Defence Technology 24 (2023) 269e284

Contents lists available at ScienceDirect

Defence Technology
journal homepage: www.keaipublishing.com/en/journals/defence-technology

Inherent relationship between process parameters, crystallization and

mechanical properties of continuous carbon fiber reinforced PEEK
composites
Xiao-long Ma a, Li-hua Wen a, Shi-yu Wang a, *, Jin-you Xiao a, Wen-hao Li a, Xiao Hou b
a
School of Astronautics, Northwestern Polytechnical University, Xi'an, 710072, China
b
China Aerospace Science and Technology Corporation, Beijing, 100048, China

a r t i c l e i n f o a b s t r a c t

Article history: High-performance thermoplastic composites have been developed as significant structural materials for
Received 9 January 2022 cutting-edge equipment in the aerospace and defence fields. However, the internal mechanism of pro-
Received in revised form cessing parameters on mechanical properties in the manufacturing process of thermoplastic composite
6 April 2022
structures is still a serious challenge. The purpose of this study is to investigate the process/crystalli-
Accepted 21 April 2022
Available online 27 April 2022
zation/property relationships for continuous carbon fiber (CF) reinforced polyether-ether-ketone (PEEK)
composites. The composite laminates are fabricated according to orthogonal experiments via the ther-
moforming method. The mechanical performance is investigated in terms of crystallization properties
Keywords:
Polymer matrix composites
and fracture morphology characterizations. Experimental results show that the mechanical performance
Thermoplastic and crystallization properties of thermoplastic composites are significantly affected by the coupling of
Processing parameters processing parameters. The increased molding temperature, pressure, and holding time improve the
Mechanical properties degree of fiber/matrix infiltration and affect the crystallinity and crystalline morphology of the matrix,
Crystallization which further influences the mechanical properties of the composites. This is reflected in the test results
that crystallinity has an approximately linear effect on mode-I interlaminar fracture toughness and
transverse flexural modulus. As well as the higher molding temperature can destroy the pre-existent
crystals to improve the toughness of the matrix, and the well-defined crystalline structures can be
observed when fabricated at higher temperatures and longer periods of holding time.
© 2022 China Ordnance Society. Publishing services by Elsevier B.V. on behalf of KeAi Communications
Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction structure of thermoplastic resins such as PEEK and PPS, the high
viscosity of the molten resin made it difficult to fully infiltrate the
In recent years, high-performance thermoplastic composites fibers [14e18]. And thermoplastic resin contains amorphous do-
have been successfully applied in the aircraft industries because mains and crystallites, which can convert to each other upon
they exhibit numerous advantages, such as excellent specific stiff- temperature change or shear flow [19]. Therefore, the molding
ness, high toughness and ductility, outstanding thermal stability process conditions can significantly affect the performance of the
(PEEK is exceeding 300 C), and shorter manufacturing cycles, etc thermoplastic composites. In addition, thermoplastic resins have
[1e10]. In particular, thermoplastic composites have excellent high melting temperatures and a narrow processing window. These
impact resistance, corrosion resistance, and high service tempera- harsh conditions have brought considerable challenges to the
ture, which have attracted the attention of military applications molding process of high-performance thermoplastic composites.
before commercial use and have been widely used in military Therefore, it is necessary to optimize the molding processing of
aircraft [11e13]. However, due to the long molecular chain thermoplastic composites.
Different molding processing parameters significantly affect the
mechanical properties of thermoplastic composites. A number of
researchers have studied the influence of molding processing pa-
* Corresponding author. rameters on mechanical properties [20e26]. Fujihara K et al. [20]
E-mail address: [email protected] (S.-y. Wang).
optimized the molding temperature and holding time of
Peer review under responsibility of China Ordnance Society

https://doi.org/10.1016/j.dt.2022.04.010
2214-9147/© 2022 China Ordnance Society. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

