DOT/FAA/AR-00/46

Office of Aviation Research
Washington, D.C. 20591
Repair of Composite Laminates

























December 2000

Final Report



This document is available to the U.S. public
through the National Technical Information
Service (NTIS), Springfield, Virginia 22161.





U.S. Department of Transportation
Federal Aviation Administration
NOTICE

This document is disseminated under the sponsorship of the U.S.
Department of Transportation in the interest of information exchange. The
United States Government assumes no liability for the contents or use
thereof. The United States Government does not endorse products or
manufacturers. Trade or manufacturer's names appear herein solely
because they are considered essential to the objective of this report. This
document does not constitute FAA certification policy. Consult your local
FAA aircraft certification office as to its use.





















This report is available at the Federal Aviation Administration William J.
Hughes Technical Center's Full-Text Technical Reports page:
actlibrary.tc.faa.gov in Adobe Acrobat portable document format (PDF).




Technical Report Documentation Page
1. Report No.

DOT/FAA/AR-00/46
2. Government Accession No. 3. Recipient's Catalog No.
4. Title and Subtitle

REPAIR OF COMPOSITE LAMINATES
5. Report Date

December 2000

6. Performing Organization Code


7. Author(s)

Sung-Hoon Ahn and George S. Springer
8. Performing Organization Report No.


9. Performing Organization Name and Address

Department of Aeronautics and Astronautics
Stanford University
10. Work Unit No. (TRAIS)


Stanford, CA 94305
11. Contract or Grant No.

12. Sponsoring Agency Name and Address

U.S. Department of Transportation
Federal Aviation Administration
13. Type of Report and Period Covered

Final Report
Office of Aviation Research
Washington, DC 20591
14. Sponsoring Agency Code
ACE-110
15. Supplementary Notes
The FAA William J. Hughes Technical Center COTR was Mr. Peter Shyprykevich.
16. Abstract

The effectiveness of repair of damaged fiber reinforced composite laminates was investigated; the effectiveness of repair being
assessed by the tensile failure load of the repaired laminate.

First, tests were conducted measuring the failure loads of laminates repaired either by the scarf, the uniform lap, or the stepped lap
technique. Data were generated with the following parameters having been varied: type of material of the damaged laminate, type
of repair material, scarf angle and number of external plies in scarf repair, length and number of repair plies in uniform and stepped
lap repair, moisture content of the laminate prior to repair, moisture content of the laminate after repair, test temperature, roughness
of grinding tool used in preparing the repair surface, and the temperature applied during the cure of the repair patch. The
aforementioned parameters were varied over wide ranges, and provided systematic sets of data which shed light on the influence of
each of these parameters on the effectiveness of the repair.

Second, models were developed for calculating the failure loads of composite laminates repaired by scarf and uniform lap
techniques. The models take into account anisotropy of each ply in the laminate and in the repair ply and nonelastic behavior of the
adhesive or resin interlayer between the laminate and the repair patch. On the basis of the models, two computer codes were
written for generating numerical values of the failure loads. The first code is for calculating the failure loads of laminates repaired
by the uniform lap technique, the second for calculating the failure loads of laminates repaired by the scarf technique. Failure loads
calculated by the models and the corresponding computer codes were compared with data and good agreement was found between
the results of the analysis and the tests.













17. Key Words
Bonded repairs, Composites, Analysis methods,
Parameter studies
18. Distribution Statement
This document is available to the public through the National
Technical Information Service (NTIS) Springfield, Virginia
22161.
19. Security Classif. (of this report)

Unclassified
20. Security Classif. (of this page)

Unclassified
21. No. of Pages

85
22. Price
Form DOT F1700.7 (8-72) Reproduction of completed page authorized
iii
TABLE OF CONTENTS

