Re: Making klongs decay at a fixed position

From: Anna Ferrari <a.ferrari_at_hzdr.de>
Date: Tue, 25 Jun 2013 07:03:17 +0200

Dear Joseph,

just a comment: I completely agree with your considerations below, me too, for
example, I largely use the Monte Carlo to study detector response. I'm still
convinced, however, that a "tuned" use of the biasing techniques should give
you the same good results you obtained by modifying by yourself the decay time
(without re-weighting, I assume, or in any case re-weighting by yourself).
Maybe we can continue to discuss... if you give us more details about your
simulation.

Kind regards,
Anna


Am Mon, 24 Jun 2013 10:10:07 -0700 (MST) schrieb Joseph Comfort
<Joseph.Comfort_at_asu.edu>:
> Dear Anna,
>
> Thank you for your thoughtful reply and good explanations. I understand a
>little better now.
>
> I made the studies I needed by reducing the decay time (in PAPROP) of both
>klong and kshort, while keeping the ratio. This worked fine for my purposes.
> As I mentioned, the LAM-BIAS did not work well for me. I am not sure why,
>but perhaps because there was a broad range of klong energies in my 'beam'
>file.
>
> I recognize and understand that Fluka is designed to give reliable physics
>results as best as possible. That is good. It is put together well, and is
>exceptionally careful with biasing. I think it has very high quality and, at
>least in some of my applications, does better than some alternatives.
>
>Fluka is needed not only to simulate the reality of a physics project, but
>also to study the mechanisms and details of detector response. It is just as
>important to study how post-MC analysis codes can distinguish different types
>of responses. It is sometimes necessary to look at the signatures of
>specific particle decay modes. Backgrounds can be many orders of magnitude
>larger than the process of interest, which might be a rare decay, and one
>must be sure that an event thought to be "the real thing" does not come from
>some other process with very low probability. Hence, studies of incomplete
>or 'unphysical' processes must be done. In doing so, I only want to be sure
>that some subtle feature of the code does not unduly distort the results.
>
> Sincerely,
> Joe Comfort
>
> On Mon, 24 Jun 2013, Anna Ferrari wrote:
>
>> Dear Joe Comfort,
>>
>> Making klongs decay all at a fixed position is obviously impossible and
>>physically meaningless. "Biasing" means replacing a physical distribution (in
>>this case the distance distribution of decays f(z)=exp(-L/z)) by a modified
>>one (e.g. f'(z)=exp(-L'/z), with L'<L), and correcting for the modification
>>by multiplying the weight of the particles (and of their daughters) by the
>>ratio f(Z)/f'(Z), where Z is the sampled position.
>> In words, you obtain more decays at small z than it is physically probable,
>>but the particles get a low weight, according to the basic rule: weight x
>>probability = constant.
>> What you suggest is making f'(z) a Dirac delta function (infinite
>>probability density at z=Z), which would result in particles with weight
>>zero.
>> In my opinion the correct way to increase the statistics in your decay
>>volume is to modify the physical decay length via LAM-BIAS with SDUM=GDECAY,
>>as you already did. You should find the optimal decay length to 'fill' more
>>your decay volume.
>>
>> Concerning the second part of your question: what you say, that "klongs
>>will remain along with the decay particles, at least part of the time" is
>>again a statistical game (mathematically correct) ruled by the above formula
>>(conservation of weight x probability). The manual (LAM-BIAS card) says: " At
>>the decay point sampled according to the biased probability, Russian Roulette
>>(i.e. random choice) decides whether the particle actually will survive or
>>not after creation of the decay products. The latter are created in any case
>>and their weight adjusted taking into account the ratio between biased and
>>physical survival probability". This is correct if we want that the Monte
>>Carlo provides correct average values in the user-defined phase space
>>regions (the user-defined "detectors") . If you are not interested in
>>physically meaningful phase space densities, you can still "play" with the
>>code at your own risk...
>> In any case, you can select "the decay particles without the klongs" just
>>intercepting the tracking: you can discriminate them according to their
>>generation number (available as LLO in the calls to several user routines,
>>and as LTRACK in the TRACKR COMMON)
>>
>> Hope it helps,
>> Anna
>>
>>
>> -----Ursprüngliche Nachricht-----
>> Von: owner-fluka-discuss_at_mi.infn.it [mailto:owner-fluka-discuss_at_mi.infn.it]
>>Im Auftrag von Joseph Comfort
>> Gesendet: Dienstag, 28. Mai 2013 07:46
>> An: fluka-discuss_at_fluka.org
>> Betreff: Making klongs decay at a fixed position
>>
>> I need to look at the behavior of a detector system when a beam of klongs
>>all decay at a fixed position (ZBEAM) along the beam line. The physical
>>decay volume is too short for the klong decay length (only about 4% of the
>>klongs will decay in it, and I need to bias the rate).
>>
>> I thought LAM-BIAS could help, but I could not get it to work after trying
>>several variations. Also, the manual implies that klongs will remain along
>>with the decay particles, at least part of the time. I want the decay
>>particles without the klongs.
>>
>> I tried a workaround by redefining the klong lifetime (in PAPROP). It sort
>>of works, but gets entangled with kshort. Is there a 'clean' way to make
>>things work?
>>
>> Thank you,
>> Joe Comfort
>>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Anna Ferrari
Institute of Radiation Physics
Helmholtz-Zentrum Dresden-Rossendorf e.V.
Bautzner Landstraße 400
D - 01328 Dresden (Germany)
Tel. +49 351 260 2872
Received on Tue Jun 25 2013 - 08:12:42 CEST

This archive was generated by hypermail 2.3.0 : Tue Jun 25 2013 - 08:12:43 CEST