Dear All users,
What range of fluctuation of temperature in NVT ensemble is justifiable?
My current simulation has a wide range of this fluctuation(290<T=300<310).
Looks good.
A question to muse about …, “What should happen to these deviations from the mean value of temperature as the number of atoms in your simulation increase?”
Best Regards
Manoj
Looks good.
how can you tell without knowing the size of the system or the kind of
potential or anything else going on?
in general, the fluctuations are not that much of an issue (and indeed
inversely related to the number of atoms in the system), but rather
the average temperature over a larger window and whether it has a
significant drift or not (it should not).
axel
> Looks good.
how can you tell without knowing the size of the system or the kind of
potential or anything else going on?
By interpolating between several people who asked me similar questions ;-).
Ali, Axel is, as usual correct. I should have asked what size of system you
are simulating, which potential you have used, etc. I guess your simulation
size was several 100 atoms to a few 1000 atoms of some metal. At 300 K
(again assuming K and not degrees C) large amplitude oscillations would be
absent and it probably does not matter if it is a metal or some other
element.
Now I go back to my Zen shell and promise to keep my cup half empty ....
Manoj
in general, the fluctuations are not that much of an issue (and indeed
Dear Manjo and Axel,
Thank you very much for your replies,
Sorry for late response.
My system include a confined structure, water+nacl+co2 restricted by two solid planes in z direction. I have performed an NVT ensemble(T=300, 7400 atoms
totally).