Sunteți pe pagina 1din 28

Chapter 6

Structure Development and


Mechanical Behavior During Uniaxial
Drawing of PET
U. Gschel
1. Introduction
The knowledge of structure formation with a focus on orientation and crys-
tallization is essential for the control of polymer processing and the result-
ing, e.g., thermal and mechanical properties. Below, on the basis of a mul-
tistep uniaxial drawing procedure, the fundamental changes in structure
and deformation behavior of a homopolymer PET grade are discussed in
detail (see also Chapters 3, 4, 7, and 9).
2. Initial PET
2.1. Structure
The initial material was an isotropic (birefringence < 0.5 x 10~
3
) and
amorphous poly(ethylene terephthalate) (PET) strip (as determined by
wide and small angle X-ray scattering (WAXS and SAXS)), 0.18mm thick
and 2.8mm wide [I]. The density, p, of about 1.34gcm~
3
, as measured
by means of a density gradient column [2], is characteristic of non-ordered
PET. The weight-average molar mass, M^, of about 37000 and the poly-
dispersity, M
w
/M
n
, of about 1.69 (the number-average molar mass, M
n
, is
ascertained by means of size exclusion chromatography (SEC) with triple
Handbook of Thermoplastic Polymers: Homopolymers, Copolymers, Blends, and Composites
Edited by Stoyko Fakirov
Copyright 2002 WILEY-VCH Verlag GmbH, Weinheim
ISBN: 3-527-30113-5
290 U. Gschel
detection, z.e., by a concentration method, viscometry, and light scattering
[2]) are in the normal range for PET.
2.2. Crystallization
2.2.1. SAXS studies
SAXS experiments on the initially isotropic Sample O did not reveal long
spacing at room temperature [3]. A Kratky compact camera (Anton-
Paar K.G., Graz), Ni-filtered CuK
0
,-radiation and a one-dimensional (ID)
position-sensitive detector (OED-50-M, from Braun GmbH, Garching) were
used. The measured intensities are desmeared and corrected for absorption
and background scattering. The temperature was gradually raised from 30
to 16O
0
C in steps of 1O
0
C, the sample was annealed for 10 min and then
irradiated for 30 min at each temperature. A clear meridional long spac-
ing was observed at temperatures exceeding UO
0
C (Figure 1); the further
temperature rise to 16O
0
C resulted in an improvement of the crystalline
order, indicated by an increase in the long spacing reflection intensity and
a shift toward larger scattering vectors with a smaller half-width of the
peak intensity. The long spacing of the initially isotropic PET Sample O
after crystallization at 16O
0
C was 9.4nm. Prolonged annealing for up to
90 min at 16O
0
C did not affect the peak position and intensity.
1.0
63 (nm-
1
)
1.5 2.0
Figure 1. Development of the SAXS long spacing with temperature of annealing
for 10 min of the initially isotropic non-crystalline PET Sample O
Structure Development and Mechanical Behavior... 291
2.2.2. WAXS studies
By means of a wide angle X-ray Siemens - diffractometer D500T
equipped with a scintillation counter in transmission mode, the onset of
crystallization for the initially isotropic PET Sample O was found to be at
ca. UO
0
C at the chosen annealing and irradiation times of 10 and 60 min,
respectively.
2.3. Thermal behavior
Differential scanning calorimetric (DSC) measurements performed with a
Mettler DSC 821e/700 at a heating rate of 1O
0
CnIm-
1
(Figure 2) show a
narrow exothermic crystallization peak at about 134
0
C with an enthalpy of
about 37 Jg
-1
, followed by an endothermic melting peak with an enthalpy
of about 41Jg"
1
, which is characteristic of the isotropic non-crystalline
PET [4]. The onset of the glass transition, T
9
, was at 78
0
C.
The dynamic mechanical behavior was studied by means of a Rheovi-
bron DDV II apparatus at a small steady strain, , in the linear deformation
range superimposed by a small sinusoidal strain, , of about 0.1%. The
sample was 20 mm long and the frequency was 110 Hz [5] . The loss modulus,
Glass transition Crystallization

1
Melting
40 60 80 100 120 140 160 180 200 220 240 260
Temperature (
0
C)
Figure 2. DSC traces of the initially isotropic non-crystalline PET Sample O ()
and the cold drawn Sample 1 ( )
292 U. Gschel
Figure 4. Stress-strain curve of an initially isotropic non-crystalline PET structure
at 23
0
C for two different deformation regions
Structure Development and Mechanical Behavior... 293
E"', reveals two peaks with maxima at about 45 and 88
0
C, which were
attributed to a secondary transition and the glass transition, respectively
(Figure3). The secondary transition suggests a local mobility, which does
not affect the material stiffness, as indicated by the storage modulus E
1
.
Such mobility in a short segment of the molecular chain can be enhanced
by, e.g., an external force. The glass transition is characterized by coopera-
tive segmental mobility in the main chain and is, therefore, of fundamental
interest for the study of conformational changes during drawing.
2.4. Deformation behavior
Mechanical tensile experiments on isotropic non-crystalline PET at 23
0
C
and low strain rate, ds/dt, of 0.1 min"
1
[6] provide an elastic modulus
E = 2 GPa, a strain limit of linearity of 1.1%, a yield point at the strain,
, of 2.2% (the stress being 38MPa), a broad range of plastic deformation
from 2.5 to about 330%(where the stress as the actual force, F^ divided
by the initial cross section, A
0
, is constant at about 3OMPa), and a strain
at break reached at 400%(Figure 4).
3. Drawing procedure
A uniaxial multistep drawing procedure of at most three steps at preset
temperatures was applied to an isotropic and initially non-crystalline PET
structure, using either a homogeneous or an inhomogeneous (zone drawing)
temperature field [I]. Starting under cold drawing conditions, the temper-
ature was raised at each drawing step to approach gradually the final ulti-
mate draw ratio (Figure 5). The temperature, T^
7
., stress, ^

