MEASUREMENTS AND MONTE CARLO CALCULATIONS WITH … · Institut de radiophysique appliquée SECTION...

CER

N-T

HES

IS-2

004-

082

09/1

2/20

04

THÈSE NO 3132 (2004)

ÉCOLE POLYTECHNIQUE FÉDÉRALE DE LAUSANNE

PRÉSENTÉE À LA FACULTÉ SCIENCE DE BASE

Institut de radiophysique appliquée

SECTION DE PHYSIQUE

POUR L'OBTENTION DU GRADE DE DOCTEUR ÈS SCIENCES

PAR

diplôme en physique, Université d'Athènes, Grèceet de nationalité hellénique

acceptée sur proposition du jury:

Dr J.-F. Valley, directeur de thèseDr A. Aroua, rapporteurProf. A. Bay, rapporteurDr M. Silari, rapporteur

Lausanne, EPFL2004

MEASUREMENTS AND MONTE CARLO CALCULATIONS WITHTHE EXTENDED-RANGE BONNER SPHERE SPECTROMETER

AT HIGH-ENERGY MIXED FIELDS

Evangelia DIMOVASILI

To Professor Pavlos D. Ioannou

i

CONTENTS CHAPTER 1- Introduction to Bonner Sphere Spectrometry ..... 1

1.1 Multi sphere spectrometry ......................................................................1

1.2 Synoptic historical review and developments of multi- sphere

spectrometry..........................................................................................3

1.3 The CERN Bonner Sphere Spectrometer...............................................7

1.4 Electronics............................................................................................13

1.5 Unfolding methods ...............................................................................13

1.5.1 Introduction ....................................................................................13

1.6 Unfolding codes....................................................................................16

1.6.1 MAXED ..........................................................................................16

1.6.2 GRAVEL ........................................................................................17

1.7 MONTE CARLO method ......................................................................18

1.7.1 Introduction ....................................................................................18

1.7.2 Major components of a Monte Carlo algorithm ..............................19

1.8 The FLUKA Monte Carlo code .............................................................20

CHAPTER 2- Calibration facilities .................................................... 22

2.1 Introduction...........................................................................................22

2.2 The PTB accelerator facility.................................................................23

2.2.1 Description .....................................................................................23

2.3 Calibration of the BSS with Quasi- Monoenergetic Neutron Fields

at UCL ..................................................................................................26

2.3.1 Introduction ...................................................................................26

2.3.2 The neutron facility and the experimental set- up .........................27

2.4 The CERN-EU Reference Field (CERF)...............................................29

2.4.1 Description of the facility ................................................................29

2.4.2 Beam monitoring at CERF .............................................................32

CHAPTER 3- Calibration at reference neutron fields ................ 33

3.1 Calibration of the BSS in monoenergetic neutron fields .......................33

ii

3.1.1 Introduction ....................................................................................33

3.1.2 Calibration at PTB- Materials and methods ..................................33

3.1.3 Data analysis ...............................................................................39

3.1.4 Results .........................................................................................43

3.1.5 Comparison of experimental to MC calculated responses .............44

3.2 Calibration at Quasi- Monoenergetic Neutron Fields...........................50

3.2.1 Materials and methods...................................................................50

3.2.2 Experiment and data analysis ........................................................54

3.2.3 Results and discussion ..................................................................56

3.3 Combined results of the 2001 and 2002 calibration measurements....66

CHAPTER 4- The response of the extended-range Bonner Sphere Spectrometer to charged hadrons.................................... 74

4.1 Introduction...........................................................................................74

4.2 BSS measurements at high-energy mixed fields..................................74

4.2.a Calculation of spectral fluences of hadrons (first MC study) ..........77

4.2.b Response functions of the BSS to charged hadrons

(second MC study)..................................................................................80

4.3 Comparison of the MC studies and estimation of correction factor ....83

4.4 Experimental test with 120 GeV/c hadrons at CERF...........................85

4.5 Conclusions.........................................................................................86

CHAPTER 5- Neutron spectral measurements with a Bonner sphere spectrometer............................................................................. 87

5.1 Introduction...........................................................................................87

5.2 Experiment ...........................................................................................88

5.3 Monte Carlo simulations and spectrum unfolding.................................92

5.4 Data analysis........................................................................................93

5.4.1 Beam normalization factor .............................................................93

5.4.2 BSS response to charged hadrons ................................................94

5.5 Results and discussion.........................................................................96

5. 6. Conclusions......................................................................................104

iii

CHAPTER 6 - Upgrade and development of the CERN-EU Reference Field (CERF) .....................................................................105 PART A: The beam monitoring at CERF .....................................................105

6.1 Monitoring system ..............................................................................105

6.1.1 Introduction ..................................................................................105

6.1.2 Measurements of beam profile with a Multi-Wire Proportional

Chamber ...............................................................................................107

6.2 Performance tests of the BIG PIC .....................................................108

6.2.1 Stability tests...............................................................................109

6.2.2 Linearity test ................................................................................111

6.2.3 Region of ion saturation ...............................................................112

6.2.4 Investigation of leakage current existence ...................................113

6.3 Inter-comparison of PIC and BIG PIC monitors at CERF ...................114

6.3.1 Recombination effects for the BIG PIC .......................................116

6.4 Tests of the Triggers 4, 5, 6 in the H6 beam line................................117

6.4.1 Efficiency measurements of Trigger4...........................................118

6.4.2 Measurements of Trigger5 and Trigger6......................................121

6.5 Conclusions........................................................................................123

PART B: Monte Carlo studies ......................................................................124 6.6 Introduction.........................................................................................124

6.7 Space studies for the CERF facility ....................................................127

6.8 Conclusions........................................................................................133

CHAPTER 7- Conclusions.................................................................134 APPENDICES.........................................................................................138

iv

RESUME

L’utilisation de la spectrométrie comme moyen de renseignement dans le cadre de la protection contre le rayonnement neutronique est devenue une activité de plus en plus importante durant ces dernières années. Le besoin d’avoir des données spectrales est apparu, car ni les instruments d’inspection des zones ni les dosimètres personnels ne donnent l’équivalent de dose de maniére correcte pour toutes les énergies de neutrons. Il est donc important de connaître le spectre des champs dans lesquels ces appareils sont utilisés. L’un des appareils les plus souvent utilisés dans la spectrométrie des neutrons et dans la dosimétrie est le spectromètre de sphère de Bonner (‘Bonner Sphere Spectrometer’ or BSS). La gamme du spectromètre utilisé au cours de ce travail consiste en 7 sphères et offre une réponse aux neutrons allant jusqu’à 2 GeV. Un détecteur à 3He est utilisé comme compteur thermal au centre de chaque sphère. Dans le contexte de cette thèse le ‘BSS’ a été calibré dans un champ monoénergétique de neutrons à des énergies basses et intermédiaires. Il a été également employé pour des mesures dans différents champs mixtes à haute énergie. Ces mesures ont amené aux calculs de champs de neutrons et de fluences spectrales pour des cibles non protégées. De telles données sont très utiles pour les calculs des blindages. De plus, la réponse des ‘BSS’ aux hadrons chargés a été améliorée. Parmi les champs dans lesquels le ‘BSS’ a été testé, le complexe CERF s’est révélé d’un intérêt particulier. Il est installé dans l’un des faisceaux secondaires issus du ‘SPS’ (Super Proton Synchroton) situé sur la partie françaize du CERN. La composition en particules et les fluences spectrales au-delà du blindage de CERF simulent celles rencontrées dans les champs cosmiques situés aux altitudes des vols commerciaux. Durant la dernière décennie les instituts de recherche travaillant dans le domaine de la recherche spatiale ont montré un intérêt grandissant pour l’installation CERF. Il y a donc un besoin de mettre en place une nouvelle zone d’exposition aux radiations reproduisant le champ de radiation situé au-delà de l’atmosphère ou à l’intérieur d’un vaisseau spatial. Les simulations ‘Monte-Carlo’ menées dans cet objectif concordent avec les résultats expérimentaux pris à l’intérieur de la station spatiale MIR. Ainsi, l’installation CERF pourrait s’avérer un outil très utile pour de futurs tests avec les ‘BSS’, en vue de simulation des champs de rayonnement cosmique.

v

ABSTRACT

The use of spectrometry to provide information for neutron radiation

protection has become an increasingly important activity over recent years. The need

for spectral data arises because neither area survey instruments nor personal

dosimeters give the correct dose equivalent results at all neutron energies. It is

important therefore to know the spectra of the fields in which these devices are used.

One of the systems most commonly employed in neutron spectrometry and

dosimetry is the Bonner Sphere Spectrometers (BSS). The extended- range BSS that

was used for this work, consists of 7 spheres with an overall response to neutrons up

to 2 GeV. A 3He detector is used as a thermal counter in the centre of each sphere. In

the context of this thesis the BSS was calibrated in monoenergetic neutron fields at

low and intermediate energies. It was also used for measurements in several high

energy mixed fields. These measurements have led to the calculation of neutron yields

and spectral fluences from unshielded targets. Such data are very useful for shielding

calculations and other radiation protection measurements. Furthermore, an upgrade of

the response of the BSS to charged hadrons was achieved.

Among the fields were the BSS has been tested, the CERF facility is of

particular interest. It is installed in one of the secondary beam lines from the Super

Proton Synchrotron (SPS) in the French site of CERN. The particle composition and

spectral fluences outside the shielding of CERF simulate those in the cosmic radiation

field at commercial flight altitudes. In the last decade interest has arisen in CERF

from research institutions working in the space programme. There is therefore a need

for setting-up a new exposure area reproducing the radiation environment found

outside the atmosphere or inside a space vessel. The results from Monte Carlo studies

that have been carried out for this purpose show a good agreement with experimental

data taken inside the MIR space station. Therefore, CERF may prove to be a very

useful tool for further tests with the BSS system, towards its upgrade for

measurements in cosmic ray fields.

vi

INTRODUCTION

The aim of this doctoral thesis is to upgrade and develop the Extended- Range

Bonner Sphere Spectrometer (BSS) in high-energy mixed radiation fields. The correct

use of the BSS must be certified by calibration measurements. Three calibration

campaigns have taken place between 2001 and 2003, in order to validate the Monte

Carlo calculated response matrix of the extended BSS. The overall response of the

spectrometer has been extended to charged hadrons and a full response matrix has

been calculated. It has also been possible to calibrate the BSS in a mixed charged

hadron beam at the CERN-EU Reference Field facility.

In Chapter 1 the reader can find an overview of the history of the multi-sphere

spectrometry and a description of the instruments and methods that were used for the

purposes of this work. Chapter 2 describes the calibration facilities that have been

used in the framework of this dissertation. Calibration campaigns have been

performed at PTB (Braunschweig, Germany) and at UCL (Louvain- La Neuve,

Belgium) at monoenergetic neutrons with energies from 0.565 MeV up to 60 MeV.

The aim of the calibration was the verification of the theoretical response functions

calculated with the FLUKA Monte Carlo code and the estimation of the calibration

factor of the specific 3He detector. The data analysis and the results of the calibration

are presented in Chapter 3.

A BSS can be used to measure neutron spectra both outside accelerator

shielding and from unshielded targets. Measurements that were performed in the

NA-57/ALICE experimental area at CERN have provided evidence that the BSS

under certain conditions can show a significant response to charged hadrons. A

complete response matrix of the BSS to charged pions and protons was calculated

with FLUKA. An experimental verification was carried out with a 120 GeV/c hadron

beam at the CERF facility at CERN. The analysis and the results of these studies are

given in detail in Chapter 4.

The aim of the experiments at the NA-57/ALICE area was the calculation

of the neutron yields and spectral fluences from unshielded, semi-thick targets, in

order to provide source term data for neutron production from high-energy hadrons.

vii

This information is necessary for shielding calculations and other radiation protection

purposes. The work is analytically presented in Chapter 5.

The calibration of the BSS at the charged hadron beam of CERF provoked the

need for the further development and improvement of this facility. The studies

performed are presented in Chapter 6 that consists of two parts. In the first part, the

extensive performance tests of the CERF back-up monitor are described. This is an

ionization chamber of similar design to the reference CERF beam monitor that was

submitted to a number of tests in the SC/RP calibration laboratory. It was then tested

in real experimental conditions at the CERF facility. The second part of Chapter 6

discusses the FLUKA Monte Carlo studies that have been performed for the upgrade

of CERF. The new challenging objective was to design a different shielding

configuration inside the CERF cave that could be of interest for measurements in the

framework of the space programme but also for the future use of the BSS at high

energy fields such as cosmic ray fields.

In Chapter 7 the conclusions of the overall work are summarised and an

outlook on further development of the Extended –Range BSS and of the CERF

facility is proposed. Appendices include part of the calculations that have been done

in the context of this thesis.

1

CHAPTER 1

Introduction to Bonner Sphere Spectrometry

1.1 Multi sphere spectrometry

The term ‘radiation spectrometry’ can be used to describe the measurement of

the intensity of a radiation field with respect to energy, wavelength, momentum, mass,

angle of incidence or any other related quantity. The distribution of the intensity with

one of these parameters is commonly referred to as the spectrum. Neutron energy

spectra are frequently measured by indirect methods because of the experimental

limitations in detecting fast neutrons. The need for spectrometry stems mainly from

the fact that both area survey instruments and personal neutron dosimeters have a dose

equivalent* response which is a function of energy. Spectra thus need to be known in

order to determine precisely the dose equivalent values in fields where individuals are

exposed to neutrons, e.g. in workplaces in the nuclear industry, around accelerators, or

at aircraft flight altitudes. Spectra also need to be determined to characterize

calibration fields.

A multi-sphere spectrometer uses a thermal neutron detector at the centre of

moderating spheres of different diameters, usually made of polyethylene. A schematic

drawing of such a sphere is shown in Fig. 1.1. The fast neutrons are slowed down in

the moderator and reach the detector thermalized, while the thermal neutrons initially

present in the field are mostly captured in the moderator. Therefore, when the

moderating sphere's diameter increases, the maximum sensitivity of the system moves

to higher energies. The multi-sphere spectrometer shows several characteristics useful

for radiation protection measurements. Amongst its main advantages one can include

its functional simplicity, its wide energy domain (from thermal energies to several

MeV), the high neutron sensitivity allowing the measurement of low dose equivalent

rates encountered in radiation protection, the good discrimination of electronic noise

* See Appendix A

CHAPTER 1. Introduction to Bonner Sphere Spectrometry

2

and photon counting by a suitable choice of the counter and adjustment of the

associated electronics. The electronics modules needed to operate the detector are

relatively simple, adding to the benefits of using this spectrometer.

Fig. 1.1. Schematic representation of a polyethylene Bonner Sphere.

One drawback of this method is the low energy resolution. This is partially

due to the fact that the statistical fluctuations in the number of collisions in the

neutron slowing down processes are large, and the capture reactions are completely

indistinguishable from one another. This results in loss of information about the

primary neutron energy and, consequently, low resolution. However, the energy

resolution of the system being low can be judged as satisfactory for the evaluation of

the dosimetric quantities used in radiation protection.


3

1.2 Synoptic historical review and developments of multi- sphere spectrometry

Spectrometers for measuring photon and neutron energy spectra have been

used for almost as long as there has been an awareness of the existence of these

radiations. The devices have however improved dramatically over the years. In

parallel, the original detection mechanisms have been augmented by more modern

techniques. During this period of research and advancement of technology,

spectrometry measurements have played a vital role in understanding the nature and

origins of radiation. They have subsequently been an invaluable source of information

in atomic and nuclear physics research.

Bramblett, Ewing and Bonner, working at Rice University, developed and

tested the first multi-sphere detector in 1960, known widely since then as the Bonner

Sphere Spectrometer (BSS) [1]. The first BSS consisted of a small cylindrical (4 mm

high by 4 mm diameter) 6LiI(Eu) scintillator optically coupled to a Photo-Multiplier

placed at the centre of a series of polyethylene neutron moderating spheres (Fig. 1.2).

Fig. 1.2. Schematic drawing of the Bonner [1] experimental set –up.


4

Count rates with the LiI detector placed inside 5.08 cm, 7.62 cm, 12.7 cm,

20.32 cm and 30.48 cm diameter spheres were used to obtain information about

neutron spectra. The size of the crystal was chosen to be small so as to allow good

γ-ray discrimination. The responses of the first BSS were measured for incident

neutrons in the energy range 0.06-15 MeV and are shown in Fig. 1.3.

Fig. 1.3. The energy dependence of the relative detection efficiencies of Bonner sphere neutron detectors of various diameters up to 30.48 cm [1].

Since the first use of the LiI crystal in moderating spheres many other researchers

have shown preference in this type of detector [2-4]. In cases where higher sensitivity

is required larger scintillation crystals can be used. The main disadvantage of the use

of LiI crystals is the subtraction of γ events which requires an analysis of the light-

output spectrum and it is generally a difficult procedure. 10BF3 proportional counters have been used as an alternative to LiI crystals.

BSS sets based on such counters have not been very common, however systems using

small diameter cylindrical counters have been built and used extensively [5- 8].

During the 1970s and early 1980s many papers were published describing the

use of 9 mm diameter cylindrical 3He proportional counters (type 0.5NH10) as


5

thermal sensors (see for example [9, 10]). The use of this type of counter which

contained about 8 atm of 3He provided a system with a much higher sensitivity.

Although it is less sensitive to γ- rays than a LiI one, the low gas amplification may

restrict the discrimination of all neutron induced events, while excluding noise.

Nevertheless, this counter has been used by many groups [11-14] and it provides a

useful lower-response alternative to the more sensitive type SP9 3He counter, for use

in intense neutron fields.

In the 1980s and early 1990s, several research teams [15-19] investigated the

use of the SP9 spherical 3He proportional counter, produced and commercially

available by Centronic Ltd., UK. This counter has a lower gas pressure of about 2 atm

but a bigger diameter, resulting in a larger geometrical cross-section. As a result, the

overall fluence responses of Bonner spheres incorporating this counter are of the order

of a factor of 10 higher than for the 4 mm by 4 mm LiI system, depending on sphere

size and neutron energy. Additionally, the discrimination with respect to γ- rays and

noise is excellent, except in the highest-intensity γ- ray fields where pile-up becomes

a problem. The 3He counters are fairly insensitive to radiations other than neutrons

and their efficiency proved to be stable with time. The characteristics of BSS using

this type of counter are nowadays very well established [13, 20-26].

Although the SP9 counter provides reasonably high sensitivity, there are

situations where a higher efficiency is desirable, e.g. for the investigation of cosmic-

ray-induced neutrons at ground and flight altitudes with very low fluence rates. For

such purposes, BSS systems based on large 3He counters have been developed

[27-28].

On the other hand, there are situations where low efficiency is preferred. BSS

systems which use activation detectors as the thermal sensor have been built [27, 29-

31]. The activation material is usually gold or indium. These systems do not present

problems, such as dead-time losses, that can be of great concern for active devices in

intense fields, particularly in pulsed ones. Activation foils have proved to be more

efficient in other environments, because of their very low sensitivity to γ-rays.

However, γ-ray induced neutron production in the material of the sphere needs to be

considered in intense γ-ray fields, for example in the photon beam from a medical

electron accelerator [29].


6

The development of a lower-efficiency system based on the SP9 counter has

been of great interest. There have been few attempts made towards this direction [24],

by either covering this counter with a close-fitting thin cadmium shell, or by reducing

the gas pressure.

It is commonly known that the resolution of a conventional BS at low and

intermediate energies is particularly poor. For radiation protection the resolution

below about 10 keV, where dose equivalent conversion coefficients are roughly

constant, is not very important. However, there are applications, e.g. the

characterization of epithermal fields for boron neutron capture therapy (BNCT),

where better resolution is needed. Including an outer shell of a thermal neutron

absorber, such as boron or lithium, around the smaller spheres, enhances the response

function up in energy with increasing absorber thickness. The disadvantage of the

method is that the overall sensitivity is reduced. Nevertheless, by using absorber

layers of different thicknesses, spheres with response functions covering the low- and

intermediate-energy range can be produced. Several groups have investigated this

approach [32-34].

BSS have recently started being adapted to measure spectra up to neutron

energies in the GeV range [26, 28, 35, 36]. An interesting approach among new

designs that have evolved from the BSS is to use a single block of moderator

containing either several extended, position-sensitive thermal neutron detectors [37]

or a number of small thermal neutron detectors [38] mounted at different positions.

Passive detectors have also been used in Bonner spheres, in order to measure

very intense pulsed neutron fields such as those encountered around particle

accelerators [39-41], or in the case where a low intensity neutron field requires a very

long integration time such as in some environmental measurements. The types of

passive detectors employed include activation detectors sensitive to thermal neutrons,

pairs of 6Li and 7Li fluoride thermo-luminescent detectors, and track detectors with

radiators made of 10B, 6Li or 235U.

Interest in measurements around high-energy accelerators, and at high

altitudes in the atmosphere, have stimulated the development of spheres with shells of

lead, iron or copper within the polyethylene [26, 28, 42-44]. Neutron multiplication,

which for high-energy neutrons occurs within the metal shell, increases the high-


7

energy response. Variations in the response function shapes can be obtained by

varying the configuration. The alternative of obtaining data at higher energies is to

include in the BS analysis data from ‘threshold’ reactions like neutron induced fission

reactions, e.g. 232Th(n,f) and 209Bi(n,f) (where ‘f’ stands for ‘fission’) [41].

In summary, the Bonner Sphere spectroscopy, one of the most widely

employed methods, is basically the same today as it was when initially developed.

Different detector types are nowadays used and computers have been improved to

such a point that they can perform complex unfolding procedures very quickly.

Moreover, the new systems employ portable personal computers that perform data

analysis and display of results quickly after measurement. New developments in

technology are expected to decrease the dimensions of data analysis equipment but it

is not likely that the large moderating spheres can be reduced in size.

Bonner Sphere spectroscopy is a very useful tool to provide information that

can reduce uncertainties in dosimetric measurements. It has been the system that

researchers have shown preference at, in order to provide solutions to measurement

problems. Despite innovations, energy resolution is always going to be poor. It should

however be possible to combine BSS data with high resolution measurements to

produce spectra with good resolution in the energy region where it is important.

1.3 The CERN Bonner Sphere Spectrometer

The Radiation Protection (RP) group at CERN is in possession of an Extended

range BSS (Fig.1.4), a recently built descendent of a conventional BSS. The

conventional BSS consists of a set of five moderating spheres made of polyethylene,

having outer diameters of 81 mm, 108 mm, 133 mm, 178 mm and 233 mm. In order

to better absorb the thermal neutron component, the sphere with the smallest diameter

can be surrounded by a cadmium (Cd) shell of thickness 1 mm, therefore this

configuration is called 81cd. The active part of the spectrometer is a spherical 3He

proportional counter with a diameter of 32 mm located in the centre of each sphere.

Two 3He proportional counters are used by the RP group, with gas pressures of 2 atm


8

and 4 atm. They both use a gas mixture of He- Kr inside their effective volume. The

Kr is used in order to reduce the range of the reaction products or in other words to

increase the stopping power. A more detailed description of the BSS and of the

specific 3He counters is given in [45]. For the purposes of the present study only the

2- atm 3He detector has been used.

In order to be able to study the neutron spectra around hadron accelerators,

where the high-energy neutron component is important, it was desirable to extend the

response of the BSS to substantially higher energies. To extend the range of the

conventional Bonner Sphere Spectrometer by means of two new spheres, nineteen

configurations of different sizes and materials were thoroughly investigated by means

of Monte Carlo simulations. The reader can find the details about the design studies of

the spheres or about other influencing factors (material, density, geometry, etc.)

elsewhere [45]. The convention of labeling each sphere by its diameter in centimeters

has been adopted in the present study. The two new spheres were named Ollio (for the

255 mm diameter sphere) and Stanlio (for the 119.5 mm diameter sphere).

Fig. 1.4. The extended range Bonner Sphere Spectrometer of the RP group at CERN.


9

Ollio consists of moderator shells of 3 cm polyethylene, 1 mm cadmium, 1 cm

lead and 7 cm polyethylene (from the central 3He proportional counter outwards). The

characteristic of this configuration is that it suppresses the response to incident

neutrons with energies lower than 100 keV and increases it for energies above

10 MeV and up to 1 GeV, as compared to the 233 mm sphere of the conventional

BSS. The response function shows the peak at about 10 MeV as it is typical for all

large detectors of a BSS. The second newly built sphere, Stanlio, consists of

moderator shells of 2 cm polyethylene, 1 mm cadmium and 2 cm lead (Fig.1.5). Its

response function does not show the peak at 10 MeV. This feature makes Stanlio a

useful complement to the other detectors. At low energies it behaves like a small

Bonner sphere, but at high energies the response is increased compared to the 233 mm

sphere. The response functions of the various detectors are shown in Figs.1.6- 1.7.

The nominal response matrix can be found in Appendix B.

Fig. 1.5. Cut through the moderator of the Bonner Sphere Stanlio.


10

Fig.1.6. The calculated absolute neutron fluence response functions of the CERN

BSS at full energy scale. For clarity, the statistical errors have been omitted in the

lower graph.


11

Fig. 1.7. The calculated absolute neutron fluence response functions of the CERN

BSS at low energies (upper plot) and at high energies (lower plot).


12

The accurate evaluation of the response matrix is essential for a correct use of

the spectrometric system. The response functions of the BSS were calculated by

Monte Carlo simulations performed with FLUKA98 [46]. The neutron response of

each detector was calculated for 78 incident neutron energies. For the neutron

energies of 0.05, 0.1, 0.25, 0.5, 1 and 2 GeV, a broad parallel beam having a slightly

larger extension than the sphere was assumed. In most experimental conditions low

energy neutrons are generated by down-scattering and they are undirectional.

Therefore, an isotropic distribution of the incident neutrons was chosen for the

72 low-energy groups, between E=19.6 x 106 eV and 1x 10-5 eV [45].

As Fig.1.6 shows, for each detector the response peaks at a given energy, depending

on the moderator’s size. The response of the various polyethylene spheres decreases

rapidly for energies above a few MeV. This is due to decreasing (n, p) cross section

with increasing neutron energy. For a small sphere the degree of moderation is small,

as is the capture of thermal neutrons in the moderator. Low-energy neutrons thus have

a reasonable probability of arriving at the thermal sensor and being detected, whereas

fast neutrons tend to escape. For larger spheres there is considerably more

moderation. There is also more capture which means low-energy neutrons tend to get

absorbed in the polyethylene. It is the high energy neutrons which thus have the

greatest probability of being detected in the sensor, and the response function peaks in

the high-energy region. In order to reproduce correctly the measured responses by

calculation and supplement them in energy ranges not measured, detailed information

concerning the geometry of the moderator and the 3He counter is required.

The difference in the shapes and position of the maxima in these response

curves serves as the basis for using the set of spheres as a simple neutron

spectrometer. By measuring the count rate with each sphere individually, an unfolding

process can provide some information about the energy distribution of the incident

neutrons.


13

1.4 Electronics

The electronics used with the CERN BSS system consist of a preamplifier, a

Serial Micro Channel SMC 2100 box (a module housing an amplifier, high voltage

and multi-channel analyser) produced by MAB [47] and called the mab box, a 3H

proportional counter and power supply. The Serial Micro Channel SMC 2100

modular concept design is capable of being connected to a serial interface of a pc. The

SMC 2100 is supported by a range of modules that are necessary in certain

applications for the measurement of radiation pulses. The multi-channel analyser

operates with 2048 channels. Minimum is channel zero and maximum is channel

2047. All pulses arriving from the spectroscopy amplifier are passing the window of

an existing single-channel analyser. The analog pulses of the amplifier are digitized

by a fast 12-bit analog-to-digital converter (ADC). The recognition of the peak

maximum is made in the MCA. The ADC values are stored in the MCA and are

normalized upon 2048 channels.

A program package (AM-SMCA01) is used with the electronics for the data

acquisition and processing. The program is available with different options depending

on the purpose of application. For the purposes of this dissertation the option ‘pulse

height analysis’ has always been used. The software is compatible with Windows

95/98 and more recently it has been upgraded to Windows 2000 and NT. More details

on the software can be found in [48].

1.5 Unfolding methods

1.5.1 Introduction

One of the drawbacks of the multi-sphere spectrometer is the mathematical

problem of unfolding the neutron spectrum. For the CERN BSS applications, eight

detector/moderator configurations are used, resulting in eight values of count rate

recorded from the detector. These eight data points are then used in the unfolding

code with the response function matrix to determine the neutron flux in the 78

energy bins covering the entire range of neutron energies from thermal up to


14

2 GeV. The derivation of the neutron spectrum from the experimental data is done

as follows. If sphere i has response function Ri(E) and is exposed in a neutron field

with spectral fluence Φ(E) then the sphere reading Mi is obtained mathematically

by folding Ri(E) with Φ(E), i.e.:

dEEERii )()( Φ=Μ ∫ (1.1)

Equation (1.1), formally known as a Fredholm integral of the first kind,

extends over the range of neutron energies present in the field. This equation does not

provide a unique solution because the continuous function Φ(E) cannot be defined by

a set of n discrete measurements. It can however be approximated by equation (1.2)

where Φj is the fluence in group j extending from energy Ej to Ej+1 and Rij represents

Ri(E) averaged over group j.

j

n

jiji R Φ=Μ ∑

=1 i=1, …, m (1.2)

The degree of approximation decreases as the number of groups n increases. If

there are m spheres, equation (1.2) represents a set of m linear equations. In the case

that m ≥n, equation (1.2) can be solved by using special mathematical methods.

However, because m is usually small, of the order of 10, the solution may provide a

poor representation of the spectrum. Spectra are therefore usually represented by an

array with n> m, which means that equation (1.2) can only be solved for Φj by using

smoothing functions or so-called additional a priori information. A clear distinction

has to be made between input guess spectrum and a priori information. Knowledge of

the spectral fluence distribution of the investigated neutron field, derived either from

calculations or previous measurements, is usually called specific a priori information.

In other words, adding of specific a priori information during unfolding is equivalent

to increasing the number of measurements. This means that specific a priori

information controls the results.


15

Even in the case that specific a priori information is available, it is reasonable

to require for the output spectrum a series of criteria based on the general physical

representation of the investigated neutron field. As an example, such criteria can be:

• No negative values for the differential fluence can be accepted

• The solution spectrum should be smooth. Even if structures are present in

the real spectrum, reliable information on these structures can only come from

specific a priori information. A Bonner Sphere system provides a few channel

unfolding and it is not capable of reproducing finer structures, although it can

indicate broad peaked fluence distributions. In a case that fine structures or

oscillations that are not included in the guess spectrum appear in the solution

spectrum, one should make sure that the unfolding code has not been

improperly used.

A solution can be called exact, approximate, or appropriate [49]. Exact

solutions may have zero errors, and might look reasonable. However, they may have

unphysical characteristics, such as oscillations. Usually, the unfolded data should not

be expected to have too good a fit, at least not better than the error of the input data. In

general, the trial vectors should contain the features of the neutron spectra one can

expect from the physics of the problem. Similarly, the smoothing functions, if any,

have to be properly chosen. Appropriate solutions can be obtained from good

measurements, and considerable experience is needed to judge just when a reasonable

spectral solution is reached.

Good approximations to Ri(Ej) can be obtained from simulation calculations

supported by measurements with well characterized monoenergetic and radionuclide

source neutrons. The increased computing power available nowadays makes the

Monte Carlo method the most appropriate approach to calculating response functions.

Monte Carlo codes have the advantage of allowing the sphere plus detector to be

geometrically modeled in detail, including non-spherically symmetric features.

Knowledge of the energy-dependent response, with the sphere diameter as a

parameter, is of fundamental importance for the spectrometry using unfolding

procedures. Using these data, measurements with a BSS set in an unknown field will

allow information on Φ(E) to be extracted. Many unfolding codes employing

different mathematical techniques have been used to perform spectrum unfolding for


16

multi sphere systems. Due to the non-uniqueness of the deconvolution, there are many

different methods that are based on different mathematical principles. Some of the

mathematical methods used are: maximum entropy, least-squares iteration etc. The

unfolding codes employed for the purposes of the present work are briefly described

below.

1.6 Unfolding codes

1.6.1 MAXED

The FORTRAN code MAXED (MAXimum Entropy Deconvolution) was

developed at EML (Environmental Measurements Laboratory, DOE, USA) [50]

specifically for the deconvolution (unfolding) of multi-sphere neutron spectrometer.

The maximum entropy deconvolution algorithm used in MAXED is a modification of

the one in Wilczek and Drapatz [51]. In data analysis, the maximum entropy principle

is widely used as a general and powerful technique for reconstructing positive

distributions in situations where only incomplete information is available. It requires

the maximization of entropy S given by equation (1.3)

])ln([ iDEF

iDEFi

ii fff

ffS −+−= ∑ (1.3)

where fi is the determined fluence in group i and fiDEF is the (discretized) default

spectrum.

For this purpose the MAXED code implements a special algorithm (annealing

global optimization algorithm) described in [52]. In the case of the deconvolution of

neutron spectra, the distribution to be determined is the neutron energy spectrum f(E)

also called solution spectrum and the constraints are the measurements and the

experimental errors associated with them.

The default spectrum contains all a priori information, based on which the

code is using the maximum entropy method to derive a new probability distribution

that takes into account the new information provided by the measurements. A formal


17

argument due to Shore and Johnson [53] shows that the maximum entropy method is

the only general method of choosing the new distribution that does not lead to

inconsistencies.

The approach followed in MAXED has several features that make it attractive:

it permits inclusion of a priori information in a well-defined and mathematically

consistent way, the algorithm used to derive the solution spectrum is not ad hoc (it

can be justified on the basis of arguments that originate in information theory), and

the solution spectrum is a non-negative function that can be written in closed form.

For more information on MAXED, the reader should refer to [50].

1.6.2 GRAVEL

The GRAVEL code is based on the formalism described by M. Matzke

[54-56]. It is adapted to utilize the capabilities of the MATLAB [57] language for the

I/O of data and for the graphical representation of other relevant quantities (e.g. χ2,

total flux, etc.) during the iterations, thus allowing a better understanding of the

evolution of the procedure. This code performs a nonlinear least-square adjustment.

The constraint of non-negative particle fluences is essential for the operation of the

program. To take into account the condition of non-negative fluence, the logarithms

of fluence Φ are used instead of the fluence Φ. They are determined by a special

gradient method [58] which minimizes the χ2 value obtained by the comparison

between experimental and reconstructed data.

An initial input spectrum (guess spectrum) is needed. A solution always exists,

but the solution spectrum depends on this input spectrum. The initial spectrum has

usually the shape of 1/E. In case there exists a more precise initial spectrum provided

for example by Monte Carlo calculations, the unfolding solution is a higher resolution

spectrum. These iterations are done in a way that is not quite transparent, so that a

uncertainty propagation cannot be easily performed. For this reason, a lot of

experience is needed for a successful unfolding of neutron spectra employing the

GRAVEL code. The unfolding in the context of the present work has been performed

with GRAVEL by experts in the Polytechnic of Milan [59].