continuous CF/PEEK composite laminates with bending perfor- microscopic morphologies of the CF/PEEK specimen. The effects of
mance as an indicator, and it was pointed out that excessively high processing conditions on fiber/resin interaction and crystallization
molding temperature and too long holding time would lead to of CF/PEEK composites were systematically discussed, and the
degradation of PEEK and reduce mechanical properties. Similarly, mechanism of the effect of matrix crystallization on the mechanical
Xu Z et al. [22] found that when CF/PEEK composites via the properties of the composites was studied.
wrapped yarn method were held at higher molding temperatures
(above 420 C) for an excessive time (above 60 min), would have
2. Materials and fabrication
decreased mechanical properties due to matrix degradation. Zheng
B et al. [23] investigated the influence of molding temperature and
The material used unidirectional PEEK-carbon prepreg
molding pressure on the properties of woven CF/PEEK composites,
commercialized by VICTREX, which has a melting temperature of
and pointed out that the optimal process conditions for bending
343 C and a fiber mass fraction of 54%. For this study, CF/PEEK
performance were 400 C and 5 MPa, while the molding pressure
composite laminates were fabricated using the thermoforming
had little effect on interlaminar shear strength (ILSS). Dai G et al.
method. The CF/PEEK composite laminates of 2 mm nominal
[24] took the ILSS as the optimization objective, studied the effects
thickness were hand lay-up of 16 plies of unidirectional prepreg,
of molding temperature, molding pressure, and holding time on the
which were consolidated in a matching steel mold. A pre-crack was
mechanical properties of CF/PEEK cross-ply laminates, and pointed
created by laying a 16 mm thick polyimide high-temperature
out that the optimal molding conditions were holding time of at
resistant film on the middle plane, which was coated with a
least 20 min and molding temperature should be between 400 and
release agent.
420 C. The current research mostly phenomenologically estab-
The determination of processing parameters was closely related
lished the relationship between the molding process parameters
to the fluidity of the resin. According to the relationship between
such as molding temperature, pressure, and holding time and the
the viscosity of PEEK and the processing conditions [27], the vis-
macro-mechanical properties. However, these conclusions did not
cosity of PEEK in the molten state was not only related to molding
deeply clarify the regulation mechanism of the processing param-
temperature but also affected by holding time, as shown in Fig. 1. In
eters on the macro-mechanical properties of thermoplastic com-
addition, considering the degradation of the PEEK matrix caused by
posites. In addition, the viscosity of the molten thermoplastic resin
high molding temperature [20,22]. Therefore, this study selected
is affected by factors such as the molding temperature and holding
360e405 C as the molding temperature, 10e40 min as the holding
time [27]. Therefore, the thermoforming curing processing pa-
time, and designed the three-parameter four-level orthogonal
rameters of thermoplastic composites are not independent of each
thermoforming process conditions, as shown in Fig. 2. The exper-
other but affect each other. However, to the author’s knowledge,
imental design process and objectives of this paper were shown in
there is little information available in the literature about
Fig. 3.
comprehensively and systematically considering the coupling ef-
fect of processing parameters on the mechanical properties of
thermoplastic composites. 3. Characterization methods
The mechanical properties of semi-crystalline thermoplastic
composites, especially the fracture toughness, were mainly domi- 3.1. Mechanical characterization
nated by crystallinity, which was affected by the processing pa-
rameters [28e30]. The researchers discussed the effect of In order to characterize the influence of thermoforming process
processing parameters on the crystallization behavior of thermo- parameters on the mechanical properties of thermoplastic com-
plastic resin [31e35]. Yang X and Sun H et al. [33e34] studied the posites, double cantilever beam (DCB) tests and three-point
effect of different cooling rates on the non-isothermal crystalliza- bending tests are carried out through an electromechanical uni-
tion kinetics and discussed the effects of cooling rates on crystal versal testing machine (INSTRON 5569) at room temperature. DCB
growth and morphology. Gao S L et al. [36] further analyzed the tests are performed according to the ASTM standard D5528 with a
effect of cooling rate on the fracture toughness of thermoplastic loading speed of 2 mm=min. The nominal size of the DCB specimen
composites and revealed the mechanism of cooling rate regulating with the initial crack length of 50 mmða0 Þ is 150 25 2 mm3 , as
the fracture toughness of composites by affecting the crystallization
process. Processing parameters such as molding temperature,
pressure, and holding time also affect crystallization behavior.
However, little attention has been devoted to regulating these
processing parameters on the crystallization process of thermo-
plastic composites and their impact on mechanical properties.
Therefore, it is of great significance to study the effects of molding
temperature, pressure, and holding time on the crystallization of
composites and quantitatively discuss the relationship between
crystallization and mechanical properties.
This study aimed to investigate the effects of molding process
parameters on molding quality, mechanical properties, and crys-
tallization behavior of continuous CF/PEEK composites. Thus, the
CF/PEEK composite laminates were fabricated by three-parameter
four-level orthogonal thermoforming process parameters. Three-
point bending test, double cantilever beam (DCB) test, and differ-
ential scanning calorimetry (DSC) test were carried out to analyze
the effects of the molding process parameters (molding tempera-
ture, molding pressure, and holding time) on the mechanical
properties and crystallization behavior of the CF/PEEK composites.
The molding quality and failure mechanism were investigated by Fig. 1. Viscosity of PEEK as a function of time for different temperatures [27].

270
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

length in real-time. Regularly track the position of the crack tip and
record the applied load and opening displacement at each crack
extension. The mode-I interlaminar fracture toughness (GIC ) is
calculated from Eq. (1) [37].

3P d
GIC ¼ (1)
2bða þ jDjÞ

where P is the applied load, d is the load point displacement, b is the

width of a specimen, a is the delamination length, and D is the
corrective factor for crack length.
Three-point bending tests according to the ASTM standard
D7264 are conducted. The span-to-depth ratio is 32, and the cross-
head speed is 1 mm=min. The specimen dimension is 81 12
2 mm3 , as shown in Fig. 4 (b). The test specimens are cut out of the
laminates thermoformed with different processing parameters
using a waterjet cutting machine. Moreover, the fracture surfaces of
microstructure are observed using the scanning electron micro-
scope (SEM) to study the influence of molding process conditions
on crystallization and mechanical properties of CF/PEEK
Fig. 2. Fabrication conditions of continuous CF/PEEK composite laminates by L16 (43)
orthogonal experiments.
composites.

3.2. Thermal characterization

shown in Fig. 4(a). The digital image correlation (DIC) measurement
technique is used for crack extension length measurement, which is In order to study the crystallization behavior of CF/PEEK lami-
recorded using a HIKVISION MV-CA060-11 GM industrial camera nates under different molding process conditions, differential
(resolution 3072 2048, image element size 2.4 mm) with an scanning calorimetry (DSC) is carried out on the Mettler-Toledo
acquisition framerate of 10 s a piece. Based on the full-field DSC machine. The specimen mass is about 10 mg. The DSC tests
displacement obtained by the DIC measurement system, accu- are heated at a rate of 10 C/min from 25 to 400 C in a nitrogen
rately locate the crack tip position, and obtain the crack extension atmosphere [38]. The crystallinity degree (XC ) is determined using
the following equation [39]:

Fig. 3. Experimental design process and objectives.

Fig. 4. Sketches of specimens and testing fixtures for mechanical characterizations: (a) DCB specimen; (b) Three-point bending specimen.

