Original article

Journal of Reinforced Plastics and

Composites
Fatigue damage analysis of 0(0) 1–19
! The Author(s) 2018
fiber-reinforced polymer composites— Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
A review DOI: 10.1177/0731684418754713
journals.sagepub.com/home/jrp

Md. Touhid Alam Ansari, Kalyan Kumar Singh and

Mohammad Sikandar Azam

Abstract
Fiber-reinforced polymer composites are becoming suitable and substantial materials in the repair and replacement of
conventional metallic materials because of their high strength and stiffness. These composites undergo various types of
static and fatigue loads during service. One of the major tests that conventional and composite materials have to
experience is fatigue test. It refers to the testing for the cyclic behavior of materials. Composite materials are different
from metals, as they indicate a distinct behavior under fatigue loading. The fatigue damage and failure mechanisms are
more intricate in composite materials than in metals in which a crack initiates and propagates up to fracture.
In composite materials, several micro-cracks initiate at the primary stage of the fatigue growth, resulting in the initiation
of various types of fatigue damage. Fiber volume fraction is an important parameter to describe a composite laminate.
The fatigue strength increases with the increase of the fiber volume fraction to a certain level and then decreases
because of the lack of enough resin to grip the fibers. The fatigue behavior of fiber-reinforced polymer composites
depends on various factors, e.g., constituent materials, manufacturing process, hysteresis heating, fiber orientation, type
of loading, interface properties, frequency, mean stress, environment. This review paper explores the effects of various
parameters like fiber type, fiber orientation, fiber volume fraction, etc. on the fatigue behavior of fiber-reinforced
polymer composites.

Keywords
Fiber, polymer, fatigue life, fiber volume fraction, fiber orientation

Introduction
are affected by matrix cracking, fiber breakage, debond
Fiber-reinforced polymer (FRP) composite materials between fiber and matrix, and delamination between
are broadly used in structural applications such as the layers in the fatigue loading.
aerospace, wind turbine blade, marine industries, auto- An FRP composite is described as a polymer that is
mation, etc. due to their high specific strength, specific reinforced by a fiber. They are multi-phase materials
stiffness, and tailorability. They are fast replacing con- manufactured by adding polymer matrix–resins with
ventional metal alloys in various industries. FRP is fillers and reinforcing fibers (usually carbon, glass, or
anisotropic and inhomogeneous in nature. The aramid) to produce a bulk material with better proper-
mechanical behavior of FRP materials is different ties than those of the individual base materials. Filler
from metals as it depends on the properties of its con- materials (e.g., calcium carbonate, hydrated alumina
stituent elements. It mainly depends on the matrix
strength, fiber strength and modulus, chemical stabili-
ty, and the interface bonding between the fiber/matrix Department of Mechanical Engineering, Indian Institute of Technology
to allow stress transfer.1 Many structures are often (ISM), Dhanbad, India
subjected to vibration and other fluctuating loads,
Corresponding author:
resulting in the fatigue degradation of the material. Md. Touhid Alam Ansari, Department of Mechanical Engineering, Indian
Fatigue is a major cause for the failure of the structural Institute of Technology (ISM), Dhanbad, Jharkhand 826004, India.
components. Mechanical properties of the composites Email: touhidalam90@gmail.com
2 Journal of Reinforced Plastics and Composites 0(0)

or clay) may also be mixed to enhance specific proper- and was also described by Branco et al.15 Kadi and
ties of the composite or to lower its cost. The resins Ellyin42 have showed a fatigue failure criterion based
used in the matrix of FRPs can be classified into ther- on the input strain energy for fiber-reinforced materi-
mosetting and thermoplastics. Epoxies, polyesters, and als. A number of studies on the fatigue behavior of
vinyl-esters are the thermosetting matrices, which are composite laminates have been done with a commonly
used for composite structures and these all are crossed used stress ratio range from 1 to 1.42,55,56
linked.2 Epoxy resins are used commonly as the matrix Mortazavian and Fatemi56 observed the effect of
material in FRP. These resins are made under the influ- mean stress for the notched specimens. Based on the
ence of heat, and after formation, they do not soften or stress ratio, the fatigue loads can also be classified as
melt upon reheating and do not dissolve in solvents. On Tension–Tension (T-T) (0  R  1), Tension–
the other hand, polypropylene, polyurethane, polyeth- Compression (T-C) (1 < R < 0), and Compression–
ylene, and polyvinyl chloride are the thermoplastic Compression (C-C) (1 < R < 1) type fatigue. Due to
matrices. They have not any significant chemical casual and/or long exposure to fatigue loads, many
changes during the production and can be comfortably types of internal damages such as matrix cracks, dis-
regained to their original monomer except when they bands, and delamination develop in the composites.16–
18
are defective. However, at present, thermosetting resins In the presence of such damages, the static load-
are used extensively in FRP industries as they permit carrying ability of the material reduces. Further, these
rapid production and have better properties of impreg- damages may also propagate sub-critically under cyclic
nation and adhesion to the fibers and better mechanical fatigue loads.
performance. The presence of small amounts of single- Components made of FRP composites may also
walled carbon nanotubes (SWCNTs) and multi-walled contain geometric discontinuities such as sharp corners
carbon nanotubes (MWCNTs) has been observed or holes that can increase the local intensity of stress
to improve the matrix-dominated interlaminar and strain. Hence, these factors are of specific interest
shear strength, tensile strength, and modulus of com- in the stability design and analysis of FRP structures.
posites.3–5 Many studies have confirmed that a consid- The effect of stress concentration on the fatigue in
erable increase in the interfacial shear strength could composites is quite different from metals. Curtis19
be obtained by the addition of CNTs.6–7 Carbon nano- observed a significant effect of stress concentration
fibers (CNFs) as a filler in the polymer has been (produced by holes) in case of short fatigue lives.
observed to improve several mechanical properties, Hyakutake et al.20 presented a slight effect of the
e.g., the incorporation of 2 wt.% of CNFs in a notch-root radius on the fatigue life of a glass fiber/
carbon fiber composite increases the tensile and flexural epoxy laminate. Frequency effects have been observed
strengths.8 The compressive strength of composite lam- during the fatigue of many engineering materials. The
inates is improved by CNFs.9–10 magnitude of the test frequency effects the internal
Some of the common methods to make laminated heating of the composite in fatigue. As the frequency
products from fiber fabrics are hand layup, resin trans- of the fatigue load increases, internal heating of the
fer molding, and vacuum-assisted resin transfer mold- composite increases and fatigue life of composite
ing.11 Fiber is an important constituent in FRP decreases. The temperature rise of the specimens was
composites. Many of the mechanical properties such observed during the fatigue tests and increased with the
as stiffness, strength, and toughness are decided by number of fatigue cycles, especially close to final
the fiber volume fraction (Vf ) of the composite. The failure.21
mechanical properties of FRP can be improved with
the increase of the fiber volume fraction.12
Frequently, fiber fractions of 30–70% have been
Fiber and matrix
taken in many previous studies.13,38 Most of the com- Fibers are the main load bearers in the composites. The
posite products are in laminated forms (except whiskers fatigue behavior is affected by the type of fibers used in
and particulate). Laminate orientation together with the composites. The fiber generally occupies 30%–70%
their thickness play an important role in deciding the of the matrix volume in the composites. The fibers can
properties of the composite. Cross-ply, angle-ply, and be chopped, stitched, woven, and braided. Many
unidirectional laminates with different fiber orienta- researchers studied the effect of fiber properties on
tions have been studied previously. Commonly, the advanced polymer matrix composite. The most
angles of 0
, 45
, and 90
have been taken by many common types of fibers used in the advanced compo-
researchers in their research.13,38 The stress ratio (R) sites for structural applications are the glass fiber,
is also an important fatigue parameter. In general, an carbon fiber, and aramid fiber. The glass fiber is the
increase in the stress ratio decreases the fatigue strength least expensive and carbon being the most expensive.
of the composite. This effect was presented by Owen14 The cost of aramid fibers is about the same as the lower
Ansari et al. 3