18

1.7 MONTE CARLO method

1.7.1 Introduction

Numerical methods that are known as Monte Carlo methods can be described as

statistical simulation methods, where statistical simulation is defined in quite general

terms to be any method that utilizes sequences of random numbers. Credit for

inventing the Monte Carlo method often goes to Stanislaw Ulam, a Polish born

mathematician who worked for John von Neumann on the United States’ Manhattan

Project during World War II. Ulam is primarily known for designing the hydrogen

bomb with Edward Teller in 1951. He invented the Monte Carlo method in 1946

while pondering the probabilities of winning a card game of poker.

Ulam did not invent statistical sampling. This had been employed to solve

quantitative problems before, with physical processes such as dice tosses or card

draws being used to generate samples. Ulam’s contribution was to recognize the

potential for the newly invented electronic computer to automate such sampling.

Working with John von Neuman and Nicholas Metropolis, he developed algorithms

for computer implementations, as well as exploring means of transforming non-

random problems into random forms that would facilitate their solution via statistical

sampling. This work transformed statistical sampling from a mathematical curiosity to

a formal methodology applicable to a wide variety of problems. It was Metropolis

who named the new methodology after the casinos of Monte Carlo.

Ulam and Metropolis published the first paper on the Monte Carlo method in

1949 [60]. Monte Carlo is now used routinely in many diverse fields, from the

simulation of complex physical phenomena such as radiation transport in the earth's

atmosphere to the simulation of the esoteric sub-nuclear processes in high-energy

physics experiments. Its name does not mean to imply that the method is either a

‘gamble’ or ‘risky’. It simply refers to the manner in which individual numbers are

selected from valid ‘representative collections of input data’ so they can be used in an

iterative calculation process. These representative collections of data are some sort of

a Frequency Distribution that is converted to a Probability Distribution.


19

Monte Carlo Simulation methods are primarily used in situations where:

• The system being studied can be mathematically described by a metric,

which can be either parametric or analytic.

• The Input Data can be written as some sort of a frequency distribution.

• The calculated distribution histogram of the ‘answer’, or Output, must

accurately reflect the Input data.

• The calculated uncertainty in the ‘answer’, or Output, must be an accurate

measure of the validity of the model.

Monte Carlo Simulations could be regarded as ‘True Stochastic Simulations’ in

that they describe the final state of a model by just knowing the frequency

distributions of the parameters describing the beginning state and the appropriate

metric that maps or transforms the beginning state to the final state.

1.7.2 Major components of a Monte Carlo algorithm

The primary components of a Monte Carlo simulation method include the following:

• Probability distribution functions (pdf's)- the physical (or mathematical)

system must be described by a set of pdf's.

• Random number generator- a source of random numbers uniformly distributed

on the unit interval must be available.

• Sampling rule - a prescription for sampling from the specified pdf's, assuming

the availability of random numbers on the unit interval, must be given.

• Scoring (or tallying) - the outcomes must be accumulated into overall tallies or

scores for the quantities of interest.

• Error estimation - an estimate of the statistical error (variance) as a function of

the number of trials and other quantities must be determined.

• Variance reduction techniques - methods for reducing the variance in the

estimated solution to reduce the computational time.

• Parallelization and vectorization - algorithms to allow Monte Carlo methods

to be implemented efficiently on advanced computer architectures.


20

1.8 The FLUKA Monte Carlo code

FLUKA is a FORTRAN simulation code generally used to calculate particle

transport and interactions with matter. The development of FLUKA since its first

appearance in the sixties has seen many advances that cover an extended range of

applications. These applications include proton and electron accelerator shielding,

calorimetry, dosimetry, detector design, cosmic rays, neutrino physics, radiotherapy,

etc. A more thorough and detailed historic review of FLUKA can be found in [46].

This simulation code can perform high accuracy simulations for the interaction

and propagation in matter of about 60 different particles: photons and electrons with

energy from 1 keV to TeV, muons and hadrons with energy up to 20 TeV (to be

extended soon to 10 PeV), neutrinos, and all the corresponding antiparticles. In

particular neutrons are simulated down to thermal energies as well as heavy ions.

FLUKA can employ very large-scale and complex geometries. For this

purpose it is using an updated version of the well-known Combinatorial Geometry.

This improved version of the Combinatorial Geometry is fast, flexible and user-

friendly at a much higher degree than the older versions, allowing the users to track

correctly charged particles even in the presence of magnetic or electric fields. New

bodies have been introduced, resulting in increased rounding accuracy, speed and

even easier input preparation. Various visualization and debugging tools have also

been integrated into FLUKA thus making it a very powerful tool with many different

capabilities according to the user’s needs.

Several models are employed in FLUKA for the transport and propagation of

the different groups of particles in different energy ranges. The hadron-nucleon

interaction models are based on resonance production and decay below a few GeV,

and on the Dual Parton model at higher energies. Two models are also used in hadron-

nucleus interactions. At momenta below 3 GeV/c the PEANUT package includes a

very detailed Generalized Intra-Nuclear Cascade (GINC) model. At high energies the

Gribov-Glauber multiple collision mechanism is included in a less refined GINC. One

can also simulate photonuclear interactions with FLUKA. These are described by

Vector Meson Dominance and Delta, Quasi-Deuteron and Giant Dipole Resonance.

Multiple Coulomb scattering and ionization fluctuation are also integrated in the code,


21

allowing it to handle problems such as electron backscattering even in the few keV

energy range. FLUKA can also simulate synchrotron radiation and optical photons.

The FLUKA physical models are described in detail in several journal and conference

papers [61-63].

A very interesting and important feature of FLUKA, probably unique in

comparison with other Monte Carlo codes, is its capability to be used both in a biased

mode and in a fully analog mode. This means that on the one hand it can be used to

predict fluctuations, signal coincidences and other correlated events and on the other

hand it can be used to investigate other rare events because of its wide choice of

available statistical techniques.

22

CHAPTER 2

Calibration facilities

2.1 Introduction

Calibration is a set of operations that establish, under specific conditions, the

relationship between values indicated by a detector (dosemeter), and the

corresponding known (i.e. conventionally true [64]) values of the quantity to be

measured [65,66]. This relationship can be established by determining the response of

a device for the full range of radiation energies and angles of incidence for its

intended use. Laboratory calibrations determine the calibration factor for each

individual instrument under standard test conditions [66,67]. The calibration factor f

is the factor by which the reading of the device is multiplied to obtain the value of the

quantity to be measured.

The calibration procedure is linked to the quality assurance of an instrument.

Assuming that the available instrument is the suitable one for the specific need, its

performance depends strongly on the quality of the response matrix as well as on

other parameters (unfolding technique, accuracy of data, etc.). Therefore, the

calibration of the instrument for the verification of the calculated (usually by Monte

Carlo) response matrix is essential.

In order to calibrate the Bonner sphere detectors and to determine their

response functions, measurements with monoenergetic neutrons are required. Typical

energies used for this purpose are those recommended by the International

Organization of Standardization (ISO) [68]. An extensive calibration campaign has

been carried out to evaluate the response of the BSS to monoenergetic neutrons.

Calibrations with monoenergetic neutron beams were carried out at the Physikalisch-

Technische Bundesanstalt (PTB) in Braunschweig (Germany) in two consecutive

years. In 2001 the BSS was calibrated at energies 0.144 MeV, 1.2 MeV, 5 MeV and

14.8 MeV [45] and in 2002 the calibration was done at energies 0.565 MeV, 2.5 MeV,

CHAPTER 2. Calibration facilities

23

8 MeV and 19 MeV. For the full characterization of the BSS another calibration

campaign was performed at UCL (Louvain–la Neuve, Belgium) in 2003.

The calibration of the BSS in the hadron beam of the CERN-EU Reference

Field (CERF) was later performed in the context of the upgrade of its response to

charged hadrons (see chapter 4). A short description of all irradiation facilities that

have been used for calibration and tests of the BSS system, is given below.

2.2 The PTB accelerator facility

2.2.1 Description

The Physikalisch- Techmsche Bundesanstalt (PTB) accelerator facility [69-71]

for fast neutron research, was constructed at Braunschweig between 1971 and 1974. A

schematic drawing of the facility is shown in Fig. 2.1.

Fig. 2.1. Neutron research facility at PTB. (1) the cyclotron, (2) the quadrupole magnet, (3) the neutron producing target.


24

The cyclotron is located in the basement of the building. It is mounted on a

swivel arm, about 5 m in radius. The external beam can thus be moved in a horizontal

plane such that it is always directed to the vertical axis of rotation. The scattering

probe S (Fig 2.2) is located on the axis of rotation. The neutron producing target T, set

up at a distance of 15-25 cm from the scattering probe, can be turned around the probe

together with the whole cyclotron, thus varying the scattering angle.

For experiments involving neutron sources in the low backscatter experimental

hall, the external cyclotron beam can be transported via two 90 ° deflecting magnets

from the cyclotron room to one of the entrance ports of the switching magnet on the

ground floor. The beam then can be directed into one of the four beam lines within the

experimental hall. These beam lines are terminated by the targets, i.e. the

monoenergetic neutron sources.

Fig. 2.2. Layout of the neutron scattering experiment. C: cyclotron, Q: quadrupole magnet, T: neutron producing target, S: scattering probe (= pivot of cyclotron movement), P: polyethylene shields, W: water tank, D: neutron detector, B: concrete shield.


25

The low backscatter experimental hall has an area of 24 m x 30 m and a height

of 14 m. In order to protect the environment against radiation, the walls of the

experimental hall are 1.2 m thick, the ceiling is 0.4 m thick concrete and an additional

0.3 m water layer on the roof may be provided in case additional shielding is needed.

The four beam lines are installed on a plane at 6.25 m above ground. A low scattering

grid floor of aluminum 4.5 m above ground, allows access to the targets (Fig 2.3).

Fig. 2.3. Scheme of the experimental set-up at PTB (not to scale).


26

The reference monitors used to measure the neutron fluence (Fig. 2.3) are the

following :

• The New monitor (NM): it consists of a long cylindrical 3He proportional

counter surrounded by a cylindrical polyethylene layer 15 cm thick. It was

mounted at a distance of 5.5 m from the target at an angle of 19 degrees.

• The 3He proportional counter: A 3He proportional counter embedded inside a

polyethylene moderator (50 mm in diameter and 233 mm long) was placed on

the right hand side to the beam line, just beside it and upstream of the target.

• The Geiger-Mϋller counter (GM): In order to measure the photon contribution

a Geiger-Mϋller counter was mounted on top of the production target.

• The charge monitor: it is used for monitoring the charge and thus the intensity

of the beam impinging on the target. It was placed in the beam line but it is not

shown in Fig. 2.3.

• Precision long counter (PLC): it consists of a long BF3 proportional counter

placed inside a large, specially shaped, moderating cylinder made of

polyethylene, boron loaded polyethylene, aluminum and cadmium. It was

placed at a distance of about 5.5 m from the target, at an angle of 100 degrees.

2.3 Calibration of the BSS with Quasi- Monoenergetic Neutron Fields at UCL

2.3.1 Introduction

For the full characterization of the BSS the calibration at higher energies than

those used in PTB was considered as essential. For that purpose the calibration

campaign of the BSS was completed at the cyclotron of the Université Catholique de

Louvaine La Neuve (UCL, Belgium) at energies of 33 MeV to 60 MeV [72-76].


27

2.3.2 The neutron facility and the experimental set- up

The CYClotron of LOuvaine la NEuve (CYCLONE) is a multiparticle,

variable energy isochronous cyclotron capable of accelerating protons up to 80 MeV,

deuterons up to 55 MeV, alpha particles up to 110 MeV and heavier ions up to an

energy of 110 Q²/M MeV (where Q is the charge state and M the mass of the ion).

The UCL neutron beam facility enables the production of quasi-monoenergetic

neutron beams. Monoenergetic protons with energies between about 25 and 70 MeV,

impinging on a 3 mm or 5 mm thick Li target of natural isotope composition, are used

for neutron production. The layout of the neutron production and collimation area is

shown in Fig. 2.4. The pulsing system (D in Fig. 2.4) consists of a pair of deflecting

plates, inserted in the beam line between the cyclotron beam exit port and the

switching magnet.

Fig. 2.4. Schematic drawing of the cyclotron (left) and the neutron beam facility at UCL. D: deflector, S: switching magnet, T: neutron production target, C: collimator. The scale indicates the distance from the target in metres.

The deflecting voltage is provided by a resonance amplifier, phase-locked to

the main cyclotron oscillator. Proton bursts passing in between the plates during the

time of zero deflecting voltage arrive unaffected at the neutron target while all other

bursts observe a deflecting voltage and are stopped by a slit system in the beam line.

After passing through the target the protons are deflected with a dipole magnet to a

graphite beam dump. A collimator restricts the neutron beam to a diameter of 30 mm

at its exit. A cleaning magnet serves to remove charged particles contaminating the

neutron beam. The neutron beam is dumped in a cave downstream of the

measurement cave.


28

The calibration measurements were performed in Cave Q, a heavily shielded

concrete room (Fig. 2.5).

Fig. 2.5. The cave Q at UCL.


29

2.4 The CERN-EU Reference Field (CERF)

2.4.1 Description of the facility

The CERF facility is installed in one of the secondary beam lines (H6) from

the Super Proton Synchrotron (SPS), in the North Experimental Area on the Prevessin

(French) site of CERN (Figs 2.6- 2.8). A positive hadron beam with momentum of

usually 120 GeV/c is stopped in a copper target, 7 cm in diameter and 50 cm in length

which can be installed in two different positions inside an irradiation cave.

Fig. 2.6. Axonometric view of the CERF facility. The reference positions are also marked (iron and concrete roof and side shield). The side shielding on the Salève side is removed to show the inside of the irradiation cave with the copper target set-up.


30

Fig. 2.7. Axonometric view of the CERF facility. The reference positions are also shown. The side shielding on the Jura side is removed to show the external set- up of the facility.

The secondary particles produced in the target traverse a shielding on top of

these two positions. The shielding is made up of either 80 cm concrete or 40 cm iron

and in both cases uniform radiation fields are produced. The fields extend over two

areas of 2 x 2 m2 located at approximately 90° with respect to the incoming beam

direction. Each of them is divided into 16 squares of 50 x 50 cm2, with each element

of these ‘grids’ representing a reference exposure location. Additional measurement

positions are available behind the lateral shielding of the irradiation cave, at the same

angles with respect to the target as for the two roof positions. Shielding is either 80

cm or 160 cm concrete, and at both positions 8 additional exposure locations

(arranged in 2 x 4 grids made up of the same 50 x 50 cm2 elements) are provided. The

nominal measurement locations (the reference field) are at the centre of each square at

25 cm height above floor, i.e. at the centre of a 50 x 50 x 50 cm3 air volume, where

the radiation field is calculated.

The beam is slow-extracted from the SPS over a few seconds. During the

pulse the beam intensity is constant, thus producing a constant radiation field at the

exposure locations. Typical values of dose equivalent rates are 1–2 nSv per PIC-count


31

(see section 2.4.2) on top of the 40 cm iron roof-shield and about 0.3 nSv per PIC-

count outside the 80 cm concrete shields (roof and side). The dose equivalent rate at

the reference positions can be varied by proper adjustments of the beam intensity on

the target and it typically ranges from 25 µSv/h to 1 mSv/h on the iron roof-shield and

from 5 µSv/h to 600 µSv/h on the 80 cm concrete roof or lateral shield. The energy

distributions of the various particles (mainly neutrons, but also photons, electrons,

muons, pions and protons) at the various exposure locations, have been obtained by

Monte Carlo simulations performed with the FLUKA code [61-63]. Details of the

latest simulations are given elsewhere [77]. A more thorough description of the CERF

facility is given in [78].

Fig. 2.8. Plan and sectional views of the CERN-EC facility.


32

2.4.2 Beam monitoring at CERF

The intensity of the primary beam at CERF is monitored by a Precision

Ionization Chamber (PIC), which is installed about 405 m downstream of the T4

production target. It has been the primary monitor at the CERF facility since many

years. It is placed in the beam just upstream of the copper target, connected to a

current digitizing circuit. Its design is described elsewhere [79-80].

The PIC is an open-air ionization chamber with cylindrical shape. Its sensitive

volume is 0.86 litres (diameter: 185 mm, active length: 32 mm). The charge produced

by ionization by the beam in this volume is collected at a capacitor that is discharged

whenever the charge attains a predefined threshold. It then issues one count that is a

measure of the number of beam particles that have produced this charge. The PIC

serves to normalize the experimental data to the number of particles in the H6 beam.

One PIC-count corresponds (within 10%) to 2.2x104 particles impinging on the target.

In the first years of the CERF operation a check of the beam shape and

position just before the copper target was performed by taking images of the beam,

originally with a radiographic film (for example, ref. [81]) and in the recent past with

a Polaroid film that has the advantage of immediate development. X-ray films

required development in the laboratory of the individual dosimetry service, so that the

results were often available only after the run. This method was time consuming and

only provided rough information on the ‘beam spot’, mainly because of the resolution

of the films. The use of X-ray films allowed a reconstruction of the beam profile, but

only by an off-line analysis [82]. That method was good enough for the purpose of

checking the correct alignment of the beam in the H6 line, which usually did not

present any problem as the beam set-up was done by the operator from the

Experimental Areas control room.

In 2002 two new instruments were installed in the irradiation cave. The

advances in the beam and reference field monitoring at the CERF facility since 2002

are described in chapter 6.

33

CHAPTER 3

Calibration at reference neutron fields

3.1 Calibration of the BSS in monoenergetic neutron fields

3.1.1 Introduction

In order to calibrate the Bonner sphere detectors and to determine their

response functions, measurements with monoenergetic neutrons were performed at

Physikalisch-Technische Bundesanstalt (PTB) in Braunschweig (Germany) in 2001

and 2002. The calibration campaign was completed with the measurements at UCL

(Louvain–la Neuve, Belgium), at energies 33 MeV and 60 MeV, in 2003. The 2001

and 2002 results are presented in detail and in the following the results of all

calibrations are summarised and graphically illustrated at the end of this chapter

(Figs 3.17- 3.20).

3.1.2 Calibration at PTB- Materials and methods

Monoenergetic neutron beams were produced at the 4 MV Van de Graaff

accelerator of the PTB [69-71]. The production reactions are presented in Table 3.1.

Table 3.1. Neutron production reactions for the calibration at PTB.

Neutron Energy [MeV]

Ion energy [MeV] Reaction

0.565 2.304 7Li(p,n)7Be 2.5 3.36 3H(p,n)3He 8.0 5.2 2H(d,n)3He 19.0 2.664 3H(d,n)4He

CHAPTER 3. Calibration of the BSS at reference neutron fields

34

The widths of the neutron distributions are sufficiently narrow to allow the

neutron fields to be considered as monoenergetic. The width of the distribution is

governed mainly by the energy loss of the beam in the target and the energy

distribution thus approximates more to a rectangular shape rather than to a Gaussian.

The measurements were performed in a low scatter hall of dimensions 24 x 30

x 14 m3. The irradiation location was on a grid floor in the centre of the hall (Fig. 3.1).

The irradiations were performed in open geometry, i.e. the neutron producing target

was situated in the centre of the hall while the fluence monitors and the BSS were set

up without any shielding. The BSS support stands were light to minimize neutron

scatter and were designed to allow spheres to be changed without disconnecting the

electronics. Each sphere had its own specific stand and all stands fit on a common

base.

Fig. 3.1. The irradiation hall at PTB.


35

Measurements were performed by exposing each Bonner sphere to neutrons

emitted in the forward direction at a distance of 2 m from the production target. Solid

targets were used consisting of thin layers of metallic LiOH and Ti, loaded with

deuterium or tritium and evaporated on Ta, Ag or Al backings. For the 0.565 MeV

measurement the target consisted of Li while for the 8 MeV a gas target was used.

The target is air-cooled and rotated around the beam spot during the experiment.

For the calibration of the BSS the 2-atm 3He detector was used. The spheres

were positioned for irradiation on a support that could be moved around the target on

a circle of a maximum radius of 5.2 m. The irradiation of the spheres was always

performed in such a geometry that allowed a perpendicular orientation of the 3He

detector with respect to the beam axis. From the previous calibration in the same hall

[45] it was shown that the results from a perpendicular irradiation of the 3He detector

agree with the results of a parallel irradiation within the statistical uncertainties. In the

case of an accidental displacement of the spheres during the measurement, it was

estimated that the angle subtended was less than 5° with respect to the beam axis and

thus the anisotropy effect was judged to be negligible [83]. The charge monitor which

integrates the beam current was used for monitoring the beam.

The radiation components that contribute to the BSS reading are schematically

shown in Fig. 3.2. Various corrections have to be applied in order to derive the

response to monoenergetic neutrons.

Fig. 3.2. Radiation components contributing to the BSS reading. T: direct neutrons, B: target- scattered neutrons reaching the detector, Ain: neutrons scattered by air in the detector, Aout: neutrons scatter by air away from the detector, S: neutrons scattered by other sources (wall, etc) in the detector.


36

The contribution to the counter reading from room and air scatter neutrons was

taken into account by measurements made with a shadow cone. The shadow cone was

made of 20 cm iron and 30 cm polyethylene and was interposed between the target

and the monitor (Fig. 3.3). In the previous calibration of the BSS different shadow

cones were used, depending on the size of the sphere. For the present measurements

only one shadow cone was used. The use of a single shadow cone did not increase the

risk of overshadowing for the smaller spheres or of insufficient shadowing for the

larger ones [83]. It was thus decided to use one shadow cone that would allow a good

compromise between precision and time saving. The overall uncertainty due to this

effect did not exceed 1%.

Fig. 3.3. The shadow cone technique.


37

Neutrons were emitted from a target at all angles. Those emitted at laboratory

angles around 90o are scattered in the target backing and target mounting. The target

assembly is basically a light construction, however the mass of the tube of the vacuum

chamber, the mounting ring of the backing, and the target backing itself all give rise to

neutron scattering or (n,2n) reactions. Some of these neutrons can reach the Bonner

sphere undergoing calibration at 0o. This contribution cannot be subtracted using

shadow- cone measurements and must be evaluated separately. The number of target

scattered neutrons and their spectral distribution can be obtained from Monte Carlo

calculations that consider the detailed description of the target geometry and the

reaction kinematics [84].

For the purposes of this study, the spectral distributions of the target-scattered

neutrons were provided by Schlegel [85] and are shown in Fig. 3.4. Obviously no

scattering occurs in the gas target used for the calibration at 8 MeV.

Especially for the calibration at 19 MeV an additional correction had to be

made, the blank target correction [86]. At this energy, reactions with components of

the target layer such as other isotopes or elements can produce unwanted neutrons that

induce a considerable contribution to the reading of the BSS. To correct for this effect

another measurement was performed with a background target known as blank target.

The blank target is almost identical to the main target disc, except that the material for

the production of the primary neutrons is omitted. For instance, the blank target used

in the measurements with Tritium targets consists of a titanium layer of the same

thickness on an identical backing disc. All beam parameters were kept constant for the

subsequent measurements, always with the beam current integrator as monitor.


38

Fig. 3.4. Uncollided and target-scattered neutron spectral distributions calculated with the TARGET code [85] (upper left: 0.565 MeV, upper right: 2.5 MeV, lower left: 8 MeV, lower right: 19 MeV).


39

3.1.3 Data analysis

For the measurements with the BSS detectors, the MAB (Münchner Apparate

Bau, Germany) electronics [47] was used. The electronics consist of an amplifier,

high voltage and a multi-channel analyser housed in a box. In addition, a pre-amplifier

is connected to the box. This system allows the acquisition of a pulse height spectrum

of the 3He proportional counter. The shape of a typical pulse height spectrum of the 3He proportional counter is shown in Fig. 2.5. Such a spectrum was acquired for every

Bonner sphere. The mean count rates varied between 25 cts/sec and 1200 cts/sec for

measurements without shadow cone, 20 cts/sec and 150 cts/sec for measurements

with shadow cone and between 80 cts/sec and 325 cts/sec for measurements with the

blank target. The count rates were kept low with frequent adjustments of the beam

current by the accelerator operator. This way neither pile –up occurred or dead time

corrections had to be applied. The duration of each measurement varied between 300

sec and 600 sec. The integral detector counts above threshold (Fig. 3.5) were in all

cases above 10000 so the statistical uncertainties were kept low.

Fig. 3.5. A typical pulse height spectrum acquired with the electronics of the 3He proportional counter. A simple threshold placed at about channel 100 allows the separation of the neutron-induced events from noise and gamma ray induced events.


40

At the calibration energy of 8 MeV measurements done without gas filling are

important in order to evaluate the contribution of neutrons that are produced in the

entrance window. Because of technical reasons, these measurements could not be

done. In the calibration in 2001 the percentage of the detector’s reading from neutrons

produced in a gas target without filling to one with gas filling, was evaluated for each

of the Bonner Spheres and it was found to be of the order of 1-2 % for 5 MeV [45].

Therefore, for the calibration in 2002 it was considered that the corresponding

percentage is 1.5% for the larger spheres and 2% for the smaller ones, given the fact

that the production reactions for 5 MeV (in 2001) and for 8 MeV (in 2002) have

similar cross sections.

The only monitor used for the measurement of the spectral fluence was the

charge monitor. The fluence values during the time intervals of the measurements

were communicated by PTB, already corrected for the attenuation in air as well as for

the dead time of the fluence monitor.

The absolute fluence response of the BSS is expressed as the number of counts per

incident neutron fluence. The response RT(E) of each Bonner sphere to direct neutrons

with spectral fluence ΦT(E) is associated with the reading MT of the sphere in the

following expression:

dEEER TTT )()( Φ=Μ ∫ (3.1)

As explained in the Section 3.1.2, there are other contributions to the reading

of the sphere that have to be subtracted, either as background contributions or as

‘noise’ to the real measurement. Thus the reading of the Bonner sphere due to direct

neutrons can be expressed in the following way:

TSSCST Μ−Μ−Μ=Μ (3.2)

where MS is the reading from measurements without shadow cone, MSC is the reading

of the Bonner spheres from measurements with shadow cone and MTS is the reading

of the Bonner spheres due to target scattered neutrons. In the case of calibration of the

BSS with 19 MeV neutrons, equation (3.2) is written as:


41

BTTSSCST Μ−Μ−Μ−Μ=Μ (3.3)

where MBT is the reading from measurements with the blank target. Each of the terms

MSC, MTS, MBT expresses a fraction of the response of the BSS to unwanted neutrons,

hence they can be written as follows:

dEEER SCSCSC )()( Φ=Μ ∫ (3.4)

dEEER TSTSTS )()( Φ=Μ ∫ (3.5) dEEER BTBTBT )()( Φ=Μ ∫ (3.6)

where ΦSC(E), ΦTS(E), ΦBT(E) are the corresponding spectral fluences. For the

definition of the target-scattered fluence ΦTS(E), Monte Carlo calculations were

performed by the PTB [85,86].

To determine the fraction of the reading of each Bonner sphere due to target

scattered neutrons, the spectral fluence ΦTS(E) provided by PTB was folded with the

Monte Carlo calculated response functions according to equation (3.5). Then this

number of theoretical counts MTS was subtracted from the reading obtained during the

measurement without shadow cone, and normalized to the fluence given for the same

time interval. The magnitude of the target- scattering correction depends strongly on

the neutron energy and on the diameter of the sphere. For the present analysis it was

found that the target- scattering effect is not contributing essentially to the reading of

the BSS, however the corrections were done for a more precise determination of the

response to direct neutrons [87- 88].

Large percentages occur only where the absolute value of the response is very

small. Most of the corrections lie between 0.02% and 3.6% as shown in Table 3.2.


42

Table 3.2. Percent contribution of the target- scattered neutrons to the total reading of the BSS.

Energy [MeV] Sphere

0.565 2.5 19 81 0.02 % 2.24 % 0.05 %

81cd 0.02 % 2.63 % 0.05 % 108 0.05 % 2.11 % 0.11 % 133 0.08 % 1.72 % 0.11 % 178 0.05 % 1.70 % 0.11 % 233 0.02 % 1.36 % 0.11 %

Ollio 0.02 % 3.59 % 0.10 % Stanlio 0.02 % 2.81 % 0.06 %

The effect of scattered neutrons depends on how the response of the

instrument in the energy region where the scattered neutrons occur compares with the

response to the primary neutrons. A contribution of 2-3% due to scattered neutrons

whose spectral distribution is rather flat and extends to very low energies, can induce

an important contribution (up to 40%) in the response of a small Bonner sphere when

calibrated at high energies. This can be observed in Table 3.3, which shows the

contributions to the BSS readings due to scattered radiation. Table 3.4 shows the same

effect but in terms of fluence responses.

Table 3.3. Percent contribution of scattered radiation

to the BSS reading.

Energy [MeV] Sphere 0.565 2.5 8 19

81 18% 40% 29% 28% 81cd 18% 36% 24% 24% 108 26% 20% 9% 19% 133 32% 10% 8% 16% 178 17% 7% 5% 13% 233 5% 6% 17% 9%

Ollio 5% 9% 4% 6% Stanlio 16% 37% 20% 19%


43

Table 3.4. Ratio of the response of the Bonner spheres to direct neutrons over the response to direct plus scattered neutrons.

Energy [MeV] Sphere 0.565 2.5 8 19

81 0.85 ± 0.06 0.72 ± 0.07 0.78 ±0.06 0.04 ± 0.01

81Cd 0.87 ± 0.06 0.75 ± 0.07 0.81 ± 0.06 0.08 ± 0.01

108 0.90 ± 0.06 0.84 ± 0.07 0.89 ± 0.06 0.15 ± 0.01

133 0.92 ± 0.06 0.89 ± 0.07 0.92 ± 0.06 0.20 ± 0.04

178 0.94 ± 0.06 0.93 ± 0.07 0.95 ± 0.06 0.31 ± 0.04

233 0.95 ± 0.07 0.95 ± 0.07 0.96 ± 0.07 0.47 ± 0.07

Ollio 0.96 ± 0.06 0.96 ± 0.07 0.97 ± 0.06 0.64 ± 0.07

Stanlio 0.87 ± 0.06 0.76 ± 0.07 0.88 ± 0.06 0.31 ± 0.03

For the error analysis of the absolute fluence responses the following errors

were considered and propagated with the error propagation formula:

• Statistical uncertainty in the counting rates of the BSS (raw data). This is between

1% and 3% depending on the total counts above the threshold.

• Systematic uncertainty of the fluence monitors and statistical uncertainty in their

counting rate. The first uncertainty varied between 3.8% for the energies

0.565 MeV and 8 MeV, and 5.6% for the energies 2.5 MeV and 19 MeV.

• Especially for the measurements at 8 MeV, the contribution of neutrons produced

in the entrance window was taken into account. This was estimated to be 1.5% for

the larger spheres and 2% for the smaller spheres, since the latter are more sensitive

to scattered radiation.

3.1.4 Results

Following the procedure described above and correcting for room and in-

scatter contribution as well as for target- scattering, the experimental absolute fluence

responses of the BSS were determined. These are listed in Table 3.5 together with

their total experimental uncertainties.


44

Table 3.5. Experimental absolute fluence responses to monoenergetic neutrons with energies 0.565 MeV, 2.5 MeV, 8 MeV and 19 MeV.

Energy [MeV]

Sphere 0.565 2.5 8 19

81 0.735±0.033 0.264±0.020 0.092±0.005 0.028±0.003

81Cd 0.741±0.033 0.270±0.020 0.097±0.005 0.049±0.003

108 1.965±0.084 0.936±0.059 0.381±0.016 0.171±0.016

133 2.797±0.116 1.788±0.105 0.792±0.032 0.414±0.075

178 2.899±0.118 2.768±0.155 1.510±0.059 0.777±0.080

233 1.819±0.104 2.789±0.152 1.914±0.104 1.066±0.139

Ollio 0.591±0.024 1.237±0.067 0.974±0.038 0.786±0.071

Stanlio 0.661±0.029 0.253±0.018 0.159±0.007 0.207±0.020

3.1.5 Comparison of experimental to MC calculated responses

The calculated responses taken as references in this work are those reported as

‘mean response functions’ in reference [45]. The responses at the energies of

calibration were derived by linear interpolation. These are presented in Table 3.6.

Table 3.6. The calculated absolute fluence responses for the calibration energies (linear interpolation between the mean response functions of the BSS).

Energy [MeV]

Sphere 0.565 2.5 8 19

81 0.886 ± 0.089 0.302 ± 0.030 0.083 ± 0.008 0.036 ± 0.004

81Cd 0.860 ± 0.086 0.281 ± 0.028 0.089 ± 0.009 0.050 ± 0.005

108 2.157 ± 0.216 1.036 ± 0.104 0.352 ± 0.035 0.154 ± 0.015

133 3.017 ± 0.302 1.859 ± 0.186 0.737 ± 0.074 0.339 ± 0.034

178 3.013 ± 0.301 2.800 ± 0.280 1.489 ± 0.149 0.734 ± 0.073

233 1.877 ± 0.188 2.890 ± 0.289 2.006 ± 0.201 1.139 ± 0.114

Ollio 0.605 ± 0.061 1.301 ± 0.130 1.069 ± 0.107 0.878 ± 0.088

Stanlio 0.715 ± 0.072 0.258 ± 0.026 0.146 ± 0.015 0.209 ± 0.021


45

Any new calculation should be validated against measurements to avoid errors

due to, for example, mistakes in the computational model. To check the consistency

of the calculated responses with the measured ones, the ratio Rcalc/Rexp was

determined, where Rcalc is the calculated response of a given sphere and Rexp its

experimentally derived response. The results are presented in Table 3.7 and plotted in

Figs. 3.6- 3.8.

This ratio should be close to unity, within the uncertainties, for all spheres.

The overall ratio that is calculated as the weighted mean of all ratios, i.e. <Rcalc/ Rexp>

is the calibration factor of the specific 3He proportional counter. These ratios are

shown in Table 3.8. At the end of this chapter the reader can find the final calibration

factor for both calibration campaigns.

Table 3.7. Ratios of calculated over measured absolute fluence responses for all energies.

Energy [MeV] Sphere

0.565 2.5 8 19

81 1.205±0.132 1.143±0.144 0.906±0.101 1.295±0.192

81Cd 1.160±0.127 1.041±0.129 0.922±0.102 1.019±0.119

108 1.098±0.119 1.107±0.131 0.924±0.100 0.901±0.123

133 1.079±0.117 1.040±0.121 0.931±0.100 0.819±0.170

178 1.039±0.112 1.012±0.116 0.986±0.106 0.945±0.136

233 1.032±0.119 1.036±0.118 1.048±0.119 1.068±0.175

Ollio 1.023±0.110 1.052±0.120 1.098±0.118 1.117±0.151

Stanlio 1.082±0.118 1.022±0.126 0.920±0.100 1.010±0.141


46

Fig. 3.6. Ratios Rcalc/ Rexp of calculated over measured responses for all spheres at a given calibration energy.

Table 3.8. Weighted mean of ratio of calculated over measured responses for all spheres.

