Composite Structures 130 (2015) 1–8

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Optimisation study of tapered scarf and stepped-lap joints in composite

repair patches
Hamza Bendemra a,⇑, Paul Compston a, Phillip J. Crothers b
a
Research School of Engineering, CECS, Australian National University, Canberra, ACT 2601, Australia
b
Boeing Research and Technology, Melbourne, Victoria 3207, Australia

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

Article history: Stepped-lap repairs of composite structures can offer an alternative that is easier to perform and less
Available online 24 April 2015 time-consuming to produce than the widely used tapered scarf repair. However, the design of
stepped-lap joints must be carefully investigated in order to avoid generating stress concentration
Keywords: regions. This study investigates the influence of joint parameters on peak stresses in the adhesive bond-
Composite repair line in tapered scarf and stepped-lap joints. Linear finite element analysis was performed to conduct a
Scarf parametric study with focus on six joint design parameters: ply thickness, adhesive thickness, taper
Stepped-lap
angle, stacking sequence, overply layup, and overply lap length. Results showed that tapered scarf and
Adhesive joints
Stress concentration
stepped-lap joints have a strong sensitivity to ply thickness, taper angle, and stacking sequence. The
Finite element analysis introduction of overplies provided protection and stiffness at joint tips, and a critical overply lap length
was identified. The location of 0° plies in the composite laminates was highlighted as an important factor.
The analysis was then extended to three-dimensional FE models for verification. In conclusion, results
showed that high stress concentration in stepped-lap joints can be mitigated with the introduction of
overplies and appropriate changes in joint design parameters to reduce stress peaks at joint tips and step
corners.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction identical material properties can be considered to be uniform [4].

The increased use of composite materials in aircraft structural
parts has led to further interest in efficient composite repair meth-
ods. Adhesively bonded repair patches are considered to be
mechanically efficient and can be quickly applied depending on
the size of the repair, and the proficiency of the repair technician.
Several bonding joints can be used: tapered scarf, stepped-lap, and
overlap. In aerospace, where there often is a need for a flush sur-
face to meet aerodynamic requirements, a tapered scarf or
stepped-lap repair is applied.
In recent years, adhesively bonded repair patches have been
extensively studied [1–3]. However, the use of fibre-reinforced
composite materials adds to the complexity of designing and vali-
dating repair designs due to their anisotropic nature. The mechan-
ical integrity of a repair patch is dependent on the stress
distribution and stress peaks in the adhesive bondline between
the patch and the parent structure. The stress distribution along
an adhesive joint between two homogeneous adherends with
⇑ Corresponding author.
E-mail addresses: hamza.bendemra@anu.edu.au (H. Bendemra), paul.compston@
anu.edu.au (P. Compston), phillip.j.crothers@boeing.com (P.J. Crothers).
However, the use of composite laminates featuring different fibre
orientations through the thickness results in a varying stress distri-
bution and stress concentration in the adhesive bondline [5]. In
particular, shear stress and peel (i.e. normal) stress can signifi-
cantly vary through the thickness of the laminate [6,7].
The design of tapered repairs has typically been based on a two
dimensional (2D) approach with the results being implemented in
real-world three dimensional (3D) repair scenarios. Although
tapered scarf joint failure predictions based on 2D models can be
consistent with 3D applications, the application of 2D design
methodology can result in overly conservative design guidelines
and lead to excessive material removal [8,9], particularly when
investigating compressive loading cases. Analysis methods used
by engineers to assess joint design can be analytical [10–12] or
based on a finite element (FE) modelling approach [13–15].
Several certification frameworks for composite repairs have also
been proposed [16–18].
Flush bonded repairs, such as tapered scarf and stepped-lap
joints, are typically preferred for repairs of primary composite
structures as they are more structurally efficient than external or
bolted repair patches. However, tapered scarf repairs can be diffi-
cult to apply depending on the repair configuration. Stepped-lap

