PHOTON ENERGY RESPONSE OF SILICON DIOXIDE FIBRE OPTIC AND
TLD 100 USING MONTE CARLO SIMULATION
HAZILA BINTI ASNI
A thesis submitted in fulfilment of the
requirements for the award of the degree of
Master of Science (Physics)
Faculty of Science
Universiti Teknologi Malaysia
February 2011
iii
iv
ACKNOWLEDGEMENT
In the name of Allah, the Most Gracious, Most Merciful. Praise be to
Allah S.W.T, Peace and blessings of Allah be upon His Messenger,
Muhammad S.A.W. and all his family and companions.
A sincere gratitude first goes to Prof Dr. Husin bin Wagiran for his believe in
me and opportunities given to me to further my studies, Prof Dr Ahmad Termizi
Ramli and Dr Suhairul Hashim for their guidance, support and motivation through
completion of this research.
It was a great pleasure to acknowledge help from PM Dr Wan Saridan, UTM,
Dr Iqbal from Faculty of Engineering, UPM and Wan Haizily for giving an
outstanding guidance for me to understand the MCNP5. Special thanks also go to Dr
Iqbal for providing me the MCNP5 software.
I am sincerely thankful to the government of Malaysia for a funded Master
scholarship. Not forgotten to the Academy of Sciences Malaysia and Ministry of
Higher Education of Malaysia for providing research grants for this research.
I’m also indebted to my colleagues especially Nor Haliza Yaacob, Noor
Hidayah Che Mat, Siti Sarah Mohd Azlan, Fairuz Diyana Ismail, Suhailah Abdullah,
Siti Fatimah Abdul Rahman and Fariza Hanim for their friendship, laughter,
enjoyment and enthusiasm during my studies in Universiti Teknologi Malaysia. Only
Allah S.W.T. can repay all your kindness.
I’m also would like to acknowledge all my families especially my parents and
my husband for their hope, prays and support which they still doing it until now, no
matter what happens.
A million of thank to all.
v
ABSTRACT
Even though there are other methods in detecting and measuring radiation,
thermoluminescence (TL) is still the main choice in many areas of ionizing radiation
dosimetry. This fact leads to the countless investigation of other materials to be used
as TL phosphor. Recent material being recognised and investigated as a TL phosphor
is silicon dioxide fibre optic. This study focuses on the energy response of silicon
dioxide fibre optics and TLD 100 subjected to photon irradiation. The TL responses
for photon energies, ranging from 20 keV to 20 MeV, were investigated as energy
absorbed in the TL materials. The simulation was performed using Monte Carlo NParticle transport code version 5 (MCNP5). The input parameters included in this
study are geometry specification, source information, material information and
tallies. Tally F6 was used to obtain average fraction of energy deposited by TL
materials in the simulation. Comparisons of energy responses were made between
calculated, simulation and previous experiment. For TLD 100, calculation results
show an over responses at below 100 keV while the simulation and experiment
results shows over response at below 150 keV. In terms of energy dependence, TLD
100 has a relatively flat response since its response lies within ANSI acceptable
range. Unlike TLD 100, fibre optic had a limited range of flat response. A flat
response can only be seen at energy range of 200 keV to 10 MeV. Although
simulation results exhibit similar pattern of energy responses, the values are slightly
higher when compared to calculated and experiment results, especially at lower
energy (< 100 keV). The effect of different dopant concentrations on the energy
responses were also analysed and discussed. Result from simulation shows no
apparent effect of different dopant concentration on the energy response.
vi
ABSTRAK
Walaupun terdapat pelbagai kaedah dalam pengesanan dan pengukuran sinaran,
dosimetri termocahaya (TLD) masih menjadi pilihan utama dalam bidang dosimetri
sinaran mengion. Kenyataan ini mendorong banyak penyelidikan dilakukan untuk
mencari bahan yang sesuai menjadi fosfor termocahaya. Di antara bahan baru yang
telah dikenali berpotensi untuk digunakan sebagai TLD ialah serabut optik silikon
dioksida terdop. Kajian ini menumpukan kepada sambutan tenaga serabut optik
silikon dioksida dan TLD 100 terhadap foton. Sambutan luminesens cahaya bagi
pelbagai tenaga, dalam julat 20 keV hingga 20 MeV, dikaji sebagai tenaga terserap di
dalam bahan luminesens. Kajian ini dijalankan menggunakan simulasi komputer
Monte Carlo N-particle, versi 5 (MCNP5).Terdapat beberapa parameter dimasukkan
ke dalam simulasi seperti spesifikasi geometri, maklumat berkenaan sumber sinaran,
maklumat bahan serta tallies. Tally F6 digunakan bagi mendapatkan jumlah pecahan
purata tenaga yang diserap oleh bahan TL dosimeter dalam simulasi ini.
Perbandingan sambutan tenaga dibuat antara keputusan simulasi dengan keputusan
kiraan dan eksperimen. Bagi TLD 100, keputusan pengiraan menunjukkan “over
response” pada tenaga 100 keV ke bawah, manakala keputusan simulasi dan
eksperimen menunjukkan “over response” pada tenaga 150 keV ke bawah. Dalam
terma kebergantungan tenaga, TLD 100 mempunyai sambutan tenaga yang malar
kerana semua sambutan tenaga berada dalam julat yang diterima ANSI. Tidak seperti
TLD 100, serabut optik mempunyai julat sambutan tenaga malar yang terhad.
Sambutan malar hanya dapat dilihat dalam julat tenaga 200 keV hingga 10 MeV
sahaja. Walaupun keputusan simulasi menunjukkan corak sambutan tenaga yang
serupa, nilainya sedikit tinggi apabila dibandingkan dengan pengiraan dan
eksperimen, terutamanya pada tenaga rendah (<100 keV). Kesan perbezaan
kepekatan bahan terdop dalam serabut optik terhadap sambutan tenaga juga dikaji.
Keputusan daripada simulasi menunjukkan tiada kesan yang besar terhadap
sambutan tenaga bagi kepekatan bahan terdop yang berbeza-beza.
vii
TABLE OF CONTENTS
CHAPTER
1
2
TITLE
PAGE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENTS
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
x
LIST OF FIGURES
xi
LIST OF SYMBOLS
xii
LIST OF ABBREVIATIONS
xiii
LIST OF APPENDICES
xiv
INTRODUCTIONS
1
1.1
Overview
1
1.2
Statement of problem
3
1.3
Research objectives
4
1.4
Statement of hypothesis
5
1.5
Scope of the research
5
1.6
Significance of the research
6
LITERATURE REVIEW
7
2.1
Theory of thermoluminescence
7
2.2
Thermoluminescence materials
9
2.2.1
Lithium fluoride
10
2.2.2
Silicone dioxide fibre optic
10
2.3
ANSI N545-1975
11
viii
3
4
2.4
Energy response
12
2.5
Introduction to Monte Carlo N-Particle Code
16
2.5.1
20
2.5.2 Photon interaction data
21
2.5.3 Variance reduction
22
2.5.4
23
Statistical check in MCNP5
RESEARCH METHODOLOGY
25
3.1
Introduction
25
3.2
Creating input file
26
3.2.1
26
Geometry description
3.2.2 Data card specification
29
3.3
Running the simulation
33
3.4
Interpreting MCNP5 output data
34
RESULTS AND DISCUSSIONS
36
4.1
Introduction
36
4.2
Photon energy response of TLD 100
36
4.2.1 Photon energy response of TLD 100
by calculation
36
4.2.1 Photon energy response of TLD 100
by MCNP5 simulation
38
4.3
5
Photon interaction in MCNP5
Photon energy response of silicon dioxide
fibre optic
43
4.3.1 Photon energy response of silicon dioxide
fibre optic by calculation
43
4.3.2
45
Photon energy response silicon dioxide
fibre optic by MCNP5 simulation
4.4
Effect of different dopant concentration on
TL energy response
48
4.5
Discussions on MCNP5 simulation
49
CONCLUSION
52
5.1
Summary of findings
52
5.2
Advantages and disadvantages of
using simulation
53
5.3
Recommendation of future research
54
ix
REFERENCES
56
Appendices A-G
62-72
x
LIST OF TABLES
TABLE NO.
TITLE
PAGE
2.1
Mode of transport available in MCNP5
18
2.2
Description of information in the default output files
generates by MCNP5
19
2.3
Differences of photo atomic interaction data libraries
22
3.1
Material description
31
3.2
Photoatomic data libraries
31
4.1
Weight fraction of elements contain in TLD 100
37
4.2
Calculated value of mass energy absorption coefficient of
TLD 100
37
4.3
Relative energy response of TLD 100 as calculated using
Eq. 4.4
38
4.4
Value of () for TLD 100
39
4.5
Response of TLD 100 at incident energy by simulation
40
4.6
Relative energy response (RER) of TLD 100
40
4.7
Weight fraction of elements contain in SiO2 fibre optic
43
4.8
Calculated value of mass energy absorption coefficient
for fibre optic
44
4.9
Relative energy response of fibre optic as calculated using
Eq. 4.4
44
4.10
Value of () for fibre optic
45
4.11
Response of fibre optic at incident energy
45
4.12
Relative energy response (RER) of fibre optic
46
4.13
Result produce by MCNP5 with two different SI and SB card
51
xi
LIST OF FIGURES
FIGURE NO.
TITLE
PAGE
2.1
Simple examples of luminescence processes
8
2.2
Energy band diagram
9
2.3
Relative energy responses of several TL phosphors
14
2.4
LiF: Mg, Ti energy response to photon
14
2.5
Germanium doped fibre optic energy response to photon
15
2.6
The MCNP flow chart diagram
19
3.1
Flow chart of the work in this research
25
3.2
Problem’s geometry information as written in cell and
surface card
27
3.3
3D view of problem’s geometry.
27
3.4
X-Z side view of the phantom
28
3.5
X-Y plane view of the phantom
29
3.6
Distribution of 100 000 particles from source with
30
energy 1.25 MeV
3.7
Data cards
32
3.8
MCNP5 command prompts
33
4.1
Comparison of mass energy absorption coefficients
between calculated (MEAC) and simulation (())
41
4.2
Comparison of TLD 100 relative energy response between
calculation, experiment and simulation
42
4.3
Comparison of mass energy absorption coefficients between
calculated and simulation
46
4.4
Comparison of fibre optic relative energy response between
Calculated, simulation and experiment.
47
4.5
Effect of dopant concentration on germanium doped SiO2
fibre optic response
48
xii
LIST OF SYMBOLS
E
Energy
k
Boltzmann’s constant
p
Probability of escaping by the trap
Т
Temperature
Zeff,
The effective atomic number
SiO2
Silicon dioxide
Z
Atomic number of the atom
wi
Fraction of that element
η
The efficiency of the thermoluminescence emission
m
Mass
Gy
Gray
LiF
Lithium fluoride
Ti
Titanium
ppm
Part per million
()
Average fraction of energy deposited in TL material
Photon fluence
()
TL efficiency
( ⁄)
() The mass energy absorption coefficient for air
()
Photon energy response
()
μ
Relative Energy Response
μ
Mass energy absorption coefficient of the material
Mass energy absorption coefficient of the reference material
60
Co
Cobalt-60 source
CaSO4
Calcium Sulphate
xiii
LIST OF ABBREVIATIONS
TLD
Thermoluminescence dosimeters
TL
Thermoluminescence
PMT
Photomultiplier
PMMA
Polymethylmethacrylate
SEM
Scanning electron microscope
MCNP5
Monte Carlo N-particle code, Version 5
EGS
Electron-gamma shower computer code
EGSnrc
Electron-gamma shower computer code (maintained by NRC)
NRC
National Research Council of Canada
GEANT4
Geometry and Tracking computer code
TTB
Thick Target Bremsstrahlung
SDEF
Source specification data card
EPDL
Evaluated photon data libraries
ENDF
Evaluated nuclear data files
ACE
A Compact ENDF
SB
Source bias
SD
Source definition
SI
Source information
SP
Source probability
FOM
Figure of Merit
Vised
Visual Editor
NPS
Number of particles
IMP:P
Photon importance
ANSI
American National Standard Institute
xiv
LIST OF APPENDICES
APPENDIX
TITLE
PAGE
A
Example of input file for TLD 100
62
B
Example of input file for fiber optic
63
C
Example of output file generated by MCNP5
64
D
The mass energy absorption coefficients for air.
69
E
The mass energy absorption coefficients of
70
TLD 100 elements
F
The mass energy absorption coefficients of
71
silicon dioxide fibre optic elements
G
Publications and conferences
72
CHAPTER 1
INTRODUCTION
1.1
OVERVIEW
Over the last decade, the investigation of thermoluminescence and other
thermoluminescence dosimetry has expanded enormously. Many phosphor materials
have been studied, become commercially available and have been applied to many
areas of ionizing radiation dosimetry including personal, environmental, clinical,
charged particle and neutron dosimetry.
In the late 1940s, Daniels and his co-workers had concluded in their
investigation that lithium fluoride (LiF) from Harshaw Chemical Company was most
suitable to use in measuring ionizing radiation exposure [McKinlay, 1981]. Later in
studies of LiF in 1960, Harshaw incorporated titanium and others elements in the LiF
to produce a phosphor with a high TL sensitivity. This material is the basis of what is
now generally regarded as the ‘standard’ TL phosphor: Harshaw TLD 100. TLD 100
was often being used as comparison material in investigating others TLD [McKinlay,
1981].
More recently studies in thermoluminescence are in investigating and
developing the commercial SiO2 fibre optic as a TL material. The fibre optic has
several characteristics such as respond monotonically to gamma photon radiation and
in some dose regions it responds linearly, carries a low residual TL signal, can be
reused several times and low degree of fading. As claimed by Espinosa et. al. [2006]
2
fibre optics could be very attractive for used in variety of radiation dosimetry
applications due to its small size, flexibility, low cost and commercially available.
TL performance of an irradiated fibre optic is depends on the type of fibre
and by the radiation parameters [Hashim, 2009]. Different impurities in the fibre
could give different outcome in the fibre optic performance as TL material. Abdulla
et al [2001a] has carried out a TL study on commercially available ge-doped silica
based fibre optic in dose range of 1-1230 Gy. It was found that fibre optic was
response linearly from 1 to 120 Gy and has fast fading rate (2% within 6 hours and
6% within 30 days). Besides germanium, Abdulla et al [2001b] also studied on
commercially available erbium doped fibre optic. Dose ranges investigated was 2 to
400 Gy, linear dose response was found at up to 250 Gy. Despite its wider range in
linear dose response, erbium doped fibre optic had rapid rate of fading compared to
germanium doped. As reported, nearly 30% of TL signals being lost after the first 24
hours and a total loss of 58.6% of TL signal were reported after 20 days storage at
room temperature.
Hashim [2009] carried out work also on commercially available germanium
doped fibre optic compared mainly to aluminium doped fibre optic. This TL
dosimeter irradiated by a broad range of sources, from low energy photons to
megavoltage, through to neutrons and charged particles (from E-sources, accelerated
electrons and D-particles). Germanium doped fibre optic was found to have linear
dose response until at least 4 Gy for 6 MV photons, and up to 3.5 Gy for 6-, 9- and
12 MeV electrons irradiation. Besides photon and electron irradiations, a linear dose
response was also observed for 2.5 MeV protons irradiations. Exposure to 0.18 Gy of
the 90Sr / 90Y E-ray source; it was found that germanium doped fibre had higher TL
response compared to the Al-doped fibres. Germanium doped fibre was also prove to
have strong TL response to fast neutron irradiation, whereas for aluminium doped
fibres the TL response was practically negligible.
