NEUTRON AND GAMMA RADIATION EFFECTS

ON THE VISCOELASTIC BEHAWOUR OF

POLY(ARYL ETHER ETHER KETONE)

Captain Danny J.Y.S. Pagé, CD,B.Eng

Submitted to the Department of

Chemisw and Chernical Engineering as a requirement for the degree of
Master of Engineering in Chernical and Materials Engineering

Royal Military College of Canada - Collège militaire royal du Canada

Kingston, Ontario

7 May 1997

Copyright O 1997, Danny J.Y.S. Pagé, DND

1*1 National Library
of Canada
Bibliothèque nationale
du Canada
Acquisitions and Acquisitions et
Bibliographic Services setvices bibliographiques
395 Wellington Street 395. rue Wellington
Ottawa ON K I A O N 4 Ottawa ON K1A ON4
Canada Canada
Your fi& votre réfdrence

Our tüe Nom dfdmm¶

The author has granted a non-
exclusive licence dowing the
National Library of Canada to
reproduce, loan, distnbute or seIl
copies of this thesis in microfom,
paper or electronic formats.
L'auteur a accordé une licence non
exclusive permettant à la
Bibliothèque nationale du Canada de
reproduire, prêter, distribuer ou
vendre des copies de cette thèse sous
la forme de microfiche/fïlrn, de
reproduction sur papier ou sur format
électronique.

The author retains ownership of the L'auteur conserve la propriété du

copyright in this thesis. Neither the
thesis nor substantial extracts fi-om it
may be printed or otherwise
reproduced without the author's
permission.
droit d'auteur qui protège cette thèse.
Ni la thèse ni des extraits substantiels
de celle-ci ne doivent être imprimés
ou autrement reproduits sans son
autorisation.
Je dédie cette thèse à mes parents et ma soeur.
 ABSTRACT

The effects of combined neutron and gamma radiation on the viscoelastic behaviour of
two industrial semi-crystalline PEEK grades (VICTREX 150P and 450P) were investigated.
Tensile test samples were processed on an ENGELS5 injection moulder, then irradiated in a
reactor pool environment, agahst the reactor vesse1 wali of the SLOWPOKE-2facility at RMC,
for exposures resulting m dosages ranghg fkom 10 kGy to 320 kGy. First, the rnorphology of the
irradiated resins was characterized by differential s c h g calorimetry @SC), X-ray m a c t i o n
and density meamements. Secondly, the viscoelastic behaviour of the material was studied at
three thermodynamic states, m order to define the nature of the structural damage suffered by
PEEK, by comparing t s degree of molecular motion with the dose received. Therefore, standard
mechanical testing was performed at room temperature, stress relaxation near the glass transition
temperature(T$, at 140°C, and melt viscosity above the melting point (T& at 350 OC.

Results confnmed that PEEK is highly resistant to radiation, but also demonstrated that
PEEK degrades faster under a combined neutron and gamma flux, in a reactor pool environment,
t h observed in previous work under electron beam or gamma afone. The density increased
slightly over the range of irradiation (450P: 1296 to 1299 kg/m3; 150P: 1300 to 1303 kg/m3),
while the crystallmty (450P: 24 %, 150P: 27 %), and the thermal properties, like Tgand Tm ,
remained d e c t e d . Therefore, the hcreases in density were attriibuted to transfomations
occtmhg mostly m the amorphous phase, which is also consistent with previous work. The
viscoelastic behaviour analysis provided the following damage assessnent for PEEK. the
amoqhous phase is subject to both crosslinking and chah scission; the eEects of chain scission,
on the viscoelastic behaviour of PEEK, are predomhant for doses higher than 100 kGy; tie-
molecules, at the crystalline and amorphous phase mterface, are more sensitive and are proue to
chin scission under radiation; and, crosslinkmg occurs predoxnhantly near chain-ends. OveraU,
this research confirmed PEEK as a good candidate material for applications in nuclear reactor
pool environments and for the disposal of radioactive waste.
 1 thank the personnel nom the Chemistry and Chernical Engineering Departrnent at
RMC for their pro fessionalism and dedication to a job weU done. Without the assistance
and patience of many staff and fiiends, this thesis, dong with its research work, would
have not been possible.

Special thanks to my thesis supervisors, Dr. Hughes Bonin and Dr. Van Tarn Bui,
who ûusted my initiative and provided me with sound guidance when required

Special thanks to the following technicians, Mrs. Kathy Neilsen, Mr. John
Perreault, Mr. Bob Whitehead, Mr. Helmut Weiland, Mr. James Irving and Mrs. Lynn
Keams, for their assistance and cooperation.

Special thanks to Captain Jean-François Legault and Major Dave Gosseh for their
participation in some fniitfùl discussions.

And finaIly, special thanks to my cornmittee members, Dr. M. Erki, Dr. K.

Machildon, Dr.P.J. Bates and my supervisors, who reviewed, challenged and accepted my
work as a fùltillment of the requirement for the obtainment of the degree o f Master of
Engineering in Chemical and Materials Engineering.

...
lll
 Staff Officer, Canadian Forces Comunications Command,
Ottawa, Mar 88-Jul90

QuaIifications : Basic Officer Training Course, CFB Chilliwack, 1983

Second Language Training (English), BFC Saint-Jean, 1984

Communications and Electronics Engineering Officer Course

(Part 1), CFB Kingston. 1985

Basic Parachutist Course, CFB Edmonton, 1986

Young Officer Tactics Course, CFB Kingston, 1986

Communications and Electronics Engineering Officer Course

(Part II), CFB Kingston, 1987

Land Basic Signal Officer Course, CFB Kingston, 1988

Promotion to the rank of Lieutenant, 1988

Promotion to the rank of Captain, 1990

Intermediate Signal Officer Course, CFB Kingston, 1993

Intermediate Tactics Course, CFB Gagetowa, 1994

Space Mission Analysis and Design Course, CMR Saint-Jean, 1995

Abbreviations and acronyms

ASTM Amencan standard test method

DSC differential scanning calorimeky
rn Fourier transfarmed in£rared
CY Gray (J/W
LET hear energy transfer
LEU low enrichecl Uranium
LDPE low density polyethylene
MPa Mega Pascal
MW molecular weight (MW)
MWD molecular weight distniution (Mw/Mn)
NAA neutron activation analysis
Pa Pascal (N/rn2)
PEEK poly (aryl ether ether ketone)
PMMA poly (methyl methacrylate)
PSU polysulfone
rad physical dose unit (6.24 x 1 oL3eV/g)
S second
SP gr speci.fïc gravity relative to water
TEM transmission electron microscopy
150P lower molecular weight VICTREX PEEK grade
450P higher molecular weight VICTREX PEEK grade

vii
 TABLE OF CONTENTS
..
Abstract............................................................................................................... II
...
Acknowledgments................................................................................................ 111

Vita .............................................................................................................. iv
Symbols & Abbreviations .................................................................................... vi
...
Table of Contents................................................................................................. viii

ListofIllustrations............................................................................................... xii

m oDUCTION.........................................................................

CHAPTER 1 .STATE OF THE ART

. .
1.1 In the Beginning...................................................................
1.2 Previous Work on the Irradiation of PEEK...................................
1.3 Scope of the Present Work ......................................................................

CHAPTER 2 .THEORY
2.1 Radiation............................................................................
2.1.1 Interaction with matter...................................................
2.1.2 Energy transfer............................................................
3.1.3 Doses and units............................................................................
2.2 Effects of Radiation on Polymers............................................................

2.2.2 Induced chemical changes ...............................................

2.2.3 Impact on macroscopic properties...............................................
2.2.4 Degradation.................................................................................
2.3 Poly(Ether Ether Ketone) (PEEK)...........................................................
2.3.1 Chemical characteristics...............................................................
2.3.2 Physical properties.......................................................................
2.3.3 Processing...................................................................................

...
vlll
CHAPTER 3 .METHODOLOGY
0utli.e.....................................................................................................
3.1.1 Production ofthe test samples......................................................
..
3.1.2 Expainlental conditions...............................................................
. .
3.1.3 Charactenzation...........................................................................
. .
Injection Moulding...................................................................................
. .
3.2.1 Description...................................................................................
3.2.2 Apparatus.....................................................................................
3.2.3 Procedure.....................................................................................
Neutron and Gamma Irradiation...............................................................
. .
3.3.1 Description...................................................................................
3.3.2 Apparaîus.....................................................................................
3.3.3 Procedure.....................................................................................
Density by Water Displacement................................................................
. .
3.4.1 Description...................................................................................
3 .4.2 Apparatus.....................................................................................
3.4.3 Procedure.....................................................................................
Stress Relaxation at Constant Strain.........................................................
. .
3.5 .1 Description...................................................................................
3.5.2 Apparatus.....................................................................................
3.5.3 Procedure.....................................................................................
Standard Test for Tensile Properties .........................................................
. .
...................................................................................
3.6.1 Descnp~on
3.6.2 Apparatus.....................................................................................
3.6.3 Procedure.....................................................................................
.............................................................
Differential Scanning Calorimet~y
. .
3.7.1 Description...................................................................................
3 .7.2 Apparatus.....................................................................................
3.7.3 Procedure.....................................................................................
 X-ray Diffraction......................................................................................
.
3.8.1 Description ...................................................................................
3.8.2 Apparatus.....................................................................................
3.8.3 Procedure.....................................................................................
3.9 Rheological Analysis.................................................................................
. .
3.9.1 Description...................................................................................
3.9.2 Apparatus .....................................................................................
3.9.3 Procedure.....................................................................................

CHAPTER 4 .RESULTS AND DISCUSSION

4.1 Genera1.................................................................................................
4.2 Density.....................................................................................................
4.3 Glass Transition, Crystallization and Melting Temperatures......................
4.4 Crystallinity ........................................................................................
4.5 Tende Properties.....................................................................................
4.5.1 Young's modulus and strength..........................................
4.5.2 Maximum strain and energy at break ...................................
4.6 Stress Relaxation at Constant Strain.........................................................
4.7 MeIt Viscosity ...........................................................................................

CHAPTER 5 .SYNTHESIS AND ANALYSlS

5.1 Radiation Damage Assessment..................................................... 69
5.2 Role Played by the Morphology ................................................. 70
5.3 Error Analysis ........................................................................ 72

CONCLUSION................................................................................................... 76

RECOMMENDATIONS...................................................................................... 77

References............................................................................................................. 78
APPENDICES:

A. INJECTION MOULDING PARAMETERS

B. IRRADIATION PROFILE
C. DENSITY CALCLKATION
D. RELAXED YOUNG'S MODULUS CALCULAï?ON
E. STANDARD TENSILE TEST REPORT
F. DIFFERENTIAL SCANMNG CALORIMETRY REPORT
G. X-RAY DIFFRACTION REPORT
H. RHELOGICAL ANALYSIS REPORT
 LIST OF ILLUSTRATIONS

Figure 1.1: Chemical structure of poly (aryl ether ether ketone) (PEM) .....................3
Figure 1.2: Radiolysis mechanism of PEEK .............................................................. 6
Figure 1.3: Selected temperatures for viscoelastic tests...............................................8
Figure 2.1: PEEK crystal larnellar structure.............................................................. 22
Figure 3.1: SLOWPOKE-2irradiation set-up...........................................................31
Figure 3.2: Typical tensile test stress-strain cuve for a polymer...............................36
Figure 3.3 : Tende test bars dimensions.................................................................. -37
Figure 4.1 : Density as a function of dose for PEEK 1SOP.........................................45
Figure 4.2: Density as a function of dose for PEEK 450P......................................... 45
Figure 4.3 : PEEK molecular arrangement in its semi-crystalline state...................47
Figure 4.4: Typical DSC heat flow diagram for the processed PEEK.....................48
Figure 4.5: Glass transition temperature as a fùnction of dose for PEEK 150P..........50
Figure 4.6: Glass transition temperature as a h c t i o n of dose for PEEK 450P..........50
Figure 4.7: Crystalhation and melting temperatures as a function of dose for
PEEK 15OP...........................................................................................S l
Figure 4.8: Crystaiiization and melting temperatures as a function of dose for
PEEK 450P...................................................................................... 5 1
Figure 4.9: C~stallinityas a function of dose for PEEK 150P...................................53
Figure 4.10: CrystaUinity as a function of dose for PEEK 450P...................................53
Figure 4.11 : Stress-strain curve at 24OC and Smm/min strain rate for PEEK 150P.......55
Figure 4.12 Stress-strain curve at 24°C and 5 d m i n strain rate for PEEK 450P......55
Figure 4.13: Young's Modulus at 24°C as a function of dose for PEEK 150P.............57
Figure 4.14: Young's Modulus at 24OC as a function of dose for PEEK 450P.............57
Figure 4.15: htramolecular bond o c c h g in the PEEK molecular backbone ..........58
Figure 4.16: Tende strength at 24°C as a function of dose for PEEK 150P.................60
Figure 4.17: Tende strength at 24OC as a function of dose for PEEK 450P.................60
Figure 4.18: Maximum strain at 24°C as a function of dose for PEEK 150P................62
Figure 4.19: Maximum strain at 24°C as a fiinction of dose for PEEK 450P................62
Figure 4.20: Energy at break at 24°C as a function of dose for PEEK 150P...............63
Figure 4.2 1 : Energy at break at 24OC as a function of dose for PEEK 45OP...............63
Figure 4.22: Relaxed Young's modulus at 140°C as a function of dose for
PEEK f 5OP...........................................................................................-65
Figure 4.23: Relaxed Young's modulus at 140°C as a fûnction of dose for
. . .........................-65
PEEK 4SOP.......................................................,.. .... ..
Figure 4.24: Melt viscosity at 350°C and 100 s-' as a function of dose for
PEEK 150P...........................................................................................-68
Figure 4.25: Melt viscosity at 350°C and 100 s-' as a function of dose for
PEEK 450P.........................................,.................................................-68

LIST OF TABLE

Table 4.1: Power law factor (n) for both resins in fhction of dose.........................67
 INTRODUCTION

The aerospace, communications and nuclear industries rely increasingly on new

p o l y k c materials for advanced engineering applications demanding high resistance to
radiation. Among the most promishg polymeric materials is poly (aryl ether ether ketone)
(PEEK), a semi-crystalline therrnoplastic which is resistant to high temperatures in
addition to displayhg excellent mechanical properties and good stability in both chemically
active and radiation environments. These characteristics make PEEK an excellent
candidate material for nuclear applications such as waste disposal, reactor components and
telescopic tools.

The specÿil mechanical properties of polymers are driven by their viscoelastic

atmbutes. The viscoelasticity of a material c a . be characterized through a variety of
methods including tende, stress relaxation and rheological tests. The changes in the
viscoelastic properties of polymers exposed to aggressive environrnents such as radiation
are often an excellent indicator of the ability of polymeric materials to maintain the
mecbanical resistance required kom them throughout the lifetime of thek application The
viscoelasticity of a polymer is dependent on its molecular structure, defjned by chernical
bonds and chah length, but also by its morphology which is dete-ed by the cvstalline
and amorphous phase composition and interface characteristics.

The role played by the morphology in semi-crystalline PEEK, during radiation-

induced molecula. transformations, is not weil understood. Radiolysis descnbes those
chernical reactions occurring in the polymer as a result of irradiation('"'. Radiolysis is
dependent on the reactive species created by radiation but also by the fieedom of
movement of those reactants. Molecular movement hcreases with temperature as the
polymer goes through various thermodynamic transition state~(~'.
Molecular movement in
a polymer can therefore be used to study the role played by the morphology during
radiolysis, and also to assess the damage caused by radiation on the structure of the
poiymer.

Past research work has concentrated mostly on the study of the effects of electron
on PEEK among other thermoplastics. No research has
and gannna radiati~n''~)
bea~n(~-'~'
been carried out on the effects of fast and thermal neutrons on this materid, especially on
bs viscoelastic behaviour in the temperature region close to the glass transition
temperature. In order to assess molecular darnage and to better analyze the specific role
played by the amorphous phase, the crystalline phase and their interface during radiolysis,
the morphology of the PEEK resins were characterized using density measurements,
differential s c a . g calorhetry @SC) and X-ray difnaction. Then, the Mscoekstic
behaviour of the polymer was studied at three different thermodynamic states. The
methodology used for this part of the research consisted 6rst to perfom stress relaxation
testing just below Tg(140°C) to study the viscoelastic behaviour of the irradiated PEEK,
then, these tests were complemented with tende testing at room temperature to assess the
elastic and plastic behaviour of the material finally rheology tests were performed above
T,(35O0C) to assess changes in the viscous behaviour of PEEK.

This document is divided in five chapters. Chapter 1 reviews the research

performed so far on the effects of radiation on PEEK, and, presents the scope of this
thesis. Chapter 2 covers the important theoretical principles behind the interaction of
radiation with matter, its effect on polymers and the special characteristics that rnake
PEEK an engineering thermoplastic. Chapter 3 desmies the methodology used for the
experimental work and the basic principles applied with the various techniques used in this
research. Chapter 4 presents the results obtained for the characterization of the
morphology and the study of the viscoelastic behaviour of PEEK Each series of results,
related to the same studied material characteristic, is foilowed by a preluninary discussion.
Finally, Chapter 5 provides a synthesis of all the major discussion points and analyzes the
potential sources of emor.
 Chapter 1

STATE OP THE ART

1.1 In the Beginning ...

PEEK was designed in 1978 by Imperia1 Chemical Industries (ICI) as a tough high
temperature-resistant cable insulating material'? Since its commercialVation in 1981,
PEEK's excellent mechanical stabilRy at high temperatures coupled with its ease of
processing has made it a material of choice in a number of high technology applications in
the aerospace, automotive, electrical and chemical mdustries. PEEK is generally a tough,
semi-crystallinethennoplastic with the following chmical structure (Figure 1.1) (16):

Figure 1.1: Chemical structure of poly (aryl ether ether ketone) (PEEK).

