Accepted Manuscript

Effect of load amplitude change on the fatigue life of cracked Al plate repaired
with composite patch
A. Albedah, Sohail M.A. Khan, F. Benyahia, B. Bachir Bouiadjra
PII:
DOI:
Reference:

S0142-1123(16)30007-X
http://dx.doi.org/10.1016/j.ijfatigue.2016.03.002
JIJF 3881

To appear in:

International Journal of Fatigue

Received Date:
Revised Date:
Accepted Date:

7 November 2015
26 February 2016
1 March 2016

Please cite this article as: Albedah, A., Khan, S.M.A., Benyahia, F., Bachir Bouiadjra, B., Effect of load amplitude
change on the fatigue life of cracked Al plate repaired with composite patch, International Journal of Fatigue (2016),
doi: http://dx.doi.org/10.1016/j.ijfatigue.2016.03.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effect of load amplitude change on the fatigue life of cracked Al plate repaired with
composite patch
A. Albedaha*, Sohail M. A. Khana, F. Benyahia a, B. Bachir Bouiadjra b, a
b LMPM,

aMechanical Engineering Department, College of Engineering, King Saud University, Riyadh, Saudi Arabia
Department of Mechanical Engineering, University of Sidi Bel Abbes, BP 89, Cit Ben Mhidi, Sidi Bel Abbes 22000, Algeria
*corresponding author: Tel. +966114677111, albedah@ksu.edu.sa

Abstract
In this study, the fatigue behavior of aluminum alloy 2024T3 V-notched specimens repaired
with composite patch under block loading was analyzed experimentally. Two loading blocks
were applied: increasing and decreasing at two stress ratio: R=0 and R=0.1. Failed samples
were examined under scanning electron microscope at different magnifications to analyze
their fractured surfaces. The obtained results show that under increasing blocks, the crack
growth is accelerated for both repaired and un-repaired specimens. This is attributed to the
increase of the loading amplitude in the second block. A retardation effect was observed for
decreasing blocks loading in unrepaired specimens. However, this retardation effect is
attenuated by the presence of the patch which lead to lower fatigue life for repaired
specimens.
Key words: Bonded composite repair, block loading, Retardation effect, Fatigue life.

1. Introduction
Aged aircraft structure may contain fatigue cracks resulting from their long service. Fatigue
loads, of constant and variable cyclic load amplitudes, are the major contributors to the
service induced damages in the aerospace structures. Bonded composite patches have been
successfully used to repair the damaged structures and it is structurally very efficient, can be
applied rapidly and are cost effective [1-3]. The technology involves adhesively bonding
patches of advanced fiber composite materials to repair damaged aircraft structures and
to prevent stress corrosion cracking. Once repaired using bonded composite patch, there is no
guarantee that the structure continue being subjected to the same loading amplitudes. Several
studies of unrepaired cracks [4, 5] show the significant effect of the load history on the fatigue
life of metallic structures. Tensile and compressive overload cycles of near yield magnitude
have been shown to accelerate crack growth [6, 7] while crack retardation has been observed
on long cracks following tensile overloads that were well below the yield strength[7].
Experiments composed of two constant amplitude loading blocks changing from a low to a
higher stress level (LH block) or vice versa (HL block) are usually employed to study the
block loading in materials. However the results obtained from these experiments are not
consistent, showing a greater damaging effect due to the LH blocks and retardation effect
due to H-L blocks, depending on the material and loading parameters [4, 8]. This is generally
explained by the size of the plastic zone created at the crack tip as well as by the residual
stresses and crack branching in other cases.
In the case of repaired cracks, the interaction of the load change with the stress bridging and
redistribution, caused by the composite patch, may have a considerable impact on the crack
repair.

Several works have been published on the effect of variable amplitude loading, cycle mix and
the variation in mean stress on the adhesively-bonded joints [9-11], the effect of overloads
and loading sequence on adhesively bonded double-lap joints [12-14] .
Khan et al [15] analyzed the fatigue life of Al 7075-T6 cracked specimens repaired with
adhesively bonded composite patches for different load ratios and compared with that of
unrepaired specimen under the same cyclic block loading for two different sequences
(increasing and decreasing amplitude). They showed that the patch repair efficiency is not
significant for increasing blocks of loading, whereas for decreasing blocks of loading, the
improvement is relatively noticeable. The combination of the fiber bridging and the
retardations effect leads to the significant improvement of the fatigue life for the repaired
structures.
The repair performance of cracked aluminum plates using bonded composite patches can be
evaluated by monitoring, among others, the repaired crack growth under different loads. It has
been shown in previous studies [16-18] that the smaller the initial size of the repaired crack
the higher is the efficiency of the patch repair. However, to the knowledge of the authors,
there is no information in the literature on the behavior of a patch repair relating the load
amplitude leading to the crack generation and that applied after the repair.
Behavior of cracks repaired with bonded composite patch under variable amplitude loading is
more complex than under constant amplitude loading [19]. To understand the more realistic
phenomenon which occur during the aircraft services such as the transient crack closure and
the retardation effects associated with overloads, it is necessary to study the bonded
composite repair under representative flight load spectra. The first case of such repair was
executed on the wing skin of RAAF Mirage III aircraft [20], subjected to an operational flight
load spectra and the other major case was composite repair on USAF C-141 aircraft [21]. The

