RE: [fluka-discuss]: RE: spectrum normalization and actual dose

From: Francesco Cerutti <>
Date: Wed, 30 Jan 2019 14:28:34 +0100

Dear Alex,

deposited energy [MeV] and tracklength [cm] scale with the volume (the
larger the volume, the larger their value), while dose [Gy], energy
density [GeV/cm3] and fluence [cm-2] don't, since they are the above ones
divided by the respective volume.

Now, technically:

USRTRACK divides by the respective region volume only if you put the
latter in the USRTRACK card. Otherwise, you can easily divide yourself as
you handle the results. Assuming that this division is made (by the code
or by yourself), if you eventually integrate over energy (something that
actually you do not really need to do, since you find it already done in
the _sum.lis file), you will indeed get GeV/cm^2 for ENERGY scoring (that
is a rather weird scoring choice for USRTRACK, since in that case it's not
deposited energy, rather particle kinetic energy). But you will get
photon/[cm^2] for PHOTON scoring, and not photon/[cm^2*MeV], since you
integrated over energy (DX*).
As for USRBIN region ENERGY scoring, if you divide by the region material
(air) density, you do not get dose [GeV/g], you get something [GeV*cm3/g]
that needs to be further divided by the region volume to become (the
average) dose.
Finally, the 3D integral (DX*DY*DZ*) of the fluence scored by USRBIN
Cartesian PHOTON gives definitely the same photon tracklength as obtained
by USRBIN Region PHOTON, provided that the region coincides with the
integral volume. If instead you plot in Flair the photon fluence as a 1D
distribution (e.g. in X) and you multiply by DX only, you are producing a
quantity of less evident meaning.

I think I answered at the beginning your first question. The ICRP units
refer to dose and fluence, which do not scale with the reference volume.

About "per primary", it does not refer to tracklength, but to the 'beam'
particle. This way, one can scale results with the actual source (beam)
intensity (i.e. physical number of primary particles).

I have no idea how you integrated the photon fluence spatial distribution.
As indicated above, a proper 3D integration must give back the tracklength
in the region, as it does.



Francesco Cerutti
CH-1211 Geneva 23
tel. +41 22 7678962

On Wed, 30 Jan 2019, Ševčik Aleksandras wrote:

> Thank you very much Francesco,
> Yes, now everything is starting to make sense. I know that all this is written in the manuals, however, until one starts to make the actual simulations, it sometimes hard to grasp some concepts at once. Just to be totally sure, I run kerma simulation for 1x1x1 rpp air body body with 1 MeV and emfcut. The value I got is equal to air kerma free-in-air value in ICRP119 so I suppose it’s ok. Now scoring:
> USTRACK ENERGY integrated fluence DX*Y and normalizing for MeV - the result is ok , y= 1 MeV/cm^2 per primary
> USTRACK PHOTON integrated fluence DX*Y , the result is ok , y=1 photon / [cm^2*MeV] per primary (hope I got the units correctly).
> USRBIN Region Photon - I got 1, so photon track length 1 cm
> USRBIN Region Energy - value in GeV for total region volume, dividing by air density gives GeV/g (dose(kerma)) equal to ICRP119 value.
> USRBIN X-Y-Z Energy - spatial distribution GeV/cm^3 per primary, integrating over it gives the same as USRBIN Region Energy.
> USRBIN X-Y-Z Photon - spatial distribution of photon fluence photons/cm^2 per primary. Integrating over it gives ~11 (?)
> Could you please:
> 1) When I run the same simulation with different volumes, I can see that kerma value is directly proportional to photon track length, but not to *average* photon fluence [cm-2] over the region. For example rpp 2x2x2 gives tracklength 2 cm, 4x4x10 gives tracklength 10 cm, while average photon fluences would be 2/8 and 10/160 accordingly. But the kerma value is exactly 10 times higher with larger volume. Why then ICRP gives this value in pGy*cm^2 ? Just trying to understand the rationale behind the measurement units,
> 2) In measurement units, when we state "per primary", what does it actually, in layman terms, means? Does it relates to tracklength as well?
> 3) Why integrating USRBIN X-Y-Z Photon I got 11 ?
> I will greatly appreciate your help as in my educational environment I work only with experimental guys who don't have a clue about actual mc modelling.
> Regards
> Alex
> -----Original Message-----
> From: Francesco Cerutti <>
> Sent: Wednesday, 30 January, 2019 10:36
> To: Ševčik Aleksandras <>
> Cc:
> Subject: Re: [fluka-discuss]: RE: spectrum normalization and actual dose
> Hi Alex,
> you do not get the photon number, you get a more meaningful quantity that is the photon tracklength [cm], namely the path travelled by photons over that region. If you divide this quantity by the region volume, you get the *average* photon fluence [cm-2] over that region. Other kinds of USRBIN (Cartesian and cylindrical) give you directly photon fluence, since they already divide by the respective regular bin volume. If in turn you multiply that by the latter, then you get again photon tracklength. All physical effects (energy deposition, i.e. dose, and detector response) depend on the particle tracklength, not the particle number.
> Kind regards
> Francesco
> **************************************************
> Francesco Cerutti
> CH-1211 Geneva 23
> Switzerland
> tel. +41 22 7678962
> On Wed, 30 Jan 2019, Ševčik Aleksandras wrote:
>> Dear all,
>> It seems that my question most likely unclear, so let me rephrase it
>> to be more specific and technical:
>> When scoring PHOTONS using usrbin region, do you get the total or
>> average photon number in the region for that specific simulation? Then
>> total number of photons in the region can be get by integrating usrbin
>> photon x-y-z scoring?
>> A.
>> From: <>
>> On Behalf Of Ševcik Aleksandras
>> Sent: Friday, 25 January, 2019 01:13
>> To:
>> Subject: [fluka-discuss]: spectrum normalization and actual dose
>> Dear experts,
>> I would like to get the actual dose values from the simulation and
>> want to know if the approach below is valid.
>> I am using source.f to generate realistic x-ray beam from specialized
>> x-ray simulation software which gives me the spectrum and fluence.
>> Integrating it I can get the estimated actual number of photon, let’s
>> say 10E7. In the simulation I can use USRBIN for all regions to score
>> the overall number of photons detected in the simulation, let’s say
>> 10E2. So in the same simulation I can use this factor 10E7/10E2 for
>> dose values as well to get the actual dose values for the specific x-ray tube.
>> I have tried to run the test simulation with 100x100x100 water cube in
>> the vacuum. In the beginning I was integrating usrtrack fluence energy
>> spectrum but later saw that usrbin region photon scoring gives the
>> same value.  So multiplying usrbin total dose value by this fluence
>> actual/simulated factor and converting it to Gy gives me around 7E5 Gy
>> – as it was expected from 80 kvp  *1 mAs_at_1meter x-ray beam.
>> However if this approach is not valid, I would appreciate any comment
>> what would be the correct way to do this.
>> Regards
>> Alexander

You can manage unsubscription from this mailing list at
Received on Wed Jan 30 2019 - 15:38:44 CET

This archive was generated by hypermail 2.3.0 : Wed Jan 30 2019 - 15:38:52 CET