Sphere <Rcalc /Rexp>

81 1.076 ± 0.066 81cd 1.022 ± 0.059 108 0.997 ± 0.058 133 0.983 ± 0.060 178 0.999 ± 0.058 233 1.043 ± 0.064

Ollio 1.066 ± 0.061 Stanlio 0.999 ± 0.059

<Rcalc /Rexp> 1.021 ± 0.021


47

In order to illustrate the behavior of each sphere at the energies of the present study,

the ratios (calculated/ experimental response) are plotted versus sphere in Fig. 3.7.

Fig. 3.7. Mean ratios <Rcalc/Rexp> of calculated over measured responses if all calibration energies for a given sphere are considered. The bars represent the total uncertainty. The ratio is the calibration factor of the 3He proportional counter as derived from the four calibration energies.


48

Fig. 3.8. Ratios Rcalc/Rexp (calculated over measured responses) for all calibration energies for each Bonner Sphere. The average ratio <Rcalc/Rexp> over all energies is also given.


49

The spread of the data points in Fig. 3.6- 3.8 does not show any systematic

tendency and is generally well described by their uncertainties. The somewhat larger

deviations for the smallest spheres at 19 MeV can be attributed to additional

experimental uncertainties on the fluence determinations which could not be corrected

and thus were not reflected in the evaluated uncertainties. This could not be a

systematic error because it would have appeared in the results for all spheres. It is also

likely to be an uncertainty linked to the behavior of the small spheres at high energies.

As known, the small Bonner spheres are mostly sensitive to thermal neutrons. In the

case of the calibration at 19 MeV the thermalization is incomplete for high-energy

incident neutrons. However, this results only in a small energy dependence of the

response ratio, as reported elsewhere [88].

There are other factors that could influence the response of the Bonner

spheres, thus also the ratio of the calculated over the measured response. A very

important factor is the density of the polyethylene. As it is reported in ref. [88], a

small variation in the density can lead to a large change in the response. The

magnitude of this effect varies with energy. Another influencing factor could be the

geometry of the 3He detector, or the gas pressure adopted in the MC calculations.

These characteristics are not exactly known and a slight deviation from their true

value may provoke a large uncertainty in the estimation of the response function of

the sphere. Figs 3.6- 3.8 show that in general the calculated responses, Rcalc(E), are in

very good agreement with the measurements, Rexp(E).


50

3.2 Calibration at Quasi- Monoenergetic Neutron Fields

3.2.1 Materials and methods The measurements were performed in cave Q that is shown in Fig. 3.9. The beam

radius was 2.78 cm as measured at a distance of 6.1 m from the target. The neutron

production reactions are given in Table 3.9.

Fig. 3.9. Experimental set up of the BSS in Cave Q at UCL.

Table 3.9. Main characteristics of the neutron production reactions of the quasi- monoenergetic beams.

Reaction 7Li(p,n)7Be

Proton energy Ep [MeV] 36.4 62.9

Energy loss of protons in target ∆Ep MeV] 3.7 1.4

Nominal neutron energy Eo [MeV] 32.9 60.6


51

A set of different neutron detectors was used for the characterization of the

neutron beam. The properties of these detectors can be found in ref. [75]. A short

overview is given below.

• A proton recoil telescope (PRT) was used for the measurement of the peak

fluence, i.e. the fluence of neutrons within the high-energy peak. This

instrument is based on the detection of recoil protons produced by neutrons

that are scattered off hydrogen nuclei at center- of- mass scattering angles

greater than 148o.

• A U-238 fission chamber (FC), consisting of a stack of ionization chambers.

Due to their large energy deposition, the fission products can be easily

discriminated from other events by setting a threshold in the pulse-height

spectrum.

• A NE102 transmission detector was permanently installed in the beam. This

detector has a fast (NE102_f) and a slow (NE102_s) component.

• A beam charge (Q) monitor. It was not suited for use as a high-precision beam

monitor because of a large leakage current varying in time.

All spectral neutron fluences were measured with a NE213 liquid scintillation

detector 102 mm in length and 51 mm in diameter, with the time of flight method

[76]. They are shown in Fig. 3.10. Of the five monitors, only the fast component of

NE102 was used for the normalization of the present data. The charge monitor and

the slow component of NE102 presented linearity problems, while the fission

chamber had poor statistics. Improvement of the charge measurement was not

attempted because of lack of access to the beam dump and its high radiation level.

The measurement of the spectral fluence at each calibration energy was essential

because the neutron beam is not monoenergetic.


52

Fig. 3.10. Spectral fluence measured by the NE213 scintillator for the 33 MeV and 60 MeV neutron beams.

To determine the response function of the BSS, each detector has to be

exposed to a broad parallel field, uniform over the whole sensitive area of the

instrument. The area of the BSS to be irradiated was larger than the cross section of

the beam, therefore a scanning procedure had to be applied in order to simulate a

uniform irradiation by a broad beam. For the purposes of the present work the PTB

scanning system which is based on a cross scanning movement was used (Fig. 3.11).


53

Fig. 3.11. The PTB scanning system used for the irradiation of the BSS at UCL.

The scan was achieved by continuously moving the monitor in front of the

beam, according to a law that is the superposition of two motions, perpendicular to

each other, as in the case of a Lissajous figure. A Lissajous figure is obtained by

combining two harmonic motions, orthogonal to each other and having different

frequencies, non-multiple and out of phase (Fig. 3.12). More details on the scanning

procedure are given in ref. [89].

Fig. 3.12. Lissajous figure.


54

Apart from the complete set of the BSS, the LINUS (Long Interval Neutron

Survey-meter) detector was also used in this calibration experiment. LINUS is a

neutron rem counter (a detector with high efficiency to thermal neutrons) of the

Andersson-Braun (A-B) type. The structure of the moderator-attenuator is modified in

such a way that the instrument has a response comparable to conventional monitors in

the low and medium energy range, i.e. from thermal to 7- 8 MeV, and much increased

at the higher energies. The specific moderator for LINUS consists of polyethylene,

borated plastic and a shell of Pb. The counter is a spherical 3He proportional counter

(3.2 cm active diameter, filled with 304 kPpa 3He and 101 kPa Kr). For more details

on the LINUS the reader can see references [90-93]. LINUS is a fully characterized

instrument and its configuration is very similar to the configuration of Ollio. For these

reasons and in order to inter-compare the responses of the two detectors,

measurements were performed with LINUS during the experiments at UCL.

3.2.2 Experiment and data analysis

The procedure followed for the calibration at UCL is different from the one

followed at PTB (see section 3.1.2). No shadow cone measurements were needed

because the contributions from neutrons scattered in the target, in the collimators and

in the air, are negligible. However, background measurements were performed for

each sphere. Another contribution of scattered radiation can be produced during the

scanning procedure from the sphere itself. For this purpose special measurements

were performed and it was found that this contribution does not exceed 2%.

The mean count rates varied between 35 cts/sec and 325 cts/sec for the

calibration measurement and between 7 counts/s and 45 counts/s for the background

measurements. No pile –up or dead time corrections had to be applied. Measurements

lasted between 150 sec and 1600 sec. The total number of counts for each

measurement provided good statistics.

The fluence and spectral fluence measurements were performed by the PTB

metrology team [94]. In contrast to the PTB monoenergetic neutron fields,

the 33 MeV and 60 MeV fields are quasi-monoenergetic fields [95]. The spectral


55

distributions of the neutron beams used are characterized by a high-energy peak and a

continuum that extends to low energies. The peak results from transitions to the

ground state and the first excited state in 7Be (Ex = 0.429 MeV), which are stable

against particle emission. The continuum is caused by breakup reactions and by

interactions of neutrons with the collimator, etc.

The low-energy cut-off of the spectral fluence measured with the scintillation

detector and the fission chamber is about 3 MeV. Below this energy the spectral

fluence is unknown. For the determination of the absolute fluence response of the

BSS the knowledge of this information is essential because the BSS responds to this

part of the spectra as well. Especially for the smaller Bonner spheres this response is

assumed to be higher in comparison to their response to the peak neutrons. A flat

extrapolation to thermal energies could be a good approximation to the real case

because of the absence of scattered neutrons in the cave, however a 1/E extrapolation

should also be considered.

In order to be able to extrapolate, an unfolding with the experimental data

taken with the BSS at the energies of 33 MeV and 60 MeV was performed. The

unfolding provided the spectral fluence at all energy intervals above zero. The

solution spectrum indicated that a constant extrapolation to lower energies seemed to

be justified. For the analysis of the irradiations of the Bonner spheres the spectral

fluence measured with the reference monitor NE102 should be used. The number of

peak neutrons N0 per unit monitor count Mf of the fast NE102 detector can be

determined with the following formulas:

(N0/Mf) = 2.78(10)⋅102 at 33 MeV (3.7)

and

(N0/Mf) = 1.72(14)⋅102 at 60 MeV (3.8)

The reference distance from the Li-target is 6 m. The fluence monitor data

communicated by PTB provide only the fluence in the peak. Therefore, the total

fluence was ‘reconstructed’ by using the ratio ‘peak/total’ from the spectra provided

by PTB (Fig. 3.10).


56

3.2.3 Results and discussion

For the calculation of the Bonner sphere absolute fluence response, two

methods can be used.

The first method has already been used in the past in similar cases [93]. It

involves Monte Carlo simulations for the determination of the absolute fluence

response of each Bonner sphere to the total spectrum and to the peak of the spectrum

only. Then the experimental absolute fluence response to monoenergetic neutrons of

33 MeV and 60 MeV can be derived by equation (3.9):

QQQ

QRR

R

RRR FLUKA

peakFLUKA

totalFLUKAtotal

totaltotalpeak

ΦΦΦ

Φ−×−= ΦΦ

Φ

ΦΦΦ

0

0,,

,

exp,exp

,exp

, )]([ (3.9)

where exp, peakRΦ is the experimental absolute fluence response to the high- energy peak

of the spectrum, exp,totalRΦ is the experimental absolute fluence response to the total

spectrum, FLUKApeakR ,Φ is the FLUKA [61-63] Monte Carlo calculated response to the high-

energy peak and FLUKAtotalR ,Φ is the FLUKA Monte Carlo calculated response to the total

spectrum. The coefficients Φ0/Q and Φ/Q are the neutron fluence in the high-energy

peak per unit charge in the primary beam and the total neutron fluence per unit charge

in the primary beam, respectively.

The second method is less direct than the first one: if the Monte Carlo

calculated response functions of the BSS are folded with the experimental spectral

fluences, they should reproduce the experimental reading (i.e, the counts) for each

detector. Then, if the ratios of the experimental counts over the theoretical ones is

close to unity within the statistical uncertainties, this means that the shape of the MC

calculated response functions is correct and therefore the response of the BSS at the

calibration energies should be close to the MC calculated values.

The first method is generally accepted as more precise in comparison to the

second one but the latter is less time-consuming. Therefore, in the context of this

thesis and because of a big delay in the communication of the monitor data from PTB,

it was decided to use the second method.


57

The first step was to re-adjust (‘re-bin’) the width of the energy bins of the

fluence spectra from PTB, so as to obtain the energy binning used in the FLUKA

Monte Carlo code. This was necessary in order to fold the spectra with the MC

calculated response functions of the BSS, since they were also calculated with the

FLUKA code [96]. The response functions involved in the study presented in this

chapter are the ones reported as ‘nominal’ in [45]. The reader can address to this

reference for more details and can find the response matrix in the Appendix B.

It was also necessary to extrapolate the fluence spectra to energies lower than

3.5 MeV and 5 MeV for the spectra of 33 MeV and 60 MeV respectively. As

mentioned in section 3.2.2, a justified extrapolation would be a flat one for both

energies. However it was decided to use several types of extrapolations for each

spectrum. This allowed a number of conclusions to be drawn that are presented later

in this chapter. The extrapolations that were used are summarised in Table 3.10.

Table 3.10. Extrapolations used for the fluence spectra of 33 and 60 MeV in order to use them in the unfolding program.

Extrapolation type Peak

Energy

[MeV] Flat 1/E

33 3.5 MeV 0 1 keV 0 10 keV 0 100 keV 0 500 keV 0 1 MeV 0

60 5 MeV 0 1 keV 0 100 keV 0 150 keV 0 220 keV 0 300 keV 0 500 keV 0 1 MeV 0

As seen from Fig. 3.10 (section 3.2.1), the cutoff energy E1 for the peak region

is 29.0 MeV for the 33 MeV while for the 60 MeV it is 54.5 MeV. Knowing the

partial contribution of each part, i.e. of the tail and of the high-energy peak to the total

and the number of neutrons in either of the two parts, one can easily reconstruct the

total number of neutrons in the whole spectrum. For both energies it was calculated

that the tail contributes about 50% to the total spectrum. It is obvious that for every

type of extrapolation, the relative contribution of the tail to the total spectrum

changes. The percentage of neutrons included in the extrapolated part over the number

of neutrons in the total spectrum, varies between 0.03% (for the 1/E extrapolation

between 1 keV and zero, for the 33 MeV) and 32% (for the 1/E extrapolation between

1 MeV and zero, for the 60 MeV).


58

In the next step, the different spectra were folded with the MC calculated

responses and the sum of the products for all energy bins was divided by the total

number of neutrons in the full spectrum. The results are presented in Tables 3.11 and

3.12 for all types of extrapolation.

Table 3.11. Results of folding the MC responses with different types of spectra, for the 33 MeV calibration energy. The values given are counts/neutron/cm2.

Type of extrapolation Sphere Flat

< 3.5 MeV 1/E

< 1 MeV 1/E

< 0.5 MeV 1/E

< 0.1 MeV 1/E

<10 keV 1/E

< 1 keV 81 0.093 0.799 0.503 0.182 0.101 0.094

81cd 0.110 0.751 0.481 0.190 0.117 0.110 108 0.324 1.018 0.719 0.406 0.331 0.324 133 0.594 1.078 0.863 0.647 0.598 0.594 178 1.033 1.097 1.060 1.035 1.033 1.033 233 1.338 1.084 1.187 1.305 1.335 1.337

Stanlio 0.264 0.740 0.541 0.325 0.270 0.265 Ollio 0.874 0.627 0.732 0.844 0.872 0.874

LINUS 0.377 0.279 0.320 0.365 0.376 0.377

Table 3.12 Results of folding the MC responses with different types of spectra, for the 60 MeV calibration energy. The values given are counts/neutron/cm2.


<5 MeV 1/E

< 1 MeV 1/E

<500 keV 1/E

<300 keV 1/E

<220 keV 1/E

<150 keV 1/E

<100 keV 1/E

<1 keV 81 0.051 0.476 0.280 0.193 0.155 0.122 0.097 0.051

81cd 0.068 0.455 0.276 0.197 0.163 0.132 0.110 0.068 108 0.192 0.626 0.422 0.332 0.294 0.261 0.236 0.192 133 0.366 0.693 0.535 0.468 0.439 0.415 0.397 0.366 178 0.677 0.778 0.725 0.704 0.696 0.689 0.684 0.677 233 0.933 0.854 0.888 0.905 0.912 0.919 0.924 0.933

Stanlio 0.309 0.583 0.457 0.401 0.377 0.355 0.339 0.309 Ollio 0.886 0.738 0.807 0.838 0.851 0.862 0.871 0.886

LINUS 0.346 0.293 0.317 0.328 0.333 0.337 0.340 0.346


59

The ratios of the experimental total counts over the peak fluence were

calculated for each BSS detector. The peak fluence was calculated by dividing the

peak number of neutrons with the field size used at each measurement. Then the total

fluence was calculated by the peak fluence multiplied by the factor (Total/Peak)

derived from the spectra. The reader should keep in mind that the factor (Total/Peak)

varies because the total fluence varies according to the type of extrapolation used.

From these ratios one can calculate the ratios Exp.counts/total fluence using the

following formula (3.10)

total

peak

peaktotal

countsExpcountsExpΦ

Φ⋅

Φ=

Φ.. (3.10)

The results are shown in Tables 3.13 and 3.14 for 33 MeV and 60 MeV

respectively.

Table 3.13. Ratios (Exp.counts/ Total fluence) for 33 MeV calculated according to formula (3.10). The values given are counts/neutron/cm2.

Exp. Total (33 MeV) Sphere Flat

< 3.5 MeV 1/E

<1 MeV 1/E

<0.5 MeV 1/E

<0.1 MeV 1/E

< 10 keV 1/E

< 1 keV 81 0.043 0.030 0.035 0.041 0.042 0.043

81cd 0.040 0.028 0.034 0.039 0.040 0.040 108 0.138 0.097 0.114 0.133 0.137 0.138 133 0.250 0.176 0.208 0.242 0.250 0.250 178 0.443 0.311 0.368 0.428 0.442 0.443 233 0.577 0.405 0.479 0.557 0.575 0.557

Stanlio 0.118 0.082 0.098 0.113 0.117 0.118 Ollio 0.422 0.296 0.351 0.407 0.421 0.422

LINUS 0.115 0.081 0.096 0.111 0.115 0.115


60

Table 3.14. Ratios (Exp.counts/ Total fluence) for 60 MeV calculated according to formula (3.10). The values given are counts/neutron/cm2.

Exp. Total (60 MeV) Sphere Flat

< 5 MeV 1/E

<1 MeV 1/E

<500 keV 1/E

<300 keV 1/E

< 220 keV 1/E

< 150 keV 1/E

< 100 keV 1/E

< 1 keV 81 0.100 0.082 0.090 0.094 0.096 0.097 0.098 0.100

81cd 0.100 0.082 0.091 0.094 0.096 0.097 0.098 0.100 108 0.156 0.129 0.142 0.148 0.150 0.152 0.154 0.156 133 0.260 0.214 0.236 0.245 0.249 0.253 0.256 0.260 178 0.407 0.335 0.369 0.384 0.390 0.396 0.400 0.407 233 0.539 0.444 0.489 0.509 0.517 0.525 0.530 0.539

Stanlio 0.228 0.188 0.207 0.215 0.219 0.222 0.224 0.228 Ollio 0.583 0.480 0.529 0.550 0.559 0.567 0.572 0.583

LINUS 0.146 0.120 0.132 0.138 0.140 0.142 0.143 0.146

The aim is to obtain the ratios of the experimental counts over the counts

calculated by the folding. These results are presented in Tables 3.15 and 3.16 for 33

MeV and 60 MeV respectively.

Table 3.15. Experimental counts / calculated counts for 33 MeV.


< 3.5 MeV 1/E

<1 MeV1/E

<0.5 MeV1/E

<0.1 MeV1/E

< 10 keV 1/E

< 1 keV 81 0.462 0.038 0.070 0.225 0.416 0.457

81cd 0.364 0.037 0.071 0.205 0.342 0.364 108 0.426 0.095 0.159 0.328 0.414 0.426 133 0.421 0.163 0.241 0.374 0.418 0.421 178 0.429 0.284 0.347 0.414 0.428 0.429 233 0.431 0.374 0.404 0.427 0.431 0.417

Stanlio 0.447 0.111 0.181 0.348 0.433 0.445 Ollio 0.483 0.472 0.480 0.482 0.483 0.483

LINUS 0.305 0.290 0.300 0.304 0.306 0.305


61

Table 3.16. Experimental counts / calculated counts for 60 MeV.


< 5 MeV1/E

<1 MeV 1/E

<500 keV1/E

<300 keV 1/E

<220 keV1/E

<150 keV 1/E

< 100 keV1/E

< 1 keV81 1.961 0.172 0.321 0.487 0.619 0.795 1.010 1.961

81cd 1.471 0.180 0.330 0.477 0.589 0.735 0.891 1.471 108 0.813 0.206 0.336 0.446 0.510 0.582 0.653 0.813 133 0.710 0.309 0.441 0.524 0.567 0.610 0.645 0.710 178 0.601 0.431 0.509 0.545 0.560 0.575 0.585 0.601 233 0.578 0.520 0.551 0.562 0.567 0.571 0.574 0.578

Stanlio 0.738 0.322 0.453 0.536 0.581 0.625 0.661 0.738 Ollio 0.658 0.650 0.656 0.656 0.657 0.658 0.657 0.658

LINUS 0.422 0.410 0.416 0.421 0.420 0.421 0.421 0.422

The data presented in Tables 3.15 and 3.16 are the final results of the method

that was followed for the analysis of the calibration experiment of the BSS. They are

plotted in Figs 3.13 and 3.14 respectively. The horizontal axis represents the upper

limit of the energy interval in which the extrapolation 1/E is done.

Fig. 3.13. Ratios Experimental counts / calculated counts for 33 MeV.


62

Fig. 3.14. Ratios Experimental counts / calculated counts for 60 MeV.

As shown in Tables 3.15 and 3.16 the ratios (experimental/calculated counts)

are far from unity. This big discrepancy was first attributed to a possible error in the

definition of the field size that was used in each measurement for the determination of

the total neutron fluence. This argument was discarded after a thorough investigation

of the experimental data. First, any change of the field size or of any other detail in the

set-up was recorded in the logbook, and second, data obtained for Ollio for different

field sizes gave consistent experimental results. In Table 3.17 the experimental data

for Ollio along with the results are shown.

Table 3.17. Extract from the logbook of the experimental data obtained for Ollio. Energy (MeV)

Field size (cm2)

Net Ollio counts

Peak fluence (1/ cm2)

Counts/peak fluence (cm2)

31 x 30 421592 425552 0.991 33 42 x40 192421 190489 1.010 31 x 30 502605 292573 1.718 60 42 x40 235042 135512 1.734


63

As seen from Table 3.17 the results for Ollio for both field sizes are consistent

within a statistical uncertainty of maximum 2%. This provides evidence of the correct

recording of the field size during the experiment. Therefore it was decided to

investigate other possible sources of error.

The possibility that the MC calculated response matrix of the BSS system is

not correct can be excluded. The reason is that the MC responses have been verified

by the two previous calibrations of the BSS at PTB and the agreement between the

experimental and the MC response functions is very good (see Figs. 3.17-3.20 at the

end of this chapter). There is no reason to believe that the MC simulations provide

correct results up to 20 MeV but not at 33 MeV and 60 MeV.

Taking into account that the experimental procedure did not include any other

source of error and that the monitor data provided by PTB are correct, it is clear that

the discrepancies observed are due to a factor that had not been noticed and hence not

taken into account up to now. The only possible source for the big discrepancies thus

seems to be a malfunction of the electronics and data acquisition system of the BSS.

The task of the electronics is to ensure the correct determination of the number

of neutron events produced in the 3He proportional counter. For that purpose, it is

recommended that the pulse height -spectrum (the spectrum of the proton recoils in

the 3He counter) obtained from the detector system be recorded. This has always been

done for all the calibration campaigns of the BSS (at PTB 2001, 2002 and at UCL

2003) since the same electronics is used. By comparing the shapes of pulse height

spectra obtained by the same electronics one can see that the UCL spectra seem

distorted. There is no definite plateau region while the thermal peak is also not well

defined. This may provide an indication of a malfunction of the data acquisition

system and may therefore explain the big discrepancies observed in the experimental

results. Examples of the spectra acquired in the several calibration experiments of the

BSS are shown in Fig. 3.15.


64

Fig. 3.15. Spectra from the PTB 2001 calibration (upper left), the PTB 2002 calibration (upper right), and the UCL 2003 calibration (lower left: a normal measurement, lower right: a background measurement).

A typical 3He pulse height spectrum is shown in Fig 3.5 (section 3.1.3). By the

comparison of that spectrum with the ones shown in Fig. 3.15 it was concluded that

the MAB data acquisition electronics box was not operating correctly during the UCL

calibration. To verify the problem with the MAB electronics, measurements with a

second set of electronics were performed long after the calibration. It was not possible

to make this test immediately after the calibration at UCL, because the reference

fluence data was communicated with a big delay and until the moment that this data

arrived and the analysis could be done, there was no evidence that the electronics was

not functioning properly. As soon as the analysis of the experimental data was

finished, it was obvious that there was a source of error. The mab box was then sent

for repair and only very recently it could be tested again in the calibration hall of the

SC-RP group.


65

The electronics used for this purpose, is the so called LINUS electronics. It is

based on NIM standard and it consists of an amplifier (Ortec 570), a single channel

analyser (Ortec 550A), a counter (Ortec 994) and the power supply assembled in a

portable NIM crate. This electronics are used with the LINUS that includes the same

type of 3He proportional counter as the Bonner spheres. The only difference is the

operating high voltage of the detector. Normally the comparison of the results

obtained with the two different electronics systems, should give the ratio unity. This

must be the case when the two systems are operating correctly. From the

measurements in the calibration hall, it was found that the ratio of the two systems is

approximately 0.3. This means that the mab box counted about three times less that it

would, if it worked properly. This test can only prove the initial hypothesis about the

malfunction of the electronics and cannot provide a correction factor to be applied on

the experimental data from UCL. This might have been possible if the tests had been

done before the shipping of the electronics to the MAB company or immediately after

the calibration.

From Figs 3.13 and 3.14 one can draw some additional interesting conclusions.

• The ratios of the experimental /calculated (convoluted) counts for each

Bonner Sphere vary as a function of the extrapolation used for the part of the

spectrum where the shape was not precisely defined. It is obvious that the

type of extrapolation affects seriously the sphere count-rate, especially for the

smaller ones. This is due to the fact that these spheres are more sensitive to

the thermal and low- energy neutrons and this is the energy region where the

extrapolation is done. From this behavior one can conclude that the Bonner

spheres respond correctly to the energy spectrum, but for reasons mentioned

above (failure of electronics) their correct experimental responses could not

be derived.

• From the same graphs it can be seen that the sensitivity curves of LINUS and

Ollio are parallel but not identical. This is expected because the 3He

proportional counters used inside them as active sensors, are not the same.

This fact can be considered as an additional proof of the correct behavior of

the spheres.

• The ratios calculated for the flat extrapolation at 33 MeV for each sphere are

more ‘centered’ around a certain value (0.45). This is not the case for all other


66

extrapolations where the ratios fluctuate randomly. This agreement indicates

that these ratios would be close to unity if the electronics had not failed. It

also provides evidence that the spectrum is flat below 3.5 MeV. Although the

poor energy resolution of the BSS does not allow a detailed description of the

spectral fluence to be given, it has been shown [97] that a well characterized

set of spheres allows the integral neutron fluence to be determined within 4%.

• For the same reason, for the 60 MeV the 1/E extrapolation between zero and

220 keV seems more correct than the flat extrapolation given in [95].

3.3 Combined results of the 2001 and 2002 calibration measurements - conclusions

The BSS was calibrated for the first time at PTB in 2001 with monoenergetic

neutrons with energy 0.144 MeV, 1.2 MeV, 5 MeV and 14.8 MeV. The full

description of the data analysis and the presentation of the experimental results are

given in [45]. Apart from the verification of the MC calculated response functions of

the BSS, the second task of the calibration experiments is the determination of the

calibration factor of the 2-atm 3He proportional counter used with the BSS. In Table

3.18 the experimental absolute fluence responses of both calibration experiments of

2001 and 2002 are listed. The MC calculated absolute fluence responses are given in

Table 3.19.

For the determination of the calibration factor of the 3He proportional counter

with the given set of spheres, the weighted ratios Rcalc/Rexp are needed. First, the ratios

of the calculated over the experimental fluence responses were calculated by simply

dividing the data given in Tables 3.18 and 3.19. The results are shown in Table 3.20.

Then, the mean weighted ratios over all energies for every sphere were calculated

(Table 3.21) and plotted in Fig. 3.16.

From Table 3.21 and Fig. 3.16 one can see that the weighted ratios for all

spheres are slightly higher than unity. This may indicate a systematic uncertainty that

is due to factors influencing the MC calculated responses. For example, the density of

the polyethylene used in the FLUKA MC simulations is a factor that can introduce an

uncertainty in the calculations. It has already been stressed [45] that the polyethylene

density is not the same for all spheres and for all the filler pieces used to match the


67

3He shape to the central hole in each sphere. It has been reported by other researchers

[87] that small uncertainties in the determination of the density can result in large

discrepancies in the MC calculations.

Another important parameter is the 3He detector’s geometry that was used in

the MC simulations. The geometry used in the calculations was not the precise one.

This was verified in 2003 when the technical drawings of the 3He proportional counter

were confidentially communicated by Centronics Ltd [98].


68

Table 3.18. Experimental absolute mean fluence responses from the calibration experiments in PTB (2001 & 2002).

Counts per unit neutron fluence [cm2] Sphere

0.144 MeV 0.565 MeV 1.2 MeV 2.5 MeV 5 MeV 8 MeV 14.8 MeV 19 MeV

81 1.199±0.033 0.735±0.033 0.421±0.008 0.264±0.020 0.119±0.008 0.092±0.005 0.041±0.002 0.028±0.003

81Cd 1.197±0.028 0.741±0.033 0.404±0.008 0.270±0.020 0.113±0.002 0.097±0.005 0.051±0.002 0.049±0.003

108 2.386±0.052 1.965±0.084 1.137±0.022 0.936±0.059 0.476±0.037 0.381±0.016 0.178±0.005 0.171±0.016

133 2.848±0.102 2.797±0.116 2.201±0.030 1.788±0.105 1.008±0.052 0.792±0.032 0.396±0.007 0.414±0.075

178 2.182±0.051 2.899±0.118 2.808±0.039 2.768±0.155 1.808±0.086 1.510±0.059 0.835±0.013 0.777±0.080

233 0.984±0.030 1.819±0.104 2.263±0.031 2.789±0.152 2.226±0.113 1.914±0.104 1.241±0.016 1.066±0.139

Stanlio 0.951±0.020 0.661±0.029 0.369±0.008 0.253±0.018 0.159±0.005 0.159±0.007 0.179±0.004 0.207±0.020

Ollio 0.192±0.006 0.591±0.024 0.909±0.012 1.237±0.067 1.085±0.039 0.974±0.038 0.816±0.010 0.786±0.071

Table 3.19. Calculated absolute fluence responses of the BSS at monoenergetic neutrons for the calibration energies.

Counts per unit neutron fluence [cm2] Sphere 0.144 MeV 0.565 MeV 1.2 MeV 2.5 MeV 5 MeV 8 MeV 14.8 MeV 19 MeV

81 1.306±0.149 0.886±0.089 0.524±0.094 0.302±0.030 0.144±0.029 0.083±0.008 0.043±0.008 0.036±0.004

81Cd 1.260±0.104 0.860±0.086 0.515±0.069 0.281±0.028 0.148±0.023 0.089±0.009 0.053±0.008 0.050±0.005

108 2.588±0.200 2.157±0.216 1.539±0.115 1.036±0.104 0.544±0.042 0.352±0.035 0.182±0.014 0.154±0.015

133 3.005±0.199 3.017±0.302 2.519±0.149 1.859±0.186 1.114±0.064 0.737±0.074 0.403±0.022 0.339±0.034

178 2.359±0.181 3.013±0.301 3.220±0.233 2.800±0.280 2.042±0.131 1.489±0.149 0.858±0.053 0.734±0.073

233 1.109±0.120 1.877±0.188 2.617±0.222 2.890±0.289 2.533±0.186 2.006±0.201 1.291±0.087 1.139±0.114

Stanlio 1.065±0.145 0.715±0.072 0.442±0.076 0.258±0.026 0.185±0.038 0.146±0.015 0.199±0.037 0.209±0.021

Ollio 0.214±0.014 0.605±0.061 1.085±0.065 1.301±0.130 1.263±0.076 1.069±0.107 0.896±0.057 0.878±0.088

Table 3.20. Ratios of calculated over measured (experimental) absolute fluence responses for all energies.

Rcalc/Rexp Sphere

0.144 MeV 0.565 MeV 1.2 MeV 2.5 MeV 5 MeV 8 MeV 14.8 MeV 19 MeV

81 1.089±0.108 1.205±0.132 1.245±0.145 1.144±0.144 1.210±0.175 0.902±0.101 1.049±0.183 1.286±0.192

81Cd 1.053±0.082 1.161±0.127 1.275±0.106 1.041±0.129 1.310±0.119 0.918±0.102 1.039±0.150 1.020±0.119

108 1.085±0.074 1.098±0.119 1.354±0.057 1.107±0.131 1.143±0.096 0.924±0.100 1.022±0.080 0.901±0.123

133 1.055±0.071 1.079±0.117 1.144±0.053 1.040±0.121 1.105±0.070 0.931±0.100 1.018±0.056 0.819±0.170

178 1.081±0.074 1.039±0.112 1.147±0.064 1.012±0.116 1.129±0.071 0.986±0.106 1.028±0.062 0.945±0.136

233 1.127±0.100 1.032±0.119 1.156±0.074 1.036±0.118 1.138±0.078 1.048±0.119 1.040±0.066 1.068±0.175

Stanlio 1.120±0.123 1.082±0.118 1.198±0.145 1.020±0.126 1.164±0.179 0.918±0.100 1.112±0.168 1.010±0.141

Ollio 1.115±0.065 1.024±0.110 1.194±0.051 1.052±0.120 1.164±0.060 1.098±0.118 1.098±0.059 1.117±0.151


69

Table 3.21. Weighted mean of ratios of calculated over experimental responses for all spheres.

Sphere <Rcalc /Rexp>

81 1.106±0.048 81cd 1.097±0.039 108 1.136±0.031 133 1.063±0.027 178 1.070±0.029 233 1.090±0.033

Ollio 1.132±0.026 Stanlio 1.056±0.046

Fig. 3.16. The weighted average ratios (Calculated response / Experimental response) for all energies and all spheres.

The calibration factor of the 2- atm 3He proportional counter is calculated as

the mean weighted factor for all spheres, as derived by the data of Table 3.21, and it

has the value:

The uncertainty 0.012 is the weighted average of all uncertainties shown in Table

3.20. The value of the χ2 indicates that the errors were not overestimated.

fc= 1.096 ± 0.012 with χ2=0.811


70

Fig. 3.17. MC calculated and experimental response functions for the spheres 81 mm and 81Cd.


71

Fig. 3.18. MC calculated and experimental response functions for the spheres 108 mm and 133 mm.


72

Fig. 3.19. MC calculated and experimental response functions for 178 mm and 233 mm.


73

Fig. 3.20. MC calculated and experimental response functions for Stanlio and Ollio.

74

CHAPTER 4

The response of the extended-range Bonner Sphere Spectrometer to charged hadrons

4.1 Introduction

A BSS can be used to measure neutron spectra both outside accelerator

shielding and from an unshielded target. In the former case the contamination of the

neutron field with other types of hadrons is usually comparatively small. On the

contrary, when measuring the neutron emission from an unshielded target bombarded

by a high-energy hadron beam, a large contribution of hadrons other than neutrons

may be present. These secondary hadrons may interact with the moderator and

generate neutrons, which are in turn detected by the 3He proportional counter. This

study investigates the importance of this effect when using an extended-range BSS in

a mixed high-energy radiation field. It also provides a response matrix to charged

hadrons along with an experimental verification.

4.2 BSS measurements at high-energy mixed fields

Measurements with the extended range BSS were performed at SPS (Super

Proton Synchrotron) at CERN of the neutron yield and spectral fluence from 50 mm

thick copper, silver and lead targets bombarded by a hadron beam (composed of about

80% positive pions and 20% protons) of 40 GeV/c. The neutron emission per incident

hadron on target was measured at 60 cm from the target in the angular range from 30º

to 135º with respect to the beam direction [99]. This experiment is described in

chapter 5.

