[gmx-users] NPT equilibration with small cell

[Tsjerk Wassenaar]

Hey :)

Since pressure is so ill-defined at those volumes, you may as well run NVT
then at a more or less plausible volume. That will relieve the worry that
the deviations from the intended physics (in light of the force field
chosen) will cause the system to expand or contract out of control and you
can cherish some hope that the aberrations in the interactions sort of
balance. Of course, part of that should later be corrected by DFT geometry
optimization. As for the question which volume you need to use then, I can
only say that there is no answer. There is a distribution of possible
volumes for a small system at a given temperature. That is also true for
big systems, but there the distribution is much sharper.

Hope it helps,

Tsjerk

On Wed, Feb 17, 2016 at 1:47 PM, André Farias de Moura <moura at ufscar.br>
wrote:

> maybe just a historical curiosity, but small systems with only 100
> molecules or so and cutoffs of 0.7 nm were considered as best practices,
> maybe even state-of-art, in computer simulations of liquids:
>
> J. Chem. Phys. 79, 926 (1983); http://dx.doi.org/10.1063/1.445869
>
> nowadays you have to use systems like those for your AIMD for the same
> reasons they had too back in the day: not enough computer power to treat
> the system of the size you would like to.
>
> so I would do just like Vitaly suggested: classical MD with small
> system/small cutoffs, DFT geometry optimization and then your goal, the
> AIMD simulation.
>
> Andre
>
> On Wed, Feb 17, 2016 at 9:44 AM, VITALY V. CHABAN <vvchaban at gmail.com>
> wrote:
>
> > you are on the right way.
> >
> > set high pressure.
> >
> > set cut-offs to half box side.
> >
> > run classical MD, ~10 000 steps, and, afterwards, manually increase the
> box
> > sides to mimic the experimental density
> >
> > in the plane-wave DFT, run local optimization prior to AIMD.
> >
> >
> >
> >
> >
> >
> >
> >
> > On Tue, Feb 16, 2016 at 4:01 PM, Miguel Caro <miguel.caro at aalto.fi>
> wrote:
> >
> > > Hello,
> > >
> > > I was wondering if there is any reason why one should expect a constant
> > > pressure simulation not to work for a small cell. I am running a very
> > small
> > > cell (1.2 nm^3) simulation because I need it as input for ab initio
> MD. I
> > > am trying to simulate NPT at 1 atm and it seems like my cell, a mixture
> > of
> > > water and methanol, is expanding uncontrollably (to 10 nm^3).
> > >
> > > Is it not possible to do NPT for cells this small? I should add that
> NVT
> > > seems to work just fine for the same system.
> > >
> > > These are the settings I am using for the pressure coupling:
> > >
> > > Pcoupl = MTTK
> > > tau_p = 0.2
> > > compressibility = 4.5e-5
> > > ref_p = 1.0
> > >
> > > I have tried to do the same simulation with a supercell made up of
> 2x2x2
> > > periodic copies of my small cell and that seems to work fine. For that
> > > simulation, the only difference besides the cell size was the neighbors
> > > cutoffs, which I could increase from rlist = rcoulomb = 0.45, rvdw =
> 0.5
> > > for the small cell to rlist = rcoulomb = 0.8, rvdw = 1.1 for the
> > supercell.
> > > For the small cell the cutoff radii are rather limited by the box size.
> > >
> > > Thanks in advance,
> > > Miguel
> > >
> > > --
> > > *Dr. Miguel Caro*
> > > /Postdoctoral researcher/
> > > Department of Electrical Engineering and Automation,
> > > and COMP Centre of Excellence in Computational Nanoscience
> > > Aalto University, Finland
> > > Personal email: *mcaroba at gmail.com*
> > > Work: *miguel.caro at aalto.fi*
> > > Website: http://mcaroba.dyndns.org
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > > send a mail to gmx-users-request at gromacs.org.
> > >
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-request at gromacs.org.
> >
>
>
>
> --
> _____________
>
> Prof. Dr. André Farias de Moura
> Department of Chemistry
> Federal University of São Carlos
> São Carlos - Brazil
> phone: +55-16-3351-8090
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>



-- 
Tsjerk A. Wassenaar, Ph.D.