From: LEWYN LI (ll2150_at_columbia.edu)
Date: Thu Dec 09 2004 - 11:17:33 CST
Vivek,
When I simulated a water droplet in vacuo without periodic
boundary condition, I saw water molecules evaporating from the droplet.
Then my guess is that both N and E of the droplet is no longer constant,
and the droplet is out of equilibrium. Then equilbrium statistical
mechanics no longer applies.
On a practical level, I guess that if you run a "NE" or "NT"
simulation for only a short duration, you might still be able to get some
reasonable results. For example, if you simulate a large water droplet
for not too long, maybe the evaporation rate is slow enough that the
droplet is meta-stable. Then you might still be able to estimate some
properties of bulk water from your simulation.
Having said that, I still wonder if the water dynamics in such a
simulation might be in serious error, because the water at the droplet
boundary will be seeing a vacuum on one side but other water molecules on
the other. I don't know how much this would affect the properties of the
droplet.
Hope this helps!
LEWYN
On Thu, 9 Dec 2004, vivek sharma wrote:
> Hello everybody,
>
> Thanks for the help.
> So this means a run (constant temperature/presuure options 'off' +
> boundary conditions 'on')will give NVE ensemble simulation and a run
> (constant pressure option 'off' + constant temperature option 'on' +
> boundasry condition 'on') will give NVT ensemble simulations.
>
> But out of curiosity I ask , if I keep boundary conditions 'off' or say do
> not use them in the above mentioned two cases , then such ensembles will
> be of type NE (N and E constant) and NT (N and T constant), respectively.
> But theoretically, such ensemble does not exist. Statistical Mechanics
> uses the concept of volume of system in picturising the ensemble (all
> replicas of system are confined in some volumes), then is it right to run
> such 'NE' / 'NT' type of simulation? Ofcourse then the system to be
> simulated will lie in infinite vacuum.
>
> Please reply, waiting for your suggestions.
>
> Thanking you once again,
>
> sincerely,
>
> Vivek Sharma
>
> Quoting nordgren_at_sas.upenn.edu:
>
>> Hi Vivek,
>>
>> I can't really comment on your first question, since I have no
>> experience
>> with NAMDplot. But as for constant-volume dynamics, well, basically
>> that's
>> what you get unless you specifically use constant pressure. In other
>> words,
>> if you don't invoke one of the NAMD options to use constant pressure,
>> then
>> the size of your MD box will remain static, i.e., constant volume.
>>
>> Of course, the concept of a "box" and a "volume" only make sense if you
>> are
>> using some sort of periodic boundary conditions for your system...
>> otherwise,
>> you have a cluster of atoms surrounded by infinite vacuum....
>>
>> So, just use one of the constant-temperature methods, along with
>> periodic
>> boundaries (but without constant pressure) and you have NVT dynamics.
>>
>> good luck,
>> Erik
>>
>>
>> C. Erik Nordgren, Ph.D.
>> Department of Chemistry
>> University of Pennsylvania
>>
>>
>> Quoting vivek sharma <vivek.sharma_at_helsinki.fi>:
>>> 2. Also I want to know is there a way to run NVT ensemble in NAMD, I
>> see
>>> the
>>> methods and options for constant Temperature and constant Pressure,
>> but
>>> not
>>> for constant volume. Please explain briefly.
>>>
>>> Thanking you,
>>>
>>> sincerely
>>>
>>> Vivek (NAMD newbie)
>>
>>
>
>
> --
> VIVEK SHARMA
> HBG BI
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:02 CST