If a TL material is to be used for any dosimetric applications in the field of
photon radiation, one of the main characteristics that must be known is its energy
response. The energy response is a measure of an energy absorbed in TL material
3
used in comparison to the energy absorbed in a material taken as the reference, when
irradiated at the same exposure. Normally air is used as a reference material in
dosimetry. Energy response is not easily measured, but its theoretical value is very
helpful in selecting a particular TL material for any special application. [Oberhofer et
al, 1981]. Other than energy response, the absorbed dose response was also one of
the thermoluminescence characteristics that can be examined. It is very useful for a
phosphor to have a linear TL-absorbed dose response over measurement and
calibration ranges of absorbed dose [McKinlay, 1981].
Nowadays, other than carried out by experiment, the response studies also
could be done by simulation. One of the famous simulation methods is Monte Carlo
simulation. Monte Carlo simulation is a stochastic technique which is based on
random numbers and probability statistics.
This simulation is one of the most
important tools to study particle transport and interaction with matter as well as
radiation protection and dosimetry. Various Monte Carlo simulation programs are
available for different user needs. For instance, Monte Carlo N-Particle (MCNP),
Geometry and Tracking computer code (GEANT4) and Electron- Gamma Shower
computer code (EGSnrc).
Monte Carlo N-Particle version 5 (MCNP5) will be used for this research.
MCNP5 is a general purpose Monte Carlo n-particle transport code that is
continuous-energy, generalized-geometry, time-dependent code, and can be used for
single or coupled neutron/photon/electron transport.
1.2
STATEMENT OF PROBLEM
There are several studies showing the useful of fibre optic as TL
material [Houstan,et.al, 2002; Abdulla, 2003; Espinosa et.al, 2006; Hashim, 2009] in
terms of TL sensitivity, fading and dose response. This study will give more attention
to the fibre optic energy response as from the review; none of these studies discuss
enough about energy response of TL dosimeter. Energy response has been studied by
Abdulla [2003] and Safitri [2006]. However, their studies on energy response of TL
4
dosimeter only limited to accelerated photons and electrons from linear accelerator,
which means only energy more than 1 MeV were applied. On the other hand, this
studied will applied energy of 20 keV until up to 20 MeV which cover from
environmental dosimetry energy ranges to clinical applications. With the aim to
understand the dependency of fibre optic response on broad energy ranges, studying
this energy ranges is essential as this study may provide a basis for energy correction
method needed for fibre optic later.
Previous studies in thermoluminescence were carried out mainly by
experiment. However this research tries alternative approach which is Monte Carlo
simulation, specifically MCNP5. It is a common but not widely used method in
thermoluminescence dosimetry area. With the aim to prove the capability of
MCNP5, this research investigates the energy response of TLD 100 and
commercially available SiO2 fibre optic at energy ranges of 20 keV to 20 MeV.
1.3
RESEARCH OBJECTIVES
The objectives of this research are as follows:
i.
To simulate the response of TLD 100 to various energy of
photon using MCNP5 simulation.
ii.
To simulate the response of commercial germanium doped
SiO2 fibre optic to various energy of photon using MCNP5
simulation.
iii.
To study the effect of different dopant concentration in SiO2
fibre optic on energy response using MCNP5 simulation.
5
1.4
STATEMENT OF HYPOTESIS
Over-response (when the ratio of the amount TL from a dose of a test
radiation to the TL from the same amount of
60
Co source is greater than 1.0) in the
150 keV ranges for TLD 100 has been well documented [McKinlay, 1981,;
McKeever et al, 1995; Glennie, 2003; Davis, 2003; Hranitzky et al, 2006]. It is
hypothesize that simulation response of TLD 100 will have similar response as
measured by experiment as well as calculated.
Energy response of fibre optic will also be comparing with calculated and
experiment response measured by Abdulla [2003]. It is expected that fibre optic
response by simulation depict similar response with experiment and calculation.
Abdulla [2003] studies shows over response at energy 100 keV and lower, and like
most of TL material energy response has a flat energy response at high energy
(greater than 100 keV).
1.5
SCOPE OF RESEARCH
These studies will applied Monte Carlo N Particle code version 5 (MCNP5).
MCNP was chosen because of the simplicity and ease when using the code compared
to the other codes. Many of the published papers perform Monte Carlo simulation
using EGS code in the area of thermoluminescence dosimetry. Thus, the lack of
published literature referencing MCNP shows that there is a need for more research
to be conduct with MCNP.
Energy response of fibre optic will be compare with calculated response and
experiment response as measured by Abdulla [2003]. Even though there is various
type of fibre optic with different dopant available, germanium doped fibre optic was
chose as investigated material because of its usefulness in terms of sensitivity and
dose response was proved by Abdulla [2003] and Hashim [2009]. Since only one
type of fibre optic was used in this research, any fibre optic mention was referred to
germanium doped fibre optic hereafter.
6
Other than germanium doped fibre optic, as a standard material, TLD 100
will be used as a comparison material. TLD 100 was chose as its information on its
energy response were extensively studied by several group [Davis, 2003; Glennie,
2003 and Hranitzky et. Al, 2006]. For more specific comparison, studies by Glennie
[2003] were chose as its thesis provided complete and comprehensible setting of the
experiment as well as result of energy response. The experiment set up by Glennie
will be a reference for simulation setup of this research later. Energy response of
TLD 100 from simulation studies will also be compared with result of TLD 100
energy response from Glennie [2003].
1.6
SIGNIFICANCE OF THE RESEARCH
This research will give an insight of commercial SiO2 fibre optic for a
candidate as a TL material in terms of its energy response. The fibre optic can be
widely used not only for clinical dosimetry but also for environmental and personal
dosimetry.
This research utilized MCNP5 simulation in order to study the energy
response with the aim to promote the usage of computer simulation as well as
provide basic knowledge in MCNP5 application. Moreover, by using MCNP5
simulation; it is much easier to study the relationship between response and TL
material. Hopefully this research will contribute to the knowledge that can create
new experts in thermoluminescence area as well as Monte Carlo area.
CHAPTER 2
LITERATURE REVIEW
2.1
THEORY OF THERMOLUMINESCENCE
When certain materials are irradiated, the absorption of energy from the
radiation induces instability in their structure. These materials can, however, return
to their normal structure if the energy absorbed is released again. The measurement
of the released energy provides an indication of how much radiation dose was
absorbed in the material.
Figure 2.1 showing the example of luminescence processes. As shown in
Figure 2.1(a), the prompt return of an electron from excited state E (valence band)
either directly to the ground state G (conduction band) or via an allowed transition
from intermediate state S (relaxation) is called fluorescence. However, due to the
presence of an electron trap known as metastable state M, the return of electron to the
ground state is delayed. The metastable state represents a shallow electron trap and
electrons returning from it to the excited state require energy as shown in Figure 2.1
(b).
8
E
E
(a)
(b)
S
G
M
G
E
M
(c)
Heat In
G
Figure 2.1
Simple examples of luminescence processes.
The probability p of escape of an electron from a metastable state to an
excited state is governed by the Boltzmann equation.
= −∆
(2.1)
Where s is a constant, ΔE is the energy difference between states E and M or
commonly known as trap depth, k is Boltzmann’s constant, and T is the temperature
in Kelvin. By raising the temperature, the probability of escape of an electron is
increased. This process is called thermoluminescence as illustrated in Figure 2.1 (c).
The presence of defects or so-called dopant in a material is important for
thermoluminescence process. The conduction band and the valence band are widely
separated in energy by the so-called ‘forbidden gap’ (Figure 2.2 a). Without the
influence of the external forces, it is improbable for an electron to be able to cross the
forbidden gap from the valence band to conduction band. The presence of dopant
allowed energy level to exist in the forbidden region as illustrated in Figure 2.2 (b),
subsequently make it possible for electrons to be trapped. Thus, the material becomes
more sensitive as the energy level trapped more electrons during irradiation. For
instance, Harshaw Chemical Company had incorporated Titanium (Ti) in LiF to
produce a phosphor with a high TL sensitivity [McKinlay, 1981].
9
(b)
(a)
Conduction Band
Conduction Band
Energy level
Without the influence of the
external forces, it is not easy
for an electron to be able to
cross this gap.
E
L
H
Valence Band
Valence Band
Figure 2.2
2.2
Energy Band Diagram.
THERMOLUMINESCENCE MATERIAL
A large number of materials have been discovered exhibit a good
characteristic of thermoluminescence. For used in radiation dosimetry, only some
material are suitable. As discussed by Regula and Driscoll [1993] and McKeever et.
al. [1995], the characteristics of TL material which are most important for radiation
dosimetry are:
1. Linear dose dependence. A good TL material has a wide interval of linear
dose dependence.
2. A sufficiently high sensitivity, i.e., a high TL signals per unit absorbed
dose.
3. A suitable flat energy response over a wide range of energy of the
incident radiation. In other word, the TL response should have low
dependence on the photon energy.
10
4. Low rate of fading. The TL material must be capable to store dosimetric
information for a long time.
5. Simple TL curve.
6. TL material should be mechanically strong, chemically inert and radiation
resistant.
2.2.1
Lithium Fluoride
The first basic TL material commercially known as TLD 100 has the
composition of ~92.14% 7LiF and ~7.36% LiF incorporated with ~ 200 ppm
Magnesium and ~10 ppm Titanium [Saint, 2000] is generally regarded as a standard
TL dosimeter. TLD 100 properties are still extensively studied since 1960’s and was
often being used as a comparison material in investigating other TL materials. TLD
100 characteristics that made it good TL dosimeters include the wide useful ranges
(10μGy – 10Gy), low fading (5 – 10% per year) and sensitivity to small doses
[Kortov, 2007]. TLD 100 arising factor that made it interesting is its tissue
equivalence (Zeff = 8.04), which is an important factor in personnel dosimetry as
well as medical application. Lithium Fluoride is also available with different
composition that is suitable for measuring mix radiation fields (TLD 600, TLD 700).
2.2.2 Silicon Dioxide Fibre Optic
Recently, several research groups have reported a number of studies on
application of silicon dioxide (SiO2) fibre optic as a dosimeter. Justus et. al. [1997]
and Houston et. al. [2002] with their research group have reported the development
of doped optical fibre material for use as a dosimeter. In 2003, Abdulla studied the
effect
of
different
doped
(Germanium
doped
and
Erbium
doped)
in
thermoluminescence response to photon irradiation with the focus more on usage of
optical fibre dosimeter in radiation therapy. Later on, Hashim [2009] had studied the
thermoluminescence response of commercially optical fibre to ionizing radiation
11
such as low-energy X-rays, megavoltage photons (6 MV), accelerated electrons (up
to 12 MeV), accelerated protons (2.5 MeV), D-particles (5.954 MeV) and fast
neutrons (10 keV - 10 MeV). His study had proved the doped SiO2 fibre optic are
capable of measuring the radiation emission from a range of radionuclide sources
(E’s,D’s and neutron-emitting), low-energy X-rays and megavoltage photons and also
accelerated particles (electrons and protons).
2.3
ANSI N545-1975
Since a suitable flat energy response is insisted for a good TL dosimetry, a
standard to show the degree of the flat energy was chose known as ANSI N5451975. Its title “Performance, Testing, and Procedural Specifications for
Thermoluminescence Dosimetry (Environmental Applications)” and was coded as
ANSI N545-1975 is a standard code prepared by American National Standard
Institute (ANSI) in order to fulfil the need for an accurate, sensitive and reliable for
monitoring environmental radiation. The standard specifies performance criteria for
TLD systems used for the measurement of environmental exposures levels of X and
gamma radiation.
One of the dosimeter performance criteria specified in ANSI N545 that relate
to energy response is “The response of TLD to photon shall be determine for several
energies between 30 keV and 3 MeV. The response shall not differ from that obtained
with calibration source by more than 20 percent for photon with energies greater
than 80 keV and shall not be enhanced by more than a factor of two for photons with
energies less than 80 keV”.
Even though this standard was prepared for environmental application, it was
also suitable for this research since range of energy investigated for this research
covered for all kinds of dosimetry.
12
2.4
ENERGY RESPONSE
The photon energy response may be expressed in many different ways.
According to Oberhofer et. al. [1981], photon energy response is the measurement of
energy absorbed dose in the TL material compared to energy absorbed dose in a
reference material. In other word, for a fixed dose, the energy response is the
variation of the detected TL output as a function of the energy absorbed radiation.
The photon energy response or known as energy dependance, SE (E) is
defined as [McKeever et al, 1992; Horowitz, 1984 ]
μ
!" #
() = μ
&()
!
" $%
(2.2)
('! ⁄")*+$() is the mass energy absorption coefficient of the material, subscript ‘m’
refer to TLD material and subscript ‘ref’ refers to reference material or medium. Air
is normally taken as a reference medium because the ratio between absorbed dose
and exposure for air is constant. ('! ⁄") is a measure of the average fractional
amount of incident photon energy transferred to kinetic energy of charged particles
in the medium. The variable η(E) is energy-dependent relative TL efficiency. The
light output from a given TL material is dependent primarily on total energy
deposited within a material and it also depend to some degree on the energy of
incident radiation. Horowitz [1984] explains this effect is called energy-dependent
relative TL efficiency and suggests that grain size and ionization density effects
could be contributing factor to its value. Since available data of η(E) for TLD 100
and fibre optic is scarce and inconsistent, it is assumed to be unity in all of the
calculation in this study.
When discussing about energy response, it is unavoidable to discuss also
about effective atomic number, Zeff. The effective atomic number, Zeff dictates the
amount of energy absorbed by the TL material in a given radiation field. When
photons interact with matter, the energy loss of photon can take place by
photoelectric absorption, Compton scattering and pair production where the
dominated process depends upon the energy of incident photon. However, the exact
energy ranges for which these processes predominate are governed by Zeff.
13
According to McKeever et al [1992], combination of higher value of Zeff and lower
energy of incident photon will produce larger TL response to a given dose
For the purpose of personnel or medical dosimetry, the tissue equivalency is
an essential consideration. A material is consider as “tissue equivalent” if it have the
Zeff near the human tissue Zeff which is 7.4. Zeff for germanium doped SiO2 fibre optic
is in the range of 11.9 – 13.4 [Hashim et. al., 2007] whereas Zeff for TLD100 is 8.2
[McKinlay, 1981].
Non-tissue equivalent materials tend to exhibit an “over-response” or “under
response” when it interacts to photon. The “over response” is when the ratio of the
amount TL from a dose of a test radiation to the TL from the same amount of
60
Co
source is greater than 1.0. Since Zeff for SiO2 fibre optic and TLD 100 is larger than
tissue, it is expect that it will tend to over-response when interact with low energy of
photon. This is the result of photoelectric absorption coefficient which is a function
of the effective atomic number (Zeff) of the TLD material to the third power and the
energy of the photon inversely to the third power (Zeff3/E3) [Horowitz 1983,
Horowitz 1984, McKeever et. al. 1995]. For example, as shown in Figure 2.3, CaSO4
(Zeff = 15.3) has an over response at an energy of 100 keV, while LiF (Zeff = 8.2) has
almost no over response at the same energy.
