My question is how to equilibrate the system at exactly a series of designated temperatures using NPT ensemble.
I want to simulate thermal expansion of uranium dioxide. The reference paper states “MD calculation was performed at desired temperature and pressure (NPT ensemble). After passing enough time for the system to reach the equilibrium state, ………………”
I used the fix command with arguments specified as below.
fix all all npt temp 300 300 0.001 iso 0 0 0.5 drag 0.0
I increase the temperature parameters (both) by an interval of 50K and relax the system for 100ps after each increase.
But the simulation results indicated that both the temperature and volume fluctuate (±5k for temperature ) even though a considerable amount of time (100ps, used by the reference) has passed.
Therefore, I wonder how I can get the equilibrium state of a system exactly at a specific temperature using NPT ensemble.
Any comment or suggestion would be gratefully appreciated.
Some small fluctuations in temperature are to be expected when running NPT, even if it is well equilibrated. The issue is that the temperature should be constrained within a very narrow band around the chosen temperature.
However, one issue is that it looks like you have a very aggressive temperature correction: that may be causing the system to “overreact” to damp out the fluctuations, and thus cause bigger fluctuations than if you used a larger value for the damping constant (which is measured in time units).
My question is how to equilibrate the system at exactly a series of
designated temperatures using NPT ensemble.
I want to simulate thermal expansion of uranium dioxide. The reference paper
states “MD calculation was performed at desired temperature and pressure
(NPT ensemble). After passing enough time for the system to reach the
equilibrium state, ……………..”
I used the fix command with arguments specified as below.
fix all all npt temp 300 300 0.001 iso 0 0 0.5 drag 0.0
I increase the temperature parameters (both) by an interval of 50K and relax
the system for 100ps after each increase.
But the simulation results indicated that both the temperature and volume
fluctuate (±5k for temperature ) even though a considerable amount of time
(100ps, used by the reference) has passed.
two remarks.
a) 100ps is not at all a lot of time.
b) the only way to reduce temperature fluctuations is to increase the
system size. temperature is not a well defined property for a microscopic
number of particles, i.e. anything that can be handled by an MD simulation.
your description of your steps and input is not very detailed, so it is
not possible to give more detailed answers. here are a few questions
to think about:
how many particles does your system have?
for how long did you equilibrate your system with nvt?
how well does it conserve energy at nve?
how large is your initial pressure at nve/nvt?
perhaps you need to first relax the box with a
suitable drag factor to avoid oscillations?