CHAPTER 4. The response of the BSS to charged hadrons

75

The unfolding of the data obtained with the set of eight Bonner spheres was

performed with a code based on the GRAVEL formalism [54], using as pre-

information the neutron energy distribution calculated with the FLUKA Monte Carlo

(MC) code [61-63]. For the MC simulations a pencil beam of 40 GeV/c protons/pions

impinging on a 50 mm thick Ag, Cu and Pb target and the spectral fluence of the

generated secondary neutrons was scored in void spheres with diameter of 133 mm

(corresponding to that of the medium- size Bonner sphere). The irradiation locations

for the simulations were the same as in the experiment, i.e. at angles of 30°, 45°, 60°,

75°, 90°, 105°, 120°, 135°, with respect to the direction of the incoming proton/pion

beam. The full results of the experiment are given in chapter 5. In the present chapter

only the response of the BSS to charged hadrons is discussed.

As an example, the neutron spectral fluences at 30° resulting from unfolding of the

experimental data and from the simulations are shown in Fig. 4.1 for a silver target.

The experimental spectrum shows the two peaks predicted by the MC guess spectrum;

an isotropic evaporation component centred at 3 MeV and a high-energy peak situated

around 100- 150 MeV. However, the unfolded spectrum shows a much more

pronounced high-energy peak and a comparatively smaller evaporation peak with

respect to the distribution predicted by the MC simulations. Precise background

corrections were done but the repeated unfolding showed that the discrepancy

remained large.


76

10-6 1x10-5 1x10-4 10-3 10-2 10-1 100 101

0.0

1.0x10-5

2.0x10-5

3.0x10-5

4.0x10-5

5.0x10-5

6.0x10-5

7.0x10-5

8.0x10-5

9.0x10-5

E*M

(E) p

er p

rimar

y pa

rticl

e (c

m-2)

Guess Spectrum Experimental (uncorrected) Experimental (corrected)

Neutron energy (GeV)

Fig. 4.1. Neutron spectral fluences from 40 GeV/c positive hadrons (80%

pions + 20% protons) on a 50 mm thick silver target, at 30° with respect to beam

direction, resulting from the MC simulations and from the uncorrected and corrected

experimental data unfolding (see text).

This discrepancy was eventually attributed to the contribution of secondary

neutrons produced in the BSS detectors by charged hadrons coming from the target.

This contribution cannot be suppressed by background subtraction. The charged

hadrons (mainly charged pions and protons) undergo inelastic interactions in the

moderator of the spheres, especially in the lead shell that is a part of Stanlio’s and

Ollio’s configuration. They produce secondary neutrons which are subsequently

slowed-down and thermalized by the polyethylene. In this way the 3He detector shows

an enhanced count- rate without of course being possible to discriminate the unwanted

counts due to charged hadrons. The determination of the response functions of the

BSS to charged hadrons is discussed in this chapter. In order to verify this hypothesis,

two studies involving FLUKA Monte Carlo simulations were performed.


77

4.2.a Calculation of spectral fluences of the hadrons (first Monte Carlo study)

The main task of the first study was to calculate the spectral fluence of all

secondary hadrons. These calculations were performed using the same geometry as

for the estimation of the energy distribution of neutrons emitted from the target.

As an example the spectral fluences of neutrons and charged hadrons at 30°

are shown in Fig. 4.2. To save computing time, the calculations were only made for a

primary pencil beam of 40 GeV/c positive pions that was the dominant component of

the primary beam. It was found that the most intense components of the spectral

fluence at all angles are protons and charged pions. Positive and negative pions have

very similar energy distributions, with a peak at almost 1 GeV. Protons show a two-

peak distribution centred at about 200- 300 MeV and 500- 600 MeV.

The ratios of each of the response to charged hadrons and neutrons over the

response to the total (all charged hadrons + neutrons) are listed in Table 4.1 for the

silver target. Similar tables for the copper and the lead target are given in Table C13

in the Appendix C. The ratios of the neutron response to the total represent the

correction factors that have to be applied to the experimental data in order to suppress

the charged hadron contribution.

As shown in Table 4.1 the response of Stanlio and Ollio to charged hadrons is

quite similar and contributes for more than 50% of the 3He detector counts at 30°. The

correction for the contribution of charged hadrons was also applied at large angles

because as shown from Table 4.1, these particles are responsible for more than 6% of

the counts even at 135°.


78

Fig. 4.2. Spectral fluences of neutrons and charged hadrons at 30° from a 40 GeV/c beam of positive pions impinging on a 50 mm thick Ag target.

The correction factors of Table 4.1 (section 4.2) were applied to the

experimental data and a new unfolding was performed with the corrected data. The

corrected neutron spectral fluence at 30° is shown in Fig. 4.1. The agreement with the

simulated spectrum is now satisfactory.

The procedure adopted for estimating these correction factors is very time-consuming.

It consisted of ninety- six sets of simulations (8 angles × 4 particles × 3 number of

spheres) per each target material, with each set consisting of five individual runs, and

it is obviously specific to the present experimental conditions. On the other hand, if

the response matrix of the BSS to charged hadrons is known, the correction factors

can be determined directly by folding the response functions with the spectral fluence

of charged hadrons coming from the target. Ideally one should be able to determine

the spectral fluence or at least the yield of neutrons and charged hadrons from the

experimental data, the response matrices of the BSS to the various particles and some

pre- information on the relative importance of the various contributions. The

calculation of the response matrix of the BSS to charged hadrons is discussed in the

next section.


79

Table 4.1. Ratios of the response of three Bonner spheres to p, n, π+ and π- to the total, for a 50 mm thick silver target (first MC study).

Angle Particle 233 mm Stanlio Ollio

neutron 0.902±0.002 0.439±0.001 0.456±0.001

π+ (2.572±0.011)x10-2 0.204±0.001 0.194±0.001

π- (5.688±0.019)x10-2 0.262±0.001 0.263±0.001

proton (1.537±0.004)x10-2 (9.427±0.005)x10-2 (8.700±0.014)x10-2

30°

π++π-+ proton (9.797±0.031)x10-2 0.561±0.001 0.544±0.001

neutron 0.951±0.001 0.638±0.001 0.653±0.001

π+ (1.023±0.005)x10-2 0.117±0.001 0.109±0.001

π- (3.085±0.006)x10-2 0.182±0.001 0.183±0.001

proton (7.879±0.063)x10-3 (6.255±0.005)x10-2 (5.410±0.010)x10-2

45°

π++π-+ proton (4.897±0.011)x10-2 0.362±0.001 0.347±0.001

neutron 0.970±0.001 0.766±0.001 0.778±0.001

π+ (5.049±0.017)x10-3 (6.570±0.006)x10-2 (6.042±0.011)x10-2

π- (1.994±0.003)x10-2 0.124±0.001 0.125±0.002

proton (4.909±0.023)x10-3 (4.449±0.004)x10-2 (3.674±0.010)x10-2

60°

π++π-+ proton (2.991±0.005)x10-2 0.234±0.001 0.222±0.001

neutron 0.979±0.002 0.840±0.001 0.848±0.002

π+ (2.807±0.014)x10-3 (3.919±0.003)x10-2 (3.560±0.006)x10-2

π- (1.438±0.003)x10-2 (8.829±0.007)x10-2 (8.999±0.017)x10-2

proton (3.442±0.017)x10-3 (3.262±0.004)x10-2 (2.630±0.007)x10-2

75°

π++π-+ proton (2.061±0.004)x10-2 0.160±0.001 0.152±0.002

neutron 0.985±0.002 0.885±0.001 0.890±0.001

π+ (1.766±0.008)x10-3 (2.609±0.002)x10-2 (2.359±0.003)x10-2

π- (1.114±0.002)x10-2 (6.771±0.008)x10-2 (6.941±0.008)x10-2

proton (2.214±0.011)x10-3 (2.188±0.002)x10-2 (1.731±0.004)x10-2

90°

π++π-+ proton (1.512±0.003)x10-2 0.115±0.001 0.110±0.001

neutron 0.988±0.002 0.910±0.001 0.914±0.002

π+ (1.275±0.007)x10-3 (1.927±0.002)x10-2 (1.743±0.003)x10-2

π- (9.589±0.018)x10-3 (5.640±0.004)x10-2 (5.820±0.010)x10-2

proton (1.396±0.008)x10-3 (1.377±0.002)x10-2 (1.042±0.004)x10-2

105°

π++π-+ proton (1.226±0.002)x10-2 (8.944±0.009)x10-2 (8.605±0.014)x10-2

neutron 0.990±0.002 0.927±0.001 0.929±0.002

π+ (9.328±0.041)x10-4 (1.480±0.001)x10-2 (1.314±0.003)x10-2

π- (8.492±0.018)x10-3 (4.955±0.004)x10-2 (5.126±0.007)x10-2

proton (8.972±0.062)x10-4 (8.697±0.001)x10-3 (6.469±0.018)x10-3

120°

π++π-+ proton (1.032±0.002)x10-2 (7.305±0.006)x10-2 (7.086±0.009)x10-2

neutron 0.991±0.002 0.936±0.001 0.938±0.002

π+ (7.991±0.038)x10-4 (1.287±0.001)x10-2 (1.139±0.002)x10-2

π- (7.591±0.015)x10-3 (4.468±0.004)x10-2 (4.643±0.007)x10-2

proton (6.055±0.043)x10-4 (6.094±0.008)x10-3 (4.448±0.018)x10-3

135°

π++π-+ proton (8.996±0.019)x10-3 (6.364±0.006)x10-2 (6.227±0.009)x10-2


80

4.2.b Response functions of the BSS to charged hadrons (second Monte Carlo study)

A second set of simulations was performed to estimate the contribution to the 3He count-rate of the secondary neutrons generated by the charged hadrons in the

moderator. Stanlio and Ollio were first investigated because they include a shell of

lead in their design. The two spheres were assumed to be irradiated by parallel

monoenergetic beams of neutrons, protons and charged pions, with energies between

50 MeV and 150 GeV (see Tables in the Appendix C). For each energy, five

independent simulations were performed in order to achieve good statistics. The

statistical error in all cases did not exceed 5%. The results are shown in Figs. 4.3-4.5.

The response functions to charged hadrons were calculated with the FLUKA

code for the complete BSS, for monoenergetic broad parallel beams of protons,

positive and negative pions with energy in the range 50 MeV– 150 GeV. For each

particle the neutron spectral fluence was scored in the 3He detector and subsequently

folded with the 3He(n,p)3H cross sections [100] to evaluate the response (see section

4.3). The same calculations were repeated for the 233 mm sphere. This sphere showed

a definitely smaller response to charged hadrons (about 10% at 30° and below 1% at

135°), confirming that this effect is far less important for polyethylene. Following this

result it was decided not to investigate the response to charged hadrons of the other

polyethylene spheres of smaller diameter, since their response would be negligible.

The ratios of the response to charged hadrons and neutrons to the total (all charged

hadrons + neutrons) are listed in Table 4.1 (section 4.2.a) for the silver target.

For the various detectors, the response functions to protons and positive pions

is quite similar, with a rapid increase up to about 100 MeV followed by a flat trend.

This is due to the fact that above this energy the inelastic cross sections tend to the

geometrical value. As expected, Ollio and Stanlio have a much higher response with

respect to the polyethylene spheres. Negative pions show for most spheres an

increased response below about 100 MeV. This can be because some π− are slowed

down to rest inside the sphere, they are captured and produce spallation reactions.

This effect is expected to be larger for larger spheres. As it can also be noted from the

graphs, the addition of a 1 mm cadmium shell to the 81 mm sphere increases its

response to charged hadrons by a factor of 5 to 10.


81

102 103 104 105

10-3

10-2

10-1

100

101

81 81cd 108 133 178 233 stanlio ollio

Res

pons

e [c

m2 ]

Energy [MeV]

Fig. 4.3. Fluence response of the extended BSS to protons.

These response functions were folded with the spectral fluences of neutrons

and charged hadrons calculated for the 40 GeV/c positive pion beam on a silver target

(Section 4.2.a), to verify the consistency with the former procedure used for

estimating the percent contribution of neutrons and charged hadrons to the total 3He

count-rate. For this purpose the response functions of the BSS to neutrons, previously

calculated up to 2 GeV had to be extended to 40 GeV. The results of such a

comparison are given in a next section.


82

102 103 104 105

10-3

10-2

10-1

100

101

81 81cd 108 133 178 233 stanlio ollio

Res

pons

e [c

m2 ]

Energy [MeV]

Fig. 4.4. Fluence response of the extended BSS to positive pions.

102 103 104 10510-3

10-2

10-1

100

101

81 81cd 108 133 178 233 stanlio ollio

Res

pons

e [c

m2 ]

Energy [MeV]

Fig. 4.5 Fluence response of the extended BSS to negative pions.


83

4.3 Comparison of the Monte Carlo studies and estimation of correction factor

Two methods were used to determine the correction factor to be applied to the

experimental results. They can be summarised as follows:

• Simulation of parallel beam irradiation by neutrons, protons, π+ and π− with

source spectra obtained by scoring yields at various angles in a previous

simulation of a target yield experiment. The correction factor for each angular

interval is obtained as the ratio of the neutron response to the sum of the four

responses to neutrons, protons, π+ and π− (Table 4.1, section 4.2.a).

• Calculation of energy response functions for parallel monoenergetic beams of

protons, π+ and π− and when not already done, neutrons (see section 4.2.b).

Folding these response functions with the spectra obtained at various angles in

the same previous target yield simulation as before, one obtains the

contribution of each component. The correction factor for each angular interval

is obtained as the ratio of the neutron contribution to the sum of the four

contributions, neutrons, protons, π+ and π- (Table 4.2).

The ratios of the results obtained with the two methods are listed in Table 4.2.

The agreement is satisfactory for the 233 mm sphere and for Ollio. For Stanlio the

agreement is satisfactory for neutrons, whilst data for charged hadrons show

discrepancies up to a factor of 2. Although one is essentially interested in the

correction factors for neutrons, this discrepancy is not completely negligible. One

should also note that there are systematic trends in the discrepancy also for the other

two detectors (233 and Ollio), which cannot be explained as random fluctuations but

may rather point to a systematic error in one of the two procedures, possibly under-

sampling of particles in the tail of the spectra. It should also be mentioned that the two

procedures are based on the same calculated spectra, and therefore they are not

completely independent of each other.


84

Table 4.2. Ratios of the percent contribution by neutrons, π+, π- and p to the total 3He count-rate determined by the full MC simulations of the experimental set-up (section 4.2.a ) to that determined by folding the detector response function with the particle spectra (section 4.2.b) .

Angle Particle 233 mm Stanlio Ollio neutron 0.974 ± 0.061 0.713 ± 0.080 1.026 ± 0.058

π+ 1.342 ± 0.046 1.483 ± 0.043 0.982 ± 0.060 π- 1.319 ± 0.047 1.446 ± 0.043 0.987 ± 0.060

30°

proton 1.320 ± 0.047 1.440 ± 0.044 0.950 ± 0.062 neutron 0.988 ± 0.060 0.815 ± 0.071 1.020 ± 0.058

π+ 1.360 ± 0.046 1.699 ± 0.038 0.974 ± 0.060 π- 1.281 ± 0.048 1.659 ± 0.039 0.970 ± 0.061

45°

proton 1.358 ± 0.046 1.620 ± 0.040 0.916 ± 0.064 neutron 0.993 ± 0.060 0.881 ± 0.066 1.019 ± 0.058

π+ 1.390 ± 0.045 1.833 ± 0.036 0.966 ± 0.061 π- 1.252 ± 0.049 1.788 ± 0.037 0.944 ± 0.062

60°

proton 1.324 ± 0.047 1.728 ± 0.038 0.891 ± 0.065 neutron 0.995 ± 0.059 0.920 ± 0.064 1.017 ± 0.058

π+ 1.385 ± 0.045 1.897 ± 0.035 0.946 ± 0.062 π- 1.214 ± 0.050 1.848 ± 0.036 0.917 ± 0.064

75°

proton 1.350 ± 0.046 1.784 ± 0.037 0.871 ± 0.067 neutron 0.997 ± 0.059 0.943 ± 0.062 1.015 ± 0.058

π+ 1.375 ± 0.045 1.941 ± 0.034 0.939 ± 0.062 π- 1.184 ± 0.051 1.888 ± 0.035 0.893 ± 0.065

90°

proton 1.346 ± 0.046 1.809 ± 0.036 0.854 ± 0.068 neutron 0.998 ± 0.059 0.955 ± 0.062 1.014 ± 0.058

π+ 1.385 ± 0.045 1.959 ± 0.034 0.932 ± 0.063 π- 1.161 ± 0.052 1.900 ± 0.035 0.868 ± 0.067

105°

proton 1.364 ± 0.046 1.824 ± 0.036 0.824 ± 0.070 neutron 0.999 ± 0.059 0.964 ± 0.061 1.012 ± 0.059

π+ 1.350 ± 0.046 1.969 ± 0.034 0.911 ± 0.064 π- 1.153 ± 0.052 1.908 ± 0.035 0.855 ± 0.068

120°

proton 1.379 ± 0.045 1.817 ± 0.036 0.812 ± 0.071 neutron 0.999 ± 0.059 0.968 ± 0.061 1.011 ± 0.059

π+ 1.356 ± 0.046 1.977 ± 0.034 0.909 ± 0.064 π- 1.141 ± 0.053 1.915 ± 0.035 0.853 ± 0.068

135°

proton 1.348 ± 0.046 1.819 ± 0.036 0.801 ± 0.072


85

4.4 Experimental test with 120 GeV/c hadrons at CERF

A real ‘calibration’ of the BSS response to charged hadrons would be ideal for

the verification of the response matrix to charged hadrons but it would require the

availability of broad beams of protons and charged pions of several defined energies

up to tens of GeV. Since this was not feasible, an experimental verification of the

response of the BSS to a monoenergetic beam of high-energy hadrons was performed

at the CERF facility at CERN [78].

Each detector of the BSS was exposed to a 120 GeV/c positive hadron beam.

The beam is composed of 1/3 protons and 2/3 pions and it shows a Gaussian profile

with FWHM of 30.5 mm and 31.7 mm in the horizontal and vertical planes,

respectively (measured with a multi-wire proportional chamber). The beam impinged

on each sphere at 25 mm from its centre because the irradiation of the spheres at their

centre provoked a very high count-rate that could damage the detector or result in

pile- up effects. MC simulations with the FLUKA code were performed reproducing

the exact experimental conditions.

The experimental data (counts per beam particle) are compared with the MC

results in Table 4.3. The agreement is rather good except for the two smaller spheres

where the experimental value is almost twice than the MC value. This may be due to

the fact that, given the dimension of the sphere and the beam size, the sphere was not

fully intercepting the beam. Even a small variation in the beam position and/or

dimension would cause a significant variation in the count-rate.

The good agreement between the MC predictions and the experimental data

give confidence in the Monte Carlo calculations of the BSS response functions.


86

Table 4.3. Counts per beam particle of the BSS detectors exposed to a narrow beam of 120 GeV/c protons/positive pions. The beam impinged on each sphere at 25 mm from its centre. The MC results are the weighted sum of the response to protons (1/3) and pions (2/3).

Sphere Monte Carlo result Experimental result

81 (4.95±0.04) x10-5 (7.81±0.78) x10-5

108 (2.24±0.02) x10-4 (3.41±0.34) x10-4

133 (4.30±0.04) x10-4 (3.88±0.39) x10-4

178 (8.97±0.10) x10-4 (1.16±0.12) x10-3

233 (1.31±0.01) x10-3 (1.69±0.17) x10-3

Stanlio (2.59±0.03) x10-2 (2.72±0.27) x10-2

Ollio (4.47±0.01) x10-2 (3.60±0.36) x10-2

4.5 Conclusions

The studies presented in this chapter show that the extended range BSS can be

used to measure around unshielded targets. An important condition that has to be met

is that the response functions of the lead-enriched detectors must be corrected for the

contribution due to charged hadrons. This fact has been verified by neutron spectral

measurements performed at CERN in a high-energy radiation field. Stanlio and Ollio

were found to have a significant response to the charged hadron component

accompanying the neutrons emitted from the target. The other detectors of the

extended-range BSS, that are composed only of polyethylene, showed a similar

behavior although of lesser importance.

The complete response matrix of the extended BSS to charged pions and

protons that was calculated with FLUKA, can be generally used whenever the BSS is

used to measure at fields where a large contribution of charged hadrons is present.

Some discrepancies for Stanlio have to be thoroughly studied, although they are not

associated with the corrections that need to be made for its response to hadrons.

The Monte Carlo calculated response functions were experimentally verified,

so this agreement can additionally be considered as another successful benchmark test

of the FLUKA code itself.

87

CHAPTER 5

Neutron spectral measurements with a Bonner sphere spectrometer

5.1 Introduction

This chapter presents experimental measurements of the neutron yield and energy

distribution from comparatively thin copper, silver and lead targets bombarded by high-

energy hadrons. The experiments performed at CERN were intended to provide source

term data for neutron production from high-energy hadrons, necessary for shielding

calculations and other radiation protection purposes.

Measurements of the spectral fluence and of the ambient dose equivalent of

secondary neutrons produced by 250 GeV/c protons and 158 GeV/c per nucleon lead ions

were performed around a thick beam dump a few years ago [101]. The results showed

that the spectral fluence of the secondary neutrons outside a thick shield is similar for

light (protons and pions) and heavy (lead) hadrons of comparable energy per nucleon

stopped in a massive target. It was also shown that the approach of considering a high-

energy lead ion as a group of independent protons is sufficiently accurate for the purpose

of evaluating the ambient dose equivalent of secondary neutrons outside a thick shield.

While in a massive dump the development of the hadronic cascade results in a

neutron yield Y which varies with energy E as 8.0EY ∝ , with a thin target the high-energy

cross-sections vary much slower and are roughly equal to their geometrical values [102].

Measurements of the neutron emission from comparatively thin copper and lead targets

bombarded by beams of high-energy protons/pions and lead ions have been discussed in

[103,104]. A comparison between a Monte Carlo simulation for protons and the

experimental results for lead ions has shown that the neutron yield for lead ions can be

reasonably predicted by scaling the result of a Monte Carlo calculation for protons by the

projectile mass number of the ion to the power of 0.85 to 0.95 for a lead target and 0.88 to

1.03 for a copper target.

CHAPTER 5. Neutron spectral measurements with the BSS

88

Following the above results, further measurements were conducted to extend these

data to different energies and target materials over a wider angular range. The present

chapter discusses measurements that were performed with the extended range Bonner

Sphere Spectrometer (BSS) and the experimental apparatus used also in [68].

5.2 Experiment

The neutron yield and spectral fluence from 50 mm thick copper, silver and lead

targets bombarded by a mixed beam of protons and pions (about 75% pions and 25%

protons) of 40 GeV/c momentum was measured at CERN in one of the secondary beam

lines of the Super Proton Synchrotron (SPS). The target thickness is about 1/3 of

interaction length, as the interaction length of 40 GeV/c protons and pions is 15 cm in Cu,

14.6 cm in Ag and 17 cm in Pb [105]. The neutron emission per incident particle on target

was measured in the angular range from 30º to 135º with respect to the beam direction.

The experimental data were compared with the results of Monte Carlo simulations

performed with the FLUKA code [61-63]. The simulations were also used as a priori

information for the unfolding of the experimental data.

The measurements were performed in the experimental area used by the NA57

[106,107] and ALICE [108] collaborations. The experimental apparatus used for the latest

measurements with lead ions was also used in this experiment. The equipment is provided

with remotely- controlled multiple target-holder and detector support, so that the

measurements could be conducted in a purely parasitic mode during the normal running

of NA57. Access to the experimental area was thus only needed to change the Bonner

detector every 32 measurements (8 angles x 4 target positions). The system was installed

in the same position as in the previous experiments, i.e. a few metres downstream of the

NA57 apparatus (see Fig. 5.1), on a concrete platform normally employed by the ALICE

collaboration. The zone is laterally and top shielded by 80 cm concrete.


89

Fig. 5.1. The NA-57 set –up and the exit pillars of the GOLIATH magnet.

The experimental device (Fig. 5.2) consists of a supporting frame (A) over which

a fan-shaped platform (B) slides horizontally, perpendicularly to the beam axis, to

position the target holder in and out of the beam. An arm, with the detector support at its

end, moves on the platform swinging around the target holder. The target to detector

distance is 600 mm, which permits a good angular resolution of the measurement yet

keeping the device compact. Four targets with diameter of up to 25 mm can be mounted

on the holder (Fig. 5.3), but one position is left empty for background measurements. The

holder moves vertically to position the selected target in the beam. The target holder has a

vertical excursion of 126 mm and a positioning accuracy of better than 0.1 mm. Its

position is determined by a rotative transducer fitted to a worm screw driven by a

stepping motor. A CsI scintillator coupled to a photodiode is mounted at the centre of the

holder. This scintillator is used as a reference to correctly align the targets in the beam.

The 150 mm horizontal excursion of the platform B is driven by a stepping motor coupled

to a worm screw, with a resolution of better than 0.1 mm. The speed can be adjusted

between 0.1 and 4 mm s-1.


90

Fig. 5.2. The experimental device used for the neutron spectrometry measurements, allowing the semiautomatic positioning of target and detector.

The detector arm moves over an angular range of 120º in 15º steps, from 15º to

135º with respect to the beam direction. The arm is driven by a stepping motor fitted

directly to a wheel (Fig. 5.3); the positioning accuracy is 0.05º.

The entire system is remotely controlled; the correct positioning of the detector

arm at the various angular locations is sensed by micro-switches and indicated on the

control panel by leds. The stepping motors for the three independent movements

(horizontal of the platform, vertical of the target holder and angular of the detector arm)

are driven by four-phase unipolar control boards of Eurocard format, operating at low

voltage and installed in an Europac standard rack close to the device. The transducer

signals are processed by boards installed in the same rack and sent to the control panel.

The actual positions of the platform, of the detector arm and of the target holder are

displayed on the control panel, from which the three movements are also controlled. The

correct operation of the equipment is additionally surveyed by a video camera.


91

Fig. 5.3. Left: detail of the driving mechanism of the swinging arm for detector positioning. Right: sketch of the multiple target holder.

The intensity of the primary beam was monitored by an air- filled Precision

Ionization Chamber (PIC) at atmospheric pressure, placed in the beam just upstream of

the target, connected to a current digitising circuit. One PIC-count corresponds to

2.45x104 ± 675 particles impinging on the target [109]. The extended-range neutron rem

counter LINUS [90-93] was placed in a fixed position on the measurement platform and

used as a reference monitor to verify the stability of the experimental conditions.

The spectrometer uses a Centronics 3He proportional counter connected via a

preamplifier to a MAB [47] acquisition system (amplifier, high voltage, multi channel

analyser housed in a single box). More details for the electronics are given in chapter 1 as

well as in [45].

For each target, measurements were performed in the angular range from 30º to

135º in steps of 15º. The count rate was sufficiently high, hence the statistical uncertainty

on the single measurements was always below a few percent. Measurements without

target were performed with each detector for background subtraction. The background is

generated by interaction of the beam with upstream components (although the overall

thickness of the material present in the beam path is small) and by radiation scattered

from the local shielding and the dump (situated approximately 20 m downstream of the


92

target). Interactions with air are negligible, as the interaction length for protons in air is

about 750 m (and about 2/3 of this value for pions [104]) and the air path upstream of the

target is about 20 m.

5.3 Monte Carlo simulations and spectrum unfolding

The Monte Carlo simulations were performed with the FLUKA code [61-63]. The

spectral fluence of secondary neutrons generated by 40 GeV/c protons striking 50 mm

thick copper, silver and lead targets was scored in void spheres with diameter 133 mm

(corresponding to that of one of the Bonner spheres used for the measurements) placed at

the same positions used in the experiment (i.e., at angles of 30°, 45°, 60°, 75°, 90°, 105°,

120°, 135°).

Separate simulations were performed considering only the target and reproducing

strictly the geometry of the concrete structure which shields locally the experimental set-

up. This allowed a part of the contribution of the scattered neutrons to be estimated. A

detailed treatment of the diffused component would have requested the simulation of all

the interactions of the primary beam upstream of the target, the contribution of other

experiments running in parallel and of the beam dump placed about 20 m downstream of

the target. The complexity of the geometry of some structures upstream of the target

together with their large distance from the scoring positions are difficult to simulate and

make almost impossible to converge to sufficiently accurate results without using

variance reduction techniques. Some structures (such as the beam dump) were considered

in a set of simulations done in the past [103], showing that, as expected the local concrete

shield is responsible for most of the scattered neutrons contributing to the spectral fluence

at the scoring positions.

The calculated spectra were used as pre-information for unfolding the

experimental data. The statistical uncertainty of the simulations is well below 1%.

As in previous experiments [103,104], the unfolding of the data obtained with the

set of eight Bonner spheres was performed with a code based on the GRAVEL formalism

[54]. The code employs the MATLAB™ capabilities for the I/O of data and results and

for the graphical presentation of other relevant quantities (e.g. chi-square, total fluence,

dose, etc.) during the iterations.


93

The sensitivity of the unfolded spectra against the statistical fluctuations of the

experimental data was checked by randomly varying the experimental counts of the

Bonner spheres according to a normal distribution with a standard deviation of 3%.

5.4 Data analysis

The uncertainty on the experimental data is in the range 6- 9%. However, two

important corrections had to be applied to the experimental data: a normalization factor

required by the fact that the beam size was larger than the target cross-section, and a

correction factor on the counts of Ollio and Stanlio to account for the neutron production

in the lead shell of these two spheres by the charged hadrons coming from the target.

5.4.1 Beam normalization factor

Since the beam was larger than the neutron-producing target, the fraction of the

beam hitting the target, the surrounding material or not interacting, had to be estimated. In

order to correctly assess these values, a program was written [110] which calculates the

beam fractions for given beam profiles and offsets. The beam was considered as Gaussian

shaped with σhorizontal = 9.3 mm and σvertical = 7.8 mm. The cylindrical target of 10 mm

radius was mounted inside a plastic ring with a lateral thickness of 2 mm. This ring was in

turn embedded in a second ring made of aluminum, which had a lateral thickness of

3 mm.

As mentioned above, the beam was centred on target with the help of a

scintillator. Assuming that the beam centre can be aligned with the target axis with an

uncertainty of 2 mm, the beam distribution on the target and target holder is given in

Table 5.1. The maximum neutron contribution coming from the target holder can be

estimated by taking into account the nuclear interaction lengths of the various materials

constituting the targets and holder, the fractions of the beam hitting the various parts and

the thickness of the plastic and aluminum rings (3.9 mm) as compared to the target

(50 mm). For all targets, the neutron production in the holder is less than 1.5% of the


94

neutron production in the target and can therefore be neglected. The experimental results

were thus multiplied by a factor of 1/0.487 as deduced from Table 5.1.

It must be noted that the uncertainly in the calculation of the σhorizontal and σvertical is

of the order of 10 – 20%. A measurement of the beam size 20 m upstream of the

experimental position confirmed the computed values within 10%. In addition, the beam

distribution is not exactly Gaussian. By decreasing both σhorizontal and σvertical by 10% to

20%, the fraction of the beam hitting the target increases approximately in the same

proportion. This uncertainty in the correction to be applied may explain the systematic

error found in the comparison of the experimental results with the Monte Carlo

predictions (see section 5.5).

Table 5.1. Fraction of beam hitting the target, the components of the target holder or missing the system, assuming a precision of 2 mm for aligning the beam on-target [courtesy of Helmut Vincke, CERN].

5.4.2 BSS response to charged hadrons

The second important correction to be applied became manifest while performing

the unfolding of the experimental data. The experimental spectra showed the two peaks

predicted by the Monte Carlo guess spectrum, an isotropic evaporation component

centred at 3 MeV and a high-energy peak sitting around 100-150 MeV (Fig. 5.4).

However, the unfolded spectrum showed a much more pronounced high-energy peak and

a comparatively smaller evaporation peak with respect to the distribution predicted by the

MC simulations.

Component Beam fraction (%)

Target 48.68 ± 0.91

Plastic ring 12.94 ± 0.06

Aluminum ring 15.86 ± 0.07

Lost particles 22.52 ± 0.91


95

Fig. 5.4. Neutron spectral fluences from 40 GeV/c p/π+ on a 50 mm thick silver target, at 30° with respect to beam direction (from MC simulations and uncorrected experimental data unfolding).

This discrepancy can be explained by the contribution of secondary neutrons

produced in the BSS detectors by charged hadrons coming from the target, a contribution

which cannot be suppressed by background subtraction. These charged hadrons (mainly

charged pions and protons) undergo inelastic interactions in the moderator (mainly in the

lead shell of Stanlio and Ollio), producing secondary neutrons which are subsequently

slowed-down and thermalized by the polyethylene giving rise to additional counts in the 3He detector. The importance of this effect when using an extended-range BSS in a mixed

high-energy radiation field is thoroughly discussed in chapter 4 (see also ref. [111]).

A set of FLUKA simulations was performed to determine the correction factor.

First the spectral fluence of all secondary hadrons was calculated using the same

geometry that was implemented for estimating the energy distribution of neutrons and

charged hadrons emitted from the target. A second set of simulations was then performed

to estimate the contribution to the 3He count-rate of the secondary neutrons generated by

these charged hadrons in the moderator. The reader can find a table with the correction

factors for the response of the BSS to charged hadrons and a full description of this part

of the analysis in chapter 4.


96

5.5 Results and discussion

The neutron spectral fluences for the three targets are shown in Fig. 5.5- 5.7

at angles of 30°, 60°, 90° and 120°. The spectra show the two peaks predicted by the

Monte Carlo a priori information: an isotropic evaporation component at 3 MeV and a

high-energy peak around 100- 150 MeV. The valley between the evaporation and the

high-energy components is due to a minimum in the hadron-nucleon cross sections at

intermediate energies. It should be noted that the relative importance of the high-energy

peak over the evaporation peak is larger for the copper target. The same effect was

observed with a lead ion beam [104].

Fig. 5.5. Neutron spectral fluence [EΦ(E)] per primary hadron from 40 GeV/c p/π+ on a 50 mm thick copper target, at emission angles of 30°, 60°, 90° and 120°.


97

Fig. 5.6. Neutron spectral fluence [EΦ(E)] per primary hadron from 40 GeV/c p/π+ on a 50 mm thick silver target, at emission angles of 30°, 60°, 90° and 120°.

Fig. 5.7. Neutron spectral fluence [EΦ(E)] per primary hadron from 40 GeV/c p/π+ on a 50 mm thick lead target, at emission angles of 30°, 60°, 90° and 120°.


98

The neutron yields resulting from the unfolding of the experimental data are listed

in Tables 5.2 – 5.4. The results are given in four energy bins and as total yield, in the

angular range 30° – 135°. The total yield is compared with the results of the Monte Carlo

simulations. The simulations included the shielding around the target, so that

experimental and simulation data can be directly compared. The ratio between the

experimental results and the calculations is given in the last column of Tables 5.2 – 5.4.