For dosimetric purposes, a material with an energy response which is as
constant as possible over the energy range of interest is desirable. Information of the
deviation in the energy response (over-response) indicates how much the dose
deviates from the dose to be determined. Thus, information of energy response is
very helpful in selecting a particular TL material for any special application.
[Oberhofer et. al., 1981],
14
Figure 2.3
Relative energy responses of several TL phosphors.
[McKinlay, 1981]
The photon energy response of LiF: Mg, Ti (TLD 100) has been measured by
various groups since its development. Result measured by Davis [2003], Glennie
[2003] and Hranitzky et. al. [2006] shows the same pattern of TLD 100 response to
photon. Glennie [2003] reported that TLD 100 tend to over response in the 10 – 150
keV range as shown in Figure 2.4.
1.6
Relative Response
1.4
1.2
1.0
0.8
0.6
10
100
1000
10000
100000
Energy (keV)
Figure 2.4
LiF: Mg, Ti Energy response to photon. [Glennie, 2003]
15
For practical use of photon energy response, the relative energy response
(RER)E definition is used. The relative energy response is defined with respect to
1.25 MeV which is average photon energy from a
60
Co source (1.17 MeV and 1.33
MeV photon per 60Co disintegration), thus
(--) =
()
.. 01 23 6745
(2.3)
The energy response in Glennie [2003] was compared to 6MV photons
produce on the 6/100 linear accelerator rather than normalizing to average energy
from
60
Co source. Normalizing the result between
60
Co and 6MV would only give
slightly difference result [Glennie, 2003].
Due to its status as recent TL material, there is not much literature available
for germanium doped fibre optic studies on energy response. Abdulla [2003]
reported energy response of this fibre optic meanwhile Safitri [2006] reported the
energy response of erbium doped fibre optic. Figure 2.5 shows energy responses as
measured by Abdulla [2003] normalize to 1.25 MeV. It obviously shows fibre optic
had large over response at lower energy.
4.50
4.00
Relative Response
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
10
100
1000
10000
100000
Photon Energy (keV)
Figure 2.5
Germanium doped fibre optic energy response to photon [Abdulla,
2003]
16
2.5
MONTE CARLO N-PARTICLE CODE
Monte Carlo particle transport techniques were first introduced in the 1940s
for the design of atomic weapons. They have since evolved into many different areas
of application, such as the much more peaceful applications in radiotherapy physics.
It is well accepted now that Monte Carlo methods offer the most powerful tool for
modelling radiation transport. Nowadays, various Monte Carlo simulation programs
are useful for different user needs such as EGS, GEANT 4 and MCNP. With so
many codes available, there have been increasing numbers in papers published on the
application of Monte Carlo methods specifically dealing with external photon beam
modelling as well as thermoluminescence dosimetry [Verhaegen, 2002].
A number of paper and thesis were reviewed and analyzed. Many of research
group use EGS Monte Carlo system as a tool to investigate their problem. Miles
[1994] studied the photon energy response for filtered CaSO4 dosimeter using EGS4
Monte Carlo system. Since the response of CaSO4 is not air-equivalent, it needs
filtration or any other energy correction method if dosimeters are to be exposed at
lower energy photon. Miles compared response of the dosimeter from simulation
with response from experiment and also with those calculated based on simple linear
attenuation. His finding shows EGS4 calculated results were in close agreement with
experimental data than the calculation based on attenuation. Miles reported
discrepancies in the EGS4 calculation are believed to be due to energy dependent
thermoluminescence efficiency of the phosphor.
Meanwhile, Davis [2003] in his research utilizes EGSnrc, the extended code
of EGS4. Davis studied the potential of TLD 100H as a detector for environmental
dosimetry. TLD 100H was encapsulated in Teflon and mounted in aluminium card
(The Harshaw Type 8855 environmental dosimeter) which had four different filters.
Davis compared the result with between measured value and EGS calculated value
and had found Monte Carlo calculations were in good agreement with the measured
result. He suggests that Monte Carlo techniques could prove useful for future
dosimeter studies.
17
In 1996, Mobit et. al. applied the EGS4 dosimeter in investigating quality
dependence factor of LiF TLD in megavoltage photon beams. Later in a same year,
Mobit et. al. carried on research with studying energy correction factor of LiF TLD
in megavoltage electron beams. Both of this study have proved the result provided by
simulation were in agreement with the experiment.
In the area of thermoluminescence dosimetry, application of MCNP was not
as popular as EGS Monte Carlo system. Olsher [1993] in his studied used MCNP
code (version 4.2) to investigate the photon energy response of aluminium oxide.
With the aim to used aluminium oxide as environmental dosimeter, photon energy
response of the dosimeter was calculated over the range from 10 keV to 1 MeV. Its
specific goals include comparison with TLD 100 and effect of copper filtration to
flattening response. This research found that 0.5 mm copper filter was effective to
flatten the response of the aluminium oxide dosimeter.
Other group that utilized MCNP was Eakins et. al. [2008]. Using MCNP4C2, this group aim to redesign photon/electron personal dosemeter in order to
adopted TLD 700H instead of using usual TLD 700. Calculated (MCNP-4C2)
relative response characteristics at different angles of incidence and energies between
16 and 6174 keV are presented for this new dosemeter configuration and compared
with measured type-test results. Measured results from experimental type testing of
the dosemeter validated the calculated data to an adequate extent and, consequently,
served as additional confirmation of the MCNP simulation employed.
Monte Carlo N-Particle version 5 (MCNP5) will be used for this research.
MCNP5 is a general purpose Monte Carlo n-particle transport code that is
continuous-energy, generalized-geometry, time-dependent code, and can be used for
single or coupled neutron/photon/electron transport. One main difference of MCNP
from other Monte Carlo code is MCNP can be run in several different modes. By
default, mode N is used; neutron transport only. Other mode that can be used in
Monte Carlo listed in Table 2.1 [X-5 MC Team, 2005].
18
Table 2.1
Mode
NP
P
E
PE
NPE
Mode of transport available in MCNP5
Description
neutron and neutron-induced photon transport
photon transport only
electron transport only
photon and electron transport
neutron, neutron-induced photon, and electron transport
The user of MCNP5 need to creates an input files that is subsequently read by
MCNP. The file contains information of the problem such as:
a)
The geometry specification,
b)
The material description,
c)
The characteristic and information of the source,
d)
The type of answers or tallies desired, and
e)
Any variance reduction technique used to improve efficiency.
Figure 2.6 showed the MCNP flow chart diagram. The input file created
consisted of cell card, surfaces card and data card. The cell card and surface card
contain the information about geometry specification of the problem. Meanwhile the
data card contains information about the material description, the localizations and
characteristics of the source, the type of tallies and variance reduction technique
used. To create input files, there are two options available. User can create manually
or using MCNP Visual Editor (MCNP Vised). Creating manually means, user must
use their own capability to visualize the geometry of the problem, setting parameter
according to MCNP manual and key in input parameter manually by line editor.
Second options available in creating input file is by using MCNP Vised.
MCNP Vised is the interactive software developed to assist the user in the
creation of MCNP input files. The Visual Editor allows the user to easily set up and
modify the view of the MCNP geometry and to determine model information directly
from the plot window. User also can interactively create an input file with the help of
two or more dynamic cross sectional views of the model. A wide selection of menu
options enables rapid input of information and immediate visualization of the
geometry and other information being created. The Visual Editor code became part
of the MCNP package with the release of version 5 of MCNP.
19
The user must also certain that all required files are in the working directory. This
is important because these files contain the cross section libraries for photon,
neutron, materials cross section and also information about the simulation. More
detailed regarding photon interaction is discussed in subtopic 2.4.1 and subtopic
2.4.2.
Create the input
file
Ensure all the
required files are in
working directory
Run the
simulation
Output files
generated by
MCNP
Figure 2.6 The MCNP flow chart diagram
Upon completion of the simulation, MCNP5 will generate an output file. An
output file provides information regarding the simulation. The user may find the
following information will be printed automatically in the default output file:
x
A listing of the input file,
x
The problem summary of particle creation and loss,
x
KCODE cycle summaries,
x
Tallies
x
Tally fluctuation charts, and
x
Information listed in Table 2.2
Table 2.2
Description of information in the default output files generates by
MCNP5
Table
Number
Type
Description
60
62
72
100
126
160
161
162
175
Basic
Basic
Basic
Basic
Basic
Default
Default
Default
Shorten
Cell importance
Forced collision and exponential transform
Cell temperatures
Cross-section tables
Particle activity in each cell
TFC bin tally analysis
f(x) tally density plot
Cumulative f(x) and tally density plot
Estimated keff results by cycle
20
190
200
Basic
Basic
Weight window generator summary
Weight window generated window
Information indicated as ‘Basic’ in table 2.1 can be switched off by user. User
just has to specify in MCNP command prompt which table number they don’t need.
Besides ‘Basic’ and ‘Default’ table, user may also print other information such as
universe map, source distribution information table and etc if they need it. The list of
other optional tables can be found in MCNP manual (Vol. II), chapter 3, page
number 3-146 and 3-147.
2.5.1
Photon Interaction in MCNP5
MCNP has two photon interaction models: simple and detailed.
The simple physics treatment is intended for high-energy photon problems or
problems where electrons are free and is important for next event estimators.
Coherent (Thomson) scattering and fluorescent photon from photoelectric absorption
is ignored in this model.
Different from simple model, the detailed physics treatment includes coherent
(Thomson) scattering and accounts for fluorescent photons after photoelectric
absorption. Form factors and Compton profiles are used to report for electron binding
effects. Analogue capture is always used. The detailed physics treatment is almost
always used by default. It is the best treatment for most applications, particularly for
high Z nuclides or deep penetration problems.
The generation of electrons from photons is managed in three ways. These
three ways are the same for both the simple and detailed photon physics treatments.
1) If electron transport is turned on (Mode P E), then all photon collisions
except coherent scatter can create electrons that are banked for later transport.
21
2) Thick-target Bremsstrahlung model (TTB) is used in condition of no electron
transport (no E on the Mode card). This model however generates electrons,
but assumes that they are locally slowed to rest. Any Bremsstrahlung photons
produced by the non-transported electrons are then banked for later transport.
Thus electron-induced photons are not neglected, but the expensive electron
transport step is omitted. (The TTB production model contains many
approximations compared to models used in actual electron transport. In
particular, the Bremsstrahlung photons inherit the direction of the parent
electron.) The TTB approximation is the default for MODE P problems.
3) If all electron production is turned off, no electron-induced photons are
created, all electron energy is assumed to be locally deposited. This can be
done by adding IDES = 1 on the PHYS:P card in the input file. In MODE P E
problems, TTB approximation plays a role when the energy cut-off for
electrons is greater than that for photons. In this case, the TTB model is used
in the terminal processing of the electrons to account for the few low-energy
Bremsstrahlung photons that would be produced at the end of the electrons’
range.
If the user did not alter anything, MCNP will take default setting of photon
interaction. By default, the upper energy limit for detailed photon physics treatment
will be 100 MeV. In handling production of electron, MODE P will use TTB
approximation. There is no photonuclear collision by default and Doppler energy
broadening will occurs.
2.5.2 Photon Interaction Data
Photon interaction cross sections are required for all photon problems. Photon
interactions can now account for both photo atomic and photonuclear events. There
are currently four photo atomic interaction data libraries. The difference between 4
libraries is presented in Table 2.3.
22
By default, photon transport in MCNP5 will utilized “04p” table unless
specified by the user. The “04p” ACE tables were introduced in 2002 and contain the
first completely new data set since 1982. [White, 2002]
Table 2.3
Table
“01p”
“02p”
“03p”
“04p”
Differences of photo atomic interaction data libraries.
Library / Source
Storm & Israel
Description
Z = 84, 85, 87, 88,
89, 91 and 93
Energy
1 keV to 15 MeV
ENDF/B-IV
Z = 1 to 94, except
stated above
1 keV to 100 MeV
Identical to “01p”
Livermore Evaluated
Photon Data Libraries,
(EPDL 89)
Identical to “02p” with
addition of probability
and momentum profile
data from Biggs et. al
ENDF/B-VI.8
Below 10 MeV
10 MeV and above
(up to 100 GeV)
1 keV to 100 GeV
Z = 1 to 100
1 keV to 100 GeV
2.5.3 Variance Reduction
Variance reduction in MCNP5 minimizing the computer time needed to
obtain result with acceptable relative error as well as pass other statistical check. In
order to reduce this computational effort, there are two basic approaches that can be
applied in MCNP:
1. Simplifying the MCNP geometry and physics model
2. Use non-analogue simulation.
The first approach is the basic technique in reducing the time consuming
simulation. One may want to model their problem’s geometry in the simulation as
close as possible to the real problem. However, it is waste of time to model the
geometry that has a little influence or significance to the desired result. Hence, the
user can just simplify or truncated any geometry and the physics used in the particle
simulation.
23
Non analogue simulation is second basic approach that offers the user to
reduce the variance of the tally by modifying the simulation process itself. MCNP5
offers a variety of variance reduction method which had been categorized as follows:
1. Population Control Methods: These methods control the number of
particle in spatial or energy regions that are relevant / irrelevant to the
tally score by artificially increase / decrease the number. Example of
population control methods are Geometry splitting and Russian roulette
(IMP), Energy splitting / roulette (ESPLT), weight cutoff (CUT, WWT)
and weight windows (WWE, WWN, WWP, WWG, WWGE)
2. Modified Sampling Method: These method increase the probability of a
particle reaches the tally region. Include in the MCNP are Exponential
transform (EXT, VECT), Implicit Capture (PHYS), Source direction and
energy biasing (SDEF, SP, SI, SB), Forced collision (FCL),
Bremsstrahlung
biasing
(BBREM)
and
Neutron-induced
photon
production biasing (PWT).
3. Partially Deterministic Method: These methods utilized deterministic
process instead of random-walk process in moving particles between the
regions. Examples of this methods are Point and ring detectors (F5a),
DXTRAN spheres ( DXT, DXC) and correlated sampling (PD).
2.5.4 Statistical check in MCNP5.
Output result generate by MCNP5 provides rich of information about
the simulation that allow user to assess the precision of the result. Statistical
test summarized ten statistical checks it performs on the tally. The 10 test are
summarized below. Although MCNP5 provide this test, it is not foolproof.
Users still have to rely on their understanding to check for the reliability of
the simulation.
24
8:
Tally Mean, 1. The means must exhibit, for the last half of the problem, only
random fluctuation as number of histories (N) increase. No up or
down trend must be exhibited.
Relative Error, R:
2. R must be less than (0.05 for point/ring detectors)
3. R must decrease monotonically with N for the last half of the
problem.
4. R must decrease as 1⁄√; for the last half of the problem.
Variance of the Variance, VOV:
5. The magnitude of the VOV must be less than 0.1 for all types of
tallies
6. VOV must decrease monotonically for the last half of the
problem.
7. VOV must decrease as 1⁄√; for the last half of the problem.
Figure of Merit, FOM:
8. FOM must remains statistically constant for the last half of the
problem.
