From: Giuseppe Battistoni (Giuseppe.Battistoni@mi.infn.it)
Date: Mon Dec 04 2006 - 17:56:41 CET
If you use REAL*4 variables you are probably missing precision
with that geometry.
Please, as recommended within the all FLUKA code: always use
REAL*8 variables, in particular for position and angles.
Can you try this?
Giuseppe B.
On Mon, 4 Dec 2006, Olaf Hartmann wrote:
> Hello.
>
> Giuseppe Battistoni wrote:
> > Quite strange: the size that you are describing is not critical for FLUKA.
> > Have you used double precision everywhere in your routine?
>
> All variables are initialized as REAL*4.
>
> > What is the size of your beam spot?
>
> It's a pencillike beam.
>
> > Are you usinb a magnetic field?
>
> No magnetic field.
>
> > Did you already check for trivial errors in your alignment?
> >
> Well, with the USRBIN card I see pions, nucleons etc. also when applying
> the "thin" target. Just the JTRACK variable in that case is always =2.
> With the "big" target instead (as in the attached input), it ranges from
> -5 to 25.
>
> Ciao,
> Olaf.
>
This archive was generated by hypermail 2.1.6 : Mon Dec 04 2006 - 18:20:55 CET