University Of Cape Town

Radiation Protection Assignment Two

Radiation Transport Report
2013
Courage Mahuvava (MHVCOU001)
April 7, 2013

Question 1 : Radiation transport theory

Part (a)
Present a brief overview of the strengths and weaknesses of deterministic and
stochastic (Monte Carlo) radiation transport methods.

Monte Carlo refers to a statistical method wherein expected characteristics of particles (e.g.
particle fluence) are estimated by sampling a large number of individual particle histories whose
trajectories are simulated by a digital computer whilst deterministic techniques involve nu-
merical solutions to the Boltzmann transport equation (BTE), that is, the radiation field is
characterised by functions describing the phase-space density of particles at a given point and
its distributions with respect to energy, angle and particle type.

The fundamental advantage of Monte Carlo techniques over deterministic techniques is that
Monte Carlo techniques more accurately represent the geometry and the nuclear data than de-
terministic techniques. Deterministic techniques require reasonably elementary geometries for
the numerical technique to work, and use the multigroup approximation to cross-section data.
Monte Carlo techniques can handle complex geometries and continuous cross-section data as
well as simple geometry and multigroup data.

Monte Carlo simulations of radiation transport may be used to obtain response functions for
operational dose quantities for example the effective dose response function. Monte Carlo codes
and cross-section libraries also improve every decade. Monte Carlo codes under active devel-
opment improve quite significantly in about a 4-year development cycle. Monte Carlo also
allows calculation of any deterministic quantities as they average the corresponding one over
its probability distribution.

The disadvantage of the Monte Carlo technique is that it is statistical in nature and does
not provide an exact solution to the problem. All results represent estimates with associated
uncertainties. Also, Monte Carlo techniques can be quite time-consuming on a computer if
small uncertainties are required.

1
The relationship between Monte Carlo techniques and deterministic techniques is that deter-
ministic techniques provide a highly exact solution to a significantly simplified approximation
of the problem, while Monte Carlo techniques provide an approximate solution to a highly
exact representation of the problem.

Part (b)
Present a brief conceptual overview of the calculational methodology followed in
Monte Carlo radiation transport codes.

Monte Carlo method in determining the effect of water content on its neutron
shielding ability
• The Monte Carlo radiation transport code is used to model a hypothetical point source emit-
ting neutrons surrounding by a 100 cm thick spherical concrete sphere
• The effective dose on the outer surface of the sphere is tallied using MCNP surface fluence
rate tally modified by effective dose rate response functions for the neutrons and the ionizing
photons tally i.e. the source strength the energy spectrum of the emitted radiation
• The effective dose rates are calculated as a function of the water content of the concrete.
Taking into account the effect of the water content of the concrete on the concrete density.
• The mass density of ordinary concrete is then calculated as a function of its water content
using the normal partial density in ordinary concrete and also mass density of concrete at the
water content.
• The neutron energy spectrum is analysed, as well as the transmitted dose rate to determine
if concrete is an optimal shielding material.

Question 2 : Response functions

Part (a)
Present a rough sketch of the response functions <n (E) and <γ (E) for fluence rate
to effective dose rate, for neutrons (n) and photons (γ).

2
Figure 1: The above figure is a a rough sketch of the response functions <n (E) and <γ (E) for
fluence rate to effective dose rate, for neutrons (n) and photons (γ) .

Part (b)
Present a brief overview of how the above response functions <n (E) and <γ (E) are
determined by using the ICRP definitions for effective dose, ICRP radiation &
tissue weighting factors, and calculational anthropomorphic voxel phantom models
in Monte Carlo radiation transport calculations.

The response function for effective dose may be determined in two ways: experimentally or
by calculations employing accurate radiation transport methods. An anthropomorphic phan-
tom is placed in a radiation field, e.g. in a uniform, parallel field that is incident onto the
front of the phantom. This is called an anterior-posterior or AP-exposure geometry. A known
fluence-rate of mono-energetic radiation quanta having an energy E0 is incident on the phantom
as a parallel beam:

ψ(x, Ω̂, E) = ψ0 δ(E − E0 ) δ(Ω̂ − Ω̂0 )

where ψ0 is the particle fluence-rate (unit: cm−1 s−1 ) and Ω̂0 is the direction of movement of
the radiation quanta. This function describes a perfectly collimated, perfectly monoenergetic
incident beam in a vacuum. The average absorbed dose in every organ is now experimentally
measured or calculated. Equivalent doses to the organs are determined by weighting with
the ICRP’s latest set of recommended radiation weighting factors, wR . Finally, the effective
dose to the human body is determined by adding the equivalent doses to the different organs,
weighing the sum with the ICRP’s latest set of recommended tissue weighting factors, wT .
The experiment (whether using a physical anthropomorphic phantom or an anthropomorphic
calculational phantom used in a Monte Carlo simulation) is now repeated for different incident
energies, E0 , and possibly also for different exposure geometries. The calibration function

3
<effdose (E) that satisfies the relationship :

R = h<effdose , ψi

represents the response function for effective dose, for the exposure geometry of choice.