Table 5.2. Neutron yield (per energy group and total yield) for 40 GeV/c p/ π+ on a 50 mm thick copper target. Data are neutrons per primary particle on target.

Angle

(o)

<100

keV

0.1-20

MeV

20-500

MeV

0.5-2

GeV

Total Yield

(experimental)

Total Yield

(Monte Carlo)

Exp./Calc.

Yield

30 0.056 0.42 0.26 4.7E-02 0.78 0.57 1.37

45 0.054 0.39 0.29 1.5E-02 0.74 0.50 1.48

60 0.041 0.37 0.27 3.4E-03 0.68 0.46 1.48

75 0.041 0.37 0.19 7.5E-04 0.60 0.41 1.46

90 0.038 0.35 0.18 3.9E-04 0.57 0.37 1.54

105 0.050 0.34 0.12 1.5E-04 0.51 0.34 1.50

120 0.048 0.33 0.12 1.4E-04 0.50 0.31 1.61

135 0.046 0.30 0.15 3.1E-04 0.50 0.29 1.72

Table 5.3. Neutron yield (per energy group and total yield) for 40 GeV/c p/ π+ on a 50 mm thick silver target. Data are neutrons per primary particle on target.

Angle

(o)

<100

keV

0.1-20

MeV

20-500

MeV

0.5-2

GeV

Total Yield

(experimental)

Total Yield

(Monte Carlo)

Exp./Calc.

Yield

30 0.091 0.67 0.41 7.4E-02 1.24 0.87 1.43

45 0.089 0.70 0.29 1.2E-02 1.09 0.81 1.35

60 0.078 0.69 0.27 2.3E-03 1.04 0.77 1.35

75 0.083 0.67 0.25 6.5E-04 1.0 0.72 1.39

90 0.086 0.64 0.19 2.4E-04 0.92 0.67 1.37

105 0.082 0.64 0.14 9.6E-05 0.87 0.62 1.40

120 0.080 0.62 0.16 1.5E-04 0.86 0.58 1.48

135 0.085 0.58 0.19 1.9E-04 0.85 0.53 1.60


99

Table 5.4. Neutron yield (per energy group and total yield) for 40 GeV/c p/ π+ on a 50 mm thick lead target. Data are neutrons per primary particle on target.

Angle

(o)

<100

keV

0.1-20

MeV

20-500

MeV

0.5-2

GeV

Total Yield

(experimental)

Total Yield

(Monte Carlo)

Exp./Calc.

Yield

30 0.14 1.29 0.50 8.7E-02 2.01 1.31 1.53

45 0.15 1.27 0.44 1.6E-02 1.87 1.25 1.50

60 0.14 1.26 0.39 2.8E-03 1.79 1.21 1.48

75 0.13 1.24 0.33 6.2E-04 1.70 1.16 1.47

90 0.14 1.20 0.30 2.2E-04 1.64 1.10 1.49

105 0.13 1.18 0.24 1.1E-04 1.56 1.04 1.50

120 0.13 1.14 0.26 1.4E-04 1.54 0.98 1.57

135 0.13 1.10 0.27 2.0E-04 1.50 0.92 1.63

From Tables 5.2- 5.4 one can observe a systematic discrepancy in the range 30% to

40% between experimental and simulation data. The discrepancy may be possibly due to

an uncertainty in the beam normalization factor as discussed in section 5.4.1. The ratios of

the experimental to the MC calculated yields are plotted in Fig. 5.8 for all targets.

Fig. 5.8. Ratio of experimental to simulation neutron yield data for the three targets in the angular range 30° – 135°.


100

For the copper target (Table 5.2 and Fig. 5.5) one notices that the high-energy peak (the

20 – 500 MeV bin) is larger at 60° than at 30°, contrary to what one would expect. A

similar trend is observed for all targets at angles of 105° to 135° (not shown in Figs. 5.5 –

5.7). On the other hand, for all targets the total yields regularly decrease with increasing

emission angle as expected (Tables 5.2 – 5.4). This unexpected behavior can be explained

by the influence of the statistical fluctuations of the experimental data on the unfolding

process.

The sensitivity of the unfolded spectra against the statistical fluctuations of the data

was checked by randomly varying the experimental counts of the Bonner spheres

according to a normal distribution with a standard deviation of 3%, as done in a previous

work [104]. It was found that the neutron fluences in large energy groups vary from 3%

up to 12% for each group. These variations may be considered as a lower limit of the total

uncertainties, which can be estimated as lower than 20% in each of these energy groups,

if all other sources of errors are taken into account. The total fluences are obviously more

stable, as shown in Tables 5.2 – 5.4. The variation of the yield per energy group versus

angle is shown in Fig. 5.9- 5.11.

Fig. 5.9. Neutron yield per energy group as a function of emission angle for 40 GeV/c p/ π+ on a 50 mm thick copper target.


101

Fig. 5.10. Neutron yield per energy group as a function of emission angle for 40 GeV/c p/ π+ on a 50 mm thick silver target.

Fig. 5.11. Neutron yield per energy group as a function of emission angle for 40 GeV/c p/ π+ on a 50 mm thick lead target.


102

The variation of the integral yield versus angle is shown in Fig. 5.12 for all targets.

Fig. 5.12. Neutron yield as a function of emission angle for 40 GeV/c p/ π+ on 50 mm thick copper, silver and lead targets. The lines are fits to the data according to the law Y =Y0 + α·exp(-θ/t) (expression (5.1)).

The contribution of the high-energy component at the forward angles is obvious

and results in an increase of the yield although it is not as pronounced as in the case of

heavy ions [104]. The neutron yield Y versus angle can be fitted by the expression:

Y = Y0 + α·exp (– θ/t) (5.1)

in which θ is the emission angle. The parameters of the fits Y0, α and t are given in

Table 5.5 for the three targets.

Table 5.5 Parameters of the fit (expression 5.1) for the total neutron yield as a function of emission angle for the three targets (Fig. 5.12).

Target Fit parameter

Cu Ag Pb Y0 0.35 ± 0.05 0.78 ± 0.05 1.36 ± 0.12 α 0.69 ± 0.02 0.79 ± 0.07 1.00 ± 0.06 t 77.11 ± 12.02 53.11 ± 12.79 69.26 ± 24.39


103

The variation of the integral neutron yield Y versus target mass number A is shown in Fig. 5.13.

Fig. 5.13. Neutron yield as a function of target mass number for 40 GeV/c protons/pions in the angular range 30° – 135°. The lines are fits to the data according to the law Y = a·Ab (expression (5.2)).

Fig. 5.13 shows that the yields increase by a factor in the range 2.5 – 3 from copper

to lead. This trend is explained by the partial development of the hadronic cascade in the

targets, both in the radial and in the longitudinal direction. The data at the eight angles

approximately follow the same trend and can be fitted by the expression:

Y = a·Ab (5.2)

The parameters of the fits are given in Table 5.6. They take into account a maximum 10%

uncertainty in the experimental data. The exponential b varies from 0.77 to 0.92, which is

not too far from the value one would theoretically expect from the expression of the

geometrical cross-section of a proton striking a target with mass number Atarget :

3/2arg0arg tettp Arπσ =→ (5.3)

where r0 = 1.3- 1.5x10-15 m.


104

Table 5.6. Parameters of the fit (expression 5.2) for the total neutron yield as a function of target mass number in the angular range 30o – 135o (Fig. 5.13).

Emission angle (o)

30 45 60 75 90 105 120 135

a 0.033±0.006 0.028±0.003 0.023±0.001 0.018±0.004 0.015±0.001 0.012±0.002 0.011±0.002 0.013±0.002

b 0.769±0.037 0.789±0.021 0.816±0.005 0.848±0.045 0.876±0.014 0.915±0.031 0.924±0.043 0.897±0.039

5. 6. Conclusions

The neutron yields and spectral fluences from 50 mm thick copper, silver and lead

targets bombarded by a mixed beam of protons and pions with momentum of 40 GeV/c

were measured in the angular range from 30º to 135º with respect to the beam direction.

These experiments are intended to provide source term data for neutron production from

high-energy heavy ions, necessary for shielding calculations and other radiation

protection purposes. Two peaks can be observed in the neutron spectral fluence,

according to the different steps of the intranuclear cascade model. The high- energy peak

at around 100- 150 MeV is due to direct hadron-nucleon reactions and pre- equilibrium

emission, while the peak at about 3 MeV comes from evaporation neutrons. It was found

that the relative importance of the high- energy peak over the evaporation peak is larger

for the copper target.

The experimental results are in good agreement with predictions obtained with the

FLUKA Monte Carlo code. The 30% to 40% discrepancy between experiment and

simulations can most likely be imputed to the uncertainty in the beam normalization

factor. It can be stated that the values of total yields given in Tables 5.2 – 5.4 lie in

between the experimental and the Monte Carlo data.The results complement previous

measurements of the neutron emission from comparatively thin targets bombarded by

beams of high-energy protons/pions and lead ions [103,104], as well as measurements of

neutron production from a massive dump [101].

105

CHAPTER 6

Upgrade and development of the CERN-EU Reference Field (CERF)

PART A: The beam monitoring at CERF

6.1 Monitoring system

6.1.1 Introduction

Two new instruments were installed in the beam line in the CERF irradiation

cave before the start of the June 2002 run at CERF. The first is a Multi-Wire

Proportional Chamber (MWPC) for accurate measurements of the profile of the

120 GeV/c hadron beam impinging on the copper target (Fig. 6.1). The second is a

5- litre ionization chamber of similar design as the reference CERF beam monitor (the

PIC), put in place as a back-up instrument to the PIC (Fig. 6.2). This chamber was

first submitted to extensive performance tests with 137Cs sources in a calibration

laboratory at CERN and later tested and inter-compared with the PIC in the CERF

hadron beam.

CHAPTER 6. Upgrade and development of CERF

106

Fig. 6.1. The wire chamber installed in the CERF cave.

Fig. 6.2. The BIG PIC (foreground) and the PIC (background) monitors in the CERF cave.


107

6.1.2 Measurements of beam profile with a Multi-Wire Proportional Chamber

A more reliable control of the beam position and profile had always been a

general claim by the CERF users. A solution was identified in a MWPC, which is one

of the standard monitors used to measure the beam profile in the SPS secondary beam

lines.

The MWPC was installed between the two CERF target positions,

approximately one metre upstream of the one below the concrete roof-shield. Its effect

on the beam is negligible, so it was decided that the MWPC be mounted on a static

support and stay in the beam during the whole CERF run period. To measure the

beam profile, the CERF copper target must obviously be removed from the upstream

support (the one under the iron roof-shield). It was verified (by tests during the beam

set-up) that backscattering from the target installed under the concrete roof-shield

does not affect the profile measurement, so in this case one can measure the beam

profile on-line while taking data. An example of the horizontal and vertical beam

profiles measured with the MWPC is shown in Fig. 6.3. Changing collimators C3 and

C5 from ± 11 mm to ± 14 mm modifies the rms width of the beam in the two

transverse dimensions from 10.4 mm (horizontally) and 9.1 mm (vertically) to

11.5 mm (horizontally) and 9 mm (vertically).

Fig. 6.3. Horizontal and vertical beam profiles measured with the MWPC installed one meter upstream of the copper target under the concrete roof-shield. The beam intensity was 3x107 particles per SPS pulse.


108

6.2 Performance tests of the BIG PIC

The Precision Ionization Chamber (PIC) [80], the primary beam monitor at

CERF, had until now no back-up instrument. Since all measurements at CERF are

normalized to unit beam particle incident on the copper target, the importance of the

PIC is apparent. Therefore, a second beam monitor was tested, installed in the facility

and inter-compared to the PIC, in order to have it characterized and ready for use in

case of a PIC failure in any future run.

This second device was built at CERN about 25 years ago, at the same time as

the standard PIC. It is a 4.9-litre effective volume open- air ionization chamber of the

same design as the PIC (i.e., of cylindrical shape), with identical external cross-

sectional area. Knowing the effective volume and the diameter of the chamber one

can calculate its effective length, which is 99.8 mm. This is 3.1 times the effective

length of the PIC, which is 32 mm [80]. For this reason this device is called here the

‘BIG PIC’. From this result it was expected that the sensitivity of the BIG PIC is

approximately three times the sensitivity of the PIC. This was later verified by the

experimental results discussed in section 6.3. The technical characteristics of the BIG

PIC are summarised in Table D1 in the Appendix D.

The device is connected to a high-voltage power supply and digitizer housed

in a small metal box. In this box, a small electronic device manufactured by

HAMAMATSU and referenced C4960-1 provides the high-voltage supply with a very

good stability and linearity. The voltage can be adjusted from 0 Volts to –1250 Volts

(in the following, all voltage values mentioned are negative polarity). Under normal

conditions the chamber is operated at 600 Volts. Two S-HVS connectors are used for

the power supply of the chamber and for the control of the output voltage. A digitizer,

with a sensitivity of 1 pC/digit, converts the input current to an output frequency.

There are 2 connectors LEMO 00, one for connecting the chamber and the other for

checking and recording the output frequency.

The BIG PIC was submitted to a number of performance tests similar to those

made in the past on the PIC [112,113]. Measurements were performed in the SC/RP

calibration laboratory to verify the region of ion saturation, the linearity of the

response versus intensity of the radiation field, the stability and the leakage current of

the chamber. The BIG PIC was then tested in the hadron beam at the CERF facility


109

during the CERF run, and its response was compared to that of the standard PIC.

Measurements of the instrument stability were repeated in the calibration laboratory

after the CERF runs.

All measurements in the SC/RP calibration laboratory were performed at the

nominal voltage of 600 V, except for the determination of ion saturation for which the

voltage was varied between 10 V and 1000 V. Air pressure and ambient temperature

were sufficiently stable throughout the measurements, hence no corrections had to be

applied. There were some variations in the humidity, but this parameter has only a

minor influence on the operation of the chamber [114].

6.2.1 Stability tests

To investigate the warming up time of the electronics, i.e. the time needed to

reach stable operating conditions, measurements were performed for seven hours after

switching on the power. Measurements were performed with 137Cs sources of

different activities providing air kerma rates varying between 10 µGy/h and

30 mGy/h. For the air kerma rates of 10 µGy/h up to 300 µGy/h the measuring time

was 1000 s for each measurement. The measurements with air kerma rates between

500 µGy/h and 3 mGy/h lasted 300 s each, while measurements at the highest air

kerma rates lasted 100 s each. The sources used, the values of air kerma rates and the

duration of the measurements are summarised in Table D2 in the Appendix D. The air

pressure, temperature and humidity in the calibration laboratory were recorded before

each measurement. The pressure and the temperature were fairly stable,

P = (968.7 ± 1.0) hPa and T = (19.8 ± 1.0) 0C, respectively. The humidity varied in

the range 43.3% to 73%, which translates in a variation of 2% on the reading of the

chamber [114].

This first stability test indicated that after several hours the chamber might still

not have reached stable operating conditions. The stability test was therefore repeated

over a longer period (72 hours). For this series of measurements only one source was

used, providing an air kerma rate of 3 mGy/h. The measuring time for the first four

measurements was 1000, 900, 300 and 360 s, respectively, and 300 s for all the other

measurements. The air pressure and temperature in the calibration laboratory were


110

recorded before each measurement and were found again stable, P = (960.3 ± 1.0) hPa

and T = (19.4 ± 1.0) 0C, respectively. The humidity varied in the range 53.4% to

72.3%. The results of the different series of measurements are plotted together in

Fig. 6.4. The slight difference between the values of air pressure in the two periods

corresponds to a variation in the sensitivity of the chamber of approximately 1%.

Since this is a minor correction, it was not accounted for. The results of the second

series of measurements alone are plotted in Fig. 6.5. The results of the two series of

measurements indicate that the BIG PIC needs almost ten hours to stabilize. For

comparison, the standard PIC needs about two hours to reach conditions of operation

[112].

Fig. 6.4. Stability test of the BIG PIC: sensitivity of the chamber as a function of time after switching it on. The error bars include the statistical uncertainties of the measurement, the 2% variation with humidity and the 3% uncertainty on the source output.


111

Fig. 6.5. Stability test of the BIG PIC: sensitivity of the chamber as a function of time after switching it on. The error bars include the statistical uncertainties of the measurement, the 2% variation with humidity and the 3% uncertainty on the source output. The line is only to guide the eye.

6.2.2 Linearity test

To test the linearity of the response of the BIG PIC, the chamber was

irradiated with 137Cs sources providing air kerma rates in the range 10 µGy/h to

30 mGy/h. Table D2 in the Appendix D lists the relevant parameters of the

irradiations (source identification, air kerma rates, number and duration of

measurements). The results, shown in Fig. 6.6, indicate a good linearity of the

response in the entire interval investigated. The parameters of the linear fit are given

on the plot. The chi-square (Χ2) divided by the degrees of freedom N-m (where N is

the number of data points and m is the number of parameters estimated by the fit) is

the reduced chi-square. The values of the reduced chi-square corresponding to the

probability Px(x2; N-m) of exceeding x2 versus the number of degrees of freedom are

tabulated in [115].


112

Fig. 6.6. Linearity test of the BIG PIC: measured count-rate (counts per second) versus air kerma rate. The line is a linear fit Y =A +B X to the experimental data, with Y = log10·y and X = log10·x.

6.2.3 Region of ion saturation

The region of ion saturation (i.e., the region in which there is no ion

recombination in the chamber), which determines the operating region of the

instrument, was evaluated with a series of measurements made with a 137Cs source.

The chamber was exposed to an air kerma rate of 30 mGy/h at a distance of 1.56 m

from the source. The voltage applied to the chamber was varied from 10 V to 1000 V

in steps of 50 V (except for the first two measurements, where the step was 20 V). For

each voltage setting three measurements were made of duration 20 to 60 s. The results

are listed in Table D3 in the Appendix D and are plotted in Fig. 6.7. The figure shows

that the nominal operating voltage of 600 V lies well within the region of ion

saturation for photon air kerma rates of up to 30 mGy/h, which is the maximum one

available in the calibration laboratory.


113

Fig. 6.7. Count-rate versus applied voltage, showing the region of ion saturation of the BIG PIC. The line is only to guide the eye.

6.2.4 Investigation of leakage current existence

Another test made on the BIG PIC was the investigation of leakage current.

A leakage in the capacity of the digitizer would modify the response of the chamber to

a pulsed radiation field like at CERF, where the beam comes in pulses lasting a few

seconds and spaced by about 13 seconds. To investigate this effect, the data from the

linearity test were divided by the air kerma rate and plotted as a function of kerma rate

in Fig. 6.8. The data and the fit seem to indicate the absence of leakage current in the

interval of air kerma rate investigated, in agreement with the results of the linearity

test shown in Fig. 6.6. However, in the region of very low air kerma rates (below

10 µGy/h) no conclusions can actually be drawn. To verify the absence of a small

leakage current, which would cause a deviation from linearity, additional

measurements at low kerma rates (< 10 µGy/h) would be needed, to simulate the

behavior of the chamber for very low intensities of the CERF hadron beam.


114

Fig. 6.8. Sensitivity versus air kerma rate of the BIG PIC. The line is a fit y = A, to the experimental data.

Long-term measurements also showed that the chamber did not give any

spurious counting in the course of the test. This means that the electronics of the

chamber is not affected by any noise.

6.3 Inter-comparison of PIC and BIG PIC monitors at CERF

The BIG PIC was tested at CERF during the June and July 2002 runs. The

chamber was installed in the beam about one metre downstream of the standard PIC.

Preliminary tests were carried out in June and measurements were performed in July,

to inter-compare the response of the two beam monitors and to verify the presence of

any recombination effect while operating the BIG PIC in a hadron beam.

In order to inter-compare the response of the BIG PIC to that of the PIC, the

readings of the two instruments were recorded over a number of SPS pulses and for

several beam intensities. The results are given in Table D4 in the Appendix D

and are plotted in Fig. 6.9. Since the distance traversed by the particles in the active


115

volume of the BIG PIC is 3.1 times the distance traversed in the standard PIC (see

section 6.2), one expects that the ratio of the readings of the two instruments is about

3, within an uncertainty that is ± 5% for the PIC [109] and can reasonably be assumed

of the same order for the BIG PIC.

From Table D4 (Appendix D) and Fig. 6.9 one sees that the ratio between the

readings of the BIG PIC and the PIC ranges from 3.09 at low beam intensities (200

PIC-counts per pulse) down to 2.95 for very high intensities. These variations are well

within the experimental uncertainties and in agreement with the above factor.

Nonetheless, the ratio seems to show a slightly decreasing trend with increasing

intensity of the hadron beam starting at about 6000 PIC/ pulse (Fig. 6.9).

Another inter-comparison of the two PIC monitors was performed in August

2003. For another verification of their ratio, new measurements were performed at

CERF with intensities varying from about 70 PIC-counts/pulse to 4,000 PIC/ pulse.

The raw data are shown in Table D5 in Appendix D. From this table one can see that

the ratio between the readings of the BIG PIC and the PIC ranges from 2.98 at low

beam intensity (75 PIC-counts/pulse) up to 3.02 at ~500 PIC-counts/pulse. Then it

slightly decreases again for higher beam intensities. These variations are well within

the experimental uncertainties and in agreement with the factor found in the previous

year.

Fig. 6.9. Ratio of the readings BIG PIC and PIC as a function of beam intensity. The lines are fits Y = A and Y = B + C·x to the experimental data.


116

6.3.1 Recombination effects for the BIG PIC

It has previously been demonstrated that the standard PIC is not subjected to

charge recombination when exposed to the CERF beam (120 GeV/c positive hadrons)

and that its response remains constant with increasing beam intensity [113]. The

above results may therefore suggest a small recombination effect at high intensity in

the BIG PIC. To verify this hypothesis, a series of measurements were performed to

obtain saturation curves of the chamber (i.e, voltage characteristics curves) for

different beam intensities.

The beam intensity was varied by adjusting collimators C3 and C5; all the

other collimators were left at their nominal CERF settings (i.e., collimators C1 and C2

set at ± 20, and C8 to C11 fully open). The beam intensity was varied between 200

and 14000 PIC-counts (of the standard PIC monitor) per pulse. These intensities

correspond to three times higher count- rates of the BIG PIC. Beam intensities and

fluctuations are summarised in Table D6 in the Appendix D.

For each setting of the collimators, data were taken by varying the voltage

applied to the BIG PIC from 10 V to 1000 V. The reading of the chamber was

corrected for slight beam intensity fluctuations by recording the number of primary

protons impinging on the T4 production target according to the expression:

collsetTVT

pulseppulsep

pulsecountsPICBIG //

/4

4⋅

− (6.1)

For each voltage setting, five readings of the BIG PIC per pulse and the

corresponding proton intensities on the T4 target were recorded. Expression (6.1)

represents the average of the count-rates (BIG PIC-counts/pulse) normalized to the

proton rate on T4 (pT4/pulse) for one voltage setting, multiplied by the mean proton

rate for each collimator setting.

The above quantity is plotted in Fig. 6.10 as a function of the voltage applied

to the BIG PIC, for various beam intensities. The uncertainties associated to the data

points range from 0.14% to 2.13% and are too small to appear in the plot. The graph

seems to confirm the existence of a slight ion recombination effect in the chamber for

beam intensities above about 6000 counts/pulse.


117

0 200 400 600 800 1000

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

3.5x104

4.0x104

4.5x104 14000 counts/pulse 12000 counts/pulse 10000 counts/pulse 8000 counts/pulse 6000 counts/pulse 4000 counts/pulse 2000 counts/pulse 1000 counts/pulse 450 counts/pulse 200 counts/pulse

BIG

PIC

cou

nts

corr

ecte

d fo

r T4

read

ing

Voltage (V)

Fig. 6.10. Voltage characteristic curves of the BIG PIC for different beam intensities at

CERF. The count-rates are corrected for slight beam intensity fluctuations. The lines are

sigmoidal fits to the data to guide the eye.

6.4 Tests of the Triggers 4, 5, 6 in the H6 beam line

The Trigger4 is one of the scintillators installed in the H6 beam line very close to the

PIC. It is routinely used for the verification of the calibration factor of the PIC and

subsequently of the BIG PIC. More precise calibration methods for the PIC were used

in the years 1998 and 1999 [109,116], but a routine check is done before each CERF

run. The characteristics of Trigger4 are given in [117].

The Trigger5 and Trigger6 are two scintillation counters located 15 m

downstream of the iron beam dump installed at the end of the CERF area. These

counters measure the muons produced by the pion component of the beam, and thus

provide an indirect measurement of the beam intensity. Therefore they can be useful

for a more precise estimation of the PIC response.


118

6.4.1 Efficiency measurements of Trigger4

A routine check of the PIC calibration factor is performed according to the

procedure described briefly in the following. The PIC-counts are read out online via a

LabView program running on a PC. A certain number of cycles are preset and the

PIC-counts are accumulated during the same time interval. The reading of Trigger4 is

received directly by the SPS beam-control program. The beam cycle lasts 16.8 s while

the pulse extraction lasts 4.8 s (August 2003). After selecting a number of beam

cycles, the Trigger4 counts the beam particles and after the last cycle the beam-

control program gives the average number of particles. An estimation of the

calibration factor is obtained from the ratio of the average number of particles to the

average PIC-counts.

First measurements showed a significant deviation from the known calibration

factor, i.e. 31,000 instead of the expected ~23,000 particles per PIC-count. The reason

of such discrepancy was an inappropriate high voltage (HV) of the photomultiplier of

Trigger4 (-1.85 kV). By varying the HV of the photomultiplier and recording the

counts of the scintillator, it was found that a good operating HV value (i.e., in the

middle of the so-called “plateau”) is -1.73 keV (Fig. 6.11).

1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

0.0

0.5

1.0

1.5

2.0

2.5

Even

ts p

er p

rimar

y pa

rticl

e [x

10-6]

HV of PMT of Trigger 4 [-kV] Fig. 6.11. Characteristic curve of the Trigger4 scintillator. Events on Trigger4, normalized to the protons on T4 production target, as a function of high-voltage (HV) of the photomultiplier (PMT).


119

With the HV set to the new value of -1.73 kV, the counts per pulse of Trigger4

were recorded for different apertures of collimators C3 and C5, and compared to the

counts per pulse of the PIC-counter (Table 6.1).

Table 6.1 Ratio of Trigger4-counts over PIC-counts for different beam intensities.

PIC-counts /SPS

pulse

(mean over 5 cycles)

Trigger4 response

per SPS pulse

(mean over 5 cycles)

Ratio

(Trigger4-counts

/ PIC-counts)

73.6 1,697,103 23,058

165.4 3,742,178 22,625

293.8 6,521,542 22,197

458 9,953,216 21,732

652.8 13,853,153 21,221

1,160.4 23,141,498 19,943

2,132.2 33,055,128 15,503

As noted in [117], Trigger4 is not reliable above 1700 PIC-counts per second

because in this region the limit of the photomultiplier is reached. Results in Table 6.1

show that at intensities around 300 PIC-counts per pulse it is already important to

correct the response of the scintillator for dead time losses.

In order to check if these data are consistent with the usual calibration factor

of 23,000, the Trigger4 response was corrected for dead time losses. When calculating

the effects of dead time, the entire detector system must be taken into account. There

are two models for dead time behavior: paralyzable and nonparalyzable response

[118]. The expressions of the true interaction rate n as a function of the recorded

count rate m and the system dead time τ are:

τnenm −= , paralyzable model (6.2)

and τn

nm+

=1

, nonparalyzable model (6.3)


120

Each element of a detector system usually has its own dead time, which can be

extendable (paralyzable model) or non extendable (non-paralyzable model) [119]. The

two models predict the same first-order losses and significantly differ for high true

event rates (i.e., higher than 1/τ, where τ is the system dead time). Since no

information about the detector components and their dead times was available, data

analysis was done with both models. It was assumed that they represented the two

extremes of the experimental set- up and the real set- up lies in between.

The data presented in Table 6.1 have been normalized to a pulse length of

4.8 s in order to obtain the recorded count rate m for Trigger4, the PIC-counts per

second and their ratio Γ. The dead time of the system and the ratio Γ have been

estimated by fitting the experimental data with expressions (6.2) and (6.3), as shown

in Fig. 6.12.

The mean value of the calibration factors obtained from the two fitting models,

results in the value Γ=23,640. This value is consistent within 3% with the calibration

factor in use of 23,000 particles per PIC-count.


121

0 50 100 150 200 250

0

1x106

2x106

3x106

4x106

5x106 Γ 23607.9 ±38.9τ 29 ns ±0.2

Res

pons

e Tr

igge

r4 (p

artic

les/

s)

Intensity (PIC-counts/s)

Paralyzable model

0 50 100 150 200 250

0

1x106

2x106

3x106

4x106

5x106 Γ 23672.1 ±66.8τ 32 ns ±0.6

Res

pons

e Tr

igge

r4 (p

artic

les/

s)

Intensity (PIC-counts/s)

Nonparalyzable model

Fig. 6.12. Measured count rate by Trigger4 as a function of the true count rate by the PIC (symbols). The solid line is a fit according to the paralyzable model (upper plot) and the nonparalyzable model (lower plot).

6.4.2 Measurements of Trigger5 and Trigger6

The response of the PIC was compared with the reading of the scintillation

counters Trigger5 and Trigger6, installed downstream of the CERF area. The Trigger5

and Trigger6 measure the muons produced by the pion component of the beam that is

roughly proportional to the beam intensity in the H6. The muon flux is about three

orders of magnitude lower than the beam intensity, thus it allows monitoring over a

wider intensity range than Trigger4 [116]. The response of the scintillation counters

versus PIC-counts is shown in Fig. 6.13 for different beam intensities.


122

0 500 1000 1500 2000 2500 3000 3500 40000

20

40

60

80

100

120

140

160

180

200

Trig

ger5

[x1,

000

muo

ns p

er s

pill]

PIC [counts per pulse]

Trigger5 Linear fit: y = 0.04111 x + 6.24887

0 500 1000 1500 2000 2500 3000 3500 40000

20

40

60

80

100

120

140

160

180

200

Trig

ger6

[X1,

000

muo

ns p

er s

pill]

PIC [counts per pulse]

Trigger6 Linear fit: y = 0.0373 x + 4.99083

Fig. 6.13. Response of Trigger5 (upper plot) and Trigger6 (lower plot) versus PIC-counts. The straight line is the fit in the range of linearity (squares).

The experimental points in Fig. 6.13 were fitted by a linear function in the non-

saturated range. The data points above 2,000 PIC-counts (about 105 muons) were

excluded from the fit because at these intensities the Trigger5 and Trigger6 start to

saturate. The calculation of the average of the linear fits shown in Fig. 6.13, results in

the expression:

T5/6 = (0.039205 ±0.0011) · P + 5.61985 ± 1.25. (6.4)

This expression gives the expected number of Trigger5/6-counts (T5/6) as a

function of the PIC-counts (P). In 1998 a similar expression was calculated in order to


123

correct the response of Trigger4 up to 2000 PIC-counts per pulse and may be

compared to the expression that was calculated in July 1998 [116] for correcting the

response of Trigger4 up to 2,000 PIC-counts per pulse:

T5/6 = 0.04078 · P + 4.57536 (6.5)

The equations (6.4) and (6.5) agree well within the experimental uncertainties.

6.5 Conclusions

The beam set- up at CERF has been made much easier by the MWPC installed

in the irradiation cave close to the copper target. The beam profile in the horizontal

and vertical planes can thus be measured on- line.

The BIG PIC seems slightly less performing than the standard CERF beam

monitor (the PIC): it shows a comparatively long warm- up time and a slight

recombination at high beam intensities. However, the present results seem to indicate

that its characteristics and performance are sufficient to establish it as a back-up

instrument to the primary CERF beam monitor, but some corrections for ion

recombination might have to be applied at high beam intensity.

Trigger4 has proved to be a reliable instrument to perform calibration tests for

intensities below 2,000 PIC-counts per pulse, if the appropriate corrections for dead-

time losses are made. Measurements taken during the CERF run are consistent with

the calibration factor in use (23,000 Trigger4-counts/PIC-counts), the correlation

factor between the PIC and BIG PIC (~3) and the expected Trigger5/6-counts per

PIC-counts within the experimental uncertainties.


124

PART B: Monte Carlo studies

6.6 Introduction

Under the term cosmic radiation one usually means the flux of energetic

particles that enter the Earth’s atmosphere from outer space [120]. Extra solar cosmic

rays, usually called Galactic Cosmic Rays (GCR), represent the dominant component

of the cosmic rays flux in the Earth vicinity for energies above ~ 10 MeV. These are

particles accelerated at distant sources, which propagate in the galaxy through

relatively weak magnetic fields and experience electromagnetic and nuclear

interactions with photons and nuclear matter in the interstellar medium. Approaching

the heliosphere they interact with the magnetic fields carried by the solar wind, that

effectively modify their flux up to the energy region of several GeV.

Over the past 10 years, there has been increasing concern about the exposure

of air crews to atmospheric cosmic radiation [121-123]. At aviation altitudes, the

neutron component of the secondary cosmic radiation contributes about half of the

dose equivalent, but until recently it has been difficult to accurately calculate or

measure the cosmic- ray neutron spectrum in the atmosphere to determine accurate

dosimetry [124-127]. Dose rates from atmospheric cosmic radiation at commercial

aviation altitudes are such that crews working on present day jet aircraft are an

occupationally exposed group with a relatively high average effective dose

[123,125,128]. Crews of future high-speed commercial aircraft flying at higher

altitudes would be more exposed [126].

The CERF particle composition and spectral fluences are similar to those in

the cosmic radiation field at commercial flight altitudes between 10 and 20 km

(Fig.6.14). A large fraction of the exposure of aircrew is due to neutrons with a

comparable energy spectrum to that found around high- energy hadron accelerators.

The CERF facility provides a radiation field for the inter-comparison and the

calibration of instrumentation that is used for in- flight measurements. Moreover, the

fast neutron component at CERF is considered to be more similar to that in space.

Therefore, interest has arisen from several International Institutes to study whether a


125

different shielding configuration at CERF could produce a radiation field, rich in

high-energy particles, as the one found inside a space station.

Fig. 6.14. Comparison of the energy spectrum at commercial flight altitude to the one on the concrete roof-shield at CERF [45].

The radiation field in low-Earth orbit consists of three main components:

1. GCR components trapped in the Earth’s magnetic field;

2. Neutrons, as secondary particles created in GCR interactions with spacecraft

materials, which could scatter back from the Earth (albedo neutrons);

3. Radiation coming from Solar Particle Events (SPEs).

The relative intensities of the components and the fluence rates of the different

particle types depend on the orbit altitude and inclination, on solar activity, and on the

occurrence of SPEs. Both inside and outside a spacecraft, the radiation field at a point

in tissue or in free air is affected by the shielding, scattering and generation of

secondary particles by surrounding material including an astronaut’s and/or

colleagues’ bodies. For the altitude range and inclination of a manned space station,

secondary neutrons can be a major contributor to effective dose, up to a few tens of

percent of effective dose rates. The exact proportion is very dependent on the amount


126

of shielding. Most of the neutrons inside the station result from nuclear interactions of

charged particles (mainly galactic protons and protons trapped by the earth's magnetic

field) with the wall material of the vessel.

