how does lammps deal with the LJ and Coulomb potential at the cutoff

Dear LAMMPS developers,

At the cutoff, are the LJ and Coulomb potential by default shifted to zero by subtracting the value at the cut-off to ensure that the energy is the integral of the force? or there are small forces at the cut-off?

Thanks,

Jibao

Dear LAMMPS developers,

At the cutoff, are the LJ and Coulomb potential by default shifted to zero

no. that statement is not correct. by default both are truncated.

by subtracting the value at the cut-off to ensure that the energy is the
integral of the force? or there are small forces at the cut-off?

the shifting of the potential only affects energies, *not* forces.
even with a shifted potential, you have a discontinuity in the forces.
however, that one is usually very small for typical LJ (4*sigma)
cutoffs and with long-range electrostatics, coulomb interactions are
almost exact.

if you want smooth truncation of forces without a long-range solver,
you need to use a pair style that does that, e.g. the charmm variants.

axel.