grades of the carbon fiber. Some classic glass fibers are life. Also, tougher resins have higher interlaminar frac-
C-glass, E-glass, ECR-glass, and S-glass fibers. Boller22 ture toughness which will contribute to an increased
studied the comparative fatigue performance of S-glass fatigue resistance against delamination.30 Boller22 stud-
and E-glass fiber-reinforced composites for same matri- ied the comparative fatigue behavior of laminates of
ces and concluded that S-glass composites have a better polyester, epoxy, phenolic, silicone, and matrices rein-
fatigue performance than E-glass composites. S-glass forced with E-glass fabric and found that epoxy matrix
fibers have higher strength and stiffness in comparison composites have a better fatigue behavior, as shown in
to the other glass fibers. Under an corrosive environ- Figure 3. The fiber–matrix interface affected the frac-
ment, these glass fibers react differently.23 The effect of ture toughness of FRPs. The weaker interface prevents
the fiber type on the fatigue behavior of carbon FRPs the cracks to propagate through the matrix and hence
(CFRPs) composites was studied by Jones et al.24 It enhances the fracture toughness and reduces the effec-
was concluded that the S-N curve for the low modulus tiveness of the stress transfer.31 The mixing of rubber
CFRP is steeper as shown in Figure 1. The high mod- particles in the resin can also increase the resistance to
ulus CFRP has a better fatigue performance than the crack growth in the matrix material. The mixing of
low modulus CFRP. T300, AS4, and IM6 are some nanoparticles in the matrix material can considerably
types of carbon fibers. Plots of stress versus log cycles improve tensile strength and impact strength and
to failure for three commonly used FRP composites are fatigue life of the composite.32 Table 1 presents a con-
shown in Figure 2. It was found that the S-N curve of solidated view of fibers and matrices with different fiber
glass FRPs (GFRPs) indicated more severe fall in its
fatigue strength than that of CFRPs. The use of carbon
fiber confines the strain in the composite and thereby
limits vast distortion in the matrix. On the contrary, the
use of glass fiber permits for large distortion in the
matrix giving rise to the fatigue failure. The fatigue
behavior of Kevlar FRPs (KFRPs) is more complex
than that of CFRPs and GFRPs, due to which
Kevlar fiber is fatigue sensitive.25
The resin is another important constituent in the
composite. The fatigue strength of fiber-reinforced
composites is significantly influenced by the properties
of the resin.26 Crack initiation generally starts in the
matrix region in the form of a fatigue damage. Many
researchers have shown that the thermoplastic resins
have better properties than the thermosetting resins
with regard to toughness and ductility.27–29 Therefore, Figure 2. S-N fatigue curves for unidirectional composite
thermoplastic resins exhibit significantly longer fatigue materials.25

Figure 1. Normalized S-N curves for different unidirectional CFRPs.24

4 Journal of Reinforced Plastics and Composites 0(0)

volume fractions used by different scholars in their epoxies. Jen and Wang34 examined the influence of
works. CNTs on the fatigue behavior of the MWCNT/epoxy
Knoll et al.33 studied the effect of carbon nanopar- nanocomposite with and without central hole on speci-
ticles on the fatigue behavior of carbon fiber-reinforced men. Newaz35 studied the fatigue behavior of unidirec-
tional glass fiber-reinforced epoxy and vinyl ester
composites. Unidirectional specimens were made with
48% and 50% volume of fiber for E-glass/vinyl
ester and E-glass/epoxy composites, respectively.
Unidirectional E-glass/vinyl ester and E-glass/epoxy
specimens were showed similar fatigue damage mecha-
nisms at high as well as at low deflection levels under
flexural loading. However, E-glass/vinyl ester specimen
early attained the similar damage stage as in the
E-glass/epoxy specimen due to more rapid accumula-
tion of multiple cracking. Manjunatha et al.36 studied
the tensile fatigue behavior of both the neat epoxy-
polymer and the hybrid epoxy polymer with 9 wt.%
of carboxyl-terminated butadiene-acrylonitrile
(CTBN) rubber microparticles and 10 wt.% of silica
Figure 3. S-N curves for E-glass fiber composites with different nanoparticles at a stress ratio of 0.1 and frequency of
matrices.22 1 Hz to 3 Hz. The test frequency was maintained below

Table 1. Summary of materials used.

S. no. Author Matrix Fiber Fiber vol/wt.%

1 Owen and Howe26 Polyster Glass 35

2 Jones et al.24 Epoxy Carbon and Glass 58, 64
3 Newaz35 Epoxy, vinyl ester Glass 48, 46
4 Ahlborn64 Epoxy Carbon 60
5 Kadi and Ellyin42 Epoxy Glass –
6 Abd Allah et al.39 Polyster Glass 15.8, 31.8, 44.7
7 Caprino and D’amore27 Polyproplene Glass 20
8 Soden et al.72 Epoxy Glass 60
9 Ganstedt and Talreja52 PEEK Carbon –
10 Bureau and Denault29 Polypropylene Glass 60
11 Shan and Liao68 Epoxy Carbon and Glass 30
12 Barron et al.60 Epoxy Carbon –
13 Kawai et al.43 Epoxy Carbon –
14 Mallick and Zhou57 Polyamide Glass 33
15 Nasr et al.55 Polyster Glass 55 to 65
16 Hojo et al.30 Epoxy Carbon –
17 Kumagai et al.63 Epoxy Glass 47
18 Zhung and Wightman 31 Epoxy Carbon –
19 Bernasconi et al.44 Polyamide Glass 30
20 Nagalingam et al.32 Polyster Carbon and Glass 30
21 Manjunatha et al.36 Epoxy Glass 57
22 Bizeul et al.50 Epoxy Glass 50
23 Huh et al.45 Epoxy Glass –
24 Mini et al.40 Epoxy Glass 34.4, 38.6, 48.9, 57
25 Nixon-Pearson et al.49 Epoxy Glass –
26 Fatemi et al.59 PBT, PA6 Glass 30, 35
27 Brunbauer et al.13; Brunbauer and Pinter38 Epoxy Carbon 30, 55
28 Eftekhari and Fatemi61 Polypropylene Glass 30
29 Mortazavian and Fatemi56 PBT, PA6 Glass 30, 35
30 Belmonte et al.41,53 Polyamide Glass 35
PBT: polybutylene terephthalate; PA: polyamide.
Ansari et al. 5

