Fibers and Polymers 2017, Vol.18, No.

8, 1614-1622 ISSN 1229-9197 (print version)

DOI 10.1007/s12221-017-7384-4 ISSN 1875-0052 (electronic version)

Effect of Autoclave Pressure on Interfacial Properties at Micro- and Macro-

Level in Polymer-matrix Composite Laminates
Tengfei Chang1, Lihua Zhan1,2*, Wei Tan3*, and Shujian Li1
1
Light Alloys Research Institute, Central South University, Changsha 410083, China
2
School of Mechanical and Electrical Engineering, Central South University, Changsha 410083, China
3
Engineering Department, Cambridge University, Cambridge CB2 1PZ, UK
(Received April 24, 2017; Revised July 4, 2017; Accepted July 6, 2017)

Abstract: Variable autoclave pressures were applied on polymer-matrix composite to evaluate the effects on interfacial shear
strength (IFSS) and interlaminar shear strength (ILSS) in [0]10 T800/X850 composite laminates. The characterizations of
microstructure of the laminates were obtained using optical digital microscope and scanning electron microscope. The IFSS
and ILSS were measured by fiber push-in and short beam three-point bending tests respectively. Results have shown that
there exists an optimum range of autoclave pressure. The delamination and voids are mainly located at the plies of laminates,
where the shapes of voids are changeable. The number and size of voids decrease as the pressure increases, while the
impregnation of resin on fibers is improved continuously, leading to an increase for the fiber volume fraction. The IFSS and
the ILSS are related with the pressure all in the polynomial increasing models. For void contents are above 1 %, the ILSS
increases progressively as the pressure increases. Conversely, when void contents are less than 1 %, the changes of ILSS are
negligible and the ILSS is insensitive to the voids. For T800/X850 composite, the autoclave pressure in the range of 0.4 MPa
to 0.6 MPa is recommended.
Keywords: Polymer composites, Autoclave pressure, Void content, Fiber volume fraction, Interfacial properties

Introduction important role in autoclave process. In the early stage of

curing process, the temperature can reduce the resin viscosity,
Carbon fiber-reinforced polymer-matrix composites are which is beneficial to the resin flow and infiltration on the
widely used in aeronautical components with their performance fibers surface. At the same time, it can also promote the
advantages of light weight, high specific strength and development of composite curing cross-linking reaction that
modulus [1,2]. Autoclave molding process is the major way changes the resin from linear structure to three-dimensional
for fabricating the main-bearing or secondary-bearing network structure, but it is only related to the resin system
composite components of aircraft. The composite prepregs itself. The pressure is an important optimizing parameter
were laid on the mold with a vacuum state, and undergoing controlled by the high-temperature compression gas in an
the simultaneous application of uniform temperature and autoclave. In the process of forming the composite laminates,
pressure in an autoclave, it was ultimately consolidated to the pressure can evenly apply to the prepreg sheets to drive
the shape and quality [3,4]. However, the disadvantages also the resin flow between the fibers and exclude the retained
exist in autoclave process because of the long process cycle, water bubbles and volatile gas out of the laminates. If the
high energy consumption and low efficiency. Some scholars external pressure is too large, the resin would flow faster
have carried out new non-autoclave forming process one than normal and result in unexpected defects in local area,
after another, such as resin film infusion process [5], resin such as poor glue, rich fat and fibers’ distorting etc.
transfer molding process [6] and pultrusion process [7]. But Meanwhile, the excessive resin would be sucked from the
for forming the high-performance complex elements, external bleeder cloth to cause the oversize in thickness of laminates
pressure is required to eliminate defects (e.g. voids and and affect its assembly precision. If the external pressure is
delamination) and improve interface bonding state, so as to too small, the resin cannot flow sufficiently in the laminates,
ensure materials with excellent mechanical properties. which would weaken the bonding interfaces. Then, the
Those techniques mentioned above cannot often meet the delamination defect would be produced that have a severe
mechanical performances obtained in autoclave process. effect on the mechanical properties. Besides, the water
Therefore, it is necessary to study the optimization of bubbles and volatile gas would not be eliminated timely in
autoclave molding process route and make it toward the the curing process. There is a high possibility to form the
direction of energy-saving, cost-control and efficiency voids in the resin matrix. Therefore, it is very important to
maximization. understand the influence mechanism of autoclave pressure
Autoclave pressure and curing temperature plays an on the forming quality and precision of composite laminates.
Many scholars studied the effect of autoclave pressure on the
*Corresponding author: [email protected] mechanical properties of composites mainly in the macroscopic
*Corresponding author: [email protected] level [8-10]. They all pointed that the void content decreased