9. FOM must exhibit no monotonic up or downs trend in the last half
of the problem.
Tally PDF, %():
10. The SLOPE determined from the 210 largest scoring events must
be greater than 3.
CHAPTER 3
RESEARCH METHODOLOGY
3.1
INTRODUCTION
As mention earlier, this research will apply Monte Carlo n-particle code
version 5 (MCNP5) as tool to investigate the photon energy response of two different
TLD material which are TLD 100 and SiO2 fibre optic. The results from the MCNP5
were then compared with the experimental result from calculated value and Glennie
[2003]. Thus, the simulation setup used in this research will follow the experimental
setup by Glennie [2003]. Figure 3.1 shows the flow chart of this research.
Fatal error
Collect information
for input file
Create new input
file or modified
using Notepad and
MCNP Vised
Run the simulation
using MCNP5
Success
Obtained simulation
result and compare
with experiment
and calculation
Figure 3.1
Manipulate data
using Microsoft
Excel
Retrieve data from
output file
Flow chart of the work in this research
26
3.2
CREATING INPUT FILE
3.2.1 Geometry Description
The geometry is specified by a phantom, TLD tray and the TLDs. TLD are
placed in the plastic tray. A plastic tray contains 50 wells, so 50 TLDs can be held in
a single tray. The wells are 1 cm x 1 cm square and 1.0 mm deep. The sizes of the
well are large enough to hold the TLDs and small enough to not significantly affect
radiation fluence [Glennie, 2003]
The phantom is a standard PMMA phantom with external dimensions of
30 cm x 30 cm x 23 cm. There is a rectangular cut-out on the phantom that is
sufficient enough to place the TLD tray. For the purpose of simulation, the TLD tray
is assumed to be built inside the phantom itself.
Two types of TLD were used in this simulation, which are TLD 100 and
silicon dioxide fibre optic (SiO2 fibre optic). The TLD 100 has a dimension of
3.3mm x 3.3mm square and 0.9 mm thick and it has densities of 2.64 g cm-3. For the
SiO2 fibre optic, the initial length is 10 meter long but for this research the fibre was
cut into 5mm length. The fibre had 125.0 ± 0.1 μm cores and the mass of each fibre
was (0.20 r 0.02) mg. SiO2 fibre optic has densities of 2.32 g cm-3. [Hashim, 2009].
Figure 3.2 show cell card and surface card in the MCNP5 input file. All the
geometry information was written in these cards. All the information for TLD 100
simulation and fibre optic simulation were remains same except for yellow
highlighted information shown in Figure 3.2.
27
TLD 100 RESPONSE TO PHOTON
c Beginning of cell cards
c
10 1 -2.64
-1
u=1
20 2 -0.00192 1:2
u=12
21 2 -0.00192 -3
fill=1 u=2 lat=1
30 2 -0.00192 -2
fill=2
40 3 -1.19
#30 -4
50 2 -0.00192 4 -5
60 0
5
c
c end of cell cards
c
c
1
2
3
4
5
c
c
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=0
$TLD-100
$Septa
$lattice
$Tray
$Phantom
$Sphere
$Universe
Beginning of surfaces cards
rpp
rpp
rpp
rpp
so
-0.165 0.165 -0.165 0.165 -0.045 0.045
-4.5 5.5 -2.5 2.5 -0.25 0.25
-0.491 0.499 -0.491 0.499 -0.249 0.249
-15 15 -15 15 -21.5 1.5
120
$TLD-100
$Tray dimension
$Lattice window
$Phantom
$sphere
End of surfaces card
Figure 3.2
Problem’s geometry information as written in cell and surface card
Z-axis
Source
X-Y View
TLD’s
Tray
X-Z View
Phantom
X-axis
Figure 3.3
Y-axis
3D view of problem’s geometry.
28
Figure 3.3 depict the problem’s geometry in XYZ axis. TLD’s tray was
located 1.5 cm under the phantom surface and the point source is 100 cm from
phantom surface. Figure 3.4 and 3.5 show side view and plane view of the problem’s
geometry. Both of these figures were printed from MCNP Vised and were used to
look out any discrepancies in the cell card or surface card. MCNP Vised will
automatically assign colour to distinguish materials in the geometry. Further
information about materials will be discussed in the next section.
Any geometry error can be easily detected using MCNP Vised. Any cell or
surface that cannot be read or interpret, MCNP Vised will make dashed line to that
particular cell or surface. A complete cell or surface will be indicated by full outline.
Tray
Phantom
Dry Air
Figure 3.4
X-Z side view of the phantom.
A TLD tray can be seen in the phantom.
It is located 1.5 cm from the phantom surface.
29
TLD
Tray
Phantom
Dry Air
Figure 3.5
X-Y plane view of the phantom.
The tray’s wells can be seen here. The tray has a size 10 cm x 5 cm external
dimension and contained 50 square wells with dimension of 1 cm x 1 cm.
3.2.2 Data card specification
Data card in the MCNP input file contains information such as source,
material, tallies and variance reduction. The source and type of radiation particles for
an MCNP problem are specified by the SDEF command. Only one SDEF card is
allowed in an input file. In this simulation, the source is photon and defined as a point
source collimated into a cone of direction. The source energy is between 20 keV
(0.02 MeV) to 20 MeV and was placed 100 cm from the phantom surface with each
run ending at 10 million histories. Figure 3.6 describe the direction and distribution
of particle from source to the phantom.
30
Point source.
100 cm from
phantom surface
Figure 3.6
Distribution of 100 000 particles from source with energy 1.25 MeV.
It can be seen that the particle were distribute into a cone of direction.
The TLD 100 or LiF: Mg,Ti contains 73.28 % Fluorine 7, 26.72 % Lithium,
200 ppm Magnesium and approximately 10 ppm of Titanium. [Saint, 2000] Type of
fibre used in this research is germanium doped SiO2 fibre optic. The fibre had
average weight fraction as follows:
a) Silica
: 53.6 %
b) Oxygen
: 46.1 %
c) Germanium
: 0.3 %
As mention before, the fibre initial length is 10 m, but for the purpose of this
research, the fibre was cut into 50 tiny pieces. Each pieces had 5 mm long. The
fractions of elements in each fibre were measured by SEM technique as measured by
Yaakob [2011]. In Figure 3.4 and 3.5, reader may find three difference colour which
are red, blue and yellow. The colour in those figures represents a difference material
of the geometry. Table 3.1 below listed the material description used in this research.
31
Table 3.1
Number of
Material
Material Description
Material Name
m1
TLD 100
m1
SiO2 Optical fiber
m2
Dry Air
m3
PMMA Phantom
Material Fraction
Lithium
Fluorine
Magnesium
Titanium
Silica
Oxygen
Germanium
Carbon
Nitrogen
Oxygen
Argon
Hydrogen
Oxygen
Carbon
Color
Red
Red
Blue
Yellow
User could also specify the photon interaction data used in the data card.
Photoatomic data are stored on ACE tables that use ZAIDs with the form
ZZZ000.nnP. For example, for material m1, one of the element content in m1 is
lithium where atomic number, Z = 3. Table 3.2 shows example how the photoatomic
data libraries were written and used in MCNP. Refer to chapter 2 for more details on
source of the photoatomic data.
Table 3.2
ZAIDs
3000.01p
3000.02p
3000.03p
3000.04p
3000
Photoatomic data libraries
Description
Photoatomic data libraries from table “01p” were used
Photoatomic data libraries from table “02p” were used
Photoatomic data libraries from table “03p” were used
Photoatomic data libraries from table “04p” were used
Photoatomic data libraries from table “04p” were used
The other important parameter in data card is tallies. MCNP provides seven
standard tally types. For this research tally F6 was used. According to the MCNP
User’s Manual, tally F6 is the track heating tallies modified to tally a reaction rate
convolved with an energy-dependant heating function instead of flux. The units of
this tally are MeV/g.
There are two variance reduction technique were applied in this simulation.
First technique was a geometry splitting and Russian roulette where the photon
32
importances (IMP:P) are assigned to each cell and weighted it. A sphere with radius
120 cm was also created in this model. This sphere would separate the desired
regions of simulation with the universe. IMP: P was set to zero at regions outside
sphere. One would say this sphere act as a boundary for MCNP tracks the particle.
Any photon that reach the outside boundaries, will be not be tallied by MCNP.
IMP:P was set to unity at all cells inside the boundary. In Figure 3.6 obviously shows
the sphere. MCNP Vised coded anything outside sphere with white colour to denote
zero importance of photon.
Second technique applied was source biasing. Biasing the source is one of the
easiest non-analog variance reduction techniques. MCNP5 source distribution would
be isotropic by default. Source can be biased by the user in the SDEF card by using
SI card, SB card and SP card. The source was biased so it will collimate into a cone
of direction instead of isotropic as depict in Figure 3.6.
c Beginning of data cards
mode p
sdef pos=0 0 101.5 erg=20 PAR=2 vec=0 0 -1 dir=d1
SI1 -1 0.9999 1
$Histogram for cosine bin limit
$Fraction solid angle for each bin
SP1 0 0 1
SB1 0. 0. 1.
$Source bias for each bin
m1 3000 -0.2672
$TLD-100
9000 -0.7328
12000 -0.0002
22000 -0.00001
m2 6012 -0.000124
$Dry air
7014 -0.755268
8016 -0.231781
18000 -0.012827
m3 1001 -0.0805259 1002 -0.0000121
$Phantom
8016 -0.3194907 8017 -0.0001217
6012 -0.593190 6013 -0.00665831
f6:p 10
nps 1e7
Figure 3.7 Data card
Figure 3.7 show data card in the MCNP5 input file. All the information
discussed beforehand was written in this card. Input line for TLD 100 simulation and
fibre optic simulation were remains same except for yellow highlighted line. Green
highlighted line indicates the energy of the problem and that information can be
change to desire energy of the research (in unit of MeV). Example of a complete
input file can be found at Appendix A for TLD 100 and Appendix B for fibre optic.
33
3.3
RUNNING THE SIMULATION
After the input file is completed, the simulation can be run using MCNP5
command prompt. Figure 3.8 show the captured image of MCNP5 command prompt.
This figure shows example of simulation for file “tldnew” at energy of 1.25 MeV.
The simulation was terminated after 10 million particle histories were done and it
took 9.55 minutes of computer time to complete. If there any major discrepancy in
the input file because of user violates a basic constraint of the input specification,
MCNP5 will terminate the simulation before running any particle. This is called fatal
error. If fatal error occur, the input file must be examine once again and search for
any misinformation that contribute to fatal error.
Most MCNP error messages are either warnings or comments and are not
fatal. Warnings are intended to inform the user about unconventional input
parameters or running conditions and may need further attention. Comments relay
useful additional information to the user. The user should not ignore these messages
but should understand their significance before making important calculations. [X-5
MC Team, 2005].
Figure 3.8
MCNP5 command prompts
34
After the simulation was done, MCNP5 will generate an output file for each
input file. User can named the output file or MCNP5 will name it automatically.
Example in Figure 3.8 shows the output file was given name as “1250kev”.
Appendix C is the example of output file generated by MCNP5.
3.4
INTERPRETING MCNP5 OUTPUT DATA
TLD response to photon was determined as energy absorbed in the TL
material, which assumed to be proportional to the light output. This relationship is
given by [Miles, 1994]
<> -5! ∝ %() @ &()
(3.1)
Where,
%() = average fraction of energy deposited in TL material, for any given energy
@
= photon fluence
= incident photon energy
&() = energy dependent relative TL efficiency, assumed to be unity
Radiation exposure is proportional to the photon fluence, photon energy and
the mass energy absorption coefficient for air. Thus,
5A$ ∝ @('! ⁄")*+$()
(3.2)
where
('! ⁄")*+$() =
the mass energy absorption coefficient for air as a function of
the incident photon energy
Combining Eq. (4) and Eq. (5) gives
%()@
<> -5!
∝
@('! ⁄")*+$()
5A$
(3.3)
35
which reduces to
%()
('! ⁄")*+$()
(3.4)
And normalizing to 1.25 MeV gives
%()
H
('! ⁄")*+$()
-B*C+D <> -5!, -- =
%(.. 01 23)
H
G
('! ⁄")*+$(..01 23)
G
(3.5)
From the above expression, the %() values from the MCNP calculation can
be used to obtain the relative energy dependence of various types of TL materials
with the results normalize to 1.25 MeV. The relative energy response is defined with
respect to 1.25 MeV which is average photon energy from a 60Co source (1.17 MeV
and 1.33 MeV photon per
60
Co disintegration). In TLD applications, TL output of
TLDs exposed to X-ray and gamma rays of different energies are usually
standardized against average energy of 60Co source.
CHAPTER 4
RESULTS AND DISCUSSION
4.1
INTRODUCTION
The MCNP5 simulation only produces the raw data. A number of information
was retrieved, being used and analyzed for the purpose of this research. The example
output file generated by MCNP5 can be found in Appendix C. There are two
methods in determine photon energy response, which are by calculation and by
MCNP5 simulation.
4.2
PHOTON ENERGY RESPONSE OF TLD 100
4.2.1
Photon energy response of TLD 100 by calculation
Photon energy response can be determined by calculation using Eq. 4.1.
μ
!" (4.1)
#
() = μ
!
" $%
The parameter required to calculate photon energy response is mass energy
absorption coefficient of material and reference material. In this research, air is used
as a reference material and mass energy absorption coefficient obtain by calculation
will be denoted as MEAC. The mass energy absorption coefficients for air were
obtained from Tables of X-Ray Mass Attenuation Coefficients and Mass Energy-
37
Absorption Coefficients by Hubbell and Seltzer [1995] and were attached as
Appendix D.
Table 4.1 lists weight fraction, I+ of the element contains in TLD 100 used
for this research. This list was used to determine the mass energy absorption
coefficient for mixture material as shown in Eq. 4.2
(
'!
'!
)#*C$+*B = J(
) I
"
" + +
(4.2)
+
Weight fraction of elements contain in TLD 100
Table 4.1
TLD 100 Weight Fraction,W
Lithium
Fluorine
Magnesium
Titanium
0.2672
0.7328
0.0002
1.0E-05
Using Eq. (4.2), the mass energy absorption coefficient of TLD 100 can be
calculated as follows:
'!
"
<> .77
= K(
'!
)
× 7. 06L0M + K(
" <+
'
+ K( ! )2R
"
'!
)
" O
'!
× 7. 7770M + K(
× 7. LP0QM
(4.3)
)+ × . × .7S1 M
"
The mass energy absorption coefficient each element in the TLD 100 was also took
from Tables of X-Ray Mass Attenuation Coefficients and Mass Energy-Absorption
Coefficients [Hubbell and Seltzer, 1995] and was attached as Appendix E.
Calculated value of mass energy absorption coefficient of TLD 100
Table 4.2
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
0.20
'!
"
<> .77
2
-1
(cm g )
0.6502
0.1828
0.0790
0.0455
0.0323
0.0239
0.0223
0.0233
0.0248
Energy, E
(MeV)
0.30
0.40
0.50
0.60
0.80
1.00
1.25
1.50
2.00
'!
"
<> .77
2
-1
(cm g )
0.0266
0.0273
0.0275
0.0274
0.0267
0.0259
0.0247
0.0236
0.0217
Energy, E
(MeV)
3.00
4.00
5.00
6.00
8.00
10.00
15.00
20.00
'!