271
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

extension. The load for crack extension increases with the increase
DHm of molding temperature and holding time. The GIC of CF/PEEK
XC ¼ 100% (2)
Hf ð1 aÞ composites is calculated and plotted in Fig. 5(c) and Fig. 5(d). It can
be clearly found that the GIC of CF/PEEK composites increases with
where Hf is the melting enthalpy of 100% crystalline PEEK, Hf ¼ increasing molding temperature, pressure and holding time, and a
130 J=g[40], DHm is value of the melting enthalpy, and a is the fiber similar trend is proposed in other Ref. [41].
mass fraction. The relationship between the GIC of CF/PEEK composites and the
processing parameters is summarized in Fig. 6. In order to more
intuitively study the influence of the molding process coupling
4. Results and discussion
effect on GIC, the orthogonal test results are interpolated to obtain
the contour map of each two processing parameters on GIC, as
4.1. Mode-I interlaminar fracture toughness
shown in Fig. 6(b), Fig. 6(c) and Fig. 6(d). From the experimental
data in Fig. 6(a), it can be found that the influence of processing
4.1.1. The influence of processing parameters on the fracture
parameters on the GIC of CF/PEEK composites does not change
toughness of CF/PEEK composites
monotonously, which is reflected in the maximum and minimum
Fig. 5 illustrates the curves of typical DCB force-displacement
values occurring at 375 C, as shown in Fig. 6(b). The GIC of CF/PEEK
and GIC-extended crack length of CF/PEEK specimens fabricated at
composites results from the coupling effect of various processing
different molding process conditions. During the load loading
parameters. The increase of the molding temperature reduces the
process, the interlayer of composites undergoes damage initiation
viscosity of the resin, and the increase of the molding pressure
and damage evolution until the load suddenly drops at the peak
promotes the permeability of the fiber/matrix, thereby improving
point and the crack extends. Then the slope of the force-
the interlayer performance of the composite laminate, as shown in
displacement curve decreased as the number of crack propaga-
Fig. 6(b) and Fig. 6(c). Similarly, increasing the molding time allows
tion increased, as shown in Fig. 5(a) and Fig. 5(b). It is observed that
for better impregnation of the resin with the fibers, improving the
the processing parameters have a significant effect on the crack

Fig. 5. Mode-I interlaminar fracture toughness of specimens fabricated at different molding process conditions: (a) Force-displacement curves; (b) Force-displacement curves; (c)
GIC-extended crack length curves; (d) GIC-extended crack length curves.

272
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 6. Relationship between GIC and molding process parameters of CF/PEEK composites: (a) Orthogonal test results of GIC; (b) Contour map of molding temperature and molding
pressure; (c) Contour map of molding pressure and holding time; (d) Contour map of holding time and molding temperature.

fiber/resin bond and thus the mechanical properties, as shown in influence of the processing parameters on the failure mechanism in
Fig. 6(d). However, the PEEK matrix evolves over long periods at Fig. 7. As shown in Fig. 7(a) and Fig. 7(b), the exposed fibers and the
high temperatures. Chain oxidation and crosslinking are the pro- voids can be clearly seen at the fracture of the specimens. This is
posed mechanisms for its thermal degradation, leading to a mainly due to the poor fluidity of the resin when the molding
decrease in the properties and an increase in the viscosity of the temperature is low, and the lower molding pressure is not enough
PEEK matrix [27], which is detrimental to the impregnation of the to promote penetration of the resin between the fibers, which leads
fibers with the PEEK matrix. to a weaker bonding interface of the CF/PEEK composites. As the
The curves above the contour maps in Fig. 6(b)eFig. 6(d) are the molding temperature and pressure increase, as shown in Fig. 7(c)
trend graphs of the three process parameters on GIC according to and Fig. 7(d), a large amount of resin residue on the surface of the
range analysis. It can be seen that the recommended molding fiber and a well-bonded fiber/matrix interface can be observed at
process parameters are 405 C, 1.5 MPa, and 40 min with the the fracture of the specimens. These clarify that the fiber/resin
objective of GIC. The range analysis results indicate that the influ- forms a strong interfacial bond under high temperature and pres-
ence degree of process parameters on GIC is molding pressure, sure. More energy is needed for interlaminar crack propagation,
temperature, and holding time in descending order. The detailed and the composite material exhibits higher interlaminar fracture
analysis is carried out in Section 4.4. toughness.
The fracture morphologies of the DCB specimens fabricated at
4.1.2. The influence of processing parameters on the fracture different holding times are observed in Fig. 8. As seen in Fig. 8(a)
morphology of DCB specimens and Fig. 8(b), exposed carbon fibers and separate resin clusters are
The fracture morphologies of the DCB specimens fabricated at clearly visible on the fracture surface of specimens. This is because
different temperatures and pressures are observed to identify the the short holding time leads to insufficient infiltration of the resin
273
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 7. SEM micrographs of the DCB fracture surfaces of specimens fabricated at different molding process conditions: (a) 360 C, 0.25 MPa (40 min); (b) 375 C, 0.5 MPa (20 min);
(c) 390 C, 1 MPa (40 min); (d) 405 C, 1.5 MPa (20 min).

Fig. 8. SEM micrographs of the DCB fracture surfaces of specimens fabricated at different process conditions: (a) 10 min (360 C, 0.5 MPa); (b) 20 min (390 C, 0.25 MPa); (c) 30 min
(375 C, 1 MPa); (d) 40 min (405 C, 0.5 MPa).