1614
Pressure on Interfacial Properties of Composites Fibers and Polymers 2017, Vol.18, No.8 1615

as the increase of the pressure, and the mechanical properties composite panels cured at different autoclave pressure
can be increased by reducing the amount of voids. But the conditions using a precision diamond saw. The samples were
in-situ effect of autoclave pressure on the microscale fiber/ filled in an epoxy resin to facilitate the polishing. Surfaces
matrix interfacial properties is not considered, and its perpendicular to fibers were polished with a series of the
relationship with interlaminar shear failure is rarely mentioned. silicon carbide abrasive papers of 600, 1000, 1500 grit size,
A better understanding of these effect is crucial to optimize then finished with 1 µm diamond paste, followed by a
the autoclave pressure conditions in an autoclave process of ultrasonic cleaning. The distribution, the number and the
composites. Analysis of their failure mechanisms provides a size of the internal defects in composite laminates were
good basis for multiscale modelling. observed using the optical digital microscope (ODM, model:
In this paper, the fiber/matrix interface shear strength OLYMPUS DS×500, Japan). The spatial distribution and
(IFSS) and the interlaminar shear strength (ILSS) were volume fraction of fibers in the intraply treated by spray-
investigated by push-in and short-beam tests. The influence gold were analyzed using the scanning electron microscope
mechanism of autoclave pressure on microscopic defects (SEM, model: TESCAN MIRA3 LMU, Czech).
and structure state of composite were analyzed in detail using
optical digital and scanning electron microscopic techniques, Mechanical Testing
respectively. The quantitative evaluations of the relationship Interlaminar Shear Strength Test (ILSS)
between autoclave pressure, microscopic characteristics and To assess the effects of autoclave pressure on the
the interfacial properties were established. Moreover, the interlaminar interface properties of composite laminates, the
interfacial failure modes of composites under different interlaminar shear strength tests were selected according to
autoclave pressure were also discussed. Chinese standard JC/T 773-2010. Samples of 20 mm (length)×
10 mm (width), and the thickness was the full thickness of
Experimental cured laminates, were cut from the center of the composite
panels and tested using the three-point short-beam bending
Material method with a span ratio of 7:1, as is shown in Figure 1. Five
The experimental composite material under research was tests for each group of samples were performed under the
the T800/X850 carbon fiber reinforced epoxy prepreg stoke control at a constant speed of 1 mm/min using the
(purchased from Commercial Aircraft Co., Ltd., China). The CMT5105 tensile testing machine (produced by Sansi Taijie
Cytec’s CYCOM X850 is a solvent-free epoxy resin which Co., Ltd., China). The interlaminar shear strength (ILSS,
has been added the thermo-plastic particles to enhance the MPa) of the composite samples was determined according to
fracture toughness and impact properties, applied to the equation (1),
aviation fiber prepreg manufacturing. The carbon reinforcement
3 Pmax-
used in the prepreg is a high strength carbon fiber of T800 τILSS = --- × --------- (1)
4 bh
with a fiber diameter around 5.1 µm. The initial resin
content of 35 % and the 190 g/m2 prepreg areal density were where Pmax is the maximum load (N), b is the sample width
all offered by the manufacturer. (mm) and h is the sample thickness (mm). The fracture
modes of samples after ILSS tests were examined by optical
Fabrication of Unidirectional Composite Laminates digital microscope (OLYMPUS DS×500, Japan) to analyze
In order to investigate the effect of autoclave pressure on the main fracture mechanism.
microscopic structures and interfacial properties of the Interfacial Shear Strength Test (IFSS)
composites. Four unidirectional composite laminates were In order to understand the effects of autoclave pressure on
fabricated by different kinds of the external pressure in the fiber/matrix interface properties of composite, fiber
autoclave process. The pressures were set as 0.0 MPa, 0.2 MPa, push-in tests were conducted by a triboindenter instrument
0.4 MPa and 0.6 MPa respectively. Under a given temperature (model: Hysitron TI 950, US) with a diamond flat punch of
cycles, the laminates were heated to the 180 oC at the rate of 5 µm in diameter. Micromechanical tests were performed
1.5 oC/min, and then were held for 150 min. Finally it was under displacement control at the 40 nm/s with the
cooled down to room temperature by means of air cooling. maximum displacement of 1200 nm and then unloading was
Each kind of pressure was controlled precisely by the preset conducted to the original position. The schematic and the
program of the autoclave system, and began to boost when experimental setup for the push-in test are shown in Figure
the temperature was reached to 30 oC. The dimension of the 2. The sample was inspected under the optical microscopy to
cured square panels was 100 mm×100 mm, and the final find the position of typical fiber packing patterns. Afterwards,
thickness of these panels was approximately 2.0±0.2 mm. the sample was moved beneath the low-load with maximum
load of 120 mN indenter probe. The in-situ images are
Microstructure Observation obtained by raster scanning the low-load indenter probe over
Samples of 10×10 mm2 were cut from each piece of the sample surface to allow for precise test placement and
1616 Fibers and Polymers 2017, Vol.18, No.8 Tengfei Chang et al.

