Dear Niklas
the best option (and also the one more sound physics-wise) is to plot
the neutron fluence with the energy in log scale, having care of
multiplying the fluence value for each bin by the median energy of the
bin.
This is normally referred to as a "lethargy" plot, and you can easily
check that plotting E x dPhi/dE is the same as plotting dPhi/dlog(E).
This kind of plot has the advantage that the integral below a part of the
plot/spectrum is directly comparable to the integral below another part
(again assuming the X-axis is logarithmic). It also enhances the features
of the neutron spectrum. For a "perfect" moderator (eg hydrogenated
materials between 1 eV and 1 MeV) it can be demonstrated that the
spectrum in lethargy represention should be flat (or, if you wish,
dPhi/dE should be proportional to 1/E => E x dPhi/dE = constant).
Concerning the 0 bins, simply skip them. You have zero bins above 70 MeV
(obviously given your proton beam energy) and in the thermal region since
Ta is a poor moderator and anyway your target is thin => the produced
neutrons did not undergo enough scattering to be thermalized.
Multiplying by the bin width, with unequal bins, will result in unphysical
discontinuities as you correctly pointed out, and it does not make any
physical sense.
Have a nice day
On Sat, 14 Mar 2020, Ophoven, Niklas wrote:
> Dear Fluka users,
> I want to plot neutron spectra for different primary proton energies
> impinging on a simple and cylindrical tantalum target. However, when I plot
> it, some problems arise.
>
> I know that Usrtrack will deliver me the differential fluence dÖ/dE with the
> unit neutrons/cm^2*GeV^-1*primary in the third column of the tab.lis file,
> which was obtained by processing the corresponding fort files by means of
> the ustsuw command (please find that file enclosed). Im not totally sure if
> a semi- or double-log plot would be best suited here. The semilog plot would
> show a steeper peak, the double log plot would however also show the
> variation of the fluence at very small and high energies with very small
> fluences. If I want to plot the data in the form of a double-log plot I have
> to process the zero values in order to avoid gaps in between. So far I
> decided to delete these values, this however leads to a higher bin width at
> some points and an unequal bin width in general then. When the bin width
> increases, one could obtain a somewhat misleading impression I think since
> the plot then pretends at some points that there is a fluence but indeed it
> is not.
>
> Another common option to plot my data would be to multiply all the y-values
> by the energy bin width, so that I get neutrons/cm^2*primary on the y-axis.
> However, I have realized that this will usually always lead to irritating
> results since the bin width from 1E-14 GeV up to 20 MeV (which is fixed due
> to the multi-group transport of neutrons in FLUKA) differs from the bin
> width after 20 MeV, where the program considers the chosen bin width taken
> by the user. Consequently, an irritating "jump" up or down after 20 MeV
> would be visibile.
> Apart from that I know that I could move from a stairstep graph to a common
> line graph if I take the middle of each bin as my x-value.
>
> I would be very thankful if you could help me and tell me what would be the
> best way to plot the data correctly.
>
> With best regards,
>
> Niklas
>
>
> ---------------------------------------------------------------------------
> ---------------------
> ---------------------------------------------------------------------------
> ---------------------
> Forschungszentrum Juelich GmbH
> 52425 Juelich
> Sitz der Gesellschaft: Juelich
> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
> Vorsitzender des Aufsichtsrats: MinDir Volker Rieke
> Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
> Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
> Prof. Dr. Sebastian M. Schmidt
> ---------------------------------------------------------------------------
> ---------------------
> ---------------------------------------------------------------------------
> ---------------------
>
>
>
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Mon Mar 16 2020 - 12:06:00 CET