Question 3 : Physics of radiation shielding; shield optimisation

Part (a)
Referring to a supplied table, explain how the physics of the underlying reaction
mechanisms lead us to the conclusion that two material properties of a fixed thick-
ness shield against ionising photons must be maximised in order to maximise the
dose rate reduction provided by the shield. Name these two material properties.
Name 3 excellent shielding materials against ionising photons.

Ionising photons interact and lose energy by 3 principal mechanisms: photo-electric absorption
(PEA), Compton scattering (CS) and pair production (PP). The table shows the dependence
of the linear interaction coefficients on the two material properties, density (ρ) and atomic
number (Z) for the three main attenuation mechanisms for ionising photons.

Interaction Linear interaction coefficient µ

Photo-electric absorption (PEA) µP EA ∝ ρ( EZ0 )3
Compton scattering (CS) µCS ∝ ρ
Pair production (PP) µP P ∝ ρZ

It follows clearly from the above Table that there exists a simple recipe for maximising the
efficiency of a photon shield Maximise the atomic number, Z, and maximise the mass-density,
ρ, of the shielding material.

The general trend is that the linear interaction coefficient µ(E) for ionising photons, which
measures the ability of shielding materials to interact with ionising photons and thereby lower
their energy, increases as photon energy decreases.

µ(E) = µP EA (E) + µCS (E) + µP P (E) (1)

In the high photon-energy region, where pair production is dominant, both ρ and Z should
be maximised with equal weight to maximise shield efficiency. In the low-energy region, where
photo-electric absorption of photons is dominant, it pays significantly more to maximise Z than
to maximise ρ. In the intermediate photon energy region, where Compton scattering dominates,
the atomic number Z has very little effect on shield efficiency, and the parameter that should be
maximised to maximise shield efficiency, is the mass-density ρ of the shield. Gamma-photons
are very penetrating and best shielded by lead or depleted uranium. Materials of high mass
density, ρ, and high atomic number, Z, will be the best shields against ionising photons, these
are:
• depleted uranium (DU)
• tungsten (W)
• lead (Pb).

4
Part (b)
Briefly present the fundamental principles of neutron shielding and neutron shield
optimisation. Define the concept complementary shielding materials. Use rough
sketches in support of your presentation.

Neutron shielding principle 1: Slow neutrons down to make them easier to ab-
sorb
The general rule is that the lower a neutron’s energy, the easier is it absorbed; to ”get rid of”
neutrons, we first slow them down, because lowering its energy makes the neutron easier to
absorb and so to eliminate it from the radiation field. The secondary types of ionising radiation
produced by neutron interactions in matter such as e.g. photons via (n, γ) interactions, alpha
particles via (n, α) interactions etc are significantly easier to shield than neutrons themselves.
It is therefore essential to absorb neutrons in a shield; slowing neutrons down will make them
more absorbable.

Neutron shielding principle 2: Slow neutrons down to make them less harm-
ful
It is clear that the harmfulness of neutrons will become less and less as their energy E decreases.
The material damage done by low-energy neutrons is substantially less than the damage done
by fast neutrons. Detriment is therefore minimised by minimising the fast-neutron spectral
component of the neutron fluence-rate function φn (E) in the spatial region of interest, and this
is principally achievable by slowing down fast neutrons as effectively as possible. A fundamental
principle of effective neutron shielding follows from the above: select shielding materials that
will optimally slow neutrons down, because slow neutrons are (1) more easily absorbed and (2)
less harmful to people and materials, than fast neutrons. In this way, the fluence-rate function
φ(E) will be lowered, especially in the high-energy region, so that dosimetric functionals will
have lower values.

Neutron shielding principle 3: Use shielding materials rich in 1 H to efficiently

slow down neutrons with energies below circa 4 MeV , via elastic scattering
1
H is the most efficient isotope to slow neutrons down by elastic scattering-the kinematic ef-
ficiency of elastic scattering by 1 H as a mechanism for neutron slowing down, is very high:
the average energy loss for the neutron is 50% per elastic scattering event- the mass of the
neutron is practically the same as the mass of the proton that constitutes the nucleus of 1 H.
The number of successive elastic scattering events off a target nucleus with mass number E,
that is equally effective in slowing down neutrons, as a single elastic scattering of a neutron off
1
H is given by, 2
log( 21 )

A−1
n(A) = Where α(A) = (2)
log( 1+α(A) )
2
A+1
Thus, heavy elements are quite inefficient at slowing down neutrons by elastic scattering;

5
Figure 2: The number of successive elastic scattering events off a target nucleus with mass
number A, that will be equally effective in slowing down neutrons as a single elastic scattering
of a neutron with 1 H

