Dear users
I would like to point to an important issue concerning dose conversion
factors. It has been reported that in particular cases equivalent doses
scored with DOSE-EQ and fluka2023.3 are significantly larger than those
scored with DOSE-EQ and fluka2021.2 in the same setup.
Those increases, sometimes large, are "real" and not due to physics,
or fluka2023.3, rather the culprits are ICRP/ICRU...
Let me try to explain since this is a "surprise" which could occur
at whichever electron facility.
ICRU95 changed the operational quantity from Ambient Dose Equivalent to
Ambient Dose, and as a consequence Fluka2023.3 uses by default Ambient
Dose conversion coefficients (H*, AMBDS in FLUKA jargon) instead of
Ambient Dose Equivalent (H*(10), AMB74 in FLUKA jargon) ones, unless
the user specifies otherwise.
Ambient Dose (H*), despite the similar name, is a completely different
quantity wrt Ambient Dose Equivalent (H*(10)). The latter was defined as
the equivalent dose measured at 1 cm depth on the axis of the ICRU sphere
for a parallel beam etc etc. Ambient Dose is the maximum Effective Dose
among the possible irradiation geometries (AP, PA, ROT). As such it is NOT
measurable, it is rather computable if one can measure/simulate the
radiation field, by folding the measured/simulated radiation field with
the ICRP recommended conversion coefficients for Effective Dose, exactly
how it is done in the Monte Carlo.
Quoting from ICRU95:
"The quantity for assessing potential effective doses in a
given area (predominantly prospectively) is named ambient
dose H*, and is defined at all energies of the particles consid-
ered as the maximum of the values of effective dose for the
different directions of an incident radiation field on an
anthropomorphic phantom as published in ICRP Publication
116 (2010)." (end quotation)
One of the reasons for this change was the criticism to Ambient Dose
Equivalent put forward by M.Pelliccioni and myself who demonstrated in
the 90's that Ambient Dose Equivalent was not at all conservative
wrt Effective Dose for high energy electrons and photons.
The new conversion coefficients can be MUCH larger at GeV energies wrt
the old ones, to get a feeling look at page 118 of ICRP116, fig. 5.2, where
the maximum of the possible effective dose irradiation geometries (which
is now by definition the Ambient Dose), is at 10 GeV photon energy a
factor 23 (twentythree) higher than the old H*(10).
In specific problems, where for example a magnetic field is wiping away
the charged particles and hence the radiation field mostly consist of
photons, the energetic ones will have much larger conversion coefficients
for H* wrt the "old" ones for H*(10). A similar effect exists for electrons
as well, even though less dramatic (a factor 2 at 10 GeV), look at
figure 5.4 in ICRP116.
I hope this explain why in some situations with GeV electron/photon beams
Ambient Dose can come up significantly higher than Ambient Dose Equivalent.
Of course, concerning FLUKA calculations, one can still use AMB74 (the old
H*(10) conversion coefficients), by using an AUXSCORE card, instead
of the AMBDS new default. However if/when your applicable regulatory
authorities will adopt the ICRU95 recommendations, one should move to
H* (Ambient Dose). As a consequence a shielding which was good enough with
H*(10) in some circumstances could no longer be adequate wrt H*.
Computed gas bremsstrahlung dose equivalents can be of course quite affected
if one uses H* rather than H*(10) with multi-GeV beams.
All the above is due to the changes in the operational quantities
dictated by ICRU, not to Fluka. On the contrary, fluka2023.3 contains
completely new (and much better!) physics for photonuclear interactions
above ~100 MeV, therefore for electron facilities in excess of that energy,
and where photoneutrons are an issue, it is highly recommended to upgrade
to fluka2023.3 as soon as possible.
Ciao
Alfredo
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Sat Oct 21 2023 - 12:24:23 CEST