Figure 1. (a) Schematic diagram of the three point bending method and (b) experimental setup with the ILSS test.

Figure 2. (a) Schematic of the fiber push-in test [12] and (b) experimental setup with the Hysitron TI950 Triboindenter.

Figure 3. (a) Optical micrograph of composite cross-section and (b) in-situ imaging showing the detail of one hexagonal packing before and
after fiber push-in tests.

microstructure identification. Afterwards, the sample was by its six nearest neighbors, as is show in Figure 3. The
tested under the high load indenter probe with maximum interfacial shear strength (IFSS, MPa) was determined
load of 950 mN to obtain the critical load Pc at the onset of according to equation (2) based on the shear-lag model [11,12],
interface failure. nP
For experimental comparison, samples with size of τIFSS = ----------c-2 (2)
10 mm×10 mm×2 mm were embedded into epoxy resin to 2πr
conduct the IFSS tests. Over 15 fibers per sample groups where Pc is the critical load at the onset of interface failure
were tested and these fibers were performed with a typical (mN), r is the fiber diameter (µm), and n is a parameter that
hexagonal packing pattern formed by one fiber surrounded depends on the fiber/matrix elastic properties and the
Pressure on Interfacial Properties of Composites Fibers and Polymers 2017, Vol.18, No.8 1617

constraint caused by the surrounding fibers of composite. n

can be determined according to equation (3) [11],
S0
n = ---------- (3)
πrEf
where S0 is the slope of the fiber/matrix linear elastic region
during the fiber push-in process, Ef is the longitudinal elastic
modulus of the fiber.

