Dear all,
I have a nanoparticle and water as a nanofluid system (nanoparticle in
water) which is to be equilibrated at the targeted temperature. I read in
one of the paper that water is independently equilibrated using NPT
ensemble and PBC were imposed, while for nanoparticle NVE ensemble was used
and free boundary conditions was imposed and then berendson thermostat was
used to maintain the nanoparticle
you cannot have two ensembles in the same simulation. if you are using fix
npt, that does not automatically mean, you are running in an NPT ensemble,
and similarly when using fix nve, you are not automatically in an NVE
ensemble. and when you add a berendsen thermostat, you do manipulate the
kinetic energy to match a particular target temperature, but you are in no
defined statistical mechanical ensemble at all. similarly, when you mix
integrators and thermostats. if you think the authors of the paper you are
following claim this, it either means, you are not reading/understanding it
correctly, or it casts a serious doubt on their understanding of
statistical mechanics. also, you cannot have non-periodic and periodic
boundary conditions in the same system.
fix npt means, you are using a nose-hoover thermostat and a nose-hoover
barostat.
fix nve means, you are doing simple timestepping without any modification
of kinetic energy or volume.
fix temp/berendsen means, you are adjusting the kinetic energy according by
smooth velocity rescaling according to an exponential decay.
temperature to the targeted one. Once the water system and nanoparticle
were equilibrated separately, they were combined and then the system formed
is allowed for post processing. (No overlapping of water molecule and
naoparticle during combination as water system already has nanoparticle
cavity when it was equilibrated individually).
My query: What is wrong if I have a system combined with (water +
nanoparticle) and equilibration is done all together as a system rather
than doing it individually as done in the paper? Does it make any
difference?
that depends on the practical details and convenience. actually, when
joining separately equilibrated parts to form a compound system, you add a
huge disruption, that can cause as much issues as starting the compound
from scratch. if the equilibration after the joining is not done correctly,
you can have long-lasting disrupting phonons in your system (as with any
multi-component system, that is assembled this way).
using separate thermostatting for different subsystems can help reaching
equipartitioning of kinetic energy in the system faster as exchange of
kinetic energy between the two subsystems may be slow. but, e.g., using a
dissipative thermostat on the whole system can be used for the same purpose.
ultimately, equilibrium is equilibrium. ...and as such it is hard to
detect reliable whether you have reached it or not, regardless of the way
how you get there. thus people tend to do what they have the best
experiences with.
axel.