274
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

and fiber. The molten resin is unable to infiltrate the fiber bundles the crystal gap, and further increase the contact boundary area of
between the different layers, resulting in poor fiber/matrix inter- crystals. Due to the high stiffness of the crystal boundary structure,
face bonding. As the holding time increases, as shown in Fig. 8(c) which improves the modulus of the CF/PEEK composites. Similar to
and Fig. 8(d), the exposed fibers are significantly reduced, and the the results of the GIC analysis, the longer holding time at higher
improved fiber/resin bonding interface can be observed at the temperatures results in PEEK degradation and crosslinking [27],
fracture of the specimens. These indicate that a longer holding time which reduces the Eyy, as shown in Fig. 10(d). The detailed analysis
ensures adequate fiber and resin infiltration and interface strength, of the influence of molding process parameters on crystallization
thus improving the GIC of CF/PEEK composites. and the effect of crystallization on mechanical properties is carried
out in Sections 4.3 and 4.4.
The curves above the contour maps in Fig. 10(b)eFig. 10(d) are
4.2. Transverse bending performance the trend graphs of the three process parameters on Eyy according
to range analysis. The results show that the influence order of
4.2.1. The influence of processing parameters on the flexural process parameters on Eyy is the same as GIC, and the recommended
modulus of CF/PEEK composites molding process parameters are 405 C, 1.5 MPa and 30 min with
Fig. 9 presents the typical bending force-displacement curves of the objective of Eyy.
CF/PEEK composites specimens fabricated at different processing
conditions. The CF/PEEK specimens show linear elastic character-
istics until brittle failure. The different slopes of the curves reveal 4.2.2. The influence of molding parameters on the fracture
the variation of transverse flexural modulus (Eyy) of CF/PEEK morphology of three-point bending specimens
composites with the processing parameters. With the increase of To analyze the effects of the processing parameters on the fail-
molding temperature and molding pressure, the Eyy of the CF/PEEK ure mechanism of CF/PEEK composites, the fracture morphologies
composites increases, which is consistent with the experimental of the three-point bending specimens fabricated at different tem-
results of Zheng B et al. [23]. peratures and pressures are observed using SEM. As shown in
The relationship between the Eyy of CF/PEEK composites and the Fig. 11(a) and Fig. 11(b), exposed carbon fibers and the fibers
processing parameters is summarized in Fig. 10. In order to more debonding can be observed in the fracture of the specimens. This
intuitively investigate the coupling effect of processing parameters reflects that the movement of PEEK resin molecular chain segments
on mechanical properties of CF/PEEK composites, the contour map of PEEK resin under lower molding temperature and pressure is
of the Eyy under different processing parameters is obtained by impeded. Higher energy is required to overcome the movement
interpolating the results of the orthogonal experiments, as seen in between the chain segments, which is manifested by high resin
Fig. 10(b)eFig. 10(d) From the experimental data in Fig. 10(a), it can viscosity and poor fluidity, resulting in incomplete penetration of
be found that the effect of processing parameters on Eyy is not the matrix, high porosity of the composites, and poor fiber/matrix
linear. According to Fig. 10(b) and Fig. 10(c), the increase of molding interface strength. As the molding temperature and pressure in-
temperature reduces the viscosity of the resin, and the increase of crease, Fig. 11(c) and Fig. 11(d) show that the exposed carbon fibers
molding pressure and holding time promotes the infiltration of fi- and the voids are significantly reduced. These indicate that
bers with resin, thus improving the Eyy, and a similar trend is found increasing the molding temperature and pressure can ensure the
in other literature [22]. Meanwhile, increasing the molding pres- fiber/resin interface strength and improve resin and fiber infiltra-
sure can increase the density of the crystallization domain, reduce tion while making the composite laminate denser and reducing
porosity, thus enhancing the bending performance of the CF/PEEK
composites.
The fracture morphologies of the three-point bending speci-
mens fabricated at different holding times are observed in Fig. 12.
As shown in Fig. 12(a) and Fig. 12(b), the fiber/matrix debonding
can be observed at the fracture of specimens. This is mainly due to
the inability of the fibers and the matrix to fully infiltrate in a short
time, which leads to a weaker bonding interface of the CF/PEEK
composites. As the holding time increases, as shown in Fig. 12(c)
and Fig. 12(d), it can be observed that the resin attached to the fi-
bers has increased. These clarify that increasing the holding time
can increase the fiber/matrix interface strength, thereby improving
the transverse bending performance of the CF/PEEK composites.

4.3. The influence of processing parameters on the crystallization of

CF/PEEK composites

Generally, the decrease in crystallinity leads to a reduction in

stiffness and strength of semi-crystalline polymers but an increase
in ductility due to fewer secondary intermolecular bonds and an
increase in disorder. Therefore, it is essential to investigate the ef-
fect of molding process parameters on the crystallization behavior
Fig. 9. Three-point bending typical stress-displacement curves of specimens fabri- of semi-crystalline polymer composites. To reveal the effects of the
cated at different molding process conditions. processing parameters on CF/PEEK composites, DSC tests are
applied on the CF/PEEK specimens after thermoforming. The DSC

275
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 10. Relationship between Eyy and molding process parameters of CF/PEEK composites: (a) Orthogonal test results of Eyy; (b) Contour map of molding temperature and molding
pressure; (c) Contour map of molding pressure and holding time; (d) Contour map of holding time and molding temperature.

curves are shown in Fig. 13(a). The crystallinity of CF/PEEK speci- affects the number and size of crystalline entities but does not
mens is calculated according to Eq. (2), as shown in Fig. 16(a). change the ultimate crystallinity level of the composite material
As shown in Fig. 13, the melting temperature decreases with the [42].
increase of molding temperature and increases with the increase of The melting temperature of semi-crystalline polymer compos-
molding pressure. This is because, on the one hand, the melting ites is determined by factors such as crystal size, folding surface
temperature decreases as the molding temperature increases from energy and plate thickness [38]. Fig. 14(a) shows the microscopic
360 to 405 C, and a similar trend has been found in other studies morphology of the CF/PEEK fracture fabricated at 405 C, 1.5 MPa,
[42]. The decrease in melting temperature may be related to the and 20 min process parameters. It can be observed that high-
reduction of the number of remaining crystalline entities and the oriented spherulites are distributed on the fiber surface, and
change in the molecular structure of the PEEK matrix. On the other these spherulites constitute the columnar crystalline structure of
hand, molten resin mobility increases with molding pressure, high quality. Schematic diagram of the high-oriented spherulites
accompanied by increased melt shear effects leading to enhance generated on the fiber surface, as shown in Fig. 14(b). This is also
molecular chain orientation, thereby increasing the melting tem- consistent with the conclusions of other studies that the lower the
perature of the PEEK matrix. In addition, as shown in Fig. 13(b), it matrix melting temperature, the larger the size of the spherical
can be seen that the crystallinity remains relatively constant as the crystal, and the higher the molecular integrity. Conversely, the
molding temperature increase. This is also consistent with the higher the matrix melting temperature, the smaller the spherical
conclusions of other studies that the molding temperature mainly crystal size [43,44]. The crystal structures formed at lower molding

276
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 11. SEM micrographs of the three-point bending fracture surfaces of specimens fabricated at different molding process conditions: (a) 360 C, 0.25 MPa (40 min); (b) 375 C,
0.5 MPa (20 min); (c) 390 C, 1 MPa (40 min); (d) 405 C, 1.5 MPa (20 min).