http://dx.doi.org/10.1016/j.compstruct.2015.04.016
0263-8223/Ó 2015 Elsevier Ltd. All rights reserved.
2 H. Bendemra et al. / Composite Structures 130 (2015) 1–8
joints offer an alternative that is easier to perform and less time-
consuming to produce [19]. However, the design of stepped-lap
joints must be carefully investigated in order to avoid generating
stress concentration at step corners.
Hart-Smith [20] provided early analytical models for analysis of
tapered scarf and stepped-lap joints which considered potential
adherend stiffness- and thermal-mismatch effects. Gunnion and
Herszberg [21] performed a parametric study where they investi-
gated the influence of tapered scarf joint parameters on the stress
distribution in the adhesive bondline. Their investigation mainly
focused on stress distribution and average stresses along the cen-
treline of the adhesive layer. However, local peak stresses may also
contribute to failure if stress limits are exceeded before the average
stress reaches allowable stress levels. As was noted by Gunnion
and Herszberg, further investigation in local peak stresses is
needed, particularly as local yielding of the adhesive may affect
the fatigue and creep performance of the bonded joint.
The aim of this paper is to present an investigation of the influ-
ence of joint parameters on peak stresses in the adhesive bondline
in tapered scarf and stepped-lap repairs. FE models of tapered scarf
and stepped-lap joints were built, and peak peel and shear stresses
for several joint configurations were compared. A parametric study
was performed with focus on six joint design parameters: ply
thickness, adhesive thickness, taper angle, stacking sequence, over-
ply layup, and overply lap length. The analysis was then extended
to three-dimensional FE models to further investigate stress
concentration regions across the adhesive bondline.
2. Geometry and material properties
Two-dimensional CAD models for the tapered scarf and
stepped-lap joints were built as shown in Fig. 1. The models con-
sisted of eight-layered composite laminates for the parent struc-
ture and the repair patch bonded by an adhesive placed along a
taper. Individual plies were discretely modelled, and a ply-by-ply
match was assumed between the parent structure and the repair
patch. The radius for the top layer of the repair patch was set at
150 mm and the length of the top layer in the parent structure
was set at 50 mm. In the stepped-lap joint, for comparison
Fig. 1. Design parameters for (a) tapered scarf joint, and (b) stepped-lap joint.
purposes, the taper angle (a) was defined as the angle between
the top surface of the joint and the centreline of the adhesive bond-
line, while adhesive thickness (ta) was defined as the distance
between the patch and the parent structure, as shown in Fig. 1.
The material for the parent structure and the repair patch was car-
bon/epoxy AS4 3501–6, and the adhesive material was FM300.
Material properties for AS4 3501–6 and FM300 are shown in
Table 1 [21].
3. Methodology
3.1. Parametric model
A parametric study was undertaken to investigate the influence
and sensitivity of joint parameters on stresses in the adhesive in
tapered scarf and stepped-lap joints. Peak shear and peak peel
stresses in the adhesive bondline were evaluated with respect to
changes in six design parameters: ply thickness (tp ), adhesive
thickness (ta ), taper angle (a), stacking sequence, overply layup,
and overply lap length. Design parameters were defined for the
tapered scarf and stepped-lap joints as shown in Fig. 1, with
baseline values as listed in Table 2. A parametric study was then
performed, as listed in Table 3. Except when under investigation,
the baseline parameters listed in Table 2 were used.
3.2. Finite element model
FE models for the tapered scarf and stepped-lap joints were
built in ANSYS. The models featured three main sections: the par-
ent structure, the adhesive layer, and the repair patch. Shell ele-
ments were used which featured four-node elements with six
degrees of freedom at each node (translations in the x, y, and z
directions, and rotations about the x, y, and z axes). The model also
included contact elements between the adhesive and the compos-
ite plies. Each composite ply and the adhesive layer had four ele-
ments through the thickness. Face meshing was applied across
the composite plies and the adhesive layer to enhance aspect ratio
and mesh uniformity. Local coordinate systems were created for
each fibre orientation present in the laminate (i.e. 0°, +45°, 45°,
 H. Bendemra et al. / Composite Structures 130 (2015) 1–8 3

