pair_style none used in conjunction with pair_style hybrid/overlay

LAMMPS LAMMPS Mailing List Mirror
1

Hello,

Can “pair_style none” be used to “reset” all previously set pair_styles between atoms i,j with pair_style hybrid/overlay so that new pair_styles can subsequently be applied?

For example, suppose I specify a custom pair_potential (3spn2) between sites using Pair Coeffs in a file read by read_data like so:

1 3spn2 0.184000 4.500000 2.250000 6.000000 -0.994563
2 3spn2 0.184000 6.200000 3.100000 4.500000 0.781250
3 3spn2 0.184000 5.400000 3.120000 4.500000 0.547323
4 3spn2 0.184000 7.100000 3.860000 4.500000 0.468451
5 3spn2 0.184000 4.800000 3.100000 4.500000 0.480402
6 3spn2 0.184000 6.900000 3.520000 4.500000 0.616635
7 3spn2 0.184000 5.400000 3.120000 4.500000 0.547323
8 3spn2 0.184000 7.100000 3.860000 4.500000 0.468451
9 3spn2 0.184000 4.800000 3.100000 4.500000 0.480402
10 3spn2 0.184000 6.900000 3.520000 4.500000 0.616635
11 3spn2 0.184000 5.400000 3.120000 4.500000 0.547323
12 3spn2 0.184000 7.100000 3.860000 4.500000 0.468451
13 3spn2 0.184000 4.800000 3.100000 4.500000 0.480402
14 3spn2 0.184000 6.900000 3.520000 4.500000 0.616635
15 3spn2 0.184000 6.900000 3.520000 4.500000 0.0
16 3spn2 0.184000 6.900000 3.520000 4.500000 0.0
17 3spn2 0.184000 6.900000 3.520000 4.500000 0.0

This potential would normally be applied applied to all sites using mixing rules. I would like to turn off a subset of these pair_style interactions and then specify new pair styles between this subset of sites.

This is what I am currently using:

pair_coeff 1 117 none
pair_coeff 217 15*17 none

pair_coeff 1 1517 coul/long
pair_coeff 1517 15*17 coul/long

pair_coeff 1 1 coul/diel 78. 6.86 0.5

…

Thank you,

Dan

2

Pair style and pair coeff commands are like any others in LAMMPS,
you can re-specify them, e.g. between runs, to change whatever
you want. If you re-specify the pair style command, all previous
coeff settings are wiped out, since they are stored inside the previous
pair style. The hybrid/overlay pair style is a bit unique in that
it is OK to specify multiple coeff commands for one pair of atom types.
So if you want to wipe out all previous settings for types 1,2, you
would do pair_coeff 1 2 none, and then assign the new ones. This
is documented on the pair hybrid doc page. If you are doing complex
changes, it is probably safest to re-specify the pair style, and re-define
all the coeff commands.

Steve