, and velocity,
Vdr, are regarded as the drawing parameters.
3.1. Cold drawing
In the first drawing step, a temperature in the vicinity of the glass transi-
tion (T
9
w 7O
0
C) or even lower was chosen in order to avoid crystallization.
Such a drawing with propagation of the neck through the material is very
sensitive to the stress applied. As seen in Figure 6, the stress range, where
drawing can be performed, decreases with decreasing temperature [I]. A
minimum stress is required to initiate drawing. The maximum stress results
in inhomogeneous deformation and crazing. The bold lines in Figure 6 char-
acterize the regions of uniform drawing at two different temperatures of 68
and 8O
0
C. At high temperatures (16O
0
C) and low stresses (< 2MPa), the
additional phenomenon of chain slippage, characterized by an extraordinar-
ily high draw ratio (5 to > 8) without any significant orientation, can be ob-
served. The stress of uniform drawing using zone drawing at 100 mm min"
1
was determined to be ca. 20, 10, and 2 MPa at the drawing temperatures
of 160, 80, and 68
0
C, respectively. A broad range of drawing stresses is of
294 U. Gschel
8.0
7.0
6.0

Q
5.0
4.0
J
Isotropie and
non- crystalline PET
Figure 5. Scheme of the drawing procedure
great importance for industrial processing to meet the requirements for a
low failure drawing by making use of a less critical ratio of winding speeds.
Starting from an isotropic and non-crystalline PET material, three dif-
ferent Samples 1,6, and 8 were obtained as a result of a one-step drawing at
the temperatures of 68, 80, and 23
0
C, respectively (Figure5 and Table 1).
Sample 1 was prepared by a zone drawing procedure at a low Vd
r
=
5.5 mm min"
1
, defined as the relative velocity between the undrawn ma-
terial and the locally fixed band heater. The chosen drawing stress ^

=
15MPa is within the narrow range of uniform drawing (Figure 6).
Structure Development and Mechanical Behavior... 295
Table 1. Drawing conditions in the first step
Sample
1
6
8
Drawing
procedure
Zone drawing
Zone drawing
Drawing in a
homogeneous
temperature
field
Drawing
temperature,
T
dr
(
0
C)
68
80
23
Drawing
stress,
&dr
(MPa)
15
38
30
Drawing
velocity,
Vdr
(mm/min)
5.5
100
3.0*
Draw
ratio,

4.3
5.6
4.2
Bire-
fringence,

0.177
**
0.200
* Corresponds to a traverse velocity of 3 mm% *
**None, opaque sample
Sample 6 was zone drawn at Td
r
= 8O
0
C and Vd
r
= 100 mm min"
1
.
The applied drawing stress of 38 MPa is extremely high, beyond the range
of uniform drawing (Figure 6) and close to the ultimate stress Fi/A
0
in
Figure 4. These drawing conditions resulted in an opaque and wavy sample
with an extraordinarily high draw ratio of 5.6 [I].
Using a homogeneous temperature field and a tensile machine equipped
with a temperature chamber, Sample 8 was prepared at a low traverse
velocity of 3mm%~
1
and a temperature of 23
0
C (Table 1). With respect to
the achieved draw ratio of 4.2, the neck has propagated through the whole
material with a velocity of 0.71 mm%~
1
[I].
6
O
5
I
Q
4
3
2
T
dr
= 80 C
_L
10 20 30 40
Drawing stress, <

(MPa)
50
Figure 6. Draw ratio vs. drawing stress of PET at 68 and 8O
0
C using zone drawing
conditions and drawing velocity of 100 mm min"
1
. The bold lines illustrate the
regions of uniform drawing
296 U. Gschel
3.2. Hot (post)drawing
The one-step drawn Samples 1,6, and 8 were used as starting materials in
further zone drawing steps (hot drawing) at elevated temperatures (160 to
23O
0
C, see Figure 5). Each drawing step is characterized by a certain draw-
ing temperature. Drawing in the second and further steps was performed
at different drawing stresses, gradually increased in 5 to 10 stages toward
the maximum attainable value at the preset drawing temperature. There-
after, a higher drawing temperature was applied. After the second drawing
step (16O
0
C), the gradual increase in the drawing stress from about 70 MPa
made it possible to approach its ultimate value almost linearly. Compar-
ison of the first and the final (third) drawing step reveals an increase in
the drawing stress from 15MPa (Sample 1) to 40OMPa (Sample 4), which
caused a significant increase in the draw ratio from 4.3 to 7.3, respectively
(Table 2).
Table 2. Drawing conditions
Sample
1
2
3
4
5
6
7
8
9
Drawing temperature,
T
dr
(
0
C)
68
68 and 160
68, 160, and 200
68, 160, and 230
68 and 200
80
80 and 160
23
23 and 180
Maximum drawing stress,

(MPa)
15
350
385
400
400
38
250
30
100
Draw ratio,

4.3
6.3
7.1
7.3
7.2
5.6
6.2
4.2
5.4
Samples 2,3,4, and 5 originate from Sample 1 by reaching the maximum
attainable drawing stress in the second and the third drawing steps. In the
case of Sample 5, a high ^ = 7OMPa was suddenly applied to the non-
crystalline and one-step drawn Sample 1 at the very high temperature of
20O
0
C. The questions arose of whether Sample 1 can withstand the loading
(7OMPa at 20O
0
C) and undergo an enormous structural change. In the
further stages at 20O
0
C, the drawing stress was increased to 35OMPa in
increments of 50 MPa.
Drawing of Samples 6 and 8 led to Samples 7 and 9, respectively (Fig-
ure 5). Sample 6 had already experienced a high drawing stress ^

