NaCl.

Well, in the work of Fusi and Tomi (M. P. Tosi and F. G. Fumi, J. Phys. Chem. Solids, 1964, 25, 31) the potential form is:

and the corresponding parameters for molten NaCl are as follows:

i.e. in the reference, the sign in front of D is plus, but the D is negative. The reference is also cited by LAMMPS. In fact, in many other references, the sign in front of D is minus and the D is positive.

So I think that there is some contradiction between LAMMPS and the most references, right?

What’s more, when I calculate molten NaCl, if i set the parameters with metal units as follows (D is positive), I can do the calculation and reproduce the values of reference.

pair_style born/coul/long 10.0 8.0
pair_coeff 1 1 0.2640625 0.317 2.340 1.05 0.5
pair_coeff 1 2 0.21125 0.317 2.755 7 8.6875
pair_coeff 2 2 0.1584375 0.317 3.170 72.5 145.625
kspace_style ewald 1.0e-5

But if I set D as negative, thus the net effect of the last term is subtract the 8-order dispersion interaction, which is consistent with the reference, the calculation comes up with “nan”.

I don’t know why, I am confused about it and I can’t explain it.
Thanks very much.

Jia Wang

You’ve posted this message a million times. While waiting for Steve or any of the main developers to answer try to be a bit creative

about how to move forward. The pair_born.cpp file lists Sai Jayaraman as the author of such a package. A quick search on the web shows Sai listed as a group member of the Ceder group at MIT: http://ceder.mit.edu/people.php

Has to be the same guy 'cause his Linked in profile shows the same picture and identifies him as a former Sandia postdoc.

Carlos

You've posted this message a million times. While waiting for Steve or any
of the main developers to answer try to be a bit creative
about how to move forward. The pair_born.cpp file lists Sai Jayaraman as
the author of such a package. A quick search on the web shows Sai listed as
a group member of the Ceder group at MIT: http://ceder.mit.edu/people.php
Has to be the same guy 'cause his Linked in profile shows the same picture
and identifies him as a former Sandia postdoc.

more importantly, how about to make an effort to determine if it is
actually the initial configuration that is at fault (like it almost always
is in cases of NaN) and not the potential?

the born potential in LAMMPS has been around for quite a while and been
used by people successfully, so it is highly likely that it doesn't have
such a fundamental flaw that is insinuated in this (annoyingly) repeated
posting.

reposts of the same question without edits only irritate people and make it
much less likely to get help.

axel.

Yes, listen to Axel carefully who seem to have Coulombic ability (long-ranged) to sniff problems with people’s input files

Carlos