Preliminary Monte-Carlo simulations were performed [45] with FLUKA in

order to investigate this possibility. In order to keep computing time reasonable,

simulations were based on a simplified spherical geometry (Fig. 6.15) rather than

modeling the complete facility. The aim was to understand whether a given

target/shielding combination and angular scoring region would indicate a promising

situation which could subsequently be further investigated. The preliminary studies

revealed that the real geometry of the facility has to be slightly modified and

implemented in the Monte Carlo calculations.

Fig. 6.15. Simplified geometry for the preliminary CERF studies.

The results of the simulations confirmed that there is no need of changing the

‘standard’ copper target or the present beam momentum of 120 GeV/c at CERF. The

new exposure area will be a high-intensity one, hence an additional back shield will

be needed; the combination of iron backed by concrete may be able to reduce the low-

energy neutron component. It was also shown that a 40 cm thick aluminum slab


127

placed at the forward direction fulfills satisfactorily the requirement for high

intensities, high energy particles and high neutron fluence.

6.7 Space studies for the CERF facility The most recent geometrical set- up of the CERF facility was implemented in

Monte Carlo FLUKA simulations with main focus on the new exposure area in the

forward direction (only the area under the concrete roof was modeled). The exact

shape of the external shielding was not implemented since it does not affect the

radiation field inside the cave. The geometry used for the simulations is shown in Figs

6.16- 6.18. An aluminum slab of 40 cm thickness 75 cm width and 2.3 m height was

added downstream of the Cu target. The slab was placed at the forward direction (30

degrees from the Cu target, with respect to the beam axis) at a distance 3.5 m from the

target. In addition, an iron slab of 40 cm thickness and of 240 cm height was added

behind the aluminum slab, attached to the existing concrete wall as a back shield for

the low energy neutron component that contributes significantly to the increase of the

dose equivalent. The fluence of neutrons, protons, pions, muons, photons and

electrons was scored behind the aluminum slab. The calculated spectra of all particles

are shown in Fig. 6.19.


128

Fig. 6.16. Top view of the modified CERF facility. The position of the Al and the Fe slab are indicated. The

external concrete walls on the left (with respect to the beam axis) and the roof have been excluded for clarity.

Fig. 6.17. Side view of the simulated CERF facility. The external concrete wall on the right and the roof

are not shown.


129

Fig. 6.18. Side view of the simulated CERF facility. The external concrete wall on the left and the roof

are not shown.

Fig. 6.19. Monte Carlo FLUKA calculated energy spectra of different particles at the modified CERF facility.


130

In Fig 6.19 the E·M(E)/dE known as fluence per unit lethargy is plotted versus

log(E). The lethargy representation is often the most appropriate, since spectra, when

plotted with a linear energy scale, very often show a steep negative slope at lower

energies. The proton spectrum in the forward direction (from 0o to 30o) is similar to

the energy spectrum of the galactic cosmic protons inside the MIR space station [129]

(see also Fig 6.20). In order to emphasize the similarity of experimental and

calculated spectra, a polynomial fit was done for the calculated spectra, that leads to

the comparison shown in Fig. 6.21. The uncertainty of the MC calculated data (up to

20%) and the uncertainties of the polynomial fit (10%) have been omitted for reasons

of clarity.

In order to compare computed results with experimental data many parameters

have to be taken into account, such as solar modulation, the inclination of orbit of the

space vessel, the type of the space vessel etc. For the present comparison, data taken

inside the module ‘Spektr’ of the MIR space station [130] during a solar minimum

and a solar maximum were used. The module which was added at the MIR complex

in 1995, contained equipment that was used for atmospheric research and surface

studies.

Fig. 6.20. FLUKA calculated and experimental proton spectra for minimum and maximum solar activity.


131

Fig. 6.21. Comparison of experimental proton spectra for minimum and maximum solar activity with a polynomial fit of the FLUKA calculated spectrum.

The FLUKA calculated proton spectrum at CERF, as shown in Fig. 6.21, is

more similar to the GCR spectrum as measured at MIR under conditions of minimum

solar activity, particularly for the energy range below 1 GeV. For higher energies the

curves converge for both conditions of minimum and maximum solar activity. The

explanation for this phenomenon may be that the low energy portion of the cosmic ray

spectrum is subject to strong variability due to solar modulation, following the 11 year

solar cycle. The modulation effects decrease with increasing particle rigidity 1 and

become negligible for particles with rigidities of a few GV, or approximately 5 GeV

per nucleon. During high solar activity the solar wind is stronger and so are the

magnetic fields transported by it, resulting in a decrease of the cosmic ray flux. The

contrary is true during low solar activity.

It must be stressed that the FLUKA calculated spectra had to be scaled in

order to be compared with the experimental data. The scaling factor corresponds to

1 The penetrating ability of a cosmic ray is quantified by the rigidity that it is expressed as the GCR’s momentum divided by its charge.


132

intensities at the CERF facility in the range of 600- 1100 PIC/ spill. The agreement

can be considered as satisfactory, judging especially from the shape of the two spectra

that extend over the same energy range.

It is also possible to find the measured neutron spectrum in the MIR station

[129] for minimum solar activity and compare it with the calculated neutron spectrum

at CERF. The comparison shown in Fig 6.22 indicates that there is a good agreement

between experimental and calculated spectra. This is even more important if one takes

into account the lack of statistical uncertainties in the experimental data.

Fig. 6.22. Comparison of experimental neutron spectra for minimum solar activity.


133

6.8 Conclusions

The modified geometry of CERF can provide a radiation field whose proton

and neutron components are very similar to the ones found inside a space station. For

the other radiation components no conclusions can be drawn since it was not possible

to find experimental data in the same energy range. The hadron beam at CERF

consists of protons, pions and kaons and it is known that the behavior of the pions is

similar to the protons’ at high energies. For this reason it can be concluded that the

CERF facility can be widely used for tests also for the pion component.

Aside from protons, neutrons and pions, the radiation environment inside a

space station is also composed of heavy ions which primarily originate from the GCR

and secondly produced by interactions of the GCR with the walls of the space vessel

or with other materials. At the CERF facility heavy ions are not present.

It can be concluded that the optional new exposure area at the CERF facility

may be very useful for the tests and inter-comparison of detectors used in space

stations or for other measurements in the framework of the space program.

134

CHAPTER 7

Conclusions

The present dissertation focused on the upgrade and development of the

extended- range Bonner Sphere Spectrometer (BSS). The BSS has gone under an

extensive calibration campaign that started in 2001 and was completed in 2003. It has

always been highlighted by many researchers that significant errors may be

introduced in spectroscopy performed by Bonner Spheres, by using incorrect response

functions. It is therefore vital to obtain the most appropriate response matrix. The only

way to accomplish this task is to validate the Monte Carlo calculated response

functions against calibration measurements.

The BSS has also been successfully used in measurements of neutron yields

and spectral fluences from unshielded targets. In addition, it has been calibrated at the

CERN- EU Reference Field facility (CERF), in the context of the upgrade of its

response to charged hadrons. CERF is a special irradiation facility that simulates the

radiation environment at commercial flight altitudes. The BSS has already been tested

in that field, several times in the past. A continued need for tests at CERF has been

expressed by many scientists worldwide. Hence, the improvement of the facility is

essential for its future existence and use. In consideration of this fact, extensive

performance tests were applied to an ionization chamber in order to certify its

capacity to serve as back- up to the reference monitor used at CERF. Moreover, the

potential development of the BSS towards its deployment for fluence measurements at

cosmic fields has arisen the need for Monte Carlo studies for the modification of the

facility.

The following conclusions can be drawn.

CHAPTER 7. Conclusions

135

Studies with the Extended– Range Bonner Sphere Spectrometer

1. The fluence response functions obtained from the analysis of the calibration

of the BSS at PTB (Germany) are in a very good agreement (within 20%)

with the Monte Carlo calculated fluence response functions. The calibration at

UCL (Belgium) verified that the Bonner spheres respond correctly to the

energy spectra (33 MeV and 60 MeV), but due to the electronics’ failure, the

correct experimental responses could not be derived.

2. The ratio of the calculated fluence response over the measured one is defined

as the calibration factor for the specific 2- atm 3He proportional counter and

its value is : fc= 1.096 ± 0.012.

3. The BSS can be used to measure around unshielded targets but the response

functions of the lead-enriched detectors must be corrected for the contribution

due to charged hadrons. This fact has been verified by neutron spectral

measurements performed at CERN in a high- energy radiation field. The other

detectors of the extended-range BSS that are composed only of polyethylene,

showed a similar behavior but of lesser importance.

4. The complete response matrix of the extended BSS to charged pions and

protons was calculated with FLUKA. It can be applied whenever the BSS is

used to measure at fields where a large contribution of charged hadrons is

present.

5. The neutron yields and spectral fluences from 50 mm thick copper, silver and

lead targets bombarded by a mixed beam of protons and pions with

momentum of 40 GeV/c were measured. Two peaks can be observed in the

neutron spectral fluence, according to the different steps of the intranuclear

cascade model. A high-energy peak at around 100 – 150 MeV, that is due to

direct hadron-nucleon reactions and pre-equilibrium emission, and a peak at

about 3 MeV that comes from evaporation neutrons. The experimental results

are in good agreement with predictions obtained with the FLUKA Monte

Carlo code.


136

The upgrade and development of the CERF facility

1. The beam set-up at CERF has been made much easier by the Multi Wire

Proportional Counter (MWPC) that is installed in the irradiation cave close to

the copper target. The beam profile in the horizontal and vertical planes can

thus be measured on-line. The ionization chamber (BIG PIC) reached a

sufficient level of performance so as to be established as a back-up instrument

to the primary CERF beam monitor. However, some corrections for ion

recombination might have to be applied at high beam intensity.

2. Trigger4 has proved to be reliable for performing calibration tests at

intensities below 2,000 PIC-counts per pulse, with the appropriate corrections

for dead-time losses.

3. The modified geometry of CERF can provide a radiation field whose proton

and neutron components are very similar to the ones found inside a space

station. Therefore, the facility can be very useful for future tests and inter-

comparison of detectors used in space stations.

Outlook

In the framework of this thesis it was shown that there have been significant

improvements to the extended BSS, in particular to the knowledge of its

characteristics and performance capabilities. The BSS is a reliable system that has

been successfully used in mixed radiation fields. This fact becomes more important if

one takes into account that one of the major difficulties in neutron spectrometry for

protection purposes is the very wide range of energies which need to be covered.

Hence, the ability of the BSS to operate reliably can be employed in future

measurements at experimental areas or for the control of the radiation monitors

around the Large Hadron Collider (LHC).

The Monte Carlo studies for the design of a new exposure area at the CERF

facility are very promising. The first results showed a good agreement between

calculated and experimental data. Since there is no other known facility where to

perform measurements on Earth with detectors that will be used in space, it might be


137

worth to upgrade the CERF facility by constructing this new area. In order to achieve

higher reliability further investigations will be needed. The new calculations will have

to be based on inputs that consider spectra of primary cosmic rays, other parameters

such as solar modulation or geomagnetic fields and atmospheric models. It might be

more interesting to include different atmospheric profiles, in order to reproduce the

expected conditions at different latitudes. Although these studies may not be easy to

implement, they could be valuable for the future development of the CERF facility.

Towards the extension of the use of the BSS in the region of cosmic ray energies, the

CERF facility may prove to be very useful for future tests and measurements.

APPENDIX A – Physical and Dosimetric quantities

138

A.1 Introduction

In 1990, the International Commission on Radiological Protection (ICRP)

[131] published the definition of two new quantities to be used for protection

purposes. These are the protection quantities equivalent dose, HT,R, and effective

dose, E. Since equivalent dose and effective dose are not directly measurable, the

International Commission on Radiation Units and Measurements (ICRU) [132]

defined the quantities ambient dose equivalent, H*(d), and directional dose

equivalent, H'(d, W), for area monitoring, and personal dose equivalent, HP(d), for

determining the dose to the individual. These quantities are measurable and are

termed operational quantities. ICRP introduced new energy and radiation type

dependent radiation weighting factors, wR, to account for differences in the biological

response associated with different radiation qualities (photons, neutrons, electrons and

heavy charged particles).

A.2 BASIC PHYSICAL QUANTITIES

Energy fluence, Φ the quotient of dN by da, where dN is the number of particles incident on an

elementary sphere of cross-sectional area da.

dadN=Φ

The most frequently used unit is the reciprocal square centimeter [cm-2].

Energy fluence rate The quotient of dΦ by dt, where dΦ is the increment of energy fluence in the time

interval dt.

dtdad

dtd

⋅Ν=Φ=Ψ

2

The unit is [cm-2·s-1].


139

Absorbed dose D is the energy absorbed per unit mass . It is defined as the quotient of the mean energy

imparted, εd , to matter of mass dm,

dmdD ε=

Its unit is the Joule per kilogram (J·kg-1) which is given the special name Gray (Gy).

Kerma The quantity kerma relates to kinetic energy of the charged particles, released in

matter by uncharged particles. It is the quotient of dEir by dm, where dEir is the sum of

the initial kinetic energies of all the charged particles liberated by uncharged particles

in the mass dm of material, thus:

dmirdΕ

=Κ

A.3. DOSIMETRIC QUANTITIES

A.3.1. Protection quantities The protection quantities give a measure of the risk of damage due to ionising

radiation and are suitable for defining exposure limits.

Organ absorbed dose

It is defined as the mean absorbed dose, DT, in a specified tissue or organ of the

human body, T, given by

∫ ⋅= dmDTmTD 1

where mT is the mass of the tissue or organ, and D is the absorbed dose in the mass

element dm. Unit: Gray [Gy].


140

Equivalent dose, HT, of an organ or tissue

It is defined as the absorbed dose averaged, DT,R , over the tissue or organ, T,

due to radiation R, multiplied by the relevant radiation weighting factor, wR , for

radiation, R. When the radiation field is composed of radiations with different values

wR , the absorbed dose is subdivided into blocks, each multiplied by its own value of

wR and summed to determine the total equivalent dose, i.e.,

RTDR RwTH ,∑=

Unit: Sievert [Sv], 1 Sv = 1 Jkg-1

The radiation weighting factor, wR, reflects the different radiobiological

effectiveness of the various types and energies of radiation. Table A1 provides the

values of radiation weighting factor used for radiological protection purposes as now

recommended by ICRU [133]. Table A.1. Radiation weighting factors according to ICRU 60.

Types and energy range of radiation Radiation weighting factor wR

Photons (all energies) 1

Electrons and muons (all energies) 1

Neutrons

< 10 keV 5

10-100 keV 10

> 100 keV- 2 MeV 20

>2 -20 MeV 10

>20 MeV 5

Protons (not recoil), energy > 2 MeV 5

Alpha particles, fission fragments, heavy nuclei 20


141

Effective dose, E, is a summation of the equivalent doses in tissue or organs, each multiplied by the

appropriate tissue weighting factor. It is given by the expression

RTDR Rw

T TwTHT TwE ,∑∑ ⋅=∑=

where HT is the equivalent dose in tissue or organ, T, and wT is the tissue weighting

factor for tissue, T, and accounts for the different susceptibilities of different organs to

radiation damage.

Unit: Sievert [Sv], 1 Sv = 1 J·kg-1

A.3.2. Operational quantities

Operational quantities were designed by the ICRU to provide appropriate

estimates of the protection quantities and to serve as calibration quantities for

dosimetric devices. As stated by the ICRP : ‘The probability of stochastic effects is

found to depend not only on the absorbed dose but also on the type and energy of the

radiation causing the dose’. Operational quantities are based on the dose equivalent

concept which combines the absorbed dose at the point of interest, D, and Q, the

quality factor at that point. Q is related to the type and energy of the radiation via the

unrestricted linear energy transfer, L, of charged particles in water.

Dose equivalent, H derives from the relation

∫ ⋅⋅=⋅= dLLDLQDQH )()( where DL is the distribution of dose D in linear energy L and Q(L) is the quality factor

as a function of L in water. Unit: Sievert [Sv], 1 Sv = 1 Jkg-1


142

Area monitoring

For area monitoring the phantom of definition is the ICRU sphere, a 30 cm

diameter sphere made of a four- element tissue- like material, with a mass density of

1 g/cm3 and a mass composition of 76.2% oxygen, 11.1% carbon, 10.1% hydrogen

and 2.6% nitrogen. In addition, certain mathematic conventions, namely expansion

and alignment of the actual radiation field, are included in the definitions of quantities

to be used for area monitoring.

Different dose equivalent quantities are defined for strongly and weakly

penetrating, radiation and for area and personal monitoring. Radiation is strongly

penetrating if the dose equivalent received by the terminative layer of the skin

(0.07 mm) at normal incidence to a broad radiation beam is lower than ten times the

effective dose. Radiation is weakly penetrating if for normal incidence the skin dose is

higher than ten times the effective dose. The area dose provides an estimate of the

effective dose that a person would receive if he stayed at a particular location while

the personal dose is a measure of the exposure of an individual to external radiation.

The human body influences the radiation field so they are not in general equivalent.

Area dose The area dose is the dose equivalent of soft tissue measured at a specific point

(Unit: [Sv]). For strongly penetrating radiation the ambient dose equivalent H*(10) is

used while for weakly penetrating radiation the relevant quantity is the directional

dose equivalent H’(0.07 Ω).

Ambient dose equivalent, H*(10), at a point of interest in a real radiation field is the dose equivalent that would be

produced by the corresponding expanded and aligned radiation field at a depth

of 10 mm in the ICRU sphere in the opposing direction to the aligned field.


143

The directional dose equivalent, H’(0.07 Ω), at a point of interest in a real radiation field is the dose equivalent that would be

produced by the corresponding expanded radiation field in the ICRU sphere at a depth

of 0.07 mm on a radius in a specified direction Ω. Often, the maximum value is used

as it is not known a priori what orientation a person will have in the radiation field.

Personal dose is the dose equivalent in soft tissue measured at a point on the body surface

representative of the radiation conditions prevailing (Unit: [Sv]). Again different

quantities are used for strongly and weakly penetrating radiation. The personal depth

dose equivalent, Hp(10), is the dose equivalent in ICRU soft tissue at a depth

of 10 mm in the body at the location where the personal dosimeter is worn.

Personal surface dose equivalent, Hp(0.07), is the dose equivalent in ICRU soft tissue at a depth of 0.07 mm in the body in the

location where the personal dosimeter is worn. These values may vary between

individuals and depend on the part of the body to which the dosimeter is attached.

APPENDIX B – Nominal Response functions of the BSS to neutrons

144

Nominal response functions of the extended range BSS. The group response is given.

E (MeV) 81 81cd 108 133 178 233 Stanlio Ollio 1.00E-11 4.14E-07 1.24E+00 4.32E-02 8.38E-01 5.56E-01 2.47E-01 8.47E-02 3.89E-02 3.22E-03 6.83E-07 2.50E+00 1.00E+00 1.71E+00 1.15E+00 5.05E-01 1.75E-01 9.92E-01 5.38E-03 1.13E-06 2.75E+00 2.46E+00 2.01E+00 1.34E+00 6.14E-01 1.97E-01 2.13E+00 6.33E-03 1.64E-06 3.05E+00 2.77E+00 2.23E+00 1.50E+00 6.67E-01 2.25E-01 2.52E+00 6.38E-03 2.38E-06 3.13E+00 3.15E+00 2.35E+00 1.58E+00 7.42E-01 2.47E-01 2.72E+00 6.55E-03 3.47E-06 3.19E+00 3.06E+00 2.46E+00 1.68E+00 7.76E-01 2.59E-01 2.91E+00 8.32E-03 5.04E-06 3.31E+00 3.10E+00 2.61E+00 1.87E+00 8.21E-01 2.83E-01 2.96E+00 6.58E-03 7.34E-06 3.31E+00 3.38E+00 2.72E+00 1.92E+00 8.53E-01 2.88E-01 2.89E+00 6.25E-03 1.07E-05 3.24E+00 3.24E+00 2.83E+00 1.98E+00 9.22E-01 3.05E-01 2.92E+00 8.23E-03 1.55E-05 3.28E+00 3.13E+00 2.93E+00 2.05E+00 8.99E-01 3.30E-01 2.56E+00 1.26E-02 2.26E-05 3.17E+00 3.06E+00 2.97E+00 2.08E+00 9.93E-01 3.45E-01 2.89E+00 1.25E-02 3.73E-05 3.19E+00 2.75E+00 2.97E+00 2.14E+00 1.04E+00 3.54E-01 2.70E+00 1.17E-02 6.14E-05 3.08E+00 3.13E+00 2.94E+00 2.26E+00 1.08E+00 3.75E-01 2.46E+00 1.12E-02 1.01E-04 3.19E+00 2.64E+00 3.00E+00 2.29E+00 1.15E+00 3.87E-01 2.22E+00 1.50E-02 1.67E-04 2.93E+00 2.70E+00 3.05E+00 2.37E+00 1.20E+00 4.08E-01 2.09E+00 1.54E-02 2.75E-04 2.84E+00 2.67E+00 3.11E+00 2.39E+00 1.17E+00 4.33E-01 2.30E+00 1.88E-02 4.54E-04 2.82E+00 2.58E+00 2.91E+00 2.35E+00 1.24E+00 4.37E-01 2.06E+00 2.52E-02 6.89E-04 2.62E+00 2.73E+00 3.02E+00 2.45E+00 1.32E+00 4.60E-01 2.22E+00 2.73E-02 1.04E-03 2.45E+00 2.40E+00 2.92E+00 2.53E+00 1.32E+00 4.80E-01 2.11E+00 3.08E-02 1.58E-03 2.36E+00 2.42E+00 2.95E+00 2.56E+00 1.35E+00 4.88E-01 1.92E+00 3.73E-02 2.31E-03 2.42E+00 2.16E+00 2.99E+00 2.54E+00 1.43E+00 5.22E-01 1.89E+00 3.81E-02 3.35E-03 2.33E+00 2.17E+00 2.88E+00 2.52E+00 1.42E+00 5.31E-01 1.88E+00 4.68E-02 4.88E-03 2.21E+00 2.04E+00 2.84E+00 2.62E+00 1.44E+00 5.39E-01 1.70E+00 5.03E-02 7.10E-03 2.16E+00 2.00E+00 2.81E+00 2.51E+00 1.47E+00 5.59E-01 1.87E+00 5.58E-02 1.03E-02 2.07E+00 1.94E+00 2.75E+00 2.49E+00 1.55E+00 5.83E-01 1.63E+00 5.43E-02 1.50E-02 1.96E+00 1.89E+00 2.73E+00 2.61E+00 1.56E+00 6.05E-01 1.48E+00 6.39E-02 2.19E-02 1.87E+00 1.82E+00 2.73E+00 2.57E+00 1.62E+00 6.26E-01 1.53E+00 7.68E-02 3.18E-02 1.85E+00 1.70E+00 2.69E+00 2.62E+00 1.63E+00 6.80E-01 1.42E+00 8.28E-02 5.25E-02 1.72E+00 1.61E+00 2.74E+00 2.66E+00 1.76E+00 7.54E-01 1.40E+00 9.64E-02 8.65E-02 1.70E+00 1.46E+00 2.64E+00 2.82E+00 1.85E+00 7.60E-01 1.26E+00 1.31E-01 1.23E-01 1.54E+00 1.45E+00 2.56E+00 2.86E+00 2.04E+00 8.96E-01 1.20E+00 1.65E-01 1.50E-01 1.47E+00 1.42E+00 2.51E+00 2.89E+00 2.14E+00 9.43E-01 1.26E+00 1.96E-01 1.83E-01 1.42E+00 1.30E+00 2.54E+00 2.90E+00 2.27E+00 1.03E+00 1.17E+00 2.14E-01 2.24E-01 1.33E+00 1.27E+00 2.51E+00 2.94E+00 2.29E+00 1.11E+00 1.15E+00 2.88E-01 2.73E-01 1.34E+00 1.21E+00 2.50E+00 2.95E+00 2.39E+00 1.22E+00 1.06E+00 3.27E-01 3.34E-01 1.21E+00 1.10E+00 2.32E+00 2.88E+00 2.64E+00 1.33E+00 9.68E-01 3.86E-01 4.08E-01 1.15E+00 1.09E+00 2.34E+00 2.98E+00 2.69E+00 1.48E+00 9.80E-01 4.58E-01 4.98E-01 1.07E+00 1.02E+00 2.23E+00 3.01E+00 2.76E+00 1.73E+00 8.96E-01 5.39E-01 6.08E-01 9.85E-01 9.06E-01 2.11E+00 2.93E+00 2.92E+00 1.78E+00 7.66E-01 6.25E-01 7.43E-01 8.91E-01 8.19E-01 1.97E+00 2.77E+00 2.90E+00 1.97E+00 6.59E-01 7.72E-01 8.21E-01 8.05E-01 7.38E-01 1.87E+00 2.77E+00 2.98E+00 2.17E+00 6.72E-01 8.43E-01 9.07E-01 7.56E-01 7.28E-01 1.85E+00 2.72E+00 3.06E+00 2.29E+00 5.90E-01 8.77E-01 1.00E+00 7.05E-01 7.18E-01 1.82E+00 2.73E+00 3.01E+00 2.35E+00 5.37E-01 9.24E-01

APPENDIX B – Nominal Response functions of the BSS to neutrons

145

Nominal response functions of the extended range BSS (continued).

E (MeV) 81 81cd 108 133 178 233 Stanlio Ollio

1.11E+00 6.78E-01 5.98E-01 1.69E+00 2.58E+00 3.15E+00 2.47E+00 5.66E-01 1.00E+00 1.22E+00 6.25E-01 6.12E-01 1.59E+00 2.53E+00 3.07E+00 2.42E+00 5.15E-01 1.05E+00 1.35E+00 5.87E-01 5.11E-01 1.44E+00 2.35E+00 3.07E+00 2.50E+00 5.03E-01 1.14E+00 1.50E+00 5.31E-01 4.91E-01 1.45E+00 2.32E+00 3.09E+00 2.68E+00 4.14E-01 1.19E+00 1.65E+00 5.07E-01 4.59E-01 1.38E+00 2.20E+00 3.09E+00 2.76E+00 4.26E-01 1.25E+00 1.83E+00 4.66E-01 4.16E-01 1.30E+00 2.22E+00 2.86E+00 2.66E+00 3.85E-01 1.27E+00 2.02E+00 4.22E-01 3.77E-01 1.21E+00 2.05E+00 2.85E+00 2.79E+00 3.63E-01 1.27E+00 2.23E+00 3.80E-01 3.59E-01 1.12E+00 1.90E+00 2.92E+00 2.71E+00 3.20E-01 1.28E+00 2.47E+00 3.67E-01 3.08E-01 1.10E+00 1.87E+00 2.68E+00 2.74E+00 3.01E-01 1.28E+00 2.73E+00 3.05E-01 3.06E-01 9.42E-01 1.65E+00 2.62E+00 2.82E+00 3.15E-01 1.31E+00 3.01E+00 3.04E-01 2.68E-01 9.00E-01 1.60E+00 2.41E+00 2.53E+00 2.63E-01 1.28E+00 3.33E+00 2.69E-01 2.44E-01 7.89E-01 1.50E+00 2.35E+00 2.59E+00 2.60E-01 1.26E+00 3.68E+00 2.32E-01 2.18E-01 7.51E-01 1.39E+00 2.25E+00 2.41E+00 2.50E-01 1.20E+00 4.07E+00 2.10E-01 1.98E-01 6.84E-01 1.28E+00 2.17E+00 2.38E+00 2.38E-01 1.21E+00 4.49E+00 1.98E-01 1.75E-01 6.76E-01 1.18E+00 2.08E+00 2.45E+00 3.69E-01 1.29E+00 4.97E+00 1.77E-01 1.75E-01 5.68E-01 1.12E+00 2.02E+00 2.43E+00 2.14E-01 1.27E+00 5.49E+00 1.62E-01 1.55E-01 5.20E-01 1.06E+00 1.88E+00 2.37E+00 2.31E-01 1.24E+00 6.07E+00 1.50E-01 1.44E-01 4.99E-01 9.74E-01 1.75E+00 2.35E+00 1.85E-01 1.20E+00 6.70E+00 1.34E-01 1.31E-01 4.44E-01 9.10E-01 1.70E+00 2.30E+00 1.94E-01 1.17E+00 7.41E+00 1.05E-01 1.04E-01 3.91E-01 8.17E-01 1.58E+00 2.11E+00 1.59E-01 1.17E+00 8.19E+00 9.76E-02 1.02E-01 3.61E-01 7.16E-01 1.42E+00 1.86E+00 1.72E-01 1.04E+00 9.05E+00 8.23E-02 8.65E-02 3.06E-01 6.35E-01 1.27E+00 1.76E+00 2.29E-01 1.00E+00 1.00E+01 7.06E-02 7.46E-02 2.63E-01 5.62E-01 1.08E+00 1.62E+00 2.14E-01 9.56E-01 1.11E+01 6.67E-02 7.28E-02 2.33E-01 5.00E-01 1.08E+00 1.57E+00 2.04E-01 9.54E-01 1.22E+01 6.20E-02 7.30E-02 2.23E-01 4.87E-01 1.00E+00 1.47E+00 2.11E-01 9.61E-01 1.35E+01 5.05E-02 6.30E-02 2.03E-01 4.18E-01 8.79E-01 1.28E+00 2.28E-01 9.02E-01 1.49E+01 5.12E-02 5.95E-02 1.93E-01 4.09E-01 8.32E-01 1.26E+00 2.26E-01 9.10E-01 1.75E+01 5.00E-02 6.17E-02 1.77E-01 3.96E-01 8.31E-01 1.21E+00 2.38E-01 9.01E-01 1.96E+01 3.83E-02 5.23E-02 1.47E-01 3.05E-01 6.44E-01 1.05E+00 2.35E-01 8.81E-01 2.10E+01 2.90E-02 4.66E-02 1.44E-01 3.06E-01 6.54E-01 1.03E+00 2.49E-01 9.18E-01 2.50E+01 1.77E-02 3.79E-02 1.28E-01 2.77E-01 6.04E-01 9.31E-01 3.01E-01 9.55E-01 3.50E+01 1.20E-02 3.01E-02 8.84E-02 1.96E-01 4.41E-01 7.02E-01 3.64E-01 9.56E-01 5.00E+01 7.91E-03 2.51E-02 5.38E-02 1.24E-01 2.89E-01 4.79E-01 4.01E-01 9.61E-01 7.50E+01 7.03E-03 2.34E-02 4.12E-02 9.31E-02 2.16E-01 3.60E-01 4.21E-01 9.72E-01 8.50E+01 6.46E-03 2.32E-02 3.59E-02 8.02E-02 1.83E-01 3.05E-01 4.30E-01 9.90E-01 1.00E+02 6.15E-03 2.29E-02 3.32E-02 7.36E-02 1.68E-01 2.86E-01 4.38E-01 1.02E+00 1.35E+02 5.72E-03 2.12E-02 2.99E-02 6.43E-02 1.52E-01 2.58E-01 4.54E-01 1.05E+00 1.75E+02 5.18E-03 2.06E-02 2.54E-02 5.78E-02 1.36E-01 2.33E-01 4.78E-01 1.10E+00 2.50E+02 4.94E-03 2.04E-02 2.29E-02 5.40E-02 1.24E-01 2.11E-01 5.15E-01 1.17E+00 3.00E+02 4.93E-03 2.04E-02 2.25E-02 5.04E-02 1.15E-01 1.95E-01 5.59E-01 1.25E+00 4.00E+02 4.67E-03 2.06E-02 2.18E-02 4.86E-02 1.14E-01 1.96E-01 6.18E-01 1.37E+00 5.00E+02 4.27E-03 2.10E-02 2.02E-02 4.73E-02 1.13E-01 1.96E-01 6.99E-01 1.54E+00 6.50E+02 4.18E-03 2.20E-02 1.91E-02 4.48E-02 1.09E-01 1.90E-01 7.85E-01 1.71E+00 7.50E+02 4.15E-03 2.31E-02 1.89E-02 4.29E-02 1.06E-01 1.87E-01 8.61E-01 1.87E+00 1.00E+03 4.02E-03 2.34E-02 1.87E-02 4.17E-02 1.03E-01 1.83E-01 9.38E-01 2.04E+00 1.50E+03 3.95E-03 2.42E-02 1.89E-02 4.14E-02 1.01E-01 1.80E-01 1.09E+00 2.36E+00 2.00E+03 3.98E-03 2.53E-02 1.96E-02 4.26E-02 1.01E-01 1.79E-01 1.25E+00 2.69E+00

APPENDIX C - Response functions of BSS to positive pions

146

TABLE C1

Sphere 81 mm

Sphere 81Cd Energy

[MeV] Response

[cm2] Uncertainty Response [cm2]

Uncertainty

50 4.43E-04 7.81E-06 2.61E-03 2.86E-05

70 6.90E-04 1.16E-05 4.19E-03 4.30E-05

80 8.59E-04 1.25E-05 5.44E-03 5.21E-05

100 1.57E-03 1.84E-05 5.98E-03 4.63E-05

120 1.69E-03 1.84E-05 7.27E-03 4.92E-05

150 1.94E-03 1.87E-05 7.57E-03 4.85E-05

180 1.84E-03 1.74E-05 7.58E-03 5.52E-05

200 1.96E-03 1.99E-05 7.40E-03 5.07E-05

220 1.79E-03 1.88E-05 7.60E-03 5.00E-05

250 1.86E-03 2.00E-05 7.24E-03 4.95E-05

280 1.55E-03 1.83E-05 6.78E-03 4.85E-05

300 1.54E-03 2.47E-05 6.57E-03 4.72E-05

400 1.34E-03 1.84E-05 6.27E-03 5.83E-05

500 1.10E-03 1.38E-05 5.59E-03 4.91E-05

700 1.09E-03 1.33E-05 5.66E-03 5.34E-05

1000 1.16E-03 1.29E-05 6.16E-03 4.98E-05

2.00E+03 1.07E-03 1.17E-05 6.75E-03 4.55E-05

3.00E+03 1.16E-03 1.44E-05 6.17E-03 4.41E-05

4.00E+03 1.45E-03 1.70E-05 7.01E-03 4.90E-05

5.00E+03 1.51E-03 1.74E-05 7.29E-03 4.82E-05

7.00E+03 1.49E-03 1.59E-05 8.05E-03 6.91E-05

1.00E+04 1.74E-03 2.06E-05 7.93E-03 6.30E-05

5.00E+04 1.65E-03 2.02E-05 8.30E-03 6.35E-05

1.00E+05 1.70E-03 1.89E-05 8.88E-03 8.19E-05

1.50E+05 1.77E-03 1.92E-05 8.71E-03 7.45E-05


147

TABLE C2

Sphere 108 mm

Sphere 133 mm Energy

[MeV] Response

[cm2]