Neutron shielding principle 4: The isotope 1 H is only able to effectively slow

down neutrons with energies below about 4 MeV; when higher energy neutrons
are to be shielded, a higher-Z element such as Fe must be included in the neutron
shield to slow neutrons down via inelastic scattering
1
H has a relatively high cross-section for scattering neutrons with energies below approximately
0.1 MeV, then the cross-section σ(En ) for elastic scattering by H begins to drop steeply with
increasing neutron energy. Even though the kinematic effectiveness of elastic scattering by 1 H
is high, the low cross-section σ(En ) at higher neutron energies, makes hydrogenous shielding
materials ineffective at higher incident neutron energies. High-Z elements such as Fe are used
to slow down the high-energy neutrons (which are outside the cross-sectional ”grasp” of elastic
scattering by 1 H)in the incident neutron field, by inelastic neutron scattering. The inelastic
scattering by the high-Z element lowers neutron energy to within the energy range where 1 H
has a proper cross-sectional ”grip” on them and can efficiently slow them down to low energies,
where they can be easily absorbed.

Neutron shielding principle 5: Use an isotope with substantial (n, p), (n, α) , (n, d)
(n,3 He) and (n, t) reaction cross sections to remove neutrons from the radiation
field, and with substantial (n, np), (n, nt) , (n, nd), (n, nα) and (n, 2n) reaction cross
sections to help to slow neutrons down
Nuclear reactions such as (n, p), (n, α) , (n, d) and (n, t) reactions, removes a highly penetrat-
ing, energetic neutron from the radiation field and replaces it with a minimally penetrating
light ion whilst nuclear reactions such as (n, np), (n, nt) , (n, nd) and (n, nα) serve as highly
effective neutron slowing down interactions, because the energy of the neutron in the exit chan-
nel will be significantly lower than that of the neutron in the entrance channel. The higher
these cross-sections are, and the lower their threshold energies, the more they can contribute
to neutron removal and slowing down. In (n, 2n), (n, 3n) neutron nuclear reactions, which are

6
always endo-thermic, energy conservation ensures that the energy of the neutrons in the exit
channel will be substantially lower than the neutron energy in the entrance channel. There-
fore even these neutron multiplication reactions can contribute to neutron slowing down, and
therefore to neutron attenuation.

Neutron shielding principle 6: Use heavier elements such as e.g. C and Fe

to scatter neutrons through large angles, i.e. to reflect neutrons
Heavy nuclides such as Fe are able to deflect neutrons through significantly larger angles than
light isotopes such as H. These ”large angle scatterers” such as Fe are valuable neutron reflec-
tors. 1 H can not scatter neutrons through angles larger than 90-only scattering angles between
0 and 90 are allowed, while heavier elements such as Fe can scatter a neutron through an angle
as large as 180 in a single scattering interaction. This better backscattering property is valu-
able in neutron shielding, because reflection through large scattering angles will make it more
difficult for the incident neutrons to penetrate through the shield.

Figure 3: The number of successive elastic scattering events off a target nucleus with mass
number A, that will be equally effective in slowing down neutrons as a single elastic scattering
of a neutron with 1 H

Neutron shielding principle 7: Suppress the production of energetic ionising pho-

tons produced by thermal neutron capture reactions in a shield, by mixing a slow-
neutron absorber such as boron carbide (B4 C) into the hydrogenous layer(s) of
the shield
The lower the energy of a neutron, the higher the value of the neutron absorption cross-section
σa (E) will be. The cross-sections for (n, γ) -reactions are much higher for low energy neutrons
than for high energy neutrons. Therefore one effective way to suppress the production of po-
tentially problematic ionising photons from (n, γ) -reactions, is to suppress the slow-neutron
fluence rate φ by strongly absorbing slow neutrons by mixing a slow-neutron absorber such as
e.g. B4 C or ammonium pentaborate into the hydrogenous layers of a neutron shield, whilst
producing only a low energy ionising photon, which is quite easy to absorb.

7
Neutron shielding principle 8: Suppress the production of radioactive neutron
activation products by including a slow neutron absorber such as boron carbide
(B4 C) in the neutron shield
The use of a thermal neutron absorber such as B4 C will lower the thermal neutron fluence-rate,
which can substantially reduce the reaction rates of all reactions having an elevated cross-section
at low neutron energies, because reaction rates are proportional to,
Z
dEφ(E)σ(E)

Thus, B4 C suppresses the production of radioactive neutron activation products which are
formed by (n, γ) reactions. This will lower radiation levels that will be encountered during
maintenance work in e.g. (1) radionuclide production vaults at a particle accelerator facility,
(2) in regions inside the containment of a PWR close to the reactor, and (3) in regions close to
the core of a MTR reactor.

Shield optimisation
The kind of shield optimisation problem that will be solved is: A given thickness is available
to place shielding materials in. Given that four shielding materials are available, namely paraf-
fin wax, boron carbide (B4C), iron and lead, determine thicknesses and placements of these
material layers that will minimise the transmitted dose rate for a given shield thickness. A tech-
nique that will be often used in this study, is parametric shield optimisation. A single material
property is varied as parameter, and a radiation transport calculation is done for each shield
configuration. The transmitted dose rate is plotted as a function of the material property that
is varied as parameter, and the optimum is the material configuration where the transmitted
dose rate is a minimum.

