[gmx-users] Infinite cutoffs with periodic boundary condition

[Mark Abraham]

On Mon, Jan 12, 2015 at 11:53 AM, 이종화(Jong Wha Lee) <jongwha12 at postech.ac.kr
> wrote:

> Dear Gromacs users,
>
>
>
> Hello everyone. I am currently trying to simulate a molecule in carbon
> dioxide gas. To maintain carbon dioxide near my molecule of interest at a
> well-defined pressure, it was necessary to apply periodic boundary
> condition. Then, I tried to apply infinite cutoffs (calculate all
> interactions) as this would be more accurate with small increase in the
> computational costs,


... maybe... if you were running in a non-periodic vacuum, which you aren't.


> considering that I am simulating the gas phase with
> small number of atoms.
>

Periodic => infinite number of atoms.


>
>
>
> However, I found that it was not possible to give infinite cutoffs with
> periodic boundary condition. Consequently, I decided to use a very large
> cutoff (20 nm)


OK, but now it's on your head to prove that your choice of model physics is
a good representation of reality. If you want a full representation, that's
why Gromacs has PME implementations (for Coulomb and VdW).

and update neighbor list every step to minimize possible
> errors from simulation. But this resulted in high computational cost. Could
> you suggest any more efficient and accurate methods that I could try?
>

Just like simulations of condensed-phase systems, you want a buffered
neighbourlist with a suitable update frequency. This is all done for you in
the Verlet cutoff scheme, but you are using the group cutoff scheme (so
don't do that).

Mark


> I attached my .mdp file below. I’m performing NVE simulation after
> equilibration with NPT.
>
>
>
> integrator           = md
>
> nsteps                 = 50000000
>
> dt                       = 0.001
>
> ; Output control
>
> nstxout               = 50000
>
> nstvout               = 50000
>
> nstenergy           = 5000
>
> nstlog                 = 50000
>
> nstxtcout            = 5000
>
>
>
> ; Noncovalent interactions and neighbor search
>
> cutoff-scheme    = group
>
> coulombtype      = cut-off
>
> vdwtype            = cut-off
>
>
>
> ns_type              = simple
>
> nstlist                 = 1
>
> rlist                     = 20
>
> rcoulomb            = 20
>
> rvdw                  = 20
>
>
>
> ; No temperature and pressure coupling
>
> tcoupl                 = no
>
> tc-grps                = System
>
> tau_t                  = 0.1
>
> ref_t                   = 298
>
> pcoupl                = no
>
>
>
> pbc                     = xyz
>
> gen_vel              = no
>
> continuation      = yes
>
>
>
> Many thanks in advance,
>
>
>
> Jong Wha
>
> --
> 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.
>