Dear FLUKA experts,
I am running a simulation that bombards a sodium iodide detector with aluminum casing with photons at various energies. When scoring the particles using the USRBDX card, the expected output value in the *_sum.lis file should be around 4.3E-4. I get this expected value for all photon energies except for 9MeV and 700MeV where the value is instead of magnitude E-5. I do not understand why these two input photon energies produce this discrepancy. I’ve attached my input file. If anyone would be able to help me understand it would be much appreciated.
detector.inp (6.7 KB)
Dear @tkoletic ,
Thank you for your question.
After having a look to your input file, I have noticed you use 1E6 primaries in your simulation. Obtaining 1E-5 particles/primary when scoring fluence over a surface means that only ten particles cross the surface in your simulation. Hence the statistics is considerably low if you use only 1E6 primaries. I ignore the number of cycles and repetitions you use in your simulation, which can effectively increase your statistics.
I have run your input file locally, using 5 cycles, obtaining the following outcome, and as you can see the statistical error is large:
Tot. resp. (Part/cmq/pr) 6.0800005E-05 +/- 8.984540 %
( --> (Part/pr) 6.0800005E-05 +/- 8.984540 % )
However, using 10 times more primaries (5 cycles x 1E7 primaries), the error is reduced and the scored quantity seems to converge towards what you might expect:
Tot. resp. (Part/cmq/pr) 4.3094004E-04 +/- 0.6569377 %
( --> (Part/pr) 4.3094004E-04 +/- 0.6569377 % )
Therefore, I suggest you increase the number of primaries. Besides, it seems you are using an isotropic source considerably far from the detector. This means that most of the primaries simply do not interact with the detector volume, explaining why so many primaries are needed.
I hope this helps.
Thanks for your fast reply.
I changed the number of primaries and did multiple cycles which helped with the error, but I am still not getting the 4E-4 value that you seem to have got. Could you share the input file you modified?
Dear @tkoletic ,
I simply ran your input file using Flair user interface, which by default defines 5 cycles. Initially exactly the same file is used and secondly the number of primaries is changed from 1E6 to 1E7. I did not change any other field.
However, I am revisiting the file right now since I do not get the same results either. I will come back to you as soon as possible.
Thank you for your patience. After careful study of your input file, it seems the problem may come from the fact that there are no explicit energy boundaries in the USRBDX scoring card. In such case, the default values are Emin = 0 and Emax = Eprimary. Besides, given the high energy photons you are using and the thin slabs of material, it can happen that photons arrive to the detector unaffected.
These two points might be leading to a FLOAT==FLOAT evaluation which is in general computationally challenging and, for the case of 700 MeV, seems to be larger than the binning maximum limit, hence not scored. I have successfully obtained the 4E-4 result by setting Emin = 0, Emax = 1 GeV.
Please let me know if this is helpful.
That is very helpful and has fixed my issue. Thank you for taking the time to help me.