The special properties of this polymer can be attriiuted to its chemical stnicture.
The aromatic ring is the domhant component in PEEK's composition, as shown above,
and is mostly responsible for the strength, the high temperature and radiation resistance of
the material? The ether bond provided by oxygen in the backbone contnbutes toughness
and fiexhility. The Iast feature, is the ketone side goup which provides some degree of
intennolecula. spacing, reducing both the density and glass transition temperature
 (Tg= 143OC), and also contnbuting to the toughness of the material by increasing the fiee
volume. The repetitive sequence in PEEK's molecular structure, alIows aystaliization to
O ccur, enhancing both the strength and the chernical resistance. Detailed characteristics of
PEEK are provided in Chaprer 2.

The potential of PEEK for use in the nuclear and space industries was identified
early and research work was initiated by Yoda in 1983 with the study of the effect of high
energy electron beam radiation on the structure and physical properties of the polymer'g'.
ïhis effort was the initial thrust on the nudy of the effects of radiation on PEEK m a d y in
four areas: morphology, t h e m 1 properties, gas evolution and mechanical properties. The
source of radiation most often us& was high energy electron beam radiation and on some
occasions, gamma radiation.

1.2 Previous Work on the Irradiation of PEEK

In the mid 80's, Temey and Slemp fkom NASA studied a number of polymeric
materials targeted for space applications. including PEEK composites'n. The scope of the
research was to shidy the effects of thermal cychg and electron radiation on the tensile
and shear properties of the studied materials. Materials were exposed to 1 MeV electrons
' rad/%(5xlos Gyh) for doses up to 1~ 1 0 ' 'rads (1 xlo8
at a maximum dose rate of 5x10
Gy). An inrrease in Young's rnodulus was observed for the AS/4 PEEK and graphite
composite. This observation confirmed Sasuga's et al. research done in 1984 ''L',who
studied the degradation of tensile properties of both morphous and semi-crystalline
PEEK, among other thermoplastics, exposed to high dose of electron beam radiation.
They observed that the elongation at break and the tensile strength decrease and the
Young's modulus tends to increase for doses up to 50 MGy.

In 1991, Hegazy et al. investigated the effect of gamma and electron b a r n

irradiation under vacuum on the gas evolution of several arornatic polymers at doses up to
25 MGy. The two types of radiation seemed to generate comparable damages to the
PEEK rnolecular structure. Crystalline PEEK yielded less gas under irradiation than
amorphous PEEK"~),and disphyeti better resistance to gamma radiation due to the
presence of crystais. CO2 and CO gas evolution showed that the ether and ketone groups
were the most affected by radiation, however the evohitioa of H2alSO attested for the
occurrence of crossiinking which is the creation of new intermolecular bonds. From this
work, Hegazy coocluded that aryl ether ketone linkages were seen as exhibiting a
protective effect aga& radiation(13)and proposed a radiolysis mechanism for irradiated
PEEK as shown in Figure 1.2. Furthermore, Hegazy et al. confirmecl, by ushg DSC,
that PEEK creates crosslinks under radiation below its glass transition temperature('?
Crosslinlsng was attri'buted predomkmtly to the amorphous phase between the crystal
, while Tm and the crystallization temperature (Tc)
lamellar structure, which increased T
were lowered.

Yoda obsemed that amorphous PEEK suffered enough defects to be unable to

crystallize after receiving doses of 50 MGy. Reduction in crystahity and of the rnelting
temperature were observed upon recrystallization after the material had been exposed to
an electron beam radiation dose higher than 10 MGy. Yoda also ob~erved(~),
following
electron beam irradiation at room temperature up to 50 MGy, that the average crystafite
size was reduced by 15% while the c-axh of the orthorhombic czystal ceii structure of
PEEK remained d d l y affected.

Vaughan et aL observed that PEEK aystals are slightly aEected up to 66 MGy.

At higher electron beam doses changes are so important that at 260 MGy only a small
fkaction of crystals can recrystallize(? Vaughan et aL proposed that the formation of
intermolecular crosslinks above 260 MGy can occur to such an extent that no rnelting is
possible and no endothemi can be detected on DSC. However, Maction patterns and
transmission electron microscopy (TEM) would stiiI indicate some local ordering in the
form of lamellae that could be associated with the crystailine structure of PEEK?
Vaughan concluded that c r o s s ~ gcould occw to such an extent in the amorphous
phase as to prevent melting of the remaining crystals.

Figure 1.2: Radiolysis mechanism of PEEK as proposed by Hegazy.

Sasuga and Hagiwara proposed a highex electron beam radiation resistance for
amorphous PEEK than semi-crystalline PEEK based on dynamic relaxation results. They
observed important variations on the material relaxation behaviour at the y-transition
(-1OO°C), p '-transition (40°C), @-transition(140°C) and a-transition (180°C). They
suggested that the disintegration of tie molecules at the amorphous and crystahe
interface would explain the lesser radiation resistance of semi-crystalline PEEK(''.'~).

1.3 Scope of the Present Work

Past research work has conceneated moaly on the study of the effects of electron
beam and gamma radiation on PEEK in vacuo or in air. Previous work on e p o v and
polyimide composites clearly demonstrated that the mecùanical properties of the studied
materials were more sensitive to neutron radiation as opposed to an equivalent physical
dose of gamma radiation"? No research has been carried out on the effects of fast and
thermal neutrons on the mechanical properties of PEEK, especidy in a reactor pool
environment. Since PEEK is known to bz a radiation-resistant p o b e r , such information
may be valuable to assess PEEK's potential for employment in nuclear reactor or
applications involving neutron radiation.

The mechanical properties of poly-mers are dependent on their viscoelastic nature.

The viscoelastic behaviour of polyrners is best observed starting from their glass transition
region up to their melting point, where both the eflects of the elasticity provided by
molecular conforïnation changes and the Mçcous flow created by chah segment
movements, takes place. In order to ver@/ if viscoelastic changes could be observed
foUowing irradiation, a series of stress relaxation tests were performed on Vradiated
polycarbonate and polystyrene for doses between 10 kGy and 100 kGy at temperatures
going fiom (Tg-20°C)to (Tg+20°C). The best temperature region, providing for the
optimal observation of stress relaxation and the repeatability of the results with a minimum
-
standard deviation, was (Tg 10°C) to (Tg-3°C).

Consequently two grades of PEEK, 150P and 450P.were processed in the form of
tende test bars and irradiated with a combined neutron and gamma flux m the pool of the
SLOWPOKE-2 reactor facility at the Royal Military CoIlege of Canada (RiMc). ïhe
studied dose range was set between 10 kGy and 320 kGy in order to identify early
degradation in serni-crystalline P E E L and also to m e s s molecular damage before the
occurrence of significant morphological changes. The morphology was characterized
with density measurements, differential scanning calorimetry @SC) and X-ray
diffraction. Then, the viscoelastic behaviour of the h d i a t e d resins was investigated at
three thermodynamic states as demonstrated in Figure 1.3, with stress relaxation tests
conducted 140°C, standard tensile tests at 24OC , and melt viscosity measurements at
350°C.

24°C 140°C 350°C

Tende Stress Relaxation Melt Viscosity

Figure 1.3: Selected temperatures for the three viscoelastic tests, and, the impact of the
phase transitions of PEEK on its Young's modulus and thermal properties.
 THEORY

2.1 Radiation

Radiation is radiant energy emitted by a radioactive substance. Radiation can be

divided in two groups. Fust, particdate radiation which consists in moving matter
particles, me electron(e3, beta(P), neutron(n), proton@) and alpha(a) particles'*). Then,
electromagnetic radiation, which consists m energy photons, like gamrna and X-rays.
Radiation propagates through the environment surrounding the source and transfers 3 s
energy to the various media in its path, resdting either in absorption, dissipation as heat or
ionkation which is the formation of electncdy charged radicals. This research, is mostly
concemed with the most damaging process, being ionization. Ionization &tes physical
changes in the imdiated medium that may potentially affect its mechanical, electrical and
chemical propertiesB). In assessing the potential impact of ionization on the macroscopic
properties of the irradiated material several factors must be considered mcludmg the
composition and the density of the mednrm, the energy, flux among and other
characteristics of the ionizing radiation

2.1.1 Interaction with matter

Nuclear reactors produce a large variety of radiation. The nuclear reaction is

initiated h the core by the spontaneous fission of Uranium-235, and leads to a series of
interactions producing radiation in the fuel rods, surroundhg water and reactor vesse1
 materiah. While electron, bet a, alpha particles and heavy ions are attenuated rapidly,
neutron and gamma radiation can interact with mataials fàr fiom the reactor core. This
radiant behaviour is the r e d t of different interaction mechanhm between radiation and
matter''').

Photons ionize atoms through three miportant energy tramfer processesRO'. Fifst,
the photoelechic process,(21"4)where low energy photons (< 0.05 MeV) interact with a
tightly bound electron fiom an huer shell of the atom and causes one electron to be
ejected out. Second, the Compton scattering, m which a higher energy photon (0.1-10
MeV) interacts with several atoms, each time scatterhg off an orbital electroq which goes
on to ionize more atoms. Compton scattering ends when the photon has decreased its
energy low enough so that the photoelectric process may occur. The thUd process is the
pair-production, where a photon of energy 1.02-20 MeV, converts t s energy in the
presence of an atomic nucleus to a positron-electron pair, which proceeds to interact with
more atoms and rnoleculesN. In most cases, Compton scattering is the principal radiation
interaction mechanism for organic po lpers. Overall, photon energy deposition tends to
be fairly unZorn within the irradiated material.

Neutrons ionize atorns along their paths by ejecting electrons, but also collide with
and produce energetic recoil nucleus by hock-un process. In a polymer, fast neutrons (1-
2 MeV), mostly produce energetic protons fiom the collisions with hydrogen atoms
which, in turn, interact and cause far more damage than the incident neutronN. Neutrons
of thermal energy (0.0253 eV) cm a b be absorbed by other nuclei which then become
unstable radionuclides and undergo nuclear reaction or further radiation deca$lg). This
type of reaction cm induce radiation that can remairi active a long tmie d e r the irradiation
exposure is completed, especially m metals. Again, in most pure polymers, induced
radiation is short-lived, because the radionuclides generated by C,O and H possess short
y the emission of beta and gammii-rays. The end result of
half-lives, and decay ~ a l with
neutron radiation is the generation of recoa protons and heavier ions, producing a different
geometrical distri'bution of the energy deposition which tends to concentrate damages
along specific tracks and in more localized sitesN.

The level of damage caused by radiation depends not ody on the quantity of the
energy deposited in the material but also on how concentrated the energy deposition is,
resulting in fiee-radical pairs along the path taken by the energized particle or photon'2).
The destructive potential of radiation is assessed in temis of hear mergy tramfer (LET)
which is the amount of energy deposited per unit length of the ionization trackCJ"',
expressed as dE/dx For example, gamma-radiation has a low LET of 3.5 keV1p.m in
water, whereas energetic protons and alpha-radiation have a high LET of 53 keV/pm in
water. Since fast neutrons can generate several recoil protons along theù paths, they are
also attnbuted with a LET of 53 k e V / p m water. If the incident neutrons have energies
lower than 10 keV, they are attributed with a Iower LET of 7-10 k e V / p in water'lg).

Another important factor is the abildy of the radiation to penetrate the irradiated
material. Gamma and X-rays can deeply penetrate polymers, as opposed to electron
beams or alpha-radiation which are strongly attenuated, resulting only in suface effects(*).
Conversely, if an irradiated sample is too thin, less than 1 nim, the radiation goes through
the material, for the most part, with little interaction with the medium. Therefore, the
dosage for thin specimens c m be large@ overestimated? In nuclear reactors, protons or
alpha-particles generated in the fûel do not &se beyond the reactor core. However, fast
neutrons can travel much fûrther and induce high LET recoil protons when they collide
with hydrogen atoms in the materials surrounding the core. Therefore, nuclear reactor
radiation can combine both the effects of high LET and deep penetration, and produce a
hi& concentration of radicals throughout the irradiated material.
2.13 Doses and units

When measuring the effect of radiation on a specific materiai, not ody the energy
and the type of radiation rmist be considered but also the intensity of the emitting source.
Those parameters dong with the distance fkom the source and the radiation absorbing
properties of the imadiated medium are used to measure the absorbed dose. The SI unit
for radiation intensity is the Becquerel (Bq) or 1 disintegratiods, the energy of the
particles is expressed in eV or MeV and the absorption properties are r;hacterized with
absorption coefficients. The Gray (Gy) and the rad are the most commonly used dose
units to measure the amount of energy transferred by radiation to an irradiatecl material(".

1 Gray(Gy) = 100 rad = 1 Jkg = 6.24 x 10" eV/g

Other units are also used to express the chernical yields of certain radicals or
chernical species, produced by radiation. Among those, is the G value, as defincd
bel~w('.~*).

G = number of changes1100 eV of irradiation

G(S) = number of scission!100 eV of irradiation
G ( X ) = number of cross-linksll00 eV of irradiation

The dose descnied above is the physical dose which is based only on the energy
absorption by the mediun In biology, ceU fiinctions are encoded in their molecular
arrangement. I f the radiation energy deposition is diffised. most of the damage caused is
of a rninor nature and self repair mechanisms can re-establish the original molecular code
of the celL Concentrated energy deposition creates major darnagesc5) beyond the repair
capability of the ceil, therefore killing it. For that reason, in biology, the physical dose is
multiplied by a quality factor to give a biological dose. That quaiity factor is based on the
energy transfer capab* of a type of radiation which provides an indication of the cell
 killBig potential based on the type of radiation. The resultmg dose units are the Sievert
and the rem

Like living ce&, poiymers are also structureci with long mlecular chains.
Although there is a siniilarity between polymers and biological systems with their
sensitivity to the nature of the incident radiation(2),a quality factor system is harder to
define for polyrners because they react differently to irradiation due to their wide variety of
molecular structure.

2.2 Effects of Radiation on Polymers

Polymers are being used increasingly in environments exposed to hi& levels of

ionking radiation. The properties of polyrners, such as elasticity, light weight, strength,
resistance to corrosion and formability are ideal for use m the space, communication,
electical and nuclear industries. Ionizing radiation cm alter the molecuiar structure and
macroscopic properties of the polymer~(202? It is mostly the levels of chah scission
which is the breaking of macromolecules, and crosslinking, that determine the extent of
the degradation of those properties. Several factors play a key role in predictïng and
understandmg the behaviour of polymeric materiais under radiation. Some of them are the
molecular structure, the crystallinay, the density of the polymer, also the type, energy and
flux of the incident radiation, just to name a few.

The ionization of molecules is the initial step in a series of events that lead to
chemical changes in the material. Irrelevant of which pdcular radiation interaction
mechanism applies, as described m section 2.1, the prirnary consequence is the ionization
of the affected molecules. The absorption of energy causing ionization depends sinq>ly on
the electron density of the material. Therefore, the Hiitial radical species induced by
radiation are distributecl m a fâirly random m e r m the material, concentrated along the
tracks of the penetrating radiation. Ionization occlils by the ejection of a high energy
electron Eom the molecule (Equation 2.1) in the fkst 10'~ s following the interaction with
radiation. The ejected electrons ionize M e r molecules (Equation 2.2) until they
become thermalized about 10"' s Iater. Then, mtra atoniic spur activity goes on for
another 10"s and radiolysis species start to migrate and mteract with other moleculesW.
Excited molecules (Equatiom 2.3 and 2.4) may undago radiationless decay to the ground
state, with heat being released, which may result in a siificant rise in the temperature of
the materials mder high radiation flux'20'.

radiation
R+ (ion) + e

- radiation
R+ + e R* (excited state)

radiation
R R*

Although excited molecules can trigger dnectly chemical reactions, the

predominant mtennediate outcome is the creation of fiee radicals (Equation 2.5). The
foxmation of radiation-induced fkee radicals is similar to chemical radicals except that
radiation-induced radicals are not f o d in randomly dispersed pairs but rather in
localized sets of pairs within tracks. The localised radical concentration can fàvour
recombination reactions as opposed to reactions wah other chemical species present in the
polymer. The radical chemical species resulting f?om ionization remain active for some
iime and induce various chemical transformations in the polymer at rates dependent on
both tïme (dictated by the rate of diffusion and the concentration of the reactive species)
and temperature (dictated by kinetics based on the Amhenius equatim)(2).

Radicals in a solid polymeric material are not as free to migrate as a liquid or gas(l':
the radical lifetime can thus be very long, up to weeks and months. The amorphous phase
of a polymer is more Milnerable to degradation than the crystalline phase, because in the
amorphous phase, the fiee vohme is larger, therefore hcreasmg the mobility of fiee
radicals which c m propagate degradation more easily. The crystalline phase is tighter.
Radicals caught in the crystahe phase either stay m place with a long half-Me, or
transfer their reactive sites, molecules to molecules, untLL they migrate out of the crystal
into the amorphous phase(". For t h reason, semi-aystalline polymers generaliy have a
better resistance to radiation than glassy and amorphous polymers.

2.2.2 Induced chemical changes

The initial ionization, resulting from the radiation interaction with matter, is
distniuted randomly. The radiochemistry that follows is more selective. It is caused by
the transfer of energy of f?ee radicals, ions and excited molecules and attacks relatively
weak chemical bonds. However, m radiation-resistant polymers, that energy is entrapped
in radiation-resistant functional groups that undergo efficient non-reactive decay to the
ground state'? The extent to which a polymer is chemicalîy transformed by radiation is
Weak and strong chemical bonds along
primarily determined by its molecular stru~ture~?
the chab detamine the nature of the reactive species induced by radiation and also the
most likely chemical reactions to fo~ow(~).