"
<> .77
(cm2 g-1)
0.0191
0.0173
0.0162
0.0153
0.0142
0.0135
0.0126
0.0122
38
Table 4.2 shows MEAC as calculated using Equation 4.3. Using Equation
4.1, photon energy response for TLD 100 can be calculated. For an easy comparison,
the photon energy response was normalized to 1.25 MeV to obtain relative energy
response. Equation 4.4 shows the equation used to acquire relative energy response,
RER.
-B*C+D !$RT -5! (--) =
Table 4.3
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
0.20
4.2.2
()
(.. 01 23)
(4.4)
Relative energy response of TLD 100 as calculated using Eq. 4.4
RER
1.30
1.28
1.25
1.20
1.15
1.07
1.04
1.01
1.00
Energy, E
(MeV)
0.30
0.40
0.50
0.60
0.80
1.00
1.25
1.50
2.00
RER
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Energy, E
(MeV)
3.00
4.00
5.00
6.00
8.00
10.00
15.00
20.00
RER
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Photon energy response of TLD 100 by MCNP5 simulation.
For the response of TLD 100 by simulation, 22 input file were run by
MCNP5 with each of the file contain same information except for the source energy.
The energy used were 20, 30, 40, 50, 60, 80, 100, 150, 200, 300, 400, 500, 600, 800
keV and 1, 1.25, 2, 3, 6, 10, 15 and 20 MeV. Tally F6 was utilized to generate the
photon energy response of TLD. The output data from the simulation contains much
information that can be used. But the value retrieved from the file is energy absorbed
normalized to a unit mass of the material, Eabs.
Value of () is average fraction of energy deposited in TL material
obtained from MCNP simulation and it is actually the mass energy absorption
coefficient of the TLD 100. () will be used afterward to denote the mass energy
absorption coefficient obtain by simulation. The result extract from MCNP is in unit
39
of MeV/g and it means average energy deposited from a particle to a unit gram of the
dosimeter, Eabs. The following equation is used to obtain mass energy absorption
coefficient of the dosimeter by simulation.
%() = *U ×
Where %() =
.V#0
23
(4.5)
average fraction of energy deposited in TL material, for any
given energy (cm2 g-1)
*U
=
23 =
Energy absorbed normalized to unit mass (MeV g-1)
Incident photon energy (MeV).
Table 4.4 shows () value, calculated using Eq. (4.5) whereas Figure 4.1
depicts mass energy absorption of TLD 100 with comparison between calculated and
simulation.
Table 4.4
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
f(E)
Value of () for TLD 100
f(E)
(cm g )
Energy, E
(MeV)
(cm g )
0.3638
0.1998
0.1070
0.0667
0.0483
0.0342
0.0302
0.0290
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
0.0297
0.0308
0.0312
0.0311
0.0308
0.0300
0.0290
0.0277
2
-1
2
-1
f(E)
Energy, E
(MeV)
(cm2 g-1)
2.00
3.00
6.00
10.00
15.00
20.00
0.0245
0.0216
0.0176
0.0158
0.0151
0.0150
Table 4.5 shows the photon energy response of TLD 100 at incident energy, S(E) and
were obtain using Eq. (4.6).
% ()
() = μ
!" *+$ ()
(4.6)
40
Response of TLD 100 at incident energy by simulation
Table 4.5
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
()
0.736
1.433
1.713
1.748
1.671
1.450
1.307
1.163
Energy, E
(MeV)
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
()
1.113
1.073
1.057
1.047
1.044
1.040
1.037
1.036
Energy, E
(MeV)
2.00
3.00
6.00
10.00
15.00
20.00
()
1.039
1.047
1.073
1.092
1.119
1.131
Using Eq. (4.7), the responses were then normalized to average energy of
60
Co source. Table 4.6 shows the relative energy response, (RER) of TLD 100.
-B*C+D !$RT -5! (--) =
Table 4.6
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
-0.71
1.38
1.65
1.69
1.61
1.40
1.26
1.12
()
(.. 01 23)
(4.7)
Relative energy response (RER) of TLD 100.
Energy, E
(MeV)
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
-1.07
1.04
1.02
1.01
1.01
1.00
1.00
1.00
Energy, E
(MeV)
2.00
3.00
6.00
10.00
15.00
20.00
-1.00
1.01
1.04
1.05
1.08
1.09
41
Mass Energy Absorption Coefficient (cm2 g-1)
1.0E+00
TLD 100 (calculated)
TLD 100 (Simulation)
1.0E-01
1.0E-02
10
100
1000
10000
100000
Photon Energy ( keV)
Figure 4.1
Comparison of mass energy absorption coefficients between
calculated (MEAC) and simulation(()).
Equation 4.1 obviously shows the mass energy absorption coefficient was
needed in order to have photon energy response of dosimeters. Thus, it is essential to
compare mass energy absorption coefficient obtain from calculation, MEAC with
mass energy absorption coefficient from simulation,(). From the graph, it shows
value of () is slightly higher than MEAC except for energy of 20 keV. It is
common for a result from MCNP slightly higher than calculated since in MCNP
there are a few simplifying assumption made from the actual model. This simplifying
assumption could be a factor for a difference value between calculated and
simulation.
42
........... ANSI acceptable range
Figure 4.2
Comparison of TLD 100 relative energy response between
calculation, experiment and simulation.
Figure 4.2 depict the comparison of relative energy response between
calculated, simulation and experiment. Also shown in Figure 4.2 as dotted line, is the
acceptable response range defined in ANSI N545-1975. These line was use to
indicate constant response range that can satisfy dosimeter performance criteria as
specified in ANSI N545. One of the dosimeter performance criteria specified in
ANSI N545 that relate to energy response is “The response of TLD to photon shall
be determine for several energies between 30 keV and 3 MeV. The response shall not
differ from that obtained with calibration source by more than 20 percent for photon
with energies greater than 80 keV and shall not be enhanced by more than a factor
of two for photons with energies less than 80 keV”.
From Figure 4.2, calculated response shows that TLD 100 has constant
response at as low as 100 keV and slightly over response at energy lower than 100
keV. However simulation and experimental response were in agreement as it both
shows over response in the 150 keV range. This over response also has been
documented by Jones [1989], McKinlay [1981], McKeever et. al. [ 1995], and Mobit
et. al. [1996]. Simulation response shows 20 keV energy response yield less TL than
43
Co-60 which is in agreement with other research [Barber et. al., 1976; Ahmed et. al.,
1989; Olivera et. al., 1994; Aschan et. al., 1999]. This is because low energy photon
(< 20 keV) are significantly attenuated within the thickness of TLD, resulting in a
self shielding effect [Morgan et. al., 1977; Horowitz et. al., 1978; Olivera et. al.,
1994; Aschan et. al., 1999]
Graph in Figure 4.2 had shows TLD 100 was relatively had constant
response. In terms of energy dependence, TLD 100 is sufficiently flat to make it
attractive for environmental and personnel dosimetry, where a wide ranges of
exposures energies is possible.
4.3
PHOTON ENERGY RESPONSE OF SILICON DIOXIDE FIBRE
OPTIC
4.3.1
Photon energy response of silicon dioxide fibre optic by calculation
Same procedures as TLD 100 were applied in order to calculate photon
energy response for fibre optic. Table 4.7 lists weight fraction of the element
contains in TLD 100 used for this research. This list was used to determine the mass
energy absorption coefficient for mixture materials.
Weight fraction of elements contain in SiO2 fibre optic
Table 4.7
Fibre Optic Weight Fraction, W
Oxygen
Silicon
Germanium
0.4612
0.5364
0.00234
Using Eq. (4.8), the mass energy absorption coefficient of SiO2 fibre optic
can be calculated as follows:
'!
"
+W0 O+U$
= K(
+ K(
'!
"
'!
"
)+ × 7. 1P6XM + K(
)Y × 7. 770PXM
'!
"
)W × 7. X6.0M
(4.8)
44
The mass energy absorption coefficients of each element in the SiO2 fibre
optic were also took from Tables of X-Ray Mass Attenuation Coefficients and Mass
Energy-Absorption Coefficients [Hubbell and Seltzer, 1995] and was attached as
Appendix F.
Table 4.8
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
0.20
Table 4.9
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
0.20
Calculated value of mass energy absorption coefficient for fibre optic
'!
"
Z[
(cm2 g-1)
2.5454
0.7303
0.3032
0.1571
0.0955
0.0500
0.0360
0.0284
0.0281
Energy, E
(MeV)
0.30
0.40
0.50
0.60
0.80
1.00
1.25
1.50
2.00
'!
"
Z[
(cm2 g-1)
0.0291
0.0296
0.0297
0.0295
0.0288
0.0278
0.0266
0.0254
0.0235
Energy, E
(MeV)
3.00
4.00
5.00
6.00
8.00
10.00
15.00
20.00
'!
"
Z[
(cm2 g-1)
0.0208
0.0193
0.0182
0.0175
0.0167
0.0163
0.0159
0.0158
Relative energy response of fibre optic as calculated using Eq. 4.4
RER
4.74
4.76
4.45
3.84
3.15
2.08
1.44
1.07
0.98
Energy, E
(MeV)
0.30
0.40
0.50
0.60
0.80
1.00
1.25
1.50
2.00
RER
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Energy, E
(MeV)
3.00
4.00
5.00
6.00
8.00
10.00
15.00
20.00
RER
1.02
1.03
1.05
1.07
1.10
1.13
1.18
1.20
45
4.3.2
Photon energy response of silicon dioxide fibre optic by MCNP5
simulation.
For the response of fibre optic by simulation, same procedures were applied
in order to obtain mass energy absorption coefficient, energy response and relative
energy response.
Table 4.10 shows () value, calculated using Eq. (4.5) for fibre optic
whereas.
Table 4.10
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
f(E)
Value of () for fibre optic
f(E)
(cm g )
Energy, E
(MeV)
(cm g )
1.4312
0.7949
0.4104
0.2361
0.1518
0.0816
0.0564
0.0389
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
0.0355
0.0342
0.0340
0.0336
0.0332
0.0321
0.0310
0.0296
2
-1
2
-1
f(E)
Energy, E
(MeV)
(cm2 g-1)
2.00
3.00
6.00
10.00
15.00
20.00
0.0262
0.0234
0.0202
0.0192
0.0193
0.0197
Table 4.11 shows the photon energy response of fibre optic at incident
energy, S(E) and were obtain using Eq. (4.6).
Table 4.11
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
()
2.656
5.172
6.006
5.761
4.993
3.390
2.425
1.558
Response of fibre optic at incident energy
Energy, E
(MeV)
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
()
1.328
1.192
1.153
1.134
1.124
1.114
1.111
1.110
Energy, E
(MeV)
2.00
3.00
6.00
10.00
15.00
20.00
()
1.117
1.139
1.224
1.327
1.428
1.506
46
Mass Energy Absorption Coefficient (cm2 g-1)
1.0E+01
Fibre Optic (Calculated)
Fibre Optic (Simulation)
1.0E+00
1.0E-01
1.0E-02
10
Figure 4.3
100
1000
Photon Energy ( keV)
100000
10000
Comparison of mass energy absorption coefficients between
calculated and simulation.
Figure 4.3 depicts mass energy absorption coefficient of fibre optic with
comparison between calculated and simulation. Both result of mass energy
absorption coefficient for TLD 100 and fibre optic show similarity. From Figure 4.3,
just like TLD 100, () of fibre optic is slightly higher than MEAC except for
energy of 20 keV. As TLD 100, difference values between calculated and simulation
could be due to simplifying assumption in generating simulation.
Using Eq. (4.7), the responses were then normalized to average energy of
60
Co source. Table 4.12 shows the relative energy response, (RER) of fibre optic.
Table 4.12
Energy, E
(MeV)
0.02
0.03
0.04
0.05
0.06
0.08
0.10
0.15
-2.39
4.66
5.41
5.19
4.50
3.05
2.18
1.40
Relative energy response (RER) of fibre optic
Energy, E
(MeV)
0.20
0.30
0.40
0.50
0.60
0.80
1.00
1.25
-1.20
1.07
1.04
1.02
1.01
1.00
1.00
1.00
Energy, E
(MeV)
2.00
3.00
6.00
10.00
15.00
20.00
-1.01
1.03
1.10
1.20
1.29
1.36
47
........... ANSI acceptable range
Figure 4.4
Comparison of fibre optic relative energy response between
calculated, simulation and experiment.
From Figure 4.4, calculated response shows that fibre optic has constant
response in the range 150 keV to 10 MeV whereas simulation response started
constant at 200 keV and up to 10 MeV. Calculated and simulation response were in
agreement as it both shows over response below 150 keV range. Unlike TLD 100,
fibre optic response started to over response at energy higher than 10 MeV.
As mention before, ANSI acceptable range was significant in order to obtain
information of over-response energy. Figure 4.2 shows TLD 100 was relatively had
constant response over wide range of energy. On the other hand, fibre optic had a
limited range of constant response. Flat response of fibre optic is ranging from 200
keV to 10 MeV. This means, energy correction method will be needed if fibre optic
want to be used at energy outside of these ranges. For instance, perhaps a suitable
metal filter may be needed if fibre optic is to be used as environmental dosimeter
because at energy lower than 200 keV, the measured dose is deviates drastically from
the dose to be determine.
48
4.4
EFFECT OF DIFFERENT DOPANT CONCENTRATION ON TL
ENERGY RESPONSE
In Chapter 3, it was mentioned that fibre optic used in this research was cut
into 5 mm long from the 10 meter of fibre optic. Studying the effect of dopant
concentration arise from the fact that the concentration of dopant along this 10 meter
of fibre optic may be varies in the ranges zero to 0.8% mol. Thus, 30 input file were
simulated using tally F6 and photon energy of 1.25 MeV, in order to study the
behaviour of different concentration of germanium in the fibre optic.
Figure 4.5 depict the responses of the fibre optic with different germanium
concentration. It is obviously shows that photon energy response of fibre optic had
not affected even though the concentration of the germanium is different along the
fibre for dopant concentrations of zero percent – 0.8 percent mol.
1.15
Photon Energy Response at 1.25 MeV
(Simulation)
Response
1.10
1.05
1.00
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Dopant Concentration ( % mol)
Figure 4.5
Effect of dopant concentration on germanium doped SiO2 fibre optic
response
49
4.5
DISCUSSIONS ON MCNP5 SIMULATION
Note that in Figure 4.2 and 4.4 that the simulation response is higher than the
experiment and calculated response especially at lower energy. This discrepancies
could be due to value of η(E) or energy-dependent relative TL efficiency. The
available data were inadequate and inconsistent. For example, TLD 100 values for
η(100 keV) have been reported to be 1.05 [Horowitz, 1999], 1.06 [ Davis et al,
2003] and 1.09 [Hranitzky et. al., 2006]. Due to this fact, value of η(E) in this study
was assumed to be unity for all calculations of TLD 100 or fibre optic responses.
This assumption could be the primary cause of discrepancies between simulations
with calculated and experiment.
Preparing input file was started with modelling the geometry of the problem.
To get geometry modelling as close as the actual was quite challenging at first, but
with the aid of MCNP Vised, the modelling was easily constructed and modified.
