From owner-chemistry %-% at %-% ccl.net Tue Nov 25 06:16:02 2008
From: "Vincent Leroux vincent.leroux*o*loria.fr" <owner-chemistry ~~ server.ccl.net>
To: CCL
Subject: CCL: Ewald Sum
Message-Id: <-38171-081125050537-3912-dZAAYqL1AYVv6agw14YamQ ~~ server.ccl.net>
X-Original-From: Vincent Leroux <vincent.leroux###loria.fr>
Content-Transfer-Encoding: 8bit
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Date: Tue, 25 Nov 2008 10:24:14 +0100
MIME-Version: 1.0


Sent to CCL by: Vincent Leroux [vincent.leroux,+,loria.fr]
Hi Leyla,

You really should not try to understand Ewald sum method straight from 
source code, even if in this case it is well documented. The Ewald sum 
is done in Fourier space so first make sure you master this mathematical 
concept.
http://en.wikipedia.org/wiki/Fourier_transformation
http://en.wikipedia.org/wiki/Ewald_summation

Then if you want, try to see how the conversion to Fourier space is 
handled in the code, how the summation is done, how code is optimized so 
that terms that can be computed once and for all are grouped in the 
setup routine, etc. The routine here is particularly simple! Actual MD 
programs such as NAMD rely on the same principles but with a good number 
of additional optimizations, in particular the code is heavily 
parallelized. This is much more complex, hopefully you do not need to 
master the code in order to understand and perform MD...

For theoretical basis of molecular mechanics I strongly recommend you 
read A.R.Leach's book, if not done already.
http://www.booksites.net/leach2/molecular/molecular_modelling_2.html

Good luck for your PhD ^^
VL


Leyla Ramin lram5228%%usyd.edu.au a écrit :
> Sent to CCL by: "Leyla  Ramin" [lram5228[A]usyd.edu.au]
> I am PhD student in Sydney university and I am working in molecular dynamic simulation, The codes in fortran language in your website in part:sources, allen-tildesley-book "http://www.ccl.net/cca/software/SOURCES/FORTRAN/allen-tildesley-book/index.shtml>"
> is so useful for me.
> In the "FICHE F.22.  ROUTINES TO PERFORM THE EWALD SUM" the routine is 
> so complicated to understand the method. I am wondering if I can get more information about this code.
> Please kindly advise me and thanks for your consideration.
>
> Best Regards
> Leyla Ramin
>
>