Is it possible to fix an NPT or an NVT ensemble in non-poreiodic and shrink-wrapped condition?
If so, how?
Yes it is possible.
The boundary conditions are specified in the boundary command (https://lammps.sandia.gov/doc/boundary.html) while the NPT/NVT in the fix nvt command (https://lammps.sandia.gov/doc/fix_nh.html)
For NVT there are no restrictions. For npt, the restrictions are listed in the respective section of the fix nvt :
X, y, z cannot be barostatted if the associated dimension is not periodic.
Hope it helps
Evangelos
Στις Παρ, 6 Σεπ 2019 στις 7:11 μ.μ., ο/η Azmain Faek Islam <azmainfislam@…24…> έγραψε:
Is it possible to fix an NPT or an NVT ensemble in non-poreiodic and shrink-wrapped condition?
If so, how?
as just explained a few e-mails earlier, there is no such thing as an NVT or NPT ensemble without fully periodic boundary conditions. that is how those ensembles are defined.
that said, you can run fix nvt with non-periodic boundary conditions. however while it may make sense to use it for an initial equilibration procedure, you have to ask yourself what is the meaning coupling a system to a thermostat bath, when you have - for example - a droplet which isn’t in contact with anything to exchange energy with.
similarly, you can use fix npt, but it is only halfway meaningful when only coupled to periodic boundaries, and when simulating a solid. since otherwise you have free surfaces, that will relax by themselves and any residual pressure would not be due to external coupling but through surface tension, which would be intrinsic to the system.
axel.
Thank you for the clarification, Professor Kohlmeyer.
Hi Axel,
that said, you *can* run fix nvt with non-periodic boundary conditions.
however while it may make sense to use it for an initial equilibration
procedure, you have to ask yourself what is the meaning coupling a system
to a thermostat bath, when you have - for example - a droplet which isn't
in contact with anything to exchange energy with.
Why can't you have a closed (not isolated) system in an NVT ensemble? It could
be that the heat exchange is happening through the walls, or via an implicit
bath of particles that is assumed to pass through the walls?
Is NVT inconsistent with such an interpretation?
Thanks!
Vishnu
Hi Axel,
that said, you can run fix nvt with non-periodic boundary conditions.
however while it may make sense to use it for an initial equilibration
procedure, you have to ask yourself what is the meaning coupling a system
to a thermostat bath, when you have - for example - a droplet which isn’t
in contact with anything to exchange energy with.Why can’t you have a closed (not isolated) system in an NVT ensemble? It could
be that the heat exchange is happening through the walls, or via an implicit
bath of particles that is assumed to pass through the walls?
this is a thermalized system, but not an NVT ensemble since you don’t have a fixed volume. with a free surface you have effectively an infinite volume and the atoms can expand freely, thus thus lots of thermodynamic properties would need to be defined differently than for a fixed, confined volume. that kills calling this an NVT ensemble.
in addition one could argue that the thermalization im such a system us geometry dependent and not across all boundaries thus the heat flux with the embedded heat bath not comparable.
Is NVT inconsistent with such an interpretation?
yes. besides, the clean way of doing such a system would be to only do thermostat coupling to the bulk part of the wall and let the rest of the system equilibrate to the state defined by the explicit heat exchange through the direct contact with other simulated atoms.
axel.