Results and Discussion

Microscopic Structural Analysis

The microscopic defects and the fiber compactness
degrees can be observed clearly by the application of optical
digital microscopy (ODM) and scanning electron microscopy
(SEM), which was illustrated in Figure 4 (Direction
perpendicular to the fibers). ODM images in Figure 4(a)-(d)
showed the microscopic defects of each specimen under
different autoclave pressure. As can be seen from it, serious
microscopic defects in the composite laminate appeared
when the autoclave pressure was 0 MPa. The mainly types
of the defects were the delamination and voids, which were
all along the thickness direction of the laminates. The
number of voids was numerous and their size was different.
Most of them were the shape of long strips located mainly in
the interface between plies. The reasons of defects formation
can be summed up in two aspects: firstly, the entrapment of
air during resin impregnation or during the lay-up; and
secondly, the volatiles arised from the resin system during
curing process. Because of without the external pressure on
the specimen, the entrapped gas and volatiles cannot be
excluded out of the composites and all were remained in the
matrix to shape a large number of bubbles. As the thermal
diffusion produced by the resin itself, these bubbles would
be assembled each other to form the long-strips voids or
delamination defects. From the Figure 4(b), the delaminations
were disappeared, the size and the number of voids decreased
significantly with the autoclave pressure of 0.2 MPa. There
was still a certain amount of voids existed in the resin
matrix, which were mainly the forms of columnar shape or
globular shape. Autoclave pressure played a role to condense Figure 4. The ODM and SEM micrographs of composites with
the carbon fibers and force the bubbles to dissolve in the different autoclave pressures; (a) 0.0 MPa, (b) 0.2 MPa, (c) 0.4 MPa,
resin matrix. Due to the limits of compaction for the lower and (d) 0.6 MPa.
external pressure, some bubbles cannot be dissolve timely
and were stuck in the resin matrix. As a result, this portion of the distribution of fibers in the resin was heterogeneous and
the bubbles would ultimately be formed in the columnar the fiber/resin interface gaps were evident, which indicated
shape or the globular shape. When the autoclave pressure that the interface bonding quality between fiber and resin
reached to the range of 0.4-0.6 MPa (Figure 4(c), Figure was very weak, and the compactness degree of fibers was
4(d)), it was evident that the voids reduce dramatically and low as well. Upon the application of pressure, most of the
there were almost negligible of voids between the piles. external loads were imposed on the composites to drive the
The SEM images of the fiber compactness degrees at the resin flow between the fibers, and the infiltration of resin to
intraply area were shown in Figure 4(a')-(d') for different the fibers had been improved. Figure 4(b') and Figure 4(c')
autoclave pressures. Figure 4(a') showed the microstructure showed the microstructure states with the autoclave pressure
states cured under the autoclave pressure of 0 MPa, where of 0.2 MPa and 0.4 MPa, respectively. In these figures, the
1618 Fibers and Polymers 2017, Vol.18, No.8 Tengfei Chang et al.

distribution of fibers in the resin was more uniform and the

degrees of compactness were also enhanced. As the autoclave
pressure reached to 0.6 MPa (Figure 4(d')), the uniformlty of
the fiber distribution in the resin was more compact, which
meant the degree of compactness had increased further.

Void Content and Fiber Volume Fraction Analysis

In order to establish the quantitative relationship of
autoclave pressure on the void content and the fiber volume
