[gmx-users] constant area and normal pressure

[David L. Bostick]

Hi Erik,

I found this message lying around in my inbox and thought I'd respond.  ..

P-R coupling seems to work great on my membrane-type systems .. isotropic,
completely anisotropic, and constant area and normal pressure alike.
Thanks for the advice on the coupling time constant.  I'm using 2.0 ps now
and the fluctuations in pressure are much smaller :)

David

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
David Bostick					Office: 262 Venable Hall
Dept. of Physics and Astronomy			Phone:  (919)962-0165
Program in Molecular and Cellular Biophysics
UNC-Chapel Hill
CB #3255 Phillips Hall				dbostick at physics.unc.edu
Chapel Hill, NC 27599	           		http://www.unc.edu/~dbostick
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

On Fri, 30 Aug 2002, Erik Lindahl wrote:

> Hi,
>
> >
> > pcoupl                 = Parrinello-Rahman
> > pcoupltype              = anisotropic
> > tau_p                   = 0.5
> > compressibility         = 0.00 0.00  4.5e-5 0.0 0.0 0.0
> > ref_p                   = 0.00 0.00 0.00 0.00 0.00 0.00
> >
> > or something similar with semiisotropic coupling.  Would this constitute a
> > proper constant area and constant normal pressure (i.e. NP[n]AT ) ensemble?
> >
>
> Yep, it should. To nitpick you should probably couple the z-direction to
> 1.0 bar, but with the fluctuations on a molecular level this is
> essentially the same as 0.0 :-)
>
> I would probably select a larger time constant than 0.5 too, something
> in the order of 2-10 ps. With Parrinello-Rahman, this will be the period
> of the oscillations at equilibrium pressure, and 0.5 is short enough
> that it could affect the individual atom motions significantly.
>
> Let me know how it works; I've mainly tested P-R on crystal systems with
> a completely flexible box. It should work for your case too, but
> membranes can be quite sensitive to pressure coupling.
>
> Cheers,
>
> Erik
>
>
>