Professional Documents
Culture Documents
Zhaozeng Tao,1 Yatao Wang,1,2 Jianhua Li,2 Xiaodong Wang,1 Dezhen Wu1
1
State Key Laboratory of Organic-Inorganic Composites, Beijing University of Chemical Technology, Beijing
100029, China
2
Hebei Provincial Engineering Technology Research Center of Coal-based Materials and Chemicals,
Tangshan, Hebei 063018, China
INTRODUCTION
Short glass fiber reinforced thermoplastic composites
as one of the most important organicinorganic compo-
POLYMER COMPOSITES2014
FIG. 1. Scheme of thermoplastic pultrusion for long glass fiber reinforced POM composites. [Color figure
can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
EXPERIMENTAL
Materials
The POM resin (commercial grade: BS270) used in
this work was commercially obtained from Shanghai
Bluestar New Chemical Materials. This product is a lowviscosity typed POM copolymer with a number average
molecular weight of 22,500 g/mol and a melt flow index
of 27.0 g/10 min. Direct-drawn glass fiber roving was
kindly supplied by Chongqing Polycamp International,
China. This product has a linear density of 2,400 tex and
a filament diameter of 14 lm, and the fibers have been
surface-treated with 1,2-ethylenebis(trimethoxysilane)
coupling agent. Tetrakis[methylene-b-(3,5,-di-tert-butyl-4hydroxyphenyl) propionate]methane and tris-(2,4-di-tertbutylphenyl)phosphite as antioxidant and thermal stabilizer were commercially obtained from Beijing Additive
Institute, China.
Processing of LFT Composites
The pultrusion equipment used in this work was
custom-designed and consisted of a fiber creel, a preheating chamber, an impregnation apparatus connecting with
a screw extruder, a wind cooler, a pulling machine, and a
custom tailored pelletizer as illustrated in Fig. 1. The
glass fiber roving was pulled by the pulling machine
through a preheating chamber and then pulled into the
impregnation apparatus over a number of rollers, and an
extruder provides a molten POM into the chamber of the
impregnation apparatus. The temperature in the chamber
of the impregnation machine was set to 175180 C, and
the pulling speed of the rovings was limited in the range
of 1015 m/s. Impregnation occurs during contact with
the rollers, where the fiber roving is more or less spread
out and open up to receive the molten POM. Another
important effect is that pulling the fiber rovings over the
roll, in the presence of the viscous molten POM, causes
DOI 10.1002/pc
local high shear rates between the bundle and the bar,
resulting in high shear forces. These forces oppose the
moving direction of the rovings and are perceived as a
viscous drag on the rovings. The perceived total viscous
drag increases with increasing bundle speed and with an
increasing number of rollers and of course equalizes the
pulling force on the fiber rovings. After a wind cooling,
the pultruded braces were cut into pellets at a length of
1012 mm by a custom-tailored pelletizer. The resulted
pellets were dried under vacuum at 80 C overnight and
then were stored in the sealed aluminum foil bags.
Characterization
Measurement of Mechanical Performance. All of the
pelletized samples were dried at 80 C in a vacuum oven
for 8 h prior to injection molding, and then were
injection-molded into the test bars with the different
shapes required for mechanical and heat-resistant measurements. Notched Izod impact strength was measured on
a SANS ZBC-1400A impact tester equipped with a pendulum of 2.75 J according to ISO 180 standard. The
impact test bars were notched on a milling machine to
achieve a V type notch in a depth of 2 6 0.2 mm. The
tensile and flexible properties were measured with a
SANS CMT-4104 universal testing instrument using a 10
kN load transducer according to ISO 527 and ISO 178
standards, respectively. The dumbbell tensile specimens
were fabricated with a dimension of 150 mm 3 10 mm
3 4 mm, and the tensile speed was set to 5 mm/min. All
the measurements were performed at a constant temperature of 23 C, and the reported values reflected an average
from five tests.
Polarized Optical Microscopy. Polarized optical
microscopy was performed to observe the distribution of
glass fibers and the crystalline morphology of LFT POMbased composites on an Olympus BX51 polarizing microscope equipped with a Linkam THMS 600 temperature
POLYMER COMPOSITES2014 3
t
dHc
dt
dt
0
Xt 5 1
3100%
dHc
dt
dt
0
(1)
where Hc is the crystallization enthalpy during the infinitesimal time interval dt.