Uncertainty Response

[cm2]

Uncertainty

50 2.97E-03 5.96E-05 5.89E-03 1.20E-04

70 3.67E-03 5.38E-05 9.77E-03 1.75E-04

80 4.66E-03 7.43E-05 1.14E-02 1.70E-04

100 7.11E-03 7.93E-05 1.68E-02 1.79E-04

120 9.92E-03 9.67E-05 2.38E-02 2.33E-04

150 1.19E-02 1.14E-04 2.81E-02 2.67E-04

180 1.20E-02 1.11E-04 2.94E-02 2.79E-04

200 1.07E-02 9.54E-05 3.12E-02 3.10E-04

220 1.18E-02 1.04E-04 2.84E-02 2.26E-04

250 1.03E-02 8.73E-05 2.75E-02 2.41E-04

280 9.22E-03 8.66E-05 2.45E-02 2.65E-04

300 9.70E-03 1.10E-04 2.41E-02 2.27E-04

400 7.65E-03 9.04E-05 2.00E-02 2.50E-04

500 7.49E-03 9.31E-05 1.83E-02 2.27E-04

700 7.43E-03 7.09E-05 1.83E-02 1.81E-04

1000 7.76E-03 9.60E-05 1.95E-02 1.81E-04

2.00E+03 7.67E-03 1.02E-04 2.06E-02 2.13E-04

3.00E+03 7.09E-03 8.51E-05 2.01E-02 2.29E-04

4.00E+03 8.60E-03 9.86E-05 2.13E-02 2.56E-04

5.00E+03 9.51E-03 1.07E-04 2.58E-02 2.83E-04

7.00E+03 1.01E-02 1.10E-04 2.77E-02 2.46E-04

1.00E+04 1.06E-02 1.02E-04 2.93E-02 2.79E-04

5.00E+04 1.15E-02 1.37E-04 3.00E-02 3.46E-04

1.00E+05 1.10E-02 1.34E-04 3.37E-02 3.33E-04

1.50E+05 1.22E-02 1.23E-04 3.22E-02 2.51E-04


148

TABLE C3

Sphere 178 mm


[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 1.06E-02 1.89E-04 50 1.43E-02 5.64E-04

70 2.20E-02 3.32E-04 65 2.78E-02 6.42E-04

80 2.78E-02 4.05E-04 80 4.49E-02 8.95E-04

100 4.11E-02 3.60E-04 100 6.55E-02 1.10E-03

120 5.58E-02 4.58E-04 110 8.44E-02 9.41E-04

150 6.89E-02 6.60E-04 120 9.17E-02 1.18E-03

180 7.53E-02 5.43E-04 130 1.09E-01 1.57E-03

200 7.88E-02 5.75E-04 150 1.25E-01 1.77E-03

220 7.42E-02 5.03E-04 200 1.47E-01 1.50E-03

250 7.22E-02 5.89E-04 250 1.40E-01 2.01E-03

280 6.73E-02 6.71E-04 300 1.22E-01 1.25E-03

300 6.57E-02 4.99E-04 400 1.02E-01 9.01E-04

400 5.54E-02 5.99E-04 500 9.46E-02 1.87E-03

500 5.09E-02 4.52E-04 700 9.87E-02 1.96E-03

700 5.36E-02 5.51E-04 1000 1.15E-01 2.63E-03

1000 5.81E-02 5.26E-04 1.50E+03 1.18E-01 2.04E-03

2.00E+03 5.95E-02 6.02E-04 2.00E+03 1.15E-01 1.50E-03

3.00E+03 5.80E-02 6.32E-04 3.00E+03 1.13E-01 1.49E-03

4.00E+03 6.52E-02 7.16E-04 4.00E+03 1.29E-01 1.39E-03

5.00E+03 7.09E-02 5.88E-04 5.00E+03 1.41E-01 1.47E-03

7.00E+03 8.00E-02 6.43E-04 7.00E+03 1.58E-01 1.72E-03

1.00E+04 8.32E-02 9.60E-04 9.00E+03 1.61E-01 1.68E-03

5.00E+04 8.89E-02 8.40E-04 1.00E+04 1.66E-01 3.38E-03

1.00E+05 9.87E-02 8.95E-04 5.00E+04 1.99E-01 1.74E-03

1.50E+05 1.10E-01 7.85E-04 1.00E+05 2.02E-01 1.85E-03

1.50E+05 2.25E-01 2.23E-03


149

TABLE C4

STANLIO

OLLIO Energy

[MeV] Response

[cm2] Uncertainty

Response

[cm2] Uncertainty

50 1.54E-02 1.59E-04 3.67E-02 7.84E-04

100 1.37E-01 5.13E-04 6.92E-01 4.65E-03

150 2.69E-01 6.63E-04 1.35E+00 5.45E-03

250 3.03E-01 7.64E-04 1.62E+00 6.41E-03

300 3.05E-01 7.14E-04 1.64E+00 5.15E-03

500 3.15E-01 6.42E-04 1.68E+00 7.39E-03

700 3.43E-01 7.03E-04 1.88E+00 6.45E-03

1000 3.85E-01 7.58E-04 2.09E+00 6.63E-03

2.00E+03 5.00E-01 8.75E-04 2.67E+00 7.02E-03

3.00E+03 5.47E-01 9.03E-04 2.92E+00 7.57E-03

4.00E+03 6.20E-01 8.99E-04 3.30E+00 9.14E-03

5.00E+03 6.52E-01 9.72E-04 3.47E+00 7.96E-03

7.00E+03 7.56E-01 1.07E-03 3.95E+00 9.91E-03

9.00E+03 8.45E-01 1.26E-03 4.36E+00 9.28E-03

1.00E+04 8.76E-01 1.10E-03 4.53E+00 1.44E-02

5.00E+04 1.32E+00 2.57E-03 6.66E+00 2.21E-02

1.00E+05 1.59E+00 3.11E-03 8.03E+00 2.77E-02

1.50E+05 1.81E+00 3.08E-03 8.95E+00 2.41E-02

APPENDIX C - Response functions of BSS to negative pions

150

TABLE C5

Sphere 81 mm

Sphere 81Cd Energy

[MeV] Response [cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 2.12E-03 1.87E-05 50 8.71E-03 7.51E-05

60 2.30E-03 2.10E-05 60 8.50E-03 7.22E-05

80 2.87E-03 2.51E-05 80 9.21E-03 7.65E-05

100 3.43E-03 2.53E-05 100 1.07E-02 8.44E-05

110 3.77E-03 2.52E-05 110 1.16E-02 8.23E-05

130 3.80E-03 2.74E-05 130 1.17E-02 7.64E-05

150 3.83E-03 2.74E-05 150 1.11E-02 8.09E-05

200 3.20E-03 2.48E-05 200 1.03E-02 7.70E-05

250 2.74E-03 2.33E-05 250 9.26E-03 5.68E-05

300 2.38E-03 1.97E-05 300 8.99E-03 5.68E-05

400 2.14E-03 2.19E-05 400 7.59E-03 4.31E-05

500 1.91E-03 1.79E-05 500 7.39E-03 4.87E-05

700 1.82E-03 1.73E-05 700 7.53E-03 6.12E-05

1000 1.80E-03 2.05E-05 1000 7.38E-03 5.72E-05

2.00E+03 1.57E-03 1.66E-05 2.00E+03 6.80E-03 5.35E-05

3.00E+03 1.34E-03 1.67E-05 3.00E+03 6.81E-03 6.06E-05

5.00E+03 1.56E-03 1.80E-05 5.00E+03 7.36E-03 5.82E-05

7.00E+03 1.69E-03 1.73E-05 7.00E+03 8.01E-03 6.86E-05

1.00E+04 1.68E-03 1.71E-05 1.00E+04 8.15E-03 6.54E-05

5.00E+04 1.80E-03 2.16E-05 5.00E+04 8.03E-03 6.66E-05

1.00E+05 1.81E-03 1.94E-05 1.00E+05 8.16E-03 6.00E-05

1.50E+05 1.57E-03 1.70E-05 1.50E+05 8.60E-03 6.86E-05


151

TABLE C6

Sphere 108 mm Sphere 133 mm Energy

[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 6.26E-02 2.99E-04 50 4.90E-01 1.18E-03

55 1.58E-02 1.21E-04 60 7.73E-02 4.60E-04

60 1.50E-02 1.46E-04 80 4.53E-02 3.61E-04

80 1.73E-02 1.32E-04 100 5.45E-02 3.11E-04

100 2.13E-02 1.39E-04 110 5.83E-02 3.91E-04

110 2.27E-02 1.45E-04 130 6.37E-02 5.17E-04

130 2.44E-02 1.60E-04 150 6.34E-02 4.15E-04

150 2.35E-02 1.46E-04 200 5.42E-02 3.69E-04

200 2.12E-02 1.61E-04 250 4.63E-02 3.84E-04

250 1.70E-02 1.13E-04 300 3.90E-02 2.97E-04

300 1.48E-02 1.16E-04 400 3.12E-02 2.77E-04

400 1.25E-02 1.03E-04 500 3.18E-02 3.30E-04

500 1.16E-02 1.14E-04 700 3.08E-02 3.44E-04

700 1.23E-02 1.32E-04 1000 3.12E-02 2.64E-04

1000 1.17E-02 1.08E-04 2.00E+03 2.67E-02 2.50E-04

2.00E+03 1.03E-02 1.23E-04 3.00E+03 2.32E-02 2.37E-04

3.00E+03 8.46E-03 8.65E-05 5.00E+03 2.86E-02 3.18E-04

5.00E+03 1.05E-02 1.20E-04 7.00E+03 3.00E-02 2.54E-04

7.00E+03 1.11E-02 1.00E-04 1.00E+04 3.02E-02 3.30E-04

1.00E+04 1.10E-02 1.01E-04 5.00E+04 3.02E-02 2.63E-04

5.00E+04 1.15E-02 1.19E-04 1.00E+05 3.26E-02 3.38E-04

1.00E+05 1.15E-02 1.08E-04 1.50E+05 3.53E-02 3.62E-04

1.50E+05 1.16E-02 1.18E-04


152

TABLE C7

Sphere 178 mm


[MeV] Response

[cm2]

Uncertainty

Energy

[MeV] Response

[cm2]

Uncertainty

50 1.54E+00 3.38E-03 50 2.48E+00 5.40E-03

60 9.49E-01 2.99E-03 60 2.13E+00 5.97E-03

70 2.98E-01 1.68E-03 70 1.26E+00 4.62E-03

80 1.21E-01 1.06E-03 80 5.71E-01 3.17E-03

90 1.30E-01 1.26E-03 90 2.60E-01 2.86E-03

100 1.45E-01 1.04E-03 100 2.72E-01 2.05E-03

110 1.58E-01 8.69E-04 110 2.91E-01 3.85E-03

130 1.66E-01 9.05E-04 120 3.10E-01 2.15E-03

150 1.74E-01 1.09E-03 130 3.23E-01 1.99E-03

200 1.50E-01 8.93E-04 140 3.21E-01 2.10E-03

250 1.26E-01 1.07E-03 150 3.27E-01 2.45E-03

300 1.05E-01 7.59E-04 200 2.94E-01 1.80E-03

400 8.89E-02 8.44E-04 250 2.56E-01 2.47E-03

500 8.41E-02 6.63E-04 300 2.07E-01 1.67E-03

700 8.66E-02 7.60E-04 400 1.67E-01 2.03E-03

1000 8.96E-02 7.20E-04 500 1.71E-01 2.05E-03

2.00E+03 7.61E-02 6.70E-04 700 1.67E-01 1.35E-03

3.00E+03 6.91E-02 6.74E-04 1000 1.65E-01 4.31E-04

5.00E+03 8.03E-02 6.77E-04 2.00E+03 1.54E-01 1.35E-03

7.00E+03 8.24E-02 6.77E-04 3.00E+03 1.36E-01 1.64E-03

1.00E+04 8.89E-02 7.20E-04 5.00E+03 1.58E-01 1.67E-03

5.00E+04 9.35E-02 8.87E-04 7.00E+03 1.68E-01 2.04E-03

1.00E+05 9.98E-02 1.09E-03 1.00E+04 1.74E-01 2.67E-03

1.50E+05 1.07E-01 8.06E-04 5.00E+04 2.02E-01 1.56E-03

1.00E+05 2.14E-01 1.47E-03

1.00E+05 2.28E-01 1.86E-03


153

TABLE C8

STANLIO OLLIO Energy

[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 9.00E-01 2.31E-03 50 6.63E+00 1.03E-02

60 9.93E-01 1.19E-03 60 6.74E+00 1.63E-02

65 8.00E-01 2.04E-03 65 5.93E+00 1.67E-02

70 7.09E-01 9.72E-04 70 4.84E+00 8.62E-03

75 7.61E-01 2.07E-03 75 4.51E+00 1.64E-02

80 8.66E-01 1.25E-03 80 4.45E+00 9.16E-03

85 9.51E-01 2.34E-03 85 4.91E+00 2.93E-02

90 9.40E-01 1.16E-03 90 5.54E+00 1.35E-02

100 8.32E-01 1.98E-03 100 5.04E+00 1.31E-02

110 7.00E-01 1.86E-03 110 3.71E+00 1.06E-02

120 5.77E-01 1.73E-03 120 3.00E+00 7.64E-03

130 5.30E-01 1.63E-03 130 2.70E+00 7.11E-03

150 4.84E-01 1.36E-03 150 2.50E+00 7.19E-03

200 4.30E-01 1.39E-03 200 2.38E+00 6.94E-03

300 3.97E-01 1.53E-03 300 2.17E+00 7.76E-03

500 3.69E-01 1.37E-03 500 2.04E+00 8.84E-03

700 4.12E-01 1.57E-03 700 2.20E+00 7.48E-03

1000 4.52E-01 1.40E-03 1000 2.40E+00 8.25E-03

2.00E+03 5.54E-01 1.50E-03 2.00E+03 2.96E+00 1.06E-02

3.00E+03 5.80E-01 1.61E-03 3.00E+03 3.12E+00 8.55E-03

5.00E+03 6.90E-01 1.99E-03 5.00E+03 3.59E+00 1.08E-02

7.00E+03 7.61E-01 1.87E-03 7.00E+03 4.07E+00 7.97E-03

1.00E+04 9.02E-01 2.18E-03 1.00E+04 4.64E+00 8.36E-03

5.00E+04 1.33E+00 2.66E-03 5.00E+04 6.69E+00 2.69E-02

1.00E+05 1.59E+00 3.12E-03 1.00E+05 7.87E+00 2.72E-02

1.50E+05 1.81E+00 3.32E-03 1.50E+05 8.97E+00 2.28E-02

APPENDIX C - Response functions of BSS to protons

154

TABLE C9

Sphere 81 mm

Sphere 81Cd Energy

[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 4.13E-04 7.63E-06 50 2.14E-03 2.80E-05

60 6.06E-04 7.67E-06 60 2.60E-03 2.68E-05

80 9.43E-04 1.01E-05 80 3.15E-03 3.22E-05

100 1.07E-03 1.26E-05 100 4.43E-03 3.78E-05

150 9.91E-04 1.26E-05 150 4.61E-03 4.91E-05

200 9.60E-04 1.36E-05 200 4.34E-03 4.72E-05

300 1.04E-03 1.41E-05 300 4.90E-03 6.20E-05

500 1.18E-03 1.54E-05 500 5.26E-03 5.70E-05

700 1.20E-03 1.60E-05 700 5.71E-03 5.36E-05

1000 1.56E-03 2.01E-05 1000 6.51E-03 5.83E-05

2.00E+03 1.67E-03 2.15E-05 2.00E+03 7.22E-03 6.55E-05

5.00E+03 2.16E-03 2.43E-05 3.00E+03 7.09E-03 5.92E-05

7.00E+03 2.13E-03 2.24E-05 5.00E+03 9.08E-03 6.58E-05

1.00E+04 2.19E-03 2.13E-05 7.00E+03 9.31E-03 7.17E-05

5.00E+04 2.09E-03 1.87E-05 1.00E+04 9.68E-03 7.18E-05

1.00E+05 2.10E-03 1.95E-05 5.00E+04 1.06E-02 7.07E-05

1.50E+05 2.35E-03 2.03E-05 1.00E+05 1.08E-02 7.60E-05

1.50E+05 1.15E-02 7.23E-05


155

TABLE C10

Sphere 108 mm


[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 1.09E-03 2.75E-05 50 2.35E-03 7.69E-05

60 2.11E-03 3.65E-05 60 4.07E-03 9.91E-05

80 4.26E-03 5.67E-05 80 8.91E-03 1.27E-04

100 6.11E-03 7.57E-05 100 1.33E-02 1.43E-04

150 6.29E-03 8.52E-05 150 1.65E-02 2.28E-04

200 6.40E-03 9.18E-05 200 1.54E-02 1.75E-04

300 6.57E-03 8.14E-05 300 1.63E-02 2.74E-04

500 7.44E-03 8.32E-05 500 1.84E-02 2.12E-04

700 7.97E-03 8.64E-05 700 2.18E-02 2.52E-04

1000 9.61E-03 1.06E-04 1000 2.46E-02 2.25E-04

2.00E+03 9.45E-03 8.52E-05 2.00E+03 2.54E-02 2.38E-04

3.00E+03 9.30E-03 1.08E-04 3.00E+03 2.59E-02 3.07E-04

5.00E+03 1.20E-02 1.06E-04 5.00E+03 3.17E-02 2.82E-04

7.00E+03 1.27E-02 1.15E-04 7.00E+03 3.38E-02 2.88E-04

1.00E+04 1.37E-02 1.13E-04 1.00E+04 3.66E-02 3.38E-04

5.00E+04 1.41E-02 1.30E-04 5.00E+04 4.10E-02 2.98E-04

1.00E+05 1.58E-02 1.34E-04 1.00E+05 4.44E-02 3.32E-04

1.50E+05 1.63E-02 1.27E-04 1.50E+05 4.65E-02 3.04E-04


156

TABLE C11

Sphere 178 mm Sphere 233 mm Energy

[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 4.51E-03 1.72E-04 50 6.35E-03 2.62E-04

60 7.61E-03 3.32E-04 60 9.98E-03 3.49E-04

80 1.57E-02 2.65E-04 80 2.10E-02 4.54E-04

100 2.58E-02 3.58E-04 100 3.77E-02 6.68E-04

150 4.25E-02 4.91E-04 150 7.63E-02 8.51E-04

200 4.32E-02 5.38E-04 200 8.12E-02 1.38E-03

300 4.36E-02 6.11E-04 300 8.14E-02 1.24E-03

500 5.09E-02 6.26E-04 500 9.99E-02 9.99E-04

700 5.75E-02 6.40E-04 700 1.11E-01 1.34E-03

1000 6.61E-02 6.02E-04 1000 1.22E-01 1.26E-03

2.00E+03 7.24E-02 7.46E-04 2.00E+03 1.45E-01 1.52E-03

3.00E+03 7.32E-02 6.58E-04 3.00E+03 1.46E-01 1.38E-03

5.00E+03 8.92E-02 8.00E-04 5.00E+03 1.77E-01 1.65E-03

7.00E+03 9.43E-02 7.52E-04 7.00E+03 1.88E-01 1.44E-03

1.00E+04 1.04E-01 7.88E-04 1.00E+04 2.08E-01 1.80E-03

5.00E+04 1.25E-01 1.22E-03 5.00E+04 2.62E-01 2.21E-03

1.00E+05 1.36E-01 1.36E-03 1.00E+05 3.00E-01 1.82E-03

1.50E+05 1.46E-01 8.52E-04 1.50E+05 3.25E-01 2.29E-03


157

TABLE C12

STANLIO

OLLIO

Energy

[MeV] Response

[cm2] Uncertainty

Energy

[MeV] Response

[cm2] Uncertainty

50 5.97E-03 1.90E-04 50 8.47E-03 3.27E-04

100 3.46E-02 4.37E-04 100 9.14E-02 1.29E-03

150 9.56E-02 7.27E-04 150 4.50E-01 3.52E-03

300 2.10E-01 1.25E-03 300 1.11E+00 9.47E-03

500 2.69E-01 1.10E-03 500 1.42E+00 1.62E-02

700 3.21E-01 1.21E-03 700 1.76E+00 1.11E-02

1000 3.69E-01 1.32E-03 1000 2.06E+00 1.06E-02

2.00E+03 5.24E-01 1.71E-03 2.00E+03 2.79E+00 1.45E-02

3.00E+03 6.17E-01 1.72E-03 3.00E+03 3.24E+00 1.54E-02

5.00E+03 7.61E-01 1.97E-03 5.00E+03 4.08E+00 1.89E-02

7.00E+03 8.71E-01 2.08E-03 7.00E+03 4.63E+00 1.89E-02

1.00E+04 1.01E+00 2.26E-03 1.00E+04 5.32E+00 1.86E-02

5.00E+04 1.67E+00 2.85E-03 5.00E+04 8.38E+00 2.32E-02

1.00E+05 2.07E+00 3.57E-03 1.00E+05 1.03E+01 3.27E-02

1.50E+05 2.35E+00 3.90E-03 1.50E+05 1.18E+01 2.88E-02

APPENDIX C - Response functions of the BSS to charged hadrons

158

102 103 104 105

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

2.5x10-3

3.0x10-3

3.5x10-3

4.0x10-3

positive pions negative pions protons

Response of 81 mm sphere to charged hadronsR

espo

nse

[cm

2 ]

Energy [MeV]

102 103 104 105

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

1.0x10-2

1.2x10-2


Response of sphere 81cd to charged hadrons

Res

pons

e [c

m2 ]

Energy [MeV]


159

102 103 104 105

10-3

10-2

10-1


Response of 108 mm to charged hadronsR

espo

nse

[cm

2 ]

Energy [MeV]

102 103 104 10510-3

10-2

10-1

100


Response of sphere 133 mm to charged hadrons

Res

pons

e [c

m2 ]

Energy [MeV]


160

102 103 104 10510-3

10-2

10-1

100



Res

pons

e [c

m2 ]

Energy [MeV]

102 103 104 105

10-2

10-1

100



Res

pons

e [c

m2 ]

Energy [MeV]

APPENDIX C - Response function of the BSS to charged hadrons

161

102 103 104 105

0.0

5.0x10-1

1.0x100

1.5x100


Response of Stanlio to charged hadrons

Res

pons

e [c

m2 ]

Energy [MeV]

102 103 104 105

0

1x100

2x100

3x100

4x100

5x100

6x100

7x100

8x100


Response of Ollio to charged hadrons

Res

pons

e [c

m2 ]

Energy [MeV]

APPENDIX C - Response function of the BSS to charged hadrons

162

TABLE C13

Cu target Pb target Angle Particle Stanlio Ollio Stanlio Ollio

neutron 0.462±0.026 0.386±0.053 0.522±0.029 0.563±0.085 π+ 0.196±0.010 0.219±0.024 0.173±0.008 0.155±0.018 π- 0.249±0.010 0.294±0.039 0.228±0.010 0.214±0.027 p 0.092±0.005 0.101±0.018 0.077±0.005 0.067±0.015

30°

π++π-+ p 0.538±0.019 0.614±0.058 0.478±0.017 0.437±0.044 neutron 0.658±0.040 0.566±0.088 0.719±0.047 0.754±0.124

π+ 0.111±0.006 0.138±0.022 0.090±0.005 0.077±0.012 π- 0.169±0.008 0.225±0.027 0.147±0.008 0.134±0.016 p 0.061±0.005 0.071±0.012 0.045±0.004 0.035±0.011

45°

π++π-+ p 0.342±0.015 0.434±0.046 0.281±0.013 0.246±0.030 neutron 0.782±0.057 0.699±0.108 0.828±0.061 0.854±0.156

π+ 0.062±0.004 0.083±0.012 0.047±0.003 0.039±0.009 π- 0.115±0.007 0.168±0.022 0.094±0.006 0.084±0.012 p 0.041±0.004 0.050±0.011 0.031±0.003 0.022±0.006

60°

π++π-+ p 0.218±0.012 0.301±0.033 0.172±0.009 0.145±0.019 neutron 0.749±0.051 0.782±0.138 0.886±0.059 0.905±0.135

π+ 0.064±0.004 0.053±0.010 0.027±0.002 0.021±0.003 π- 0.140±0.010 0.130±0.017 0.064±0.004 0.058±0.007 p 0.046±0.004 0.034±0.008 0.024±0.002 0.016±0.004

75°

π++π-+ p 0.251±0.014 0.218±0.028 0.114±0.006 0.095±0.011 neutron 0.807±0.057 0.834±0.142 0.919±0.068 0.933±0.209

π+ 0.046±0.003 0.037±0.007 0.017±0.001 0.013±0.003 π- 0.117±0.006 0.108±0.017 0.047±0.003 0.042±0.007 p 0.029±0.003 0.021±0.006 0.018±0.002 0.012±0.005

90°

π++π-+ p 0.193±0.010 0.166±0.023 0.081±0.005 0.067±0.012 neutron 0.846±0.059 0.867±0.138 0.938±0.066 0.949±0.147

π+ 0.035±0.003 0.028±0.005 0.012±0.001 0.009±0.002 π- 0.102±0.006 0.093±0.013 0.037±0.002 0.033±0.005 p 0.018±0.002 0.012±0.003 0.013±0.001 0.008±0.002

105°

π++π-+ p 0.154±0.009 0.133±0.017 0.062±0.003 0.051±0.007 neutron 0.868±0.063 0.887±0.147 0.951±0.070 0.960±0.238

π+ 0.029±0.002 0.022±0.004 0.009±0.001 0.007±0.002 π- 0.092±0.006 0.084±0.011 0.031±0.002 0.028±0.005

proton 0.012±0.001 0.007±0.002 0.009±0.001 0.006±0.002

120°

π++π-+ p 0.132±0.008 0.113±0.015 0.049±0.003 0.040±0.008 neutron 0.885±0.062 0.899±0.139 0.958±0.070 0.966±0.199

π+ 0.024±0.002 0.019±0.003 0.008±0.001 0.006±0.001 π- 0.083±0.006 0.074±0.009 0.028±0.002 0.025±0.004 p 0.009±0.001 0.008±0.002 0.006±0.001 0.004±0.001

135°

π++π-+ p 0.115±0.007 0.101±0.012 0.042±0.002 0.034±0.006

APPENDIX D – Upgrade and Development of CERF

163

Table D1. Technical characteristics of the BIG PIC.

Parallel Plate Chamber for monitoring

high energetic beams of moderate intensity

Plate spacing 50 mm

Effective Diameter 250 mm

Aperture 185 mm

Effective volume 4.9 litres

Surface area 490 cm2

Electrodes Aluminized mylar foil

Electrode thickness 2.5 mg/cm2

Total Chamber thickness 35.5 mg/cm2

Table D2. 137C source identification, air kerma rate, duration and number of measurements

performed for the July 2002 stability test and the linearity test of the BIG PIC performed in the

SC/RP calibration laboratory.

Stability test Linearity test Source

Air kerma rate

(µGy/h) Duration (s) Duration (s) Repetition

10 1000 10 30

20 1000 10 20

30 1000 10 20

40 1000 10 15

Cs3739

50 1000 10 15

70 1000 10 15

80 1000 10 15

90 1000 10 15

100 1000 10 10

Cs3740

300 1000 10 10


164

Table D2. (continued). 500 300 10 10

700 300 10 10

1000 300 10 10 Cs3609

3000 300 100 2

10000 100 100 2

20000 100 100 2 Cs2045

30000 100 100 2

Table D3. Average count rate versus applied voltage for the determination

of the region of ion saturation of the BIG PIC.

Voltage

(V)

Average count rate

(counts per s)

Voltage

(V)

Average count rate

(counts per s) 10 623.47 ± 5.58 500 1167.43 ± 4.41

30 1058.23 ± 7.27 550 1160.83 ± 4.40

50 1137.08 ± 7.54 600 1167.14 ± 4.41

100 1154.63 ± 7.60 650 1161.59 ± 4.40

150 1165.75 ± 7.63 700 1160.16 ± 4.40

200 1164.86 ± 6.23 750 1162.48 ± 4.40

250 1167.46 ± 6.24 800 1160.88 ± 4.40

300 1167.40 ± 6.24 850 1157.83 ± 4.39

350 1168.63 ± 4.41 900 1161.94 ± 4.40

400 1164.32 ± 4.41 950 1159.96 ± 4.40

450 1165.14 ± 4.41 1000 1160.80 ± 4.40


165

Table D4. Inter-comparison of the two CERF beam monitors (the standard PIC and the BIG PIC) in the hadron

beam at CERF. The total PIC and BIG PIC values are the sum of the readings of the single pulses. The uncertainty

associated to the ratio in the last column has been computed by the usual error propagation formula.

Beam intensity

(PIC/pulse)

Average beam intensity

(PIC/pulse) BIG PIC Total PIC

Total

BIG PIC BIG PIC/PIC

11598 34259

11759 34761

11732

11696 ± 86

34661

35089 103681 2.95 ± 0.05

10507 31154

10648 31553

10293 30510

10461 31038

10661

10514 ± 151

31445

52570 155700 2.96 ± 0.07

7978 23754

7991 23802

8229 24504

8371 24887

8304

8175 ± 181

24722

40873 121669 2.98 ± 0.09

6057 18217

6238 18743

6105 18372

5889 17721

6003

6058 ± 129

18388

30292 91441 3.02 ± 0.10

3951 12060

3994 12220

4025 12314

3927 12023

4008

3981 ± 41

12275

19905 60892 3.06 ± 0.13

1991 6116

2000 6142

2002 6140

2028 6234

1987

2002 ± 16

6101

10008 30733 3.07 ± 0.18

212 656

212 654

213 657

212 655

211

212 ± 1

651

1060 3273 3.09 ± 0.55


166

Table D5. Raw data of the PIC and BIG PIC for different beam intensities, taken at CERF in

August 2003.

Approximate beam

intensity

(PIC/pulse)

Total

PIC-counts

in 5 cycles

Total

BIG PIC-counts

in 5 cycles

BIG PIC counts

/ PIC counts

75 366 1,090 2.98

165 815 2,447 3.00

290 1,469 4,419 3.01

460 2,290 6,907 3.02

645 3,264 9,843 3.02

1,160 5,802 17,472 3.01

1,780 8,884 26,734 3.00

2,480 9,880 29,571 2.99

3,280 16,516 49,426 2.99

4,060 20,349 61,081 3.00

Table D6. Parameters during the measurements of the voltage characteristic curves of the

BIG PIC in the hadron beam at CERF.

Collimator settings Nominal beam

intensity

(PIC/pulse)

Actual beam

intensity

(PIC/pulse)

Beam fluctuation

(%) C3 C5

14,000 13,650 2.14 ± 22 ± 22

12,000 11,650 1.91 ± 19 ± 19

10,500 10,600 2.00 ± 17 ± 17

8,000 8,250 1.36 ± 14 ± 15

6,000 6,100 0.86 ± 12 ± 12

4,000 4,000 0.54 ± 9 ± 10

2,000 1,960 0.21 ± 7 ± 6

1,000 950 0.47 ± 5 ± 4

450 455 0.04 ± 3 ± 3

200 210 0.03 ± 2 ± 2

BIBLIOGRAPHY AND REFERENCES

167

1. R. L. Bramblett, R. I. Ewing and T. W. Bonner, A new type of neutron

spectrometer, Nucl. Instr. and Meth. 9, pp. 1-12 (1960).

2. G.J.H. Jacobs, R.L.P. van den Bosch, Nucl. Instr. and Meth. 175, pp. 483, (1980).

3. R.S. Sanna, US AEC Report HASL-267, (1973).

4. N.E. Hertel, J.W. Davidson, The response of bonner spheres to neutrons from

thermal energies to 17.3 MeV Nucl. Instr. and Meth. A 238, pp. 509-516 (1985).

5. D. E. Hankins, The Substitution of a BF3 Probe for the LiI Crystal in Neutron rem-

Meters, Health Phys. 14, pp. 518-520, (1968).

6. M.P. Dhairyawan, P. S. Nagarajan and G. Venkataraman, Response functions of

spherically moderated neutron detectors, Nucl. Instr. and Meth. 169, pp. 115-120

(1980).

7. J.C. Liu, F. Hajnal, C.S. Sims, and J. Kuiper, Neutron Spectral Measurements at

ORNL, Radiat. Prot. Dosim. 30, pp. 169-178 (1990).

8. W.F. Harvey, F. Hajnal, Los Alamos Report LA 1253s, (1993).

9. G.L. Watkins and G.R. Holeman, The evaluation of an iterative technique’s use in

unfolding neutron spectra data, Health Phys. 15, pp. 535-539, (1968).

10. G.R Holeman, K.W Price, L.F Friedman and R.Nath, Neutron spectral

measurements in an intense photon field associated with a high-energy x-ray

radiotherapy machine, Med. Phys. 4 (6), pp. 508-515, (1977).

11. M. Bricka, et al., Proceedings of a Symposium on Neutron Monitoring for

Radiation Protection Purposes, December 1972, IAEA, Vienna, pp. 279 (1973).

12. M. Mourgues, Proceedings of the Second Symposium on Neutron Dosimetry in

Biology and Medicine, Munchen, September–October 1974, Vol. II, CEC,

Luxembourg, pp. 907 (1975).

13. A.V. Alevra, M. Cosack, J.B. Hunt, D.J. Thomas, and H. Schraube, Experimental

Determination of the Response of Four Bonner Sphere sets to Monoenergetic

Neutrons, Radiat. Prot. Dosim. 23, pp. 293-296 (1988).

14. A. Aroua, M. Grecescu, P. Lerch, S. Prêtre, J. -F. Valley and V. Vylet, Evaluation

and test of the response matrix of a multisphere neutron spectrometer in a wide

energy range Part I. Calibration, Nucl. Instr. and Meth. A 321, pp. 298-304 (1992).

15. P. M. Thomas, K.G. Harrison and M.C. Scott, A multisphere neutron spectrometer

using a central 3He detector, Nucl. Instr. and Meth. 224, pp. 225-232 (1984).


168

16. A. V. Alevra, M. Cosack, J. B. Hunt, D. J Thomas and H. Schraube, Experimental

determination of the response of four Bonner sphere sets to monoeergetic neutrons

(I) Radiat. Prot. Dosim. 23, pp. 249-252, (1988).

17. A. V. Alevra, M. Cosack, J. B. Hunt, D. J Thomas and H. Schraube, Experimental

determination of the response of four Bonner sphere sets to monoenergetic

neutrons (II), Radiat. Prot. Dosim. 40 (2), pp. 91-102, (1992).

18. D.J. Thomas, A. V. Alevra, J. B. Hunt and H. Schraube, Experimental

determination of the response of four Bonner sphere sets to thermal neutrons

Radiat. Prot. Dosim. 54 (1), pp. 25-31, (1994).

19. IAEA, Radiological safety aspects of the operation of electron linear

accelerators, Technical report series no. 188 (1979).

20. P. M.Thomas, K. G.Harrison and M. C.Scott, A multisphere neutron spectrometer

using a central 3He detector Nucl. Instr. and Meth. 224, pp. 225- 232 (1984).

