Hi all,
When I use kspace = ewald/n, could the cutoff to do real space calculation be larger than half of my box size?
yes - I think the cutoff can be whatever you want. However,
the code may run slow if you choose a long cutoff.
Steve
Penghao:
In principle, the code may allow this, but it’s not necessarily a smart idea to do so. Why do you want to do this?
–AEI
Penghao:
In principle, the code may allow this, but it’s not necessarily a smart idea to do so. Why do you want to do this?
–AEI
Hi Ahmed,
I want to use a very small box which contains only 8 atoms using “lj/coul long long” potential. But I found if I increase cutoff, total energy will increase and finally converge to some value. This converged value consists with the result of larger box with more atoms.
One more question: For short range interaction, usually minimum image convention is employed, which requires cutoff within half of the box.Does LAMMPS requires this for lj/cut potential?
There is no restriction in LAMMPS that the cutoff be less than
the box length or half the box length. If your system is periodic,
you should get the same answer (for a snapshot of energy, forces,
pressure, etc) when using a specific cutoff for any size periodic box,
assuming you have an identical config in the small vs large box.
Steve