=
38MPa, which is beyond the normal range at the applied T^ = 8O
0
C.
As a result of such ultimate conditions, a very high draw ratio = 5.6 was
achieved at this low drawing temperature. However, Sample 6 is opaque
and wavy, compared to the transparent and flat Samples 1 and 8. Further
drawing of Sample 6 by increasing the stress from 38 to 25OMPa led to
the opaque but flat Sample 7 at the draw ratio = 6.2 which is rather low
Structure Development and Mechanical Behavior... 297
compared to > 7 for Samples 3, 4, and 5 (Figure 5).
Sample 9 originated from the cold drawn (23
0
C, = 4.2) and trans-
parent Sample 8 by suddenly applying a stress ^
7
. = 100 MPa at 18O
0
C
without further drawing stages.
4. Structure of drawn PET
4.1. Structure after cold drawing
The multistep drawing procedure was mainly performed via an oriented
and non-crystalline intermediate structural stage in order to achieve, finally,
extraordinarily perfect chain orientations in a crystalline morphology as the
prerequisite for high mechanical stiffness and strength coupled with good
dimensional stability [7].
Drawing in the first step of an isotropic non-crystalline PET material
led to a well oriented non-crystalline structure in Samples 1 and 8, whereas
Sample 6 is semicrystalline according to WAXS [8,9] and SAXS studies
[3,8]. With respect to the molecular order, represented by the degree of
crystallinity, Di
05
/L, of about 28%(as determined by the average crys-
tallite size from the (05) reflection, >
0
5, and the meridional SAXS long
period, L [8]), Sample 6 will be discussed together with the semicrystalline
samples from the hot drawing.
WAXS experiments on Samples 1 and 8, using an ISkW rotating Cu
anode X-ray source (Rigaku RU-300) together with a Nicolet area detector,
showed the appearance of a broad amorphous peak along the equator, ie.,
perpendicular to the draw direction, with a maximum at the scattering
angle 2#cu = 21.6 [9]. The weak reflection along the meridian at 2#cu =
25.7 can be attributed to a conformational regularity [9].
The birefringence, , determined at 23
0
C and a wavelength of 551 nm
using a Zeiss polarization microscope together with an Ehringhaus com-
pensator [1], is 0.177 and 0.200 for Samples 1 and 8, respectively (Table 1).
These values are very high, as compared with the birefringence (0.275) of
a perfectly oriented amorphous PET sample, according to Dumbleton [1O].
Despite almost the same draw ratio of the order of 4.2, the chain molecules
in Sample 8 are significantly more highly oriented than those in Sample 1.
4.1.1. Thermally stimulated structural changes in oriented non-
crystalline PET
Oriented PET structures without any indication of crystallization from
WAXS and SAXS experiments can be prepared under certain conditions of
cold drawing [U]. It is well known that these structures are highly sensitive
to thermal and mechanical loading. Especially, processes such as crystalliza-
tion and relaxation will take place. Crystallization in PET upon orientation
298 U. Gschel
was investigated by several authors [12-17]. The onset of crystallization de-
creases significantly with deformation, as discussed by Alten and Zachmann
[14]. Thompson [18], as well as Smith and Steward [19], reported crystal-
lization times of the order of milliseconds in the highly oriented state,
compared to minutes under isothermal conditions in the isotropic state.
Using infrared spectroscopy and DSC, Galli et al [20] suggested that PET
chain segments in the trans conformation start to crystallize already below
the glass transition. Nobbs et al. [21] observed crystallization of oriented
PET at low temperatures in the vicinity of 6O
0
C. Oriented non-crystalline
polymers show a structural stability which is determined by the molecular
mobility of the chain segments and, consequently, dependent on the temper-
ature and the magnitude of internal stresses [22,23]. According to Trznadel
and Kryszewski [23], the appearance of shrinkage forces is mainly associ-
ated with the gradual relaxation of internal stresses frozen in the sample
after deformation. Pakula and Trznadel [24] studied the mechanism of gen-
eration of shrinkage forces by proposing a model to describe the shrinkage
phenomenon. Despite the large number of investigations, the temperature-
dependent behavior of oriented chains in a non-crystalline phase is not yet
completely understood. The question arises of how relaxation and crystal-
lization affect each other, depending on temperature and microstructure.
In the following, the cold drawn Samples 1 and 8 will be used as an example
of an oriented non-crystalline PET structure. As seen in Table 1, the chain
orientation is higher in the more oriented Sample 8. A variety of methods
were applied to study the effects of temperature on the changes in chain
mobility, molecular order, and orientation [U].
Dynamic mechanical behavior. Using a Rheometrics mechanical
spectrometer (RMS-800), the temperature dependence of the storage mod-
ulus, E
1
', loss modulus, E", and loss tangent, tan = E"/E
1
', were studied
[U]. A torsion movement at a frequency of = lOrads"
1
was applied
to the 20mm long sample in order to create a mechanical loading along
the draw direction. A heating rate of 3
0
CnIm"
1
was chosen to run the
experiment from 100 to 27O
0
C. Figure 7 describes the characteristic dy-
namic mechanical behavior of oriented and non-crystalline PET structures
in Sample 1.
The loss tangent reveals two maxima at about 80 and 127
0
C. A com-
pletely isotropic non-crystalline PET structure used in earlier studies [6]
shows an identical peak at 8O
0
C with a maximum in the loss modulus of
about 50OMPa. Consequently, the dispersion region in the vicinity of 8O
0
C
(Figure 7) corresponds to the glass transition of the non-crystalline phase
without any hindrance of the molecular mobility. The second dispersion re-
gion in the vicinity of 127
0
C is attributed to the mobility of chains, which
is hindered by orientation and crystallization. Such an interpretation is as-
sociated with separation of the non-crystalline phase into a mobile and a
Structure Development and Mechanical Behavior... 299
10
10
E'
10
9

