Pair_write with Coulomb potential

Dear all.

  I want to generate a potential which is a composition of Buckingham, Coulomb and ZBL potentials. I am checking the potential using pair_write command, but there is a discrepancy with Coulomb potential (coul/long).

  Part of the commands is:

set group o charge ${qo}
pair_style hybrid/overlay coul/long 10.0
pair_coeff * * coul/long

pair_write 1 1 10000 r 0.1 10.0 fits_zbl.r OO \{qo\} {qo}

However, when I plot the pair_write values comparing with Coulomb potential values,

Vc = C * qO * qO / r

where C is the energy-conversion constant in metal units (~ 14.39), I found a big difference. The plot is attached. For Buckingham potential, I did not see any difference when I compare the potential function values with LAMMPS tabulated ones.

  Can anyone figure out what I am missing?

  Thanks,

  Pedro

pot_OO_zbl.png

Do you get a different answer when you don’t use pair hybrid (just coul/long)?
What units are you using in your input script when you produce the
plot for coul/long and for Buckingham?

Steve

Hi, Steve.

  There is a figure with 3 plots: (1) Coulomb potential function; (2) one generated by "pair_style hybrid/overlay coul/long 10.0" and (3) one by "pair_style coul/long 10.0". The last two present differences for distances lower than 2 angstrons.

  All results are in metal units.

  Thanks,

  Pedro

Citando Steve Plimpton <[email protected]>:

pot_OO_zbl.png