[fluka-discuss]: Energy loss of relativistic nuclei in a crystal (update)

From: maestro <paolo.maestro_at_pi.infn.it>
Date: Thu, 28 Nov 2019 20:28:49 +0100

Hello Paola,
thank you for your check; it was very useful to give me some hints how to proceed.
It took me some time to make my checks and understand better the problem I reported in a previous mail of mine (see mail exchange below).

First, with respect to my previous simulation, I simplify the geometry in order to reproduce your result.
Now, I simulate nuclei of Oxygen (two energies: 1 and 10 TeV/n) impinging on a single crystal of PWO (2 cm thick), not surrounded by any other material.
Hadronic interactions are switched off (using THRESHOLD card).
As shown in annexed plot1, the energy deposited in a bar is the same, independently on the kinetic energy of the particle (black and green histograms).

As a second step, I modified the geometry by inserting a slab of tungsten (W, 0.35 mm thick) in front of the PWO crystal, as it was in my original simulation.
In this case, the energy deposit (only in PWO, not in W) distribution is shifted and gets broader (blue and red lines in plot1). The effect is more and more evident as the particle kinetic energy increases.
That was what puzzled me: why adding a slab of tungsten caused this significant change in the distributions of the energy deposited in PWO, and their shift with particle energy?

After some investigation, I discovered (by tracing particle with a modified mgdraw.f) that the effect is due to electron-positron pair production (PP) by heavy ions
at high energy (I am using CALORIMETRY as default).
As a check, I switched off PP with PAIRBREM card, and the distributions with W slab in front of PWO (blue and red in plot2) turned out to be similar (slight differences are due do dE/dx in W)
to the ones with PWO only (green and black).
Moreover, the energy deposit distributions in case of W slab (i.e. red and blue histograms in plot2) are not depending on the projectile particle energy, as expected if PP is not taken into account.

Now my point is the following. I understood the reason of the energy loss increase at high energy (i.e. it is due to PP in W).
However, why is PP so relevant in O nuclei interactions with W, but not in interactions with PWO?
In fact, I checked (with mgdraw) that PP is also produced in PWO (with similar kinetic energy spectrum of e+ e- pair to that in W),
but it seems having no effect on the energy deposit in the crystal as shown in plot3: energy deposit distributions in PWO crystal look all equals
in this case independently on whether PP is on or off and on projectile kinetic energy.
I would have expected PP in PWO is similar to that observed in W.

Do you have any explanation for that ?
Moreover I would like to know how PP of heavy nuclei is treated in FLUKA.
I have found in the manual neither the description of PP implementation nor references to papers describing this specific process.
Many thanks for your help.
Best regards
                Paolo














> On 1 Jul 2019, at 15:21, Paola Sala <paola.sala_at_mi.infn.it <mailto:paola.sala_at_mi.infn.it>> wrote:
>
> Hello
> Sorry, but I still do not reproduce your problem.
> I attach a simple input file, that has the same physics/transport cards
> as yours, runs on monochromatic ions, and uses the THRESHOLD card to
> inhibit interactions. Results are collected with a DETECT card. Running at
> 300 GeV/nucleon and 30 TeV/nucleon gives comparable results, as in the two
> plots attached (energy deposition in 2 cm PWO, no normalization).
> Maybe you could try to run your case adding the THRESHOL card, to be sure
> that there is no contamination by inelastic interactions? Although I do
> not understand why this should change with energy..
> Your source.f seems correct, but maybe try a monoenergetic run?
> I'll keep thinking..
> Paola
>
>> Hello Paola,
>> I set the BEAM card at -30000, i.e. 30 TeV/nucleon, if I correctly
>> understood the card instructions.
>> Oxygen kinetic energy is extracted from a E^-1 distribution up to 200 TeV
>> of particle energy, then it should be fine.
>> I noticed the same behaviour also using carbon nuclei; the dE/dx
>> “anomalous" rise starts around 10-15 TeV of particle energy.
>> I send you annexed my input and geometry cards for physics and the
>> source.f file I have modified for spectrum generation.
>> Thank you for your help
>> Paolo
>>
>>
>>
>>
>>
>>
>>
>>
>>> On 24 Jun 2019, at 11:47, Paola Sala <paola.sala_at_mi.infn.it <mailto:paola.sala_at_mi.infn.it>
>>> <mailto:paola.sala_at_mi.infn.it <mailto:paola.sala_at_mi.infn.it>>> wrote:
>>>
>>> Hello
>>> From a fist check, with oxygen ions at fixed energies, I cannot
>>> reproduce
>>> your results. The dE/dx stays fairly constant, as expected.
>>> One doubt: did you set the energy in the BEAM card to a sufficiently
>>> high
>>> value to cover all the source ions? If not, de/dx tabulations will not
>>> be
>>> correctly initialized. If yes.. could you please send me your input, or
>>> at
>>> least the card you use for physics and transport settings?
>>> Thanks
>>> Paola
>>>
>>>> Hello fluka developers,
>>>> I am doing simulations of the energy deposit of oxygen nuclei at high
>>>> energy in a bar of PWO crystal (density 8.28 g/cm^3, 2 cm thick).
>>>> Oxygen nuclei have generated according to KE^-1 spectrum, where KE is
>>>> the
>>>> kinetic energy per particle spanning from 178 GeV to 316 TeV.
>>>> I made plots (see attached figure to this email) of the energy loss
>>>> (dE/dx) distributions (computed from the energy deposited in the bar,
>>>> and
>>>> dividing by
>>>> the square of the particle charge) for different energy intervals of
>>>> KE.
>>>> In these plots, nuclei undergoing hadronic inelastic interaction
>>>> (somewhere in the crystal) are not included.
>>>> As you can see, dE/dx has a nearly constant value (around 2 MeV
>>>> cm^2/g,
>>>> as expected) up to KE ~ 10 TeV, then it increases dramatically for
>>>> higher
>>>> KE.
>>>> I wonder which might be the physical reason for this increase of dE/dx.
>>>> I would have expected an almost constant value for relativistic charged
>>>> particles (in the Fermi plateau).
>>>> Are there any radiative effects turned-on for ultra-relativistic
>>>> charged
>>>> particles?
>>>> Thanks for your help.
>>>> Paolo Maestro
>>>>
>>>>
>>>>
>>>>
>>>>


__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Nov 28 2019 - 21:39:57 CET

This archive was generated by hypermail 2.3.0 : Thu Nov 28 2019 - 21:40:01 CET