Hi Luigi,
you are right, the divergence is working correct. My Problem was also that
I was expecting an initial Gaussian Profile using the Source.f Routine, but
actually one has to programm this by himself. Now everything is clear,
thank you for your help!
Regards
Yuri
2015-08-18 11:17 GMT+02:00 Luigi Salvatore Esposito <
luigi.salvatore.esposito_at_cern.ch>:
> Hi Yuri,
> I executed your input with WHAT(3) = -1 with your source routine and I got
> the result in attachment,
> that is what one would expect.
> I dumped the direction cosine just after they are computed in the source
> routine.
>
> Your USRBIN scoring does not provide any information on the beam
> divergence, but only a spatial information.
> With 1 mrad FWHM ~ 0.4 mrad, the beam will open by about 0.13 mm over a 30
> cm. This implies that you will see
> almost all beam particles inside the central bin of your USRBIN scoring,
> which has a bin width of 0.2 mm.
>
> If you want to check the beam parameters at z = 30 cm, you could intercept
> the particles at that boundary with the mgdraw routine.
>
> Best regards, luigi
>
> On 18 Aug 2015, at 10:29, Yuri Simeonov <yuri.simeonov_at_gmail.com> wrote:
>
> Hi Luigi,
>
> thank you for your effort.
> The source routine I am going to use for simulating intensity modulated
> raster scanning from a particle accelarator, this is just the divergence
> part.
>
> Regards
> Yuri
>
> Luigi Salvatore Esposito <luigi.salvatore.esposito_at_cern.ch> schrieb am
> Di., 18.08.2015, 10:22:
>
> Sorry, Yuri, but I was too impulsive in my last email.
> The conversion is actually taken in charge by FLUKA itself when it reads
> the BEAM card.
> I need more time to understand what’s wrong.
>
> On 18 Aug 2015, at 10:02, Luigi Salvatore Esposito <luillo_at_cern.ch> wrote:
>
> Dear Yuri,
> in your code, you should convert the beam divergence from mrad (as it is
> expected in WHAT(3) of the BEAM card) to radiant.
>
> Btw, I see that your source routine re-implements only the beam
> divergence, otherwise everything is like the default FLUKA executable.
> It is strongly recommended not to use customised user routines especially
> in case one can use what is already
> available in FLUKA. This would be less error prone.
>
> Best regards, luigi
>
> On 17 Aug 2015, at 18:46, Yuri Simeonov <yuri.simeonov_at_gmail.com> wrote:
>
> Dear Luigi,
>
> thank you very much for your answear!
>
> Divbm is exactly WHAT(3) with a negative sign, Beam is along the positive
> Z axis.
>
> When I use your code with WHAT(3) = -100 mrad I get almost the same
> Results (look at FlukaDiv_vs_MyDiv.jpg, 1E7 Particles)
>
> But with WHAT(3) = -1 mrad I get a strange Peak (look at 1mrad.jpg), which
> i cannot understand.
> I plot the lateral Profile at z = 30cm.
>
> I have attached my .inp file, Source file and .flair.
>
> Regards
> Yuri
>
>
>
>
> 2015-08-15 20:31 GMT+02:00 Luigi Salvatore Esposito <
> luigi.salvatore.esposito_at_cern.ch>:
>
> Yuri,
> for your email, it’s not clear how big is the difference between the two
> cases…
>
> In any case, I assume that Divbm in your piece of code is the value you
> have set in WHAT(3) (with a negative sign)
> when you run FLUKA without a customised source routine.
>
> Then, if the Divbm is small and the beam is along the positive z axis (see
> note 4 of the BEAM card entry
> in the manual), the results should be almost the same.
>
> - If the divergence is not small, the code should be (TXFLK, TYFLK, TZFLX
> are cosine directions):
> …
> …
> TXHLP = TAN( DIV_VECTORX )
> TYHLP = TAN( DIV_VECTORY )
> THELP = SQRT( TXHLP*TXHLP + TYHLP*TYHLP + ONEONE )
> TXFLK (NPFLKA) = TXHLP / THELP
> TYFLK (NPFLKA) = TYHLP / THELP
> TZFLK (NPFLKA) = SQRT ( ONEONE - LTXFLK (NPFLKA)**2 - TYFLK (NPFLKA)**2 )
>
> In the limit of small Divbm, the two pieces of code should give the same
> results.
>
> - If the beam is not along the z-axis, than you have also to take into
> account an additional rotation of TXFLK (Y,Z)
> with the matrix BMAXIS contained in the common (BEAMCM).
>
> Best regards, luigi
>
> On 14 Aug 2015, at 17:07, Yuri Simeonov <yuri.simeonov_at_gmail.com> wrote:
>
> Dear Fluka Experts,
>
> I am using the source.f Routine and trying to sample my own Gaussian
> angular divergence. But somehow the results I get (1D X Profile) are
> different from the one I get when I use the original Fluka without Source.f
> .
>
> I use the following code:
>
> Divbm_Sigma = Divbm*0.4246609D+00 *Convert from FWHM to sigma
>
> CALL FLNRR2 (RGAUSSDIVX,RGAUSSDIVY)
> DIV_VECTORX = RGAUSSDIVX*Divbm_Sigma
> DIV_VECTORY = RGAUSSDIVY*Divbm_Sigma
>
> TXFLK (NPFLKA) = DIV_VECTORX
> TYFLK (NPFLKA) = DIV_VECTORY
> TZFLK (NPFLKA) = SQRT ( ONEONE - TXFLK (NPFLKA)**2
> & - TYFLK (NPFLKA)**2 )
>
> What am I missing?
>
> Regards
> Yuri
>
>
>
> <FlukaDiv_vs_MyDiv.jpg><1mrad.jpg>
> <Source_Sampling_Sph_400MeV_D0.008_5cm_20MP_Nucl_Edge.flair>
> <Source_Sampling_Sph_400MeV_D0.008_5cm_20MP_Nucl_Edge.inp><SourceFluka.f>
>
>
>
>
>
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Wed Aug 19 2015 - 16:21:06 CEST