[gmx-users] very small nonzero charge

[Alex]

Hi Mark,

Thanks. If the answer to q. 1 is no, then I will just stick with something
like 3-4 digits.

Alex

On Thu, Sep 27, 2018, 3:56 AM Mark Abraham <mark.j.abraham at gmail.com> wrote:

> Hi,
>
> 1. No. 2. From the look of
>
> https://github.com/gromacs/gromacs/blob/master/src/gromacs/gmxpreprocess/nm2type.cpp#L91
> ,
> you could supply as many as fits in a double-precision value (maybe about
> 15), but whether the subsequent calculations retain the precision is
> unclear without a detailed look at the code. I suggest you construct a toy
> case with three particles whose charge is 0.333333333333333 and one of them
> has a last digit of 4 and see how you go.
>
> Mark
>
> On Thu, Sep 27, 2018 at 11:40 AM Alex <nedomacho at gmail.com> wrote:
>
> > Hm, i thought it was a question someone could answer...
> >
> >
> > On Sat, Sep 22, 2018, 10:52 PM Alex <nedomacho at gmail.com> wrote:
> >
> > > Hi all,
> > >
> > > I am dealing with a solid structure with partial charges initially
> > > obtained from DFT and adjusted to work with MD. It just so happens that
> > > the total charge of the system, however small, will always be nonzero.
> > > The reason is trivial and unfortunately unavoidable, given the
> > > geometries. It sounds silly, but this is the situation. It has to do
> > > with division by three and one of the partial charges will have to have
> > > the form of e.g. 0.333333... Hence, a few questions:
> > >
> > > 1. Is the total charge of the system of the order ~1e-6 expected to
> > > produce artifacts with the latest versions of GMX (we're using 2018.1)?
> > >
> > > 2. The total charge will go down with the number of digits readable by
> > > x2top from atomname2type, so what's the max number of digits I could
> > > supply?
> > >
> > > Thanks,
> > >
> > > Alex
> > >
> > >
> > --
> > 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.
>