Durmg that stage, some bonds and chemical groups are more vulnerable than
others and easily react with the radiation-induced species. The most radiation sensitive
chemical groups kiclude COOH, C-Halogen, -S02-, -NI& and C=C. On the other hand,
among the most resistant groups are the methyl group and, m particuIar, the arornatic ring
which can trap the energy transferred and undergo efficient non-reactive decay to the
 ground state'? For that reason, aromatic polymers cm be 10 to 100 t h e s more resistant
to radiation than aliphatic polymer~(2).Although the C-Hbond is not a radiation sensitive
group, it plays an important role due to the fhct that it is a major component of most
polymers. The bond scission liberates a hydrogen atom which is then transferred arnong
several molecules creating radicals of lower energy that d l play an important role in
cross~g.

Radiation-induced chemically active species can react in Mirent ways resulting m

structural and rnolecular weight changes. The first change is crosslinking which increases
the rnolecular weight. Crosslinking can eventually result m the formation of an important
network which radically alters transition temperatures, chemical and mechanical properties
of the rnaterial? The second change is cham scission, where fkee radicals recombine with
low molecular weight molecules, which causes a decrease in the overall molecular weight
of the material. The general properties of the rnaterial are altered but in a more gradua1
way than crosslinkmg. The third change is the creation of low molecular weight products
resulting fiom ch& scission and followed by the abstraction and recombination reactions.
These products are usuafIy gases such as Hz. C02,CO and C&, but some polymers with
particular structural groups may release other gases like SOz, in the case of polysulfones,
or HCl for polyvinylchlonde(? Such small products impact on the overaU molecular
weight and create smaU gas pockets in the polymer which reduce mechanical properties.
The last induced chemical change is the actual structural modifications by the appearance
of new bonds creating new chemical groups or the elimiriation of existing ones. Such
changes can alter both chemical and physical properties of the polymer.

An appropriate mixture of both sensitive and resistant groups in the structure of a

polymer can lead to interestmg behaviour under radiation, like the promotion of
mt ermolecular crosslinking? For example, irradiated polystyrene f o m primarily two
types of radicalsBq: the first radical is produced on the backbone temmy carbon; the
second type has a radical site on the ring at either the ortho, metha or para position. The
first type can interact with a simüar radical to f o m a crosslink or it can react with the
second type to also fom a crosslink. For aliphatic and aromatic polyesters, the t e e
radicals occur primarîly on the carbon or ring adjacent on either side of the esta group. A
similar behaviour is expected with carbonate groups although they are more sensitive to
radiation than esters. For aliphatic chahs, fkee radicals occur primarily on ternary and
secondary carbon because methyl groups are more difKcult to ionize. The ionization of
polyethers occurs prmiarily on the carbon on either side of the ovgen, however breakage
of the ether bond c m also happen.

2.2.3 Impact on macroscopic properties

Chernical changes induced by radiation also impact on the macroscopic properties

of the material. Those changes are due primarily to the simultaneous occurrence of
crosslinking and chah scission. The yieield of a polymer for each reaction is dictated by its
structure and morphology therefore respective G(X) and G(S) are difFerent for each
material. For crosslinking to predominate, yields require to be 4G(X) > G(s)'~'. Since
both chin lengthening and shortenhg occur, the molecular weight h t i u t i o n (MWD)
wiU tend to widen. However, if scission is largely predorninant, the MWD narmws dovm
to a decreasing MW average. Crosslinking can also alter thermal properties by increasing
the glas transition, crystallization and melting ternperatures, while chah scission achieves
the opposite.

A polymer with a tendency to crossluik increases its viscosity until the formation of
a gel and then adopts the properties of a themosetn' which is a polymer that does not
possess a rnelting point like a thennoplastic. The polper becomes harder, stronger and
more brittle. Polymers prone to scission decrease their viscosity, become sofier and
weaker.
 Structural changes c m also decrease the crystallindy of a polymer, thereby
reducing its strengd2? Crystal defects caused by radiation can migrate to the surfàce of
the m a l resulting in a reduction m the crystal thickness or size. Crystal defects also
create coloured centers by the entrapment of electrons and ions m the crystal lattice,
inducing a colour change m the overall appearance of the polymero. For example, this
radiation-induced coloration is responsible for poly (methyl methacryhte) (PMMA)
tuming to yellow and polycarbonate tirming to green after irradiationRo!

2.2.4 Degradation

Environmental conditions are very important when considering the darnage

perfiormed by radiation on materials. When contaminants infiltrate or corne in contact
with the polymer during irradiation, the impurïties scavenge polymer radicals and the
crosslinking process can be severely disadvantaged as opposed to chah scission and the
polymer degrades very rapidly. Chain scission is the primary cause of degradation in a
polymer, however degradation c m be accelerated if inrpurities, like water, are present in
the polymer by terminahg reactive sites by combination with polyrneric fkee radicalsB8).
Also, extensive oxihtion c m occur deep in the materia1 if it is eqosed to air during or
after radiation.

Oxygen reacts with fkee-radicals generated by radiation m the miuiner shown

below (Equatiom 2.6-2.11). Due to its high diffusivity and its reactivity with radicals,
oxygen greatly impedes the radical reaction pathways that occur in the absence of oxygen
and force reactions into primarily oxibtive routes. Peroxides can also be formed and c m
breakdown into more radicals and propagate the creation of other peroxides. This cycle
can increase considerably the amount of free-radicals in the polymer structure and greatly
accelerate the degradation process40:
 The presence of water during or after irradiation can also accelerate the
deterioration of polymers m the same maaner as oxygen does, by propagatmg Eee-
radicals, as shown belowR8'(Equations 2.12-2.16):

+ H20 -
- OH* + H,O+ (2.13)

-
radiation
H20 H202, HZ,Ha , eq (2.14)
radiation
H20 H20*

2.3 Poly Ether Ether Ketone (PEEIC)

PEEK is a general-purpose mouiding and extrusion materiai, in both natural and

reinforced forms, with many high perfiormance applications"? Typical industries using
the polymer are the electrical (cable, insulation), automotive (bearings, piston skirts),
aerospace (structural components), oil (valves, pumps, seals), and chernical industries
(compressor valve plates, pH meters). PEEK and new polyaromatics korn the same
f d y , are engineering plastics increasingly in demand by high performance polymer users.

2.3.1 ChemicaI characteristics

High molecular weight PEEK can be produced by poIycondensation under

anhydrous conditons of a biphenyl and an aromatic dihaiide, in which the halogen atoms
can be activated by carbonyl groups. The polymerization is accomplished at a temperature
range of 200400°C with an aromatic sulfone solvent'? For example, 0.1M 4,4-
difluorobenzophenone and hydroquinone are condenseci up to 320°C in the presence of
diphenyl sulfone and anhydrous potassium carbonate. Of course, mare complex fonns of
PEEK can be produced by c h g i n g the reactantsB9'. One major constramt of PEEK is
that it is a dficult resin to manufacture because of bat ch procedures and insolubility.

PEEK has no c o m o n room-temperature solventS. Concentrated sulfiuic acid

dissolves and sulfonates the polper. At higher temperatures, dilute solutions of PEEK
can be fabricated Li hydrofluoric acid, trifiuoromethane~onic acid, or
dichlorotetrduoroacetone monohydrate, and in phenol 1,2,4-trichlorobenzene, a solvent
system used for gel permeation chromatography of PEEK. Benzophenone also produces
a stable hi& temperature solution. PEEK is highly resistant to attack by hot water and
stem up to 280°C at 18 MPa and for 3000 hours with almost no sigdcant
degradation"? PEEK has a high oxidative and thermal stabiiity. This T a consequence
of the molecular structure, which allows substantial electron overlap dong the molecuie.
Its melt is thermaliy stable at 400°C for over one hour m the absence of air. PEEK is also
non-flammable without retardant(3o!

2.3.2 Physical properties

PEEK resin is gray and normaIly crystahe at 35% with a density of 1.32 g/m3.
PEEK's mechanical and chernical resistance properties(l" are highly dependent on its
 crystallinity, which can be reduced or degraded by the presence of ortho and meta links in
the polymer cham versus para W. PEEK has a glas transition temperature (Ta of
143°C and a melting temperature (Tm)of 334°C. PEEK's other transition temperatures
include the y-transition (not to be confused wah the gamma-radiation) starting about
-lOO°C, which is characterfred by molecular movement near chabends and molecular
segments at the crystalline/amorphous phase mterface; the p-transition startmg about
140°C, corresponding to the glas transition, which is characterized by the movement of
chah segments of 10 to 50 backbone atoms m the amorphous phase; and, the a-transition
starting around 180°C,corresponding to the beginnmg of the qstallization temperature
r e g i ~ n ' ' ~ ~Although
~~'. the cxystallization temperature (Tc)of PEEK is defined as being
200°C, PEEK c m crystallize to various degree, at any temperature between 180°C and
320°C because of a cornplex multiple crystallization behaviour which is explained further
in this chapter. Ln the amorphous resin, Tc d e s t s itself as an exotherm during
dserential scanning calorimetry @SC), while in the semi-crystalline resin, Tcappears in
the f o m of a second melting point displayed by a smali endothemi, attesting of
transformations occurkg in the crystal population0 3 - 3 3 .

PEEK is suitable for ali electicai applications with the possible exception of those
where electrical loss is a highly critical consideration. PEEK maintains its electrical
properties up to 200 C(16) .
0

One of PEEK's major attniutes is its mechanical pdomuuice. Howeva an Uiitial

drop by a factor of about 3 in the strength occurs between TA143"C) and 175"C, then it
remains fairly stable up to 315"C, where the m e h g process begins, accompanied by a
rapid loss in mechanical properties. Reinforcement of PEEK with &es or particulate
filks improves si@cantly m a properties. The morphology of PEEK is indeed really
particular and plays an important role in the propaties of this polymer. The morphology
is highly dependent on fabrication, processing, molecular weight and most of dl,
temperature history of the structure. Three parameters corne into play when setting the
crystallinity of a semi-crystahe polymer like PEEK: Tg,Tm and T
!
'
?
 Normal processmg leads to a crystallinity of 20-35%. This cm be changed by
quenching to obtain as little as 5% crystallinity or as much as 50% crystallinay by
anneabg. Amorphous PEEK has a brown glassy appearance and a density of 1.263 g/cm3
and cryçtalline PEEK has a density of 1AOO g/cm3O').

PEEK has an orthorhombic crystal cell structure with its strongest reflection
intensities at (1 10),(111),(200) and ceil dimension of a = 7.55-7.88 A , b = 5.86-5.94 A ,
c = 9.88-10.07 A containing two repeat units per cell. The PEEK crystallization follows a
complex process in which crystal fibrils are sepanited by broad amorphous pockets.
Those fibrils are composed of stacks of lameiiae separatecl by a t h layer of amorphous
phase, as shown in Figure 2.1, where L (about 80 A) is de- as the long period of
thiclmess of both phases subdivided into a larger 11 (about 60 A) being the lamellar
thickness and h (20 A) being the interlamellar nonrrystalline thicknessP2). If the cell is
situated in relation to the hellar structure, the b axis represents the sphenilitc growth
direction, the c axis represents the lamellar tfückness direction, and a is orthogonal to both
b and c.

+
#.

Figure 2.1: PEEK crystal lameLiar structure dimensions and cell orientation
 Studies suggest a complex two-st age crystallization mechanism in which primary
crystallization would occur m pockets of stacks of lamellae separated by thin layns of
amorphous phase(33-35) . Those pockets would be spaced by pockets of amorphous phase
only. During slow cooling or annealing in the crystailization temperature region of
PEEK, the presence of crystals next to the amorphous phase pockets wodd produce a
negative pressure on the amorphous pockets causing them to crystaiiize in stacks of
larnellae, but of rmich mialler thickness. (Thismode1 is fairy new and is being challenged
by other rnodels). Structural changes under Tc are considered to be attributable to
expansion and are reversible. However, changes at Tc and above are considered
irreversible, because of the crystallization. It is unclear whether different brands of
crystals grow and melt simultaneously or in series. It is also suspected that PEEK may
have several T~'son.

Molecular weight also plays a key role on crystallinity. The higher the molecular
weight, the lower the maximum crystaüinity due to the higher level of chain entanglement.
However, the toughness of the material is improved due to chah length and because
cracks propagate better through crystallites(36-38) .

2.3.3 Processing

PEEK has a mould shrinkage of 1%. Processing of PEEK requires melt

PEEK shows little shear sensitivity and possesses a high
temperatures of 370-380"~"~).
thermal stability, however, in the presence of oxygen, the melted polyrner begins to
crosslink within 30 minutes at 400°C. Typical melt viscosity for natural PEEK at 380°C
and 1000 s" shear rate ranges fiom 100 to 300 Pa.s.

Processing of large extruded or moulded parts can be a problem due to high stress
build up during crystallization. The differential cooling of the rod core and surface can
lead to cracking, however, this can be resolved through high temperature annealing
(300 O C for several hours). Mould temperatures of 150-170°C are required to produce
fUy crystalline, stress-fiee parts. Heating during processing helps to control crystallinay.
In cornparison, heated mould on the other hand, produces opaque, light gray parts with a
consistent crystalIinity of 35%. Cold mould, results in translucent brown parts with 0-5%
crystallinity(30.39) .

The conventional processabiiity of PEEK has resulted in a wide range of modified

products or composite materials incorporathg fibrous mers such as glass or carbon, or
particulate miers, like d e d glass, micas etco0). Epoxy, cyanoacrylate and silicone
adhesives provide a high bond strength with PEEK; however, to optimize adhesion,
careful attention has to be paid to surFdcc preparation. Chromic acid etching, mechanical
abrasion, and plasma etching have al1 been proven successful methods(30.39).
 METHODOLOGY

Outline
The methodology used in this research work was broken down mto three steps.
The first step consisted of producing a quality test specimen that could display a constant
performance through a senes of tests. The second step consisted of irradiating the
material at various dosages within a standard set of environmental conditions. Lastly, the
airn of the third step was to characterize the material through t s macroscopic properties in
order tO identdjr rnorphological and radiation-mduced structural changes. Special
attention was paid to the viscoelastic behaviour of the irradiated material through different
thermodynamic states in an attempt to identify the nature of mlecular structure changes
of PEEK in relation with its morphology. The experimental techniques and procedures
described in this chapter are introduced in the sequence they were perfiormed during the
experimental manipulations.

3.1.1 Production of the test samples

The studied resins were industrial PEEK grades namely the 450P(MW 43,000
&noie, MWD 2.5) and 150P(MW 14,000 ghole, MWD 2.5). purchased fi-omVICTREX
USA. The resins were dned then processed on an ENGELS5 mjection modder to
produce tende test samples and then annealeci up to 200°C, for seven hours, to relieve
interna1 stress. Processing is further discussed in section 3.2 and the samples dimensions
are described in details m section 3.6. Some processed samples were selected and
 submitted to neutron activation analysis (NAA) to determine the irnpurity content and also
to establish the level of induced radiation m the kadiated material in order to set safety
handling procedures. Radiation safety is further addressed in section 3.3.

3.1.2 Experimental conditions

The tende test samples were then irradiated with a combmed gamma and neutron
flw positioned against the wall of the SLOWPOKE-2reactor vesse1 as descnied in
section 3.3, and submitted to doses ranghg fkom 104(10,000) Gy to 10" (320,000) Gy.
Specimm were irradiated in contact with the pool water which provided an oqgenated
environment more tàvourable to degradation. Samples were wrapped in polyethylene bags
to E t water circulation and provide standard flow conditions around the specimens. The
specimens were stored in a lead box for a cooling poa irradiation period of three weeks.
Characterization and testing of the irradiated samples by various technics followed, which
is the subject of sections 3.4 to 3.9.

Morphological and thermal properties were detennined by density measurements,

dinerential scanning calorimetry and X-ray difiaction which are discussed in sections 3.4,
3.7 and 3.8 respectively.

The molecular changes were studied through the analysis of the viscoelastic
behaviour of the polyrner at different themodynamic States. Three types of tests were
carried out on non-irradiated and irradiated PEEK as foiiows: stress relaxation at 140°C,
discussed in section 3.5; tensile strength at 24 O C , discussed in section 3.6; and melt
viscosity measurements at 350 OC which are discussed in section 3.9. The stress relaxation
test was performed on tensile test bars after applying a 4.8% strain.
 3.2 Injection Moulding
3.2.1 Description

Injection mouiding is an miportant processmg method for which the reciprocating-

screw mechanism is the most common. In a series of events, the screw f h t plasticizes
the solid polymer mto a metered volume of homogeneous melt. The screw then ceases to
rotate and moves forward to inject the melt, under high pressure, into the mould. M e r a
short c o o h g period, the solidifieci polyiner part is ejected out of the mould and the cycle
continues with new material entering fiom the feed'40). The mould temperature and the
cooling period in the mould affect h t l y on the crystallinity of the sample.

3.2.2 Apparatus

The ENGEL EC88XC90 injection moulding machine employs an hydraulic system

to provide the extremely high clamping pressures required in the production of modem
thennoplastic products. The machine consists of two different parts. One is a horizontal
clamping unit, and the other is an extruder with an injection cylinder behind it. The
hydraulic pump is dnven by an electric motor and the plastic mjection process is achieved
through a screw drive mechanisa The general purpose screw was used to mould al1
thennoplastic tende test bars.

3.2.3 Procedure

The PEEK 450P and 150P resins were predried in thin cans in an oven for four
hours at a temperature of 180°C. Resin powders were stirred every hour with a glass rod
to let trapped gases escape.