Figure 3.2, 3.3 and 3.4 in the chapter 3 are the example of geometry as displayed by
MCNP Vised. Despite of wide selection of menu options that possible for immediate
visualization of the geometry, MCNP Vised was easily crash. The creation of this
research input file were not done fully using MCNP Vised. It was created manually
using line editor (Microsoft Notepad) at first, and then MCNP Vised was used to
display the 2D plane of the problem’s geometry in order to look upon any error in the
input file. Other than displaying geometry, MCNP Vised was also used to view at
distribution of particles from the source and the position of the source from the
target.
Time taken to run the simulation is varied with photon energy, variance
reduction technique used and also the computer capability. For TLD 100 simulation,
average time taken to complete the simulation is 8.34 minutes whereas average
computer time taken to simulate fibre optic is 7.81 minutes. Fibre optic simulations
were generated using personal computer meanwhile TLD 100 files were simulate
using laptop. If these two files were run using same computer, the difference in
computer time would be slightly difference. For photon energy, MCNP5 took most
time to simulate problems in energy range of 100 keV to 800 keV. This is probably
50
because within that energy, interactions processes take place in the material involve
photoelectric absorption and also Compton scattering.
After the simulation is complete, MCNP5 produced two files namely runtp
file and output file. Both file can be renamed by the user in the command prompt
line. The () values were retrieved from the output file. Before the data were taken,
the output files were analyzed first in order to determine the reliability of the
calculation made by MCNP5. Hence, table of 10 statistical checks was used.
In preliminary TLD 100 simulations, only 10 out of 22 files were passing all
ten statistical checks. The other files however, fail to pass Figure of Merit (FOM)
checks. FOM gave the indicator whether the tally is well behaved or not. FOM must
remains statistically constant and exhibit no monotonic up or downs trend in the last
half of the problem [X-5 MC Team, 2005]. Since this is the random process, the
failed files were simulated again using different seed number with the aim that the
file will pass all ten statistical checks. Fortunately, all files that were failed before
had pass all ten statistical check with the result differ less than 1 percent. Unlike
TLD 100, only 5 fibre optic simulations were failed to pass all ten statistical checks.
Same procedure was applied and produces better statistical checks.
Even though MCNP5 provides statistical check, this test was meant for the
whole geometry of the problems, not the desired cells. To get uncertainties of the
specified cells (cell 10 or the TLD response), all files were simulated 3 times, with 3
different seed. The data presented in the previous sub chapter was the average result
from 3 different simulations of the files.
All data and result presented before this is the chosen and successful one.
However, several attempt were made before that data were obtained and being
analyzed. This section will present the parameters that involve in the changes made
from the beginning until this simulation generates desired and reliable result.
1. Number of particles, NPS:
Number of particle history (NPS) used for this simulation are 10
million and it was written as NPS 1e7 in the input file. Number of particles
can be assumed as a given dose. For energy response study, the dose given
must be constant for all sources. Thus NPS used throughout of this research is
51
constant. However, MCNP5 will give same result even if the values of NPS
were different, but the error and computer time will be different. The more
histories were run, the better of the precision will be, but the greater computer
time will be expense. Preliminary simulations were made with three different
NPS which are NPS 1e6, NPS 1e7 and NPS 1e8. NPS 1e7 were chose because
it produces result with reliable error and the computer time expenses were
reasonable.
2. Source Distribution
Since size of both dosimeter used was small compared to the size of
the phantom, very few particles made it to the smaller TLD elements.
Changing the importance of the cells was found to be insufficient to increase
the number of particles passing through the elements. The photon source
was biased so that more particles can go towards the TLD. Source and type
of radiation in MCNP were specified by SDEF cards in MCNP5. In this
research, the source is photon and was defined as a point source collimated
into a cone of direction. By default, radiation distribution will be isotropic.
Hence, to direct source into the specific direction, user can used SI card, SB
card and SP card with the SDEF command. Biasing the sources make
significance changes to the tally statistical check as well as tally result. Table
4.13 shows result of three attempt made by this research with different SI
and SB card for TLD 100 with 1.25 MeV photon energy.
Table 4.13 Result produce by MCNP5 with two different SI and SB card.
3rd attempt result were better as more energy were absorbed.
1st
Attempt
2nd
Attempt
3rd
Attempt
SI
SB
Energy
absorbed,
MeV/g
Tracks
Entering
Error
-1 0.99 1
0 0.95 0.05
9.9888 E-04
108 133
0.0034
-1 0.999 1
0 0.95 0.05
8.87202E-03
841 423
0.0011
-1 0.9999 1
001
3.4589 E-02
1 327 769
0.0008
CHAPTER 5
CONCLUSION
5.1
SUMMARY OF FINDINGS
This research has successfully utilized MCNP5 simulation in determine
photon energy response of TL materials. Photon with energy ranges from 20 keV –
20 MeV were used to irradiate to two types of dosimeter, TLD 100 and SiO2 fibre
optic. Using tally F6, MCNP5 simulate photon energy deposited in both dosimeter.
Output result generate by MCNP5 were analyzed and being used to calculate photon
energy response. Simulation results were compared with calculation of relative
energy response. Besides calculation, TLD 100 energy responses were also
compared with experimental responses measured by Glennie [2003] whereas fibre
optic energy responses were compared with experiment measured by Abdulla [2003].
A good TL material should have a constant or flat response over wide energy
ranges. Owing to this fact, ANSI N545-1975 was chose to show the degree of energy
response deviation. First aim of this research is to determine photon energy response
of TLD 100 using MCNP5 simulation. Simulation response was in agreement with
existing data of experiment energy response [Barber et. al., 1976; Ahmed et. al.,
1989; Olivera et. al., 1994; Aschan et. al., 1999], where it shows over response in the
150 keV range. However, despite of over response in lower energy, TLD 100 has
relatively flat response because all responses were in the ANSI acceptable range. In
terms of energy dependence, it is suitable to be dosimeter in wide energy ranges
without deviates dose to be determined significantly.
53
TLD 100 was proven to have relatively constant response over wide range of
energy. Unlike TLD 100, flat response of fibre optic can only be seen in the range of
200 keV to 10 MeV. Obviously, an energy correction method will be needed if fibre
optic wants to be used at energy outside of these ranges.
Last objective of this research are to study the effect of germanium doped in
the fibre optic. The result shows that the different concentration of germanium along
the fibre does not affect the photon energy response. Discrepancies in the simulations
as compared to experiment and calculations are believed to be due to η(E) or energydependent relative TL efficiency of the dosimeters phosphor.
Overall, the research goals of this work have been well met. The energy
response of TLD 100 and fibre optic had been determined using MCNP5. This
research had proved the capability of using simulation in investigating TL material.
Moreover, the fibre optic energy responses acquire from this research provides more
information about its usefulness and indirectly evokes other idea for future research.
5.2
ADVANTAGES & DISADVANTAGES OF USING SIMULATION
The obvious advantages in using MCNP5 is time took to obtain the result.
Average time taken to simulate photon energy response was 8.1 minutes. This is
average time for one input file or energy. Thus it took a total of 176 minutes or equal
to 2.93 hour to simulate 22 file. All the analysis hereafter would only took maximum
of 2 hours in order to produce a photon energy response of any dosimeter. In contrast
with simulation, investigating the photon energy response by experiment would take
more time from simulation. Using real dosimeter, researcher would have to face
several procedures such as annealing; expose dosimeter to the source and also
reading the dosimeter. For example, annealing the TLD would took 4 hours, to
expose the TLD to the desired source is about 5 minutes per energy and lastly
reading the TLD is about 3 minutes per TLD. All of this processes will took about 7
hours to complete, excluding the analysis.
54
Beside time saving, there are several advantages that can be offered when
using simulation, specifically MCNP5, in investigating photon energy response or
any TL research. MCNP5 provides interactive software, called MCNP Vised, which
can assist the user to create the input file, as well as gave the user visual of the
geometry created. Other than that, MCNP also can be cost saving. For example, in
investigating new TL material, researcher can utilize simulation in study the photon
energy response of any material desired. Simulation can provide preliminary result of
photon energy response of certain material. In contrast with experiment, any desired
material, have to be prepared or bought first. This could be wasting of time and also
money, if later, the material was found out to be incapable of being good TL
material. In other words, simulation can give the researcher expected result, before
the experiment was carried out.
Despite of its time saving advantages, MCNP5 also had its own
disadvantages. MCNP5 will produce the result even if the parameter or geometry
were wrong or having errors. It will provide results for any inputs. Thus, to operate
MCNP5, the user must fully understand of MCNP5 feature as well as their research
aim. Failure of the user to fulfil these criteria will result in misinterpretation of
MCNP5 data and consequently generate the incorrect result.
Other problem in using MCNP5 or simulation is lack of expertise to refer
locally. There only a few local experts were known used MCNP5 for their research
and not one of them is in thermoluminescence research area. Most of TL research
groups still prefer using conventional method in investigating and studying
thermoluminescence. There a numbers of international TL research group that
applied simulation, however, MCNP5 were not the preferred software.
5.3
RECOMMENDATION FOR FUTURE RESEARCH
Applying simulation, specifically MCNP5, in investigating and studying TL
material has given a lot of idea for future research. This thesis had proved that
MCNP5 is capable of predicting TL responses to photon energy. Further research can
be conducted; perhaps using the same input file is investigating others promising TL
55
material response subjected to photon irradiation. All the information needed is just
atom fraction or weight fraction of the material, and its density. However, before
using existing input file, future research may start with analyzing the existing input
file and improve the percentage difference of the simulation result with theoretical
value.
Fibre optic had a limited range of flat response. Owing to that fact, this
research can be extend by finding the suitable filter for fibre optic, so it could had
more constant response especially at energy lower than 200keV. This could be done
with only a few modifications to the existing input file.
One knows that MCNP5 also capable of simulate singled or coupled neutron,
photon and electron transport. Thus, it is recommended for future research to apply
all available transport offered by MCNP5 in study the energy response of TL
material. Even though it may take some times to reconstruct new input file as the
electron and neutron transport were different from photon transport, it is then worth
trying.
More research can be conducted in future is by using others Monte Carlo
simulation available such as Geometry and Tracking computer code (GEANT4) or
Electron- Gamma Shower computer code (EGSnrc). GEANT4 is a platform for the
simulation of the passage of particles through matter using Monte Carlo methods.
Two interesting fact about GEANT4; it is the first to use object oriented
programming (in C++) and plus, it is free software. Mean while, EGS is a general
purpose package for the Monte Carlo simulation of the coupled transport of electrons
and photons in an arbitrary geometry for particles with energies from a few keV up
to several TeV. EGS software is developed by The SLAC National Accelerator
Laboratory, and was now maintained by National Research Council of Canada
(NRC).
56
REFERENCES
Abdulla, Y.A., Amin, Y.M. and Bradley, D.A. (2001a). The thermoluminescence
response of Ge-doped optical fibre subjected to photon irradiation.
Radiation Physics and Chemistry, Vol. 61, pp 409−410.
Abdulla, Y.A, Amin, Y.M and Bradley, D.A (2001b). The effect of dose and
annealing on TL sensitivity of germanium and erbium doped optical fibres.
Jurnal Fizik Malaysia, Vol. 22, No. 3 & 4, pp 49-53.
Abdulla, Y.A. (2003). The thermoluminescence response of Ge-doped and Er-doped
optical fibres in radiation therapy.
PhD Thesis. University of Malaya, Kuala Lumpur.
Ahmed, A., Barber, D., (1989), Sensitivity of LiF Thermoluminescent Dosemeters to
6 – 18 keV Photons. Physics in Medicine & Biology. Vol 34, pp 343-352.
Aschan. A., Toivonen. M., Lampinen. J., Tenhunen. M., Kairemo. K.,
Korppitommola. T., Jekunen. A., Sipila. P., Savolainen. S.,(1999). The Use of
TL Detectors in Dosimetry of Systemic Radiation Therapy. Acta Oncologica,
Vol 38, pp 189-186.
Attix, F.G. (1986). Introduction to Radiological Physics and Radiation Dosimetry.
New York : John Wiley & Sons
Barber. D.E., Moore. R., and Hutchinson. T. (1975). Response of LiF to 1.0 – 4.0
keV Electrons. Health Physics, Vol 28, pp 13-15.
57
Cameron, J. R.; Sunthralingam, N.; Kenney, G.N. (1968). Thermoluminescent
Dosimetry. Milwaukee, WS. The University of Wisconsin Press. pp. 483–
486.
Cullen D.E., Hubbel J. H., and Kissel L.D. (1997), EPDL97: The Evaluated Photon
Data Library, '97 Version. Lawrence Livermore National Laboratory report,
UCRL-50400, Vol. 6, Rev. 5.
Davis,S.D (2003). High Sensitivity Lithium Fluoride As a Detector For
Environmental Dosimetry.
Master Thesis. McGill University, Montreal.
Davis, S.D., Ross, C.K., Mobit, P.N., Van der Zwan, L., Chase, W. J., and Shortt,
K.R. (2003). The response of LiF thermoluminescent dosemeters to photon beams in
the energy range from 30kV x rays to Co-60 gamma rays.
Radiat. Prot. Dosim., Vol 106, pp33-43.
Eakins J.S, Bartlett D.T., Hager L.G., C. Molinos - Solsona and Tanner R.J., (2008).
The MCNP-4C Design of a Two Element Photon/Electron Dosemeter that
uses Magnesium/ Copper/ Phosporus Doped Lithium Fluoride.
Radiat. Prot. Dosim. Vol 128. pp 21 – 35.
Espinosa, G., Golzarri, J.I. et al. (2006). Commercial optical fibre as TLD material.
Radiat. Prot. Dosim. Vol. 18, pp 1-4.
Glennie, G.D (2003). A Comparison of TLD Dosimeters: LiF:Mg,Ti and LiF:Mg,
Cu, P For Measurement of Radiation Therapy Doses.
PhD Thesis. University of Virginia, USA.
Hashim, S. (2009). The thermoluminescence response of doped silicon dioxide
optical fibres to ionizing radiation.
PhD Thesis. Universiti Teknologi Malaysia, Skudai.
58
Horowitz Y.S., Muscovitch, M., Dubi. (1983). Modified General Cavity Theory
Applied to The Calculation of Gamma Dose in Co-60 Thermoluminescence
Dosimetry. Phys. Med. Biol. Vol 28 (7). Pp 829 – 840.
Horowitz Y. S., (1984). Thermoluminescence and Thermoluminescent Dosimetry.
Vol. I and Vol. II. Boca Raton, Florida. CRC Press Publishing.
Horowitz Y.S. (1999). The average distance between Mg-based trapping strructures
in LiF:Mg,Ti and LiF:Mg, Cu, P and the relevance microdosimetry.
Radiat. Prot. Dosim. Vol 182, pp 51-54.
Houston, A.L, Justus, B.L, Falkenstein, P.L., Miller, R.W., Ning, J., Altermus, R.,
(2002). Optically Stimulated Luminescent Glass Optical Fiber Dosimeter.
Radiat. Prot. Dosim. Vol 101. pp 23 – 26.
Hranitzky, C., H. Stadtmann and P. Olko. (2006). Determination of LiF:Mg, Ti and
LiF: Mg, Cu, P. TL Efficiency for X-rays and Their Application to Monte
Carlo Simulations of Dosemeter Response.
