Re: Geometry problems

Hi Stefan,
   thanks for your answer!
I increased the statistics and checked consistency of results by looking at the standard output (plus one
USRBIN card).
Results are consistent,thanks!
   In order to sum the histograms of the energy deposited in different regions I used the Add method in
ROOT, but you're right: results are not additive.
   I'm sorry...I had forgotten to attach an input file.
Results of the "SCORE" card in the standard output of this input file (in attachment) and of 235U.inp, are
practically the same
(1.99511...E-02 against 1.9934...E-02 for the 2 *001.out).

Thanks for your help!
Bye,
Francesca

----- Messaggio Originale -----
Da: Stefan Roesler <sroesler_at_mail.cern.ch>
Data: Martedi', Aprile 21, 2009 1:37 pm
Oggetto: Re: Geometry problems
A: francesca.belloni_at_ts.infn.it
Cc: fluka-discuss_at_fluka.org

> (smtp1.mi.infn.it [192.84.138.69]); Tue, 21 Apr 2009 11:20:57
> +0200 (CEST)
> Sender: owner-fluka-discuss_at_mi.infn.it
>
> Hi Francesca,
>
> Regarding your second question, I don't see anything wrong in
> the input,
> just a few comments:
>
> What is the statistic uncertainty of your results? Energy
> deposition in
> the cone is at a few-percent level of the total. Thus, 50000.0
> primaries,
> as specified in the inputs, doesn't seem sufficient(?)
>
> Charged particles below transport cutoff are ranged out. If the
> particlecrosses a region boundary when ranged out I think that
> FLUKA apportions
> the energy deposition. Thus, with one region you get the total
> while with
> two regions both get a share but this time in different energy
> bins (at
> lower energies). Thus, I am not sure if results are additive.
>
> If you have not yet done so, please also check the printout of
> the total
> energy deposition by region as given in the FLUKA standard
> output as the
> first printout after the run. The sum (on case of two regions)
> must be
> consistent (within errors) with the value obtained with one region.
>
> If you join the regions by OR (you didn't post that input?) you
> probably
> get numerically identical values to the case with one region
> (without OR),
> no matter what statistical uncertainty. Please check.
>
> Cheers
> Stefan
>
>
> On Fri, 17 Apr 2009 francesca.belloni_at_ts.infn.it wrote:
>
> > Dear FLUKA users.
> > I'm trying to score energy
> deposition as a function of the angle
> > between the z-axis and the direction of the generated
> particle. In order
> > to define different angular regions around the z
> axis,truncated cylinders
> > (TRC) were drawn.
> > These TRC, in my intentions,
> should have the same height parallel to
> > the z-axis, the same coordinates of the centre of the major
> circular base
> > (and hence the same vertice), but different major radii.
> > As a first step, in order to
> debug my geometry I begun with
> > "macroscopic" bodies (for example, the target thickness was
> set equal to 1
> > cm instead than equal to 4.2689E-5) and just 2 cones. My input file
> > (235U_old.inp) gave a geometry error, which I can't see.
> >
> >
> > **** Geometry debugging requested and activated ****
> > from X= 0.000000000E+00 to X=
> 1.000000000E+01 in step dX= 1.000000000E-01
> > from Y= 0.000000000E+00 to Y=
> 1.000000000E+01 in step dY= 1.000000000E-01
> > from Z= 0.000000000E+00 to Z=
> 1.000000000E+01 in step dZ= 1.000000000E-01
> >
> > **** Lookdb: Geometry error found ****
> > **** The point: 1.1 1.1 1.1 ****
> > **** is contained in more than 1 region ****
> > **** (regions: 5 Reg.#5 7
> Reg.#7 ) ****
> >
> > **** Lookdb: Geometry error found ****
> > **** The point: 1.4 1.4 1.4 ****
> > **** is contained in more than 1 region ****
> > **** (regions: 5 Reg.#5 7
> Reg.#7 ) ****
> >
> > **** Lookdb: Geometry error found ****
> > **** The point: 1.7 1.7 1.7 ****
> > **** is contained in more than 1 region ****
> > **** (regions: 5 Reg.#5 7
> Reg.#7 ) ****
> >
> > **** Lookdb: Geometry error found ****
> > **** The point: 2. 2. 2. ****
> > **** is contained in more than 1 region ****
> > **** (regions: 5 Reg.#5 7
> Reg.#7 ) ****
> >
> >
> > I modified the geometry in 2
> different ways:
> > 1. setting the minor radii to a number different from 0
> (235U_radii.inp)> 2. subtracting every time all inner cones
> (thought I thought it was not
> > necessary) (235U_subtracting.inp)
> >
> > In both cases the debugging gives no mistake anymore.
> > Could someone tell me what was wrong in my first input file, please?
> >
> > As a second step, I came back to the real
> sizes and scored energy
> > deposition in the whole gas around the target (235U.inp), with no
> > attention to angles. Then I scored energy deposition in the
> whole gas
> > again, but dividing the gas in 2 regions according to angles
> > (235U_mic_trc.inp). I expected that the sum of 2 histograms of
> energy> deposition (in region #4 and #5 of the 235U_mic_trc.inp)
> had to be equal
> > to the energy deposited in region #4 of the file 235U.inp. But
> I obtain
> > something different ("Results.ps").
> > Thinking this could be due to some energy cut, I changed
> 235U_mic_trc.inp> into 235U_mic_trc_OR.inp. In this file
> the region #4 and #5 of
> > 235U_mic_trc.inp are unified in one region using "OR".
> > The histogram of the deposited energy in
> region #4 of 235U.inp and of
> > 235U_mic_trc_OR.inp are exactly the same now. Is the
> difference (in the
> > deposited energy) between the results given by 235U.inp and
> > 235U_mic_trc.inp really given by energy cuts, or do I make a
> geometry> mistake, please? Could someone give a look, please?
> >
> > Thanks.
> > Best regards,
> > Francesca Belloni
> >
> >
> > P.S.
> > As for my previous message, thanks for help!
> > As for sampling only in forward direction,excuse me: I had to speak
> > about scoring,not about sampling.
> > I meant that FFs are emitted on 4pi, but the gas is situated
> only in
> > the forward direction with respect to the U3O8 target. Since I score
> > energy deposition only inside the gas, I'm scoring it in the forward
> > direction.
> >
>
> --
> ___________________________________
>
> Stefan Roesler
> CERN, SC/RP
> CH-1211 Geneva 23
> Switzerland
>
> Phone: +41-22-7679891
> Fax: +41-22-7669639
> E-mail: Stefan.Roesler_at_cern.ch
>
>
>