Complementary Shielding
When neutrons are an important component of the source-term, complementary shielding ma-
terials are needed, because no single shielding material is good enough in isolation. A combina-
tion of paraffin wax, Fe, a thermal neutron absorber such as boron carbide (B4C) and Pb, are
needed to build a really effective shield against neutrons with a wide range of energies. These
materials are called complementary shielding materials, because they complement each other
by compensating and making up for each other’s weaknesses. Where the one shielding material
is weak the other one is strong. For neutrons, one must use complementary shielding materials
to build an optimal shield. So materials must always be used in well-chosen combinations in
order to complement each other, to construct a truly effective neutron shield with an optimised
dose rate reduction ability for a constrained shield thickness.

Part (c)
A neutron shield is to be built. Your calculations show that laminated layers of
steel and paraffin wax will be the shield of choice. Describe how you will minimise
activation product formation, and minimise the production of energetic capture
gamma-rays, by (1) the selection of steel alloy and (2) by adding a suitable addi-
tive to the paraffin wax. Name this additive. Explain the underlying physics.

(1) Steel is made of varying quantities of Fe, C, Mn, P, S, Si, O, N, Al, etc. We will con-
sider Fe, C, Mn to explain their effect on the steel.

8
Iron is the main component in steel, so lets investigate how it interacts with neutrons: Fe
produces a relatively high abundance of high energy photons in (n, γ) reactions. Therefore to
decrease the number of energetic capture gamma-rays, we should decrease the amount of iron
in the steel.
The next most common element is Carbon, which inelastically scatters neutrons. Mild steel is
made up of up to 2.5% Manganese by mass, whose only stable isotope 55 M n has a significant
(n, γ) cross-section and will activate in a neutron field to produce 56 M n (half-life 2.58 hrs).
This presents an intense short-term field of ionising photons around Mn-alloy steel structures
that have been exposed to a high thermal neutron fluence rate. The activation product of
56
M n can delay the time before it is safe to begin with maintenance operations by up to 12 to
15 hours, due to the high dose rates caused by the presence of the neutron activation product
56
M n. When shielding neutrons, use a steel alloy with a low manganese (Mn) and cobalt (Co)
content in order to minimise the production of the unwanted radionuclides 56 M n and 60 Co via
neutron activation these activation products will lead to higher dose rates during maintenance
work.

(2) Production of energetic capture γ-rays is suppressed by mixing a neutron absorber, boron
carbide B4 C , into a hydrogenous material. The use of a thermal neutron absorber such as
B4 C will lower the thermal neutron fluence-rate, which can substantially reduce the reaction
rates of all reactions having an elevated cross-section at low neutron energies, because reaction
rates are proportional to, Z
dEφ(E)σ(E)

Thus, B4 C suppresses the production of radioactive neutron activation products which are
formed by (n, γ) reactions. The neutron capture reactions (”radiative neutron-capture reac-
tions’):
1
H + n → 2 H + γ(2.205MeV) and 10 B(n, α)7 Li + γ(0.48Mev)
are crucial in neutron shield design. Neutron sources require (inter alia) hydrogenous materials
for shielding. However, the first of the above listed reactions, produces relatively energetic
2.2 MeV capture gamma-photons. Therefore, secondary gamma radiation is produced in the
shield. The gamma photons produced by the second capture reaction have photons with an
energy of only 0.48 MeV, i.e. they are approximately 4.6 times less energetic than the capture
gamma-photons produced when 1 H absorbs neutrons. B4 C (boron carbide) is often used in
hydrogenous neutron shields. Boron-10 is a very useful absorber of neutrons, especially slow
neutrons. Because it is rather impractical to mix the boron carbide into metallic layers, it
is normally only mixed into the hydrogenous shield layers. The gamma-photon emitted upon
neutron capture, has a relatively low energy, and is quite easily absorbed. In this way, adding
boron to a neutron shield suppresses the production rate of high-energy ionising photons. The
mixing ratio of B4 C in paraffin wax or polyethylene is determined by the requirement that the
reaction rates, and accordingly, the linear interaction coefficients for the favourable and the
µ10
unfavourable absorption reactions should be related as µ1 B(n,α) ≈ 25. Thus the ideal volume
H(n,γ)
percentage of B4 C in paraffin wax or polyethylene should be between circa 2.5% and 5 %.

Part (d)
(i) Why is a thick Fe shield very good at ”eliminating” neutrons having E & 0.9MeV,
from the spectrum of neutrons emerging on the personnel side of the shield?
Iron slows down neutrons by inelastic scattering- the excitation of every discrete energy level in

9
the 56 Fe nucleus by inelastic scattering of neutrons has a characteristic threshold energy value
: which lies at about 0.861 MeV for the excitation of the first discrete energy level in 56 Fe via
inelastic neutron scattering. Fe can therefore effectively retard neutrons with energies above
0.861 MeV. Fe is rather bad at shielding neutrons below 0.861 MeV, since it is very inefficient
at slowing down neutrons via elastic scattering. Also heavy elements such as Fe can scatter
a neutron through an angle as large as 180◦ in a single scattering interaction- making it more
difficult for the incident neutrons to penetrate through the shield.