Thermogravimetric Analysis. Thermogravimetric analysis (TGA) was performed in a nitrogen atmosphere using
a TA Instruments Q50 thermal gravimetric analyzer. Samples were placed in a platinum crucible and ramped from
room temperature to 750 C at a heating rate of 10 C/min
while a flow of nitrogen was maintained at 50 mL/min.
Mechanical Performance
Figure 2 shows the mechanical properties of the
injection-molded specimens derived from POM-based LFT
composites, and the experimental data from short fiber reinforced POM composites are also presented in the figure as
a reference. It appears that these mechanical properties
achieved a significant improvement as a result of long glass
fiber reinforcement on POM. The tensile strength is shown
to increase rapidly with increasing the glass fiber content
and obtains an increment of 172% compared to pure POM
when the fiber content is increased to 40 wt%. Analog
trends are observed in the flexural strength and flexural
modulus, which are also found to increase by 325% and
266%, respectively, for the samples containing 40 wt% of
glass fibers. Although the fiber content of long glass fiber
reinforced thermoplastic composites could be increased over
60 wt% through the melt pultrusion process, the specimens
for mechanical tests must be injection-molded by the
injection-molding machine specially designed. In this case,
no further mechanical data were provided. However, it was
reported that the higher long fiber content could cause a further improvement in mechanical performance of the long
glass fiber reinforced thermoplastic composites [19], and the
POM-based LFT composites may follow this trend. It is
also notable from Fig. 2a and b that the POM/short glass
fiber composites present much lower tensile and flexural
properties in comparison with the POM/long glass fiber
ones, and the tensile strength of the composites containing
30 wt% of short glass fiber only increases by 72% compared to pure POM. This indicates the reinforcing effect of
short fiber on POM is inferior to that of long glass fiber. It
seems that the POM composites almost could not be processed through melt extrusion with addition of 40 wt% short
glass fiber because of the highly thermal sensitivity of POM
resin, which may result in a thermal degradation during the
melt process of its composites with high fiber content.
Therefore, the mechanical data were unavailable for the
DOI 10.1002/pc
FIG. 2. Mechanical parameters of POM-based composites as a function of fiber content: (a) tensile
strength, (b) flexural strength, (c) flexural modulus, and (d) notched Izod impact strength. [Color figure can
be viewed in the online issue, which is available at wileyonlinelibrary.com.]
that the residual fiber length of the long glass fiber reinforced
POM composites is much longer than that of the short fiber
ones. The residual fiber length was also evaluated by observing on optical microscope and then was calculated statistically
by the Nano-Measurer software. Figure 4 shows the fiber
length distributions of these two injection-molded bars with
the corresponding optical images as insets. It should be noted
that the residual fiber lengths in the POM-based LFT composite are still longer than those in the short fiber reinforced one
after injection molding process. As observed from the long
glass fiber reinforced POM specimen, the most glass fibers
show the lengths ranging from 2.5 to 6.5 mm, which seems to
be reduced from 10 to 12 mm of the feedstock during the
injection molding on account of the strong shear effect of
injection screw at the plasticization stage of the composites.
As a reference, the short glass fiber reinforced POM specimen
injection molded with the same processing parameters only
shows a fiber length mostly distributed in the range of 0.1
0.4 mm. The number average fiber length could be calculated
POLYMER COMPOSITES2014 5
FIG. 3. Digital optical photographs of the features of the original pellets, injection-molded bars, and ashing specimens of (a) POM-based
LFT composite and (b) short glass fiber reinforced POM composite.
[Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
X sli Vi X
lc
1rum 12Vf
1
rf Vj 12
df l >l
2lj
l <l
i
(2)
rf df
2s
(3)
FIG. 4. Fiber length distribution curves of (a) POM-based LFT composite and (b) short glass fiber reinforced POM composite at the same
fiber contents of 30 wt%. Inset: optical micrographs of residual fibers.
[Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
DOI 10.1002/pc
FIG. 5. SEM micrographs of tensile fracture surfaces for POM-based LFT composites with (a, b) 10 wt%,
(c, d) 20 wt%, and (e, f) 40 wt% of long glass fiber and (g, h) for the POM composites containing 30 wt%
of short glass fiber.
pared to pure POM. Such an evident toughness enhancement with respect to the fiber length and content may be
attributed to the fracture-transformation mechanism. It
POLYMER COMPOSITES2014 7
FIG. 6. SEM micrographs of impact fracture surfaces for POM-based LFT composites with (a, b) 20 wt%
and (c, d) 30 wt% of glass fiber.
2
Vf lUd
Vf l sf
1
;
df
6df
when
l < lc
(4)
2 3
Vf l2lc Uf
Vf lc Ud
Vf lc sf
Uc 5Um 12Vf 1
1
1
;
l
df l
6df l
when l > lc
(5)
where Um, Uf, and Ud are the fracture energy for matrix,
fibers, and interfaces, respectively, and sf is the interfacial
friction during fiber pull-out. For the composite systems
with a fiber length lower or close to the critical fiber
length like short fiber reinforced thermoplastic composites, the impact energy dissipation is estimated by Eq. 4,
where the three terms cover matrix fracture, fiber-matrix
debonding, and fiber pull-out. The fiber pull-out is the
8 POLYMER COMPOSITES2014
FIG. 7. Development of the relative degree of crystallinity as a function of crystallization time for (a) pure
POM and its LFT composites with (b) 10 wt%, (c) 20 wt%, and (d) 40 wt% of glass fiber in the isothermal
process. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
FIG. 8. Avrami plots of log[ln(1 Xt)] vs. log t for the isothermal crystallization of (a) pure POM and its
LFT composites with (b) 10 wt%, (c) 20 wt%, and (d) 40 wt% of glass fiber in the isothermal process.
[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Samples
Pure POM
POM 1 10 wt%
glass fiber
POM 1 20 wt%
glass fiber
POM 1 30 wt%
glass fiber
POM 1 40 wt%
glass fiber
Tc
( C)
k
(1/minn)
t0.5
(min)
1/t0.5
(1/min)
134
138
142
144
146
134
138
142
144
146
134
138
142
144
146
134
138
142
144
146
134
138
142
144
146
2.02
2.04
1.74
1.48
1.73
1.76
1.74
2.02
1.93
1.78
1.77
1.66
1.51
1.49
1.73
2.19
2.05
2.13
1.44
1.68
1.78
1.64
1.43
1.41
1.34
8.90
6.76
3.43
1.54
0.31
8.13
5.49
4.78
2.23
0.47
7.13
4.71
3.14
1.81
0.42
10.28
5.37
3.00
1.43
0.63
5.60
3.14
3.12
1.85
0.47
0.28
0.3
0.37
0.48
0.56
0.25
0.29
0.32
0.41
0.43
0.27
0.29
0.37
0.35
0.49
0.26
0.25
0.26
0.39
0.63
0.33
0.31
0.32
0.33
0.46
3.57
3.33
2.70
2.08
1.79
4.19
3.45
3.125
2.44
2.33
3.70
3.45
2.70
2.86
2.04
3.85
4.05
3.85
2.56
1.59
3.03
3.23
3.13
3.03
2.17
development of relative degree of crystallinity and crystallization time as expressed by [40, 41]:
12Xt 5exp 2ktn
(6)
where Xt is the relative degree of crystallinity at the crystallization time t, n the Avrami exponent depending on
the nature of nucleation and the growth geometry of the
crystals, and k the crystallization kinetic constant involving both nucleation and growth rate parameters. Figure 8
shows the typical Avrami plots of pure POM and its LFT
composites. These plots do not show a well linear relationship in the whole range. However, they are found to
show a fairly good linearity at the primary and middle
crystallization stage but only to present a deviation at the
final crystallization stage, which may be ascribed to the
secondary crystallization at that stage. Therefore, the
Avrami model seemed to successfully fit the crystallization data of POM domain for the relative degree of crystallinity value up to 90 %. In this case, the Avrami
parameters n and k were obtained from the linear regression of the Avrami plots in the relative crystallinity range
of 590%, because the determination coefficients of linear
fits in this regime were greater than 0.92.