fraction of the composites, the voids distribution of the
specimens and the fiber compactness degrees at the intraply
area of one layer were analyzed by the image analysis
software GM-JX2000 with the Chinese standard GB3365-
82. The void content and the fiber volume fraction were then
calculated by statistics.
Figure 5 showed the relationship between the autoclave
Figure 6. Correlation of fiber volume fraction and autoclave
pressure and the void content of the specimens. The void pressure.
content declined with the increase of autoclave pressure and
the void content was the largest with the autoclave pressure
of 0 MPa, and its value was 8.04 %. When the autoclave range was up to 15.4 %. With the increasing of the autoclave
pressure was increased to 0.2 MPa, the void content decreased pressure to 0.4 MPa, the value of fiber volume fraction was
significantly, the value decreased from 8.04 % to 3.16 % and 80.92 % and its growth range was only 2.18 % as the
the corresponding range decreased by 60.7 %. As the pressure increased from 0.2 MPa up to 0.4 MPa. But when
autoclave pressure continued to rise, the decreasing trend of the autoclave pressure reached to 0.6 MPa, the value
the void content decreased slowly. When the autoclave pressure increased from 80.92 % up to 84.61 %. In which the growth
reached to 0.4 MPa or higher, the values of void content range expanded further and the value was 4.56 %.
were all less than 1 %. void fractions were 0.51 % and To understand the influence mechanism of autoclave
0.36 % respectively with the pressure of 0.4 MPa or 0.6 MPa. pressure on the void content and the fiber volume fraction of
The results of the fiber volume fraction under different composites. Assuming that the composite pregreg is a kind
autoclave pressure was shown in Figure 6, clearly indicating of porous grid structure, which is composed together by
that the fiber volume fraction increased as the increasing of reinforcing carbon-fibers and resin fluid with the saturation
the autoclave pressure and there existed a multilevel growth wetting viscosity [13]. According to Darcy’s law,
trend in change range. It can be seen from Figure 6, the fiber
K
volume fraction was the lowest under the autoclave pressure V = ---- ∇Presin (4)
μ
of 0 MPa and its value was 68.6 %. When the autoclave
pressure increased to 0.2 MPa, the value of fiber volume where V is the velocity vector of resin flow, K is the fiber
fraction increased from 68.6 % to 79.19 % and the change permeability, μ is the resin viscosity, ∇Presin is a pressure
gradient that along the thickness of the resin. It can be seen
that the velocity of resin flow was proportion to the pressure
gradient of resin, and inversely proportion to the viscosity of
resin. In curing process of the composites, the viscosity of
the resin reduced continuously as the increase of temperature.
When the external pressure was applied on the composites,
the pressure difference of the resin fluids would exist along
the thickness direction. And then the resin began to flow into
the fibers constantly, the excess resin and internal gas were
excluded out of the laminate and the compactness degree of
fibers had been improved. Although the low autoclave
pressure (below 0.2 MPa) can still provide the driving force
of the resin flow through the pressure gradient. In which, the
infiltration of resin on the fibers was improved and the
delamination defects were eliminated, shown in Figure 4(b),
but there were still some voids in the laminates. This can be
Figure 5. Correlation between void content and autoclave pressure. explained by the formula mentioned in the literature [14]
Pressure on Interfacial Properties of Composites Fibers and Polymers 2017, Vol.18, No.8 1619