90°), and for the adhesive layer in line with the bondline. In an
effort to validate the chosen FE methodology, a verification study
was undertaken by replicating a previous study by Gunnion and
Herszberg [21] using identical shell elements and meshing strategy
that were used in this study. Gathered results successfully con-
curred with the published results.
Peel stress was considered along the Y-axis, and shear stress
was taken in the X-Y plane, as shown in Fig. 2. The adhesive mate-
rial was assumed to remain elastic. Linear static analyses only were
performed. In terms of boundary conditions, the laminate was con-
sidered to be fully fixed on one side, and with nodes only free in the
X-direction on the outer edge where a uniform pressure was
applied, as shown in Fig. 1. Peak shear and peak peel stress values
gathered were normalised to the applied stress for comparison
purposes.
Table 1
Material properties for AS4 3501–6 and FM 300.
Property AS4 3501–6
E1 (GPa) 128
E2 = E3 (GPa) 13
G12 = G23 = G33 (GPa) 7.2
v12 0.30
4. Results
4.1. Effect of ply thickness
This section investigates the influence of ply thickness on peak
shear and peak peel stresses in the adhesive bondline in tapered
scarf and stepped-lap joints. Note that changes to ply thickness
were applied equally to the parent structure and the composite
patch to preserve a ply-by-ply match. Ply thickness was varied
from 0.13 to 0.39 mm. According to plotted results in Fig. 3, in both
tapered scarf and stepped-lap joints, normalised peak shear and
peak peel stresses decrease as ply thickness increases. Increasing
ply thickness from 0.13 to 0.39 mm, reduces normalised peak peel
stress by 81% and 53% in the tapered scarf joint and stepped-lap
joint respectively. Peak shear stress shows a similar trend with a
77% and 54% decrease in the tapered scarf joint and stepped-lap
joint respectively. Results also highlight that the tapered scarf joint
exhibit a significantly higher reduction in peak stress than
stepped-lap joints when increasing ply thickness. In both tapered
FM 300 scarf and stepped-lap joints, the increase in ply thickness trans-
1.02 lates into an increased load bearing capacity by the composite lam-
inate and hence a reduction in peak stress in the adhesive.
0.39
0.30
4.2. Effect of adhesive thickness

Adhesive thickness (ta ) was increased from 0.13 to 0.39 mm,

Table 2 while ply thickness (t p ) was kept at 0.13 mm. Results were plotted
Baseline model properties. according to the (t a =t p ) ratio, as shown in Fig. 4. According to gath-
Property Value
ered results, tapered scarf joints show a significant increase in nor-
malised peak shear and peak peel stress as adhesive thickness
Ply thickness tp (mm) 0.13
Adhesive thickness ta (mm) 0.13
increases. Indeed, as the adhesive thickness is increased from
Taper angle a° 3 0.13 to 0.39 mm, peak peel and peak shear stress in tapered scarf
Number of layers 8 joints show an increase of 161% and 170% respectively. The load
Repair patch layup [45/0/ 45/90]S being transferred and carried by the adhesive layer increases with
Parent structure layup [45/0/ 45/90]S
its thickness resulting in stress concentration forming in the
Overply thickness (mm) 0 (i.e. no overply)
bondline.
However, stress peaks in the stepped-lap joint show low sensi-
tivity to adhesive thickness. Indeed, an increase of 26% in peak peel
Table 3
stress, and a decrease of 10% in peak shear stress, was seen as the
Parametric model. (ta =tp ) ratio was increased from 1 to 3 in stepped-lap joints. Due
the geometrical differences between tapered scarf and stepped-
Parameter Values
lap joints, and how adhesive thickness was defined in this study,
Ply thickness tp (mm) 0.13–0.39 the amount of adhesive material in the stepped-lap joint only
Adhesive thickness ta (mm) 0.13–0.39
Taper angle a(°) 3–12
experienced small changes as the adhesive thickness is increased.
Stacking sequence [45/0/ 45/90]S, [ 45/90/+45/0]S Indeed, the step length is modified to compensate for the increased
[0/90]2S, [90/0]2S adhesive thickness in order to maintain the taper angle, which
Overply lap length (mm) 0–15 explains the results gathered. Fig. 5 shows the adhesive surface
Overply layup [0/0], [0/90], [90/0], [45, 45]
area as a function of adhesive thickness.