3 Hz to prevent the thermal effects leading to reduced their corresponding places. So, the matrix also plays an
fatigue life. GFRP made of these two polymers were important role in the load-carrying capacity. The ideal
also separately investigated for the fatigue behavior. value for the fiber volume fraction is 45–60%. In addi-
The cyclic fatigue life of a hybrid epoxy polymer/ tion, exceeding the fiber volume about 60–70%,
hybrid GRPF composite was about six to ten times strength of composite will reach a peak and then start
higher than that of the neat-epoxy polymer/GFRP– to reduce because of the lack of sufficient resin to grip
neat epoxy matrix composite. The fatigue life was the fibers together perfectly. A number of researchers
increased because of reduced crack growth rates in have contributed with different fiber volume fractions
hybrid epoxy polymers. This reduced cracking in the as shown in Table 1.
hybrid epoxy matrix which leads to lower degradation Brunbauer and Pinter38 performed a relative analy-
of GFRP with the hybrid matrix under fatigue loading sis of carbon/epoxy composite with fiber volume frac-
when compared to the neat matrix. tion of 30% and 55% having different fiber
Wu et al.37 investigated the fatigue performance of orientations of 0
, 45
, and 90
. In TTF test, an
various FRP and hybrid FRP composites. The lami- increase in fiber volume fraction from 30 to 55%
nates, used in this study, were prepared using carbon, volume content results in an increase in the nominal
glass, basalt, and polyparaphenylenl benzobisoxazole stress of 40% in 90
unidirectional laminate (Fig 4).
(PBO) fiber with epoxy matrix. Hybrid composite This effect was observed as the dominant damage
specimens were prepared with different layers of two mode in low fiber fraction composites is matrix crack-
different materials. For example, carbon/basalt and ing and fiber-matrix debonding. On the contrary, the
carbon/glass FRP layers were taken in the ratio of damage mode in high fiber fraction composites was
1:1, i.e., one layer of carbon was combined with one fiber pullout. However, in the T-C test, there was no
layer of basalt or glass with 50% nominal fiber volume significant change in the results for both the samples
fraction. Wu et al. presented in their study that CFRP with 30% and 55% fiber content. In 45
Unidirectional
and PBOFRP composites had better fatigue perfor- laminate, higher fiber volume fraction resulted in
mance compared to the GFRP and BFRP composite higher fatigue strength. In 0
Unidirectional laminate
laminates. The CFRP/BFRP hybrid composite had T-T loading gives rise to fiber pullout fracture or single
comparatively better fatigue resistance compared to fiber breakage. However, in T-C loading, fiber bundle
the BFRP composite. TTF tests were carried out with breakage along the fractured plane was observed. Abd
run out cycles of 2  106 cycle. Maximum fatigue limit Allah et al.39 studied the fatigue behavior of GRP rods
of 83.7 and 76.7% of the ultimate strength were at zero mean stress. Specimens were prepared by a
showed by CFRP and PBOFRP composite, respective- modified pultrusion method with different fiber
ly. For the GFRP and BFRP composites, the equiva- volume fractions (vf =0, 0.158, 0.318 and 0.447) in
lent limit was at 61.3 and 55%. In case of the hybrid the form of rod having diameter 7.9 mm and length
composite, CBFRP composite had a higher limit of 300 mm. Frequency tests were conducted at a frequen-
73.6% when compared to C-GFRP hybrid composite, cy of 25 Hz. Stiffness of specimens decreases rapidly
which had a fatigue limit of only 58% of the ultimate and reached approximately to one-fourth of the initial
strength. specimen stiffness at initial phase of the fatigue test. It
may be due to the reason such as cutting of fiber at the
surface of the specimen during the machining, presence
Fiber volume fractions of shear crack at waist of the sample and surface flaws
The volume fraction is the quantitative expression of in glass fibers. Mini et al.40 studied the fatigue behavior
the fiber present in a composite material. An important of glass/epoxy composite together with different fiber
parameter to characterize a composite laminate is the volume fractions of range 34% to 57%. The test was
fiber volume fraction. Fiber volume fraction is carried at loading frequencies between 7 and 9 Hz and
expressed in terms of the relative proportion of the stress ratio ðRÞ ¼ –1, that is, completely reversible in
fibers, and the matrix is present in the laminate. Fiber nature. It was observed that composite with low fiber
volume fraction affects approximately all properties in volume fraction lost its stiffness slowly owning to
a given material system, e.g., the static and fatigue matrix cracking and attained a constant value upon
strengths. Generally, stiffness, strength, and the load- increase in loading cycles. However, rapid decrement
carrying capacity of a laminate will be higher with was occurred in stiffness of composite having higher
higher fiber content. However, when the fiber volume fiber volume fraction. This is because a small quantity
fraction (Vf) is adequately high, the ultimate strength of matrix is present in the composite to avoid shearing
of the composite decreases. The amount of matrix of the fibers. Hence, fiber cracking is responsible for the
decreases as the fiber content increases. When load is reduction of stiffness. Belmonte et al.41 studied the
applied on the composites, the matrices grip fibers in effect of the fiber volume fraction on damage
6 Journal of Reinforced Plastics and Composites 0(0)

Figure 4. Comparison of fatigue life with two different fiber volume fraction.38

Figure 6. Tensile coupon specimens ruptured due to fatigue at

room temperature.43

Figure 5. Variation of monotonic strength with fiber orienta-

tion angle.42 Fiber orientation
Fiber orientation with respect to loading direction is
mechanisms in a notched, short glass, fiber-reinforced one of the most important parameters affecting
polyamide (PA66) specimens under fatigue loading mechanical properties of fiber-reinforced composites.
with various fiber volumes (0%, 15%, 25%, 35%, Kadi and Ellyin42 studied the fatigue behavior of uni-
and 50%). Field Emission Scanning Electron directional glass fiber/epoxy composite laminates. T-T
Microscopy (FESEM) was used for damage analysis. and T-C fatigue tests were carried out with fiber orien-
The fiber orientation at crack initiation is affected by tations, 0
,19
, 45
, 71
and 90
at positive and negative
the insert used for generating the notch in the sample. stress ratios of R =0.5, 0, and 1. Experimentally cal-
The crack path is strongly affected by the fiber volume culated strength was plotted against fiber orientation as
fraction. With high fiber contents, fibers tend to shown in Figure 5. It shows that the strength depends
agglomerate forming clusters deviating the crack on the fiber orientation of laminate. The strength is
plane more effectively than isolated fibers thereby decreased with the increase of fiber orientation.
reducing the crack growth rate. An increase in the Kawai et al.43 studied the effect of fiber orientation
fiber volume enhanced both the fatigue strength to on the fatigue behavior of unidirectional T800H/
crack initiation and to the failure of the composite Epoxy composite. Laminates with nine different fiber
materials. The relative contribution of the crack prop- orientation angles having 64% fiber volume fraction
agation to the total lifetime increases with increasing were used as a sample. TTF tests were conducted up
the fiber volume fraction. to 106 cycles at R=0.1 and frequency of 10 Hz. S-N
Ansari et al. 7

Figure 7. Dimensions of the injection molded plates and definition of the orientation angle (h) of samples.44

Figure 9. S-N curves for composite with different laminates at a

stress ratio of 0.1.45
Figure 8. S-N curves for different sample orientations.44
orientations. Slope of S-N curve decreased with the
curve for different fiber orientations is shown in Figure increase in the number of cycles. Huh et al.45 studied
17(a). As the fiber orientation angle increases the ten- the fatigue properties of GFRP with three laminate
sile fatigue strength reduces. The macroscopic failure orientation of 0
, 45
, and 0
45
at three different
morphology of the unidirectional T800H/Epoxy is sub- stress ratios (R) of 0.2, 0.1, and 0.5. It was concluded
jected to tensile fatigue as shown in Figure 6. for h ¼ 0
that the stiffness of the composite was decreased as the
in the beginning parallel splitting arises and they pro- angle of fiber increases from 0
to 90
. This dependency
longed to end tap section and finally pulled out. For the of the fiber can influence the fatigue behavior of the
other orientation angles ð0
< h > 90
Þ, the final sep- composites. Figure 9 shows the fatigue life curves (S-N)
aration happens in a single cross-section parallel to for three different composites used in this study.
fiber and thickness direction. Bernasconi et al.44 stud- Table 2 presents a consolidated view of fiber orienta-
ied the influence of fiber orientation on the fatigue tions used by different scholars in their works.
strength of a short glass fiber-reinforced polyamide-6
composite. T-T axial fatigue tests were carried out with
samples having orientations 0
, 30
, 60
and 90
, Damage mechanism
extracted from injection molded plates. Samples were Damage mechanism in an FRP composite is a complex
taken out from the plates by water jet cutting with process to describe. Reifsnider et al.46 identified three
different orientations with respect to the plate axis, stages of fatigue life. World wide Failure Excercise47
indicated by the angle h as shown in Figure 7. started to study on most effective damage mechanism
Reduction in the fatigue strength was observed with theory for the FRP composites. Major damage mech-
the increase in the orientation angle of the samples. anisms in the fatigue of fiber-based composites are
Figure 8 shows the fatigue life of sample for different matrix cracking, fiber-matrix debonding, fiber
8 Journal of Reinforced Plastics and Composites 0(0)

Table 2. Summary of laminate orientations used.

