[lammps-users] Turn off certain force calculations.

LAMMPS LAMMPS Mailing List Mirror
1

My simulation set up is as follows:

I have a fluid layer sandwiched between an upper and lower surface . My surface is modelled by a pairwise interaction (modified Buckingham) where the surface atoms have partial charges. Electrostatics are calculated using the USER-EWALD-N routine. I am looking do perform shear simulations where I simply assign a fixed velocity to the atoms in the upper and/or lower surface. My boundary conditions are periodic in the x&y direction and non-periodic fixed in z direction.

I am wondering whether it is possible to turn off the calculations of forces on the surface atoms with each other, including electrostatics. I am not interested in the dynamics or vibrations of the surface atoms. After I assign initial velocities to the surface atoms (e.g to shear in the x or y direction) I want the surface slab structure to remain 'fixed'. i.e just the initial surface slab configuration moving with a fixed velocity in a particular direction.

I would still want have the pairwise interaction of fluid/fluid, fluid/surface, and the electrostatics of fluid/fluid, fluid/surface.

I am hoping if I can turn off these surface/surface calculations I can speed up my simulations, as it I will not be spending a significant amount of time calculating the forces/ vibrations of surface atoms with each other, which is not important to the simulations that I am trying to do.

Sorry for the long question, but I was wanting to make it as clear as possible.

Thank you,
Michael.

2

You can turn-off pairwise calculation of forces using the neigh_modify exclude
command. You can turn off the effect of force on atoms, by
simply not integrating them in the group you give to fix nve/nvt/etc.

It makes no sense that I can follow to turn off the contribution of some
set of atoms to long-range electrostatics. If you want to
do this, just set their charge to 0.

Steve

3

You can turn-off pairwise calculation of forces using the neigh_modify exclude
command. You can turn off the effect of force on atoms, by
simply not integrating them in the group you give to fix nve/nvt/etc.

It makes no sense that I can follow to turn off the contribution of some
set of atoms to long-range electrostatics. If you want to
do this, just set their charge to 0.

moreover, it is unlikely to get a significant speedup.
the main time eater is likely the (plain) ewald summation,
and that will not take less time due to exclusions.

axel.