Polysulfone(PSU), with a Tgof 190°C,was used to purge the ENGEL 55 injection

moulder at a temperature of 350°C. Processing of the lower molecukr weight PEEK,
 150P, was perfonned first at 370°C. 800 gram of resin was used to purge the PSU and
adjust the injection settings, then 26 good samples were produced. The samples were
numbered as they were ejected fxom the mould. Processing of the m e r rnolecular
weight PEEK, 450P, followed. A 100 grams of resm was used to purge the injection
moulder then 4 1 good samples were produced ushg the same injection parameters. The
machine settings are enclosed in Appendix A.

Since the required mould temperature conditions of 150°C could not be met due to
machine limitations, annealing was performed on all samples to homogenize crystallinity
and relieve any bdt-up stress. VICTREX annealing recommendations were followed. Tg
being 143OC, the temperature profile was one hour at 150°C, halfan hour at 170°C, half
an hour at 180°C,three hours at 200°C, balfan h o u at 18U°C,half an h o u at 160°C, half
an hou- at 150°C, half an hour at 140°C and then slowly cooled to room temperature.
The surface appearance changed due to amealing. A transparent brown coating
representative of the amorphous phase, was showing on au samples after processing.
Following anneahg, the brown amurphous coating became of a d o m light gray colour,
representative of the appearance of crystals in the amorphous phase.

3.3 Neutron & Gamma Irradiation

3.3.1 Description

Nuclear reactors are used as sources of energy but also in research as intense
sources of neutron and gamma radiations. Although many types of radiation are present
in the reactor during operation, the most dominant ones are themial neutrons, fàst
neutrons and gamma photons of various energies. The fluxes and doses of those
radiations Vary with the reactor's geometry and its shield composition. Uranium-235 is a
fissile nuclide and c m undergo fission when it captures a neutron of any energy. The
resulting fission reaction produces on average 2.42 neutrons, 7.4 prompt gammas, 2.5
dehyed gammas, and numerous beta particles and neutrinos. The majority of the energy
of fission (198 MeV) appears as kinetic energy of the fission Eagments which is
transfomed immediately as heat(19! The rest of the fission energy is c h e d by the beta
particles, neutrinos, and gamma rays and is ultimately transfomed into heat as these
radiations intenct and are absorbed by matter.

3.3.2 Apparatus

The SLOWPOKE-2nuctear reactor, at the Royal MZt~ilitaryCollege of Canada

(RMC), provides an intense neutron source for teaching, trainmg and research. The
SLOWPOKE-2 nuclear reactor uses low e ~ c h e dUranium (LEU) fuel emiched at
19.89% with Uranium-235, and can produce neutron fluxes up to 1 x1012 ncm-2 .s-1 when
at fidl power, 20 kW. The reactor is characterized by mherent safety permitting
unattended operation in automatic mode, providmg excellent power repeatability and
stability with fluctuations of less than 0.5%.

3.3.3 Procedure

The irradiation of a material can induce radioactivity that c m last for a long time
after the exposwe. This phenornenon presents some handling s a f i problems, especially
with neutron-induced activation. However, for most polymers, the induced radioactivity
is minimal and short-lived. Polymers are mostly composed of C, O and H and can
sometimes contain S and CL The induced radiation produced by those atoms c m be
reduced well below safe public dose limits by letting the radioactivity decay for a few days
foilowing the end of irradiation.

For safety reasons, NAA was performed on some polymer samples to detect and
assess impurities left fiom the nihication process or metal traces resulting fiom the
erosion of the processmg equipment. SmaU mounts of Sodium (Na) were found as weii
as traces of Copper (Cu),Aluminum (Al), Bromine (Br), Manganese (Mn), Tin (Sn) and
Antimony (Sb). The NAA analysis mdicated at &st the requirement for longer cooluig
times, but also that the radiation hazards were much lower than those for wbich hot cell
would be needed.

Samples were put in doubled low density polyethylene W P E ) plastic bags and
installed on a PMMA rack at an equai distance fiom the reactor vessel wall as illustrated in
Figure 3.1. AU samples were irradiated in the pool, aga& the reactor vessel at mid-
reactor plane, 3 1.4 cm fkom the center core. The neutron fluxes at that irradiation site
were previously detemimed by experiments experimentally m 1989 by W.S.Andrews to
be 5.1 xl0" n.cm-2.S1 for thermal neutrons and 0.18 xloL0n.~rn-~.s-'
for fast neutrons with
an uncertainty of s%(~? In 1996, Dr. H.W. Bonin calculated physical dose rates, at the
3 1.4 cm radius pool site, fkom the experimental mixes for neutrons and also mcluded
theoretical calculations for gamma dose rates'42!

The specimens irradiation pronle, described m Appendk B, consisted in several

exposures between 4 and 16 hours a day on weekdays untii the targeted cmulated dose
was achieved. Those doses were O Gy, 104 Gy (10 kGy), 104.5 Gy (32 kGy), 10' Gy
(100 kGy) and 10S-SGy (320 kGy). A standard cooling period of three weeks was
aLlowed for each series of samples to k g the mduced radiation to a . acceptable level.
The induced radioactivity of the hottest samples was 20 mre.dh on contact, irmnediately
afier irradiation, after 132 hours of exposure in the reactor pool. Three weeks were
required to bring the induced gamma and beta dose exposure below the operational limit
of 2.4 m r d . A lead container was used to store the irradiated samples. Lead lined
protective apron and gloves were used when extended or repetitive handling was required.
 PMMA
Front view

PEEK Samp

Figure 3.1 : SLOWPOKE-2 irradiation set-up.

Dose rate (thermal neutrons) 607Gyh k 8.6 %

Dose rate (fast neutrons) 11.6Gyh + 37 %
Dose rate (prompt and delayed y) 1786 G y h + 35 %
Dose rate (backscattered y) 72.2Gylh t 35 %
Total physical dose rate 2477 G y h f 28 %

3.4 Density by Water Displacement

3.4.1 Description

The specific gravity (Sp gr) or density (p) of a solid is a property that cm be
determined as indicative of physical changes in a sarnple. Changes in density of the
material may be due to changes in crystalhity, molecular weight, crosslinking, molecular
configuration, loss of plasticizer, absorption of solvent, etc. The determination of the
d e n s e by water displacement consists in evaluatmg the specific gravity of a specimen by
fist measuring its mass m air and then measuring its apparent mass in water by a cornpiete
immersion of the specirnen in the liquid.

From the density of the material, its cryst;iuinity c m be determined with Equation
3.1 if the respective densities of both phases are known and given that the other causes of
change of densities, like voids and impurities, are negligible.

3.4.2 Apparatus

The specific gravity(31)of all the polymer samples, non-irradiated and irradiated,
was detemMed following the procedures laid out in ASTM D792-A, with a Mettler
H35AR scale, a container filled with distilled water and a corrosion-resistant wke for
suspending the specimens.

3.4.3 Procedure

First, the samples were conditioned for 48 hrs at 23°C *l°C and 50% &2%relative
humidity. Specimens of about 10 gram were used for the measwements. They were
weighted first in air, then in water. The Equation 3.2 was used to determine the specifïc
gravity of the samples.
 where:
a is the apparent mass of specimen, without wire in air,
b is the apparent mass of specimen completely immersed and of the wire
partially inimersed in liqyid, and
w is the apparent mass of the partially immersed wire.

The results for the density of the polymer are presented in Appendix C and were
obtained using the Equation 3.3:

3.5 Stress Relaxation at Constant Strain

3.5.1 Description

The viscoelastic properties of a material c m be stuclied by observing the r e h t i o n

of t s strength over tirne. At a given temperature, a stress a. is rapidly applied on a tensile
bar up to a strain E, and then this strai. is kept constant wMe the stress relaxation is
monitored as a f'unction of time t and its relaxation time s in accordance with the general
Equation 3.4:
 However polymer relaxation can be fàr more complex than the above general
description, which is not part of this research. The stress relaxation is smiply monitored
with time to calculate the relaxed Young's Modulus to compare the viscoekstic properties
of the materials irradiated at different dose. The relaxed Young's modulus is then
obtained following Equation 3.5:

3.5.2 Apparatus

An INSTRON Mode1 4206 tende test machine equipped with the Senes IX
Automated Materials Testing System V.5.25 was used with an INSTRON Environmental
Chamber System Mode13 11 1 to control environmental conditions. A thermorneter was
h d e d inside the chamber to closely monitor the temperature vanations during the
experiment and a Precision ScientSc Co. tinier model wah a precision of 0.1 second,
was used to record the INSTRON load applied as a function of tirne.

3.53 Procedure

The dimensions of the specimem cross section were measured with a micrometer.
The ASTM Do38 standard for tende test was followed. The stress relaxation test was
performed at 140°C. The samples were allowed to warm up for 30 mhs (15 mins being
the minimum recommended time in the maflual Histnictions) iu the chamber before any
stress was applied in order for the material to be d o r m l y conditioned at the test
temperature. A 1.2 mm elongation was rapidly applied at a strain rate 60 d m i n on
each specimen which corresponded to a 4.8 % strain, just below the maximum strain of
4.9 % allowable for PEEK, without signincant deformation to o~cut'5~).Then the load
was monitored for 2,000 seconds, and up to 60 % decrease in the applied stress was
obsenred. A sample relaxation curve is provided m Appendix D. The specimens were then
rernoved frorn the chamber and allowed to cool d o m to room temperature.

3.6 Standard Test for Tende Properties

3.6.1 Description

This test rnethod is used for the determination of tende properties of polymers in
the f o m of standard dumbbeii-shaped test specimens. The sample is stretched at a
constant strain rate until a break occurs in the materid (Figure 3.2). During the tes, the
load (F) applied by the instrument on the specimen is recorded agalist the elongation (L-
L ) . From these data and the sample cross section (A), several properties (Equations 3.6-
3.9) can be calculated such as the stress at yield(oY),the stress at break(ob), the strain at
yield($), the strain at break(~b),the energy at b r e a k 0 and the Young's Modulus(E).

a = FIA

E = (L-L.)/ Lo

E = Cr/€

W = AJadL
 K Break

Figure 3.2: Typical temile test stress-strain curve for a polymer.

3.6.2 Apparatus

The INSTRON Mode1 4206 tende test apparatus equipped with the Series IX
Automated Materials Testing System V.5.25 was used. The Plastics Tende Test Program
(PLA) Software was operated on a IBM compatible cornputer system. The cahiation
program (CAL) was also used to detemine the physical units selected in the test
instrument. Dimensions of the tensile test bars are shown in Figure 3.3.
 Figure 3.3: Tende test bars dimensions (mm). A 4 0 ; B4.4;C=19; D=70;
E=2.5; F=137 and L=25.

3.6.3 Procedure

The cross section of the gauge was measwed for each specimen with a micrometer
and values were averaged with a standard deviation mor. The mstniment was cahibrated
and the test parameters were entered in the PLA software. The ASTM D638 standard
test for tensile properties at room temperature was fouowed. A mainrate of 5 d m i n
was appiied until breakage of the specimen. A slow strain rate was chosen to observe the
changes in the maxirnum strain. Raw experimental data and a sample report are provided
in Appendix E.

3.7 DWerential Scanning Calorimetry

3.7.1 Description

DifTerential scanning calorimetry @SC) is a doniiriant technique for the t h e m l

analysis of polymer properties. This method measures the difference in energy inputs into
a material and a standard reference sample during a controlied temperature program.
Most physical and chemical processes mvolve changes m enthalpy or specitic heat. DSC
can record those changes for almost any liquid or solid materials withm equipment
temperature Mations. The measured changes in enthalpy are quantitative and are
usually in direct relation with the transitions or reactions occurring m the analyzed

DSC c m be used to determine heat capacity, heats of transition, heats of reaction

and temperatures of transitions of polymeric materials. The biggest advantages of
applying DSC are t s short time required to run an expriment and its accuracy with a
relatively m l 1 quantity of materid.

3.7.2 Apparatus

A DBerential Scanning Calorheter @SC) 20 10 was used with a TA Instnunent

9900 computer/themal analyzer and the Data Analysis Program version 6.1. Alufniflum
pans were used for sample preparation.

3.7.3 Procedure

Before any test was performed, the instrument was caliiated. Baseline caliiration
and cell constant cah'bration were performed witb empty aliiminum pans at a heating rate
of 10°C/mb fiom room temperature to 420°C. Temperature calibration was also
performed, based on the fusion temperature of Indium. 10 to 20 mr?ligrams samples were
cut fiom the gr@ part of the tensile bars, gauge samples were avoided because of the
possibility of stress-induced crystals which couid render the detection of crystallinity
changes caused by irradiation undetectable. AU the samples were submitted to two
consecutive heatmg nuis up to 420°C, then aiiowed to cool at 2O0CImin, Tg,Tc and Tm
were recorded on every run,some raw data and a sample report are provided in Appendix
F . Two consecutives DSC runs were performed on each sample, to compare changes in
the thermal behaviour of the material due to diffèrent cooling conditions occinring during
processing and the first DSC run.

3.8 X-ray Diffraction

3.8.1 Description

X-ray difiaction is a popular method for the study of the structure and
morphology of polymers. The technique consists m exposmg a target material to an X-ray
beam wbich is difiracted as a result of the interaction with the electrons of the material.
The ditnaction wiU. occur at specific angles (8) with inteushies (I)
producing
, an
absorption spectnim with peaks of area (A). The crystallinay of the sample was
determined by iden-g the absorption peaks belonging to the crystahe phase with the
Bragg's la^'^^) (Equation 3.10):

where dm is the interplanar distance between crystallographic

planes for a set of Miller indexes, hkl,
A. is the wavelength of the incident beam, and
& is the scattering angle at which W c t i o n intensity
c m be observed.

and, the crystal cell dimensions, abc, and the crystalline structure, being
orthorhombic for PEEK:
 --
then, the q s t d h t y was determined by the ratio of the sum of the areas of the crystalline
peaks over the total area of the spectnim.

3.8.2 Apparatus

The WAXS Scmtag X-ray difnlictometer wïth a theta-theta geometry goniorneter

and the DMS 2000 DifEaction Management System software were used. The Scintag
equipment uses a sealed Cu anode X-ray tube.

3.8.3 Procedure

A piece measuring 13 x 25 x 2.5 mm was cut fiom the tende specimen and
scanned transversally to the axis of injection, at a rate of 2 "/minand kom 10" to 35".
Peaks were recorded for theta values of 18.7 O, 20.2 O, 20.6 O, 22.6 O, 28.7 and 32.7
O O.

Peaks integration was performed usbg the Wmen Averbach analysis on the DMS
software version 3.16 in order to calculate the percentage of crystalhty. A ssample report
is included in Appendix G.

3.9 Rheological Analysis

3.9.1 Description

The viscous behaviour of a material can be studied by forcing the melted substance
to flow through a capillary of known dimensions. For a Newtonian fluid, the viscosity
would be determined with Equatiotls 3.12-3.1 4, following '40943):
 with

and

where q is the viscosity, r is the shear stress, is the shear rate, F is the force applied on
the plunger area A, of diameter D , L, and Dc are the capîilary length and diameter
respectively, and Vxhis the crosshead speed.

However, polymer melts behave as non-Newtonian fluids. In order to find the true
viscosity of a polymer melt, the corrected shear rate icm u t be calculated ushg the
Rabinowitsch correction factor (3n + 1 )/4n with Equation 3.15 as follows:

1)/4n (3.15)

which is obtained fiom the power law (~stwald)'~~)

in Equation 3.1 6:

where n is the power law factor and k the fluid consistency.

3.9.2 Apparatus

The Capillary Rheometer Instron Model 3210 with a Barber Colman Model 560
series and the Rheology Test (REO) Software were used. The Tmgsten carbide capillary
# 351 was inserted in the rheorneter, wah the following dimensions : D, = 0.1527 r
0.0005 cm, and L, = 5.0940 I 0.0003 cm.

3.9.3 Procedure

Irradiated tende test samples were cut hto small pieces weighmg about 1 to 2
grams. The rheometer was cleaned with cotton swabs and by purging the barrel with
PSU. The caiibration of the rheometer was performed for the range O to 50 IcN at the
temperature of 350r 1OC. According to the instrument specifications, the material
inserted in the barrel usually takes three minutes to reach the set temperature. A pre-heat
period of at least 10 min was allowed for the melt to reach the set temperature and to
ensure full melting of the materiaL The sample pieces were placed inside the barrel of the
extrusion assernbly and forced down the capîüary. Measurements were carried out at
plunger speeds ranghg âom 1 m d m i n to 32 d m i n . The shear viscosities were
obtained at six different shear rates, fkom a smgle charge of rnateriaL The Bagley's
correction, for the pressure drop at the capillary entrance, was deemed negligiile because
of the high L/Dratio (> 10). A sample report is included in Appendix H.
 RESULTS and DISCUSSION

4.1 General

This chapter presents the results and a prelmunary discussion for each experimental
technique descnied in the previous chapter. The results related to the characterization of
the morphology of the irradiated PEEK samples are first presented in sections 4.2 to 4.4,
descnbing changes in the density, thermal properties and the crystallinity in function of the
dose reccived for both the 150P and 450P resins. Sections 4.5 to 4.7 present respectively
the elastic, viscoelastic and viscous behaviour of the studied PEEK grades for the sarne
dose range. The important discussion points are synthesized furher in Chapter 5 where a
cornparison is done between the results obtained with the dinerent experimental
techniques. Chapter 5 also analyzes the potential sources for experimental error.

Since most results are presented in terms of the radiation dose received by the
PEEK samples, it is important to define the uncertainty related with a given exposure.
The radiation dose at the irradiation site is large and is estimated at 40%. Tnis uncertainty
is largely due to the unknown wide range of possible energies of the gamma photons and
neutrons particles at the Uradiation site(4'). However, the SLOWPOKE-2 produces a
highly reliable, repetitive and stable power with fluctuations of less than 0.5%. By
establishing the uncertainty related to the time for each exposure to be at the most, a few
minutes, then the uncertainty on the comparative dose received by the samples between
exponires, is less than 1% percent (see Appendk B, a 2 minutes uncertainty for a 4 hours
 exposure yields a 0.8% error). Therefore, the uncertainty on the comparative dose is
minimal and too s m d to appear on any of the graphs of this chapter.