S. no. Author UD/BD Orientation

1 Kadi and Ellyin42 UD 0 , 19 , 45 , 71 , 90
2 Shan and Liao68 UD –









3 Kawai et al.43 UD 0 , 5 10 15h , 30 , 45 , 60 , 90




4 Barron et al.60 UD, BD 0 , 45h , 90 0






5 Nasr et al.55 BD 45 , 90

0





44
6 Bernasconi et al. – 0, 30 , 60 , 90




7 Huh et al.45 UD, BD 0, 45 , 045





8 Brunbauer et al.13; Brunbauer and Pinter38 UD 0, 45 , 90



9 Mortazavian and Fatemi56 – 0, 18 , 45
, 90




10 Fatemi et al.59 – 0, 18 , 45
, 90

Gamstedt and Talreja52 studied the microscopic and

macroscopic fatigue behavior for unidirectional CF/
epoxy and CF/PEEK loaded in T-T along the fiber
direction. Small localized fiber-bridged cracks propa-
gated in CF/epoxy, while CF/PEEK showed more
extensive and dispersed damage with fiber breakage
induced by debonding or matrix cracking. Due to the
relative progressiveness of damage and brittleness of
the two materials, CF/PEEK has a steeper slope in
the fatigue life diagram than that of CF/epoxy as
shown in Figure 11. Belmonte et al.53 presented a char-
acter of damage initiation and propagation of short
glass fiber-reinforced polyamide (PA66-GF35) for the
plain and notched sample. Electron microscopy was
Figure 10. Three stages of damage mechanism in unidirectional used to investigate the fracture surface of the failed
composite.46 samples, and the crack path of sample undergo inter-
rupted fatigue tests. Plain samples fail due to unstable
fatigue crack propagation (FCP). In plain specimens,
breakage, delamination between fiber layers, and frac- micro-cracks were observed at the end of fibers associ-
ture as an entire component. Based on the approach ated to the loading direction. Failure of notched sam-
given by Vasiukov et al.,48 damage mechanism can be ples proceeds in three stages: (1) Crack initiation; (2)
considered as a nonlinear function of a series of cycles. Stable crack propagation; and (3) Unstable crack prop-
Damage propagation is rapid at the beginning and at agation. The fracture surface due to stable FCP is stress
the end of the cyclic loading if no stress concentrations whitened and smoother than fracture surface due to
are present in the sample.49,50 However, damage unstable FCP (Figure 12(a)). The fracture surface mor-
growth is linear in nature, in the middle of the phology varies according to the crack growth mode
stage.51 The magnitude of damage influenced by the (stable/unstable FCP). No sign of ductile matrix defor-
property of each lamina, interface property of the lam- mation was detected near the notch tip (Figure 12(b)).
inate and stacking sequence. Generally, there are three Matrix fracture behavior is micro-ductile on the frac-
stages of failure. Figure 10 shows damage initiation ture surface due to stable FCP (Figure 12(c)) and brittle
and growth process during the fatigue life of unidirec- on the fracture surface due to unstable FCP (Figure 12
tional composite materials. Stage I initiates by debond- (d)). Consequently, unstable FCP goes towards brittle
ing of the matrix-fiber interface of the matrix material fracture surface irrespective of the sample geometry.
at a local micro-defect like resin-rich regions, mis- Zhao et al.54 studied the fatigue behavior of the
aligned fibers or voids formed during the manufactur- basalt fiber-reinforced epoxy polymer composites and
ing. Stage II starts by the initiation and propagation of showed the degradation mechanism of BFRP at a dif-
delamination and almost linear in nature. Stage III is ferent stress level of cyclic loading. T-T fatigue tests
the last stage of damage growth. In this stage, fiber were carried out up to 107 cycles. Scanning electron
breakage occurs shortly after the damage has been microscopy technique was used for investigation. It
accrued during stages I and II. was observed that BFRP composites experienced
Ansari et al. 9

Figure 11. Fatigue life diagrams of CF/epoxy and CF/PEEK.52

Figure 12. Fracture surface morphology of a notched specimen; (a) failed specimen “s” and “u” indicate the fracture surface caused
by stable and unstable FCP respectively; (b) fracture surface at crack initiation; (c) fracture surface at the end of the stable FCP;
(d) fracture surface caused by unstable FCP.53

transverse matrix cracking during T-T fatigue loading patterns arose during fatigue loading. Fracture dia-
contrary to longitudinal matrix cracks occurred during gram for BFRP composites under fatigue loading is
same loading condition in unidirectional CFRP. The shown in Figure 13. In region 1, fracture of the samples
SEM images showed that three types of damage took place in the trend of progressive fiber breaking.
10 Journal of Reinforced Plastics and Composites 0(0)

Figure 13. Schematic drawing of the S-N curve of the BFRPs with damage patterns.54

Within the first a few cycles, small matrix crack and epoxy specimens subjected to off-axis tensile loading
weakest fiber breaks were occurred due to stress con- with different stress ratios. The results showed that
centration and fiber strength. In region II, stress inten- for off-axis loading cases, the slope of the S-logN line
sity was not caused progressive fiber breaking because for R = 0.5 is considerably smaller than that for R =
of low stress level. Matrix cracks propagated trans- 0. That is, the rate of fatigue strength degradation
versely,either along the fiber or fiber-matrix interface decreases when the stress ratio (R) increases.
which led to progressive interface debonding. Fiber Nasr et al.55 studied the effect of torsional mean
breakage due to wearing decreased the fiber strength stress on the fatigue behavior of GFRPs. This study
and instigated damage growth; which eventually was done by testing samples with two fiber orientations



caused total fracture under fatigue loading. In region [45
]2s and [0 , 90 ]2s under five negative stress ratios
III, the residual strength was increased or recovered to (R) of range between 1 and 0 with increment of 0.25.
its original value. The discontinuous fiber/matrix inter- Samples with volume fraction ranging from 55% to
face debonding and fiber strengthening during fatigue 65% were tested under the same loading conditions.
were taken place in this region. Due to which fiber Strength of the sample was only slightly affected and
might be exhibited its intrinsic strength. no failure mechanisms were noticed. In other words,
samples with different volume fractions within the
used range did not cause any change in failure mecha-
Mean stress and stress ratio nisms; they only affected the strength of the specimen
Mean stress is the average of maximum stress ðrmax Þ to a limited degree. It was found that at all stress ratios,
and minimum stress ðrmin Þ induced in the component due to the local stress component, samples with
and stress ratio (R) is the ratio of the minimum stress [45
]2s fiber orientation have higher fatigue strength



ðrmin Þ to the maximum stress ðrmax Þ. The stress ratio than the sample with [0 , 90 ]2s fiber orientation. The



(R) and mean stress rmean are ideally applied to explain [0 , 90 ]2s samples were subjected to pure local shear
fatigue loading. The stress ratio (R) decides the fatigue stress while [45
]2s samples were subjected to T-C
loading nature, with 1 < R < 1 representing that the local stress component. Mortazavian and Fatemi56
cycle lies in the C-C region, 1 < R < 0 represents studied the effect of mean stress on the fatigue behavior
that the cycle lies in the T-C region and 0  R  1 of two short glass fiber-reinforced thermoplastic com-
represents that the cycle lies in the T-T region. Kadi posites, polybutylene terephthalate (PBT) and
and Ellyin42 studied the effect of the stress ratio (R) on polyamide-6 (PA6). Fatigue tests were carried out at
the fatigue behavior of unidirectional laminates. A three different stress ratios R of 1, 0.1, and 0.3, in
number of cyclic tests were conducted with different different fiber orientations. Figure 14 shows the effect
fiber orientation angles and under various values of of stress ratio on samples. A significant reduction of
the stress ratio of R ¼ l, 0, 0.5. They used glass– fatigue strength was noticed under R ¼ 0.1 compared
Ansari et al. 11

Figure 14. Effect of R ratio or mean stress on the fatigue behavior of (a) PBT and (b) PA6 in the longitudinal and transverse directions
at 23
C.56

Figure 15. S-N plots (a) rmax vs. N and (b) ra vs. N for short E-glass fiber-reinforced polyamide.57

with R ¼ 1, in both longitudinal and transverse direc- GF/PA-6,6 composite are shown in Figure 15. As the
tions. However, the fatigue strength reduction from stress ratio increases, the fatigue strength increased
R ¼ 0.1 to R ¼ 0.3 was not as considerable as compared while the stress amplitude decreased. It was also
to from R ¼ 1 to R ¼ 0.1. This could be due to large observed that the fatigue life of the composite is signif-
stress ratio(R) difference in 1 to 0.1 range compared icantly affected by the mean stress. Thus, the fatigue
to 0.1 to 0.3 range. The mean stress effect was more strength of the composite decreased with increasing
pronounced in the low cycle fatigue (LCF) reign com- mean stress ratio.
pared with the high cycle fatigue (HCF) reign, due to
the increased dependency of fatigue life to the maxi-
mum stress in this life reign. Mallick and Zhou57 stud- Frequency and hysteresis effect
ied the effect of mean stress on the high cycle fatigue Testing frequency or loading frequency has a signifi-
strength of short E-glass fiber-reinforced polyamide-6,6 cant effect on the fatigue life of FRP composites.
composite. TTF tests were performed at different mean Strain rate is used somewhere instead of loading fre-
stresses to analyze the influence of mean stress on the quency. Hysteresis heating or self-heating of the speci-
fatigue strength of the composite. The S-N curves of men is one of the most important effects of loading
12 Journal of Reinforced Plastics and Composites 0(0)