Fig. 2. Finite element models of (a) tapered scarf joint, and (b) stepped-lap joint, with highlighted adhesive joint.
4 H. Bendemra et al. / Composite Structures 130 (2015) 1–8

Fig. 3. Normalised peak peel and shear stress in the adhesive bondline for ply thickness ranging from 0.13 to 0.39 mm.

Fig. 4. Normalised peak peel and shear stress in the adhesive bondline for (ta =tp ) ratio ranging from 1 to 3.

Fig. 5. Normalised adhesive surface area (2D FE model) for adhesive thickness ranging from 0.13 to 0.39 mm.

4.3. Effect of taper angle
Taper angle was varied from 3° to 12°, while other parameters
were maintained as per the baseline model. As with the adhesive
thickness, the geometrical differences between tapered scarf and
stepped-lap joints result in a different amounts of adhesive mate-
rial in each joint for a given taper angle. However, the increase of
adhesive material was similar in between both designs as can be
seen in Fig. 7. The decrease in taper angle results in decreased
adhesive bondline peak stresses as shown in Fig. 6, but it also
results an increased loss of pristine material. As the taper angle
is increased from 3° to 12°, normalised peak peel and peak shear
stresses show an increase of 200% in tapered scarf joints, and an
increase of 232 and 123% respectively in stepped-lap joints.
4.4. Effect of stacking sequence
Identical stacking sequences were used in the parent structure
and the composite repair patch. Gathered results, as plotted in
Fig. 8, highlight that stacking sequence does have an impact on
peak peel and peak shear stresses in the adhesive. Rotating fibre
angles by 90° from [ 45/0/45/90]S to [45/90/ 45/0]S resulting in
 H. Bendemra et al. / Composite Structures 130 (2015) 1–8 5

Fig. 6. Normalised peak peel and shear stress in the adhesive bondline for taper angle ranging from 3° to 12°.

Fig. 7. Normalised adhesive surface area (2D FE model) for taper angle ranging from 3° to 12°.

Fig. 8. Normalised peak peel and shear stress in the adhesive bondline for varying laminate stacking sequences.

a decrease in peak peel stress of 12% and 44% in the tapered scarf
joint and stepped-lap joint respectively, and a peak shear stress
decrease of 16% and 41% in the tapered scarf joint and stepped-
lap joint respectively. Similarly, peak stress decreases in the
tapered scarf joint and the stepped-lap joint decreased when the
stacking sequence was changed from a [0/90]2S layup to a [90/
0]2S layup. Results suggest that the location of 0° plies has a signif-
icant impact on the magnitude of stress concentration, with 0°
plies placed near or at at the free edge resulting in localised peak
stress regions. These results concur with previously published
results [22,23].
4.5. Effect of overply
The overply consisted of two plies with layup [ 45/+45], and ply
thickness of 0.13 mm. The overplies were assumed to be perfectly
bonded to the top layer of the repair patch. Loading and boundary
conditions were only applied to the original repair patch. The influ-
ence of the overply lap length was investigated with lap length var-
ied from 0 mm (i.e. the overply meets the outer boundary of the
adhesive) to 15 mm. Results show that peak peel stress shows a sig-
nificant increase in the tapered scarf joint when no lap length is
applied, as shown in Fig. 9. Indeed, as the adhesive is directly
6 H. Bendemra et al. / Composite Structures 130 (2015) 1–8

Fig. 9. Normalised peak peel and shear stress in the adhesive bondline for overply lap length ranging from 0 to 15 mm.