(ii.) Why is there a marked ”buildup” of neutrons in the energy region 0.01M eV &
E . 0.9M eV ? Why can even a very thick Fe shield not shield neutrons in this
energy region successfully, i.e. why does ”energy streaming” of neutrons occur
through this ”energy window”?
The only mechanism available to moderate neutrons below 0.861 MeV, is elastic scattering
and the probability of this occurring is very low, making iron very inefficient at moderating
neutrons by elastic scattering. Twenty elastic scattering events are required to reduce the
neutrons energy to half of its original value. Even a very thick iron shield cannot shield neu-
trons in this energy range successfully because the probability of elastic scattering is too low,
a characteristic feature of iron in the shielding of neutrons. Therefore, Fe is very good in
lowering the energies of high energy neutrons to below 1 MeV, but it is unable to ”finish the
job” and there is a ”build-up” of neutrons in the energy range En ∈ (10eV, 1M eV ), and Fe
is unable to ”get rid of” the neutrons in this energy range by either slowing down or absorption.

(iii.)Why do the two above transmitted neutron spectrum graphs look practically
identical, even though vastly different incident neutron energies are involved?
Although Fe is bad at shielding neutrons with energy less than 0.861MeV, it is on the other
hand an excellent shielding material for slowing down neutrons above 0.861 MeV . So even
though vastly different neutron energies (2 MeV and 20 MeV) are involved, the spectra would
still look similar because the high energy neutrons in the 20 MeV spectrum will be eliminated,
making it practically identical to the 2 MeV spectrum i.e. there is a build-up of neutrons in
the energy range (10 eV,1 MeV) in both spectra. Fe is unable to ”get rid of” the neutrons in
this energy range by either slowing down or absorption.

(iv.)What is the characteristic ”fingerprint” of Fe as a neutron shielding mate-

rial?
• Fe is rather bad at shielding neutrons with En < 1 MeV, because it is very inefficient at
slowing neutrons down via elastic scattering. The threshold energy for the excitation of first
discrete energy level in the dominant Fe isotope, 56 Fe, via inelastic neutron scattering, lies at
about 0.861 MeV. The element Fe can therefore not effectively slow down neutrons below about
circa 0.861 MeV.
• Fe is an excellent shielding material for slowing down neutrons with En & 1 MeV because
inelastic scattering by Fe is remarkably efficient at slowing neutrons down. Fe aggressively
slows down neutrons above about 1 MeV, because the threshold energy for inelastic neutron
scattering by the most abundant iron isotope, 56 Fe, lies at 0.861 MeV.

Part (e)
A beam of neutrons spanning the energy range from 0 to 50 MeV, are upon a
paraffin wax shield having a thickness of 100 cm. Will the thick, hydrogen-rich

10
wax shield be able to optimally attenuate the neutrons, or can far better neutron
shields be designed? Explain with reference to the underlying physics.

Paraffin wax is a hydrogenous material and therefore very effective at shielding neutrons with
energies below circa 4 MeV because the kinematic effectiveness of elastic scattering with 1 H is
high, and the cross-section for this elastic scattering interaction remains reasonably high below
approximately 4 MeV. This material as a neutron shield however has several shortcomings,
namely:
• The cross section for elastic scattering interactions falls drastically at high neutron energies.
• Quite energetic capture 2.205 MeV γ rays are emitted during the nuclear reaction 1 H(n, γ)2 H
• The scattering angle is forward peaked.
Thus the paraffin wax is able to effectively shield neutrons below 4 MeV but unable to effec-
tively shield neutrons above 4 MeV. In addition, 1 H produces 2.205 MeV γ-rays in the neutron
capture reaction 1 H(n, γ)2 H which presents further shielding problems. We thus conclude that
this shield will not be able to optimally attenuate neutrons. The paraffin wax must be comple-
mented by a shielding material which is effective at shielding fast neutrons such as Fe. Further
disadvantages of paraffin wax are : paraffin wax is flammable, water can leak and evaporate
and causes corrosion in shields.

Part (f )
A beam of neutrons spanning the energy range from 0 to 2 MeV, are incident upon
a paraffin wax shield having a thickness of 100 cm. Will the thick, hydrogen-rich
wax shield be able to optimally attenuate the neutrons? Explain with reference
to the underlying physics. Propose the addition of a final shielding layer on the
personnel side of the shield.