frequency. Since the nature of the polymers is viscoelas- tests were carried out at different stress ratios (R) of
tic, glass transition temperature (Tg) is achieved by 0.1, 0.3, and 1.0 and different temperatures (T) of 23,


matrix when composites are tested at high frequency. 85 and 120 C. At higher frequency, the beneficial effect
At glass transition temperature, the polymer starts to of increasing frequency was noticed for certain material
transform into soft rubbery material from the glassy before self-heating becoming effective at higher fre-
brittle material, altering the properties of the polymer quencies. The maximum temperature rise during the



composite. The effect of a frequency other than the test was 10 C for the test frequency of 4 Hz and 2 C
self-heating effect on the fatigue behavior of the poly- for the test frequency of 2 Hz. A decrement of fatigue
mer is known as the beneficial or strengthening effect of life was noticed at a frequency of 4 Hz as compared to
frequency. Ferreira et al.58 studied the effect of fre- 2 Hz. The beneficial or strengthening effect of increased
quency on the fatigue behavior of polypropylene/ frequency on the fatigue behavior directly linked to the
glass fiber thermoplastic composites. Samples were pre- polymer matrix. Due to the viscoelasticity nature of the
pared with the fiber volume fraction of 0.338. Fatigue matrix, the damping decreases with the increase in fre-
tests were performed at stress ratio R=0,0.25 and fre- quency. Therefore, fatigue life will increase. The area of
quencies of 2 Hz and 10 Hz. The influence of frequency hysteresis loops can be related to the damping. If the
on fatigue strength of unnotched sample was negligible effect of self-heating was negligible on fatigue life, the
within the range 2–10 Hz. area of hysteresis loop decreased with increasing the
Barron et al.60 studied the influence of loading fre- frequencies, and also all samples showed enhanced
quency on the fatigue behavior of a carbon fiber/epoxy fatigue life with increasing frequency at a constant
matrix composite. TTF tests were carried out at differ- stress level. The fatigue behavior was observed to be
ent stress levels and frequencies (5, 10, and 20 Hz) for more associated with the time instead of cycles. Flore


three specimen



orientations (unidirectional, (0 )8 cross- et al.62 studied the fatigue behavior of GFRPs in the
ply 0 =90 4s and angle-ply (45
)4s). Brief results are very high cycle fatigue (VHCF) regime. Ultrasonic
shown in Table 3. Results indicate that frequency sig- fatigue testing (UFT), which decreases the testing
nificantly influence the fatigue behavior of CFRP com- time in comparison to conventional methods by a
posite. Fatigue results for the unidirectional laminates factor of 100, has been applied for cyclic tension testing
were indeterminate due to a large amount of scatter in of GFRP. UFT is used to find the fatigue strength of
S-N curve. However, both cross-ply and angle-ply lam- GFRP in the range of 106–109 cycles. The results of the
inates reveal a clear and distinct dependency of dynam- UF-tests are equated with the results of conventional
ic mechanical response. For cross-ply laminates, it was fatigue tests. It is found that the fatigue strength agrees
observed that an initial increase in loading frequency very well. The comparability between both tests is
from 5 to 10 Hz consequently led to a reduction of given since stiffness loss and plasticity is only a little
fatigue resistance. Conversely, on further increasing pronounced in the fiber direction. The fatigue limit
the loading frequency from 10 to 20 Hz, enhancement cannot be determined due to the lack of data beyond
in fatigue life of the sample was observed. Angle-ply 109 cycles. The S-N curve becomes flattered in the sec-
laminates, due to their matrix-dominated nature show tion of 108–1010fatigue cycles.
the significant dependency on test frequency. An initial
increase in test frequency from 5 to 10 Hz resulted in
enhancement of fatigue resistance of the sample. A fur-
Environmental effect
ther increase in test frequency caused a decrease in the The environmental condition of composite materials
sample fatigue resistance. has a vital role to play in the performance and the
Eftekhari and Fatemi61 studied the effect of cycling durability of composite parts. In their practical use,
frequency on the fatigue behavior of neat, talc filled, composite products are generally exposed to a range
and short GFRP composites. Samples were prepared of environmental factor, e.g., chemical environments,
with neat polypropylene (PP), 30% glass fiber with PP UV radiation, hygrothermal, moisture, etc. This causes
(PP-G), and 40 wt.% talc filled with PP (PP-T). Fatigue degradation in material properties of FRP laminates.

Table 3. Summary of effects of frequency on fatigue behavior of carbon fiber-reinforced polymer composites.60

Composite Unidirectional Cross-ply Angle-ply

Increased frequency Too 5–10 Hz 10–20 Hz 5–10 Hz (a) 10–20 Hz

Fatigue life Much Decrease Increase Increase Decrease
Max. strain Scatter Increase Decrease Decrease Increase
Sample temperature Increase Increase Increase Increase
Ansari et al. 13

The material properties degradation includes mainly temperature as shown in Figure 16 for PBT. The


chemical changes of the matrix materials and de- fatigue strength of samples at temperature 125 C sig-


bonding at fiber/matrix interface. When a material nificantly increased and conversely decreased at 23 C as


has a cyclical load in an environment, such as moisture compared to –40 C. The effect of temperature in both
and temperature, it is under environmental fatigue. fiber orientation directions and at both stress ratios in
Generally, the effect of humidity and elevated temper- PP materials was alike PBT.
ature decreases the fatigue life of fiber composite Kawai et al.43 worked on the off-axis fatigue behav-
ior of unidirectional T800H/Epoxy composite at room

Temperature effect temperature (RT) and 100 C. Nine type of plain

coupon sample, i.e., off-axis angle (h= 0, 5, 10, 15,
Polymer-matrix composites are viscoelastic materials,
20, 30, 45, 60, and 90
) with Vf =64% were made for
and their mechanical properties are significantly affect- testing. TTF tests were conducted up to 106 at R=0.1
ed by the working temperature. The rise in temperature and at a frequency of 10 Hz. The off-axis S-N data at
up to the degree of glass transition temperature (Tg)

RT and 100 C temperature for different off axes are
leads to a reduction in strength, modulus and fatigue shown in Figure 17. The fatigue failure trend at room
life. Fatemi et al.59 and Mortazavian and Fatemi56 temperature was also identical to 100
C temperature.
studied the influence of temperature on the fatigue per- The tensile fatigue strength reduces as the off-axis angle
formance of the two short GFRP composite (PBT and increases. Comparing the S-N data at room tempera-
PP). Fatigue tests were performed under different stress ture (RT) and 100
C for each off-axis angle in the
ratios (R) of 0.3, 0.1, and 1 with different fiber ori- range 0
< h  90
, we find that the fatigue strength




entations and at a temperature of –40 C, 23 C 125 C. at 100
C reduces 33–50% than those at RT.
The fatigue behavior of samples was significantly Kumagai et al.63 studied the fatigue behavior of
affected by temperature in both longitudinal (L) and GFRP woven under T-T fatigue test at low tempera-
transverse (T) directions at a different tested ture. Tests were conducted at frequencies of 1, 4, and
16 Hz at a stress ratio of 0.1. The samples were pre-
pared with 47% fiber volume fraction and tested at
room temperature that is 77 K and 4 K. The low-
temperature environment was attained by submerging
the loading apparatus and samples in liquid helium (4
K) or liquid nitrogen (77 K). A fatigue limit was
reported to be twice at 77 K compared to the room
temperature, as shown in Figure 18. S-N diagrams indi-
cated the typical tendency that as the fatigue life
decreases, the fatigue stress increases. It was also
found that runouts without failure at room tempera-
Figure 16. Effect of temperature on the fatigue behavior.58 ture and 77 K are 106 and 105 cycles respectively.

