Re: [fluka-discuss]: Energy-scoring per region and R-Phi-Z using USRBIN

From: Luigi Salvatore Esposito <luigi.salvatore.esposito_at_cern.ch>
Date: Fri, 22 Jul 2016 13:31:39 +0000

Dear Kilian,

first of all, let me say that FLUKA has been tested extensively and there are no differences in the output
of a scoring by regions or of a scoring with a geometrical binning.
More than that, no differences will magically appear by changing the number of primaries.
Of course, the values of the result and its error change, the first converging to the “true value” and the second decreasing with the
inverse of the square root of the primaries.

The *.dat files are produced by Flair and generated for plotting: his author decided that in one case was better to use
percentage error and in the other case the decimal one and also to print the results with a different number of significant digits.

In conclusion, any difference you find between the two scorings is certainly due to a wrong post-processing of the files.
There is a Clean button also in the Plot tab of Flair as well…

Best, luigi

On 22 Jul 2016, at 14:46, Kilian Baumann <baumann.kilian_at_gmail.com<mailto:baumann.kilian_at_gmail.com>> wrote:

Dear Luigi,
thank you very much for your answer!
I did clean the produced data files and reprocessed them and the interesting thing is:
For 5 runs à 10^5 primaries the results (concerning the statistics) are the same for both USRBINs.
However for 5 runs à 10^6 primaries the results (concerning the statistics) differ from each other (see attached files).
And furthermore: when using flair to "plot" the results, the *.dat-file is created, where for the R-Phi-Z USRBIN the error is given in percent (3.259) and for the region-USRBIN the error is given in decimal (0.029)....
Best,
Kilian

Am 20.07.2016 um 18:14 schrieb Luigi Salvatore Esposito:
Dear Kilian,
I used your input and Flair to make a check and as you can see from the attached files the errors are identical.
In Flair try to clean the data files produced after the merging and reprocess them again.
Best, luigi




> On 20 Jul 2016, at 18:05, Kilian Baumann <baumann.kilian_at_gmail.com><mailto:baumann.kilian_at_gmail.com> wrote:
>
> Dear Luigi,
>
> thank you very much for your answer.
> I was using flair for the merging of the data, so the uncertainties should be right... ?
>
> Best,
> Kilian
>
> Am 20.07.2016 um 16:06 schrieb Luigi Salvatore Esposito:
>> Dear Kilian,
>> for a cylindrical scoring the result is in GeV/cm3, i.e. it is divided by the bin volume,
>> whereas in the scoring by region the result is in GeV, i.e. it is not divided by the region volume.
>> See note 13 of the USRBIN section in the manual.
>> Concerning the error, they are normally the same, therefore I suspect there is an error on your side.
>> Best regards, luigi
>>
>>> On 20 Jul 2016, at 14:45, Kilian Baumann <baumann.kilian_at_gmail.com><mailto:baumann.kilian_at_gmail.com> wrote:
>>>
>>> Dear all,
>>>
>>> scoring the energy in an air-filled cavity in a waterphantom irradiated by a 1.25 MeV photon beam I habe encountered an interesting problem: Using the USRBIN-card one can score the energy either in the region (called "chamber") or use the R-Phi-Z geometry that exactly matches my region called "chamber". So since I am scoring the energy in the same region (in different ways concerning my USRBIN) I should get the same results. However the results are different: 4.64538e-10 GeV pm 0.0290214% for the region-scoring and 5.15530840E-10 GeV pm 3.25998449% for the R-Phi-Z-scoring.
>>> Do you have an idea why this is so? Especially concerning the statistics that are quite different?
>>> In the attachment you can find my input-file.
>>>
>>> Best regards,
>>> Kilian
>>> <Dosimetry_chamber.inp>
>


<Dosimetry_chamber_35bnn.dat><Dosimetry_chamber_36.bnn.lis><Dosimetry_chamber_36bnn.dat><Dosimetry_chamber_35.bnn.lis>



__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Fri Jul 22 2016 - 17:03:01 CEST

This archive was generated by hypermail 2.3.0 : Fri Jul 22 2016 - 17:03:02 CEST