The error bars appearing on all the graphs are the result of the added combination
of first, the calculated error based on the uncertainty of every parameters involveci m the
respective mathematical manipulation, and, the value of one standard deviarion obtained
from the statistics of the measurements taken over a popdation between 4 and 5
specimens. Sample calculations are provided for each set of results in Appendices C to H.
The standard deviation used is defined below:

The curves appearing on most graphs have been established by the author to show
trends. The doted line represents the baseline, obtained with the results of the non-
irradiated samples.

4.2 Densiîy

The density of the teasile bars was measured by immersion at room temperature in
distilled water following the ASTM D792-A standard as describeci in section 3.4. Five
sarnples were measured three times each, which provided for a low standard deviation on
the mean value of the density at any given dose. The measured densities are presented in
Figures 4.1 and 4.2 as a function of the irradiation dose for PEEK grades 150P and 450P
respectively. As expected, the respective densities of the two non-hadiated PEEK grades
are different, a fact explained by the difference in the molecular weight, since the
processing conditions and the cooling rate conditions were the same. Higher molecular
weight PEEK takes a longer t h e to crystallize due to its higher degree of chain
. - 1' 104.0
1o'.~ 1o5-' 1os-ï
Dose (Gy)
Figure 4.1: Density as a function of dose for PEEK 150P

Dose (Gy)
Figure 4.2: Density as a function of dose for PEEK 450P
 entanglement"", consequently, the shorter chains of the lower molecular weight grade
(150P) permitted a slightly higher degree of crystallinity under the same cooling
conditions, thus yielding a higher density for the 150P. Both grades indicate a slight
increase in density with dosage. This density increase is indicative of morphological
changes in the polymer structure, consisting mostly in crosslinking occurring in the
amorphous phase. If damages occur in the crystals, defects migrate out of the crystal to
the interface in the amorphous phase. Therefore, it is mostly in the amorphous phase that
the density is affected by c r o s s ~ g .But, defects occurring in the crystalline phase, if
that happened, may also result in a reduction in the density. This confirms that
crosslinking in the amorphous phase is the predominant phenomenon under irradiation at
the studied dose range.

The change in density of irradiated PEEK samples with dosage was more
pronounced for the 1SOP than for the 450P: this can also be explained by the difference in
molecular weight. Since the average MW of the 150P (14,000 g/mole) is three time
smaller than the 450P (43,000 g/mole), the concentration of chain-ends in the amorphous
phase is about three times larger for the 150P. Crosslinking depends on the movement of
fkee radicals within the sample and that movement, at room temperature, is limited mostly
to repeating units at the crystal interface and chain identified in Figure 4.3.
Since the ctystallinity is somewhat comparable for both grades, the crystalline interface
conditions are thus similar and the higher density change may be attributed to more
crosslinking occurring in the amorphous phase of the 150P grade as a result of a higher
concentration of chain ends.

The leveling-off of the density for both grades between 1 0 ' and
~ ~ 1 0 ' - ' ~ ~ could
,
indicate that chain scission effects on the density, eventuaIly counterbalances those of
crosslinking, and that a state of equ5'briuIll is reached in the material between the two
reactions.
 chain-ends

Qm
C
t phase interface

amorphous phase

crystalline phase

Figure 4.3: PEEK rnolecular arrangement in its semi-crystalline state.

4.3 Glass Transition, Crystallization and Melting

The glass transition (T$, melting (Tm)and crystakation (Tc)temperatures, for

both PEEK grades, remained unaffected by irradiation at the studied doses. This indicates
that structural changes induced by radiation were too small to produce detectable
variations in the macroscopic thermal properties of the irradiated material. The accuracy
on the values for measured Tg,Tcand Tm was relatively high with an uncertainty below 5
1 .O°C in mon readings, as shown in Figures 4.5, 4.6,4.7 and 4.8, respectively. For the
purpose of this work, Tc is dehed by the temperature of the endotherm produced by the
melting or reorganizing of the secondary crystals induced either by the annealing done at
200°C, or by the c o o h g on the DSC at 20°C/min. The endothenn at Tc is typical of the
secondary crystals of PEEK which are dflerent fiom the larger primary crystals which
melt at the melting temperature Tm03.WW
 Figure 4.4: Typical DSC heat flow diagram for the processed 150P and 45OP PEEK
grades. For numerical values, see Figures 4.5 to 4.8 and Appendix F.

Two differential scanning calorimetry (DSC) scans were performed on each of the
samples. However, different temperatures were obtained for Tgand Tcbetween the fint
and second runs, as demonstrated in Figure 4.4, while no variations were detected on Tm.
The recorded Tgvalues for both resins, 150P in Figure 4.5 (162 OC and 117 OC) and 450P
in Figure 4.6 (166 OC and 150 OC), were significantly higher than the known Tgvalue of
143°C for amorphous PEEK The higher values obtained for Tgare attnbuted to the semi-
crystalline state of the studied material. The amorphous phase is entrapped between
crystalline layers which restricts the molecular motion of the non-crystalline chains and
induce a shift of Tg to a higher temperature('? The higher Tg values of the 450P is
attri'buted to its higher average molecular weight, resulting in a higher level of
entanglement, than the 150P samples. The 15°C &op in Tgbetween the first and second
DSC runs is related to a different crystallinity and crystahe mangement in the sample as
a result of a different cooling rate o c c u ~ during
g processing and annealing as opposed
to the cooling rate fiom the DSC equipment which was 20°C/min. Some authors,
identified three types of crystaUinity and two types of amorphous f o m for PEEK"",
which can compose a complex morpholog. For this research, thinner secondary crystal
Iamellae, resulting fkom annealing, would appear to grow and insert themselves between
the existing primary crystal lamellae produced f?om the melt(3zt46'. This secondary
crystallization would further restrict movement in the glassy aate and would be
responsible for the higher initial Tgof the fist DSC scan. No annealing was performed
with the DSC, so the crystalline form produced ffom the glassy state was not present in
the second DSC scan which resulted in lower Tgvalues.

Recorded values for Tm were 345 O C for 150P and 341 O C for 450P. The slightly
higher values for 150P is related to its lower molecular weight. Since chahs are shorter
they can achieve crystallization faster under the sarne cooling conditionsng', resulting in a
higher crystallinity , possibly with thicker and better formed crystals. Primary crystals
forrning immediately after coohg t o m the melt, wilI melt at Tm on reheating. Secondary
crystallization can occur in PEEK immediately afier the primary crystaükation, therefore,
at a lower temperature given that the cooling rate is slow enough (c 20"~/1nin)(~'),
to
allow such crystallization to occ~r'~".This secondary crystakation occwed during the
cooling performed after the firm DSC nui and produced a srnail second endotherm during
the heating of the second DSC nin, at Tc (300°C for 15OP and 287°C for 450P), just
below the melting point, as secondary crystals were melting and changing state. The
temperatures of the secondary endotherm were dEerent for both resins, on the second
scan, and higher for 150P because of its lower molecular weight; therefore it was easier to
crystallize and crystallized at a higher temperature than the 450P did. However, that
second endotherm did not appear on the first DSC mn, because the cooling rate, upon
processing, was certainly much higher than 20°C/min.The first DSC scan also produced a
smaIl endotherm, but at a lower temperature of 22 1°C which resulted from crystallization
occuning during the annealuig. That secondary endothenn was at the same temperature
for both PEEK grades because of the same annealing temperature of 200 OC. The
endothenn is again the result of the melting or reorganization of those secondary crystals.
 - (2nddsc scan)
.--------a
Non-irradiated P P Q

Dose (Gy)
Figure 4.5: Glass transition temperahire as a funchon of dose for PEEK 150P

(1" dsc scan)

Dose (Gy)
Figure 4.6: Glass transition temperahire as a function of dose for PEEK 450P
 Dose (Gy)
Figure 4.7: Ciystallization and melting temperames as a function of dose for PEEK 150P

320

('& 2* dsc scan)

c
.C)
-------- O O O O

260

-& 240
8 Non-irradiated
--------O
(Tcla dsc SC@
3 220 O -0
I
U 20n
1 10
Dose (Gy)
Figure 4.8: Crystallization and melting temperatures as a function of dose for PEEK 450P
 The multiple-crystalline behaviour has been attnïuted by some authors to dEerent
population of mole~ules'~~'.
However, other researcher~(~~"~'
believed that the pressure
differential in the material caused by the dEerent coefficients of expansion for the
crystalline and amorphous phase could induce the secondary crystaliization process. As
temperature increases, density changes occw in the amorphous phase and the crystahe
phase. Pressure is added onto the amorphous phase or the thinner crystalline larnellae,
which also absorb heat and melt. A slow heating rate may allow those molecules to
recrystallize under thicker lamellae until the melting temperature is reached.

4.4 Crystallinity

The crystaUinity of the specimens was detennined in t b e e ways, as shown m

Figures 4.9 and 4.10. The first way was by density measurement using the respective
densitiesB1)of 1263 kg/m3 for the PEEK amorphous phase and 1400 kg/m3 for the
crystalline phase. This method was used only for non-irradiated samples since radiation
induces changes in the amorphous phase which affects the density. Secondly, the
crystalIinity was determined by DSC using the heat of fusion of crystalline PEEK, being
130 J / ~ @ ? No variations were detected outside of the experimental uncertainty. Lastly,
X-ray deaction was used. The areas given by the dsaction peaks of the (1 IO), (1 13)
and (200) reflectiom were divided by the total area of the spectnun Again, no variations
were detected outside of the experimental error. The stability of the crystallinity in both
figures confïms that the changes only occurred in the amorphous phase. Had substantial
they would have been evidenced by a reduction in the
damages occurred in the crystal~'~~~',
density, indeed, it did not occur as evidenced in Figures 4.1 and 4.2. The ASTM D792-A
standard method was used to detennine densities and yielded the most significant results.
Therefore, the crystallinity was established at 27% for the non-imadiated 150P £kom a
density of 1300 kg/& and at 24% for the non-irradiated 450P from a density of 1296 kg/
m3. Aiso, it is concluded fiom the above results that crystahity changes, if any, were not
detectable for the studied radiation dose range.
Figure 4.9: Crystallinity by DSC(O), X-ray diffraction(.) and density(V) as a function
of dose for PEEK 1SOP

2o 4 10
4a 10
4.5
I

Dose (Gy)
5.0
10
I
10
S.s
I

Figure 4.10: Crystallinity by DSC(O),X-ray diffachon(@) and density(V) as a f i c t i o n

of dose for PEEK 450P
 The DSC results presented a higher crystallinity with a larger uncertamty, because
of the existence of dEerent crystalline species and PEEK is suspected to have a high level
of ordering in its amorphous phase that also contributes sigolficantly to the heat of
transition associated with f ~ s i o r i ' ~ ~DSC
~ ~ ~yielded
). crystallinities of 34% and 30% for
150P and 450P, respectively. However, as explained above, DSC is not an accurate
method of assessing crystallinity for PEEK and values tend to be over-estimated.

Density rneasurements and DSC-determined heat of fusion yielded a higher

crystaUinity for the 150P, however X-ray dsaction produced opposite results with 24%
for 150P and 29% for 450P. This inconsistency is explained by the important
experimental error for the X-ray ditnaction method since the width at the base of each
diffraction peak was poorly consistent.

4.5 Tensile Properties

Standard Tensile Tests at a strain rate of 5 m ' m i n produced revealing stress-

strain diagrams. Figures 4.1 1 and 4.12 represent the main features resulthg fiom tensile
tests at room temperature for 1SOP and 450P samples, respectively. Both resins displayed
a yield point with similar characteristics except that the stress at yield was slightly higher
for the 150P as a result of its higher crystaUinity. In contrast, the break point was
significantly diffierent. The 450P showed the characteristics of a stronger and much
tougher material with a strain at break of 0.84 m m / m versus 0.38 mm/mrn for l5OP and
a stress at break of 97 MPa versus 84 MPa. The vertical portion on the diagram, below
the break point, is indicative of a tough break where tearing of the material occurs. The
stronger and tougher characteristics of 450P,displayed at the break point, are related to
its much higher average molecula.weight produced by longer chains.
 0.00 0.25 O .50
Strain ( d m m )
Figure 4.11: Typical stress-strain curves at 24°C and Smdmin strain rate for PEEK l5OP

E = 1200 MPa

0.00 0.50 1.O0

Sbain (&mm)
Figure 4.12: Typical stress-sûain curves at 24OC and Smmhnin strain rate for PEEK 450P
 An interesting feature appears on Figures 4.1 2 and E. 1 and may be interpreted as a
second yield point with a strain of 0.52 &mm and a stress of 98 MPa. The feature is
elongated and bas a jagged edge appearance. This second yield point is attnbuted to the
crystalline network created by molecules long enough to be part of several crystalline
layers. The jagged edge of that feature attests of crystal reorientation in the material as a
result of the achieved maximum strain allowable by the arnorphous segments which are
forcing the crystalline structure to react to the imposed stress. The break then occurs
when reonentation in the crystalline structure is no longer possible.

The higher stress at break (97 MPa) is representative of the stronger force required
to pull molecules out of their crystals and to break the tie molecules between those
crystals, as opposed to the lower force required to break entangled molecules (84 MPa) in
the amorphous phase. This theory is supported by the fact that the strain at break for the
150P is inferior to the strain of the second yield point found in 45OP. This second yield
point, attributable to the crystalline structure. also provides an indication of the extent of
chain folding achieved per rnolecule per crystalline lamella.

4.5.1 Young's modulus and tensile strength

Figures 4.13 and 4.14 represent the Young's modulus in relation with the dose for
the 150P and 450P grade sarnples, respectively. The obtained values of non-ùradiated
specimens (150P: 1193212 MPa, 450P: 1239515 MPa) are comparable with those
obtained by Sasuga et al.''L',between 1170-5 MPa and 1370-0 MPa. At e s t , both
grades of PEEK bave a comparable behaviour, a steady increase of the Young's modutus
up to a maximum reached at about 10'' Gy. Past this dose, the behaviour then becomes
quite dif5erent for both resins. For the 150P resin, a slight decrease of the Young's
modulus is followed by an increase up to Gy. The 450P decreases its modulus down
to los.' Gy. At room temperature, PEEK is above its y-transition (-40°C), where
molecular movement is lùliited to backbone rotations, chah-ends in the amorphous phase
and repeating units at the phase interface""."'.
 Dose (Gy)
Figure 4.13: Young's Modulus at 24°C as a function of dose for PEEK 150P

Dose (Gy)
Figure 4.14: Young's Modulus at 24°C as a h c t i o n of dose for PEEK 450P
 The aromatic rings and the ketone groups possess bonds with very high activation
energies and are the less probable sites for radiation damage. However, the two C-C and
four C-Obonds, in the backbone repeating unit, have the lowest activation energies behg
347 kl/mole and 360 kJ/mole, respectively, which makes them the most probable sites for
radicalization. Radicalkation Eom the backbone would create new active chain ends with
added mobility that would recombine, or, crosslink on adjacent molecules. The sensitivity
of PEEK, at those sites, is confrmied by the gas evolution work and the radiolyssis
mechanism proposed by ~ e ~ a z ~ ' ' ~The
. ' ~other
' . important radiolysis sites are the twelve
C-Hbonds, with an activation energy of 4 14 W/rnole, which play an important role due to
their number. These bonds are located at the 2,3,5 and 6 sites on the three aromatic rings.
From their respective positions, six of those bonds when irradiated favour the creation of
new intrarnolecular bonds in the backbone between aromatic rings, as descnied in Figure
4.15, and the other six when inadiated favour intemolecular crosslinkmg with adjacent
molecules.

Figure 4.15: Example of an intramolecular bond occumhg in the PEEK molecular

baci~bone'~~'.

In non-irradiated specimens, the supenor Young's modulus of the semi-crystalline

PEEK is attrîbuted to the presence of crystals in its morphology where tie molecules can
transfer stress fiom the tough morphous phase to the hard crystahe phase. In spite of
the slightly higher crystauinity content of the 150P, it displays a lower Young's modulus
due to a concentration of chain-ends about three times higher than the 450P due to its
much lower molecular weight average.

For the in-adiated specimens, the initial dose of 104 Gy induces an hcrease in
modulus for both resins. This initial increase is attnhted to the loss in backbone
flexiiility resulting from intramolecuiar bond formation combined with some crosslinking.
As the radiolysis continues, chah scission occurs and the appearance of additional chain-
ends reduces the Young's modulus. Tie molecules, which links the crystalline and
amorphous phase together, are more sensitive to radiation, especialiy at the phase
interface, due to the stress hduced by the different densities of the crystalline and
amorphous phases and to the fact that bonds under stress are more vulnerable to
radiolysis"'. The decrease in moduius is therefore attributed to the chah scissions of
those tie molecules. The différent behaviours of the modulus for the resins at doses of 10'
Gy and Gy are attributable to the concentration of chah-ends again. In 450P the
reduction in modulus is characterized predorninantly by chah scissions of the tie
molecules. However, the higher nwnber of chai.-ends in the 150P increases the
probability of crosslinking which is predominant and favours an increase in modulus for
the same doses.