repair has been applied to over 150 wings in service and over 3 years of operational history
has been observed during the investigation. Both cases have been reported to be successful.
The finite element method was also carried out during the repair of RAAF Mirage III aircraft.
It has been used to design several complex repair schemes, such as the repair of fatigue cracks
in the lower skin of Mirage aircraft and cracks on the upper surface of the wing pivot fitting
of FlllC aircraft in service with the Royal Australian Air Force (RAAF) [26]. Baker reported
that many patch repairs of the Royal Australian Air Force have been in service for longer than
20 years without durability issues arising from environmental or fatigue damage [22].
Raizenne et al [23] conducted series of fatigue tests on repaired aircraft panels using a
clipped FALSTAFF spectrum in which the negative loading has been removed. Poole [25]
reported the beneficial effect of single side patch repair on thin aluminum sheets and
suggested that the effect of variable amplitude loading spectra on patch debonding should be
studied in terms of patch efficiency. He also suggested that there is a clear requirement for a
model to predict debonding, and patch efficiency under a wide range of loading spectra.
Baker [27] conducted fatigue test on wing skin of an Australian Defense Force F-111C
aircraft repaired with an adhesively bonded boron/epoxy fiber composite patch. During these
tests, the patch was successful in preventing growth of the crack for around a further 9000
simulated flying hours.

In the work done by Walker and Rose [28], safety critical bonded composite repair to the outer

lower wing skin of an F-111 aircraft has been fully validated and substantiated. The first wing
to be repaired exceeded 665.9h of actual operational usage and a further 8074.4 h in a fullscale wing fatigue test [28].

In this study, we aim to evaluate experimentally the effect of the load amplitude change priorpost the repair on its efficiency. Two blocks of constants maximum loading set-ups are
considered in this study: Low-High (7kN-12kN) and High-Low (12kN-7kN). One of the two
loading blocks is responsible to creating and propagating a crack, from a notched specimen,
to a length of 3 millimeters representing the initial detected crack size to be repaired. After
repair, the specimen is subjected to the second block where the maximum load is changed and
the crack growth is monitored until failure. Similar tests are conducted but without patch for
the purpose of comparison. Previously published results [29] are used to highlight the material
effect on the performance of the composite patch repair through the two different load levels.

2. Experimental setup
2.1 Chemical characterization of the material
The material used in this study was in the form of thin plates. Few samples of the material
was taken and chemically analyzed on SEM equipped with Energy Dispersive X-ray analyzer
(EDS) to confirm the composition and properties of Al 2024 tempered at T3. The chemical
composition of the aluminum alloy (Al 2024-T3) is given in Table 1.

2.2 Specimen details

In this study, we used Al2024-T3 single edged notched tension (SENT) specimens (see Figure
1) of dimensions 150x50x2 mm. An initial v-notch of 6 mm depth and 60 0 angle was created,
in accordance to ASTM E647 standards[30],in the center of each specimen by milling, as
shown in Figure 1. The specimens were then pre-cracked under fatigue loading to different
initial crack lengths. The presence of the V notch facilitates the mode I propagation of the

crack. Once the desired crack length is reached, the sample was unloaded and taken for
surface preparation followed by bonding the composite patch.

2.3 Patch preparation

Composite patches are made using 8 plies of unidirectional carbon/epoxy pre-pregs. The
dimension of each ply was 250x250 mm. The pre-pregs were sandwiched in between two
Teflon sheets and woven matrix of thermoplastic polymers to absorb the extra resin which
comes out under curing. After laying up the laminates, it was then cured under a press at
1200C for 90 minutes. The composite plate is later cut into 50x50 mm square patches.

The use of double sided symmetric patch is more efficient and would annul the bending
moment due the shift of the neutral axis for single sided patch, but the main disadvantage of
the double sided patch is the difficulty to monitor the crack propagation. To minimize the
bending effect for single sided patch, we have carefully chosen the stiffness ratio between the
cracked plate and the composite patch. This ratio is defined by :

.
.

Where Er : The young modulus of the composite patch

Ep : is the young modulus of the aluminum plate

er : is the patch thickness

ep :: is the plate

which is an important non-dimensional parameter characterizing a repair.

It was shown [31] that higher stiffness ratio provided lower crack growth rate or longer
fatigue life for both thin and thick repaired plate for a given patch length. Higher stiffness
ratio gives less bending from thermal stresses and hence reduces radius of curvature due to
the neutral axis offset from application of the load [31]. Air Force and Boeing studies [32, 33]
have determined that stiffness ratios of 1.2 to. 1.5 produce effective doubler designs.
Lockheed-Martin has also used this range of stiffness ratios in military composite doubler
designs. In this study we have chosen the value of 1.48 for the stiffness ratio.

2.4 Adhesive
The surface of SENT specimens was prepared according to Bell Process Specification method
[34]. The pre-cracked specimens were repaired with the composite patches bonded to the
specimens using the bi-components Permabond ET515 adhesive such that the lay-up principal
direction is perpendicular to the loading direction, as shown in Figure 2. The specimens were
cleaned using acetone as there were no indents or scratches on the specimen and it was
received in mirror polished condition. After cleaning the specimen, Permabond ET515A, the
epoxy resin, is mixed with Permabond ET515B, the polyamine hardener, in 1:1 ratio and the
mixture is applied to pre-cracked area and the composite patch was bonded to it as shown in
Figure 2. Permabond ET515 is cured without any external heat and thus, prevents the
galvanization between the composite and the metal and, hence, avoids the formation of
thermal residual stresses. The assembly of aluminum specimen and patch (Figure 2) was
allowed to cure under pressure for 72 hours, without any heat, to obtain the full handling
strength.