and the equation was as follows, specimens was enhanced obviously. When the autoclave
pressure increased from 0 MPa to 0.2 MPa, the measured
σ
Presin = Pvoid – -------- (5) value of ILSS increased from 58.65 MPa to 86.75 MPa and
Mlv
the ILSS increased by up to 47.91 %. From Figure 7, the
where Presin is the pressure of resin, Pvoid is the internal ILSS values still increased significantly as the value of
pressure of bubbles, σ is the surface tension between resin autoclave pressure increased from 0.2 MPa to 0.4 MPa. But
fluid and bubbles, Mlv is the bubble ratio of volume to compared with the growth range of the autoclave pressure
surface area. Equation (5) demonstrated that the internal increased from 0 MPa to 0.2 MPa, the ILSS was grow
pressure of bubbles was related to the resin pressure. And slowly. In which the value of ILSS increased from 86.75 MPa
the higher of the autoclave pressure, firstly, it would enforce to 96.51 MPa and the ILSS increased by 11.25 %. However,
the resin to withstand the pressure greater and the internal when the autoclave pressure was higher than 0.4 MPa or
bubbles were dissolved more in the resin, resulting in the more, the difference of the ILSS were not significant. And
voids decreased further; secondly, the pressure difference of the values of ILSS increased from 96.51 MPa to 98.15 MPa
resin was greater that accelerated the excess resin to compact with the autoclave pressure increased from 0.4 MPa to
out of the composites under low void content, and led the 0.6 MPa, while the value of their growth range was only
compactness degree of fibers closely more. Therefore, in 1.69 %. It can be seen in Figure 7, the corresponding fitted
term of the experimental composite, a good interface quality curve for the trend of the ILSS with the aotuclave pressure
and lower void content can be obtained within the autoclave had shown the relationship between the autoclave pressure
pressure range of 0.4-0.6 MPa. and the ILSS was basically in line with the polynomial
increasing relationship, which was inversely proportional to
Interfacial Properties Analysis the relationship between the autoclave pressure and the void
Interlaminar Shear Strength Analysis content (as shown in Figure 5).
It has been pointed out in the previously literature that the To further investigate the influence mechanism of autoclave
void leads to the marked reduction in composite mechanical pressure on the ILSS of the specimens, the ODM micrographs
properties, especially those dominated by the resin failure of the interlaminar failure modes under different autoclave
behavior like the interlaminar shear strength [15,16]. Thus, pressures were shown in Figure 8. As can be seen from it,
the autoclave pressure should produce the better mechanical the failure cracks of the tested specimens distributed mainly
properties by reducing the voids of composites and the in the interlayer, and extended along the void direction when
hypothesis was tested by evaluating the ILSS of the laminates the range of the autoclave pressure was from 0 MPa to
cured under different autoclave pressure. The corresponding 0.2 MPa. Reasons resulting in such phenomena may be
measured values of ILSS with the different autoclave pressure affected by the numbers and sizes of voids between the plies.
were shown in Figure 7, and the lowest value of ILSS was Because when the bending load was applied on the specimens,
58.65 MPa with the autoclave pressure of 0 MPa. As the the stress concentration often occurred in the voids easily,
increasing of the autoclave pressure, the ILSS of the especially in the big void region. Therefore, the crack began
to develop from the big voids and then extended between the
piles, which illustrating the void could affect the initial stage
and evolutional stage of the ILSS failure. However, when
the autoclave pressure reached to 0.4 MPa or more, the crack
appeared in the intralayer of laminates and there were no
difference about the failure modes of ILSS. This is due to the
void content was less than 1 % under conditions of the
higher autoclave pressure (exceed 0.4 MPa), the numbers
and sizes of the voids were negligible and the interlaminar
interfaces between the piles were bonded closely. In these
circumstances, the voids were no longer the main cause for
the failure of ILSS and the crack was concentrated in the
interfacial layer between the fibers and the resin. ILSS with
autoclave pressure over 0.4 MPa are almost constant.
Therefore, with pressure over 0.4 MPa, it would help to
eliminate the voids of composites, and obtain higher values
of ILSS.
Interfacial Shear Strength Analysis
Figure 7. Relationship between interlaminar shear strength and The interfacial shear strength measured by push-in test can
autoclave pressure. be used to characterize the bonding strength of the fiber-
1620 Fibers and Polymers 2017, Vol.18, No.8 Tengfei Chang et al.