Figures 4.16 and 4.17 represent the tensile strength as a function of dose for the
150P and 450P respectively. Tbe tende strength was determined fiom the ultimate stress
achieved by the material. The yield point yielded higher stress results than the break point
for both resins, therefore, the tensile strength values were provided by results fiom the
yield point. The measured tende strength values for non-irradiated specirnens were
108.00.5 MPa for the 150P and 103.020.5 MPa for the 450P which compares very well
with the reported values by ViCTREX being 100 MPa and 97 MPa re~pectively"~).
Those
Obtained values were indicative of a good specimen preparation with proper processing
and annealing conditions. The variations in strength as a result of irradiation, were
insignificant, being smaller than the experimental enor. These results are attnbuted to a
high radiation resistance of PEEK displayed by a constant crystaUinity, and an equili'brhn
in mechanical properties due to the occurrence of both crosslinking and chah scission in
the amorphous phase.
 Dose (Gy)
Figure 4.16: Tensile strength at 2CC as a k c t i o n of dose for PEEK l5OP

- .

Dose (Gy)
Figure 4.17: Tensile strength at 24OC as a function of dose for PEEK 450P
4.5.2 Maximum strain and energy at break

Figures 4.18 and 4.19 represent the maximum strain achieved at break. The
maximum strah of both resins is little affected up to 1 0 ~ ~Gy,
' then, it becomes quite
different for each grade for doses up to Gy. The 150P reaches a lower maximum
strain at 1 0 % ~ then goes back to îts onginal maximum main at ioSS Gy. The 450P
keeps a constant value for its m;ucimu strain up to a dose of 10' Gy, then drops
sigoificantly at 1oS5 Gy. This différence in behaviour may be explained by the effect of
crosslinkmg and chah scission on a linear polymer and a crosslinked polymer. The 150P
behaves like a linear polymr in the amorphous phase. The 450P behaves Like a d d l y
physicaliy crosslinked polymer due to its long chahs able to be part o f several crystalline
iamellae. The molecules are packed together in crystals inducing some type of crystalline
network. The drop in m;Uamum strain for l5OP between 1o ~Gy
. and
~ 1O' Gy indicates the
occurrence of crosslinking and the formation of a network which is supported by the
increase in modulus at the same dose range in Figure 4.13. The network is short lived due
to chin scission and the maximum suain goes back to its original value at Io5-' Gy. On
the other hand, the 450P maximum main is tiiUly stable up to 10' Gy but then
demonstrates a suddm drop due to crosslinking in the amorphous phase which is also
supported by the leveling off of the Young's modulus in Figure 4.14 behveen 105 Gy and
1 GY.

Figures 4.20 and 4.21 present the energy at break m relation with the dose. The
energy at break is usually calculated &om the area under the curve of a stress-strain
diagram Since the modulus and the stress at yield Vary little with radiation for both
resins, the curves display a smiiiar trend as for the maximum strain. Therefore the same
rationale, expressed for Figures 4.18 and 4.19, is applicable for the energy at break.
 Dose (Gy)
Figure 4.18: Maximum strain at 24OC as a function of dose for PEEK 150P

Dose (Gy)
Figure 4.19: Maximum strain at 24OC as a function of dose for PEEK 450P
 I I I
4iO 4.5 5.0 5.5
10 10 10 10
Dose (Gy)
Figure 420: Energy at break at 24°C as a function of dose for PEEK 150P

Dose (Gy)

Figure 4.21: Energy at break at 24°C as a function of dose for PEEK 450P
4.6 Stress Relaxation at Constant Strain

Figures 4.22 and 4.23 present the relaxed Young's moduhis versus the radiation
dose received for the 150P and 450P resins, respectively. The stress relaxation test was
perfomed at 140 OC on tende test bars after rapidly applying a 5% stram at a rate of 60
d m i n . Then the load, subsequently converted to the relaxed modulus, was monitored
through time for a period of 2000 seconds. At 140 O C , PEEK is in às P-transition, which
corresponds to the beginnlig of the glass transition(10.17) , where relaxation is dominated by
molecular movement in the arnorphous phase of 10 to 50 backbone atoms. The
relaxation, caused by the p-transition. is responsible for the much lower modulus values
(about 900 MPa) obtained for t = O second, as compared with the ones obtained at room
temperatures (about 1200 MPa). During the short time of 1 second, while stress was
applied, relaxation occurred on a mucb larger scale than what would have happened at
room temperature. For that reason the shapes of the relaxed Young's modulus (EXt))
curves at t = O second are different than the ones obtained on Figures 4.1 3 and 4.14.

The relaxed Young's modulus Edt) was rneasured against t h e with a statistical
deviation error of 10 to 30 MPa for each pomt based on the results obtained by five
samples. The results in Figures 4.22 and 4.23, representing the measured relaxed Young's
modulus, displayed a marked decrease for samples that received doses of 10' Gy or
more. The higher molecular weight grade (450P) dispkiyed a slîght m;uàmumin ut)after
receiving a dose between lo4 Gy to 104-' Gy. The smail variations in Mt), even at a
temperature of 140°C, confirmed the very high stability of the neutron irradiated semi-
crystalhe PEEK and are indicative of the cornpetition occurring within the amorphous
phase between the crosslinking and the chai. scission processes. Although ut)seemed at
first to be only slightly affected by radiation, a rnarked decreasing trend appeared for
doses between lo4.' Gy and 10' Gy. Therefore, at this dose range, the molecular
backbone movement in the amorphous phase seems to become more important, probably
caused by an increase in lower molecular weight products resulting h m chain scission.
 300 4 ,

10
I
4.0
I
104m5
I
IO 5.0
I
5.5

Dose (Gy)
Figure 4.22: Relaxed Young's rnodulus at 140°C as a funchon of dose for PEEK 150P

800
Non-irradiated
- - . - I - - L L

f g- s -
(t = Os)
5

Dose (Gy)
Figure 4.23: Relaxed Young's rnodulus at 140°C as a function of dose for PEEK 450P
 On a segmental scale of several repeating units, the effect of chain scission in the
amorphous phase becomes predominant over crossünkuig afier a dose of 104-' Gy and
contributes more eeectively to the reduction in the mechanical properties of irradiated
PEEK. It is apparent that the 150P resin degrades at a faster rate than the 450P resin.
This is due to the k t that the 450P, wbh its higher molecular weight, can rely on its
crystahe network structure, to resist the relaxation caused by the production of lower
molecular weight radiolysis products.

Melt Viscosity

Lastly, the results fiom the rhmlogy test performed at 350 C(44.45.51) are displayed
O

in Figures 4.24 and 4.25, and represent the melt-viscosity in relation with the dose for the
15OP and 450P resins, respectively. Table 4.1 presents the obtained values for the power
law factor (n) for both resins at the studied doses, which was used to determine the
corrected viscosity as demonstrated in section 3.9. At 350°C, PEEK crystals melt and the
semi-crystalline morphology of the polyrner becomes one homogeneous viscous phase.
The limited rnolecular rnovernent caused by the presence of crystals gives place to full
molecular motion m the melt above Tm. For t
h test, both resins yielded similar results. A
maximum viscosity was reached for a dose of 105 Gy foUowed by a decrease at higher
dose. This maximum, can be attributed to the presence of larger molecules resuiting korn
a limited degree of c r o s s ~ g .A fact that is ako supported by the power law factor
values which showed a decrease for both resb at a dose of 10' Gy pointed to an increase
in molecular weight. However, fiom those results, it is not clear if c r o s s ~ occurred
g
only in the amorphous phase or also in the crystalline phase.

The mcrease in viscosity at 10' Gy confimu an increase in molecular weight. The

large error bars associated with that point, for both resins, indicate a very unstable state,
possibly due to the appearance of some micro-gels. This is plausible smce a crosslinked
polymer should behave like gels rather than a melt. However, the effects on the viscosity,
caused by the existence of those micro-gels, disappear beyond a dose of 10' Gy as
demonstrated in Figures 4.24 and 4.25. The combmed results of the following observed
tacts: no clear increase in modulus, no increase in strength at room temperature, and the
increased relaxation near T, for a dose of los-' Gy, au point towards the predomimmce
effect of chah scission on the viscoelastic properties of both PEEK resins at the studied
reactor pool conditions and dose range.

Table 4.1: Power law factor (n)at studied doses with the calculated standard deviation
based on a population of five measurements.

GRADE DOSE (Gy) POWER LAW FACTOR STD DEVIATION

150P O 0.763 0.058
1O" 0.808 0.053
1o ~ - ~ 0.822 0.026
Figure 4.24: Melt viscosity at 350°C and 100 s-[ as a funchon of dose for PEEK 150P

Dose (Gy)

Figure 4.25: Melt viscosity at 350°C and 100 s-l as a function of dose for PEEK 450P
 Chapter 3

SYNTHESIS and ANALYSIS

5.1 Radiation Damage Assessment

Two industrial PEEK grades were purchased nom V I C W X USA. the

450P(MW 43k, MWD 2.5) and lSOP(MW 14k, MWD 2.5). The resins were processed
on an ENGELS5 injection moulder to produce tende test samples and then annealed up
to 200°C. for seven hours, to relieve intemal stress. The tende test samples were then
irradiated with a combined gamma and neuaon flux, agauist the reactor vesse1 wail of the
RMC SLOWPOKE-2facility, for exposures ranging f?om 10' Gy to 10"~Gy. The total
dose rate at the irradiation site in the SLOWPOKE-2reactor pool was detemiined to be
2400 kg60 ~ ~ / h ' ~ 'and
* ' ~consisting
', of a 75% gamma radiation contribution and 25% fast
and themial neutrons contribution. FoiIowing a cooling period of three weeks,
morphological and thermal properties were detemimed by differential sc&g
calorimetry, X-ray =action and density rneasurementso? Although the density
increased slightiy over the range of irradiation (450P: 1296 to 1299 kg/m3, 150P: 1300 to
1303 kdm3)which can be attnbuted to crossiinking, changes in themial properties Like the
glass transition and melting temperatures were not detected, as weii as no changes were
detected in the crystallinity (450P: 24 %, 150P: 27 %) within the precision of the
equipment. This leads to the belief that most of the changes occurred within the
amorphous phase which is consistent with previous work(s~6v1?
 Three types of viscoelastic tens were carried out on non-irradiated and irradiated
PEEK as follows: standard tensile tests at 24"C, stress relaxation tests at 140°C and melt-
viçcosity measurements at 350°C. The results of those three tests clearly demonstrated
polymer degradation dombted by chah scission, in both resins, after receiving a
combmed neutron and gamma dose of 1 0 ' ~Gy. Although the performance of PEEK is
still outstanding in sucb irradiation conditions in a reactor pool environment, the results in
cornparison with previous work, clearly showed a more significant degradation rnechanism
than for electron beam'") or gamma radiation done'?

5.2 Role Played by the Morphology

The role of the morphology on the radiolysis of the polymer was also considered.
Standard tensile tens were performed at 24°C. At that temperature, PEEK is above its y-
transition temperature where molecular movement is limited mostly to chah-ends in the
amorphous phase and repeat units at the phase interface'? Stress-strain diagrams
revealed that the higher molecular weight PEEK (450P)possessed a second yield point
which could be associated with the presence of a crystalline network where molecular
chains are long enough to be part of several crystalline iayers. The lower molecular
weight (1 5OP) did not display that kind of behaviour which would hdicate that its strain is
limited by the effect of chah entanglement in the amorphous phase. For that reason, the
Young's modulus of the two PEEK grades, reacted differently to radiation. The 450P
displayed a clear reduction in its Young's modulus der receiving 104-'Gy while the 150P
showed a slight increase with dosage. The reduction in modulus, was explained by the
higher radiation v~lnerabiIh$'~)of the tie molecules at the phase interface due to the
presence of some tension between two phases of different densities. The behaviour of the
l5OP was attriiuted to the higher concentration of chain-ends, about three t h e s that of
450P, which, by their mobility potentiaily favoured crosslinking.
 The stress relaxation test was performed at 140 O C on tende test bars after rapidly
applying a 4.8% stram. At that temperature, PEEK is in its P-transition where movement
is domhatecl by rnolecular backbone movement in the amorphous phase. The relaxed
Young's moduhis u t ) was measured agamst t h e with a statistical deviation error of 10
to 30 iMPa based on five samples for each point, as demonstrated in Appendix D. The
results Hi Figures 4.22 and 4.23, representing the measured reiaxed Young's modulus,
displayed a marked decrease for samples that received doses of 10' Gy or more. The
higher molecular weight grade (450P)displayed a slight maxhnum in Edt) after receiving
a dose between lo4 Gy to 104" Gy. The small variations in Edt), even at a temperature
of 140 O C , confirmed the very high stability of the neutron hadiated semi-crystalline
PEEK and are indicative of the competition occurring within the amorphous phase
between the crosslinking and chah scission processes. Although ut)seemed at fint ody
slightly affected by radiation, a marked decreasing trend appeared for doses between IO^.^
Gy and 10' Gy. Therefore, at this dose range, molecular backbone movement in the
amorphous phase seems to become more important, probably caused by an increase in
lower molecular weight products resulting f?om chai. scission. On a segmental scale of
several repeat units, the effect of chain scission in the amorphous phase becomes
predominant over crosslinking afler a dose of IO*.' Gy and contriibutes more effectively to
the reduction in the mechanical properties of Uradiated PEEK

Lastly the rheology test was performed at 350 OC to measure the melt-viscosity.
At that temperature, PEEK is in one homogeneous phase and its viscosity is detemiined
by fùll molecular motion. For th& test, both resins yielded similar results. A maximum
viscosity was reached for a dose of 10' Gy foliowed by a decrease at higher dose. This
maximum, can be attri'buted to the presence of larger molecules resulting from
crosslinking. However, fkom those results, it is not clear whether crosslinkmg occurred
only in the amorphous phase or also in the crystdline phase.

This research work showed that the degradation under a combined flux of neutron
and garmna in the SLOWPOKE-2 is more significant than evidenced in previous work for
 PEEK irradiation under electron beam or gamrna. Also, the pardel work of mechanical
testing at room temperature, relaxation near Tg and melt flow above Tm provided an
effective method to better understand the role of the morphology during the radiolysis of
the semi-crystahe PEEK Mechanical testing suggested that the tie molecules between
the two phases are sensitive to chah scission, and that the chah-ends promote
crosslinking. Stress relaxation near T
, demonstrated both the occwence of crosslinking
and chah scission in the amorphous phase and greater relaxation with radiation, especially
for doses of 10' Gy and up. Chain scission appeared to induce more movement on a
molecuhr segment scale caused by the possible increase in the number of small radiolysis
products. Melt-viscosity revealed the viscous behaviour of PEEK on a full rnolecular
scale including molecuies nom the crystaUme phase. Both resins displayed the same
behaviour, with a maximum viscosity at 10' Gy followed by a decrease at higher dose.
This trend supports the occurrence of crosslinking producing larger molecules up to 1'O
Gy. The subsequent reduction in viscosity at higher dose can be attriiuted to the
predominance effect caused by chah scission over the crosslinking of only a few larger
molecules.

5.3 Error Analysis

The morphology was identifïed with confidence. The uncertainty on the

crystallinity of the samples was minimal. The set of density measurements, (five specimen
measured three tirnes each) provided for a low statistical standard deviation. Since
standard tensile tests yielded highly repeatable and superior results with a low deviation, it
can be assumed that the Ievel of defects and voids were minimal in the specimens due to
good processing and annealhg conditions. Therefore the degree of crystallinity of the
specimens, detemhed by the density, was deterrnined with a 52% relative uncertahty as
demonstrated in Appendix C.
 The degree of crystallmay was considered constant, within its initial uncertainty of
2%' throughout the irradiation dose range'slgj. Ahhough the crystallinay results
detemirtecl by DSC and X-ray =action in Figure 4.9 and 4.10 are erratic and present a
high level of uncertainty, for reasons explained in the next two paragraphs, they did not
display any trend that would indicate a deterioration of the crystak population with
incleased dosage. In addition, the density increased slightly with radiation. A significant
crystalline phase deterioration would have redted in a reduction of the density of the
material, which is not the case as demonstrated by Figures 4.1 and 4.2. The tende
strength of PEEK is dependent on Rs c~ystallinity(~~),
and displayed no reduction for the
studied dose range (Figures 4.16 and 4.17). The constant tende çtrength, the small
variations in the Young's modulus (Figures 4.13 and 4.14), the siigth hcrease m density
and the absence of trend in the crystallinay detemiined by DSC and X-ray difnaction, all
point towards a constant crystaliinity of the PEEK samples for the studied dose range.
Therefore, the crystallinity was considered constant, which provided a good base to study
the effect of radiation on the semi-crystalline rnorphology of PEEK.

The transition temperatures related with Tg,Tcand Tm were determined by DSC

with an uncertainty of -C in the wora cases as demonstrated in Table F. 1 of Appendix
F. Those temperatures were not modified by irradiation and remained within the
uncertainty of the non-irradiated PEEK values. Degradation was then considered minmial,
not causing signifïcant changes kto the molecular weight or molecular weight distniution.
The crystallinty could not be measured accurately with the heat of fusion provided by
DSC due to the multiple crystahe nature of PEEK@"~~*~?
The crystallinity values
detemiined by the heat of fusion were nevertheless plotted to detect any trend pointmg
towards a reduction in crystallinity. No patterns were observed pemiitting to conclude
into any crystallinity change in the irradiated material.

Results fiom the X-ray dsaction could not be used to determine crystalIinity
accwately because of the poor resolution of the crystalline peaks and the amorphous peak
fkom the dfiaction pattern This poor resolution cornes fiom software limitations
definmg the amorphous peak as a symmetical one, as demonstrated in Figures G. 1 and
G.2 in Appendix G, while it isn't in r e a w 4 v 4 ? The amorphous peak is more elongated
on one side as the daaction angle increases. Therefore the software attributes a portion
of the -action pattern integration at higher angles to the crystalline peaks as opposed to
the amorphous peak. This has the effect of overestirnating the crystallinity of the studied
specimens. Therefore the crystabity was determined only using the first three crystalline
peaks, at lower angles, out of the reported five. The calculated crystallinity values were
o d y used to monitor any trend pointing towards a reduction in crystallinity, which did not
occur.