Radiat. Prot. Dosim., Vol. 119, pp 1–4.
Hubbell, J.H. and Seltzer, S.M. (1995), Tables of X-Ray Mass Attenuation
Coefficients and Mass Energy-Absorption Coefficients. National Institute of
Standards and Technology, Gaithersburg, MD.
Jones, A., Ohno, A., Richter, W. (1989). A Personal Dosimeter using LiF (Mg,Cu, P)
Thermoluminescent Material. Radiat. Prot. Dosim, Vol 27, pp 261-266.
Justus, B.L., Rychnovsky, S., Houston, A.L., Merritt, C.D., Pawlovich, K.J.. (1997).
Optically Stimulated Luminescnece Dosimetry using Doped Silica Glass.
Radiat. Prot. Dosim., Vol 74, pp 151 – 154.
Kortov V., (2007), Materials for thermoluminescent dosimetry: Current status and
future trends. Radiation Measurements. Vol. 42, pp 576-581.
59
McKeever S.W., Chen, R. (1992), Theory of thermoluminescence and related
phenomena. World Scientific.
McKeever,
S.W.S.,
Moscovitch,
M.
and
Townsend,
P.D.
(1995).
Thermoluminescence Dosimetry Materials: Properties and Uses, Nuclear
Technology Publishing, Kent.
McKinlay A.F. (1981), Medical Physics Handbook 5: Thermoluminescence
Dosimetry, Bristol. United Kingdom. Adam Hilger Publication.
Miles C.J. (1994), Photon Energy Response Calculations Filtered CaSo4:Dy
Dosimeters using The EGS4 Monte Carlo Code.
San Jose State University, Master Thesis.
Mobit P N, Nahum A E and Sandison G., (2006), Comparison of the EnergyResponse Factor of LiF and Al2O3 in Radiotherapy Beam.
Rad. Prot. Dos. Vol. 119, No. 1–4, pp. 497–499
Mobit P. N., Nahum, A. E. and Mayles P., (1998). A Monte Carlo study of the
quality dependence factors of common TLD materials in photon and electron
beams.
Phys. Med. Biol. Vol 43, pp. 2015–2032
Mobit P. N., Mayles P. and Nahum A. E., (1996). Quality dependence of LiF-TLD in
megavoltage photon beams: Monte Carlo simulation and experiments.
Phys. Med. Biol. Vol. 41, pp 387–389.
Morgan. T.J., Brateman. L., (1977), The Energy and Directional Response of
Harshaw TLD 100 Thermoluminescence Dosimeters in the Diagnostic X-Ray
Energy Range. Health Physics, Vol 33. pp 339 – 342.
Oberhofer, M.; Scharmann. (1981) Applied Thermoluminescence Dosimetry. Bristol,
United Kingdom. Adam Hilger Publication
60
Olivera. G.H., Kessler. C., Sansogne. R.A., Saravi. M., (1994). Energy Dependance
of the Response of Thermoluminescent Dosimeters to photon and Electron
Beams. Nuclear Instruments and Methods in Physics Research B, Vol 84, pp
89-94.
Olsher R. H., (1993), Photon Enegy Response of an Aluminium Oxide TLD
Environmental Dosimeter. IEEE
Regulla D.F. and Driscoll C.M.H., (1993). Thermoluminescence materials and their
dosimetric characteristic. In: Oberhofer, M. and Scharmann (eds). (1993).
Techniques and Management of Personnel Thermoluminescence Dosimetry
Services, pp 63 – 123, ECSC, EEC, EaEC, Brussels.
Saint Gobain Crystals and Detector (2000), LiF: Mg, Cu, P Physical Data and
Constants, Communication, Tech Worksheet.
Shivaramu and Ramprasath,V (2000). Effective atomic numbers for photon energy
absorption and energy dependence of some thermoluminescent dosimetric
compounds. Nucl. Inst. And Meth. Phys. Research. B. Vol. 168, pp 294-304.
Verhaegen, F. (2002). Evaluation of the EGSnrc Monte Carlo Code for Interface
Dosimetry near High-Z media exposed to kilovolt and 60-Co photons.
Phys Med Biol. 47 (10). Pp 1691 – 1705.
White M.C., (2002), Photoatomic Data Library MCPLIB04: A New Photoatomic
Library Based on Data from ENDF/B-VI Release 8, Los Alamos National
Laboratory internal memorandum X-5: MCW-02-111
(Available URL:
http://www.xdiv.lanl.gov/PROJECTS/DATA/nuclear/pdf/mcw-02-111.pdf).
X-5 Monte Carlo Team, (2005). MCNP — A General Monte Carlo N-Particle
Transport Code, Version 5. Vol 1 and 2. Los Alamos.
61
Yaakob, N. (2011). Germanium and Aluminium doped Silicon Dioxide Optical Fibre
Dosimeters for Radiotherapeutic Dose Measurement.
Universiti Teknologi Malaysia: Master Thesis.
62
APPENDIX A
Example of input file for TLD 100
TLD 100 RESPONSE TO PHOTON
c Beginning of cell cards
c
10 1 -2.64
-1
u=1
20 2 -0.00192 1:2
u=12
21 2 -0.00192 -3
fill=1 u=2 lat=1
30 2 -0.00192 -2
fill=2
40 3 -1.19
#30 -4
50 2 -0.00192 4 -5
60 0
5
c
c end of cell cards
c
c
1
2
3
4
5
c
c
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=0
$TLD-100
$Septa
$lattice
$Tray
$Phantom
$Sphere
$Universe
Beginning of surface cards
rpp
rpp
rpp
rpp
so
-0.165 0.165 -0.165 0.165 -0.045 0.045
-4.5 5.5 -2.5 2.5 -0.25 0.25
-0.491 0.499 -0.491 0.499 -0.249 0.249
-15 15 -15 15 -21.5 1.5
120
$TLD-100
$Tray dimension
$Lattice window
$Phantom
$sphere
End of surfaces card
c Beginning of data cards
mode p
sdef pos=0 0 101.5 erg=20 PAR=2 vec=0 0 -1 dir=d1
SI1 -1 0.9999 1
$Histogram for cosine bin limit
SP1 0 0 1
$Fraction solid angle for each
bin
SB1 0. 0. 1.
$Source bias for each bin
m1 3000 -0.2672
$TLD-100
9000 -0.7328
12000 -0.0002
22000 -0.00001
m2 6012 -0.000124
$Dry air
7014 -0.755268
8016 -0.231781
18000 -0.012827
m3 1001 -0.0805259 1002 -0.0000121
$Phantom
8016 -0.3194907 8017 -0.0001217
6012 -0.593190 6013 -0.00665831
f6:p 10
nps 1e7
63
APPENDIX B
Example of input file for fibre optic
SiO2 FIBER OPTIC RESPONSE TO PHOTON
c Beginning of cell cards
c
10 1 -2.32
-1 u=1
20 2 -0.00192 1:2
u=12
21 2 -0.00192 -3
fill=1 u=2 lat=1
30 2 -0.00192 -2
fill=2
40 3 -1.19
#30 -4
50 2 -0.00192 4 -5
60 0
5
c
c end of cell cards
c
c
1
2
3
4
5
c
c
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=1
imp:p=0
$SiO2 Fiber optic
$Septa
$lattice
$Tray
$Phantom
$Sphere
$Universe
Beginning of surface cards
rcc
rpp
rpp
rpp
so
-0.25 0 0 0.5 0 0 0.000130
$SiO2 Fiber
-4.5 5.5 -2.5 2.5 -0.25 0.25
-0.491 0.499 -0.491 0.499 -0.249 0.249
-15 15 -15 15 -21.5 1.5
120
optic
$Tray dimension
$Lattice window
$Phantom
$sphere
End of surfaces card
c Beginning of data cards
mode p
sdef pos=0 0 101.5 erg=1.25 PAR=2 vec=0 0 -1 dir=d1
SI1 -1 0.9999 1
$Histogram for cosine bin limits
SP1 0 0 1
$Fraction solid angle for each bin
SB1 0. 0. 1.
$Source bias for each bin
m1 8000 -0.4612
$SiO2 Fiber optic
14000 -0.5364
32000 -0.0024
m2 6012 -0.000124
$Dry air
7014 -0.755268
8016 -0.231781
18000 -0.012827
m3 1001 -0.0805259 1002 -0.0000121
$Phantom
8016 -0.3194907 8017 -0.0001217
6012 -0.593190 6013 -0.00665831
f6:p 10
nps 1e7
64
APPENDIX C
Example of output file generated by MCNP5.
Thread Name & Version = MCNP5_RSICC, 1.40
_
._ _
_ ._
._
|_
| | | (_ | | |_)
_)
|
+--------------------------------------------------------------------+
|
This program was prepared by the Regents of the University of
|
|California at Los Alamos National Laboratory (the University) under |
| contract number W-7405-ENG-36 with the U.S. Department of Energy |
|(DoE). The University has certain rights in the program pursuant to|
| the contract and the program should not be copied or distributed |
| outside your organization. All rights in the program are reserved |
|by the DoE and the University. Neither the U.S. Government nor the |
| University makes any warranty, express or implied, or assumes any |
|
liability or responsibility for the use of this software.
|
+--------------------------------------------------------------------+
1mcnp
version 5
ld=11012005
06/05/10 09:37:01
*************************************************************************
06/05/10 09:37:01
i=tldnew o=1250kev
warning. universe map (print table 128) disabled.
1TLD 100 RESPONSE TO PHOTON
2c Beginning of cell cards
3c
410 1 -2.64
-1
u=1
imp:p=1 $TLD-100
520 2 -0.00192 1:2
u=1
imp:p=1 $Septa
621 2 -0.00192 -3
fill=1 u=2 lat=1 imp:p=1 $lattice
730 2 -0.00192 -2
fill=2
imp:p=1 $Tray
840 3 -1.19
#30 -4
imp:p=1 $Phantom
950 2 -0.00192 4 -5
imp:p=1 $Sphere
1060 0
5
imp:p=0 $Universe
11c
12c end of cell cards
1314c Beginning of surface cards
15c
161 rpp -0.165 0.165 -0.165 0.165 -0.045 0.045
$TLD-100
172 rpp -4.5 5.5 -2.5 2.5 -0.25 0.25
$Tray dimension
183 rpp -0.491 0.499 -0.491 0.499 -0.249 0.249
$Lattice window
194 rpp -15 15 -15 15 -21.5 1.5
$Phantom
205 so
120
$sphere
21c
22c End of surfaces card
2324c Beginning of data cards
25mode p
26sdef pos=0 0 101.5 erg=1.25 PAR=2 vec=0 0 -1 dir=d1
27SI1 -1 0.9999 1
$Histogram for cosine bin limit
28SP1 0 0 1
$Fraction solid angle for each bin
29SB1 0. 0. 1.
$Source bias for each bin
30m1 3000 -0.2672
$TLD-100
319000 -0.7328
3212000 -0.0002
3322000 -0.00001
34m2 6012 -0.000124
$Dry air
357014 -0.755268
368016 -0.231781
3718000 -0.012827
38m3 1001 -0.0805259 1002 -0.0000121
$Phantom
398016 -0.3194907 8017 -0.0001217
406012 -0.593190 6013 -0.00665831
41f6:p 10
42nps 1e7
43comment. lattice speed tally modifications will not be used.
comment. surface
surface
20 <
comment. surface
surface
20 <
21
20 and
30 in chain
2.2 appears more than once in a chain.
2.2 is in cells
<
30
comment. surface
surface
2.1 appears more than once in a chain.
2.1 is in cells
21 <
30
20 and
30 in chain
2.3 appears more than once in a chain.
2.3 is in cells
20 and
30 in chain
probid =
65
20
<
21
<
30
comment. surface
surface
20 <
21
2.4 appears more than once in a chain.
2.4 is in cells
<
30
comment. surface
surface
20 <
21
30 in chain
2.5 appears more than once in a chain.
2.5 is in cells
<
30
comment. surface
surface
20 <
20 and
20 and
30 in chain
2.6 appears more than once in a chain.
2.6 is in cells
21 <
30
20 and
30 in chain
warning.
1 materials had unnormalized fractions. print table 40.
1cells
print table 60
cell
mat
10
20
21
30
40
50
60
1
2
2
2
3
2
0
1
2
3
4
5
6
7
atom
density
gram
density
1.22512E-01
7.95005E-05
7.95005E-05
7.95005E-05
1.07374E-01
7.95005E-05
0.00000E+00
2.64000E+00
1.92000E-03
1.92000E-03
1.92000E-03
1.19000E+00
1.92000E-03
0.00000E+00
total
volume
9.80100E-03
0.00000E+00
4.88090E-01
2.50000E+01
2.06750E+04
0.00000E+00
0.00000E+00
mass
photon
pieces importance
2.58746E-02
0.00000E+00
9.37132E-04
4.80000E-02
2.46033E+04
0.00000E+00
0.00000E+00
0
0
0
0
0
0
0
1.0000E+00
1.0000E+00
1.0000E+00
1.0000E+00
1.0000E+00
1.0000E+00
0.0000E+00
2.07005E+04 2.46033E+04
minimum source weight = 1.0000E+00
maximum source weight = 1.0000E+00
***************************************************
* Random Number Generator =
1 *
* Random Number Seed
=
19073486328125 *
* Random Number Multiplier =
19073486328125 *
* Random Number Adder
=
0 *
* Random Number Bits Used =
48 *
* Random Number Stride
=
152917 *
***************************************************
2 warning messages so far.
1cross-section tables
print table 100
table
length
tables from file mcplib04
1000.04p
02/07/03
3000.04p
02/07/03
6000.04p
02/07/03
7000.04p
02/07/03
8000.04p
02/07/03
9000.04p
02/07/03
12000.04p
02/07/03
18000.04p
02/07/03
22000.04p
02/07/03
total
1898
ENDF/B-VI Release 8 Photoatomic Data for 1-H
mat 100
2339
ENDF/B-VI Release 8 Photoatomic Data for 3-LI
mat 300
3152
ENDF/B-VI Release 8 Photoatomic Data for 6-C
mat 600
3194
ENDF/B-VI Release 8 Photoatomic Data for 7-N
mat 700
3272
ENDF/B-VI Release 8 Photoatomic Data for 8-O
mat 800
3206
ENDF/B-VI Release 8 Photoatomic Data for 9-F
mat 900
3781
ENDF/B-VI Release 8 Photoatomic Data for 12-MG
mat1200
4696
ENDF/B-VI Release 8 Photoatomic Data for 18-AR
mat1800
5742
ENDF/B-VI Release 8 Photoatomic Data for 22-TI
mat2200
31280
maximum photon energy set to
100.0 mev (maximum electron energy)
tables from file el03
1000.03e
6/6/98
3000.03e
6/6/98
6000.03e
6/6/98
7000.03e
6/6/98
8000.03e
6/6/98
9000.03e
6/6/98
12000.03e
6/6/98
18000.03e
6/6/98
2329
2331
2333
2333
2333
2333
2337
2341
66
22000.03e
6/6/98
2345
*********************************************************************************************************
**************
dump no.
1 on file runtpt
nps =
0
coll =
0
ctm =
0.00
nrn =
0
2 warning messages so far.
