Académique Documents
Professionnel Documents
Culture Documents
7 May 1997
The author has granted a non- L'auteur a accordé une licence non
exclusive licence dowing the exclusive permettant à la
National Library of Canada to Bibliothèque nationale du Canada de
reproduce, loan, distnbute or seIl reproduire, prêter, distribuer ou
copies of this thesis in microfom, vendre des copies de cette thèse sous
paper or electronic formats. la forme de microfiche/fïlrn, de
reproduction sur papier ou sur format
électronique.
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.
...
lll
Staff Officer, Canadian Forces Comunications Command,
Ottawa, Mar 88-Jul90
vii
TABLE OF CONTENTS
..
Abstract............................................................................................................... II
...
Acknowledgments................................................................................................ 111
Vita .............................................................................................................. iv
Symbols & Abbreviations .................................................................................... vi
...
Table of Contents................................................................................................. viii
ListofIllustrations............................................................................................... xii
m oDUCTION.........................................................................
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............................................................
...
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.....................................................................................
CONCLUSION................................................................................................... 76
RECOMMENDATIONS...................................................................................... 77
References............................................................................................................. 78
APPENDICES:
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
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.
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.
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.
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(''.'~).
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.
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.
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
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(".
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('.~*).
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.
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*
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.
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.
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
+ H20 -
- OH* + H,O+ (2.13)
-
radiation
H20 H202, HZ,Ha , eq (2.14)
radiation
H20 H20*
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
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
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) .
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.
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.
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.
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
3.2.2 Apparatus
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.
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.
3.3.1 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
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!
PEEK Samp
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:
3.5.1 Description
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.
3.6.1 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
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.
3.7.1 Description
3.7.2 Apparatus
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.
3.8.1 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):
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
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.
3.9.1 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)
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
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
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
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
2o 4 10
4a 10
4.5
I
Dose (Gy)
5.0
10
I
10
S.s
I
E = 1200 MPa
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.
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.
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.
Table 4.1: Power law factor (n)at studied doses with the calculated standard deviation
based on a population of five measurements.
Dose (Gy)
Figure 4.25: Melt viscosity at 350°C and 100 s-l as a function of dose for PEEK 450P
Chapter 3
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.
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
Sperhg L.H., Iniroduction to Physical P d m e r Science, John Wiley & Sons, 2nd
edition, New York, 1992
Sasuga T., Hayakawa N., Yoshida K. and Hagiwara M. Polymer 1985,26, 1039
Hegazy E.-S.A., Sasuga T., Nishii M., and Seguchi T. PoZymer 1992,33,2904
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
Encyclopedia of poIymer science and engineering 2nd Ed., John Wiley & Sons,
New York, 1986, vol 13, pp667-703
Chapiro A., Radiation Chemise of Pofpneric Systems, John Wiley & Sons, Inc.,
New York, 1962
Encyclopedia ofpolymer science and engineering 2nd Ed., John Wiley & Sons,
New York, 1986, vol 4, pp418-446
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
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
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
Fougnies C., Damman P., Villers D., Dosière M. and Koch M.H.
J.
Macromolecules 1997,30, 13 83
49. Choy C.L.,Leung W.P. and Nakafuku C . J. Polym. Sci.. P o l w . Phys., 27, 1965
(1 990)
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
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.
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:
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
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
Sample crystallinity calculations fkom the density of the PEEK tensile test bars
Uncertainty on the Relaxed Young's modulus for 450P PEEK (10' Gy)
APPENDIX E
Dimensions:
Spec. 1 Spec. 2 Spec. 3 Spec. 4 Spec. 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)
Dimensions:
Spec. 1 Spec. 2 Spec. 3 Spec. 4
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
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
!-
-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 )
-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 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.
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
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.
. . - - --
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
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
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).