Re: FLUKA - sum files not reporting the final units after applying

Thank you Alberto for the corrections, I appreciate you taking the time.

In this section, I definitely did not mean to say
PSv/GeV/cm^2/prim but PSv/GeV/prim when plotting Y
and
PSv/prim for Y*DX or the integral over energy bins.
It was just a typo from my part. Sorry about that.
It is also my second day on the master cleanse fast and perhaps I am feeling the side effects! Should have read over my
message.

>>>
Here you are using a USRBDX estimator, scoring DOSE-EQ from neutrons crossing a boundary between regions sp15 and
DOS318. The area of the boundary has to be set (which in your case is 27925.268 cm^2). When you plot this distribution
in flair, the units are in PSv/GeV/cm^2/prim (if you plot Y for Y) and PSv/cm^2/prim if you plot (Y*DX for Y). The sum
of the DOSE-EQ reported in the sum.lis file should be in pSv/GeV/prim.
>>>

On another note, when one is looking at the DOSE-EQ, does the binning affect the answer, i.e. (number of bins between
min and max energies) and whether one selects lin or log scale under "Type"?

Thank you and best wishes,
Mina

On 12-06-04 01:24 PM, Alberto Fasso' wrote:
> Hi Mina and Julie,
>
> I think that it would be useful to clarify a few points.
>
>> If you used Neutrons instead of DOSE-EQ in Part and without the AUXSCORE card,
>> you would be scoring Neutrons crossing the boundary. The units in sum.lis
>> file would be Neutrons/cm^2/prim.
>
> Correct. But to understand it better, you better read it as
> Neutrons*cm/cm^3/prim.
> Why? Because fluence's dimensions cm^-2 are actually cm/cm^3: fluence
> is tracklength density, or tracklength per cm^3.
> The volume in which fluence is averaged (or tracklength is summed)
> is not necessarily finite: on the contrary, the formal definition of fluence
> is "fluence at a point", i.e. in an infinitesimal volume. With USRTRACK,
> the fluence is integrated over a region volume.
> The fact that you score fluence at a boundary (instead of in a volume) doesn't
> change the basic meaning of fluence as tracklength density. The area of the
> boundary is indeed a volume of infinitesimal thickness. Look at Fig. 6 in
> https://www.fluka.org/free_download/course/jlab2012/Lectures/07_Scoring_0412.pdf
> Then what I am going to say will be more clear.
>
>> If you don't give the area of the boundary (def.
>> 1), the units are Neutrons/prim. If you know the primary beam intensity
>> (prim/sec), you can give this as a normalization factor in the Norm box and
>> you'll get Neutrons/sec.
>
> No. If you give the area of the boundary, the units are Neutrons*cm/cm^3/prim,
> namely total neutron path per primary in an infinitesimal volume centered
> around the boundary. If you don't give the area, the units are still the same,
> but with the area assumed to be = 1.
> "Neutrons/prim" has no meaning: it could have a meaning if you scored a
> current (number of neutrons crossing the boundary). But if you score a fluence,
> you should write perhaps "Neutrons/cos(theta)/prim, which is not very nice.
>
>> Here you are using a USRBDX estimator, scoring DOSE-EQ from neutrons crossing
>> a boundary between regions sp15 and DOS318. The area of the boundary has to
>> be set (which in your case is 27925.268 cm^2). When you plot this
>> distribution in flair, the units are in PSv/GeV/cm^2/prim (if you plot Y for
>> Y)
>
> No. In this case, the units are pSv/GeV/prim. The cm^2 gets cancelled because
> the program has multiplied Neutrons/GeV/cm^2/prim by a conversion coefficient
> with units pSv/(Neutrons/GeV/cm^2): dose equivalent per unit fluence.
>
> I call your attention to the fact that, while it is always useful to look
> for an energy spectrum of fluence, it is usually less useful to look for
> a "spectrum of dose equivalent". What you want is its integral between two
> energies (the units for your integral dose equivalent will be pSv/prim,
> and you get rid of "GeV").
>
>> and PSv/cm^2/prim if you plot (Y*DX for Y).
>
> No. cm^2 disappears in the conversion, as explained above.
>
>> The sum of the DOSE-EQ
>> reported in the sum.lis file should be in pSv/GeV/prim.
>
> Correct, this time.
>
>> When scoring DOSE-EQ, you should ignore the units in the sum.lis and tab.lis files since these units do not take into
>> account the conversion factors applied (PSv*cm^2). I would assume this would be true for other quantities where a set
>> of conversion factors is applied via an AUXSCORE card. This is very confusing for a beginner (and in my humble opinion
>> not a good practice). I would hope future changes in FLUKA would correct this feature.
>
> I understand that you are annoyed by the units used in the title of the sum.lis
> and tab.lis files. But please keep in mind that the usxsuw and ustsuw auxiliary
> programs used by Flair to analyze the FLUKA binary output are designed to
> calculate fluence. They cannot "know" if you have multiplied the fluence by
> a conversion function. It is up to the user to account for that.
>
>> As a side, it is a good practice to look at the particle fluence in a given
>> volume or crossing boundaries as well as the DOSE-EQ, just to see what the
>> "raw" distribution looks like before any conversion coefficients are applied.
>
> Very good suggestion.
>
> Alberto
>
>
>>
>> On 12-06-04 01:40 AM, Julie Barbier wrote:
>>> Hi Mina,
>>>
>>> I am a new Fluka user and I saw you asked the same question I have now.
>>> I actually want to observe neutron ambient dose equivalent and I used :
>>> USRBDX 101. DOSE-EQ -38. sp15 DOS318 27925.268bdx575
>>> USRBDX 1. 0. 100. 6.283185 0. 6.&
>>> AUXSCORE USRBDX NEUTRON bdx1 bdx575 AMB74
>>> In the sum.lis it's indicates :
>>> Tot. resp. (Part/cmq/pr) 1.6257963E-03 +/- 0.6990718 %
>>> ( --> (Part/pr) 2.837550 +/- 0.6990718 % )
>>> Should I understand that is in fact pSv/cm/pr and pSv/pr ? Or is it the "real" fluence ?
>>>
>>> Best regards,
>>>
>>> Julie
>
Received on Tue Jun 05 2012 - 09:19:50 CEST