Re: Electron fluence unexplained edges

Dear George,

> 1. What is the reason for the edges in electron spectrum?

as you might have realized in the meantime, the reason is the high
production threshold you set for photons.
The electron energy loss by bremsstrahlung reflects the smooth photon
spectrum between the photon production threshold and the electron energy,
while below it takes a fixed value equal to the mean energy of the
spectrum between zero and the photon production threshold. This value adds
to the dE/dx and the sum fluctuates according to Landau.
So, if you wish to inspect the electron spectrum, you should properly
lower the photon production threshold, leaving a high transport threshold
in case you do not care about low energy photons.

> 2. What is the difference betweenBEAMPART and LLO=1?

None.
The logic implemented in your fluscw is wrong, since you include in the
first generation scoring LLO=1 and LLO>4. The IF statement(s) should read

        IF(JSCRNG.EQ.1.AND.LLO.NE.1) LSCZER = .TRUE.
        IF(JSCRNG.EQ.2.AND.LLO.NE.2) LSCZER = .TRUE.
        IF(JSCRNG.EQ.3.AND.LLO.NE.3) LSCZER = .TRUE.
        IF(JSCRNG.EQ.4.AND.LLO.NE.4) LSCZER = .TRUE.

> 3. Why is the BEAMPART spectrum equal to 0 below 490 MeV?

After a Moller scattering, on the basis of the two electron
indistinguishability, the higher energy electron is considered as the
primary one. Thus the outgoing primary cannot have less than half of the
incident primary energy. On the other hand, a bremsstrahlung event
increases the generation number (i.e. the final electron is no longer
primary).

Best regards

Francesco

**************************************************
Francesco Cerutti
CERN-EN/STI
CH-1211 Geneva 23
Switzerland
tel. ++41 22 7678962
fax ++41 22 7668854

On Thu, 19 Aug 2010, George Kharashvili wrote:

> Dear FLUKA experts,
>
> We modeled interaction of 1-GeV electrons with a 5-cm radius, 10-cm long
> cylindrical gold target (FLUKA input file attached "primaries.inp").
> Electron fluence in the target region is scored using USRTRACK card in
> 1000 bins between 8 MeV (which is also e-/e+/photon energy cutoff set using
> EMFCUT card ) and 1 GeV. The electron spectrum displays unexplained (to
> us) edges at approximately 50, 110, 210, 390, 700 MeV. You can see this in
> "allE.eps"; file in the attachment. Can these edges be explained?
>
> When we scored just the primary particles using USRTRACK and
> AUXSCORE (BEAMPART option) there were no edges and no electrons scored
> below approximately 490 MeV (primaries_beampart.eps). Is this also strange?
> In order to track down where these edges were coming from, we modified a
> user subroutine fluscw.f to distinguish between 1st, 2nd, 3rd, and 4th
> generation electrons ("llofluscw.f"). New input file is "llo.inp".
> Fluences of all 4 different generations of electrons are plotted in
> "llo_Efluences.eps". The edges are present in all 4 cases and
> also the first generation electron fluence is different from the
> fluence scored using AUXSCORE BEAMPART. Is BEAMPART different from first
> generation particles defined by LLO=1? (Using LTRACK instead of LLO
> didn’t change anything).
>
> I should also mention that turning off Bhabha/Moller scattering
> did not affect the edges and made almost no difference in case of 2nd,
> 3rd and 4th. However, 1st generation electron fluence decreased noticeably
> between 0.940 MeV and 970 MeV.
>
> So the three main questions are:
> 1. What is the reason for the edges in electron spectrum?
> 2. What is the difference betweenBEAMPART and LLO=1?
> 3. Why is the BEAMPART spectrum equal to 0 below 490 MeV?
>
> Thank you,
>
> - George
>
Received on Tue Aug 24 2010 - 19:47:05 CEST