2.5 Fatigue tests procedure

Fatigue tests were conducted on SENT specimens according to the ASTM E647 standard
[24]. The specimens were obtained in the transverse longitudinal direction from an unclad 2
mm thick aluminum sheets. The fatigue tests were performed in a load controlled mode on
100 KN servo- hydraulic INSTRON 8801 machine. The tests were conducted at room
temperature using a sinusoidal waveform at a loading frequency of 20 Hz under constant and
block loading. The crack length versus number of cycles were recorded in each case using a
digital camera (See Figure 3) and no damage was detected in the carbon fibers during the
fatigue tests.

Two cyclic block-loading sequences (increasing and decreasing), as shown in Figure 4, were
applied to each specimen in two sequences. The tests were performed using the first loading
block of constant maximum load, low or high (L or H, 7 kN or 12 kN) until the crack reached
3 mm length, then the second maximum loading block , high or low, (H or L, 12kN or 7kN) is
applied and maintained until fracture. For the repaired specimens, the same procedure is used
except that the composite patch is bonded after the crack length (a) reaches 3 mm. Two stress
ratios(R=Fmin/Fmax), R=0 and R=0.1 were used for the two loading cases. The details of
loading case, patch configurations and materials used are summarized in Table 2.

We note that tests on one sided patch repairs without bending effect being constrained to have
little effect, is similar to that performed in [35]. This was done even though the resultant test
program is believed to result in localized bending that is unrepresentative of what will happen
in an operational aircraft.

2.6 SEM observations

Fracture surfaces of selected fatigue specimens were inspected visually and with a stereo
microscope. Further, macroscopic and microscopic examination was done using SEM
(Scanning electron microscope model Joel JSM-6610LV) for the purposes of measuring
crack growth striations and identifying crack initiation sites. The specimens were prepared,
for the SEM observation, by cutting immediately below the fracture surface to reduce the
height of the samples to fit into the vacuum chamber of the SEM. The samples, after cleaned
with acetone, were then mounted on an SEM sample holder using carbon tape. An Electron
Diffraction Spectroscopy (EDS) Camera was used to gather chemical analysis of the initiation
sites in the fracture surfaces.
The observations were conducted with the electron beam in alignment with the applied axial
force. The images of the important features of the fracture surfaces were recorded. The
sections of the fatigue fractured samples were also inspected via stereo and SEM microscopes
to identify the specific fracture modes. Where appropriate, the fatigue crack propagation
modes were recorded using back scattered electron imaging. The selected samples from each
type of materials and with different loading conditions were used for fractographic analysis.

3. Results and discussion

The main motivation of this work was to quantify the retardation/acceleration effects related
to the superposition of the patch repair and the load history (amplitude) of the crack detected
during structural regular inspections. It is known that the plastic zone at the crack tip before
patching is related to the load amplitude bringing the crack to its initially detected length (3
mm in our case). In case of the unrepaired specimen, the crack growth is known to increase or
decrease consistently with the load change. The results that we discuss in this section are
9

related to the combined effect of patching and load change. These results are also compared
with the unrepaired case as a reference and known configuration. Also, we studied the effect
of patch repair under constant load and presented the result in the following sections.

3.1 Combined effect of patch repair and load increase on the crack growth
Figure 5 presents the fatigue life response (crack length vs. cycle number) of Al 2024-T3
specimen, pre-cracked to a length of 3 mm, under amplitude loading (L= 7 kN) then subjected
to a stepped load increase (H= 12 kN) up to failure. The fatigue tests were conducted for two
stress ratios R=0 and R=0.1 applied to both fatigue loading blocks (L and H).The results
shown in the Figure 5 indicate that the fatigue life of repaired specimens was improved for
both stress ratios. However, the rates of life improvements, in number of cycles to failure, are
fairly low and are limited to about 17% for R=0 and to 21% for R=0.1. In a previous study
[36],the authors showed that for specimens, with semi-circular notches, reinforced with a
composite patch and subjected to constant amplitude loading of 7 kNat a stress ratio of 0.1,
the fatigue life increases sixteen times. The poor fatigue life improvement obtained for the
repaired specimen, in our study, is mainly dominated by the load jump from 7kN to 12 kN
which accelerated the crack growth rate and almost eliminated the patch strength effect in
carrying a part of the load transferred throughout the adhesive layer. In addition, the increase
of the fatigue loading amplitude stimulates the adhesive partial or total disbond, which can
considerably reduce the overall fatigue life of the repaired structure. It is important to mention
that, the usage of V-notch specimens increases the crack growth rate and naturally reduces the
repair efficiency. The V-notch choice for this study was motivated by the need of a fast initial
crack initiation and a focus on the crack propagation in mode I. The results shown in Figure 5
are also consistent with several studies highlighting an increase of the fatigue life with the
stress ratio [5, 6, 15, 28]. However, in our case, the stress ratio has a lower effect, on the
10

fatigue life, than that obtained under constant amplitude loading. Figure 6 presents the fatigue
life of Al 2024-T3 specimens subjected to constant block loading at two different load levels
of 7kN and 12 kN, compared with stepped load increase scenario (L-H). The figure shows
that fatigue life is increased to 1.8 and 2.25 times by reinforcement of patch under constant
amplitude loading of 7 kN and 12 kN, respectively. The change in fatigue life after the
reinforcement of patch is compared and presented in Table 3. Seo and Lee [17]reported that,
for a test performed on CCT specimen at R=0.1, fatigue life ratio was about 3.5 times the
baseline specimen.

