about Coulombic energy computation with coul/long+pppm vs coul/cut

LAMMPS LAMMPS Mailing List Mirror
1

Hi all,

I’m wondering how the Coulombic energy is computed when coul/long+pppm is used, does it just consist of ecoul+elong ?

I found the Coulombic energy using pppm+coul/long(ecoul+elong) with a large vacuum space is very smaller than
that by coul/cut with large cutoff(big than the box’s diagonal). And the ecoul part isn’t same.
BUT both of the force compulation is almost the same.

Is there anythin wrong with my computation ?

Thank you!
Yours,
Cun

2

Ecoul + elong is the total Coulombic energy.
LAMMPS just calculates it in 2 pieces. Ecoul
is the short-range portion.

Steve

3

Steve,
Thank you for your quick replay.

Yes,I know. The ecoul computed with brute force using coul/cut with cutoff large enough is accurate for a

non-periodic system, is it right? 
And why the total Coulombic energy computed by pppm+coul/long is more smaller than it

when the forces are the same ?
4

Hi all,

I'm wondering how the Coulombic energy is computed when coul/long+pppm is
used, does it just consist of ecoul+elong ?

I found the Coulombic energy using pppm+coul/long(ecoul+elong) with a large
vacuum space is very smaller than
that by coul/cut with large cutoff(big than the box's diagonal). And the
ecoul part isn't same.
BUT both of the force compulation is almost the same.

Is there anythin wrong with my computation ?

have you checked the convergence of your coul/cut energy
with respect to the cutoff? i would suspect that it will
still oscillate a lot. the coulomb energy dies off very
slowly, specifically on an ordered system.

forces converge much faster.

axel.