FLUKA: EMF-BIAS
Previous Index Next
EMF-BIAS
Sets electron and photon special biasing parameters, including leading
particle biasing region by region, and mean free path biasing material by
material
See also EMF, EMFCUT, LAM-BIAS
For SDUM
= LPBEMF (default):
WHAT(1)
> 0.0: leading particle biasing (LPB) is activated. Which
combination of leading particle biasing is actually set up
depends on the bit pattern of WHAT(1)
Let WHAT(1)
be represented as:
2^0xb0 + 2^1xb1 + 2^2xb2 + 2^3xb3 + 2^4xb4 +
2^5xb5 + 2^6xb6 + 2^7xb7 + 2^8xb8 + 2^9xb9
then the meaning of the ten bits is the following:
b0 = 1 --> LPB activated for bremsstrahlung and pair production
(old default)
b1 = 1 --> LPB activated for bremsstrahlung
b2 = 1 --> LPB activated for pair production
b3 = 1 --> LPB activated for positron annihilation at rest
b4 = 1 --> LPB activated for Compton scattering
b5 = 1 --> LPB activated for Bhabha & Moller scattering
b6 = 1 --> LPB activated for photoelectric effect
b7 = 1 --> LPB activated for positron annihilation in flight
b8 = 1 --> not used
b9 = 1 --> not used
Note that WHAT(1)
= 1022 activates LPB for all physical effects
(values larger than 1022 are converted to 1022)
< 0.0: leading particle biasing is switched off
= 0.0: ignored
WHAT(2)
> 0.0: energy threshold below which leading particle biasing is
played for electrons and positrons (for electrons, such
threshold refers to kinetic energy; for positrons, to total
energy plus rest mass energy)
< 0.0: resets any previously defined threshold to infinity (i.e.,
leading particle biasing is played at all energies)
= 0.0: ignored
This value can be overridden in the user routine UBSSET (see 13})
by assigning a value to variable ELPEMF
Default
: leading particle biasing is played at all energies for
electrons and positrons
WHAT(3)
> 0.0: energy threshold below which leading particle biasing is
played for photons
< 0.0: resets any previously defined threshold to infinity (i.e.,
leading particle biasing is played at all energies)
= 0.0: ignored
This value can be overridden in the user routine UBSSET by
assigning a value to variable PLPEMF.
Default
: leading particle biasing is played at all energies for
photons
WHAT(4)
= lower bound (or corresponding name) of the region indices where
the selected leading particle biasing has to be played
("From region WHAT(4)
...")
Default
= 2.0
WHAT(5)
= upper bound (or corresponding name) of the region indices where
the selected leading particle biasing has to be played
("...to region WHAT(5)
...")
Default
= WHAT(4)
WHAT(6)
= step length in assigning indices
("...in steps of WHAT(6)
")
Default
= 1.0
SDUM
= LPBEMF (Leading Particle Biasing for EMF). This is the
default, for other values of SDUM
see below.
This value can be overridden in the user routine UBSSET by
assigning a value to variable LPEMF
For SDUM
= LAMBEMF, LAMBCOMP, LAMBBREM, LBRREMF, LBRRCOMP, LBRRBREM:
(not yet implemented for photons)!
WHAT(1)
> 0.0 and < 1.0: the interaction mean free paths for all electron
and positron electromagnetic interactions (SDUM
= LAMBEMF), or
for electron/positron bremsstrahlung only (SDUM
= LAMBBREM) are
reduced by a multiplying factor = WHAT(1)
= 0.0: ignored
< 0.0 or >= 1: resets to default (no mean free path biasing for
electrons and positrons)
WHAT(2)
> 0.0 and < 1.0: the interaction mean free paths for all photon
electromagnetic interactions (SDUM
= LAMBEMF), or for Compton
scattering only (SDUM
= LAMBCOMP) are reduced by a multiplying
factor = WHAT(2)
= 0.0: ignored
< 0.0 or >= 1: resets to default (no mean free path biasing for
photons)
WHAT(3)
= generation up to which the biasing has to be applied
Default
: biasing is applied only the first generation (i.e., the
primary BEAM or SOURCE particles)
WHAT(4)
= lower bound (or corresponding name) of the indices of materials
in which the indicated mean free path biasing has to be applied
("From material WHAT(4)
...")