7
I 10*

4
10
_ ^ tan 5
10

10
2
-120 -80 -40 O 40 80 120 160 200 240 280
Temperature (
0
C)
Figure 7. Effect of temperature on the storage modulus, E'', loss modulus, E"',
and loss tangent, tan, of the oriented non-crystalline PET Sample 1
rigid fraction (see, e.g., Schick et al [25]). Based on a concept of cooperative
length and an experimental approach by means of dielectric spectroscopy
and DSC, the glass transition is found to depend on the average thickness
of the amorphous layers, as discussed by Dobbertin, Hensel, and Schick
[26].
Comparing Samples 1 and 8, the higher chain orientation in Sample
8 causes a lower peak intensity of E" and tan at about 8O
0
C. This is in
accordance with the suggestion in [26] that the rigid non-crystalline fraction
does not participate in the glass transition.
The storage modulus in Figure 7 shows an unusual increase with tem-
perature in the range from 57 to 66
0
C. Such an increase in the material
stiffness with temperature can only be explained by crystallization, which
would strengthen the molecular network. A further increase in temperature
leads to a decrease in E", suggesting a continuation of the crystallization
superimposed by an orientational relaxation [U].
Crystallization. According to Daubeny et al [27] (see also Chapter 3),
PET crystallizes in a triclinic unit cell. In the case of oriented crystallites,
three strong (010), (10) and (100) reflections are observed, with their
lattice planes parallel to the c-axis (Figure 8).
300 U. Gschel
118'
(010) (10) (100)
Figure 8. Unit cell planes of PET corresponding to the crystalline (010), (110),
and (100) reflections. The lattice planes are parallel to the c-axis
All of the main WAXS reflections in the angular range 20 < 30 at
ACU = 0.154nm are represented schematically in Figure 9. The (05) re-
flection at the wide angle 20 w 43, which can be used to determine the
crystallite orientation [8,28,29], is not shown. Such an angle is often be-
yond the usual measuring range. This problem can be solved by inclining
-(03)
hk3
hk2
hkl
hkO
17.3 22.5 25.7'
Figure 9. Schematic representation of the main crystalline WAXS reflections of
oriented PET in the angular range 20 < 30 at ACU = 0.154nm, where the c-axis
of the crystallites is vertical
Structure Development and Mechanical Behavior.. 301
9O
0
C
Figure 10. Effect of annealing temperature on the 2D WAXS patterns of the
oriented non-crystalline PET Sample 1. The arrow denotes the draw direction
the sample or the detector with respect to the X-ray beam.
Two-dimensional (2D) WAXS experiments (Rigaku RU-300 together
with a Nicolet area detector) were performed at different temperatures of
30, 50, 60, 70, 80, 90, 100, 120, and 16O
0
C on the oriented non-crystalline
Sample 1, annealed for 2 h and then irradiated (Ac
u
) for 4 h at each tem-
perature. The sample was mounted between two clamps [14]. Annealing
temperatures in the range from 30 to 6O
0
C do not affect crystallization.
At 7O
0
C, a small increase in the peak intensity and a starting defor-
mation of the WAXS pattern, characterized by the formation of the non-
equatorial reflections (Oil) and (111), as well as one close to the (Oil)
and (111) reflections can be noticed (Figure 10). These have been verified
by peak analysis and the peak positions are shown in Figure 9. Such a
phenomenon of growing reflections relates to the onset of crystallization.
302 U. Gschel
800
600
400
200
O 100 200 300 400 500
Channel
Figure 11. Effect of annealing temperature on the WAXS equatorial intergrations
of the oriented non-crystalline PET Sample 1. The formation of the crystalline
(010), (10), and (100) reflections is seen
With orientation, this onset shifts toward lower temperatures, in agree-
ment with the observations in [14,20,21]. At low annealing temperatures,
the patterns show a predominant formation of the non-equatorial (hkl)
reflections, whereas the equatorial intensity distribution remains almost
unchanged (see [U]). With further increases in temperature, the crystal-
lization effects become more significant (Figures 10 and 11). At 8O
0
C, the
formation of the (12) reflection takes place. Starting at 9O
0
C, the three
equatorial (010), (10), and (100) reflections, which are characteristic of
unit cell lattice planes parallel to the c-axis (Figure 8), are formed. How-
ever, the crystalline PET structure is still distorted, as also reported by
Fischer and Fakirov [3O]. At 12O
0
C, a crystalline structure appears. With
the further increase in temperature, the molecular order improves.
Using the polymer beamline of HASYLAB at the DESY synchrotron
facility in Hamburg, Germany, the time dependence of the equatorial (010),
Structure Development and Mechanical Behavior... 303
(110), and (100) WAXS intensities was studied [U]. The experiments were
performed at constant sample length and different annealing temperatures
in the range from 50 to 24O
0
C in vacuum of about O.OTmbar. Heating was
performed stepwise from 2O
0
C at a rate of 34
0
CnLm"
1
. At each tempera-
ture, the intensity distribution along the equator was recorded for 2 h as a
set of 30s scans by employing a ID proportional counter of 512 channels.
The wavelength was close to that of CuK
a
. The data were corrected for the
detector sensitivity and background scattering. No smoothing procedures
were applied.
WAXS synchrotron studies on the oriented non-crystalline Samples 1
and 8 in the temperature range from 50 to 24O
0
C at constant sample
length revealed crystallization in terms of the formation of equatorial crys-
talline (010), (10), and (100) reflections at annealing temperatures above
8O
0
C. The crystallizations were almost completed within a heating period
of 7.5 min. After reaching the desired annealing temperature, virtually no
changes with time in the range from about 7.5 to 110 min are found in
the equatorial integrations, as shown in Figure 12 for the temperature of
12O
0
C. This finding suggests that oriented non-crystalline PET structures
12O
0
C
t (min)
Scattering angle
Figure 12. Synchrotron WAXS intensities along the equator perpendicular to the
draw direction of the oriented PET Sample 8 during isothermal crystallization
at 12O
0
C
304 U. Gschel
crystallize rapidly even at low temperatures such as 12O
0
C, which is in
agreement with the data reported in [13-15,18,19].
DSC measurements on the oriented non-crystalline PET Samples 1 and
8, using a Mettler DSC 821e/700 and a heating rate of 1O
0
CnIm"
1
, show
a broad crystallization area from about 75 to 18O
0
C (Figure 2). It is note-
worthy that the onset of crystallization is in the vicinity of the glass tran-
sition. The local alignment of the chain segments, caused by orientation,
plays the role of crystallization nuclei and, consequently, enhances crys-
tallization as revealed by the shift of the onset of crystallization to lower
temperatures. The phenomena of glass transition and crystallization can
hardly be separated in the thermogram. Furthermore, the attempts to per-
form temperature-modulated DSC studies failed. Because of crystallization,
the normal procedure of T
9
determination could not be applied. Therefore,
only the onset of glass transition was taken from the intercept of the low-
temperature tangents. Compared to 78
0
C for the initially isotropic Sam-
ple O, the onset of glass transition of the oriented samples is significantly
lower (66
0
C for Sample 8 and 69
0
C for Sample 1, see Table 3). These results
are reproducible and confirm our finding in [U]. However, its explanation
is still under discussion. Probably, differences in the cooling rates in the
preparation of samples of different solid state organization, faster sub-T^
aging at higher specific free volume content, effects of moisture in these
nanoscopic pseudo-ordered structures, etc., have to be considered [31].
Table 3. DSC data of non-crystalline PET
Sample
O
1
8
Birefringence
< 0.005
0.177
0.200
Glass transition
onset
(
0
C)
78
69
66
Crystallization
enthalpy
(Jg-
1
)
37
28
31
Melting
enthalpy
(Jg-
1
)
41
52
52
Shrinkage strain. The shrinkage strain, 5, was measured with an In-
stron apparatus in the temperature range from 25 to 12O
0
C [U]. During
the experiment, a small constant stress ( = F/A) of 1.8MPa was applied
with an accuracy of the force AF = 0.002 N at constant cross section, A.
A long sample (/Q = 500 mm) was used in order to ensure a high precision
of the measurement of the shrinkage strain. From /Q and I i (the initial and
the actual length, respectively), the shrinkage strain was calculated as
S(%) = 100ft - Z
0
)Ao- (1)
Shrinkage strain experiments on the oriented non-crystalline Sample 1
showed no shrinkage in the temperature range from 25 to 5O
0
C (Figure 13).
Structure Development and Mechanical Behavior... 305
.a
^
CO
b
- *--
I I
O 40
C
'"
s
s
/
/
/
/
/
/
/
s
- - tf
I I I I I I
60 80 100
Temperature (
0
C)
Figure 13. Effect of temperature on the shrinkage strain, , of the oriented non-
crystalline Sample 1
With the further rise of temperature, 5 increases abruptly. This is because
of a temperature-induced increase in the molecular mobility, which en-
ables conformational changes to occur. The shrinkage strain reaches 5.5
and 7.7%at 100 and 12O
0
C, respectively. Comparable shrinkage behavior
is found for the oriented Sample 8. As seen in Figure 13, the shrinkage
strain-temperature curve reveals an inflection point at about 7O
0
C, which
relates to an onset of crystallization, confirmed by DSC and WAXS ex-
periments. During cooling of Sample 1 from 100 down to 3O
0
C (Figure 13,
arrow a) and subsequent heating to 10O
0
C (Figure 13, arrow 6), the shrink-
age strain remains constant. Such behavior is attributed to freezing of the
shrinkage, initiated by a decrease in the chain mobility. Bonart et al [32]
already reported a thermally stimulated reversible change in the shrinkage
strain up to the previous annealing temperature for cold drawn ( = 1.6)
polycarbonate.
4.2. Structure after hot (post)drawing
Drawing can affect the chain architecture on different structural levels rang-
ing from the molecular to the macroscopic. The microstructural level in
terms of crystallites, lamellae, and fibrils is closely related to crystalliza-
tion and thermomechanical properties, and for this reason it is the main
focus of the present chapter. The macrostructural level describes phenom-
ena such as crazing, necking, and fibrillation and will be discussed only
briefly.
306 U. Gschel
4.2.1. Crystallite orientation
By applying the sample coordinate system (#1, X
2 5
^s)
5
the crystallite ori-
entation can be described by the angles < p h k i and h k i , which correspond
to the angles between x % and p h k i as well as between and the projection
of phki in the x^- x^ plane, respectively. X
3
denotes the draw direction and
Phki is the normal to the chosen (hkl) lattice planes, specified by the Miller
indices. The average orientation function (cos
n
] is determined by Eq. (2)
using the scattering intensities ( )
I