Page

EXECUTIVE SUMMARY xi

1. INTRODUCTION 1

2. THE PROBLEM 1

PART I. EXPERIMENTAL RESULTS 3

3. EXPERIMENTAL 3

4. TEST RESULTS 10

4.1 Type of Repair Material 11
4.2 Geometry Effects 16
4.3 Effects of Specimen Moisture Content and Test Temperature 22

4.3.1 Base Laminates Premoisturized 22
4.3.2 Specimen Moisture Content and Test Temperature 24

4.4 Effect of Sanding 25
4.5 Effect of Cure Cycle 26
4.6 Summary 29

PART II. ANALYTICAL MODELS 30

5. INTRODUCTION 30

6. THE PROBLEM 30

7. MODEL OF UNIFORM LAP REPAIR 32

7.1 Tensile Failure of the Laminate or Repair Patch 34
7.2 Shear Failure 35

7.2.1 Boundary and Continuity Conditions 39
7.2.2 Calculation of the Shear Failure Load 43

8. MODEL OF SCARF REPAIR 43

8.1 Interlayer Shear Strain 45

8.1.1 Boundary and Continuity Conditions 49
iv
8.1.2 Calculation of the Interlayer Shear Strain 55

8.2 Failure Load 55

9. RESULTS 57

10. CONCLUDING REMARKS 62

11. REFERENCES 63

APPENDICES

A÷Moisture Loss of Specimen During Test
B÷Failure Loads of Specimens Exposed to Humid Air
C÷Test Results
D÷Failure Loads of Specimens Under Moisture and Temperature



v
LIST OF FIGURES

Figure Page

1 Illustrations of the Scarf and Lap Repair Techniques 1
2 The Base Laminates Prior to Repair 2
3 Illustration of the Repair Techniques Used in This Study 2
4 The Temperature Cycle Used for Curing the Base Laminate and in Making the
Repair 4
5 Geometries of the Test Specimens 5
6 Application of the Adhesive Layer When Repairing With Prepreg 5
7 Test Specimens Cut Out of the Repaired Base Laminate 6
8 Specimen Conditions and Test Temperatures 6
9 Failure Loads of Different Base Laminates Repaired With Different Repair Materials
(Scarf Repair at 70°F) 12
10 Failure Loads of Different Base Laminates Repaired With Different Repair Materials
(Scarf Repair at 180°F) 12
11 Failure Loads of Different Base Laminates Repaired With Different Repair Materials
(Stepped Lap Repair at 70°F) 13
12 Failure Loads of Different Base Laminates Repaired With Different Repair Materials
(Stepped Lap Repair at 180°F) 14
13 Typical Failure Modes of Specimens Repaired by Scarf and Stepped Lap Techniques 16
14 Definition of the Geometric Factors Investigated 17
15 The Variation of Failure Load With Scarf Angle 19
16 The Effect of the Number of External Plies on the Failure Load in Scarf Repair 19
17 The Effect of Lap Length on the Failure Load in Stepped Lap Repair 20
18 The Effect of Lap Length on the Failure Load in Uniform Lap Repair 20
19 Typical Failure Modes of Specimens Repaired by Stepped Lap and Uniform Lap
Techniques With Various Lap Lengths 21
vi
20 The Effect of the Number of Repair Plies on the Failure Load 21
21 Comparisons of the Failure Loads of Specimens Repaired by the Stepped Lap and
Uniform Lap Repair Techniques 22
22 Failure Loads of Repaired Specimens When the Base Laminates Were Moisturized
Prior to Repair 24
23 The Effects of Specimen Moisture Content and Test Temperature on the Failure
Load 25
24 The Failure Loads of Specimens Sanded With Different Grit Diamond Sanders 26
25 Illustration of the Test Setup Used in Studying the Effects of Cure Cycle on the
Repair 27
26 The Cure Time Required to Reach Full Cure at Different Cure Temperatures 27
27 A Typical Output of the Microdielectrometer 28
28 Failure Loads of Specimens Repaired at Different Cure Temperatures 29
29 Models of the Uniform Lap and the Scarf Repairs 31
30 The Interlayer Between the Laminate and the Repair Patch When the Repair is
Performed With Wet Lay-Up (Lower Left) or Prepreg (Lower Right) 31
31 Illustration of the Shear Stress-Shear Strain Relationship of an Elastic-Perfectly
Plastic Interlayer 32
32 Double-Sided Uniform Lap Repair Model 32
33 Double-Sided Uniform Lap Repair Treated in the Model 33
34 Illustration of the Types of Failure Which May Occur in a Laminate Repaired by the
Double-Sided Uniform Lap Technique 33
35 The On-Axis (x, y) and Off-Axis (1, 2) Coordinate Systems 34
36 Loads on a Section dx in Length of the Laminate Repaired by the Double-Sided
Uniform Lap Technique 36
37 Deformations of the Repair Patch, the Interlayer, and the Laminate 37
38 Possible Regions of the Interlayer Under an Applied Load P 38
39 The Boundary Conditions for the In-Plane Loads in the Laminate and in the Repair
Patch and the Continuity Conditions for the Shear Strain in the Interlayer 40
vii
40 The Equations and Boundary Conditions for Calculating the Shear Strain in the
Interlayer When the Entire Interlayer Behaves in a Linearly Elastic Manner 41
41 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 End 41
42 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = d
l
End 42
43 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 and
x = d
l
Ends 42
44 Scarf Repairs Treated in the Model 44
45 The Types of Failure Which May Occur in a Laminate Repaired by the Scarf
Technique 44
46 The Scarf Repair Treated in the Model 45
47 Loads on a Section dx in Length of the k-th Segment of the Laminate Repaired by
the Scarf Technique 46
48 Deformations of the Repair Patch, the Interlayer, and the Laminate in the k-th
Segment of a Laminate Repaired by the Scarf Technique 47
49 The Compliance Matrices in the Laminate and the Repair Patch 48
50 Possible Regions of the Interlayer Under an Applied Load P 48
51 Model of the Boundary at the Outer Edge (x = x
K
) of the Repair Patch 51
52 The Boundary Conditions for the In-Plane Loads in the Laminate and in the Repair
Patch 51
53 The Continuity Conditions at the Edge of the k-th Segment 52
54 The Continuity Conditions at the Interlayer Between the Elastic and Perfectly Plastic
Regions 52
55 The Equations and Boundary Conditions for Calculating the Shear Strain in the
Interlayer When the Entire Interlayer Behaves in a Linearly Elastic Manner 53
56 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 End 53
57 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = x
K
End 54
viii
58 The Equations and Boundary and Continuity Conditions for Calculating the Shear
Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 and
x = x
K
Ends 54
59 Comparisons of the Calculated (Model) and the Measured (Data) Failure Loads as a
Function of Lap Length for 8-Ply Laminate 60
60 Comparison of the Calculated (Model) and the Measured (Data) Failure Loads as a
Function of Lap Length for 32-Ply Laminate 61
61 Comparisons of the Calculated (Model) and the Measured (Data) Failure Loads as a
Function of Scarf Angle 61
62 Comparison of the Calculated (Model) and the Measured (Data) Failure Loads as a
Function of Extra Plies 62