Default
= 3.0
WHAT(5)
= upper bound (or corresponding name) of the indices of materials
in which the indicated mean free path biasing has to be applied
("... to material WHAT(5)
...")
Default
= WHAT(4)
WHAT(6)
= step length in assigning indices
("...in steps of WHAT(6)
")
Default
= 1.0
SDUM
= LAMBEMF (LAMbda Biasing for ElectroMagnetic FLUKA): mean
free path biasing is applied to all electron, positron
and photon interactions, and both the incident and the
secondary particle are assigned a reduced weight
LAMBCOMP (LAMbda Biasing for Compton interactions): mean free
path biasing is applied only to photon Compton effect,
and both the incident photon and the secondary electron
are assigned a reduced weight
LAMBBREM (LAMbda Biasing for BREMsstrahlung interactions): mean
free path biasing is applied only to electron and
positron bremsstrahlung, and both the incident
electron/positron and the secondary photon are assigned
a reduced weight
LBRREMF (Lambda Biasing with Russian Roulette for
ElectroMagnetic FLUKA): mean free path biasing is applied
to all electron, positron and photon interactions, and the
incident particle either is suppressed or survives with
the same weight it had before collision, depending on a
random choice
LBRRCOMP (Lambda Biasing with Russian Roulette for Compton
interactions): mean free path biasing is applied only to
photon Compton effect, and the incident photon either is
suppressed or survives with the same weight it had before
collision, depending on a random choice
LBRRBREM (Lambda Biasing with Russian Roulette for
BREMsstrahlung interactions): mean free path biasing is
applied only to electron and positron bremsstrahlung, and
the incident electron/positron either is suppressed or
survives with the same weight it had before collision,
depending on a random choice
Default
: LPBEMF (see above)
Default
(option not requested): none of the above biasings apply.
Notes:
1) Depending on the SDUM
value, different kinds of biasing are applied
to the secondary particles issued from the reaction.
2) If SDUM
= LPBEMF, the interaction point of electrons, positrons and
photons is sampled analogically and Leading Particle Biasing is
applied to the secondary particles.
3) With all other values of SDUM
, the interaction point is sampled from
an imposed (biased) exponential distribution, in a manner similar to
that provided by option LAM-BIAS for hadrons and muons. Further
differences in SDUM
values allow to restrict biasing to one specific
type of interaction and/or to select different treatments of the
incident particle.
4) If SDUM
= LAMBEMF, LAMBCOM, LAMBBREM, the weights of both the
incident and the secondary particle are adjusted according to the
ratio between the biased and the physical interaction probability
at the sampled distance.
5) If SDUM
= LBRREMF, LBRRCOM, LBRRBREM, the suppression or survival
of the incident particle (with unchanged weight) is decided by
Russian Roulette with a probability equal to the ratio between the
biased and the physical interaction probability at the sampled
distance. The weight of the secondary particle is adjusted by the
same ratio.
6) When using option EMF-BIAS, and in particular when choosing the
Russian Roulette alternative, it is suggested to set also a weight
window (cards WW-FACTOR and WW-THRESh) in order to avoid too large
weight fluctuations.
7) LAMBCOMP (LBRRCOMP) and LAMBBREM (LBRRBREM) are synonyms: i.e.,
input concerning photon interaction biasing given with SDUM
=
LAMBBREM (LBRRBREM) is accepted and treated in the same
way as with SDUM
= LAMBCOMP (LBRRCOMP); and input concerning
electron/positron interaction biasing with SDUM
= LAMBCOMP
(LBRRCOMP) is the same as with LAMBBREM (LBRRBREM).