70
I ( ) cos
71
sin
(2)
/ ( ) sin
Jo
where
()= (,). (3)
Jo
In the case of deviation from the fiber symmetry, which is expressed by
a /3-dependence, the scattering intensities /(0, ) must be first integrated
over according to Eq. (3) and then over using Eq. (2) (see Alexander
[33]).
The orientation distribution can be represented in terms of spherical
harmonic functions (Legendre polynomials P
n
, see Stein et al [34] and
Windle [35]) as
( ) = ( + l/2)(P (cos^))P (cos0). (4)
n=0
For symmetry reasons, only polynomials with even order are relevant.
The second and fourth Legendre polynomials are
<P
2
(cos0)) = (3(cos
2
0)-l) (5)
and
(6) (P
4
(cos 0)) = ^(35{cos
4
) - 30{cos
2
) + 3).
8
For three different orientations (parallel, random, and perpendicular)
with respect to the draw direction, x% , the corresponding values for (cos
2
) ,
(P
2
(COSc/))), and (P
4
(cos )) are given in Table 4.
The crystallite orientation of PET in terms of the angle between the
c-axis of the unit cell and the draw direction can be calculated from the
Structure Development and Mechanical Behavior... 307
Table 4. Orientation parameters
Orientation
Parallel to draw direction
Random
Perpendicular to draw direction
(cos
2
)
1
1/3
O
(P
2
(cos 0))
1
O
-0.5
(P
4
(cos 0))
1
O
0.375
experimental average orientation function (cos
2
4 > hki) of the three (hk(f)
reflections using Wilchinsky's method [36,37].
(cos
2
) = 1 - A(cos
2
0oio) - B(cos
2