21. A.V. Alevra, M. Cosack, J.B. Hunt, D.J. Thomas, and H. Schraube, Experimental

Determination of the Response of Four Bonner Sphere Sets to Monoenergetic

Neutrons (II), Radiat. Prot. Dosim. 40, pp. 91- 102 (1992).

22. D.J. Thomas, A.V. Alevra, J.B. Hunt, and H. Schraube, Experimental

Determination of the Response of Four Bonner Sphere Sets to Thermal Neutrons,

Radiat. Prot. Dosim. 54, pp. 25- 31 (1994).

23. V.Mares, G. Schraube and H. Schraube, Calculated neutron response of a Bonner

sphere spectrometer with 3He counter Nucl. Instr. and Meth. A 307, pp. 398- 412.

(1991).

24. B. Wiegel, A.V. Alevra and B.R.L Siebert, Calculations of the Response

Functions of Bonner Spheres with a Spherical 3He Proportional Counter Using a

Realistic Detector Model Report PTB-N-21, (1994).

25. V. Mares and H. Schraube, Improved response matrices of Bonner sphere

spectrometers with 6LiI scintillation detector and 3He proportional counter between

15 and 100 MeV Nucl. Instr. and Meth. A 366, pp. 203- 206 (1995).

26. B. Wiegel, A. V. Alevra, M. Matzke, U. J. Schrewe and J. Wittstock,

Spectrometry using the PTB neutron multisphere spectrometer (NEMUS) at flight

altitudes and at ground level , Nucl. Instr. and Meth. A 476, pp. 52- 57 (2002).


169

27. Y.Uwamino, T. Nakamura A. Hara, Two types of multi-moderator neutron

spectrometers: Gamma-ray insensitive type and high-efficiency type, Nucl. Instr.

and Meth. A 239, pp. 299- 309 (1985).

28. P. Goldhagen, M. Reginatto, T. Kniss, J. W. Wilson, R. C. Singleterry, I. W. Jones

and W. Van Steveninck, Measurement of the energy spectrum of cosmic-ray

induced neutrons aboard an ER-2 high-altitude airplane Nucl. Instr. and Meth. A

476, pp. 42- 51 (2002).

29. E.J. Axton, A.G. Bardell, NBS Spec. Publ. 554, pp. 109 (1979).

30. A.V. Sannikov, Report CERN/TIS-RP/IR/94-05, (1994).

31. D. J. Thomas, A. G. Bardell and E. M. Macaulay, Characterization of a gold foil-

based Bonner sphere set and measurements of neutron spectra at a medical

accelerator, Nucl. Instr. and Meth. A 476, pp. 31- 35 (2002).

32. R.J. Henninger, et al., Nucl. Instr. and Meth. 96, pp. 515 (1971).

33. C.K. Wang and T.E. Blue, A neutron spectrometer for neutrons with energies

between 1 eV and 10 keV, Nucl. Instr. and Meth. A 290, pp. 237- 241 (1990).

34. A.Aroua, M. Grecescu, S. Prêtre, and J.F. Valley, Improved Neutron Spectrometer

Based on Bonner Spheres, Radiat Prot Dosimetry 70, pp. 285-289 (1997).

35. V. Vylet, Response matrix of an extended Bonner sphere system,Nucl. Instr. and

Meth. A 476, pp. 26- 30 (2002).

36. M. Grecescu, A. Aroua, M. Boschung, A. Fiechtner, K. Gmür, J. -P.Laedermann,

J. -F. Valley and C. Wernli, Neutron field measurements at the 590 MeV ring

cyclotron of the Paul Scherrer Institute, Nucl. Instr. and Meth. A476, pp. 58- 62

(2002).

37. H. Toyokawa, A. Uritani, C. Mori, N. Takeda, K. Kudo, A Spherical Neutron

Counter with an Extended Energy Response for Dosimetry, Radiat. Prot. Dosim.

70, pp. 365- 370 (1997).

38. S. Yamaguchi, A. Uritani, H. Sakai, C. Mori, T. Iguchi, H. Toyokawa, N.Takeda

and K. Kudo, Spherical neutron detector for space neutron measurement, Nucl.

Instr. and Meth. A 422, pp. 600- 605 (1999).

39. NCRP, Neutron contamination from medical electron accelerators, National

Council on Radiation Protection and Measurements, Report 79, (1984).

40. AAPM, Neutron measurements around high-energy X-ray radiotherapy machines,

Report 19, (1986).


170

41. H. Dinter, and K. Tesch, Determination of Neutron Spectra Behind Lateral

Shielding of High Energy Proton Accelerators, Radiat. Prot Dosim. 42 , pp 5-10

(1992)

42. V.P. Kryuchkov, G.I. Semenova, IHEP Preprint 87–191, Serpukhov, 1987.

43. H.H. Hsu, K.R Alvar and D.G Vasilik, A new Bonner-sphere set for high energy

neutron measurements: Monte Carlo simulation IEEE Trans. Nucl. Sci. NS-41,

pp. 938-940 (1994).

44. V. Mares, H. Schraube, Proceedings IRPA Symposium, Prague, September 1997,

pp. 543 (1998).

45. Angela Mitaroff, Design, Calibration and Tests of an Extended Range Bonner

Sphere Spectrometer, PhD Thesis, CERN (2002).

46. A. Fassò, A. Ferrari, J. Ranft and P. Sala, The FLUKA Manual, 1998.

47. MAB (Münchner Apparate Bau), GmbH, Wallstr.122, Marl,Germany.

48. Mr. C. Hoffmann, Fundamental Software AM-SMCA01, Manual, FREIBERGER

SENSORTECHNIK, (private communication).

49. R. Gold, An Iterative Unfolding Method for Response Matrices, USAEC Report

ANL-6984, Argonne National Laboratory, Argonne, IL (1964).

50. M. Reginatto, P. Goldhagen, MAXED, a Computer Code for Maximum Entropy

Deconvolution of Multisphere Neutron Spectrometer Data, Health Phys. 77 (5),

pp. 579-583, (1999).

51. R. Wilczek and S. Drapatz, A high accuracy algorithm for maximum entropy

image restoration in the case of small data sets. Astron. Astophys.142, pp. 9-12

(1985).

52. A. Corana, M. Marchesi, C. Martini and S. Ridella, Minimizing Multimodal

Functions of Continuous Variables with the "Simulated Annealing" Algorithm,

ACM Transactions on Mathematical Software, 13 (3), pp. 262-280 (1987).

53. J.E Shore, R.W. Johnson Axiomatic derivation of the principle of maximum

entropy and the principle of minimum cross- entropy. IEEE Trans. Info. Theory:

IT-26, pp. 26-37, (1980).

54. M. Matzke, Unfolding of pulse height spectra: the HEPRO program system, PTB

report PTB-N-19 (Braunschweig: Physikalisch- Technische Bundesanstalt)

(1994).


171

55. M. Matzke, Unfolding of particle spectra. In Proc. of Int. Conf on Neutrons in

Research and Industry, Crete 1996. Ed.G. Vourvopoulos, (Proc. SPIE 2867,

pp. 598-607, (1997).

56. M. Matzke, Propagation of uncertainties in unfolding procedures, Nucl. Instrum

Meth. A 476, pp. 230-241 (2002).

57. The MathWorks GmbH Schürmattstrasse 6+8, Gümligen, Switzerland.

58. C.C. Braga, M. S. Dias, Application of neural networks for unfolding neutron

spectra measured by means of Bonner spheres, Nucl. Instr. and Meth. A 476,

pp. 252 –255 (2002).

59. Prof. Armando Foglio Para, Polytechnic of Milan (private communication).

60. N. Metropolis and S. Ulam, The Monte Carlo method, Journal of the American

Statistical Association, 44 (247), pp. 335-341 (1949).

61. A. Fassò, A. Ferrari, J. Ranft and P. Sala: New Developments in Fluka Modelling

of Hadronic and EM Interactions. Proc. of Third Workshop on Simulating

Accelerator Radiation Environments (SARE3). KEK Tsukuba, Japan 7-9 May

1997, Ed. H. Hirayama, KEK Proceedings 97-5, pp. 32 (1997).

62. A. Fassò, A. Ferrari, J. Ranft and P. Sala, Electron-Photon Transport in FLUKA:

Status. Proc. of Advanced Monte-Carlo for Radiation Physics, Particle Transport

Simulation and Applications. Lisbon, Portugal, 23-26 October 2000. Eds. A.

Kling, F. Barao, M. Nakagawa, L. Tavora, P. Vaz, Verlag Springer, pp. 159-164

(2001).

63. A. Fassò, A. Ferrari, J. Ranft and P. Sala, FLUKA: Status and Prospective for

Hadronic Applications. Proc. of Advanced Monte-Carlo for Radiation Physics,

Particle Transport Simulation and Applications, Lisbon, Portugal, 23-26 October

2000. Eds. A. Kling, F. Barao, M. Nakagawa, L. Tavora, P. Vaz, Verlag Springer,

pp. 955-960 (2001).

64. International Organization for Standardization, Reference neutron radiations -Part

2: Calibration fundamentals of radiation protection devices related to the basic

quantities characterizing the radiation field, ISO 8529-2 (2000).

65. International Organization for Standardization. X and gamma reference radiation

for calibrating dosemeters and doserate meters and for determining their

response as a function of photon energy- Part 3: Calibration of area and


172

personal dosemeters and the measurement of their response as a function of

energy and angle of incidence, ISO 4037-3 (1999).

66. International Organization for Standardization. Reference neutron radiations - Part

3: Calibration of area and personal dosimeters and determination of response as

a function of energy and angle of incidence, ISO 8529-3 (1998).

67. International Organization for Standardization. Reference Neutron Radiations:

Part 2: Calibration fundamentals of radiation protection devices related to the

basic quantities characterizing the radiation field. ISO 8529-2 (2000).

68. International Organization for Standardization. Reference Neutron Radiations:

Part 1: Characteristics and Methods of production. ISO 8529-1 (2001).

69. H. J. Brede, M. Cosack, G. Dietze, H. Gumpert, S. Guldbakke, R. Jahr, M.

Kutscha, D. Schlegel-Bickmann and H. Schölermann, The Braunschweig

accelerator facility for fast neutron research. I. Building design and accelerators,

Nucl. Instrum. and Meth. 169 (3), pp. 349-358 (1980).

70. H. Klein, H. J. Barrenscheen, G. Dietze and B. R. L. Siebert, The Braunschweig

accelerator facility for fast neutron research II. Data acquisition and analysis,

Nucl. Instrum. and Meth. 169, pp. 359-367 (1980).

71. S. Guldbakke, S. Dietz, H. Kluge and D. Schlegel, PTB neutron fields for the

calibration of neutron sensitive devices, Strahlenschutz: Physik und Messtechnik,

Vol. Ieds. W. Koelzer and R. Maushart, pp. 240-247 (Cologne: Verlag TUV

Pheinland) (1994).

72. A. Bol, P. Leleux, P. Lipnik, P. Macq and A. Ninane, A novel design for a fast

intense neutron beam, Nucl. Instr. and Meth 214, pp. 169-173, (1983).

73. C. Dupont, P. Leleux, P. Lipnik, P. Macq and A. Ninane, Study of a collimated

fast neutron beam, Nucl. Instr. and Meth A 256, pp. 197-206, (1987).

74. I. Slypen, V. Corcalciuc, A. Ninane and J.P. Meulders, Charged particles

produced in fast neutron induced reactions on 12C in the 45-80 MeV energy range,

Nucl. Instr. and Meth A 337, pp. 431-440, (1994).

75. H. Schuhmacher, H. J. Brede, V. Dangendorf, M. Kuhfuss, J. P. Meulders, W. D.

Newhauser and R. Nolte H., Quasi-monoenergetic neutron beams with energies

from 25 to 70 MeV, Nucl. Instr. and Meth A 421, pp. 284-295 (1999).


173

76. H. Schuhmacher and W.G. Alberts, Reference Neutron Fields With Energies up to

70 MeV for the Calibration of Radiation Protection Instruments, Radiat. Prot.

Dosim. 42, pp. 287-290 (1992).

77. C. Birattari, A. Ferrari, M. Höfert, T. Otto, T. Rancati and M. Silari, Recent

Results at the CERN-EC High-energy Reference Field Facility. In: Proc. Satif-3.

Shielding Aspects of Accelerators, Targets and Irradiation Facilities, Sendai,

Japan, 12–13 May 1997. NEA/OECD, pp. 219–234 (1998).

78. A. Mitaroff and M. Silari, The CERN-EU high-energy Reference Field (CERF)

facility for dosimetry at commercial flight altitudes and in space, Radiat. Prot.

Dosim. 102, pp. 7-22, (2002).

79. A. Aroua, J. Bourges, R.C. Raffnsøe: Beam-Monitoring and Data-Logging for the

H6 Experiment. CERN/TIS-RP/TM/93-20 (1993).

80. R.C. Raffnsøe, Instrumentation for beam experiments in North Hall,

TIS-RP-DC/TM/92-27 (1992).

81. T. Otto and M. Silari, The July/August 1996 run at the CERN-CEC reference

facility, CERN/TIS-RP/TM/96-25 (1996).

82. T. Otto and M. Silari, The May 1996 run at the CERN-CEC reference field

facility, CERN/TIS-RP/IR/96-15 (1996).

83. S. Guldbakke, (private communication).

84. D. Schlegel, TARGET User’s Manual. Report PTB-6.41-1998-1 (Braunschweig:

Physikalisch- Technische Bundesanstalt), (1998).

85. D. Schlegel, (private communication).

86. R. Böttzer, S. Guldbakke, H. Klein, H. Schölermann and H. Schuhmacher.

Problems associated with the production of monoenergetic neutrons, Nucl.

Instrum. and Meth. A 282, pp. 358-367 (1989).

87. A.V. Alevra, M. Cosack, J.B Hunt, D.J Thomas and H. Schraube, Experimental

determination of the response of four Bonner sphere sets to monoenergetic

neutrons (II), Radiat. Prot. Dosim. 40 (2), pp. 91-102 (1992).

88. B. Wiegel, A.V. Alevra and B.R.L. Siebert: Calculations of the Response

Functions of Bonner Spheres with a Spherical 3He Proportional Counter Using a

Realistic Detector Model. PTB- N-21 (1994).

89. C. Birattari, G. Manfredi and M. Silari, Scanning procedures for calibration of

radiation detectors, Rev. Sci. Instrum. 66 (8), pp. 4198-4203 (1995).


174

90. C. Birattari, A. Ferrari, C. Nuccetelli, M. Pelliccioni and M. Silari, An extended

range neutron rem counter, Nucl. Instrum. Meth. A 297, pp. 250 (1990).

91. C. Birattari, A. Esposito, A. Ferrari, M. Pelliccioni and M. Silari, A neutron

survey-meter with sensitivity extended up to 400 MeV, Radiat. Prot. Dosim. 44,

pp. 193. (1992).

92. C. Birattari, A. Esposito, A. Ferrari, M. Pelliccioni and M. Silari, Calibration of

the neutron rem counter LINUS in the energy range from thermal to 19 MeV,

Nucl. Instrum. Meth. A 324, pp. 232 (1993).

93. C. Birattari, A. Esposito, A. Ferrari, M. Pelliccioni, T. Rancati and M. Silari, The

extended range neutron rem counter LINUS: overview and latest developments,

Radiat. Prot. Dosim. 76, pp. 135 (1998).

94. R. Nolte (private communication).

95. R. Nolte, M. S. Allie, P. J. Binns, F. Brooks, A. Buffler, V. Dangendorf, J. P.

Meulders, F. Roos, H. Schuhmacher and B. Wiegel, High-energy neutron

reference fields for the calibration of detectors used in neutron spectrometry,

Nucl. Instr. and Meth A 476 (1-2), pp. 369-373, (2002).

96. Prof. C. Birattari (private communication).

97. A.V. Alevra, Accurate neutron fluence measurements using Bonner spheres. In:

Reactor Dosimetry, ASTM STP 1228. Eds. H. Farrar IV, E.P. Lippincott, J.G.

Williams and D. W. Wehar, pp. 290-299 (Philadelphia, PA:American Society for

Testing and Materials) (1994).

98. Mrs. Jo Carmody, Sales Account Manager, Centronic Ltd. Centronic House, King

Henry’s Drive, Croydon CR9 OBG, England (private communication).

99. S. Agosteo, C. Birattari, E. Dimovasili, A. Foglio Para, M. Silari, L. Ulrici and

H. Vincke, Neutron production from 40 GeV/c hadrons on thin copper, silver and

lead targets in the angular range 30º – 135º, (in press in NIM B).

100. http://www-nds.iaea.or.at/ndsout/endf2784.endf

101. S. Agosteo, C. Birattari, A. Foglio Para, M. Silari and L. Ulrici, Neutron

measurements around a beam dump bombarded by high energy protons and lead

ions, Nucl. Instrum. and Methods A 459, pp. 58-65 (2001).

102. D.H. Perkins, Introduction to high energy physics, Addison-Wesley, 3rd Edition,

Reading, MA, (1987).


175

103. S. Agosteo, C. Birattari, A. Foglio Para, A. Mitaroff, M. Silari and L. Ulrici, 90º

neutron emission from high energy protons and lead ions on a thin lead target,

Nucl. Instrum. Meth. A 476, pp. 63-68 (2002).

104. S. Agosteo, C. Birattari, A. Foglio Para, L. Gini, A. Mitaroff, M. Silari and

L. Ulrici, Neutron production from 158 GeV/c per nucleon lead ions on thin

copper and lead targets in the angular range 30-135°, Nucl. Instrum. Meth. B

194, pp. 399-408 (2002).

105. Particle data group, Review of particle physics, Eur. Phys. J. C 15,

pp. 1-878 (2000).

106. V. Manzari et al., Experiment NA57 at the CERN SPS, J. Phys. G: Nucl. Part.

Phys. 25, pp. 473 (1999).

107. V. Manzari et al., Silicon pixel detectors for tracking in NA57, Nucl. Phys.

A661 (1999).

108. The ALICE collaboration, Technical Proposal, CERN/LHC95-71 (1995).

109. E. Gschwendtner, A. Mitaroff and L. Ulrici, A new calibration of the PIC

monitor in H6, CERN Internal Report TIS-RP/IR/2000-009 (2000).

110. H. Vincke, SC-RP, CERN (private communication).

111. S. Agosteo, E. Dimovasili, A. Fassò and M. Silari, The response of a Bonner

sphere spectrometer to charged hadrons, Proc. of the NEUDOS9 conference,

Delft, The Netherlands, 28 Sept - 3 Oct 2003, Radiat. Prot. Dosim. 110, pp.161-

168 (2004).

112. E. Gschwendtner, A. Mitaroff and T. Otto, Performance tests of the CERF

beam monitor, CERN/TIS-RP/TM/2000-07 (2000).

113. A. Mitaroff and T. Otto, Recombination measurements of the CERF beam

monitor for a 120 GeV/c hadron beam, CERN/TIS-RP/IR/2002-028 (2002).

114. R. G. Jaeger and W. Hübner, Dosimetrie und Strahlenschutz, 1974 (in German).

115. P.R. Bevington and D. Keith Robinson, Data Reduction and Error Analysis for

the Physical Sciences, McGraw-Hill, 2nd Edition, (1992).

116. K. Elsener, M. Heilmann and M. Silari, Verification of the calibration factor of

the CERF beam monitor, CERN/TIS-RP/TM/98-22 (1998).

117. I. Efthymiopoulos, A. Mitaroff and M. Silari, Efficiency measurements of the

Trigger4 beam monitor in H6, CERN/TIS-RP/TN/2002-027 (2002).

118. G. Knoll, Radiation detection and measurement, (John Wiley), 3rd Ed. (2000).


176

119. W. Leo, Techniques for nuclear and particle physics experiments, Springer

(1987).

120. P. K. F Grieder, Cosmic rays at earth : researcher's reference manual and data

book, North-Holland Amsterdam 2001.

121. K. O’Brien, F. Friedberg, F.E. Duke, L. Snyder, E.B. Darden, H.H. Sauer,

The Exposure of Aircraft Crews to Radiations of Extraterrestrial Origin,

Radiat. Prot. Dosim. 45, pp. 145-162 (1992).

122. G. Reitz, K. Schnuer, K. Shaw (Eds.), Proceedings of the Workshop on

Radiation Exposure of Civil Aircrew, Luxembourg, 25–27 June 1991, Radiat.

Prot. Dosim. 48 (1993).

123. P. Goldhagen, Overview of Aircraft Radiation Exposure And Recent Er-2

Measurements, Health Physics. 79(5), pp. 526-544, (2000).

124. J.W. Wilson, J.E. Nealy, F.A. Cucinotta, J.L. Shinn, F. Hajnal, M. Reginatto,

P. Goldhagen, NASA Technical Paper 3524 (1995), National Technical

Information Service, Springfield, Virginia.

125. S. Roesler, W. Heinrich, H. Schraube, Calculation of Radiation Fields in the

Atmosphere and Comparison to Experimental Data, Radiat. Res. 149, pp.87-97,

(1998).

126. A. Kurochkin, B. Wiegel, B.R.L. Siebert, Study of the Radiation Environment

Caused by Galactic Cosmic Rays at Flight Altitudes, at the Summit of the

Zugspitze and at PTB Braunschweig, Radiat. Prot.Dosim. 83, pp. 281-291,

(1999).

127. J.W. Wilson, Overview Of Radiation Environments And Human Exposures,

Health Phys. 79 (5), pp. 470-494 (2000).

128. R. A. Nymmik, M. I. Panasyuk and A. A. Suslov, Galactic cosmic ray flux

simulation and prediction, Advances in Space Research, 17 (2), pp. 19-30

(1996).

129. N.M. Sobolevsky, Space Radiation Shielding and Environment Applications

with the SHIELD Transport Code, Geant4 Space Users Forum, 20-22 January

2003 ESTEC, Noordwijk, The Netherlands.

130. http://liftoff.msfc.nasa.gov/rsa/mir.html

131. International Commission on Radiological Protection, http://www.icrp.org


177

132. International Commission on Radiation Units and Measurements,

http://www.icru.org

133. International Commission on Radiation Units and Measurements, Fundamental

Quantities and Units for Ionizing Radiation, ICRU Rep. 60, Bethesda,

Maryland (1998).

178

Acknowledgements

The work presented in this thesis has been carried out in the framework of the

Doctoral Student Programme at CERN in the SC/RP group in collaboration with the

University of Lausanne (UNIL) and the Swiss Federal Institute of Technology in

Lausanne (EPFL).

I wish to thank Professor Aurelio Bay in the Physics Department of EPFL for

giving me the possibility to get involved as a Doctoral student there. I also thank him

for the support and the freedom he gave me to pursue my diploma in my own way. I

would like to express my gratitude to Professor Jean-Francois Valley at the Institute

of Applied Radiophysics (IRA) in Lausanne for accepting me as his student and for

being always helpful, discrete and extremely kind. I would like to specially thank my

CERN supervisor Dr. Marco Silari for his support since I first came to CERN in 1999

as a summer student. I thank him also for offering me the privilege to spend three

years in the fascinating international environment of CERN. I have to extend my

thanks to the leader of the Radiation Protection Group, Dr. Hans Menzel and to the

Division leader Dr Helmut Schonbacher. I appreciate a lot their trust in me that was

expressed by their discrete presence, their support and help to extend my contract up

to three years.

I feel blessed for having met Professor Pavlos Ioannou in the Physics

Department of the University of Athens. I will always be indebted to him for his

unconditional support and encouragement. He has taught me much physics and an

entire philosophy of scientific thinking. I would have never arrived at this goal

without his constant assistance and confidence. My dear teacher and friend Alberto

Fassò has helped me in solving so many problems that I cannot list here. Above all, I

am grateful to him for being my spiritual guide and for reminding me many times to

keep ‘chin up’. He showed me what kind of physicist I want to become.

Special thanks are addressed to Professors Claudio Birattari and Armando

Foglio Para for providing the unfolding codes and for always making my stay in

Milan very pleasant. Many warm thanks go to Stefano Agosteo for the fruitful

collaboration and discussions. I wish to thank Alfredo Ferrari, Thomas Otto, Stefan

Roesler, and Professor Guizeppe Battistoni for their kind help and stimulating

discussions in many practical issues. I am grateful to Daniel Perrin, Michel Renou,

179

Michel Pangalo and Hubert Muller for always solving any mechanical and

electronics problem that occurred just before the experiments. I feel extremely

grateful to Matteo Magistris for his friendship and for his patience to help me with my

first steps in FLUKA.

My special thanks go to Mario Mueller for his kindness and friendship and for

spending with me many hours on FLUKA discussions. Many thanks also go to

Helmut Vincke for always finding time to answer my questions in his special way. I

would also like to thank Sabine Mayer for sharing overnight shifts with me during my

first CERF runs. I specially thank Anne-Laure Perrot for her kind help in writing the

French abstract of this thesis. I have been gifted with friendship and support from

many people during the course of this thesis. My greatest and sincere thanks to

Emmanuel Tsesmelis, Luisa Ulrici, Marta Sans Merce, Luca Reina, David

Emschermann, and John Antonakis for sharing joys and stress and for keeping always

the spirits up. I owe a HUGE ‘thank you’ to Dr Emmanuel Tsesmelis for his

friendship as well as for his unconditional and constant support through all these

years. I appreciate a lot the fact that despite his overloaded schedule he carefully read

and corrected the manuscripts of my thesis.

I gratefully acknowledge the financial support of IRA through my engagement

as University assistant. I am also indebted to Mrs Jo Carmody of Centronics Ltd,

U.K., for providing me confidential data about the 3He counter used in this work.

The last words are the most heartfelt and they are reserved for my family, for

providing me with the upbringing, opportunities and support that motivated me and

allowed me to pursue my interests. Very special thanks for their unfailing love and

constant encouragement. I would like to express my love to my fiancée Spiros

Kokolakis, for all the years together, for sharing worries and joys, for being a friend

and a companion.

180

CURRICULUM VITAE

EVANGELIA DIMOVASILI

Personal information

Nationality: Greek

Date of birth: 08.06.1974

Place of birth: Athens

Education

2001- 2004 Doctoral Thesis, EPFL, UNIL and CERN (Radiation

Protection Group). Measurements and Monte Carlo

calculations with the Extended- Range Bonner Sphere

Spectrometer at High- Energy Mixed Fields.

1998- 2001 Masters studies in Medical Physics, University of

Heraklio, Greece M. Sc. Diploma with grade

‘Distinction’.

1992- 1998 Studies in the Physics Department of University of

Athens, Greece, B. Sc. Diploma in Physics with grade

‘Very good’.

1991 Diploma of Proficiency in English, Univ. Ann Arbor,

Michigan.

1986- 1992 High School, Amfilohia, Greece

Apolitirio Diploma with grade ‘Excellent’.

181

Research and scientific activities Jan 2001- Aug 2001 Investigation of effects of TiN coating on the hydrogen

permeability of stainless steel and Ni membranes (in the

framework of a collaboration between the Departments

of Physics of the Universities of Athens and

St- Petersbourg, in the field of material science).

Oct 2000- May 2001 Masters thesis, Conceptus Radiation dose and risk from

chest screen-film radiography, performed at the

University Hospital of Heraklio, Crete.

Sept 1999- Oct 2000 •Studies on the biological effects of the E/M radiation

of 50/60 Hz.

•Quality control of Computed Tomograph, SPET

and film-developing machines.

•Medical image analysis (ANALYSE software).

July 1999- Sept 1999 Summer student at CERN with the TIS/RP group

Determination of calibration factors for gamma,

neutron and muon detectors.

May 1999- May 2001 Participation in the data analysis (using the STAPRE

code) for experiments of fission cross-section

measurements for 232Th, 235U, 238U and 239Pu (in the

framework of Greek- Russian collaboration).

Jan 2000 –today Member of the nTOF European Collaboration for High-

Resolution Measurements of Neutron Cross Sections

between 1 eV and 250 MeV.

Sept 1997- Feb 1998 Diploma thesis, Quality control of the computed

tomography and correlation studies between optical

density and Hounsfield units, performed at the

Aretaieion University Hospital of Athens.

182

Participation in Workshops/Conferences

• Ninth Symposium on Neutron Dosimetry, Delft, The Netherlands,

28 September – 3 October 2003.

• Workshop on Radiation Protection Issues Related to Radioactive Ion Beam

Facilities, CERN, Geneva (Switzerland), October 30 - November 1, 2002.

• Special Workshop of Marie Curie Fellows on Research and Training in

Physics and Technology, CERN, Geneva (Switzerland), October 3 - 4, 2002.

• Meeting of the Athens Dentists Association on the topic

‘Radiation Protection in Dental Radiology’, Athens, 30 January 1999.

• Fourth Hellenic Laser Congress under the auspices of the Greek Scientific

Association of Laser Applications in Medicine, Athens, 4 December 1998.

PUBLICATIONS

1. S. Agosteo, E. Dimovasili, A. Fassò and M. Silari

The response of a Bonner Sphere Spectrometer to charged hadrons, Radiat.

Prot. Dosimetry (110), pp.161-168 (2004).

2. S. Agosteo, C. Birattari, E. Dimovasili, A. Foglio Para, M. Silari, L. Ulrici

and H. Vincke, Neutron production from 40 GeV/c hadrons on thin copper,

silver and lead targets in the angular range 30º– 135º, Nucl. Instr. and Meth.

B, (in press).

3. C. Birattari, E. Dimovasili, A. Mitaroff and M. Silari, Calibration and latest

developments of the Extended range Bonner Sphere Spectrometer

(under preparation).

4. T.G. Kazantzeva, Yu.N. Koblik , V.P. Pikul, A.V. Hugaev, B.S. Yuldashev,

P.Ioannou, E.Dimovasili, Secondary electron emission from metal foils

formed by fission fragments (to be published).

5. J.Damilakis, K.Perisinakis, P.Prassopoulos, E.Dimovasili, H.Varveris,

N.Gourtsoyiannis Conceptus Radiation dose and risk from chest screen-film

radiography, European Radiology 13(2), pp. 406-412, (2003).

6. Abramovsky V.A.E., Dimovasili E, Ioannou P. Neutron - induced nuclear

fission cross- sections of 232Th, 235U, 238U and 239Pu from 1 MeV to 200 MeV

183

in quark- gluon model. LII Meeting on Nuclear Spectroscopy and Nuclear

structure, Nucleus 2002, MSU Moscow, pp. 239 (2002).

7. A.V. Khugaev, Yu.N. Koblik., V.P. Pikul, P. Ioannou, E. Dimovasili

About total kinetic energy distribution between fragments of binary fission

In Proceedings of the conference: II Eurasian Conference on Nuclear Science

and its Application , Almaty (Kazakhstan), September 16-19, 2002

Vol. I, pp. 159-162, (2003).

8. A.V. Khugaev, Yu.N Koblik , G.A. Mkrtchan, B.S Yuldashev, P. Ioannou, E.

Dimovasili, About some generalization of calculation algorithm of three-

dimensional imagesin emission tomography. In Proceedings of the conference:

II Eurasian Conference on Nuclear Science and its Application, Almaty

(Kazakhstan), September 16-19, 2002, Vol. III, pp. 11-16, (2003).

9. A.V. Khugaev, Yu.N. Koblik , G.A. Mkrtchan, B.S Yuldashev, P. Ioannou, E.

Dimovasili, Mathematical modeling of three-dimensional images in emission

tomography In Proceedings of the conference: II Eurasian Conference on

Nuclear Science and its Application, Almaty (Kazakhstan), September 16-19,

2002, Vol. III, pp. 119-125 (2003).

10. A. Mitaroff, E. Dimovasili, S. Mayer, C. Birattari, B. Wiegel, M. Silari, H.

Aiginger, Kalibrierung und Experiment eines Bonnerkugel Spectrometer mit

erweitertem Messbereich (Calibration and experiment of an extended range

Bonner sphere spectrometer), Proceedings of the Conference: Strahlenschutz

fur Mensch und Gesellschaft in Europa von morgen Gmuden, 17-21

September 2001.

11. I.E.Gabis, V.A.Dubrovsky. E.A.Denisov, E.Dimovasili, P.Ioannou,

T.N.Kompaniets, A.A. Kurdyumov, K. Ja.Polonsky, I.A.Khazov, Hydrogen

Permeability of Titanium Nitride, International Conference: Interaction of

Hydrogen Isotopes with Structural Materials IHISM'01, Sarov. Book of

abstracts, pp. 75-82, (2001).

12. E.Denisov, E. Dimovasili, P.Ioannou, T.Kompaniets, A.Kurdyumov

Interaction of Molecular Hydrogen with Solid Surfaces in the Lack of

Temperature Equilibrium Gas-Solid, International Conference: Interaction of

Hydrogen Isotopes with Structural Materials IHISM'01, Sarov.

Book of abstracts, pp.170-174, (2001).

13. P.Kipouros,A.Peris, P. Papagiannis , E.Dimovasili and C. Antypas

Monitoring radiotherapy beam characteristics using image processing

184

software, VI International Conference on Medical Physics,Patras, Greece

1999, Physica Medica, Volume XV, N. 3.

TECHNICAL NOTES AND INTERNAL REPORTS

1. E. Dimovasili, D. Macina and M. Oriunno, Energy Deposition in the

Window of the Roman Pot, CERN TS- LEA Note (2004),

(under preparation).

2. E. Dimovasili, A. Ferrari, M. J. Mueller and M. Silari, A proposal for

upgrading the CERF facility for space applications, CERN Technical Note

(2004).

3. E. Dimovasili, M. Magistris and M. Silari, Inter-comparison of the CERF

beam monitors, CERN Technical Note TIS-2003-016-RP-TN (2003).

4. E. Dimovasili and M. Silari, Beam and reference field monitoring during

the 2002 CERF runs, CERN Technical Note TIS-2002-033-RP-TN (2002).

5. E. Dimovasili, S. Mayer, A. Mitaroff and M. Silari, HANDI TEPC

measurements during the 1999, 2000 and 2001 CERF runs, CERN

Technical Note TIS-RP/TN/2002-020 (2002).

6. S. Mayer, T. Otto and E. Dimovasili, Further Investigations of the

Recombination Chamber REM-2 as a Mixed Field Dosimeter at CERF in

October 200, CERN Technical Note TIS-RP/TN/2002-06 (2002).

7. S. Mayer, M. Zielczynski, F. McLay, E. Dimovasili and T. Otto,

Measurements with a recombination chamber made at the CERN-EC high

energy reference field CERF, in August 2001, CERN Technical Note TIS-

RP/TN/2001-03 (2002).

8. The n_TOF Collaboration, Study of the Background in the Measuring

Station at the n_TOF Facility at CERN: Sources and Solutions,

CERN/INTC 2001-038.

9. The n_TOF Collaboration, Measurements of Fission Cross Sections for the

Isotopes relevant to the Thorium Fuel Cycle, CERN-INTC-2001-025.

Date post:	23-Jul-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

MEASUREMENTS AND MONTE CARLO CALCULATIONS WITH … · Institut de radiophysique appliquée SECTION...

Documents