Figure 17. Off-axis S-N relationship for unidirectional specimen (a) at RT (b) at 100
C.43
14 Journal of Reinforced Plastics and Composites 0(0)

Figure 18. S-N data for sample (a) at RT (b) 77 K.63

Table 4. Description of the tested CFRPs and components.64

Indexa Layup Fiber Matrix

1 Unidirectional T300 EP Rigid

1a Unidirectional T300 EPs—semiflexible
1b Unidirectional HTA7 PC—thermoplastic
2 Unidirectional Structural element T300 EP—rigid
3 Cross-ply(0
/90
/0
/90
/0
)s HTA7 PC—thermoplastic
4 Quasiplanar(0
/þ 45
/90) T300 EP—rigid
5 Quasiplanar(0
, 90
/þ45
)s Structural element HTA7 EP—rigid
a
Index numbers refer to numbers in figures.

Ahlborn64 studied the fatigue behavior of CFRP com-

posites at low temperature. Tensile fatigue tests were
conducted with three laboratory-type CFRPs and two
structural elements, with 60% fiber volume fraction at
a stress ratio of =0.1. Lay-up and composition of the
seven composites tested are described in Table 4.
Fatigue tests were carried out at three different low
temperatures. It was concluded that unidirectional
composites presented the superior fatigue behavior
and maximum fatigue strengths. The fatigue strengths
of the multi-directional composites were slightly lower.
Observably, fatigue strength was influenced by the fiber
orientation as shown in Figure 19. The effect of tem-
perature on S-N curves was minimum as the fatigue
behavior was mainly dominated by the fibers.

Moisture
Moisture affects the mechanical properties and glass
transition temperature (Tg) of the composites. Access Figure 19. Stress-life curves of the CFRPs.64
Ansari et al. 15

of moisture combined with a high operating tempera- with 60% fiber volume fraction were conditioned by
ture significantly weakens the materials mechanical drying, by exposing to 65% relative humidity (r.h)
properties. Blaga and Yamasaki65 studied the physical atmosphere and by boiling in water. Tensile 
fatigue


decay of GFRP composites under cyclic variations in tests were conducted on laminates in 0 =90 and


moisture, temperature, and radiation. They concluded 45 orientation. There was no 
effect of conditioning


that microcracking arisen under the combined effect of on the fatigue behavior of the 0 =90 CFRP samples.
tensile stresses were developed by radiation in the poly- However, GFRP laminates were affected by boiling,
meric surface area and physically developed stress which weakens the composite and reduces the material
fatigue. Numerous photomicrographs also presented comparatively less susceptible to further damage by
to show proof of surface cracks.66 It can be possible fatigue. Preloading to high-stress level before condi-
that water caught by surface microcracks can condense tioning had a small effect on their fatigue behavior of
to ice in cold environments and afterward lead to great- GFRP samples. The fatigue resistance of KFRP com-
er failures caused due to net volume expansion effect, posite was reduced less by boiling in water than by
or it may cause excessive vapor pressures under hot drying. At high-stress material showed excellent fatigue
environments, which can also cause deleterious effects resistance in all condition. However, at low and inter-
in the composite. Shetty et al.67 studied the effect of mediate stress level they behaved poorly due to fiber
hygrothermal environmental conditions on FRP com- damage resulting from complete unloading on the com-
posites w.r.t. aerospace applications. It was found that posite on each cycle. Hu et al.69 studied the influence of
moisture affects the mechanical properties and glass hygrothermal aging on the T-T fatigue behavior of uni-
transition temperature (Tg) of the composites. The deg- directional (UD) glass/poly dicyclopentadiene
radation of mechanical properties in hygrothermal (pDCPD) composites. Glass/epoxy was taken as a
environments allows the designer to calculate the standard for comparison of result and samples were
knockdown factor for composites like the factor of immersed in salt water and deionized Water at 60
C.
safety concept used in the metallic component design. T-T fatigue tests were carried out up to 106 Cycle at
Shan and Liao68 studied the fatigue life of unidirec- R=0.1, and loading frequency 10 Hz. Results showed
tional glass fiber-reinforced and glass carbon- that the aged samples have failed before reaching the
reinforced epoxy matrix composite in the air and dis- 106 cycle. Figure 20(a) and (b) shows the S-N curve for


tilled water at 25 C under T-T fatigue test. Both types samples in deionized water and salt water. The fatigue
of samples have nearly 30% total fiber volume fraction. strength of pDCPD composites was enhanced after one
The relative fiber volume ratio of glass to carbon is month and reduced after three months. The S-N curves
about 3:1 in the hybrid sample. Fatigue and environ- in salt water and DI water were same after three
mental fatigue were executed at a stress level of 85%, months. For the epoxy composites, fatigue strength
65% and 45% UTS. Samples were cyclically loaded at was reduced slowly with aging time in salt water.
10 Hz for 65 and 85%, and 20 Hz for 45% UTS, at However, the fatigue strength reduced significantly
R=0.1 for all glass fiber and hybrid fiber, cyclic load- after one month but enhanced slightly after three
ing at 85% UTS did not show significant change in months in DI water aging, which could increase the
fatigue life between samples tested in air (dry) and dis- strength of the matrix.
tilled water (wet). At 65% UTS some of wet samples Jen and Huang70 studied the hygrothermal effect on
were failed earlier than dry samples. Separation of fatigue strengths of multi-walled carbon nanotubes
fatigue life between wet sample and dry sample (CNTs)/epoxy composites. Fatigue tests were carried
becomes explicit at 45% UTS. Wet samples have out under two hygrothermal conditions 25
C=85%
failed before attaining 107 cycles; however, dry samples RH (room temperature/high humidity (RT/HH)) and
survived 107 cycles and no damage sign was indicated. 40
C=85% RH (high temperature/high humidity (HT/
The environmental fatigue behavior of hybrid compo- HH)) at a stress ratio (R) of 0.1, loading frequency of 5
sites is identical to that of glass composites, with the Hz. Experimental data were compared with previous
stress corrosion dominant behavior at lower stress study,71 in which tests were conducted under 25
C=60%
levels and stress dominant behavior at higher stress RH (room temperature/moderate humidity (RT/MH)).
levels. The fatigue life in cyclic loading significantly The fatigue strength of the CNT/epoxy composite under
decreases at lower stress levels in distilled water for the RT/MH environment was significantly higher than
both hybrid glass carbon fiber and all glass fiber- RT/HH environments and under the HT/HH condition
reinforced composites. decreased significantly as compared to RT conditions.
Jones et al.24 studied the effect of hygrothermal con- The fatigue strength of the CNT/epoxy composites has
ditioning on the fatigue behavior of epoxy-based com- been significantly affected by a combination of humidity
posites reinforced with carbon, glass, aromatic and temperature, however slightly affected only by
polyamide (kevlar49) fiber. These three fiber laminates humidity. Composites with 0.5 wt.% CNT have higher
16 Journal of Reinforced Plastics and Composites 0(0)

Figure 20. Evolution of S-N curves in DI water and salt water environments.69

fatigue strength than those of the 1.0 wt.%. CNT content modulus CFRP. For unidirectional fiber, carbon fiber
composite and neat epoxy composite. is preferable for use in composite, due to their very stiff
nature, which controls the strain in the composite and
thus arrests large deformation in the matrix. The
Summary
fatigue strength of glass fiber-reinforced composite is
FRP composites are now increasingly being used in especially affected by the properties of the matrix. The
high cycle fatigue applications. Composites have fatigue damage will usually initiate in the form of crack
become essential in some fields of engineering applica- initiation in the matrix region. Thermoplastic resins
tions such as aerospace and aircraft, sporting goods, have considerably longer fatigue life. Epoxy matrix
marine and automobile. In these applications, fluctuat- composites have a better fatigue behavior among the
ing and cyclic loads are common and eventually, thermoplastics matrices. For a woven fabric, warp and
fatigue in structures will induce due to this type of weft directions reinforced polymer composites both
loading condition. Higher strength, higher stiffness, have similar important. These fabrics enhance the
corrosion resistance, improved fatigue behavior, and strength in depth. On the other hand, the development
tailored properties are some advantages of polymer of the resin-rich region with these materials is a
composite materials over conventional materials. common problem. Thus, failure initiates at this
However, interchangeability, reparability, delamina- region even at low cycles. The static strength of the
tion, and discontinuous stress are the drawbacks of material will increase by adding nanoparticles in the
polymer composites. The fatigue behavior of composite matrix. It also improves the fatigue life as the load-
materials is different from that of the metals. carrying capacity of the composite increases and also
Composites are heterogeneous and anisotropic as com- changes the damage mechanism of the composite in the
pared to metals which are homogenous in nature and fatigue test. The crack path is strongly affected by the
isotropic in behavior. In fatigue, composite materials fiber volume fraction. The fatigue performance of the
experience a number of changes, that is, fiber breaking, composite is increased with an increase in the fiber
fiber–matrix debonding, matrix cracking and finally the volume fraction. The fatigue strength is also influenced
failure. Failure of notched specimens is achieved in by the fiber orientation, as the fiber orientation
three stages: (1) crack initiation; (2) stable crack prop- increases strength is decreased. The mean stress has a
agation; and (3) unstable crack propagation. Thus, major effect on the fatigue behavior of the composite.
fiber, matrix, and the interface between matrix and As stress ratio define the mean stress or vice versa, the
fibers play a vital role in determining the fatigue behav- fatigue life of the composite can be reduced by increas-
ior of FRP composite. ing the stress ratio. The fatigue performance of the
The polymer-based composites fail at low strains composite is also affected by loading or testing
due to their brittle nature, thus creating a challenge frequency.
to study the stress-controlled fatigue. In the glass Polymer-based composites, which are viscoelastic in
fiber case, S-glass has higher strength and stiffness in nature, are sensitive to temperatures as the glass tran-
comparison to the other glass fibers. High modulus sition temperature (Tg) changes the properties of the
CFRP has a better fatigue performance than Low polymer. At this temperature, the polymer transforms
Ansari et al. 17