Fig. 12. SEM micrographs of the three-point bending fracture surfaces of specimens fabricated at different molding process conditions: (a) 10 min (360 C, 0.5 MPa); (b) 20 min
(390 C, 0.25 MPa); (c) 30 min (375 C, 1 MPa); (d) 40 min (405 C, 0.5 MPa).

277
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 13. DSC results of CF/PEEK composites fabricated under different heating histories: (a) Heat flow-temperature curves; (b) Relationship between crystallinity, melting tem-
perature, and molding temperature.

Fig. 14. (a) SEM observation of the columnar crystalline structures and (b) schematic diagram of high-oriented spherulites on the fiber surface.

278
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

temperatures are not well defined, mainly due to the low mobility entities and the formation of new nuclei, but since the molding
of the polymer chains when the molding temperature is low, which temperature mainly affects the number and size of crystalline en-
limits the ability of the chains to diffuse to the growing crystal tities, it does not affect the final crystallinity of the CF/PEEK com-
fronts. Moreover, columnar crystal structures can be observed only posites. Therefore, the degree of crystallinity remains relatively
when the molding temperature is high enough to destroy the pre- constant as the molding temperature increases, as shown in
existing crystalline entities [42]. Therefore, based on the above Fig. 16(b). Molding pressure and holding time are as well important
results, it can be concluded that the higher the molding tempera- for the crystallization behavior of CF/PEEK composites. The molding
ture, the larger size of the well-defined spherical crystals generated pressure orients the polymer melt and induces nucleation, which
in the CF/PEEK composites, and the number of crystals is lower than leads to a higher crystallization rate. However, the slight difference
that of the specimen fabricated at lower molding temperatures. of each molding pressure in the thermoforming process leads to an
The relationship between molding process parameters, lamella, unobvious effect of the molding pressure on the crystallinity, as
and fiber-matrix bonding mechanisms on the degree of crystal- shown in Fig. 16(c). As the holding time increases, the chain
linity is shown in Fig. 15. The micro-morphology of the fiber surface oxidation and crosslinking of the PEEK matrix leads to a decrease in
of the specimen fabricated at 405 C, 0.5 MPa, and 40 min is shown the molecular mobility of crystallizable units in the melt, which
in Fig. 15(a). The results indicate that higher molding temperatures influences the nucleation mechanism and slows down the crystal
and longer holding time allow the PEEK matrix tends to diffuse and growth process, as shown in Fig. 16(d), and a similar trend is pro-
strongly adsorbs to the fiber surface, which is consistent with the posed in other literature [46]. In addition, appropriately increasing
conclusion of Brady R L et al. [45]. The schematic diagram of the the holding time can generate transverse crystals on the carbon
fiber-matrix interaction of the CF/PEEK composites in Fig. 15 is fiber surface [42], thus increasing the stiffness of the composites,
similar to the other studies [38]. The flattened crystalline lamellar but extremely short and extremely long holding time would reduce
chains with high crystallinity in the interfacial region enable more the transverse crystals [47].
polar components to interact functionally with the fiber surface.
The adhesion of the interfacial region is increased, which affects the
mechanical properties of the composites, such as improving shear 4.4. The relationship between molding process parameters,
performance. In contrast, holding for a longer time at a lower crystallization and mechanical properties
molding temperature reduces the viscosity of the molten PEEK
matrix. Due to the crosslinking of PEEK, the mobility of crystalliz- In the above Sections, the influence mechanism of processing
able unit molecules is further reduced, which affects the nucleation parameters on the mechanical properties and crystallinity of CF/
mechanism and slows down the crystal growth. Thus, the PEEK PEEK composites is discussed, respectively. However, it has been
matrix is unable to develop strong bonds with the fiber surface, as reported that the crystallinity of resin has a significant effect on
shown in Fig. 15(b), which is fabricated at 360 C, 0.25 MPa, 40 min. mechanical properties [31], clarifying the inherent relationship
The DSC test results of CF/PEEK composites under different between process parameters, crystallization and mechanical
processing parameters are calculated and drawn in Fig. 16(a). It can properties can further optimize the manufacturing process of
be seen that the crystallinity of the CF/PEEK composites fabricated thermoplastic composites. Schematic diagrams of deformation in
by the orthogonal process ranged from 23.5 to 33.6%. The molding the crystalline and amorphous domains of the DCB and the three-
temperature affects the destruction of the pre-existent crystalline point bending specimens are shown in Fig. 17 and Fig. 18. Due to
the macromolecular chains in the crystalline domain aligning in

Fig. 15. SEM of fracture surface and schematic diagram of fiber-matrix interaction of CF/PEEK composites under different molding process conditions: (a) SEM fracture surface of
high crystallinity specimen (XC ¼ 32.41%); (b) SEM fracture surface of low crystallinity specimen (XC ¼ 23.54%).

