Pair_style nb3b/screened

LAMMPS LAMMPS General Discussion
,
1

Dear LAMMPS users,

I am currently trying to implement a nb3b/screened pair_style but it seems that the lammps documentation has not been updated since I was unable to find what the format of the potential file should be. Therefore, I tried to do something similar to the nb3b/harmonic potential file but changing the last columns from cutoff to rho1 and rho2 (E=K*((theta-theta0)^2)*exp(-rij/rho1)*exp(-rik/rho2) :
#i j k K/2 theta0 rho1 rho2
S S S 0.00 0.00 1.0 1.0
S S Ge 0.00 0.00 1.0 1.0
S Ge Ge 10.0 90.0 1.0 1.0
S Ge S 0.00 0.00 1.0 1.0
Ge Ge Ge 0.00 0.00 1.0 1.0
Ge Ge S 0.00 0.00 1.0 1.00
Ge S S 40.0 109.00 1.0 1.0
Ge S Ge 0.00 0.00 1.0 1.0

Then, the simulation run but no angular correction is seen in my system compare to similar system but with the nb3b/harmonic pair_style.

If you have any Idea I would be very grateful !

2

The file format can be inferred by comparing the two example files in the “potentials” folder.

The file potentials/MOH.nb3b.harmonic has the entries:

[...]
# these entries are in LAMMPS "real" units:
#   k [kcal/mol/rad^2], theta_0 [degrees], cutoff [Angstroms]
#
# format of a single entry (one or more lines):
#   element 1 (central atom), element 2, element 3, 
#   k_theta, theta_0, cutoff
#
# (ijj)-->determines cutoff for i-j distance
# (jik)-->determines k, theta_0 for i-j-k angle term
#
# ielem jelem kelem     k     theta_0 cutoff
   O     O     O        0.00    0.00    0.00
   O     O     M        0.00    0.00    0.00
   O     O     H        0.00    0.00    0.00
 [...]

While the file potentials/PSiO.nb3b.screened has the entries:

[...]
# These entries are in LAMMPS "metal" units:
#
# Format of a single entry (one or more lines):
#   element 1 (central atom), element 2, element 3,
#   k_theta, theta_0, rho, cutoff
#
# (ijj)-->determines cutoff for i-j distance
# (jik)-->determines k, theta_0 for i-j-k angle term
#
# i  j  k   K     theta0   rho  cutoff
O   P   P    32.5  109.47   1.0   3.3
O   P   Si   60.0  109.47   1.0   0.000
O   P   O   0.000  0.000    1.0   0.000
O   Si  P    60.0  109.47   1.0   0.000
[...]

So your modification is apparently incorrect, since there is only one entry for \rho and your file is incorrectly setting the cutoff.

1 Like
3

Thank you very much for your prompt answer!
This answer my question about the file format and also the change of units between harmonic and screened pair_style wich was also a problem I had.
So, to make sure that I have understood the nb3b/screened pair style well, could you confirm that it basically behaves like the harmonic one, but instead of having a “brutal” cutoff of the harmonic potential, it is done more smoothly thanks to the screening of the exponential?
Thanks again.

4

The pair style does what the documentation says it does. I am not going to give you a digested version of that here. If you do not trust the documentation, feel free to set up some test inputs where you can compute the results yourself ‘by hand’.