AU the tests used to study the viscoehstics properties of the irradiated resins were
performed on the INSTRON which yielded highly consistent results. Mechamcal t esting
at room temperature yielded precise results with a low uncertainty as demonstrated in
Figures 4.13 to 4.2 1, the constant stress-strain curves of Figures E. 1 and E.2 and the
sample calculation provided in Appendix E. The obtained uncertainty for the tende
strength (OS- 1.O%) and the Young's modulus (1 -3%) results were compared with those
of Sasuga et al."') (being 2% and 6% respectively) for semi-crystahe PEEK following a
0-5 MGy electron beam irradiation dose. The relaxation curves in Figures 4.22 and 4.23
displayed a clear decrease in the relaxed Young's modulus at higher dose within the
uncertainty expressed by the error bars, for which a sample calculation is provided in
Appendix D.

The rheology measurements eqressed in Figures 4.24 and 4.25 also presented a
clear behaviour of the melt viscosity with increased dosage, however error bars were
larger for some irradiated resins. The uncertainty is caused by an unstable melt, which is
interpreted by the presence of micro-gels induced by non-homogeneous crosslinkllig
throughout the polper. A sample calculation on the uncertainty over the melt viscosity is
provided in Appendix H.
 Given the repeatable power of the SLOWPOKE-2which c m produce a constant
flow of radiation, the low uncertainty on the compared dose due to long exposures
between 4 and 132 hours, and the quality of the tende test samples as demonstrated by
the standard deviation obtained âom the mechanical testing, the results presented m this
thesis are considered highly reproducible.
 CONCLUSION

This study has confirmed PEEK as a resistant material to radiation and provided a
better understanding of the mechanical performance of semi-crystalline PEEK under the
effects of neutron and gamma radiation in a reactor pool environment. Three types of
viscoelastic tests were carried out on non-irradiated and irradiated PEEK as follows:
standard tende tests at 24"C, stress relaxation tests at 140°C and melt-viscosity
measurements at 350°C. The results of those three tests clearly demonstrated polymer
degradation dominated by chah scission, in both resins, d e r receiving a combined
neutron and gamma dose of loS.' Gy. Results also pointed toward the
crystalline/amorphous phase interface as a weak point in the morphology of semi-
crystabe PEEK. Although the performance of PEEK is stili outstanding in such
irradiation conditions in a reactor pool environment, the results in cornparison with
previous work, indicate a more significant degradation of the mechanical properties of
PEEK than for electron bearn radiatien'' ?

Tools made of PEEK could be used m nuclear reactors for long periods of tirne
without significant loss in mechanical properties. Owing to its high mechanical strength
and chernical resistance, PEEK could also be used in the coustruction of containers for
the transport of nuclear fuel waste without risking hazardous container detenoration.
PEEK could be used as a coating on metal (copper) container to prevent corrosion or as
the resin with carbon or boron fiber to produce aii made composite containers. In this
research, semi-cryst a i h e PEEK resins were studied up to a maximum dose of 320 kGy in
the SLOWPOKE-2. This dose corresponds to an quivalent eqosure of 3.23 yean for a
PEEK composite nuclear waste container's2)holding fuel bundles with glass as a mer and
700 years usmg Thorium dioxide as a mer. Composite containers would definitely
provide a handling advantage over the cwently used heavy titanium alloy containers.
 RECOMMENDATIONS

The scope of research work presented in this thesis, focused on the radiation
effects of neutron and gamma on the viscoelastic properties of PEEK in a SLOWPOKE-2
reactor pool environment, for doses of 1o4 Gy to "'01 Gy. Viscoelastic tests definitely
showed changes in the structure of the irradiated material. Although the degradation
mechankm seems to be larger under neutron radiation than gamma alone or electron
beam it is difficult to detemine what can be attriiuted to radiation alone and the
accelerated degradation mechanism caused by the presence of water. PEEK resins could
be irradiated in the h e r sites of the reactor in wet and dry capsules to measure the impact
of water on the radiation degradation of the polymer. Since radiation darnage was not
large enough to affect the thermal properties and the crystalluiity of PEEK,further study
needs to be done at higher dose possibly up to 10 MGy. Also, the gamma 6 0 ~source
o at
DREO could be used to measure the impact of gamma radiation alone on the mechanical
properties of PEEK.

The mechanical testing of this thesis pointed towards the phase interface as a weak
point in the semi-crystalline morphology of PEEK due to the higher radiation sensitivity of
the tie molecules between the ciystahe and arnorphous phases. It is known that the
interface in a polymeric composite present a similar vulnerabilitf) and that further tests
should be perfonned to measure the impact of neutron radiation on the mechanical
properties of PEEK composites.
 REFERENCES

O'Donnell J.H., "Radiation Chemistry of Polymers," in E. Reichmanis and J.H.

O 'Donnell, Eds., The Effects of Radiation on High- Technology Pol'ers.
Amencan Chernical Society, Washington, DC, 1989

O'DonneU J.H., "Chemistryof Radiation Degradation of Polymers", in R.L.

Clough and S.W. Shalaby, Eds., Radiation Effects on Polpers. A d c a n
Chemical Society, Washington, DC, 1991

O 'Donnell J.H. and Sangster DI.."Principles of Radiation Chemistry"; Edward

Arnold (Publishers) Ltd: London, 1970.

Sperhg L.H., Iniroduction to Physical P d m e r Science, John Wiley & Sons, 2nd
edition, New York, 1992

Tenney DR.and Slemp W.S., "Radiation Durability of Polymeric Matnx

Composites," in E. Reichmanis and J.H.OYDonnell,Eds.. The Effects of Radiation
on High-Technology Polymers, American Chemical Sociev, Washington, DC,
1989

Vaughan A.S. and Stevens G.C. Polymer 1995,36, 1531

Vaughan A.S. and Sutton S.J. Polymer 1995,36,1549

Yoda O. Poljm Commun 1985,26, 16

Yoda O. Polym Commun 1984,25,238

Sasuga T.and Hagiwara M. PoZper 1986,27,82 1

Sasuga T., Hayakawa N., Yoshida K. and Hagiwara M. Polymer 1985,26, 1039

Sasuga T. and Hagiwara M. PoZymer 1985,26,501

Hegazy E.-S.A., Sasuga T., Nishii M., and Seguchi T. PoZymer 1992,33,2904

Hegaq E.-S.A., Sasuga T.,Nishii M., and Seguchi T. Polymer l992,33,2897

U.S. Pat. 4,320,224 (Mar. l6,1982), J. Rose and P. Staniland (to ICI Americas,
Inc.)

Encyclopedia ofpolymer science and engineering 2nd Ed., John Wiley & Sons,
New York, 1986, vol 12, pp3 13-318
17. H e g a y E.S.A., Sasuga T. and Seguchi T. Polymer 1992,33,29 11

Egusa S., Seguchi T.,Hagiwara M., Nakajima H., Shimamoto S., Kirk M A . and
Biacher RC. "Radiation-Degradation Studies of the Mechanical Properties of
Polymer Matrix Composites Used in Fusion Magnets", in ILL. Clough and S.W.
Shalaby, Eds., Radiation Effects on PoZynters. Amerïcan Chemical Society,
Washington, DC, 199 1

Glasstone S. and Sesonske A., Nuclear Reactor Engineering, Van Nostrand

Reinhold, 3rd edition, New York, 1981, pp 438

Encyclopedia of poIymer science and engineering 2nd Ed., John Wiley & Sons,
New York, 1986, vol 13, pp667-703

Makhlis FA., Radiation Physics and Chmistry of Polymers, Halsted Press, a

division of John Wiley & Sons, Inc., New York, 1975

Chapiro A., Radiation Chemise of Pofpneric Systems, John Wiley & Sons, Inc.,
New York, 1962

Williams F. in M. Dole, ed., The Radiation Chembtv of Macromolecules, Vol. 1,

Academic Press, New York, 1972, Chapt.2

Ausloos, ed., Fundamental Processes in Radiation Chemistry, John Wiley &

Sons, Inc., New York, 1968

Sangster DE.,"Early Events in High-Energy Irradiation of Polymers," in E.

Reichmanis and J.H. O'Donnell, Eds., The Effects of Radiation on High-
TechnoZogy Polymers. Arnerican Chemical Society, Washington, DC, 1989

Dole M., The Radiation Chemistv of Macromolecules, Vo1.1, Academic Press,

New York, 1972

Encyclopedia ofpolymer science and engineering 2nd Ed., John Wiley & Sons,
New York, 1986, vol 4, pp418-446

Hawkins WL., Polymers degradation and stabilirarion, Polymers Properties and

Applications Senes, New York, 1984

Koch T.and Ritter H. Macromolecules 1995,, 28,4806

In ternationaZEnqydopedia of Composites, Lee S.M. ed., VCH Publishers, New

York, 1990, vol 6, pp258-269
31. Blundell D.J. and Osbom B.N. Polymer l983,24,953

Hsiao B.S., Sauer RB.,Vema RK, Zachmann H.G., Seifert S., Chu B. and
Harney P . Mncromolecules 1995,28,6931

Lanimer M.P., Hobbs J K , Hill M.J. and Barham P.J. Polymer 1992,33,3971

Kumar S., Anderson D.P.and ,4dams W.W. Polymer 1986,27,329

Cheng S.Z.D.,
Cao M.-Y. and Wunderlich B. Macromolecules 1986, 19, 1868

Harndan S. and Swailow G.M. J. PoZym. Sci.. Part B. PoZym. Phys. 1996,34,669

Jones D.P.,Leach D.C. and Moore DR. Polymer 1985,26, 1385

Chivers R.A. and Moore D.R. Polymer 1994,35, 1 10

International Encyclopedia of Composites, Lee S.M. ed., VCH Publishers, New

York, 1990, vol 4, pp156-168
M c C m N.G.,Buckley C.P.and Bucknall CB. Principles of Polymer
Engineering. Oxford Science Publications, New York, 1988, pp 294

Andrews W.S., "Themal Neutron Flux Mapping Around the Reactor Core of the
SLOWPOKE-2 at RMC". M.Eng. thesis, Royal Military College of Canada at
Kingston, 1989.

Bonin H. W., "Estimation of the doses £Yom the thermal and fast neutron flwr and
gamma radiation exposure in the pool ofthe SLOWPOKE-2 research reactor".
Report # RMC-CCE-SL2-96-2, Department of Chemistry and Chernical
Engineering, Royal Military College of Canada at Kingston, 1996

Kroschwitz J.L. Polynzersr Pol'er Characterizution and Analysiî. John Wiley

& Sons,New York, 1990, pp 850 & 898

Fougnies C., Damman P., Villers D., Dosière M. and Koch M.H.
J.
Macromolecules 1997,30, 13 83

Fougnies C., Damman P., Dosière M. and Koch M.H.J.Macromolecules 1997,

30, 1392

Tong Y. Ko and Woo E.M. Polymer 1996,37, 1167

Tung C.M. and Dynes P.J., J. ofAppl. P o l p . Sci.,33,505, (1987)

48. Lu S.X., Cebe P. and Capel M. Polymer 1996,37,2999

49. Choy C.L.,Leung W.P. and Nakafuku C . J. Polym. Sci.. P o l w . Phys., 27, 1965
(1 990)

50. VICTREX PEEK: A Guide to Grades for Injection Molding W 0 7 8 5 , ICI

Arnericas, Inc., WiLnington, Del., 1986

5 1. Devaux J., Delimoy D., Daoust D., Legras R,Mercier J.P., Strazielle C. and
Nield E. Polvmer 1985,26,1994

52. Bonin H.W., Bui V.T., Legault I-F., Davey A.R., Douglas T.D.,Duriez C.C.S and
MacGregor G.B." High Polymer Composites for Containers for the Long-Term
Storage of Spent Fuel and High Level Radioactive Waste", 18'~Annual CNA/CNS
Conference proceedings, Toronto, 1 997

53. Cole K.C. and Casella I.G. Po!mer 1993,34,741

 APPENDIX A

Injection moulding conditions for the processing of the PEEK tensile test bars
Table A.1: ENGEL 55 Injection moulder settings for the processing of the VICTREX 15OP
and 450P PEEK grades.

CONTROL SYSTEM PARAMETER SETTING

Barre1 Heat & Oil Temperature N o d e 380 OC
Barre1 zone 1 380 OC
Barre1 zone 2 360 OC
Barre1 zone 3 340 OC
Injection Boost Injection speed V 12-V 16 120 mm/s
Injection speed V 17-V2 1 90 d s
Injection boost pressure P6 110.0 bar
injection Hold Injection hold pressure P7-P 11 50.0 bar
Injection hold pressure P 12-P16 50.0 bar
Injection hold time 22 4.0 s
Cooiing t h e 24 35.0 s
Cushion length C h 3.5 mm
Screw Feeding Set screw feed stroke C 1 45.0 mm
Backpressure P17-P21 5.0 bar
Screw feed delay time 23 0.0 s
Mould Closing Mould open position A 114.3 mm
Start 2. Speed (V2) W3 50.8 mm
Start 3. Speed (V3) W1 19.1 mm
1. Closing speed (A-W3) VI 65 %
2. Closing speed (W3-W1) V2 35 %
3. Closing speed (Wl-GI) V3 15 %
Start mould protection G 1 12.7 mm
Mould prot. press. P2A-P2E 25 % .
Clamp force 45.0 t ,
Mould Opening Start 2. Speed (V7) W4 50.0 nim
Start 3. Speed (V8) W2 100.0 mm
1. Opening speed (G2-W4) V6 40 %
2. Opening speed (W4-W2) V7 35 %
. 3. Opening speed (W2-A) V8 10 %
 APPENDUC B

Irradiation exposures profile in the SLOWPOKE-2 for the PEEK speeimens

Exposures related to doses

The total radiation dose applied on the PEEK specimens was the result of several
exposures in the reactor pool of the SLOWPOKE-2.Most exposures ranged between 4 and
16 hours and were applied during weekdays depending on the availability of the reactor
y was applied m one exposure of 4 hr 10 min, the 1o ~ * ~ G ~
facility. Therefore, the 1 0 4 ~dose
dose was applied in two exposures totalling 13 hr 11 min in the reactor pool the 1 0 ' dose
~ ~
was applied m six exposures totalling 4 1 hr 40 nmi (Figure B. 1) and f i ~ l l ythe 1O ~ . ' Gdose
~
was applied in thirtem exposures totalling 131 hr 46 min (Figure B.2).

Uncertainty on exposures

The uncertainty on each exposure is at the most two minutes. Therefore, for a 4 hours
exposure the relative uncertainty is:

The corresponding sample calculation on the mcertainty for the comparative dose of
100 kGy is then:

6 exposures for a total of 41 hr 40 min (2500 mi.)= ((6x2)12500)x100% = 0.5%

The uncertainty related to the fluctuations of the reactor = 0.5%
Total uncertainty = 1.O%
 1 2 3 4 5 6 7 8 9 10 11
ihys in Irradiation sequenœ

Figure B.1: Irradiation profile for PEEK specimens exposed for a total dose of 10' Gy at
3 1.4 cm radius in the SLOWPOKE-2.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Days In Irradiatlon sequence

Figure B.2: Lrradiation profile for PEEK specimens exposed for a total dose of 10~-'Gy
at 3 1.4 cm radius in the SLOWPOKE-2.
 APPENDIX C

Density calculation of PEEK performed by the ASTM D792-A water immersion method
 Table C.1: Mass in air and mass by immersion in water of the non-irradiated PEEK 450P
grade tende test bars.

1 SPECIMEN NUMBER 1 MUS IN WATER WITH WiRE 0.0003 g) 1 MASS IN AIR 0.0003 g) 1

Sample density calculations of the PEEK tensiie test bars

Mass of the wke immersed in water = 0.4309 + 0.0003 g

Average mass of specimen 6 with wire in water (Table C.1) = 1.81O7 g

Uncertainty on mass of specimen 6 with wire in water

Mass of specimen 6 in water

(1.8 107 5 0.0019 g) - (0.4309 2 0.0003 g) = 1.37982 0.0022 g

SpecSc gravity of specimen 6

5.9952 1 (5.9952 - 1.3798) = 1.îWOglg of water at 23°C

 Uncertainty on the specific gravity of specimen 6

Specific gravity of the processed 450P PEEK grade (specimens 6,21,23,30and 39)

Specimen 6 +
1.2990 0.0009
Specimen 21 1.2990 5 0.0008
Specimen 23 1.2985 2 0.0007
Specirnen 30 1.2988 2 0.0006
Specimen 39 1.2996 5 0.0008
Average 1 .2990

Uncertainty on the specific gravity for the 450P PEEK grade

Density of the processed 450P PEEK grade

Uncertainty on the density of the processed 450P PEEK grade

Sample crystallinity calculations fkom the density of the PEEK tensile test bars

Density of the non-irradiated 450P PEEK grade = 1296 2 1 kg/m3

Density of the PEEK amorphous phaseCrncsisaf ") = 1263 + 1 kg/m3
Density of the PEEK crystalline phasem c ~ i31)
s ~=~ 1400 2 1 kg/m3

Uncertainty on the crystallinity

 APPENDIX D

Relaxed Young's Modulus versus time for both PEEK grades

Figure D.1: Relaxed Young's Modulus versus time for PEEK 150P specimens at 140°C.
Figure D.2: Relaxed Young's Modulus versus t h e for PEEK 450P specimens at 140°C.
Table D.l: Measured relaxed load (JNSTRON) of 45OP P E M specimens, kadiated for
a cumulative dose of 104Gy, afkr 20 seconds at 140°C following a 4.8% strain
applied in 1.2 seconds.