The isothermal crystallization kinetic parameters of
pure POM and its LFT composites were calculated from
their Avrami plots and summarized in Table 1. It is
known that the value of n strongly depends on both the
mechanism of the nucleation and the crystalline growth,
DOI 10.1002/pc
ln 2
t0:5 5
k
1=n
(7)
FIG. 9. Polarized optical micrographs of the crystalline morphology for (a) pure POM and its composite
containing (b) 20 wt% and (c, d) 30 wt% of glass fiber. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
POM domain could be accelerated due to the heterogeneous nucleating effect of glass fiber. However, it is also
observed that the values of t0.5 are almost not reduced
with the increase of fiber content, indicating that the
space hindrance from long fibers may prevent the crystalline growth and thus extends the crystallization time. This
factor evidently dominated the isothermal crystallization
rate for the LFT composite with 40 wt% of glass fiber,
resulting in an increase of t0.5 accordingly. In addition, it
is noteworthy that, for both pure POM and its LFT composites, the lower the crystallization temperature, the less
the t0.5 value. Such a trend indicates that the overall isothermal crystallization rate decreases with increasing the
crystallization temperature due to the low supercooling in
the range of given temperatures. Despite the heterogeneous nucleation of glass fiber, it is more difficult to form
the crystal nuclei at higher temperatures, especially as
close to the melting point, due to the high mobility of
polymeric chains. Furthermore, at higher temperatures,
only a small part of fibers could act as nuclei for crystallization, but the rest of them generated a retardation effect
on the movement of polymeric chains. Consequently, the
induction time became longer, and the crystallization process was postponed accordingly.
12 POLYMER COMPOSITES2014
Spherulitic Morphology
POM is well known for its high crystallinity and can
usually form well-defined spherulites when crystallizing
from the melt. These spherulites were reported to consist
of folded lamellar crystals and extended chain crystals
[46]. Figure 9 shows the crystalline morphologies of pure
POM and its LFT composites observed under a polarized
optical microscope. The polarized optical micrograph
confirms the reported result for pure POM as shown by
Fig. 9a, where the spherulitic structures based on parallelly packed-lamellar crystals with distinct boundary
could be clearly distinguished. The formation of these
large and partially interconnected spherulites is ascribed
to few of nucleation centers among the POM molecules.
Differing from the short glass fiber reinforced composites,
the residual fibers in the LFT composites have much
greater aspect ratios (ratio of fiber length to diameter).
This may reduce the heterogeneous nucleation sites but
makes it possible to induce the transcrystallinity of POM
domain. Figure 9bd shows the morphology of the POMbased LFT composites formed during isothermal crystallization, where a narrow zone of transcrystallinity along
the fibers could be clearly distinguished in addition to
bulk nucleated spherulites. The formation of
DOI 10.1002/pc
FIG. 10. TGA thermograms of pure POM and its LFT composites.
Inset: Derivate TG curve of pure POM and its LFT composites. [Color
figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
transcrystallinity may be a heterogeneous process occurring on impurities collected on the long fiber surfaces.
Furthermore, the observed transcrystallinity appears at the
surfaces of long fibers without any interruption by the
bulk-nucleated spherulites, and the transcrystallization of
POM domain is also a process of growth with spatial constraints [47]. It is no doubt that the formation of transcrystallinity during the molding of POM-based LFT
composites facilitates the interfacial adhesion between the
matrix and fibers and thus enhances the reinforcement
effect.
Thermal Stability
The effect of long glass fiber on the thermal stability
of POM-based composites has been investigated by TGA,
and the obtained thermograms are shown in Fig. 10.
According to the weight-loss profiles, pure POM and its
LFT composites exhibited a typical one-step thermal degradation behavior under a nitrogen atmosphere. Such a
pyrolysis may be attributed to the chain scission of the
POM backbone as the prevailing decomposition reaction.
Accordingly, pure POM was found to encounter an initial
decomposition at 320.4 C to generate a weight loss of 3
wt%, and then, a rapid decomposition took place with a
maximum weight loss at 418.5 C due to the random scission of CAO bonds in the molecular chains. A thermally
induced chain rupture led to complete unzipping of the
damaged chain and thus almost left no residual char. On
the other hand, it seems that the incorporation of long
glass fiber did not notably affect the thermal degradation
behavior of POM domain. However, it is noteworthy that
the characteristic temperatures at the maximum weightloss rate were considerably improved with increasing the
fiber content as shown by derivate TG thermograms in
DOI 10.1002/pc
14 POLYMER COMPOSITES2014
DOI 10.1002/pc