pressure equillibration

Hi,

I have a doubt.

I am doing an equillibration run for a liquid system. For that I have performed the following steps:

nve (with T rescaling for 50 ps) ==> nvt (100 ps) ==> npt (500 ps) ==> nvt (500 ps)

In “npt” : the Target Pressure = 1 bar. and from my calculation 1 found that the average pressure was also ~1 with some uncertainty.

But during the last “nvt” run, I found that pressure is again shooting up and finally equillibrated at 100 Bar.

It seems that system is not yet equillibrated.

can you please suggest whether a longer “npt” is required ?

-OR-

In LAMMPS, do I need to change the parameters in the “fix nvt…” command to keep the pressure near the Target value ?

Best regards,

Hi,

I have a doubt.

I am doing an equillibration run for a liquid system. For that I have
performed the following steps:

nve (with T rescaling for 50 ps) ==> nvt (100 ps) ==> npt (500 ps) ==> nvt
(500 ps)

In "npt" : the Target Pressure = 1 bar. and from my calculation 1 found that
the average pressure was also ~1 with some uncertainty.

But during the last "nvt" run, I found that pressure is again shooting up
and finally equillibrated at 100 Bar.

It seems that system is not yet equillibrated.

can you please suggest whether a longer "npt" is required ?

very likely. impossible to tell without knowing
more details of your system setup. it is recommended
to visualize the trajectory and identify which process
correlates with the increase in pressure, if any.

simple statistical mechanical considerations tell you
that in general, the larger a system is, the more difficult
it will be to guide the system to equilibrium. a lot also
depends on the nature of your system, e.g. if "jamming"
or "entanglement" scenarios happen, that will keep
you from ever reaching equilibrium, unless you run for
an impossibly long time and with a much larger system.

-OR-

In LAMMPS, do I need to change the parameters in the "fix nvt..." command
to keep the pressure near the Target value ?

this is not a problem with a simple either-or answer. there are
many ways to lead or even force the system to behave in a
particular way. whether this is a good method to get a meaningful
trajectory depends a lot on system itself, the boundary conditions
and the intended purpose.

axel.