from a hard, glassy material to a soft, rubbery material, 7. Meguid SA and Sun Y. On the tensile and shear strength
thus losing all or some of its stiffness. The higher the of nano-reinforced composite interfaces. Mater Des 2004;
frequency, the earlier the specimen will reach its glass 25: 289–296.
transition temperature (Tg). The fatigue life of 8. Zhou Y, Pervin F, Jeelani S, et al. Improvements in
mechanical properties of carbon fabric-epoxy composite
polymer-based composites is also individually affected
using carbon nanofibers. J Mater Process Technol 2008;
by temperature. At higher temperatures, the composite
198: 445–453.
fails earlier. However, at cryogenic or lower temper- 9. Iwahori Y, Ishiwata S, Sumizawa T, et al. Mechanical
atures, the fatigue strength and fatigue life of the com- properties improvements in two-phase and three phase
posite material increase. Addition of CNTs decreases composites using carbon nano-fiber dispersed resin.
the cryogenic fatigue properties of the composite. Compos Part A Appl Sci Manuf 2005; 36: 1430–1439.
Moisture in glass fiber-based composites degrades the 10. Shao Y, Yashiro T, Okubo K, et al. Effect of cellulose
fatigue performance. Matrix corrosion, interface weak- nanofiber (CNF) on fatigue performance of carbon fiber
ening, fiber swelling are common in such case. fabric composites. Compos Part A Appl Sci Manuf 2005;
Conversely, nominal deviation in the properties 76: 244–254.
occurs in case of carbon fiber-based composites. 11. Mallick PK. Fiber reinforced composites: materials,
manufacturing and design. 3rd ed. Florida: CRC Press,
Thus, usually, the addition of CNTs increases the
2008.
fatigue performance of the composites as carbon 12. Xu HHK, Ostertag CP, Braun LM, et al. Effects of fiber
fibers possess better fatigue properties as compared to volume fraction on mechanical properties of SiC-fiber/
glass fibers either at room temperature or an increased Si3N4 matrix composites. J Am Ceram Soc 1994; 77:
temperature along with the nominal effect of moisture. 1897–1900.
13. Brunbauer J, Stadler H and Pinter G. Mechanical prop-
Declaration of conflicting interests erties, fatigue damage and microstructure of carbon/
epoxy laminates depending on fiber volume content. Int
The author(s) declared no potential conflicts of interest with
J Fatigue 2015; 70: 85–92.
respect to the research, authorship, and/or publication of this
14. Owen MJ. Fatigue of carbon-fiber-reinforced plastics.
article. In: LJ Broutman (ed) Composite material fracture and
fatigue volume 5. New York: Academic Press, 1974,
Funding pp. 341–369.
The author(s) received no financial support for the research, 15. Branco CM, Ferreira JM, Richardson MOW, et al.
authorship, and/or publication of this article. Fatigue behaviour of a phenolic matrix composite. Int J
Fatigue 1992; 14: 367–376.
16. Hahn HT and Kim RY. Fatigue behaviour of composite
References laminates. J Compos Mater 1976; 10: 156–180.
1. Erden S, Sever K, Seki Y, et al. Enhancement of the 17. Reifsnider KL and Jamison R. Fracture of fatigue-loaded
mechanical properties of glass/polyester composites via composite laminates. Int J Fatigue 1982; 4: 187–197.
matrix modification glass/polyester composite siloxane 18. Talreja R. Damage mechanics and fatigue life assessment
matrix modification. Fibers Polym 2010; 11: 732–737. of composite materials. Int J Dam Mec 1999; 8: 339–354.
2. Hollaway LC. A review of the present and future utiliza- 19. Curtis PT. A review of the fatigue of composite materials.
tion of FRP composites in the Civil infrastructure with Technical Report 87031, Royal Aircraft Establishment,
reference to their importance in service properties. Constr Farnborough 1987.
Build Mater 2010; 24: 2419–2445. 20. Hyakutake H, Hagio T and Yamamoto T. Fatigue fail-
3. Wichmann MHG, Sumfleth J, Gojny FH, et al. Glass- ure criterion for notched FRP plates. JSME Int J Ser A
fiber-reinforced composites with enhanced mechanical 1993; 36: 215–219.
and electrical properties-benefits and limitations of a 21. Pink E and Campbell JD. Deformation characteristics of
nanoparticle modified matrix. Eng Fract Mech 2006; 73: reinforced epoxy resins: part 1. J Mater Sci 1974; 9:
2346–2359. 658–664.
4. Gojny FH, Wichmann MHG, Kopke U, et al. Carbon 22. Boller KH. Fatigue fundamentals for composite materi-
nanotube reinforced epoxy composites: enhanced stiff- als. ASTM STP 1969; 460: 217–235.
ness and fracture toughness at low nanotube content. 23. Myers TJ, Kytomaa HK and Smith TR. Environmental
Compos Science Techno 2004; 64: 2363–2371. stress-corrosion cracking of fiberglass: lessons learned
5. Miyagawa H and Drzal LT. Thermo-physical and impact from failures in the chemical industry. J Hazard Mater
properties of epoxy nanocomposites reinforced by single- 2007; 142: 695–704.
wall carbon nanotubes. Polymer 2004; 45: 5163–5170. 24. Jones CJ, Dickson RF, Adam T, et al. The environmental
6. Hsiao KT, Alms J and Advani SG. Use of epoxy/multi- fatigue behaviour of reinforced plastics. Proc R Soc Lond
walled carbon nanotubes as adhesives to join graphite A 1984; 396: 315–338.
fiber reinforced polymer composites. Nanotechnology 25. Curtis PT. Fatigue behaviour of fibrous composite mate-
2003; 14: 791–801. rials. J Strain Anal Eng Des 1989; 24: 235–244.
18 Journal of Reinforced Plastics and Composites 0(0)

