Re: question about NPT

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Thu Jan 19 2012 - 08:12:54 CST

On Thu, Jan 19, 2012 at 6:55 AM, Boyang Wang <pkuwangboyang_at_gmail.com> wrote:
> Hi Axel,
>
> if you use the interactive MD, you will see that the system volume is

of course the volume changes. i didn't say that it does not.
*but* you are confusing cause and action. the NPT integrator
changes the volume to react to the pressure, it is not that the
pressure fluctuates because the volume is changing like you
are implying. the pressure *always* fluctuates, regardless of
whether you keep the volume of your cell constant or not.
and the pressure fluctuations are large because water (and
most dense liquids) are not very compressible. the volume
changes are small in comparison. since we are doing simulations
in finite size boxes with periodic boundary conditions, we have
a finite size effect on top of that. the pressure fluctuations are
caused by local density fluctuations (atoms bumping into each
other) and thus the larger your system, the smaller the fluctuations
of the total pressure due to averaging over (uncorrelated) local
fluctuations. it is the same with temperature (i.e. kinetic energy),
only that the fluctuations are (relatively) much smaller.

> changing. The NPT ensemble only fixes pressure, but not the volume.

actually, NPT does not "fix" the pressure. it adjusts the
volume to according to the sign and magnitude of the pressure.

you get a "constant" pressure only for a system in equilibrium
and by averaging over a longer time.

axel.

> Boyang
>
>
> On Thu, Jan 19, 2012 at 7:14 PM, Axel Kohlmeyer <akohlmey_at_gmail.com> wrote:
>>
>>
>>
>>
>>
>> On Jan 19, 2012, at 3:16 AM, Boyang Wang <pkuwangboyang_at_gmail.com> wrote:
>>
>> Hi Oguz,
>>
>> pressure fluctuates a lot in NPT ensemble, because the volume fluctuates a
>> lot. If you would like
>>
>>
>> This is nonsense. The volume does *not* fluctuate a lot, it is the low
>> compressibility of dense fluids that causes the relatively high pressure
>> fluctuations.
>>
>> Axel
>>
>> smaller fluctuation of pressure, you can increase the Langevin piston's
>> mass in the configuration file.
>>
>> Boyang
>>
>>
>> On Wed, Jan 18, 2012 at 4:06 PM, oguz gurbulak <gurbulakoguz_at_yahoo.com>
>> wrote:
>>>
>>> Dear All,
>>>
>>> I performed md simulations in NPT ensamble and  got -1000 atm pressure
>>> value for the system. My aim is to find 1 atm pressure value instead of a
>>> negative pressure value.  Is this problem indicates that the force field I
>>> use is not appropriate for my system ? or Is this a problem caused by the
>>> langevin parameters ?
>>>
>>> why the system did not reach the target pressure value? Could you please
>>> share your thoughts about this problem ?
>>>
>>> Kind regards.
>>
>>
>>
>>
>> --
>> Boyang Wang, Ph.D.
>>
>
>
>
> --
> Boyang Wang, Ph.D.
>

-- 
Dr. Axel Kohlmeyer
akohlmey_at_gmail.com  http://goo.gl/1wk0
College of Science and Technology
Temple University, Philadelphia PA, USA.

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:08 CST