Fig. 10. Normalised peak peel and shear stress along adhesive bondline for varying overply stacking sequences.

Fig. 11. Exploded view of flushed bonded repair (3D CAD model).

adjacent with the overply, a strain incompatibility is introduced in 5. Three-dimensional analysis

that area resulting in an increased peak stress. The impact of over-
ply lap length was less significant in the stepped-lap joint. Also, a
critical lap length was clearly identified at 5 mm with little changes
in peak stresses when lap length is further increased. The influence
of overply layups was considered with several layups applied to the
overply being investigated ([0/0], [0/90], [90/0], and [ 45/45]).
According to plotted results in Fig. 10, the introduction of a [0/0]
overply results in a decrease in peak shear and peak peel stress of
19% in the tapered scarf joint, and 9% and 23% respectively in the
stepped-lap joint. However, the stacking sequence (i.e. fibre orien-
tations of each ply) in the overply showed little impact on the
resulting peak stresses in the adhesive, as shown in Fig. 10.
A three-dimensional FE study for the tapered scarf and stepped-
lap joints were performed in ANSYS to verify 2D FE results.
Similarly to previously discussed 2D models, 3D models consisted
of eight-layered composite laminates bonded by an adhesive
placed along a taper, as shown in Fig. 11. Individual plies were dis-
cretely modelled, and a ply-by-ply match was kept between the
parent structure and the repair patch. The radius for the top layer
of the repair patch was set at 150 mm, and the length of the top
layer in the parent structure was set at 50 mm. Four models were
investigated: tapered scarf, tapered scarf with overplies, stepped-
lap, and stepped-lap with overplies. Baseline joint parameters were
 H. Bendemra et al. / Composite Structures 130 (2015) 1–8 7

Fig. 12. Peel (left) and shear (right) adhesive stress distribution in tapered scarf joint (3D model).

Fig. 13. Peel (left) and shear (right) adhesive stress distribution in stepped-lap joint with overplies (3D model).