The Figure 7 presents the variation of the crack growth rate (da/dN) as a function of the crack
length (a) for repaired and unrepaired samples under constant and block loading and for the
two stress ratios R=0 and R=0.1. This figure shows that before patching (at a = 3 mm), the
crack growth rate does not exceed the value of 3.10 -7 mm/cycle. The growth rate increases
significantly when the fatigue load amplitude jumps from 7kN to 12 kN for R=0 and R=0.1
and for both configurations; patched and unpatched. The growth rate, which is practically
linear with the crack length, indicates that the fast load increase has almost abolished the
benefits of the patch repair which has compensated for the extra load to keep the fatigue life
similar to the unpatched case totally loaded with 7kN up to failure.

3.2 Combined effect of patch repair and load reduction on the crack growth
Figure 8 presents the fatigue life response of repaired and unrepaired specimens subjected to
stepped load reduction scenario (H-L)for R=0 and R=0.1. At a first glance, a significant
retardation effect of the crack growth for both repaired and unrepaired specimens is observed.
This is manifested by a substantially stopping of the crack during a significant number of
cycles. Some researchers attributed the primary cause of retardation to a residual plastic zone

11

ahead of the crack tip though to be proportional to the load amplitude[37, 38]. Other causes
may be crack tip blunting, crack branching, plasticity induced closure and roughness induced
closure [9]. Figure 8 shows an unexpected behavior; the fatigue lives of unrepaired specimens
are higher than those of repaired ones for both R=0 and R=0.1. In other words, the fatigue life
is reduced by the composite patch by 153392 cycles for R=0.1 and 113800 cycles for R=0.
Those results indicate that the composite patch caused a speedup of the crack growth rate for
the experimented H-L (12-7 kN) loading scenario. This has decreased the repaired specimen
fatigue life rather than improving it. According to several studies, available in the literature,
on the overload effects on crack growth retardation, this result can be explained by the fact
that a larger plastic zone is created by the firstly applied high load (12 kN) causing a
considerable reduction of the crack growth rate for the unpatched specimen when subjected to
the secondly applied lower load (7 kN).The plastic zone increases the resistance to the crack
propagation which provokes the observed retardation effect. For repaired specimens a
bridging effect takes place. The stress transfer between the cracked plate and the composite
patch reduces the plastic strain in the aluminum plate and consequently the extent of the
plastic zone around the crack tip is also reduced. This reduction weakens the retardation
phenomenon of the crack growth causing a significant reduction of the fatigue life of the
repaired specimen compared to the unrepaired one. The fatigue life summary is shown in
Table 3. It can be observed that the fatigue life is increased after the reinforcement of patch in
all cases except under stepped load reduction case of Al 2024-T3.

Figure 9 presents the variation of the fatigue crack growth rate (da/dN) as function of the
crack length (a) for repaired and unrepaired specimens subjected to step decreased loading
(H-L). Because of the high amplitude of loading (12 KN), the fatigue crack growth rate is
high for a crack length less than 3 mm(before patching). This rate drops noticeably when the
12

loading amplitude is reduced to 7kN for both repaired and unrepaired specimens. The crack
growth rate increases slowly as the crack length increases but at a rate 10 times lower
compared to the case of step increased loading (Figure 7) .

The results of load reduction case are compared with constant amplitude loading as shown in
Figure 10. The figure indicates that fatigue life is improved by reinforcement of patch under
constant amplitude loading, whereas in case of load reduction case, it is the reversed. The
improvement in the fatigue life is about 1.8 and 2.25 folds compared to baseline specimen at
constant amplitude loading of 7 kN and 12 kN, respectively, at a stress ratio of 0.1.

We have also conducted other tests to evaluate the impact of the first load amplitude on the
overall specimen fatigue life. Four load amplitudes were used: 9,10, 11 and 12 kN. After
patch repair the load amplitude was maintained to 7 kN. The fatigue life is shorter, see Figure
11, for smaller pre-patch load amplitude. This is because the plastic zone, produced by higher
pre-patch load amplitude, ahead of the crack front is reduced by the presence of the patch.
This reduction can attenuate the retardation effect of the crack growth and consequently, the
role of a bonded patch composites repair can be inversed.