Figure 8. The ODM micrographs of the fracture modes of specimens tested to measure the ILSS; (a) 0.0 MPa, (b) 0.2 MPa, (c) 0.4 MPa, and
(d) 0.6 MPa.

Figure 9. (a) Load-indentation depth curves vs. autoclave pressure and (b) schematic diagram of critical load calculation.

matrix interface at the microscopic scale level, which is was followed by a linear region (with slope S0) after the full
affected by different fiber packing density and polymer contact due to the elastic deformation of the fiber and the
morphology cross-link density [17]. Figure 9(a) showed the matrix; And thirdly, with the increase of the indenter load,
typical load-depth curves of single fiber debonding process the non-linear region between loading and fiber displacement
obtained by push-in tests for different autoclave pressure. As was followed which illustrated the onset of fiber debonding
can be seen from it, these curves presented an S shape failure. All of these curves presented the same initial elastic
similarly. Firstly, there was initially an imperfect contact stiffness, shown in Figure 9(a), but the non-linear regions
process between the indenter and the fiber; Secondly, this followed were different. The difference demonstrated that
Pressure on Interfacial Properties of Composites Fibers and Polymers 2017, Vol.18, No.8 1621

the critical load, which enforced the failure of the fiber- Figure 11 showed the in-situ scanning images of the fiber
matrix interface, increased with the increasing of the indentation produced by push-in tests for each sample under
autoclave pressure. The determination of the critical load different autoclave pressures. Figure 11(a) showed that the
was based on the shear-lag model [12]. In which, two fiber indentation was shallowest under the autoclave pressure
parallel lines with the initial stiffness S0 as the slope were of 0 MPa, indicating the fiber loaded had the largest amount
drawn with offset displacement of 2 % and 10 %, shown in of rebound. This meant that the impregnation of resin on the
Figure 9(b), then two points intersected between the two fibers was poor and the adhesion between the fiber and the
parallel lines and the load-depth curve were formed a line, matrix was very weak, which lead to the most non-obvious
and last the point regarded as the critical load was identified fiber indentation and the weakening of the IFSS. Upon the
by the intersection of the line and the curve. Therefore, the application of autoclave pressure, the external pressure were
values of IFSS with different autoclave pressure can be loaded along the thickness direction of the laminates and
calculated from the experimental results of Figure 9(a) and forced most of the internal resin to infiltrate the fibers
equation (2). continuously. Figure 11(b) showed the typical fiber indentation
The results of IFSS of the specimens under different with the autoclave pressure of 0.2 MPa, which meant the
autoclave pressure were shown in Figure 10, clearly indicating infiltration of the resin on the fibers had been improved and
the value of IFSS increases with the increase of the autoclave the fiber volume fraction in the intraply was increased,
pressure. As is shown in Figure 10, the IFSS was the lowest leading to the better of the IFSS. But there were still parts of
with the autoclave pressure of 0 MPa and its value was voids due to the inadequate pressure. When the autoclave
56.38 MPa. As the autoclave pressure increased from 0 MPa pressure reached to 0.4 MPa, the internal voids of composites
to 0.2 MPa, the IFSS increased significantly, their values were almost eliminated and the compactness degree of fibers
increased from 56.38 MPa up to 66.72 MPa and the IFSS was more uniform (as shown in Figure 4(b)). This confirmed
increased by 18.34 %. When the autoclave pressure increased that the high pressure which drove the resin flow to
from 0.2 MPa to 0.4 MPa, the value of IFSS increased from impregnate the fiber would bring a good bonding of fiber-
66.72 MPa to 69.39 MPa and the IFSS increased only by matrix interface and the improvement of the IFSS. Therefore,
4.00 %. But with the increasing of the autoclave pressure, the fiber indentation was obvious as well (Figure 11(c)).
the rising trend of IFSS increased rapidly. Compared the As the autoclave pressure increased to 0.6 MPa, shown in
IFSS with the autoclave pressure of 0.4 MPa, the value Figure 11(d), the fiber indentation was the deepest. This
increased to 76.89 MPa as the autoclave pressure reached to indicated that the fibers were impregnated completely by the
0.6 MPa and the IFSS increased by 10.81 %. In order to resin and the fiber-matrix bonding interface had been
show the trend of the IFSS with the autoclave pressure, the
fitted curve had been existed in the Figure 10. It can be seen
from Figure 10, the relationship between the autoclave
pressure and the IFSS was in good agreement with that of
the polynomial increasing model, which was consistent with
the effect of autoclave pressure on the fiber volume fraction.

Figure 11. The in-situ scanning images of debonding failure

Figure 10. Relationship between interfacial shear strength and surfaces of fiber for push-in tests; (a) 0.0 MPa, (b) 0.2 MPa, (c) 0.4
autoclave pressure. MPa, and (d) 0.6 MPa.
1622 Fibers and Polymers 2017, Vol.18, No.8 Tengfei Chang et al.