used in all models, with overply lap length set at 5 mm and [ 45/
+45] stacking sequence. The material for the composite plies and
the adhesive was carbon/epoxy AS4 3501–6 and FM300 respec-
tively. Local coordinate systems were created for each fibre orien-
tation present in the laminate (i.e. 0°, +45°, 45°, 90°), and for the
adhesive layer in line with the bondline. Higher-order 3D 10-node
elements were used. A patch conforming method was imple-
mented to created a mesh using tetrahedron elements with
enhanced aspect ratio. A fine mesh was applied to the adhesive
layer, and all 3D FE models consisted of about 550,000 elements
resulting in quite an expensive computational effort required. A
verification study was undertaken by replicating previous work
from Gunnion and Herszberg [21] where they focused on the stress
distribution in the adhesive bondline centreline at 30° intervals.
Using identical 3D solid elements and meshing strategy that were
used in this study, gathered results successfully concurred with
published results.
Linear static analyses only were performed. Peel stress was con-
sidered along the Z-axis, and shear stress was taken in the X-Y
plane. The adhesive material was assumed to remain elastic. The
laminate was considered to be fully fixed on one side, and with
nodes only free in the X-direction on the outer edge where a uni-
form pressure was applied along the X-axis. Peak shear and peak
peel stress values gathered were normalised to the applied stress.
As can be seen in Figs. 12, and 13, the addition of overplies pro-
vided added stiffness and protects joint tips which resulted in a
decrease in peak stress, which agrees with observations made in
the two-dimensional models. Also, the location of stress peaks at
join tips and adjacent to 0° plies concurs with results from the
two-dimensional analysis.
6. Discussion
Reducing peak stresses in the adhesive bondline thickness is
crucial to the mechanical integrity of the applied repair as peak
stresses are precursor to debonding and potential failure of the
composite patch. Results gathered from the parametric study
clearly highlight that peak stresses in the adhesive bondline in
tapered scarf joints and in stepped-lap joints are more sensitive
to certain joint parameters. Peak peel and shear stress in tapered
scarf and stepped-lap joints have shown a strong sensitivity to
ply thickness, taper angle, and stacking sequence. However,
results suggest that tapered scarf joints have a higher sensitivity
to adhesive thickness than stepped-lap joints. The introduction
of overplies provided protection and stiffness at joint tips.
However, an overply lap length must be introduced to prevent
further increasing peak stresses, particularly for tapered scarf
joints, and stacking sequence of overplies was shown to have
8 H. Bendemra et al. / Composite Structures 130 (2015) 1–8
little impact. Also, the overply lap length reached a critical lap
length over which negligible reduction in peak stress is achieved.
Gathered results have highlighted the importance of 0° plies
location in the composite laminates. Indeed, the adhesive region
adjacent to 0° plies were prone to develop stress concentration
as the majority of the load transfer, when loaded under tension,
occurs at the location near the stiffer plies. Analysis of the stress
distribution in stepped-lap joints showed that the adhesive region
placed between 0–0 plies are prone to develop stress concentra-
tion, with increased intensity if the plies are placed near free edges.
As such, a follow-up study was performed with laminate stacking
sequence with 0° placed in the centreline of the laminate (i.e. par-
ent structure and repair patch stacking sequence [-45/90/+45/0]S),
adhesive thickness in the stepped-lap joints at 0.26 mm (increased
from nominal thickness 0.13 mm), and overplies with lap length at
5 mm and stacking sequence [ 45/+45]. Compared to the baseline
tapered scarf joint, this stepped-lap joint configuration showed a
reduced increase of 10% and 20% in peak shear and peak peel stress
respectively in the adhesive bondline. This further emphasises that
stepped-lap joints can rival tapered scarf joints if appropriately
designed for stress peak reduction.
In terms of on-field application, airworthiness certification is
required for repairs applied to flight-critical structures according
to requirements set by regulatory bodies such as the Federal
Aviation Administration [24]. While the results gathered in this
study are relevant to a safe-life approach (i.e. analysis of bondline
stress without any pre-existing flaws or in-service damage), dam-
age tolerance must also be considered for a complete assessment. A
recent study by Wang et al. [19] investigated the effect of step cor-
ners on the fracture behaviour of stepped-lap joints, and found that
stepped-lap joints can be designed to meet design ultimate
strength requirements to tolerate impact damage. As such,
stepped-lap joints can in fact provide an alternative to tapered