Paraffin wax is very effective at shielding neutrons with energies below circa 4 MeV because
of the high kinematic effectiveness elastic scattering with 1 H , and the cross-section for this
scattering interaction remains reasonably high below about 4 MeV. Since the energy range
of the neutrons is En ∈ (0, 2M eV ), this shield is capable of optimally attenuating neutrons.
However, when pure paraffin wax is used to shield low energy incident neutrons, the trans-
mitted ionising photon dose rate is excessively high compared to the transmitted neutron dose
rate, because the production rate of secondary 2.205 MeV ionising photons from 1 H(n, γ)2 H
nuclear reactions are high, and the low Z atoms in paraffin wax are unable to effectively shield
these secondary photons. One effective way to suppress the production of potentially problem-
atic ionising photons from the (n, γ) reactions, is to suppress the slow-neutron fluence rate by
strongly absorbing the slow neutrons. This can be done by mixing a slow-neutron absorber
such as boron carbide (B4 C) or ammonium pentaborate into the paraffin wax layer.

Part (g)
Which are the ”strong points” of Fe as a neutron shield? Which are the ”strong
points” of paraffin wax or polyethylene as a neutron shield? Which are the ”weak
points” of Fe as a neutron shield? Which are the ”weak points” of paraffin wax or
polyethylene as a neutron shield? Introduce the notion of complementary shield-
ing materials by explaining the complementary characteristics of a Fe/Wax neutron
shield, and the necessity of using complementary shielding materials for neutron

11
shielding.

”strong points” of Fe- Fe is an excellent shielding material for slowing down neutrons with
En & 1 MeV, because inelastic scattering by Fe is remarkably efficient at slowing neutrons
down. Also, heavier elements such as Fe scatter neutrons through large angles and thus has
better backscattering ability
”weak points” of Fe- Fe is very bad at shielding neutrons with En < 1 MeV, as it very
inefficient at moderating neutrons by elastic scattering.
”strong points” of paraffin wax- Paraffin wax is very effective at shielding neutrons with
energies below circa 4 MeV because of the high kinematic effectiveness elastic scattering with
1
H , and the cross-section for this scattering interaction remains reasonably high below about
4 MeV.
”weak points” of paraffin wax-the low Z atoms in paraffin wax are unable to effectively
shield the secondary ionising photons from 1 H(n, γ)2 H nuclear reactions.
Thus, no single shielding material is good enough in isolation-no material used on its own for
shielding will effectively shield both high and low energy neutrons. To effectively shield neutrons
with a wide range of energies, complementary shielding materials are required as they make
up for each other’s weaknesses and thus together shield the whole range of neutron energies.
Through inelastic neutron scattering, Fe is therefore used to retard high-energy neutrons which
are outside the cross-sectional ”grasp” for elastic scattering by 1 H in the incident neutron field,
which lowers neutron energy to within the energy range where 1 H has a proper cross-sectional
”grip” on them and can efficiently slow them down to low energies, where they can be easily
absorbed.

Part (h)
(i)Why is an initial thick Fe layer employed on the ”source side” of the shield?

Fe is an excellent shield for slowing down neutrons with energies above 0.861 MeV due to
the fact that inelastic scattering by Fe is remarkably efficient at slowing neutrons down. A
significant fraction of the neutrons emerging from the accelerator will have fairly high energies
(>4 MeV) and will therefore need to be slowed by iron to below about 1 MeV so that it can
eventually be absorbed (by the complementary layer).

(ii)Why is the initial Fe layer followed by a wax layer?

Since Fe is very inefficient at slowing neutrons down via elastic scattering, neutrons with en-
ergies less than 0.861 MeV will be transmitted through the iron layer. Wax is very effective
at shielding neutrons with energies below 4 MeV. Adding this wax layer ensures that the low
energy neutrons that are transmitted through the Fe layer are further slowed down into the
energy band below about 1 MeV, to thermal and near-thermal energies, where neutron absorp-
tion can take place and where neutrons are biologically less dangerous.

(iii)Explain why it will be counter-productive to ’swap” the positions of the thick

Fe and wax layers.
If we swap the two layers, the low energy neutrons emerging from the accelerator will be ef-
fectively shielded in the wax layer with the high energy (> 4MeV) neutrons being transmitted
through it. The iron layer will then effectively slow down the high energy neutrons to energies

12
below 1 MeV, which are then transmitted through the iron layer. In this way, neutrons are not
effectively removed from the radiation field.

(iv)What are the functions of the two steel shells having thicknesses of 1 cm each?
For ensuring that the lowest dose rate is transmitted because close to the optimum material
variations should be finer, thus the 1cm thickness.

Part (i)
Explain why 60 Co, which emits a photons at a maximum ionising photons energy
of 1.332 MeV, can not make irradiated food and other articles radioactive.

Radionuclides emitting a significant amount of ionising photons with E & 2.3 MeV may not
be used in food preservation by irradiation-at energies above this threshold, neutrons will be
produced in (n, γ)reactions, and these neutrons-if produced in significant quantities will cause
radioactivation of the irradiated food via (n, X)-reactions. The radionuclide most widely used
in food irradiation, is 60 Co, which emits practically no ionising photons above energy 1.332
MeV, which is well below photoneutron-producing, i.e. (γ, n) reaction threshold energies, so
that food preservation by irradiation with 60 Co will cause negligible radio-activation of irradi-
ated food. The only γ-photon emitted by 60 Co that can activate food, is the 2.506 MeV photon,
but its emission yield is very low at 2e − 8 and the only isotope in food that will be activated
by this photon is 2 H which has a natural abundance of only 0.015% and a (γ, n) cross-section
of circa 1 mb. Accordingly, food will not be measurably activated by neutrons produced in the
food via (γ, n)- reactions, when 60 Co is used as the irradiator isotope.