To confirm the precedent results and conclusions, we have compared the fatigue crack growth
of aluminum alloys 2024-T3 (ductile) specimens with 7075-T6 (brittle) and presented the
comparison in Figure 12-16. The results of Al 7075-T6 were taken from our precedent work
(see Khan et al. [15]). The comparison was done for repaired and un-repaired samples
subjected to step reduced fatigue loading (H-L) with a stress ratio of 0.1 as shown in Figure
14. The retardation effect is clearly more substantial for Al 2024-T3 samples compared to
those of Al 7075. This confirms that the higher ductility of Al 2024, compared to that of Al

13

7075, as shown in Figure 15, is responsible for the reversed patch effect. In fact, the fatigue
life of Al 7075 specimens is improved by the composite patch while that of Al 2024
specimens is drastically reduced by the same patch.
Molent et al [39] studied fatigue crack growth data from a large number of fatigue tests on
different military aircraft types. The data analyzed, generally, include the primary crack(s)
leading to failure of the test article. They showed that, an exponential model fits well the
crack growth history data regardless of the design standard used in designing the airframe and
the metallic material used in its construction. Therefore, it is important to consider fatigue
tests with real loading spectrum, which will be done in future works.

3.3. Fractographic observations

After the fatigue tests, the failed samples were examined under scanning electron microscope
at different magnifications to identify and quantify different fractographic features.
Fractography was conducted on representative repaired and unrepaired Al 2024 samples
subjected to step increasing and decreasing fatigue loads (H-L and L-H). The SEM
observations were made at 3 mm from the notch tip which corresponds to the zone of
variation of the load amplitude(7 to 12kN and 12 to 7 kN). Other observations were done at
35 mm from the notch tip. Figure 17 (a) presents the SEM observation of the fractured surface
of an unrepaired specimen subjected to step increasing loading (L-H) at 3 mm from the notch
tip at higher magnification. In this figure, a brittle-ductile transition fracture, cleavage and
dimple fracture types are observed. In Figure 17 (b), we can clearly observe the fatigue
striations. These striations were formed as a result of non-crystallographic growth of the
fatigue crack, and can provide information about the crack growth direction and crack growth
rate. Figure 18 presents SEM observation of the fracture surface at 3 mm from the notch tip of
specimen subjected to step increasing fatigue load. A brittle fracture by cleavage and ductile
14

fracture by formation of dimples are observed, indicating a transition from brittle to ductile
fracture.

Figure 19 shows the fracture surface of an unrepaired Al 2024-T3 specimen under step
increasing loading. The SEM observation was taken at 35 mm from the notch tip and it shows
exclusively a ductile fracture by dimples growth. This is because as the load and crack length
increase the plastic strain around the crack tip becomes higher which induce a ductile fracture.
However, at the same distance from the notch tip, Figure 20 shows the presence of dimple and
cleavage fracture for repaired specimens. This means that the stress level is lower for the
repaired case because of the stress transfer from the repaired plate toward the composite
patch.

Figure 21 presents SEM observation, at 35 mm from the notch tip, of a fractured surface of an
unrepaired specimen subjected to step decreasing fatigue load. This figure shows a ductile
fracture with higher dimple. The plasticity strains in this case are significant which explain the
retardation effect. The same behavior can be observed for the equivalent repaired specimen
(Figure 22) but the dimples are smaller for this case. The plasticity is reduced by the bridging
effect of the composite patch and consequently the crack retardation is diminished by the
patch repair.

4. Conclusion
In this work, we have studied the effect of the load amplitude change pre and post bonded
patch repair on the fatigue of life cracked Al 2024 specimen.

15

We found that an increase of the fatigue load amplitude after a patch repair overcomes most
of its crack retardation for both Al 2024 T3 and Al 7075 T6 materials. Nonetheless, a
decrease of the fatigue load amplitude after the patch repair stops the crack growth for a
considerable number of cycles then induces a lower crack growth than the unpatched case for
Al 7075 T6 material.

A crack generated in Al 2014 T3 material under a given fatigue load amplitude has grown
faster after a bonded patch repair once subjected to lower cyclic amplitude loading. In this
case, the patch repair plays an adverse role and shorten the structure residual life.

The main explanation of the crack growth acceleration under combined patching and load
amplitude reduction for the Al 2024 T3 is attributed to the patch bridging effect of the plastic
zone at the crack tip. The effect of plastic zone causing crack growth retardation is thus
nullified causing the obtained crack growth acceleration.

The reduction of the load amplitude difference between the load block, pre and post patching
has shown a consistent reduction of the bridging effect, which confirms our finding.

Acknowledgement
The authors extend their appreciation to the Deanship of Scientific Research at King Saud
University for funding the work through the research group No. RGP-VPP-035.

References
1.

Baker, A., N. Rajic, and C. Davis, Towards a practical structural health monitoring
technology for patched cracks in aircraft structure. Composites Part A: Applied
Science and Manufacturing, 2009. 40(9): p. 1340-1352.

16

2.

3.
4.

5.
6.

7.
8.

9.
10.

11.

12.
13.

14.

15.
16.
17.
18.
19.

20.