) - <7(cos
2

00
) (7)
where the parameters A = 0.8786, B = 0.7733, and C = 0.3481 are deter-
mined by the triclinic crystal system.
2D WAXS experiments were performed by means of a Rigaku RU-
300 together with a Nicolet area detector [3]. The horizontal X-ray beam
( = 0.154nm, point focus) was placed perpendicular to the x^- x^ film
plane of the sample. The anisotropy was ascertained by measuring the
scattering intensity, /(20), by changing the rotation angle in steps of 10.
Standard procedures were used for correction of the detector sensitivity,
the primary beam profile, and the background. The scattering intensities
were separated from the overlapping and neighboring intensities. The angle
was taken from the 2D diffraction patterns as the azimuthal angle.
The results reveal near perfect alignment of the c-axis with respect to
the draw direction, expressed by an averaged angle from 2.9 to 7.7 with
only a minor dependence on the ultimate drawing conditions applied (Ta-
ble 5). According to the orientation parameters (P2(cos</>)} and (P^(COS ))
for the normals of the (010), (10), and (100) lattice planes, the orientation
distribution is very narrow. Very large drawing stresses, suddenly applied
to an oriented non-crystalline structure, can also result in a perfect crys-
tallite orientation, comparable to those reached by multistep drawing. Hot
drawing of an oriented non-crystalline PET structure leads to a higher final
crystallite orientation than that achievable by strain-induced crystallization
performed at the glass transition.
Table 5. Crystallite orientation
Sample Orientation parameters
2
3
4
5
6
7
9
0.9917
0.9942
0.9974
0.9964
0.9819
0.9925
0.9957
5.2
4.4
2.9
3.4
7.7
5.0
3.8
308 U. Gschel
4.2.2. Lamellar structure
2D SAXS studies using the Rigaku RV-300 together with a Nicolet detector
as described above showed four-point patterns for Samples 2, 5, and 9 [3],
see Figure 14. These patterns represent a lamellar structure where the
lamellae normal is inclined with respect to the draw direction [38,39].
In the case of further and ultimate drawing, the inclined lamellae turn
their orientation, the normal becoming parallel to the draw direction, with-
out any change in the crystallite orientation, as schematically shown in
6
600
500
400
300
200
100
O
-100
-100 O 100 200 300 400 500 600
Channel
500
600
400
Si
200
O 100 200 300 400 500
Channel
Figure 14. SAXS pattern of Sample 5
Structure Development and Mechanical Behavior... 309
Figure 15 by means of two- and four-point patterns. Stockfleth et al [38]
described a change from four-point into two-point patterns during PET
deformation as a consequence of different deformation modes. These au-
thors suggested that for inclined lamellar stacks and temperatures above
the glass transition, lamellar slips play an essential role.
It seems that the four-point patterns represent the characteristic struc-
tural feature of drawn PET. From the distance between the meridional
layer line and the origin of the pattern, the lamellar periodicity along the
draw direction can be determined in terms of the long period as
L = 2/&
3
,

(8)
Real space Reciprocal space
(a)
Figure 15. Lamellar structure of drawn PET; and are the angles between the
normal to the lamellae and the fibril axis, respectively, and the draw direction, #3.
The inclination angle = O results in a two-point pattern (a). An angle 0/0
gives a four-point pattern (b), whereas an additional describes inclined
fibrils composed of inclined lamellae (c)
310 U. Gschel
Taking into account the inclination of the lamellae defined by the angle
, the lamellar periodicity, L', can be expressed as
L' = 27/&'3
|
with
=&3,max/COS0.
The above SAXS parameters are given in Table 6
Table 6. Parameters determined from the 2D SAXS patterns
(9)
(10)
Sample
2
3
4
5
9
Half-width,
layer line intensity,
b* (A-
1
)
0.185
0.124
0.129
0.202
0.202
Inclination
angle,
(deg)
45.4
O
O
52.0
47.8
Long
period,
L (nm)
15.0
16.6
16.0
15.4
14.2
Lamellar
periodicity,
L' (nm)
10.6
16.6
16.0
9.3
9.5
The differences in L were confirmed by ID SAXS studies using a Kratky
compact camera (Anton-Paar K.G., Graz) [3]. Compared to the long peri-
ods for isothermally crystallized PET, which were found to be from 9.5 to
(a)
Figure 16. Structural model of oriented semicrystalline PET, related to (a) the
two-point SAXS patterns ( , = O) and (b) the four-point SAXS patterns
( = O, = O)
Structure Development and Mechanical Behavior... 311
IT.lnm by Groeninckx et al [40] and from 7 to l l nm by Cruz et al [41],
the L values of the hot drawn Samples 2-5 and 9 are higher (from 14.2 to
16.6nm) and the inclination angles vary from O to 52 (Table 6). Obviously,
the lamellar structure revealed by SAXS is strongly affected by the draw-
ing conditions, whereas the crystallite orientation is already nearly perfect
after the first hot drawing and remains almost unchanged during further
drawing, as confirmed by WAXS results (see Section 4.2.1). Figures 16a and
16b represent the structural feature behind a four-point and a two-point
SAXS pattern, respectively. The 3D chain architecture between the crys-
tallites determines to a great extent the thermomechanical properties. The
two structures shown in Figures 16a and 16b strongly differ in the spatial
arrangement of the chain molecules in the amorphous phase [1O].
5. Mechanical properties of drawn PET
5.1. Stiffness
The elastic modulus, E, was obtained at a low strain rate de/dt =1%min"
1
and T = 23
0
C in the linear viscoelastic region, <0.4%, using an Instron
tensile tester [1,2]. Figure 17 shows that the elastic moduli in the range
from 8.4 to 16.8GPa (Table 7) can be increased with each drawing step,
yielding an increase in the draw ratio, . For comparison, the isotropic
non-crystalline initial PET Sample O showed an elastic modulus of about
2GPa[ O] .
Table 7. Mechanical data
Sample
O
1
2
3
4
5
6
7
8
9
Draw
ratio,