LIST OF TABLES

Table Page

1 The Base Materials, Repair Materials, and Cure Cycles Used for Fabricating the Base
Laminate and for Making the Repair 3
2 Average Tensile Properties of Specimens Made of Materials Used for Fabricating the
Base Laminates 7
3 Average Tensile Properties of Specimens Made of Materials Used for Repair 8
4 Normalized (63% Fiber Volume) Tensile Properties of Specimens Made of Materials
Used for Fabricating the Base Laminates 9
5 Normalized (63% Fiber Volume) Tensile Properties of Specimens Made of Materials
Used for Repair 10
6 The Test Matrix Used for Evaluating the Effectiveness of Different Repair Materials 11
7 The Failure Loads of Specimens Repaired by the Scarf Technique 13
8 The Failure Loads of Specimens Repaired by the Stepped Lap Technique 14
9 Test Matrix Used for Evaluating the Effects of Different Geometries of the Scarf
and Lap Repairs 17
10 The Test Matrix Used for Evaluating the Effects of the Moisture Content of the
Base Laminate Prior to Repair 23

ix/x
11 The Test Matrix Used for Evaluating the Effects of Specimen Moisture Content and
Test Temperature 24
12 Tests With Different Preparation of the Repair Surface 26
13 The Input Parameters Required by RepairL and the Numerical Values Used in the
Present Calculations for 3k70 Plain Weave Fabric Impregnated With 9396 Resin 58
14 The Input Parameters Required by RepairS and the Numerical Values Used in the
Present Calculations for 3k70 Plain Weave Fabric Impregnated With 9396 Resin 59


xi
EXECUTIVE SUMMARY

As fiber reinforced organic matrix composites are gaining wide acceptance, it is becoming
important to develop repair techniques for structures made of such composites. The present
work studied the repair of composite laminates in a manner which shed light on significant
factors influencing the effectiveness of a repair. Specifically, the objective was to obtain a
systematic set of data which indicates how the type of repair material, the geometry of the repair,
the moisture content of the repaired area, the preparation of the surface prior to repair, and the
processing conditions used during repair affect the strength of the repaired part. In addition,
analytic techniques were developed which can be used to estimate the strengths of composite
laminates after they have been repaired.