Mall, S. and D. Conley, Modeling and validation of composite patch repair to cracked
thick and thin metallic panels. Composites Part A: Applied Science and
Manufacturing, 2009. 40(9): p. 1331-1339.
Baker, A. and R. Jones, Bonded repair of aircraft structures, 1988. Martinus Nijhoff,
Dordrecht, 1990.
Paepegem, W.V. and J. Degrieck, Effects of load sequence and block loading on the
fatigue response of fiber-reinforced composites. Mechanics of Advanced Materials
and Structures, 2002. 9(1): p. 19-35.
Sadananda, K., et al., Analysis of overload effects and related phenomena.
International Journal of Fatigue, 1999. 21: p. S233-S246.
Pompetzki, M., T. Topper, and D. DuQuesnay, The effect of compressive underloads
and tensile overloads on fatigue damage accumulation in SAE 1045 steel.
International Journal of Fatigue, 1990. 12(3): p. 207-213.
Topper, T. and M. Yu, The effect of overloads on threshold and crack closure.
International Journal of Fatigue, 1985. 7(3): p. 159-164.
Gamstedt, E.K. and B. Sjgren, An experimental investigation of the sequence effect
in block amplitude loading of cross-ply composite laminates. International Journal of
Fatigue, 2002. 24(2): p. 437-446.
Nolting, A., P. Underhill, and D. DuQuesnay, Variable amplitude fatigue of bonded
aluminum joints. International Journal of Fatigue, 2008. 30(1): p. 178-187.
Sarfaraz, R., A.P. Vassilopoulos, and T. Keller, Variable amplitude fatigue of
adhesively-bonded pultruded GFRP joints. International Journal of Fatigue, 2013. 55:
p. 22-32.
Erpolat, S., et al., A study of adhesively bonded joints subjected to constant and
variable amplitude fatigue. International Journal of Fatigue, 2004. 26(11): p. 11891196.
Cheuk, P., et al., Fatigue crack growth in adhesively bonded composite-metal doublelap joints. Composite Structures, 2002. 57(1): p. 109-115.
Skorupa, M., Load interaction effects during fatigue crack growth under variable
amplitude loadinga literature review. Part I: empirical trends. Fatigue & Fracture of
Engineering Materials & Structures, 1998. 21(8): p. 987-1006.
Skorupa, M., Load interaction effects during fatigue crack growth under variable
amplitude loadinga literature review. Part II: qualitative interpretation. Fatigue &
Fracture of Engineering Materials & Structures, 1999. 22(10): p. 905-926.
Khan, S.M., et al., Analysis and Repair of Crack Growth Emanating from V-notch
under Stepped Variable Fatigue Loading. Procedia Engineering, 2014. 74: p. 151-156.
Baker, A., Bonded composite repair of fatigue-cracked primary aircraft structure.
Composite Structures, 1999. 47(1): p. 431-443.
Seo, D.-C. and J.-J. Lee, Fatigue crack growth behavior of cracked aluminum plate
repaired with composite patch. Composite Structures, 2002. 57(1): p. 323-330.
Sun, C., J. Klug, and C. Arendt, Analysis of cracked aluminum plates repaired with
bonded composite patches. AIAA journal, 1996. 34(2): p. 369-374.
Poole P, Graphite/Epoxy Patching Efficiency Studies, Chapter 13, Advances in the
Bonded Composite Repair of Metallic Aircraft Structure, A. Baker, L. R. F. Rose and
Jones R., Elsevier Applied Science Publishers, 2002. ISBN 0-08-042699-9.
Baker AA., Callinan RJ., Davis MJ., Jones R. and Williams JG., Repair of Mirage III
aircraft using BFRP crack patching technology, Theoretical and Applied Fracture
Mechanics, 2, 1, 116,1984.

17

21.

22.
23

24

25.

26.

27.

28

29.

30.
31.
32.

33.

34.
35.

36.

37.

Butkus LM., Gaskin JA., Greer JM., Jr., Guijt CM., Jacobs NJ., Kelly DF., Mazza JJ.,
Bonded Boron Patch Repair Evaluation: Final Report, USAFA-TR-2007-06, October
2007.
Baker A.A. Bonded composite repair of fatigue-cracked primary aircraft structure.
Compos Struct 1999;47(1):43143.
Raizenne, M.D., Heath, J.B.R. and Benak, T.). TTCP PTP-4 Collaborative test
program - variable amplitude loading of thin metallic materials repaired with
composite patches, Laboratory Technical Report, LTR-ST-I 662, National
Aeronautical Establishment, Ottawa, Canada., 1988
C.H. WANG, Fatigue crack growth analysis ofrepaired structures, Advances in the
Bonded, chapter 12, Composite Repair of Metallic Aircraft Structure, A.Baker, L. R.
F. Rose and Jones R., Elsevier Applied Science Publishers, 2002. ISBN 0-08042699-9
Poole P, Graphite/Epoxy Patching Efficiency Studies, Chapter 15, Advances in the
Bonded Composite Repair of Metallic Aircraft Structure, A. Baker, L. R. F. Rose and
Jones R., Elsevier Applied Science Publishers, 2002. ISBN 0-08-042699-9.
Molent, L., Callinan, R.J. and Jones, R.), Structural aspects of the design of an all
boron/epoxy reinforcement for the F-IlIC wing pivot fitting - Final report.
Aeronautical Research Laboratory, Research Report 1, ARL-RR-I, November 1992.
See also Composite Structures, 1989, 11: . 57-83.
Baker AA., Structural Health Monitoring of a Bonded Composite Patch Repair on a
FatigueCracked F-111C Wing, DSTO-RR-0335, March 2008.
Walker KF. and Rose LRF., Case History F-111 Lower Wing Skin Repair
Substantiation, Chapter 27 , Advances in the Bonded Composite Repair of Metallic
Aircraft Structure, A. Baker, L. R. F. Rose and Jones R., Elsevier Applied Science
Publishers, 2002. ISBN 0-08042699-9.
Albedah, A., et al., Experimental analysis of the fatigue life of repaired cracked plate
in aluminum alloy 7075 with bonded composite patch. Engineering Fracture
Mechanics, 2015.145: p.210-220.
Standard, A., E647-86. 1986 Annual Book of ASTM Standards, ASTM, Philadelphia,
PA, 1986. 3: p. 714-736.
Mall S., Schubbe J. J., Bonded Composite Patch Geometry Effects on Fatigue Crack
Growth in Thin and Thick Aluminum Panels, SL, 2009, 2, p.25-47,
Schweinberg,W., Jansen, R. and Fiebig, J., Advanced composite repairs of the C-141
wing structure,Intl. Symp. on Composite Repair of Aircraft Structures in concert with
ICCM-IO, August, 1995
Klemczyk, C. and Belason, E.B. , Analysis of Maximum Stresses Associated with a
Boron/Epoxy Doubler Bonded to Aluminum, Boeing Report under contract 61 17110A3397R4, January, 1994
Specification, B.P., Surface preparation of materials for adhesive bonding. BPS FW,
1972. 4352.
L.J. Hart-Smith, Recent expansions in the capabilities of crack patching, chapter 8,
Composite Repair of Metallic Aircraft Structure, A.Baker, L. R. F. Rose and Jones R.,
Elsevier Applied Science Publishers, 2002. ISBN 0-08042699-9.
Ouinas, D., et al., Comparison of the effectiveness of boron/epoxy and graphite/epoxy
patches for repaired cracks emanating from a semicircular notch edge. Composite
Structures, 2007. 80(4): p. 514-522.
Borrego, L., J. Ferreira, and J. Costa, Partial crack closure under block loading.
International Journal of Fatigue, 2008. 30(10): p. 1787-1796.
18

