Soft-sphere potential

LAMMPS LAMMPS Mailing List Mirror
1

Dear LAMMPS users,

Does anyone know how to implement soft-sphere potential

U® = 4 e ( sigma /r )^12 ; r < rc
= 0 ; r > rc

Thanks in advance
Karteek

2

It is just a variation of the L-J 12-6 potential. You just need to turn off the ^6 term (comment it out in the src).

Ray

3

Thanks,
you want me to edit the source file.

Thanks
Karteek

4

Hello,

"comment it out in the src" suggests to go in the lammps directory, to edit the
file pair_lj_cut.cpp. Everywhere the r^{-6} terms appear, you should delete it.

You should be careful to modify the force computation, the energy computation,
the offset. From reading the file, the simplest seems to set "lj2[i][j]" and
"lj4[i][j]" to 0 in PairLJCut::init_one and delete "- pow(ratio,6.0)" from
"offset[i][j] = 4.0 * epsilon[i][j] * (pow(ratio,12.0) - pow(ratio,6.0));" This
still does not take care of the tail correction, a few more edits are needed.

Do not forget to recompile lammps afterwards.

Regards,

Pierre

5

is there a specific reason you cannot use WCA, i.e. the standard 12-6
LJ with the cutoff set at sigma*2**1/6?
for all practical purposes, it should be nearly the same.

of course, you can also always use tabulation, but that requires
careful testing of the tables...

axel.

6

Thanks Axel,

I will be using

pair_style lj/cut sigma*2**1/6
pair_modify shift yes

Thanks again
Karteek