This allows to issue just a single EMF-BIAS card requesting both
electron and photon interaction biasing at the same time.
8) Option EMF-BIAS concerns only electromagnetic interactions;
photonuclear interaction biasing is provided by option LAM-BIAS.
9) Leading particle biasing (LPB):
Leading particle biasing (available only for electrons, positrons
and photons) is generally used to avoid the geometrical increase
with energy of the number of particles in an electromagnetic shower.
It is characteristic of all electromagnetic interactions that two
particles are present in the final state: when this option is
selected, only one of them (with a probability proportional to its
energy) is randomly retained and its weight is adjusted accordingly.
Derived from the EGS4 implementation [Nel85], it has been modified
to account for the indirectly enhanced penetration potential of
positrons due to the emission of annihilation photons. The
probability of each of the two particles to be selected is therefore
not proportional to their kinetic energy but rather to their
"useful" energy (kinetic plus - in the case of positrons only -
twice the mass energy).
The weight of the particle selected is adjusted multiplying it by
the inverse of the selection probability. This kind of biasing is
aimed at reducing the mean computing time per history rather than
the variance of the scored quantities (computer cost is defined as
the product of variance times the computing time per primary
particle). It is mainly used to estimate shower punchthrough (but
comparable and even better efficiency can be obtained with
importance splitting, see BIASING), or to reduce the time spent in
simulating secondary electromagnetic showers produced by pi0 in
hadronic cascades. As any other kind of biasing, leading particle
biasing must be used with judgement, since it may lead to a slower
convergence of dose estimation in some regions of phase
space (see Note 5 to option BIASING). In particular, the fact that
the particle of highest energy is selected preferentially can have
the following effects:
a - the radial profile of the electromagnetic shower might be
reproduced less efficiently. This is in general not very
important for showers generated inside hadronic cascades, since
the overall lateral spread is governed essentially by hadrons.
b - a few low-energy particles might result with a very large weight
giving rise to strong energy deposition fluctuations (this
inconvenience can be limited to some extent by the use of a
weight window). Therefore, biasing should be avoided in scoring
regions and in adjacent ones, especially when using energy
deposition bins of very small volume.
When applied in energy deposition calculations, the use of weight
windows is recommended in order to avoid large local dose
fluctuations (see WW-FACTOR and WW-THRESh).
Example 1 (number-based):
EMF-BIAS 152. 0. 5.E-4 16. 20. 2.LPBEMF
The same example, name-based:
EMF-BIAS 152. 0. 5.E-4 Rsixteen Rtwenty 2.LPBEMF
Example 2 (number-based):
EMF-BIAS 1022. 0.0 0.0 3.0 8.0
The same example, name-based:
EMF-BIAS 1022. 0.0 0.0 thirdReg eighthRg
Example 3 (number-based):
EMF-BIAS 1022. 0.0 0.0 1.0 15.0
EMF-BIAS -1. 0.0 0.0 7.0 11.0 2.0
WW-FACTOR 0.5 5.0 1.0 1.0 15.0
WW-FACTOR 0.5 5.0 0.2 7.0 11.0 2.0
WW-THRESH 1.0 0.001 20.0 3.0 4.0
WW-THRESH 1.0 1.E-4 20.0 7.0
The same example, name-based, assuming that the 15 regions are all the
regions of the problem:
EMF-BIAS 1022. 0.0 0.0 first @LASTREG
EMF-BIAS -1. 0.0 0.0 seventh eleventh 2.0
WW-FACTOR 0.5 5.0 1.0 1.0 15.0
WW-FACTOR 0.5 5.0 0.2 seventh eleventh 2.0
WW-THRESH 1.0 0.001 20.0 ELECTRON POSITRON
WW-THRESH 1.0 1.E-4 20.0 PHOTON
Previous Index Next
