Date: Mon, 28 Oct 2013 11:07:57 -0400
Ooops, my bad! There's an '8' missing . These are the cell dimensions:
36.4353420 36.4353420 38.1764855 90.0000 90.0000 120.0000
The system is a solid, its constituents are octa-methyl sylsesquioxanes, a
type of cage-molecule. I attach a picture of a 3x3x3 crystal.
If this little crystal were to be surrounded by air, the normal pressure on
all of its surfaces should be the same. From this I had assumed that the
pressure components Px, Py, and Pz should be the same.
-Jose
On Fri, Oct 25, 2013 at 6:23 PM, case <case.biomaps.rutgers.edu> wrote:
> On Fri, Oct 25, 2013, Jose Borreguero wrote:
> >
> > I am trying to equilibrate a system in the NPT ensemble. I set the
> > following pressure settings:
> >
> > !PRESSURE REGULATION
> > ntp=1, !isotropic position scaling
> > pres0=1.013, !reference pressure
> > taup=2.0, !relaxation time (psec)
> > comp=44.6, !compressibility
> >
> > My system is not cubic, but hexagonal:
> > Lx Ly Lz alpha beta
> > gamma:
> > 36.43534 36.43534 3.17648 90.00000 90.00000 120.00000
>
> Oooh...what kind of system is this, with such a large discrpancy in unit
> cell
> lengths? I'm kind of surprised that Amber (is this sander or pmemd?) would
> even run with a cell length of only 3.2 Ang. What does it say in the
> output
> about the "Largest sphere to fit in the unit cell"?
>
> >
> > After the run, the pressures taken from file mdend average to:
> > <pressure>: 15.0
> > <press_x>: 259.9
> > <press_y>: 301.2
> > <press_z>: -515.5 !!!!!
>
> This looks about right to me: the volume scaling is only making the total
> pressure close to the target, and doesn't care about the individual terms.
> Without know what sort of system this is, it's hard to say how long it
> might
> take (if ever) to equilibrate to a system with equal pressures.
>
> ....dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
(image/png attachment: crystal.png)