scarf joints in terms of stress distribution and damage tolerance
requirements with carefully chosen joint design parameters.
7. Concluding remarks
This study presented an investigation of the influence of joint
parameters on peak stresses in the adhesive bondline in tapered
scarf and stepped-lap joints. Linear FE analysis was performed to
conduct a parametric study with focus on six joint design parame-
ters: ply thickness, adhesive thickness, taper angle, stacking
sequence, overply layup, and overply lap length. Results showed
that peak peel and shear stress in tapered scarf and stepped-lap
joints have a strong sensitivity to ply thickness, taper angle, and
stacking sequence. The introduction of overplies provided protec-
tion and stiffness at joint tips, and an overply lap length was intro-
duced to prevent further increasing peak stresses. The location of
0° plies in the composite laminates was highlighted as an impor-
tant factor as the adhesive region adjacent to 0° plies developed
high stress concentration. The analysis was then extended to
three-dimensional FE models for verification. In conclusion,
stepped-lap joints exhibit higher stress concentration than equiva-
lent tapered scarf joints, however results in this study showed that
this can be mitigated with the introduction of overplies and appro-
priate changes in joint design parameters to reduce stress peaks at
joint tips and step corners.
References
[1] Caminero M, Pavlopoulou S, Lopez-Pedrosa M, Nicolaisson B, Pinna C, Soutis C.
Analysis of adhesively bonded repairs in composites: damage detection and
prognosis. Compos Struct 2013;95:500–17.
[2] Katnam K, Da Silva L, Young T. Bonded repair of composite aircraft structures:
a review of scientific challenges and opportunities. Prog Aerosp Sci
2013;61:26–42.
[3] Pantelakis S, Tserpes K. Adhesive bonding of composite aircraft structures:
challenges and recent developments. Sci Chin: Phys Mech Astron
2014;57(1):2–11.
[4] Baker A. Joining and repair of aircraft composite structures. Mech Eng Trans
ME21 1996(1–2):1–59.
[5] Baker A, Chester R, Hugo G, Radtke T. Scarf repairs to highly strained graphite/
epoxy structure. Int J Adhes Adhes 1999;19(2):161–71.
[6] Adkins D, Byron Pipes R. End effects in scarf joints. Compos Sci Technol
1985;22(3):209–21.
[7] Johnson CL. Effect of ply stacking sequence on stress in a scarf joint. AIAA J
1989;27(1):79–86.
[8] Soutis C, Hu F. Failure analysis of scarf-patch-repaired carbon fiber/epoxy
laminates under compression. AIAA J 2000;38(4):737–40.
[9] Hu F, Soutis C. Strength prediction of patch-repaired CFRP laminates loaded in
compression. Compos Sci Technol 2000;60(7):1103–14.
[10] Barut A, Hanauska J, Madenci E, Ambur D. Analysis method for bonded patch
repair of a skin with a cutout. Compos Struct 2002;55(3):277–94.
[11] Duong C, A new analytical method for the analysis of a scarf or multiple-step
joint under general loading, In: American society for composites – 22nd
technical conference of the american society for composites 2007 –
composites: enabling a new era in civil aviation, vol. 2. 2007. p. 1442–1459.
[12] Pinto A, Campilho R, De Moura M, Mendes I. Numerical evaluation of three-
dimensional scarf repairs in carbon-epoxy structures. Int J Adhes Adhes
2010;30(5):329–37.
[13] Harman A, Wang C. Improved design methods for scarf repairs to highly
strained composite aircraft structure. Compos Struct 2006;75(1-4):132–44.
[14] Das M, Madenci E, Ambur D. Three-dimensional nonlinear analyses of scarf
repair in composite laminates and sandwich panels. J Mech Mater Struct
2008;3(9):1641–58.
[15] Tzetzis D, Hogg P. Experimental and finite element analysis on the
performance of vacuum-assisted resin infused single scarf repairs. Mater Des
2008;29(2):436–49.
[16] Jones R, Smith W. Continued airworthiness of composite repairs to primary
structures for military aircraft. Compos Struct 1995;33(1):17–26.
[17] Baker A. A proposed approach for certification of bonded composite repairs to
flight-critical airframe structure. Appl Compos Mater 2011;18(4):327–36.
[18] Baker A, Gunnion A, Wang J. On the certification of bonded repairs to primary
composite aircraft components. J Adhes 2015;91(1-2):4–38.
[19] Wang CH, Venugopal V, Peng L. Stepped flush repairs for primary composite
structures. J Adhes 2014;91(1-2):95–112.
[20] Hart-Smith L. Adhesive-bonded scarf and stepped-lap joints, Tech. Rep. NASA
CR 112237, Langley Research Center, Hampton: VA. 1973.
[21] Gunnion AJ, Herszberg I. Parametric study of scarf joints in composite
structures. Compos Struct 2006;75(1-4):364–76.
[22] Wang CH, Gunnion AJ. On the design methodology of scarf repairs to
composite laminates. Compos Sci Technol 2008;68(1):35–46.
[23] Breitzman T, Iarve E, Cook B, Schoeppner G, Lipton R. Optimization of a
composite scarf repair patch under tensile loading. Compos Part A: Appl Sci
Manuf 2009;40(12):1921–30.
[24] Federal Aviation Administration, AC 20-107B – Composite Aircraft Structure
(Advisory Circular), 2009.