Re: [fluka-discuss]: magfld subroutine - did I do it right?

From: Giuseppe Battistoni <giuseppe.battistoni_at_mi.infn.it>
Date: Thu, 4 Aug 2016 12:29:07 +0200

Hello
a couple of advices:
1) the OPEN file is correct apart from the fact that you need to assign
numerical value (between 21 and 99) to U variable (so to assign the
fortran logic unit).
For istance, U = 98.

However, using OPEN in this way, the B map file will be searched in the
same directory where fluka is running, i.e. in the temporary directory
created by rfluka script.
You have either to give an absolute path to search for the file
to be readout, or, alternatively, you can use a special
FLUKA routine: OAUXFI (Open AUxiliary File) which automatically searches
for the file in the working directory (the one from which you launch the
run), or in the $FLUPRO directory or in your home directory.
The sintax would be:
        U = 98
        ALL OAUXFI('magfld.txt',U,'OLD',IERR)
where IERR is an integer output variable returning 0 is the file
is found and correctly opened.

2) you search for the nearest integer on the basis of X,Y,Z coordinate.
In any case you should make some checks to protect yourself from the
possibility of getting an index greater to the maximum dimension
of the BX,BY,BZ errors. That would produce a run-time error.
A part from this, with such a method you will get a non-continuos
dependence of B on the coordinates. It can be acceptable in some case,
but you could instead make a better job and spend some time trying
to find the closest points in X, Y and Z and perform ab interpolation. A
linear (or I should say trilinear) interpolation would be enough to have
a smooth dependence of B on the
three space coordinates.
Surely you can find in the web examples of algorithms to perform
interpolation.

        Cheers
                Giuseppe Battistoni

On 03/08/16 22:15, mjm_at_jlab.org wrote:
> I've been brushing up on my Fortran in order to make a magfld subroutine
> for a non-uniform magnetic field, using data from a file given to me by my
> supervisor. I just wanted to get some input from more experienced FLUKA
> users before I tried adding it to Flair, to make sure it looks okay.
> Please let me know if it looks okay, or if you see any problems with it.
> Thanks in advance.
>

-- 
INFN Milano
via Celoria 16, 20133 Milano
tel: +39 02 50317307
fax: +39 02 50317617
__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Aug 04 2016 - 14:06:09 CEST

This archive was generated by hypermail 2.3.0 : Thu Aug 04 2016 - 14:06:11 CEST