279
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 16. Relationship between crystallinity and molding process parameters of CF/PEEK composites: (a) Orthogonal test results of crystallinity; (b) Contour map of molding
temperature and molding pressure; (c) Contour map of molding pressure and holding time; (d) Contour map of holding time and molding temperature.

better order and owning stronger intermolecular forces, resulting to stress induction. The higher crystallinity means a higher per-
in higher strength and rigidity [48,49], the matrix becomes more centage of brittleness in the material, while the ductility of the
stiff and brittle after crystallization, as shown in Fig. 17 and Fig. 18. matrix plays a dominant role in the crack extension of the DCB
Thus, the higher crystallinity enhances the bending performance of specimen. On the other hand, increasing the molding pressure can
CF/PEEK composites, as shown in Fig. 18. However, the ductility of effectively improve the infiltration of high-viscosity resin with fi-
the matrix plays a dominant role in the crack extension of the DCB bers and lead to a denser laminate, thus improving the GIC.
specimen. As the crystallinity increases, the percentage of ductile Therefore, for the GIC of CF/PEEK composite laminate, there has a
amorphous domains decreases, which reduces the GIC of CF/PEEK competitive relationship between the crystallinity and the molding
composites, as shown in Fig. 17. quality (infiltration, porosity, etc.). According to the data of DCB
Molding temperature has a slight effect on matrix crystallinity, range analysis results (Fig. 6), GIC increases with the increase of
but higher molding temperatures can reduce molten PEEK viscosity molding pressure, which means that high-quality laminates fabri-
and improve fiber-matrix infiltration, thereby increasing the GIC of cated under higher molding pressure contribute more to the GIC of
CF/PEEK composite laminate. Molding pressure affects the GIC of CF/ CF/PEEK composites. The crystallinity decreases with increasing
PEEK composites from two aspects. On the one hand, the higher holding time, therefore a longer holding time is beneficial to
molding pressure increases the crystallinity of the matrix PEEK due improve the matrix ductility and thus the GIC of CF/PEEK

280
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 17. Schematic diagram of mode-I delamination of semi-crystalline CF/PEEK composites specimen.

Fig. 18. Schematic diagram of semi-crystalline CF/PEEK composites specimen deformation under bending load: (a) Schematic drawing of the bending test; (b) Deformation of
crystalline and amorphous domains under bending load.

composites, which is consistent with the range analysis results the matrix crystallinity, and conversely, the Eyy is positively corre-
(Fig. 6) of the DCB test data. lated with the matrix crystallinity. Therefore, the thermoforming
The effect of molding temperature on Eyy of CF/PEEK composites process affects the mechanical properties of the composites by
is similar to that of GIC, where Eyy increases with increasing molding influencing the matrix crystallization. In addition, it is worth noting
temperature. It is worth noting that it is easier to generate well- that the crystallinity of some specimens does not match their
defined spherulites and columnar crystalline structures at higher mechanical performance, mainly due to defects caused by pro-
molding temperatures, resulting in increased Eyy. Due to the effect cessing parameters, as marked in Fig. 19. Indeed, the mechanical
of stress-induced crystallization, increasing the molding pressure properties of composites can be affected by other factors (such as
not only improves the crystallinity but also improves the molding void defects), which need to be further investigated in future
quality, thus improving Eyy. The holding time affects the Eyy of CF/ studies.
PEEK composite laminate from two aspects. On the one hand, The elucidation of the inherent relationship between molding
increasing the holding time reduces the crystallinity, which is not process parameters, mechanical properties, and crystallization is
conducive to improving Eyy. On the other hand, although the proper essential for the high-quality molding of CF/PEEK composites. The
increase of holding time can improve the molding quality, due to studies in Sections 4.1 and 4.2 show that the GIC and bending per-
the degradation and crosslinking of resin at higher temperatures, formance of CF/PEEK composites are positively correlated with the
which makes too long holding time is not conducive to the molding molding temperature, pressure and holding time, mainly due to the
quality. The above conclusions are consistent with the range anal- degree of fiber/matrix infiltration promoted by increasing these
ysis results (Fig. 10) of the three-point bending test data. process parameters. However, the effect of matrix crystallization on
The relationship between the mechanical properties and crys- the mechanical properties cannot be ignored. Although the study in
tallinity of CF/PEEK composites under orthogonal processing con- Section 4.3 shows that the effect of molding process parameters on
ditions fabricated is exhibited in Fig. 19. It can be seen that the the crystallinity of the PEEK matrix is not highly significant, but it
crystallinity of the PEEK matrix has an approximately linear effect seriously affects the crystalline morphology, which is related to the
on the mechanical properties. The GIC is negatively correlated with mechanical properties of CF/PEEK composites. On the one hand, the

281
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

Fig. 19. Trend graph of mechanical properties changing with crystallinity: (a) Mode-I interlaminar fracture toughness; (b) Transverse flexural modulus.

Fig. 20. SEM micrographs of spherulitic structures of the specimen are fabricated at 405 C, 1.5 MPa, and 20 min.

consolidation of CF/PEEK composites should be carried out at a PEEK composites were studied in this work. The CF/PEEK composite
sufficiently high temperature to destroy the pre-existing crystalline laminates were fabricated via thermoforming, using the three-
entities in the composite prepreg, thus improving the toughness of parameter four-level orthogonal test. The performance of CF/PEEK
the PEEK matrix and the GIC of the CF/PEEK composites [47]. On the composite laminates was analyzed by mechanical characterization,
other hand, the presence of columnar crystalline structures and crystallization characterization, and microscopic observations. The
spherulites improves the modulus of the composites. Although main conclusions are as follows:
PEEK resin degrades at higher temperatures, the degradation does
not prevent the formation of well-defined crystalline structures, (1) The mode-I interlaminar fracture toughness and transverse
and the crystallization domain grows with the appropriate increase flexural modulus are positively related to the molding tem-
in holding time [42,47], thus increasing the Eyy of composites. As perature, pressure and holding time. The mechanical prop-
observed for the spherulitic structures of CF/PEEK composites erties of CF/PEEK composites are not only affected by the
fabricated at 405 C, 1.5 MPa and 20 min in Fig. 20. In addition, the degree of matrix/fiber infiltration and molding quality, but
increased molding pressure not only promotes the penetration of also by potential matrix crystallization. The increased
the resin into the fiber, but also increases the density of the crys- molding temperature, pressure and holding time can
talline domain and reduces the crystal gap. Due to the stress- improve the degree of fiber/matrix infiltration and reduce
induced nucleation, the crystallization rate is accelerated, thereby molding defects such as voids, which improves mechanical
increasing the crystallinity and the mechanical properties of the properties. However, it is necessary to note that long-time
composites. Therefore, the influence of the molding process on high-temperature conditions can alter the viscosity and
crystallization behavior significantly affects the mechanical prop- properties of the matrix due to crosslinking and degradation,
erties of CF/PEEK composites. thereby reducing the molding quality.
(2) The molding process parameters affect the crystallinity and
crystalline morphology of the matrix, which in turn affects
5. Conclusions the mechanical properties of the composites. The results
show that the crystallinity of the matrix has an almost linear
In order to improve the molding quality of thermoplastic com- influence on the mechanical properties of CF/PEEK com-
posites for extension and application in the cutting-edge equip- posites. Sufficiently high molding temperature can improve
ment of the aerospace and defence fields, the process/ the toughness and mode-I interlaminar fracture toughness of
crystallization/property inherent relationships of continuous CF/
282
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