26. Owen MJ and Howe RJ. Accumulation of damage in a 43. Kawai M, Yajima S, Hachinohe A, et al. Composites at
glass-reinforced plastic under tensile and fatigue loading. room and high temperatures off-axis fatigue behaviour of
J Phys D Appl Phys 1972; 5: 1637–1649. unidirectional carbon fiber reinforced. J Comp Mater
27. Caprino G and D’amore A. Flexural fatigue behaviour of 2001; 35: 545–576.
random continuous-fiber-reinforced thermoplastic com- 44. Bernasconi A, Davoli P, Basile A, et al. Effect of fiber
posites. Compos Sci Technol 1998; 58: 957–965. orientation on the fatigue behavior of a short glass fiber
28. Bathias C. Fracture and fatigue of high-performance reinforced polyamide-6. Int J Fatigue 2007; 29: 199–208.
composite materials: mechanisms and prediction. Eng 45. Huh YH, Lee JH, Kim DJ, et al. Effect of stress ratio on
Fract Mech 1991; 40: 757–783. fatigue life of GFRP composites for WT blade. J Mech
29. Bureau MN and Denault J. Fatigue behaviour of contin- Sci Technol 2012; 26: 2117–2120.
uous glass fiber composites: effect of the matrix nature. 46. Reifsnider KL, Schulte K and Duke JC. Long-term
Polym Compos 2000; 21: 636–644. fatigue behaviour of composite materials. ASTM STP
30. Hojo M, Matsuda S, Tanaka M, et al. Mode I delami- 1983; 813: 136–159.
nation fatigue properties of interlayer-toughened 47. Hinton MJ, Kaddour AS and Soden PD. The world wide
CF/epoxy laminates. Compos Sci Technol 2006; 66: failure exercise. Oxford: Elsevier, 2004.
665–675. 48. Vasiukov D, Panier S and Hashemi A. Direct method for
31. Zhuang H and Wightman JP. The influence of surface life prediction of fiber reinforced polymer composites
properties on carbon fiber/epoxy matrix interfacial adhe- based on kinematic of damage potential. Int J Fatigue
sion. J Adhes 1997; 62: 213–245. 2015; 70: 289–296.
32. Nagalingam R, Sundaram S and Retnam BSJ. Effect of 49. Nixon-Pearson OJ, Hallett SR, Withers PJ, et al.
nanoparticles on tensile, impact and fatigue properties of Damage development in open-hole composite specimens
fiber reinforced plastics. Bull Mater Sci 2010; 33: in fatigue. Compos Struct 2013; 106: 882–889.
525–528. 50. Bizuel M, Bouvet C, Barrau JJ, et al. Fatigue crack
33. Knoll JB, Riecken BT, Kossmann N, et al. The effect of growth in thin notched woven glass composites under
carbon nanoparticles on the fatigue performance of tensile loading. Part I: experimental. Compos Sci
carbon fiber reinforced epoxy. Compos Part A Appl Sci Technol 2011; 71: 289–296.
Manuf 2014; 67: 233–240. 51. Harris B. Fatigue in composites in science and technology
34. Jen YM and Wang YC. Stress concentration effect on the of the fatigue response of fiber reinforced plastics. 1st ed.
fatigue properties of carbon nanotube/epoxy composites. Florida: CRC Press, 2003.
Compos B Eng 2012; 43: 1687–1694. 52. Gamstedt EK and Talreja R. Fatigue damage mecha-
35. Newaz GM. Influence of matrix material on flexural nisms in unidirectional carbon-fiber-reinforced plastics.
fatigue performance of unidirectional composites. Comp J Mater Sci 1999; 34: 2535–2546.
Sci Technol 1985; 24: 199–214. 53. Belmonte E, Monte MD, Hoffmann CJ, et al. Damage
36. Manjunatha CM, Sprenger S, Taylor AC, et al. The ten- mechanisms in a short glass fiber reinforced polyamide
sile fatigue behaviour of a glass-fiber reinforced-plastic under fatigue loading. Int J Fatigue 2017; 94: 145–157.
composite using a hybrid-toughened epoxy matrix. J 54. Zhao X, Wang X, Wu Z, et al. Fatigue behavior and
Comp Mater 2010; 44: 2095–2109. failure mechanism of basalt FRP composites under
37. Wu Z, Wang X, Iwashita K, et al. Tensile fatigue behav- long-term cyclic loads. Int J Fatigue 2016; 88: 58–67.
iour of FRP and hybrid FRP sheets. Compos B Eng 2010; 55. Nasr MNA, Abouelwafa Gomaa A, et al. The effect of
41: 396–402. mean stress on the fatigue behaviour of woven-roving
38. Brunbauer J and Pinter G. Effects of mean stress and glass-fiber-reinforced polyester subjected to torsional
fiber volume content on the fatigue-induced damage moments. J Eng Mater Technol 2005; 127: 301–309.
mechanisms in CFRP. Int J Fatigue 2015; 75: 28–38. 56. Mortazavian S and Fatemi A. Effects of mean stress and
39. Abd Allah MH, Abdin EM, Selmy AI, et al. Effect of stress concentration on fatigue behaviour of short fiber
fiber volume fraction on the fatigue behaviour of grp reinforced polymer composites. Fatigue Fract Eng Mater
pultruded rod composites. Compos Sci Technol 1996; Struct 2016; 39: 149–166.
56: 23–29. 57. Mallick PK and Zhou Y. Effect of mean stress on the
40. Mini KM, Lakshmanan M, Mathew L, et al. Effect of stress-controlled fatigue of a short E-glass fiber rein-
fiber volume fraction on fatigue behaviour of the glass forced polyamide-6,6. Int J Fatigue 2004; 26: 941–946.
fiber reinforced composite. Fatigue Fract Eng Mater 58. Ferreira JAM, Costa JDM and Richardson MOW.
Struct 2012; 35: 1160–1160.[CrossRef][10.1111/j.1460- Effect of notch and test conditions on the fatigue of a
2695.2012.01709.x][Mismatch] glass-fiber-reinforced polypropylene composite. Comp
41. Belmonte E, Monte MD, Hoffmann CJ, et al. Damage Sci Technol 1997; 57: 1243–1248.
initiation and evolution in short fiber reinforced polyam- 59. Fatemi A, Mortazavian S and Khosrovaneh A. Fatigue
ide under fatigue loading: influence of fiber volume frac- behavior and predictive modeling of short fiber thermo-
tion. Compos B Eng 2017; 113: 331–341. plastic composites. Procedia Engineering 2015; 133: 5–20.
42. Kadi HE and Ellyin F. Effect of stress ratio on the fatigue 60. Barron V, Buggy M and Mckenna NH. Frequency effects
of unidirectional glass fiber/epoxy composite laminae. on the fatigue behaviour on carbon fiber reinforced poly-
Composites 1994; 25: 917–924. mer laminates. J Mater Sci 2001; 36: 1755–1761.
Ansari et al. 19

61. Eftekhari MR and Fatemi A. On the strengthening effect plastic composites. In: Proceedings of 14th ISAMPE
of increasing cycling frequency on fatigue behaviour of national conference on composites (INCCOM-14),
some polymers and their composites: Experiments and Hyderabad, India, 22–23 January, 2016
modeling. Int J Fatigue 2016; 87: 153–166. 68. Shan Y and Liao K. Environmental fatigue behaviour of
62. Flore D, Wegener K, Mayer H, et al. Investigation of the unidirectional glass–carbon fiber reinforced hybrid com-
high and very high cycle fatigue behaviour of continuous posite. Compos B Eng 2001; 32: 355–363.
fiber reinforced plastics by conventional and ultrasonic 69. Hu Y, Lang AW, Li X, et al. Hygrothermal aging effects
fatigue testing. Compos Sci Technol 2017; 141: 130–136. on fatigue of glass fiber/poly dicyclopentadiene compo-
63. Kumagai S, Shindo Y and Inamoto A. Tension-tension sites. Polym Degrad Stab 2104; 110: 464–472.
fatigue behaviour of GFRP woven laminates at low tem- 70. Jen YM and Huang CY. Combined temperature and
peratures. Cryogenics 2005; 45: 123–128. moisture effect on the strength of carbon nanotube rein-
64. Ahlborn K. Fatigue behaviour of carbon fiber reinforced forced epoxy materials. Trans Canad Soci Mech Eng
plastic at cryogenic temperatures. Cryogenics 1988; 28: 2013; 37: 755–763.
267–272. 71. Jen YM and Huang CY. Fatigue characterization of
65. Blaga A and Yamasaki RS. Mechanism of surface micro- acid-treated carbon nanotube/epoxy composites.
cracking of matrix-in glass reinforced polyester by artifi- J Comp Mater 2013; 47: 1665–1675.
cial weathering. J Mater Sci 1973; 8: 1331–1339. 72. Soden PD, Hinton MJ and Kaddour AS. Lamina prop-
66. Blaga A. Weather study of glass-fiber reinforced polyes- erties, lay-up configurations and loading conditions for a
ter sheets by scanning electron microscopy. Polym Eng range of fibre-reinforced composite laminates. Comp Sci
Sci 1972; 12: 53–58. Technol 1998; 58: 1011–1022.
67. Shetty K, Srihari S and Manjunatha CM. A survey on
the environmental effects on aerospace fiber reinforced