Part (j)
Photon radiotherapy is performed with a 24 MegaVolt LINAC, housed inside a
concrete vault. After several months of heavy use, a survey indicates that radia-
tion levels inside the vault are somewhat elevated above normal background levels,
when the machine is off. Explain the reason for this observation.

Since the linac produces Mega voltage radiation of 24 MegaVolts, there will definitely be neutron
activation and since the machine is being used frequently there will still be radiation in form
of neutrons even when the machine is switched off. In this case, concrete is the material being
used for neutron shielding and after being used for several months the concrete slowly loses
crystalline water which means that its shielding ability will deteriorate over the few months.
This is why the radiation levels in the concrete vault are higher than the normal background
levels even when the machine is switched off.

Question 4 :Practical radiation shielding problems

Part (a)
Given: The lens of the eye is located 3 mm below exterior surface of the eye.
Use the reference tables to identify which of the following beta-particle emitting
radionuclides pose a hazard to the lens of the eye through the development of

13
cataracts (i.e. opacity of the lens). Explain your reasoning.

Phosphorus 32 will pose a greater hazard to the eye lens through the development of
cataracts (i.e. opacity of the lens) because radionuclides emitting beta particles with maxi-
mum energies of less than 0.7 Mev have a range of < 3mm and therefore do not contribute to
the eye lens dose in practise. Thus 32 P will pose a greater threat to the eye lens since it has
the most energy(> 0.7 MeV) and thus higher range in tissue hence greater penetration depth
to the eye lens.

Part (b)
Determine the minimum thickness of a tissue-equivalent polymer that will shield
all the emitted beta particles, i.e. electrons. A single thickness polymer plate
should be able to shield all the emitted beta particles. Explain your reasoning.

The minimum thickness of tissue equivalent polymer used to shield the beta particles is 0.8cm,
which is done by considering the range of the beta particles with reference to tissue (i.e. the
distance that the particles will travel in the tissue or tissue equivalent material before being
stopped). Since tritium (3 H), carbon-14, phosphorus-33 and sulphur-35 have energies lower than
0.7 Mev, they have a range of less than 3mm with reference to the human tissue. Phosphorus-32
however, due to its slightly higher maximum energy, has a range in tissue of about 0.76 cm
and thus choosing a tissue equivalent polymer of thickness 8cm will ensure that all the beta
particles are attenuated eefectively.

Part (c)
You have a (CH2 )n polymer plate and a thin tungsten alloy plate available to shield
a radionuclide source that emits electrons. Both plates are just thick enough to
shield all the emitted beta particles. In which order will you place the two shielding
layers so that you will minimise the transmitted dose rate? Explain.

If the tungsten is placed first then there will be secondary radiation in form of ionizing photons
which would then require further shielding. This arises as a result of electrons striking a high
atomic (Z) number material (tungsten) thus producing ionizing photon radiation field with an
energy equivalent to the incident electrons. Thus the best way to arrange the layers would be
to put the polymer first then tungsten since when the electrons strike the polymer they will be
completely stopped without production of any further radiation.

Part (d)
Give the approximate thickness of (1) tungsten, (2) lead, (3) iron and (4) concrete
needed to reduce the transmitted ionising photon dose rate from a 60 Co source by
a factor 104 .

Assuming that the cobalt-60 radiation starts at a dose rate of 104 mR /hour and is to be
reduced to 100 (=1), we use the tenth value thickness to reduce the dose rate by a factor of 10
each time. For example, tungsten which has a tenth value thickness (TVT) of 2.9 cm :

14
 Thickness tenth value dose rate
4
0 0 10 mR /hour
1TVT 2.9 103 mR /hour
1TVT 2.9 102 mR /hour
1TVT 2.9 101 mR /hour
1TVT 2.9 100 mR /hour

Thus the approximate thickness for tungsten required would be (2.9 x 4) cm = 11.6cm.
Similarly, for lead, iron and concrete with TVT’s 4.5cm, 8.2cm and 27cm respectively, the
thicknesses required to reduce the dose rate by factor 104 are 18.0cm , 32.8cm and 108cm
respectively.

Question 5 : The Code MCNP

Part (a)
Present a line-by-line explanation of the following input data set to the code MCNP.

c Cell cards
1 1 1.29E − 3 -1 imp:p=1 $ Air

These are the cell cards. Cell 1 is composed of material number 1, which has a density of
1.29E-3g cm−3 . Negative densities mean that the quantity is mass-density in unit gcm−3 whilst
positive densities mean the unit is number of atoms per barn.cm. Cell 1 consists of the volume
inside surface 1. The importance (imp) of photons (p) inside cell 1, is 1. The $ sign signifies
the beginning of an in-line comment; everything to the right of the $ sign, is only a comment.