the composites by destroying pre-existing crystallites. based composites in defence industry: a review[J]. Defence Technology; 2022.
[13] Bielawski R. Composite materials in military aviation and selected problems
Although increasing the holding time affects the nucleation
with implementation[J]. Review of the Air Force Academy 2017;(1):11.
mechanism and decreases the crystallinity due to matrix [14] Benkaddour A, Lebrun G, Laberge-Lebel L. Thermostamping of [0/90] n car-
degradation, it can form well-defined crystalline structures, bon/peek laminates: influence of support configuration and demolding tem-
which increases the flexural modulus of the composites. perature on part consolidation[J]. Polym Compos 2018;39(9):3341e52.
[15] Wang S, Zhou Z, Zhang J, et al. Effect of temperature on bending behavior of
Increased molding pressure can increase the crystallization woven fabric-reinforced PPS-based composites[J]. J Mater Sci 2017;52(24):
rate due to stress-induced nucleation, and it can increase the 13966e76.
density of the crystallization domain, which improves the [16] Tatsuno D, Yoneyama T, Kawamoto K, et al. Hot press forming of thermo-
plastic CFRP sheets[J]. Procedia Manuf 2018;15:1730e7.
mechanical properties of the composites. [17] Wang S, Wen L, Xiao J, et al. Influence of strain rate and temperature on
(3) The above results suggest that the optimal process parameter mechanical properties of carbon woven-ply PPS thermoplastic laminates
array in certain process conditions are molding temperature under dynamic compression[J]. Polym Test 2020;89:106725.
[18] Lu C, Wang J, Lu X, et al. Wettability and interfacial properties of carbon fiber
of 405 C, molding pressure of 1.5 MPa, and holding time of at and poly (ether ether ketone) fiber hybrid composite[J]. ACS Appl Mater In-
least 30 min based on the thermoforming method. This terfaces 2019;11(34):31520e31.
processing condition ensures excellent transverse bending [19] Hamdan S, Swallowe GM. Crystallinity in PEEK and PEK after mechanical
testing and its dependence on strain rate and temperature[J]. J Polym Sci B
performance and mode-I interlaminar fracture toughness of
Polym Phys 1996;34(4):699e705.
the continuous CF/PEEK composites. [20] Fujihara K, Huang ZM, Ramakrishna S, et al. Influence of processing conditions
on bending property of continuous carbon fiber reinforced PEEK composites
[J]. Compos Sci Technol 2004;64(16):2525e34.
[21] Salek MH. Effect of processing parameters on the mechanical properties of
Data availability statement carbon/PEKK thermoplastic composite materials[D]. Concordia University;
2005.
[22] Xu Z, Zhang M, Gao SH, et al. Study on mechanical properties of unidirectional
Research data are not shared.
continuous carbon fiber-reinforced PEEK composites fabricated by the
wrapped yarn method[J]. Polym Compos 2019;40(1):56e69.
Declaration of competing interest [23] Zheng B, Li M, Deng T, et al. Process-structure-property relationships of
thermoformed woven carbon-fiber-reinforced polyether-ether-ketone com-
posites[J]. Polym Compos 2019;40(10):3823e34.
The authors declare that they have no known competing [24] Dai G, Zhan L, Guan C, et al. The effect of moulding process parameters on
financial interests or personal relationships that could have interlaminar properties of CF/PEEK composite laminates[J]. High Perform
Polym 2020;32(7):835e41.
appeared to influence the work reported in this paper.
[25] Dai G, Zhan L, Guan C, et al. Optimization of molding process parameters for
CF/PEEK composites based on Taguchi method[J]. Composites and Advanced
Acknowledgements Materials 2021;30. 26349833211001882.
[26] Lessard H, Lebrun G, Benkaddour A, et al. Influence of process parameters on
the thermostamping of a [0/90] 12 carbon/polyether ether ketone laminate[J].
The authors of this paper acknowledge the financial support of Compos Appl Sci Manuf 2015;70:59e68.
the National Natural Science Foundation of China (NO. 11902255, [27] Bessard E, De Almeida O, Bernhart G. Melt state behaviour of PEEK and pro-
cessing window interpretation for fast compression moulding process[C]//AIP
U1837601 and 52090051).
conference proceedings. American Institute of Physics 2011;1315(1):167e72.
[28] Conrad TL, Jaekel DJ, Kurtz SM, et al. Effects of the mold temperature on the
Appendix A. Supplementary data mechanical properties and crystallinity of hydroxyapatite whisker-reinforced
polyetheretherketone scaffolds[J]. J Biomed Mater Res B Appl Biomater
2013;101(4):576e83.
Supplementary data to this article can be found online at [29] Chadwick AR, Kotzur K, Nowotny S. Moderation of thermoplastic composite
https://doi.org/10.1016/j.dt.2022.04.010. crystallinity and mechanical properties through in situ manufacturing and
post-manufacturing tempering: Part 1eMechanical characterisation[J]. Com-
pos Appl Sci Manuf 2021;143:106286.
References [30] Yang C, Tian X, Li D, et al. Influence of thermal processing conditions in 3D
printing on the crystallinity and mechanical properties of PEEK material[J].
[1] Saleem A, Frormann L, Iqbal A. High performance thermoplastic composites: J Mater Process Technol 2017;248:1e7.
study on the mechanical, thermal, and electrical resistivity properties of [31] Yao C, Qi Z, Chen W, et al. Experimental study on CF/PEEK thermoplastic
carbon fiber-reinforced polyetheretherketone and polyethersulphone[J]. fastener: effects of fastener matrix crystallinity and fibre content on the
Polym Compos 2007;28(6):785e96. strength of single-lap joint[J]. Compos B Eng 2021;213:108737.
[2] Yao SS, Jin FL, Rhee KY, et al. Recent advances in carbon-fiber-reinforced [32] Chen M, Chung CT. Crystallinity of isothermally and nonisothermally crys-
thermoplastic composites: a review[J]. Compos B Eng 2018;142:241e50. tallized poly (ether ether ketone) composites[J]. Polym Compos 1998;19(6):
[3] Mahesh V, Joladarashi S, Kulkarni SM. A comprehensive review on material 689e97.
selection for polymer matrix composites subjected to impact load. J]. Defence [33] Yang X, Wu Y, Wei K, et al. Non-isothermal crystallization kinetics of short
Technology 2021;17(1):257e77. glass fiber reinforced poly (ether ether ketone) composites[J]. Materials
[4] Wang WH, Zhao QC, Chang HY, et al. The development of fiber reinforced 2018;11(11):2094.
thermoplastic composite materials from German, Japan and United States [34] Sun H, Yang X, Wei K, et al. Non-isothermal crystallization kinetics of
based on information processing technology[C]//Applied Mechanics and Ma- continuous glass fiber-reinforced poly (ether ether ketone) composites[J].
terials, vol. 713. Trans Tech Publications Ltd; 2015. p. 2946e9. J Therm Anal Calorim 2019;138(1):369e78.
[5] Vaidya UK, Chawla KK. Processing of fibre reinforced thermoplastic compos- [35] Regis M, Zanetti M, Pressacco M, et al. Opposite role of different carbon fiber
ites[J]. Int Mater Rev 2008;53(4):185e218. reinforcements on the non-isothermal crystallization behavior of poly
[6] Liu H, Liu J, Ding Y, et al. The behaviour of thermoplastic and thermoset carbon (etheretherketone)[J]. Mater Chem Phys 2016;179:223e31.
fibre composites subjected to low-velocity and high-velocity impact[J]. [36] Gao SL, Kim JK. Cooling rate influences in carbon fibre/PEEK composites. Part
J Mater Sci 2020;55(33):15741e68. II: interlaminar fracture toughness[J]. Compos Appl Sci Manuf 2001;32(6):
[7] Liu J, Liu H, Kaboglu C, et al. The impact performance of woven-fabric ther- 763e74.
moplastic and thermoset composites subjected to high-velocity soft-and [37] May M. Measuring the rate-dependent mode I fracture toughness of
hard-impact loading[J]. Appl Compos Mater 2019;26(5):1389e410. compositeseA review[J]. Compos Appl Sci Manuf 2016;81:1e12.
[8] Gao X, Huang Z, Zhou H, et al. Higher mechanical performances of CF/PEEK [38] Gao SL, Kim JK. Cooling rate influences in carbon fibre/PEEK composites. Part
composite laminates via reducing interlayer porosity based on the affinity of 1. Crystallinity and interface adhesion[J]. Compos Appl Sci Manuf 2000;31(6):
functional s-PEEK[J]. Polym Compos 2019;40(9):3749e57. 517e30.
[9] Pan L, Yapici U. A comparative study on mechanical properties of carbon fiber/ [39] Kong Y, Hay JN. The measurement of the crystallinity of polymers by DSC[J].
PEEK composites[J]. Adv Compos Mater 2016;25(4):359e74. Polymer 2002;43(14):3873e8.
[10] Li H, Englund K. Recycling of carbon fiber-reinforced thermoplastic composite [40] Blundell DJ, Osborn BN. The morphology of poly (aryl-ether-ether-ketone)[J].
wastes from the aerospace industry[J]. J Compos Mater 2017;51(9):1265e73. Polymer 1983;24(8):953e8.
[11] Yadav R, Tirumali M, Wang X, et al. Polymer composite for antistatic appli- [41] Li MC, Loos AC. The effects of processing on interply bond strength of ther-
cation in aerospace[J]. Defence Technology 2020;16(1):107e18. moplastic composites[J]. J Reinforc Plast Compos 1992;11(10):1142e62.
[12] Harussani MM, Sapuan SM, Nadeem G, et al. Recent applications of carbon- [42] Denault J, Vu-Khanh T. Crystallization and fiber/matrix interaction during the

