Unexpected dip in LET spectrum

Scoring
1

Dear Dávid,
thank you very much for your previous answer.

I have another question regarding the same input I uploaded.
I am currently simulating the LET spectrum of recoil protons produced by the interaction of neutrons emitted from an AmBe or Cf sources within a polyamide converter. I am scoring this spectrum at the boundary between the converter and a second region made of PADC. However, I have observed an unexpected dip in the spectrum, which I was not expecting.

total spectrum
total spectrum1573×1015 82.6 KB

I’d like also to clarify the meaning of WHAT(6) ‘mat’ in the USRYIELD scoring card. From what I understand, this should be the material of the region where the protons enter and where LET is being evaluated. Is this interpretation correct?

To further analyze the behavior of the spectrum, I have extended my input by including a set of USRYIELD cards to study the dependence of the spectrum on the protons’angle with respect to the beam axis. I used the pseudorapidity lab. as first variable and LET (or energy) as second. However, when I plot the spectra obtained for different pseudorapidity intervals and then compare them to the total spectrum (integrating over pseudorapidity from 0 to infinity), I notice an inconsistency: the total spectrum has a very different scale on the y-axis, way much smaller.

I suspect this discrepancy is related to a normalization factor, but I have not found sufficient documentation on this. I also noticed that during the FLUKA Advanced Course, the slides on scoring mention paying attention to this factor, but without further explanation of its meaning or implications.
Here above you can see the different spectra and the total one is missing since the y scale is of the order of 10^-10 (see first picture)

pseudorapidity-LET
pseudorapidity-LET1574×1010 211 KB

Thank you very much for your support!
Cheers,
Antonella

problem.flair (41.1 KB)
problem.inp (7.2 KB)

3

Dear @Antonella_Mele,

Thank you for your question. I reply below to the three points you have raised:

  1. The behaviour you have found is due to a slight discontinuity in the shell correction term at energies close to 1 MeV. It will be addressed in future releases.
  2. The material parameter WHAT(6) in the USRYIELD card when scoring LET, as mentioned here, is the material in which the LET is calculated. This material may be different to the actual material through which the particle is being transported. For instance, the user can score the LET in water for a proton travelling through air. The sole condition about the material WHAT(6) is that it must be assigned to certain region so that the material properties are initialised by the code.
  3. Concerning the discrepancies among the pseudorapidity plots: these are due to the normalisation factor you can define in the USRYIELD card. Please note that scoring a double differential implies that the counts are divided by the binning width of the two variables, and the pseudo rapidity interval widths vary notably from case to case. You can correct this either by applying the right normalisation factors before the code execution or by including these factor ad-hoc at the plotting stage in Flair via the Y-axis norm.

I hope this helps,
Best regards
Mario

4

Dear Mario,
Thank you very much!
Your answer was crystal clear, and I truly appreciate it.

Cheers,
Antonella