38.
39.

Toribio, J., et al., Effect of sudden load decrease on the fatigue crack growth in cold
drawn prestressing steel. International Journal of Fatigue, 2015. 76: p. 53-59.
L. Molent, S.A. Barter, A comparison of crack growth behavior in several full-scale
airframe fatigue tests, International Journal of Fatigue, 2007,9,p.. 10901099

19

Table 1Chemical composition of the Aluminum alloy 2024-T3(in wt. %)

Atom

Al

Cr

Cu

Fe

Mg

Mn

Si

Ti

Zn

Other

Mass
content

90.794.7

Max
0.1

3.84.9

Max
0.5

1.21.8

0.30.9

Max
0.5

Max
0.15

Max
0.25

Max
0.15

Figure 1Specimen details

20

Figure 2 Configuration of repaired specimen assembly

Figure 3 Photo taken during the fatigue test

21

Unrepaired
Repaired

(a)

Repaired
Unrepaired

(b)
Figure 4 Variable amplitude loading sequence of two blocks
(a) Decreasing block (H-L) (b) Increasing block (L-H)

Table 2 Details of load and patch configurations

Load Case

7kN

12kN

7-12kN

12-7kN

21

Pre-patch Load

7kN

12kN

7kN

12kN

Post-patch Load

7kN

12kN

12kN

7kN

Patched at crack
length

3mm

3mm

3mm

3mm

8 Plies,
50x50mm

9 Plies,
50x50mm

10 Plies,
50x50mm

11 Plies,
50x50mm

2024 and 7075

from Ref. [19]

2025 and 7075

from Ref. [19]

2026 and 7075

from Ref. [19]

2027 and 7075

from Ref. [19]

Patch
Configuration
Material

50
R0-Unrepaired

45

R0-Repaired

40
Crack length (mm)

R0.1-Unrepaired

35

R0.1-Repaired

30
25
20
15
10
5
0
0

10000

20000
30000
40000
Number of Cycles

50000

60000

70000

Figure 5 Fatigue life of repaired and unrepaired Al 2024-T3 under increasing blocks (L-H/712kN)

23

50
L-H- Unrepaired

45

L-H -Repaired

Crack length (mm)

40

Cons 12kN- Unrepaired

Cons 12kN- Repaired

35

Cons 7kN- Unepaired

30

Cons 7kN- Repaired

25
20
15
10
5
0
0

20000

40000

60000 80000 100000 120000 140000 160000

Number of Cycles

Figure 6 Comparison of repaired and unrepaired Al 2024-T3 for increasing block loading (LH/ 7-12kN) and constant loads at R=0.1

1.E-04

L-H, UR, R0
L-H, UR, R0.1
L-H, Repaired, R0
L-H, Repaired, R0.1

da/dN (m/cycles)

1.E-05

Const 7kN, Unrepaired, R0.1

Const 7kN, Repaired, R0.1

1.E-06

1.E-07

1.E-08
0

0.005

0.01
Crack length (m)

0.015

0.02

Figure 7 Crack growth rate of 2024-T3 under increasing loading blocks

24

50
R0-Unrepaired

Crack length (mm)

45

R0-Repaired

40

R0.1-Unrepaired

35

R0.1-Repaired

30
25
20
15
10
5
0
0

100000

200000

300000

400000

Number of Cycles

Figure 8 Fatigue life of repaired and unrepaired Al 2024-T3 under decreasing blocks (H-L/127)

1.E-04

da/dN (m/cycles)

1.E-05
1.E-06
H-L, Unrepaired, R0

1.E-07

H-L, Unrepaired, R0.1

H-L, Repaired, R0
H-L, Repaired, R0.1

1.E-08

Const 12kN, Unrepaired, R0.1

Const 12kN, Repaired, R0.1

1.E-09
0

0.005

0.01
0.015
Crack length (m)

