Re: Consecutive Simulations

From: Lewyn Li (lewynli_at_gmail.com)
Date: Thu Feb 25 2010 - 10:08:47 CST

Hi Keith,

     Just my 2 cents:

1. Formally, I guess you can just run NVE for 100000 step, then take the
last output configuration and velocities, run NVT for 100000 steps, and take
the last output configuration and velocities, and run NPT for 100000 steps
i.e. just run three consecutive but separate MD simulations.

2. However, this may not be a very good idea. NVE, NVT and NPT are
different thermodynamic ensembles, and I would say that there is no
guarantee that a configuration and velocity values sets from NVE would be
stable in NVT. At the very least, I would think that some equilibration
should be done "transitioning" from NVE to NVT, and from NVT to NPT.

3. As far as I remember, in the thermodynamic limit, NVE and NVT ensembles
would be almost identical. (See MacQuarrie). It is not easy to know or
show if your MD simulations long enough to be in the thermodynamic limit.
Also, the kinetics of transition from NVE to NVT is another thing that you
may need to consider.

     Hope this helps.

LEWYN

On Thu, Feb 25, 2010 at 10:25 AM, Keith Battle <keith.battle_at_gmail.com>wrote:

> Hi NAMD world,
>
> I was wondering if it is possible to make an input file to perform a
> simulation first at NVE, then at NVT, then at NPT conditions
> consecutively. Basically I would like to execute the following
> pseudo-script:
>
> Input NVE conditions
> run 100000
> Input NVT conditions
> run 100000
> Input NPT conditions
> run 100000
>
> Best,
>
> Keith Battle
> University of South Alabama
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:50 CST