4.3
6.3
7.1
7.3
7.2
5.6
6.2
4.2
5.4
Modulus,
E (GPa)
2.0
8.4
13.4
15.0
16.8
14.3
7.1
10.2
10.2
12.0
Stress at
break,
a
br
(MPa)
650
625
800
825
775
970
570
525
640
800
Birefringence,

< 0.001
0.177
0.202
0.228
0.205
0.249
*
*
0.200
0.219
Orientation
factor,
/a
0.644
0.771
0.846
0.727
0.953
*
*
0.727
0.804
* None, opaque sample
Infrared spectroscopy under loading [42] has shown that only a minor
fraction of less than 10%of the chain molecules between crystallites in the
same fibril are taut tie molecules (t.t.m.) in a trans conformation, carrying
the load. These t.t.m. molecules determine the elastic modulus. Assuming
312 U. Gschel
18
16-
c
dn
O
14-
I 12 -
10-
o9 X
8 **
/
4.0 4.5 5.0 5.5 6.0
Draw ratio
6.5 7.0 7.5
Figure 17. Elastic modulus vs. draw ratio
1000-
900-
800-
700-
600-
500-
4.0 4.5 5.0 5.5 6.0
Draw ratio
6.5 7.0 7.5
Figure 18. Stress at break vs. draw ratio
Structure Development and Mechanical Behavior... 313
two-phase behavior, the amorphous orientation factor, /
a
, can be calculated
according to Samuels [43]
fa =
An - %An
c
/
e

(1-)
(U)
where is the experimental birefringence, An
c
= 0.220 is the intrinsic
birefringence of a perfectly oriented crystalline phase,

