[gmx-users] AdResS - adress_reference_coords

[Christoph Junghans]

2014-05-19 8:00 GMT-06:00  <gromacs.org_gmx-users-request at maillist.sys.kth.se>:
> Date: Mon, 19 May 2014 16:00:09 +0200
> From: Jakub Krajniak <jakub.krajniak at cs.kuleuven.be>
> To: gmx-users at gromacs.org
> Subject: [gmx-users] AdResS - adress_reference_coords
> Message-ID: <537A0E69.5010108 at cs.kuleuven.be>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Dear,
>
> I have a question related to the AdResS scheme. Mainly how I should
> interpret the width of ex and hy zone in the case of xsplits.
>
> Is it that adress_reference_coords define the center of the ex zone. So
> on the -x and x I have a ex/2.0 and then I have a region with width hy?
>
> Like on this scheme:
> |------| hy | ex/2.0 |rf| ex/2.0 | hy |------|
>
>
> Or I should interpreted it as:
> |------| hy | ex     |rf| ex     | hy |------|
>
> where rf is a reference coords
>
> The same question is for the sphere scheme.
It is the latter. See adress_weight in adress.c for the details, but
basically we
calculate the distance (respecting pbc) of a molecule from rf and
compare it against ex and (ex+hy).

Cheers,

Christoph
>
> Best regards,
> Jakub Krajniak
>
>
> --
> Jakub Krajniak, Computer Science Department @ KU Leuven
> mob. +32 477 688 404
> room: 200a 01.41
>
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

-- 
Christoph Junghans
Web: http://www.compphys.de