1problem summary
run terminated when
10000000 particle histories were done.
+
06/05/10 09:45:22
TLD 100 RESPONSE TO PHOTON
06/05/10 09:37:01
0
photon creation
tracks
weight
energy
photon loss
energy
(per source particle)
source particle)
source
5.2579E-01
10000000
1.0000E+00
1.2500E+00
probid =
tracks
weight
(per
escape
9404223
9.4042E-01
energy cutoff
0
0.
time cutoff
0
0.
1.3847E-05
0.
weight window
0
0.
cell importance
0
0.
weight cutoff
0
0.
e or t importance
0
0.
dxtran
0
0.
forced collisions
0
0.
exp. transform
0
0.
from neutrons
0
7.2176E-01
bremsstrahlung
529645
3.9683E-03
p-annihilation
4228
2.6401E-04
photonuclear
0
0.
electron x-rays
0
1st fluorescence
195
2nd fluorescence
0
total
10534068
1.2518E+00
0.
0.
weight window
0
0.
0.
0.
cell importance
0
0.
0.
0.
weight cutoff
0
0.
0.
0.
e or t importance
0
0.
0.
0.
dxtran
0
0.
0.
0.
forced collisions
0
0.
0.
0.
exp. transform
0
0.
0.
0.
compton scatter
0
0.
5.2964E-02
1.5786E-03
capture
4.2280E-04
2.1605E-04
pair production
0.
0.
photonuclear abs
0.
1.9500E-05
0.
1.0534E+00
0.
5.7474E-08
0.
1.2518E+00
number of photons banked
photon tracks per source particle
1.0000E+33
photon collisions per source particle
1.0000E-03
total photon collisions
-5.0000E-01
531759
1.0534E+00
total
1127731
1.1277E-01
2114
2.1140E-04
0
10534068
average time of (shakes)
escape
7.7585E-01
3.7270E+00
capture
37269656
0.
1.0534E+00
cutoffs
tco
4.5761E-01
eco
capture or escape 7.4177E-01
wc1
any termination
wc2
7.4162E-01
-2.5000E-01
computer time so far in this run
computer time in mcrun
source particles per minute
random numbers generated
history
6923210
8.34 minutes
8.32 minutes
1.2015E+06
599246882
maximum number ever in bank
bank overflows to backup file
most random numbers used was
3
0
646 in
range of sampled source weights = 1.0000E+00 to 1.0000E+00
1photon
activity in each cell
print table 126
tracks
population
collisions
collisions
number
flux
average
average
cell
track mfp
entering
* weight
weighted
weighted
track weight
(per history)
energy
energy
(relative)
1.1132E+00
1.1132E+00
1.0000E+00
(cm)
1
10
6.6986E+00
2
20
8.4795E+03
3
21
0.0000E+00
4
30
0.0000E+00
5
40
9.7442E+00
1327769
1314417
20288
2.0288E-03
33079323
9642988
751
7.5100E-05
1.1014E+00
1.1014E+00
1.0000E+00
0
0
0
0.0000E+00
0.0000E+00
0.0000E+00
0.0000E+00
0
0
0
0.0000E+00
0.0000E+00
0.0000E+00
0.0000E+00
20222896
10438706
36929030
3.6929E+00
6.9532E-01
6.9532E-01
1.0000E+00
67
6
50
7.4740E+03
total
1tally
6
19333708
10077261
319587
3.1959E-02
73963696
31473372
37269656
3.7270E+00
8.9559E-01
nps =
10000000
tally type 6
track length estimate of heating.
tally for photons
units
8.9559E-01
1.0000E+00
mev/gram
masses
cell:
cell
10
2.58746E-02
10
3.45887E-02 0.0008
=========================================================================================================
====
results of 10 statistical checks for the estimated answer for the tally fluctuation chart (tfc) bin of tally
tfc bin
behavior
--mean-behavior
desired
observed
passed?
random
random
yes
6
---------relative error--------value
decrease
decrease rate
----variance of the variance---value
decrease
decrease rate
--figure of merit-value
behavior
-pdfslope
<0.10
0.00
yes
<0.10
0.00
yes
constant
constant
yes
>3.00
3.45
yes
yes
yes
yes
1/sqrt(nps)
yes
yes
yes
yes
yes
1/nps
yes
yes
random
random
yes
=========================================================================================================
==========================
this tally meets the statistical criteria used to form confidence intervals: check the tally fluctuation
chart to verify.
the results in other bins associated with this tally may not meet these statistical criteria.
----- estimated confidence intervals:
-----
estimated asymmetric confidence interval(1,2,3 sigma): 3.4560E-02 to 3.4618E-02; 3.4530E-02 to 3.4647E02; 3.4501E-02 to 3.4676E-02
estimated symmetric confidence interval(1,2,3 sigma): 3.4560E-02 to 3.4618E-02; 3.4530E-02 to 3.4647E02; 3.4501E-02 to 3.4676E-02
1analysis of the results in the tally fluctuation chart bin (tfc) for tally
print table 160
normed average tally per history
04
estimated tally relative error
relative error from zero tallies
6 with nps =
10000000
= 3.45887E-02
unnormed average tally per history
= 8.94969E-
= 0.0008
= 0.0008
estimated variance of the variance
relative error from nonzero scores
= 0.0000
= 0.0002
number of nonzero history tallies =
1310485
history number of largest tally =
8818806
02
(largest tally)/(average tally) = 5.68515E+01
7.45030E+00
efficiency for the nonzero tallies = 0.1310
largest unnormalized history tally = 5.08803E-
(confidence interval shift)/mean
02
shifted confidence interval center
= 0.0000
(largest
tally)/(avg nonzero tally)=
= 3.45887E-
if the largest history score sampled so far were to occur on the next history, the tfc bin quantities
would change as follows:
estimated quantities
value(nps+1)/value(nps)-1.
mean
relative error
variance of the variance
shifted center
figure of merit
value at nps
3.45887E-02
8.42619E-04
5.58207E-07
3.45887E-02
1.69224E+05
value at nps+1
3.45889E-02
8.42633E-04
5.60077E-07
3.45887E-02
1.69218E+05
0.000006
0.000016
0.003351
0.000000
-0.000033
the estimated inverse power slope of the 198 largest tallies starting at 1.47216E-02 is 3.4508
the history score probability density function appears to have an unsampled region at the largest
history scores:
please examine. see print table 161.
fom = (histories/minute)*(f(x) signal-to-noise ratio)**2 = (1.202E+06)*( 3.753E-01)**2 =
(1.202E+06)*(1.408E-01) = 1.692E+05
1status of the statistical checks used to form confidence intervals for the mean for each tally bin
tally
result of statistical checks for the tfc bin (the first check not passed is listed) and error
magnitude check for all bins
6
passed the 10 statistical checks for the tally fluctuation chart bin result
passed all bin error check:
1 tally bins all have relative errors less than 0.10 with no
zero bins
68
the 10 statistical checks are only for the tally fluctuation chart bin and do not apply to other tally
bins.
1tally fluctuation charts
nps
512000
1024000
1536000
2048000
2560000
3072000
3584000
4096000
4608000
5120000
5632000
6144000
6656000
7168000
7680000
8192000
8704000
9216000
9728000
10000000
tally
6
mean
error
3.4573E-02 0.0037
3.4668E-02 0.0026
3.4654E-02 0.0021
3.4595E-02 0.0019
3.4605E-02 0.0017
3.4615E-02 0.0015
3.4655E-02 0.0014
3.4638E-02 0.0013
3.4625E-02 0.0012
3.4611E-02 0.0012
3.4604E-02 0.0011
3.4622E-02 0.0011
3.4608E-02 0.0010
3.4594E-02 0.0010
3.4589E-02 0.0010
3.4586E-02 0.0009
3.4593E-02 0.0009
3.4586E-02 0.0009
3.4590E-02 0.0009
3.4589E-02 0.0008
vov slope
fom
0.0000 4.5 168063
0.0000 4.3 168380
0.0000 4.4 168102
0.0000 4.9 167223
0.0000 5.2 167516
0.0000 4.1 167966
0.0000 3.7 168301
0.0000 4.1 168570
0.0000 4.0 168624
0.0000 3.7 168677
0.0000 4.5 168810
0.0000 4.7 168984
0.0000 4.8 169019
0.0000 4.7 169013
0.0000 5.5 169051
0.0000 4.4 169099
0.0000 4.8 169162
0.0000 4.0 169155
0.0000 3.6 169198
0.0000 3.5 169224
*********************************************************************************************************
**************
dump no.
2 on file runtpt
nps =
10000000
coll =
37269656
ctm =
8.32
nrn =
599246882
2 warning messages so far.
run terminated when
computer time =
mcnp
version 5
06/05/10 09:37:01
10000000
particle histories were done.
8.34 minutes
11012005
06/05/10 09:45:22
probid =
69
APPENDIX D
The mass energy absorption coefficients for air
[Hubbell and Seltzer, 1995]
20.0
Mass Energy Absorption
Coefficient
(μen/ρ)air
2
(cm /g)
0.53890
30.0
0.15370
40.0
0.06833
50.0
0.04098
60.0
0.03041
80.0
0.02407
100.0
0.02325
150.0
0.02496
200.0
0.02672
300.0
0.02872
400.0
0.02949
500.0
0.02966
600.0
0.02953
800.0
0.02882
1000.0
0.02789
1250.0
0.02666
2000.0
0.02547
3000.0
0.02345
6000.0
0.02057
10000.0
0.01870
15000.0
0.01740
20000.0
0.01647
Energy, E
(keV)
70
APPENDIX E
The mass energy absorption coefficients of TLD 100 elements
[Hubbell and Seltzer, 1995]
Mass Energy Absorption Coefficient (cm2/g)
Energy
(keV)
Lithium, 3
Fluorine, 9
Magnesium,
12
Titanium,
22
20
1.788E-02
8.608E-01
2.393E+00
1.455E+01
30
1.114E-02
2.401E-01
6.757E-01
4.447E+00
40
1.122E-02
1.016E-01
2.775E-01
1.895E+00
50
1.233E-02
5.652E-02
1.432E-01
9.666E-01
60
1.358E-02
3.851E-02
8.715E-02
5.594E-01
80
1.588E-02
2.656E-02
4.631E-02
2.410E-01
100
1.755E-02
2.385E-02
3.392E-02
1.308E-01
150
2.098E-02
2.415E-02
2.762E-02
5.397E-02
200
2.289E-02
2.554E-02
2.760E-02
3.736E-02
300
2.481E-02
2.730E-02
2.872E-02
3.014E-02
400
2.552E-02
2.800E-02
2.928E-02
2.871E-02
500
2.569E-02
2.815E-02
2.938E-02
2.809E-02
600
2.558E-02
2.801E-02
2.921E-02
2.761E-02
800
2.498E-02
2.733E-02
2.847E-02
2.663E-02
1000
2.418E-02
2.643E-02
2.751E-02
2.561E-02
1250
2.313E-02
2.528E-02
2.630E-02
2.445E-02
1500
2.208E-02
2.413E-02
2.509E-02
2.327E-02
2000
2.026E-02
2.223E-02
2.317E-02
2.162E-02
3000
1.751E-02
1.960E-02
2.062E-02
1.983E-02
4000
1.560E-02
1.794E-02
1.910E-02
1.907E-02
5000
1.421E-02
1.682E-02
1.814E-02
1.879E-02
6000
1.315E-02
1.604E-02
1.752E-02
1.874E-02
8000
1.166E-02
1.505E-02
1.678E-02
1.894E-02
10000
1.066E-02
1.447E-02
1.643E-02
1.928E-02
15000
9.182E-03
1.377E-02
1.610E-02
2.001E-02
20000
8.397E-03
1.366E-02
1.612E-02
2.068E-02
71
APPENDIX F
The mass energy absorption coefficients of silicon dioxide fibre optic elements
[Hubbell and Seltzer, 1995]
Mass Energy Absorption Coefficient (cm2/g)
Energy
(keV)
20
Oxygen,
8
0.61790
Silicon,
14
4.07600
Germanium,
32
31.78000
30
0.17290
1.16400
11.26000
40
0.07530
0.47820
5.15200
50
0.04414
0.24300
2.75900
60
0.03207
0.14340
1.64200
80
0.02468
0.06896
0.71840
100
0.02355
0.04513
0.38030
150
0.02506
0.03086
0.12880
200
0.02679
0.02905
0.06895
300
0.02877
0.02932
0.03891
400
0.02953
0.02968
0.03193
500
0.02971
0.02971
0.02930
600
0.02957
0.02951
0.02790
800
0.02887
0.02875
0.02618
1000
0.02794
0.02778
0.02489
1250
0.02669
0.02652
0.02353
1500
0.02551
0.02535
0.02242
2000
0.02350
0.02345
0.02094
3000
0.02066
0.02101
0.01977
4000
0.01882
0.01963
0.01962
5000
0.01757
0.01878
0.01987
6000
0.01668
0.01827
0.02027
8000
0.01553
0.01773
0.02120
10000
0.01483
0.01753
0.02208
15000
0.01396
0.01746
0.02364
20000
0.01360
0.01757
0.02452
72
APPENDIX G
PUBLICATIONS AND CONFERENCES
1.
Asni H., Wagiran H, Saripan M.I., Ramli A.T, and Yaakob N.H.
“Thermoluminescence Energy Response of Germanium Doped Optical Fibre
Using Monte Carlo N-Particle Code Simulation.” National Physics
Conference, PERFIK 2009. Institute of Physics Malaysia and Universiti
Teknologi MARA, Avilion Hotel, Melaka, Malaysia, 7 – 9 December 2009.
(Published: AIP Conference Proceedings Volume 1250, pp 420-423
ISBN: 978-0-7354-0797-8)
2.
Yaakob N.H, Wagiran H, Ramli A.T, Ali H. and Asni H. “Photon
Irradiation Response on Ge- and Al- Doped SiO2 Optical Fibre.” National
Physics Conference, PERFIK 2009. Institute of Physics Malaysia and
Universiti Teknologi MARA, Avilion Hotel, Melaka, Malaysia, 7 – 9
December 2009.
(Published: AIP Conference Proceedings Volume 1250, pp 63-66.
ISBN: 978-0-7354-0797-8)
3.
Asni H., Wagiran H, Saripan M.I., S. Hashim, Ramli A.T, and Yaakob N.H.
“Thermoluminescence Energy Response of TLD-100 Subjected to Photon
Irradiation Using Monte Carlo N-Particle Transport Code version 5”.
Second International Conference and Workshop on Basic and Applied
Sciences.
Regional
Annual
Fundamental
Sciences
Seminar
2009
(ICOWOBASS), University Teknologi Malaysia. The Zone Hotel, Johor
Bahru, Malaysia. 2 – 4 June 2009.
4.
Yaakob N.H, Wagiran H, Ramli A.T, S.Hahim, Ali H. and Asni H.,
“Electron Irradiation Response On Ge And Al Doped SiO2 Optical Fibres”.
“Second International Conference And Workshop On Basic And Applied
Sciences.
Regional
Annual
Fundamental
Sciences
Seminar
2009
(ICOWOBASS), University Teknologi Malaysia. The Zone Hotel, Johor
Bahru, Malaysia. 2 – 4 June 2009.