Dear Lammps Users,
In one of my cooling simulations, I have used the following NVT command using ramp temperature.
Dear Lammps Users,
In one of my cooling simulations, I have used the following NVT command
using ramp temperature.#-------------------------------------------
timestep 0.5 # fs
fix 40 all nvt temp 1000.0 1000.0 100.0
run 100000
unfix 40
fix 40 all nvt temp 1000.0 300.0 100.0
run 1400000
#-------------------------------------------
Simulation works fine. LAMMPS reduces the temperature from 1000 K to 300 K
within 700 ps (=1400000*0.5/1000) time. Slope of the ramp is 700K/700 ps =
1K/ps. Nose-Hoover thermostat in NVT ensemble considers constant atom
number, volume and temperature. Since I am using ramp temperature, my
understanding is that LAMMPS considers the variable ramped temperature as
constant value while time integrating. I mean at t=0 ps, constant temp
T=1000 K; at t=100 ps, constant temp T=900 K; at t=200 ps, constant temp
T=800 K and so on.One of my colleagues does not agree with me. Here are his comments.
“NVT ensemble suggests that the number of atoms = N, the volume = V and the
temperature = T are held constant for the simulation. So if just NVT is
used, all the three values should remain constant and temp is held constant
using the Nose-Hoover algorithm. I suspect that NVT is being used to have a
time integrator and then the cooling is carried out by using a ramp method
that scales the velocities at a rate of 1K/ps. So the ensemble is not NVT”I would appreciate your comments on this.
your colleague is mostly correct. to have an nvt ensemble, you need to
have a fully periodic system in equilibrium where the volume and
number of particles are constant and the kinetic energy fluctuates
around an average in a way that is consistent with a system in contact
with a sufficiently large heat bath at the desired target temperature.
your temperature ramp is a non-equilibrium simulation. now the ramp is
implemented by simply resetting the target temperature as you say, but
the fact that LAMMPS calls this fix "nvt" doesn't mean it will
automatically generate an NVT ensemble. all other required conditions
must be fulfilled as well.
axel.
Dear Axel,
Thank you for your comments.
I am using periodic system. I am wondering what would be the perfect wording for such protocol.
Should I write like –
"System is cooled from 1000 K to 300 K temperature with non-equilibrium pseudo NVT ensemble where ramp temperature is implemented by resetting target temperature in the Nose-Hoover algorithm."
Thank you in advance for your comments.
Best regards,
Sanjib
Dear Axel,
Thank you for your comments.
I am using periodic system. I am wondering what would be the perfect wording for such protocol.
Should I write like –
"System is cooled from 1000 K to 300 K temperature with non-equilibrium pseudo NVT ensemble where ramp temperature is implemented by resetting target temperature in the Nose-Hoover algorithm."
Thank you in advance for your comments.
Best regards,
Sanjib
Dear Axel,
Thank you for your comments.
I am using periodic system. I am wondering what would be the perfect wording for such protocol.
Should I write like –
"System is cooled from 1000 K to 300 K temperature with non-equilibrium pseudo NVT ensemble where ramp temperature is implemented by resetting target temperature in the Nose-Hoover algorithm."
there is no such thing as a "pseudo ensemble". just say what you did. e.g.
"The System is cooled from 1000 K to 300 K over a period of ## ps
using temperature ramp using a Nose-Hoover thermostat chain."
please note, your previous sentence was missing the very important
information of how rapidly, you were cooling the system down.
axel.
Thank you for your comments/suggestions.
Best regards,
Sanjib