2 7 -1.00 -2 +1 imp:p=1 $Water

Cell 2 is composed of material number 7, which has a density of 1.00g cm−3 (negative densities
mean quantity is mass-density with unit cm−3 ; positive densities mean unit is number of atoms
per barn. Cell 2 consists of the intersection of all space outside surface 1 and inside surface 2.
The importance (imp) of photons (p) inside cell 2, is 1. The $ sign signifies the beginning of
an in-line comment; everything to the right of the $ sign, is only a comment.

3 1 -1.29E-3 -3 +2 imp:p=1 $Air

Cell 3 is composed of material number 1, which has a density of -1.29E-3g cm−3 (negative
densities mean quantity is mass-density with unit cm−3 ; positive densities mean unit is num-
ber of atoms per barn. Cell 2 consists of the intersection of all space outside surface 2 and
inside surface 3. The importance (imp) of photons (p) inside cell 2, is 1. The $ sign signifies
the beginning of an in-line comment; everything to the right of the $ sign, is only a comment.

4 0 +3 imp:p=0 $ Outer void

Cell 4 is a void region with zero importance. It is all space outside surface 3. The moment a
particle enters this region, it is not tracked any more. The comment term, UmWelt is a nice,
short, descriptive German word that means the outside world. Some MCNP users call it the

15
graveyard or the external void.

c Surface cards
1 so 10 $ Outer surface of inner air

We now understands that cell 1 is an air-filled sphere with radius 10 cm. Surface 1 is a
sphere (s) centred at the origin (o), hence the surface mnemonic so . The radius of surface 1 is
10 cm.

1 so 69 $ Outer surface of shielding material

Cell 2 is a spherical shell consisting of ordinary concrete; it extends from 10 cm to 69 cm,

i.e. the concrete thickness is 59 cm. Surface 2 is a sphere (s) centred at the origin (o), hence
the surface mnemonic so . The radius of surface 1 is 69 cm.

3 so 100 $ Outer surface of outer air

Cell 3 is the air shell outside the concrete; it extends from radius 69 to radius 100 . Cell
4 is all space outside spherical radius 100 cm. Particles that enter into this region of space, are
not followed (tracked) any more by the Monte Carlo code.

c Data Cards
mode p

This is a data card, transports

sdef par=p erg=d1 pos=0 0 0

sdef is the source definition and the particle(par) of interest is the photon(p). The energy(erg)
of the source particles as a distribution(d1). The photon is originally at the origin.

si1 L 1.173 1.332

sp1 0.999 1.000

The Source Information card number 1, i.e. the Si1 card, shows that the energy distribu-
tion d1 is a line distribution (L), (there are discrete photon energies). There are 2 discrete line
energies in distribution d1: E1 = 1.173 MeV and E2 = 1.332 MeV. The Source Probability
card (SP1) states that, in source energy distribution d1, the photon with energy E1 = 1.173
MeV is emitted with an emission yield y1 = 0.999 photons per radioactive transition, while the
photon with energy E2 = 1.332 MeV is emitted with an emission yield of y2 = 1.000 photons
per radioactive transition event.

c Material definitions
m1 7000 -0.755 &
m1 8000 -0.232 &
m1 18000 -0.0128 $ 1 Air at STP

This shows the definitions of the materials used in the different cells. The second line of
code, shows that material: m1 is air. This is photon transport and photon energies are very

16
low, (far below the threshold for nuclear reactions). The negative number after the ZAID code
, shows the mass fraction of the material composed of that element. If the number was positive,
it would have designated a number fraction and not a mass fraction. The next line in the code
shows that there is air at standard atmospheric temperature and pressure, which is composed
of 75.5% oxygen, 23.2% Nitrogen and 1.28% Argon. The various specifications which follow
similarly describe the other materials m2, m3,..m7.

c **** Tallies *******

fc12 Tally on photon dose-rate at surfaces
f12:p 3

The tally is the requested result i.e. answer calculated by MCNP. The last number, i.e. the
2 in the tally name, f12 denotes the type of tally. The first number 1 in the tally designation
f12 is simply a counter. The tally code f2 describes the fluence-rate at a surface. F12:p 3
instructs the code MCNP to determine the fluence-rate, in unit cm2 s−1 , of photons (”p”) that
cross surface 3. The fc12 card is the ”tally comment card”, i.e. it is simply the ”name” of tally
f12; this tally comment line will be printed in the output data set where the tally results are
given for tally f12 (i.e. ”tally 1 of type 2’).

fm12 1e10
The fm12 card is the tally multiplication card for tally f12, which in this case is 1e10

nps 1e7
The number of source particles to be tracked is equal to 1e7.

ctme 10
Here it is specified that the code MCNP must execute for 10 minutes. (CPU time to be spent
in the Monte Carlo run, in unit minutes.)

17