[gmx-users] The md-vv integrator and the choice of cutoffs

[Szilárd Páll]

Have you considered increasing the verlet buffer size? 4.6.5 is rather
old, so it would be good to verify whether you see the same with 5.1,
I'd say.
--
Szilárd


On Fri, Jul 29, 2016 at 10:23 AM, Miguel Caro <miguel.caro at aalto.fi> wrote:
> Hi Mark,
>
> Thanks for your answer. I just created an issue and assigned it to you
> (I hope that's OK). Could this be a version problem? I am running 4.6.5
> but could run a number of other versions as well.
>
> Thanks again,
> Miguel
>
> On 2016-07-28 22:43, Mark Abraham wrote:
>> Hi all,
>> >
>> > I am simulating a mixture of water and acetonitrile with the OPLS force
>> > field. I am using the md-vv integrator because I want my velocities to
>> > be in synchrony with my positions (md outputs velocities and positions
>> > half a time step apart).
>> >
>> > My box is relatively small, 2400 atoms, and I am using 1.3 nm for all
>> > cutoffs. The properties of the mixture evolve smoothly with composition
>> > until the point where I reach the pure acetonitrile extreme, which also
>> > corresponds to the largest box size in the series. L increases with
>> > acetonitrile mole fraction from about 2.86 nm for pure water to 3.39 nm
>> > for pure acetonitrile (T = 323K). For pure acetonitrile the smooth trend
>> > breaks down abruptly and I can observe a severe kinetic energy transfer
>> > from the translational degrees of freedom to the vibrational degrees of
>> > freedom. However, if I increase the cutoff to 1.6 nm the issue goes away.
>> >
>> > Also, if I use md as integrator instead of md-vv, with the original set
>> > of cutoffs at 1.3 nm, the issue also goes away. I would like to
>> > understand how/why/whether the choice of integrator and the box size
>> > affect the effect of the cutoffs on the distribution of kinetic energy
>> > in the system. I've gone as far as pinpointing the possible origin of
>> > the issue and a workable solution (i.e. just switch to md), but I would
>> > like to understand the underlying cause of the problem.
>> >
>> > Many thanks,
>> > 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.