Dear all
Is there any way to equilibrate a phase, like water, under NPT ensemble while not allowing particles to cross a given periodic direction, say z.
I’d like to equilibrate a water box in NPT and then add a vacuum space over it to obtain surface tension.
Any hint is welcome
Dear all
Is there any way to equilibrate a phase, like water, under NPT
ensemble while not allowing particles to cross a given periodic
direction, say z.�I'd like to equilibrate a water box in NPT and then add a vacuum space
over it to obtain surface tension.
It is maybe too trivial for a molecular system like water, but why not
first equilibrating the system with PBC in all directions, and then
removing the periodicity in one direction and re-equilibrating to system?
Julien
Dear all
Is there any way to equilibrate a phase, like water, under NPT ensemble while not allowing particles to cross a given periodic direction, say z.
I'd like to equilibrate a water box in NPT and then add a vacuum space over it to obtain surface tension.
It is maybe too trivial for a molecular system like water, but why not first equilibrating the system with PBC in all directions, and then removing the periodicity in one direction and re-equilibrating to system?
why not start with the vacuum added right away and telling fix npt
only to change x and y? doesn't get any simpler.
axel.
Hey axel
I’ve two points on your reply:
(1) To achieve the desired density ( near experiment ), is it necessary to have an iso pressure coupling or only having x and y components of pressure will do the same?
(2) Suppose a general case, where we’d like first have a fluid equilibrated under NPT ensemble and then place it in a more complex simulation box. So, How do we should deal with some molecules crossing a periodic boundary?
Hey axel
I’ve two points on your reply:
(1) To achieve the desired density ( near experiment ), is it necessary to have an iso pressure coupling or only having x and y components of pressure will do the same?
you cannot have “iso” when you keep z uncoupled and unmodified. whether you couple x and y or not depends a bit on the system and material that you are simulating.
with an open surface, you will always get the density that is intrinsic to the water potential you choose.
(2) Suppose a general case, where we’d like first have a fluid equilibrated under NPT ensemble and then place it in a more complex simulation box. So, How do we should deal with some molecules crossing a periodic boundary?
a (full) npt equilibration is rarely needed in such case, as you will have to deal with the interfaces between the water chunk and the other parts. this is highly disruptive and will need a careful relaxation and equilibration protocol. i would just take an existing water box from some simulation software or molecule editor or similar, replicate it as needed and then delete the unwanted/overlapping molecules and then assemble the entire system. the pre-equilibration of the water is going to be the least of your problems.
axel.
why not start with the vacuum added right away and telling fix npt
only to change x and y? doesn't get any simpler.
axel.
Agreed.
> I'd like to equilibrate a water box in NPT and then add a vacuum space over it to obtain surface tension.
If this is your goal, I can't think of a reason to prevent atoms from
crossing the periodic boundary conditions.
As for how to create a slab of water with a vacuum, here is a list of
tools you can use to prepare a LAMMPS data file containing water and
other molecules, in various geometric shapes:
Cheers
Andrew