= 0.275 is the
intrinsic birefringence of a perfectly oriented amorphous phase [10], and
is the experimentally established degree of crystallinity.
5.2. Fracture behavior
The effect of the draw ratio on the stress at break is shown in Figure 18
[1,2]. In contrast to the elastic modulus, further drawing of Sample 2 in
order to obtain Samples 3 and 4 does not increase the stress at break. The
same holds for the transition from Sample 6 to Sample 7. For the sake of
comparison, the initially Isotropie non-crystalline PET Sample O reveals a
stress at break (related to the actual cross section) of about 650 MPa at a
strain of about 400%. Consequently, the stress at break has improved only
moderately under the drawing conditions applied. Under ultimate condi-
tions, a decrease in the stress at break can occur which is related to fracture
of chains and fibrils. In addition, fibrillation can take place, as revealed by
scanning electron microscopic studies (SEM, Figure 19).
Figure 19. SEM image of Sample 7
314 U. Gschel
6. Conclusions and outlook
By applying a multistep uniaxial drawing procedure to an initially isotropic
non-crystalline PET structure, the changes in morphology and resulting
thermomechanical behavior were studied in detail. Mechanical experiments
reveal a strong increase in the elastic modulus from 2 GPa for the initial
PET to 17GPa at a final draw ratio of 7.3, whereas the stress at break
decreases under the ultimate drawing conditions. The modulus increase is
related to a perfect alignment of the crystallites combined with the for-
mation of taut tie molecules. The changes in the crystallite orientation
with drawing are only minor, which cannot explain the changes in stiffness
among the oriented and semicrystalline structures by a factor of about
1.8. SAXS studies show significant differences in terms of the orientation
of lamellae. The spatial arrangement of chain molecules in the amorphous
phase is very different and can serve to explain the great differences in the
thermomechanical behavior.
Acknowledgements
The author would like to thank U. Mller at the Institute for Polymer
Testing and Polymer Science (IKP), University of Stuttgart, for performing
the SEM experiments.
References
1. Gschel U (1989) Preparation of thin strips of poly(ethylene terephthalate)
with high rigidity and strength by means of orientational drawing, Acta PoIy-
rneri ca 40:23-31 (in German).
2. Gschel U and Cools PJCH (2000) Molar mass of polyethylene terephtha-
late) (PET) during ultimate uniaxial drawing, Polym Eng Sei 40:1596-1605.
3. Gschel U (1995) Two-dimensional small-angle X-ray scattering studies on
oriented poly(ethylene terephthalate) films, Polymer 36:1157-1165.
4. Zhu P and Ma D (1999) Study on the double cold crystallization peaks of
poly(ethylene terephthalate) (PET): 2. Samples isothermally crystallized at
high temperature, Eur Polym J 35:739-742.
5. Gschel U (1992) Thermo-mechanical studies of highly oriented
poly(ethylene terephthalate), Polymer 33:1881-1886.
6. Gschel U and Nitzsche K (1985) Effect of loading on deformation behavior of
amorphous poly(ethylene terephthalate) films, Acta Polymeri ca 36:580-583
(in German).
7. Weigel P, Hirte R, Melior J P, Schulz E, Zenke D, Gschel U, Lange K and
Gei D (1989) Structure formation in polymers with flexible chains for obtain-
ing special mechanical properties, Acta Polymeri ca 40:197-206 (in German).
8. Hofmann D, Gschel U, Walenta E, Gei D and Philipp B (1989) Investiga-
tions on supermolecular structure of poly(ethylene terephthalate) samples of
higher modulus, Polymer 30:242-247.
Structure Development and Mechanical Behavior... 315
9. Gschel U, Deutscher K and Abetz V (1996) Wide-angle X-ray scattering
studies using an area detector: crystallite orientation in semicrystalline PET
structures, Polymer 37:1-6.
10. Dumbleton J H (1968) Influence of crystallinity and orientation on sonic
velocity and birefringence in poly(ethylene terephthalate) fibers, J Polym
Sei A2 6:795-800.
11. Gschel U (1996) Thermally stimulated structural changes in highly oriented
glassy poly(ethylene terephthalate), Polymer 37:4049-4059.
12. Biangardi H J and Zachmann H G (1977) Influence of the arrangement of
the crystals and the structure of the noncrystalline regions on the mechanical
properties of polyethylene terephthalate, J Polym Sei Polym Symp 58:169-
183.
13. Le Bourvellec G, Monnerie L and Jarry J P (1987) Kinetics of induced crys-
tallization during stretching and annealing of poly(ethylene terephthalate)
films, Polymer 28:1712-1716.
14. Alten G and Zachmann H G (1979) Effect of molecular orientation on
the crystallization kinetics of poly(ethylene terephthalate), Mak romol Ch em
180:2723-2733 (in German).
15. Sawatari C and Matsuo M (1985) Dependence of thermal crystallization
of poly(ethylene terephthalate) on active mobility of amorphous chain seg-
ments, Tex ti le Res J 55:547-555.
16. Quian R, He J and Shen D (1987) Crystallization of poly(ethylene tereph-
thalate) from oriented amouphous film, Polym J 19:461-466.
17. Desai P and Abhiraman A S (1985) Crystallization in oriented poly(ethylene
terephthalate) fibers. I. Fundamental aspects, J Polym Sei Polym Ph ys Ed
23:653-674.
18. Thompson A B (1959) Strain-induced crystallization in polyethylene tereph-
thalate, J Polym Sei 34:741-760.
19. Smith F and Steward R D (1974) The crystallization of oriented
polyethylene terephthalate), Polymer 15:283-286.
20. Galli R, Canetti M, Sadocco P, Seves A and Vicini L (1983) Preoriented
poly(ethylene terephthalate) yarns: influence on the gauche-trans transfor-
mation on crystallization, J Polym Sei Polym Ph ys Ed 21:717-723.
21. Nobbs J H, Bower D I and Ward I M (1976) A study of molecular orientation
in drawn and shrunk poly(ethylene terephthalate) by means of birefringence,
polarized fluorescence and X-ray diffraction measurements, Polymer 17:25-
36.
22. Trznadel M (1986) Thermally stimulated shrinkage forces in oriented poly-
mers: induction time, Polymer 27:871-876.
23. Trznadel M and Kryszewski M (1988) Shrinkage and related relaxation of
internal stresses in oriented glassy polymers, Polymer 29:418-425.
24. Pakula T and Trznadel M (1985) Thermally stimulated shrinkage forces in
oriented polymers: 1. Temperature dependence, Polymer 26:1011-1018.
25. Schick C, Dobbertin J, Ptter M, Dehne H, Hensel A, Wurm A, Ghoneim
A M and Weyer S (1997) Separation of components of different molecular
mobility by calorimetry, dynamic and dielectric spectroscopy, J Th erm Anal
49:499-511.
316 U. Gschel
26. Dobbertin J, Hensel A and Schick C (1996) Dielectric spectroscopy and
calorimetry in the glass transition region of semicrystalline poly(ethylene
terephthalate), J Th erm Anal 47:1027-1040.
27. Daubeny R, Bunn C W and Brown C J (1954) The crystal structure of
polyethylene terephthalate, PTO C Roy Soc (London) A226:531-542.
28. Ledbetter M, Cuculo J and Tucker P (1984) Structure and properties of
poly(ethylene terephthalate) crystallized by converging flow and high pres-
sure, J Polym Sd Polym Ch em Ed 22:1435-1459.
29. Dumbleton J H and Bowles B B (1966) X-ray determination of crystallinity
and orientation in poly(ethylene terephthalate), J Polym Sei A2 4:951-958.
30. Fischer E W and Fakirov S (1976) Structure and properties of
polyethyleneterephthalate crystallized by annealing in the highly oriented
state, Part 1 Morphological structure as revealed by small-angle X-ray scat-
tering, J Mater Sd 11:1041-1065.
31. Alfonso G C, private communication.
32. Bonart R, Fenzl M and Schmidt K (1987) Frozen-in deformations, 1. Defor-
mation systems, Acta Polymeri ca 38:281-286 (in German).
33. Alexander L E (1969) X-ray Diffraction Methods in Polymer Science, Wiley
Interscience, New York.
34. Stein R S and Wilkes G L (1975) Physico-chemical approaches to the mea-
surement of anisotropy, in Structure and Prop erti es of O ri ented Polymers
(Ed. Ward I M) Applied Science Publishers, London, Ch. 3, pp. 57-145.
35. Windle A H (1982) Measurement of molecular orientation and structure in
non-crystalline polymers by wide angle X-ray diffraction, in Develop ments i n
O ri ented Polymers-1 (Ed. Ward I M), Applied Science Publishers, London,
Ch. 1, pp. 1-46.
36. Wilchinsky Z W (1959) On crystal orientation in polycrystalline materials,
J Ap p l Ph ys 30:792.
37. Wilchinsky Z W (1963) Recent developments in the measurement of orienta-
tion in polymers by X-ray diffraction, in Advances i n X-ray Analysi s, Plenum
Press, New York, Vol. 6, pp. 231-241.
38. Stockfleth J, Salamon L and Hinrichsen G (1993) On the deformation mecha-
nisms of oriented PET and PP films under load, Colloi d Polym Sd 271:423-
435.
39. Paredes E and Fischer E W (1989) Small angle X-ray studies of the structure
of crazes in polycarbonate and poly (methyl methacrylate), Mak romol Ch em
180:2702-2722 (in German).
40. Groeninckx G, Reynaers H, Berghmans H and Smets G (1980) Morphol-
ogy and melting behavior of semicrystalline poly(ethylene terephthalate). I.
Isothermally crystallized PET, J Polym Sd Polym Ph ys Ed 18:1311-1324.
41. Cruz C S, Stribeck N, Zachmann H G and Balt Calleja F J (1991) Novel
aspects in the structure of poly(ethylene terephthalate) as revealed by means
of small-angle X-ray scattering, Macromolecules 24:5980-5990.
42. Gschel U, unpublished results.
43. Samuels R J (1965) Morphology of deformed polypropylene. Quantitative
relations by combined X-ray, optical, and sonic methods, J Polym Sd A
3:1741-1763.

S-ar putea să vă placă și