Hi all ,
i am trying to use an npt integrator to keep pressure and temperature at 1.0 atm and 300K respectively using
fix 1 molecules npt temp 300 300 100.0 iso 1.0 1.0 100.0
Temperature is working fine but for pressure after some timesteps i have very big fluctuations from -230 to 100.
Hi all ,
i am trying to use an npt integrator to keep pressure and temperature at 1.0
atm and 300K respectively usingfix 1 molecules npt temp 300 300 100.0 iso 1.0 1.0 100.0
Temperature is working fine but for pressure after some timesteps i have
very big fluctuations from -230 to 100.
so what?
this has been discussed without end on this list.
most of the time, the fluctuations are completely
physical, since people are simulating a system
with a low compressibility and have only a comparatively
small system cell.
in other cases, the initial system is not well equilibrated
before the npt algorithm is activated.
sometimes, those kind of fluctuations are hard to remove,
which is why the fix has a "drag" parameter to artificially
reduce fluctuations.
same as for temperature, the final target pressure is only
to be reached on average and when the system is in
well equilibrated.
cheers,
axel.
Axel is right, as usual. This fix samples volumes and energies from the NPT ensemble. As a consequence of this, the time-average instantaneous pressure <P_int> equals the specified external pressure, but the fluctuations in P_int will depend on the system size and the isothermal compressibility. Hence negative values of P_int are quite common for small systems.