improved further. But the higher autoclave pressure may However, if the magnitude of pressure exceeds the
also force the excess resin flowed so fast to exclude out of standard cure pressure, it leads to excess resin flowing out of
the laminates, and the compactness degree of fibers was composite and potential lean glue defects. For T800/X850
oversized. In which, the possibility of forming the lean composite, the autoclave pressure in the range of 0.4 MPa to
defects in local region would increase and the IFSS 0.6 MPa was recommended.
increased sharply. Therefore, for fabricating the high-quality
composite components, the autoclave pressure which is in Acknowledgements
the range of 0.4 MPa to 0.6 MPa is recommended. But too
much pressure may produce the new defects in some local The authors would like to gratefully acknowledges the
areas of components, such as lean defect, which should not funding from the National Key Basic Research Program of
be also ignored. China (973 program) under Grant No. 2014CB46502, Project
of Innovation-driven Plan in Central South University,
Conclusion 2015CX002 and State Key Laboratory of High-performance
Complex Manufacturing under Grant No. ZZYJKT2016-04.
Unidirectional laminates of a T800/X850 polymer matrix The authors would like to gratefully acknowledge the
composite were manufactured using variable autoclave composite research team members of Central South University
pressures. The influence of autoclave pressure on the micro- for their support and useful discussions in this research.
and macro- interfacial properties was assessed. Firstly, a
better characterization of the internal defects and the References
microstructure distribution in the composite were analyzed
in detail by the ODM and SEM technologies. Secondly, the 1. L. Liu, C. Y. Jia, J. M. He, F. Zhao, D. P. Fan, L. X. Xing,
influence mechanism of autoclave pressure on the void M. Q. Wang, F. Wang, Z. X. Jiang, and Y. D. Huang,
content and fiber volume fraction were illustrated. Finally, Compos. Sci. Technol., 56, 121 (2015).
the quantitatively relationships of autoclave pressure on the 2. B. Gao, R. L. Zhang, F. C. Gao, M. S. He, C. G. Wang, L.
micro- and macro-interfacial properties were established by Liu, L. F. Zhao, and H. Z. Cui, Langmuir, 32, 8339 (2016).
means of the IFSS tests and the ILSS tests respectively, and 3. I. M. Robert, Compos. Struct., 57, 3 (2002).
their interfacial failure modes were discussed. 4. G. B. Michael, Compos. Pt. A-Appl. Sci. Manuf., 33, 913
Results showed that the delaminations and the voids were (2002).
preferentially located between the piles of laminates, where 5. D. Abraham, S. Matthews, and R. Mcllhagger, Compos.
the shapes of voids were arbitrary and most of them existed Pt. A-Appl. Sci. Manuf., 29, 795 (1998).
in the forms of long-strip. Upon the application of autoclave 6. T. S. Lundström and B. R. Gebart, Polym. Compos., 15, 25
pressure, the pressure gradient was initially formed along the (1994).
thickness direction of laminates. Most of load in the higher 7. A. Singh, C. B. Saunders, J. W. Barnard, and V. J. Lopata,
pressure region could be forced the migration of resin as Radiat. Phys. Chem., 48, 153 (1996).
well as voids. The number and size of voids decreased 8. P. Olivier, J. P. Cottu, and B. Ferret, Composites, 26, 509
significantly as the increase of pressure, the impregnation (1995).
between fibers and resin was improved and the degree of 9. L. Liu, B. M. Zhang, D. F. Wang, and Z. J. Wu, Compos.
fiber compactness was increased. Struct., 73, 303 (2006).
As the pressure increased from 0.0 MPa to 0.4 MPa, the 10. S. J. Li, L. H. Zhan, R. Chen, and W. F. Peng, Fiber.
void content decreased from 8.04 % to 0.51 % and the ILSS Polym., 15, 2404 (2014).
increased by 64.55 %, while the IFSS increased by 23.08 % 11. M. Zidi, L. Carpentier, A. Chateauminois, and F. Sidoroff,
with the fiber volume fraction increased from 68.6 % to Compos. Sci. Technol., 60, 429 (2000).
80.92 %. For void contents above 1 %, the ILSS increased 12. M. Rodríguez, J. M. Molina-Aldareguía, C. González, and
progressively with the increase of the pressure, and the J. Llorca, Compos. Sci. Technol., 72, 1924 (2012).
failure cracks of ILSS appeared between the plies induced 13. S. Amico and C. Lekakou, Compos. Pt. A-Appl. Sci.
by voids. But when the pressure increased from 0.4 MPa to Manuf., 31, 1331 (2000).
0.6 MPa, the void contents were less than 1 %, there was 14. J. L. Kardos, M. P. Dudukovic, and R. Dave, Adv. Polym.
only increased by 1.69 % of ILSS and the sensitivity of ILSS Sci., 80, 101 (1986).
on void contents was much smaller. The relationship between 15. S. Hernández, F. Sket, J. M. Molina-Aldareguía, and C.
the pressure and the IFSS was in a good agreement with the González, Compos. Sci. Technol., 71, 1331 (2011).
polynomial increasing model, which was consistent with the 16. K. Bowles and S. Frimpong, J. Compos. Mater., 26, 1487
influence of the pressure on the fiber volume fraction. When (1992).
the pressure was reached to 0.6 MPa, the growth range of 17. C. Medinam, J. M. Molina-Aldareguía, C. González, and
IFSS was increased markedly further. M. F. Melendrez, J. Compos. Mater., 50, 1651 (2015).