283
X.-l. Ma, L.-h. Wen, S.-y. Wang et al. Defence Technology 24 (2023) 269e284

molding of PEEK/carbon composites[J]. Polym Compos 1992;13(5):361e71. [47] Vu-Khanh T, Frikha S. Influence of processing on morphology, interface, and
[43] Davies P, Cantwell WJ, Jar PY, et al. Cooling rate effects in carbon fiber/PEEK delamination in PEEK/carbon composites[J]. J Thermoplast Compos Mater
composites[J]. Compos Mater: Fatigue and fracture. 1991;3:70e88. 1999;12(2):84e95.
[44] Uralil FS, Newaz GM, Lustiger A. Processing effects and damage tolerance in [48] Mark JE. Physical properties of polymers handbook[M]. New York: Springer;
poly (etheretherketone) composites[J]. Polym Compos 1992;13(1):7e14. 2007.
[45] Brady RL, Porter RS. Interfacial adsorption and crystallization of polycarbonate [49] Jin L, Ball J, Bremner T, et al. Crystallization behavior and morphological
in carbon fiber composites[J]. J Appl Polym Sci 1990;39(9):1873e85. characterization of poly (ether ether ketone)[J]. Polymer 2014;55(20):
[46] Denault J, Dumouchel M. Consolidation process of PEEK/carbon composite for 5255e65.
aerospace applications[J]. Adv Perform Mater 1998;5(1):83e96.

284