0.02

0.025

25

Figure 9 Crack growth rate of 2024-T3 under decreasing loading blocks

50
H-L- Unrepaired

Crack length (mm)

45

H-L- Repaired

40

Cons 12kN- Unrepaired

35

Cons 12kN- Repaired

30

Cons 7kN- Unrepaired

25

Cons 7kN- Repaired

20
15
10
5
0
0

100000

200000

300000

400000

Number of Cycles

Figure 10 Comparison of repaired and unrepaired Al 2024-T3 for decreasing block loading
(H-L/ 12-7 kN) and constant loads at R=0.1

50

Crack length (mm)

45
40
35
30
25
20

F1 = 12 kN

15

F1 = 11kN

10

F1 = 10 kN
F1 = 9 kN

5
0
0

50000

100000

150000

200000

250000

Number of cycles

26

Figure 11 Effect of the amplitude of the first block on the fatigue life of repaired Al 2024-T3
for decreasing blocks (H-L) case at R=0.1

Crack length (mm)

50

L-H- Repaired

45

L-H- Unrepaired

40

Cons 12kN- Unrepaired

Cons 12kN- Repaired

35

Cons 7kN- Unrepaired

30

Cons 7kN- Repaired

25
20
15
10
5
0
0

20000

40000
60000
No. of Cycles

80000

100000

Figure 12 Comparison of repaired and unrepaired Al 7075-T6 for increasing block loading (LH/ 7-12kN) at R=0.1

27

H-L- Repaired
H-L- Unrepaired

50

Cons 12kN- Unrepaired

45

Cons 12kN- Repaired

Cons 7kN- Unrepaired

Crack length (mm)

40

Cons 7kN- Repaired

35
30
25
20
15
10
5
0
0

20000

40000

60000 80000 100000 120000 140000 160000

No. of Cycles

Figure 13 Comparison of repaired and unrepaired Al 7075-T6 for decreasing block loading
(H-L/ 12-7kN) at R=0.1
50

2024 - Unrepaired

Crack length (mm)

45

2024 - Repaired
7075 - Unrepaired

40

7075 - Repaired

35
30
25
20
15
10
5
0
0

100000

200000
Number of cycles

300000

400000

Figure 14 Fatigue life of specimens in aluminum alloys 2024-T3 and 7075-T6 under
decreasing blocks (H-L) at R=0.1

28

700

True Stress (MPa)

600
500
400

7075-T6
2024-T3

300
200
100
0
0

0.025

0.05
0.075
True Strain (mm/mm)

0.1

0.125

0.15

Crack length (mm)

Figure 15 Stress-strain curves of Al 2024-T3 and Al 7075-T6

50
45
40
35
30
25
20
15
10
5
0

2024 - Unrepaired
2024 -Repaired
7075- Unrepaired
7075- -Repaired

20000

40000
Number of cycles

60000

Figure 16 Fatigue life of specimens in aluminum alloys 2024-T3 and 7075-T6 under
increasing blocks (L-H) at R=0.1

29

Table 3 Summary of fatigue life

Load

Const. 7kN
Const.
12kN
L-H
(7-12kN)
H-L
(12-7kN)

Config. At
R=0.1

Unrepaired
Repaired
Unrepaired
Repaired
Unrepaired
Repaired
Unrepaired
Repaired

Al 2024-T3
No. of
cycles to
failure
(N)
77590
133867
17210
28036
54429
57579
344542
190950

Al 7075-T6

%
increase
in fatigue
life
72.53125
62.90529
5.787356
-44.5786

No. of
cycles to
failure (N)
40172
89980
13010
22149
22470
32630
77300
135060

%
increase
in fatigue
life
123.9869
70.24596
45.21584
74.72186

(a)

30

(b)
Figure 17 SEM observation of the fractured 2024-T3 surface at 3 mm of the notch tip for
unrepaired specimens after failure subjected the increasing fatigue blocks (7kN-12kN): (a)
transition brittle-ductile fracture (b) fatigue striations

31

Figure 18 SEM observation of the fracture surface at 3 mm of the notch tip for repaired 2024T3 specimens after failure subjected the increasing fatigue blocks (7kN-12kN)

Figure 19 SEM observation of the fracture surface at 35 mm of the notch tip for unrepaired
2024-T3 specimens after failure subjected the increasing fatigue blocks (7kN-12kN)
32

Figure 20 SEM observation of the fracture surface at 35 mm of the notch tip for repaired
2024-T3 specimens after failure subjected the increasing fatigue blocks(7kN-12kN)

Figure 21 SEM observation of the fracture surface at 35 mm of the notch tip for unrepaired
2024-T3 specimens after failure subjected the decreasing fatigue blocks (12kN-7kN)
33

Figure 22 SEM observation of the fracture surface at 35 mm of the notch tip for repaired
2024-T3 specimens after failure subjected the decreasing fatigue blocks (12kN-7kN)

34

Highlights

an increase of the maximum fatigue load after repair overcomes most of its crack
retardation
A decrease of the maximum fatigue load after the repair induces a lower crack growth
than the unpatched case for Al 7075
A crack generated in Al 2014 T3 material under a given fatigue load amplitude has
grown faster

35