SPECIMEN NUMBER LOAD STRAIN MODULUS

(#) (t1 N) (+ .O001 d m ) (a 7 MPa)
4 510 0.0509 646
12 525 0.0489 693
20 524 0.0506 668
32 1 518 0.0489 683

Sample calculation for the relaxed Young's modulus at time t

Measured cross section = 6.30 mm x 2.46 mm = 15.5 5 0.1 mm'

Measured strain for specimen # 4 (Table D. 1) = 0.0509 f0.000 1
Measured load for specimen # 4 (Table D. 1) = 5 10 + 1 N
Relaxed Young's modulus afier 20 seconds = 5 lO/(l5.5 x 0.0509) = 646 MPa

Uncertainty on the Relaxed Young's modulus for specimen # 4

Average Relaxed Young's modulus fkom Table D. 1 = 666 MPa

Uncertainty on the Relaxed Young's modulus for 450P PEEK (10' Gy)
 APPENDIX E

Typical standard tensile test reports for PEEK

Table E.1: Typical tende test raw data at 24°C from the 450P PEEK specimens, hadiated
for a cumulative dose of 10" Gy.

SPECIMEN ULTIMATE YOUNG'S MAXIMUM ENERGY AT

(#> STRESS (at yield) MODULUS STRAlN BREAK
(+ 0.1 MPa) (+ 1 MPa) (+ 0.0 1 rnm/rnm) (+ 0.1 J)
3 102.6 1181 0.52 17.3 -
10 103.2 1214 0.82 28.5
19 103.3 1294 0.58 19.0

Sample calculation for the Young's modulus

Average Young's modulus f?om Table E. 1 = 1239 MPa

Uncertainty on the Young's modulus for 45OP PEEK (10"' Gy)

Table E.2: Typical raw data report fiom Senes IX Automated Materials Testing System 1.16
software for PEEK 450P.
ROYAL MILITARY COLLEGE
CHEM ENG
Rm 4089
PEEK450P Tende at 24OC (1'O Gy)

Test type: Tensile W o n Corporation

Senes IX Automated Materials Testing System 1.16
Test Date: 29 Nov 1996

Sample Identification: TT45P04 Sample Type: ASTM

Interface Type: 4200 Senes
Machine Parameters of test:
Sample Rate @ts/sec): 20.000 Humidity ( % ): 50
Cmsshead Speed ( d m i n ): 5.000 Temperature (deg. O C ) : 140

Dimensions:
Spec. 1 Spec. 2 Spec. 3 Spec. 4 Spec. 5

Width (mm) 6.2600 6.2600 6.2600 6.2600 6.2600

Thickness (mm) 2.4500 2.4500 2.4500 2.4500 2.4500
Spec gauge len (mm) 25.000 25.000 25.000 25.000 25.000
Grip distance (mm) 50.000 50.000 50.000 50.000 50.000

Out of 5 specimens, O excluded.

Loiid Stress Strain Slope Energy to

at at Moddus at Break
Specimen MaxLoad MaxLoad (AutYoung) MaxLoad (AutYoung) Point
Number (KN) (MPa) (MPa) (&mm) Wm) (4
1
2
3
4
5

Mean:
Standard
Deviation:

Mean -
2.0 * Sdv:
Mean +
2.0 * Sdv:

Minimum
Maximum:

CofOfVar:
Figure E.1: Typical stress-strain curves report nom the Series IX Automated Materials
Testing System 1.16 software for PEEK 450P.
Table E.3: Typical raw data report fiom Series IX Automated Materials Testing Systern 1.16
software for PEEK 1SOP.
ROYAL MILITARY COLLEGE
CHEM ENG
Rm 4089
PEEK1SOP Tensile at 24OC (1 0' Gy)

Test type: Tensile Instron Corporation

Series IX Automated Materials Testing System 1.16
Operator name: Danny Test Date: 05 Dec 1996

Sample Identification: TT 15P03 Sample Type: ASTM

Interface Type: 4200 Series
Machine Parameters of test:
Sample Rate (pWsec): 20.000 Humidity ( % ): 50
Crosshead Speed (mmhin): 5.000 Temperature (deg. OC): 24

Dimensions:
Spec. 1 Spec. 2 Spec. 3 Spec. 4

Width (mm) 6.2800 62800 6.2800 6.2800

Thichess (mm) 2,4900 2.4900 2.4900 2.4900
Spec gauge len (mm) 25.000 25.000 25.000 25.000
Grip distance (mm) 50.000 50.000 50.000 50.000

Out of 4 specimens, O excludect

Load Stress Strain Slope Energy to

at at Modulus at Break
Specimen MaxLoad Max.Luad (AutYoung) MaxLoad (AutYoung) Point
Number (KN) (MPa (MPa (mm/-) (N/mm) (J)

1
2
3
4

Mean:
Standard
Deviation:

Mean -
2.0 * Sdv:
Mean +
2.0 * Sdv:

Miaimum:
Maximum:

CofOWar:
Figure E.2: Typical stress-~traincurves report fkom the Series IX Automated Materials
Testing Systern 1.16 software for PEEK l5OP.
 Sample raw data from DSC for both PEEK grades and
sample report from the Data AnaSisis Program version 6.1
Table F.1: Raw data for the glass transition (T$, crystallization (TJ, melting (Tm)
temperahues and the heat of fusion (AHf) of the 450P PEEK grade, irradiated
for a dose of 10' Gy.
SAMPLE MASS DSC Tg TC Tm Mf
# &O.OOlrng) RUN &O.SOC) (+ 0.5 OC) &-0.5 OC) (+ 0.5 J/g)
First DSC Run
18 10.866 1 165.5 221.4 341.O 43.2
14.125 1 165.9 219.1 340.7 40.0
22 14242 1 165.3 220.9 340.9 38.8
14359 1 166.5 220.2 340.4 41.4
Second DSC Run
18 10.866 2 149.1 287.7 340.6 44.8
14.125 2 149.4 288.2 340.7 43.6
22 14242 2 150.9 286.9 340.8 43 .O
14359 2 150.2 287.1 340.5 43.2

Sample calculation for the heat of fusion AEIfafter the Brst DSC run

Average heat of fusion fiom Table F. 1 = 40.9 J/g

Uncertainty on the heat of fusion for 450P PEEK (10' Gy)

Sample crystaiïinity calculation from the heat of fusion of the PEEK specimeos

= 130 2 1 Jlg
Heat of fusion of an all crystailine PEEK ~arnpIe@~)
Heat of fusion of the 450P PEEK (10' Gy) sample = 41 2 2 Jlg

Determineci crystallinity fiom the heat of fusion

Uncertainty on the crystaUinity

 Sampla PEEK 450P #24 (O Gy] F i l e : D: . . . \RESEARCH\bPISaP24.008
Size: 19.9950 m g
Method: TgTcTn Octemination
DSC Operator: Danny P i g e
Run D i t e : 27-Nov-96 14:20
Cornent: 10 C/min
I-

!-

-r
A O 100 200 300 400 500
ex0 Temperature (Y) Unlvsrsrl VI .61 TA fnstruments
Figure F.l: Typicai heat flow repon fiom the Data Aoalysis Program version 6.1 for the PEEK 45OP (Run 1 )

S i i p l a : P€EK 4 M P 124 ( O Gy) F i l * : O : ...\ R€SEARCH\DP450P24.001

Size: 10.5740 ng
Method: TgTcTm Oeterrrnation
DSC O p e r i t o r : Oanny P i g e
Run Date: 26-Nov-96 1 4 : 1 4

-0.B!
A O
1

50 100
4

150
I

200
t
250
I
300 350
t 1
400
0x0
T ~ n p e r a t u r c(OC) Universal V l .6I TA Instruments

Figure F.2: Typical heat flow report ftom the Data Analysis Program version 6.1 displaying from left io right,
Tg, the exothenn resuiting from Tc and the exothenn resulting nom Tm ( R n 1)
 Simple: PEEK 4 W 824 (O G y ) File: O : . . \UES&ARCii\OP45ûP24.009
Srze: 19.9950 mg
Mcthod : TgfcTm Oeterni~tion
DSC O p e r a t o r : Danny Page
Run Oate: 27-~ov-96 15:28
Cowent: 10 C/ain
3

CS0 Universil V I . 6I TA I n s t r u n m t s
Temperatura ('C)
Figure F3: Typical heat ffow report from the Data Analysis Program version 6.1 for the PEEK450P (Run 2)

Sample; P€EK 450P 124 (O G y ) F i le : D: . . . \RESEARCH\OP450P24.004

Sfzi; i8.5740mg
Method: TgTcTi Ostirmination
DSC O p e r a t o r : Danny Page
Run D i t e . 27-Nov-96 09: 06
Comment: 10 C/min
0.0

mxo Unf versa1 V I .6I TA

Figure F.4: Typical heat flow report fiom the Data Analysis Program version 6.1 displaying from left to right,
the exotherm resulting from Tc and the exotherm renilting f?om Tm (Run 2 )
 APPENDIX G

Sample raw data from X-ray diffkaction for both PEEK grades and
sample report from the DMS 2000 Diffraction Management System software
Table G.1: X-ray difnaction peak pattern integration for PEEK 450P grade, inadiated for a
cumulative dose of 104.' Gy.

SPECIMEN PEAKAT PEAKAT PEAKAT PEAKAT PEAKAT PEAKAT TOTAL

(#) 18.7' 20.2" 20.6" 22.6" 28.7" 32.8" INTEGRATION
27a 154950 552828 10591 86041 90057 4062 898529
27b 173749 528403 10194 85570 114178 5639 917733
average 164350 540616 10392 85806 102118 4851 90813 1
deviation 9400 12î13 198 h 236 12061 789 34897

SampIe crystallinity calculation from the crystalline diffraction peak integration at

18.?,20.6' and 22.6 O

Total integration for the peaks at 18J0,20.6 and 22.6 = 260548 2 10070
O O

+
Total integration of the difnaction pattern = 90813 1 34897

Determined crystallinity fkom the integration

Uncertainty on the crystaUinity

 FU: dpueek12.RD ID: PEEK SAMPLE 112 Sc1NTAo/usA
DATE: 12/11/96 T l M E 1% OB PT: O. 90000 STEP: 0.03000 Wt: 1.54060

PU9 2-THETA PK-HGT FWHM

1 10.8498 1300 0 . 7 1 3
2 20.1653 7117.224
3 22.5698 5991-O05
4 23-6A85 3 d L 2.000
5 29.6373 247 3.658
6 32.8463 53 1.596

Iteration 9 Error 5.00

Lino Tao-tneta eigma Psak (CPS) sigma ES0 Fwnm ES0 Exp Area (CPNJ
1 1e.649a 0.0041 iàï9.91 2 e . i ~ o-oi*~ 0.14 123063.OS
. 2 20.1653 0.0277 711.40 17.66 0-6616 1.05 502160 -44
3 22.5698 0 -0007 5Q9.69 12.53 0-OZ53 3.80 SQOB7.41
4 28.6484 0.0252 340.80 6.39 0.0595 0.05 112656 .SB
5 20.6379 0.0153 247 35 16.1B 0.0574 . 11.10 I4682. W
6 32.8463 O -0552 52-75 4.16 0.2244 16.30 8221.69

Figure G.1: Typical DMS 2000 report for PEEK l5OP X-ray difûaction.

FEI: peek25a .RD IO: PEEK SOLI0 (TRANSVERSE) SCIUTAG/USd

D A T E 12/03/96 TI= 09: 59 PT: 0.90000 ST€R 0.O3000 HL: 1 .5406C

1 ie.aesi 1-1 o.Ta

2 20.2216 814 7.220
3 22.5932 7Bi 1.105
I 28.7149 3602.262
5 20.6796 tdd ri.568
6 32.8385 46 1.846

ïtaration 4 Error 4.61

. . - - --

Lin. Two-theta sigma Psak (WS) sigma ESO Fuhm ESD Exp Are0 [CPM]
1 18 -6861 0 .O035 1880.73 31.50 0.0123 0.12 168t44.80
. 2 20.22i6 0.0269 814.14 19.29 0.0678 0.99 574362.25
3 22.5932 0.0078 7UL. 17 14.27 0.02üB 3.03 85550.66
4 28.7149 O.OS65 360.46 6.55 0.0663 0.05 139586-36
Ci 20.67BS 0.02Ll 188.24 18.52 0.06Q7 . 15.25 10449.88
6 32.8385 0.0709 46.44 4.45 0.2822 15.82 8368-74

Figure G.2: Typical DMS 2000 report for PEEK 450P X-ray deaction.

G-3
 APPENDDC: H

Typical rheoiogy test reports obtained for PEEK

 Table H.l:Raw data for the corrected viscosity at 350°C and 100s-'of the 150P PEEK
grade, after irradiation at a dose of 104' Gy.

1 TIME INBARREL 1 CORRECTED MELT WSCOSITY l

(+ 10 seconds) (+ 1 Pas)
1100 456
1500 457
L 800 459

SampIe calculation on the uncertainty for the melt viscosity

Average melt viscosity fkom Table H.1 = 468 Pa.s

Uncertainty on the melt viscosity for 150P PEEK ( 1 0Gy)

~~~
 RHEOLOGY TEST
Sanale Id. PEEK I W P ( 8-6Y
Temperature: 350 Capi llary Nunber: 351

Run S Ueloci t y Shear Final S hear Apparent Remeining

Rate Laad Stress uiscosity Barre1
1 /sec N Pa Pa-sec mm
1 3.401 E+000 36.25 3.811E+003 1 .121 E+B03 128.7
2 6.802€+080 6.633E+003
63.09 9.752E+002 127.1
3 1.360Et00 1 107.39 1.129E+004 8.300Et002 125.0
4 2.721€+001 194.63 2.046€+004 7.52 1 €+a02 121.1
5 5.441E+081 324.83 3.4tSE+004 6.2?7€+882 116.9
6 1.088€+002 542.28 5.701€+004 5.239€+002 112.2
Out o f 6 Runs, 6 were succtssPu1.

Pouer Law Factor: .78473

R a b i n o u i t a c h Correction F a c t o r : 1.0686
Y-Intercspt on Log-Log Plot : 3.1683

Run a Corrcc t e d Corracted Calculated Calc - Exp

Shear Rate Viscosity Shear Stress Shear S t r e s s
l /sec Pa-sac Pa Pa
3.6340E+000 1 .8488€+863 3.8496€+083 3,8342Et00t
?.2680€+000 9.1265EtB02 6.6319E+003 -.1210E+001
I.4536E+801 7.7674€+04)2 1.1425E+004 1.3451E+002
2 . gQiïZE+GM1 7.0387€+002 .
1 g683E+884 - .7799E+063
5.8144€+001 5.8736E+002 3.3909E+004 -.2426E+003
.
1 l629E+@BZ 4,9028€+002 5.8418E+004 1.4035Et003
-
RHEOLOGY TEST
PEEKiJBPC0-GY) a t Temp .
L O G L O G PLOT
350

1
1 2 3
L o g S h e a r R a t e (lfsec)

Figure H.l: Typical raw data report fkom the Rheology Test Software (REO) presenting
the corrected viscosity for PEEKISOP.
 RHEOLOGY TEST
Samplt Id. PEEK450P(0 6 Y )
Temperature: 350 Capillary Number: 351

Run # Vtlocity S hear Final S hear Apparent

Rate toad Stresa uiscosity
mm/min I /sec N Pa Pa-sec
1 1.000 3.401E+000 246. Y8 2 . WiE+884 7.63GE+003
2 2.000 6 . 802E+000 391 .Y5 4.121€+004 6.059€+003
3 4.808 1.360E+001 548.99 5 .?ïZE+884 4.243Et003
4 8.000 2.721E+001 ?83.89 8.242€+884 3.029€+003
5 16.000 5.441E+001 1 162.40 .
1 ZïZE+085 2.246E+003
6 32.000 1 .088€+002 1836.90 1 .925€+005 1 .769€+003
Out o f 6 Velocity Runs, 6 were successful.

P a w e r Law Factor: .56198

Rabinowitsch Correction F a c t o r : 1,1949
Y-Intercept on Log-Log P l o t : 4.1247

Run # Correct ad Corrected Calculated Calc -Exp

Shear Rate Uiscoalty Shcar Strass Shsar S t r e s s
1 /sec Pa-sec Pa Pa
1 4.0634€+808 6.3904E+003 2 . 6 5 1 SE+804 S.4532E+002
2 8.1269E+000 5.0707€+003 3.91 48E+684 -.2068€+004
3 1.6254€+001 3. S l 2 E + 8 6 3 5 .??82E+084 6.2599E+00 1
4 3.2507€+00 1 2.5353E+003 8 . S384€+804 2.8872E+003
5 6.501SE+001 1.8798€+003 3 .2593€+005 3.7214Ec003
6 1.3003€+002 1 .4884€+083 1.8591 €+a05 -.6581€+004
RHEOLOGY TEST - LOG LOG PLOT
P E E K 4 5 9 P C t î G Y > r t T i m p . 359

Figure H.2: Typical raw data report fiom the Rheology Test Software (REO) presenting
the corrected viscosity for PEEK450P.
 Runt Veloclty Shsar Final Shsar &parsnt Re~lntng
Rats Laad Stress Vrscosify Harrrl
rin/nln 1/sec N fa Pa-asc nn
I 1 .%O8 3.401 E+BB8 246.98 2 -557Et084 7.636E+003 31.8

Figure H.3a: Typical raw data report fiom the Rheology Test Software (REO) presenting
the apparent viscosity for PEEK450P (part 1).
 Y EXTENSIOH C i m l
1

6
+
6 32.006 .
1 €îO8E+882 l83O.98 1 .92SE+0û5 1 .70'9E+003

6
EXTENSION <ma >

Figure R3b: Typical raw data report f?om the Rheology Test Software (REO)presenting
the apparent viscosity for PEEK450P (part 2).