Tests were conducted measuring the failure loads of laminates repaired either by the scarf, the
uniform lap, or the stepped lap technique under tensile loading. Data were generated with the
parameters having been varied. The aforementioned parameters were varied over wide ranges
and provided systematic sets of data with the following findings.

• The type of wet lay-up repair material used in the repair or the type of parent laminate
material on which the repair is made does not affect the quality (failure load) of the
repaired part.

• For a scarf repair, there is a gradual decrease in the failure load with increasing scarf
angle. In the present tests the highest failure load occurred at a scarf angle of about 1
degree.

• For a lap repair, there is a limiting lap length beyond which the failure load does not
increase.

• If the parent laminate moisture content is low and the prior moisture history of the part is
known, the repair area does not need to be dried prior to repair. However, if the laminate
moisture content is high (above 1.1 percent), the repair area needs to be dried completely
before repair.

• The failure loads of repaired specimens are reduced under hot/wet conditions, i.e., when
both the moisture content and the test temperature are high.

• Preparing the repair surfaces by sanding them with diamond sander ranging from 60 to
400 grit number does not significantly affect failure load.

• Repair should be cured at the highest permissible temperature so as to achieve the
shortest cure time.

In conjunction with experimental work, models were developed for calculating the failure loads
of composite laminates repaired by scarf and uniform lap techniques. The models take into
account anisotropy of each ply in the laminate and in the repair ply and nonelastic behavior of
the adhesive or resin interlayer between the laminate and the repair patch. On the basis of the
xii
models, two computer codes were written for generating numerical values of the failure loads.
The first code is for calculating the failure loads of laminates repaired by the uniform lap
technique, the second for calculating the failure loads of laminates repaired by the scarf
technique. Failure loads calculated by the models and the corresponding computer codes were
compared with data and good agreement was found between the results of the analysis and the
tests.



1
1. INTRODUCTION.
As fiber reinforced organic matrix composites are gaining wide acceptance, it is becoming
important to develop repair techniques for structures made of such composites. Owing to the
importance of the problem, several techniques have been proposed for repairing composite
laminates. Most of these techniques were developed for repairing specific parts with specific
materials in the field [1-23], and can not readily be generalized to different repair applications.

It would be advantageous to develop repair techniques which can be employed in a wide range of
applications. With this objective in mind, the Commercial Aircraft Composite Repair
Committee (CACRC) was formed under the aegis of Air Transport Association (ATA),
International Air Transport Association (IATA), and Society of Automotive Engineers (SAE)
[24]. The major goal of the CACRC is to develop standard repair procedures for composite
structures. To develop the procedures, several parameters relevant to the effectiveness of repair
need to be thoroughly investigated.

Therefore, the objective of the present work was to study the repair of composite laminates to
shed light on significant factors influencing the effectiveness of a repair. Specifically, the
objective was to obtain a systematic set of data which indicates how the type of repair material,
the geometry of the repair, the moisture content of the repaired area, the preparation of the
surface prior to repair, and the processing conditions used during repair affect the strength of the
repaired part. In addition, analytic techniques are presented which can be used to estimate the
strengths of composite laminates after they have been repaired. The specific problems
investigated are described in next section.

2. THE PROBLEM.
When composite laminates are damaged the damage frequently occurs only in a certain region of
the laminate, as illustrated in figure 1. The damage is then repaired by removing some of the
material around the damaged area and by applying a repair patch.

Damaged laminate
Scarf Double sided lap Single sided lap


FIGURE 1. ILLUSTRATIONS OF THE SCARF AND LAP REPAIR TECHNIQUES

There are two main repair techniques, and these are referred to as scarf and lap (see figure 1). In
the scarf technique the repair material is inserted into the laminate in place of the material
removed due to the damage. In the lap technique the repair material is applied either on one or
on both sides of the laminate over the damaged area.
2
In this investigation, a repair was simulated by joining two laminates (referred to as the base
laminates, see figure 2) either by the scarf or by the lap technique (see figure 3).

• Scarf
• Lap
4 in
8 in
6 in
4 in
0.5 in
0.125 in
3.5 in
Lay-up: [(0/45)
2
]
s
Lay-up: [0]
8


FIGURE 2. THE BASE LAMINATES PRIOR TO REPAIR
(Top: laminates used for scarf repair. Bottom: laminates used for lap repair.)

Repair material Base laminate
• Scarf
Repair material
Base laminate
• Uniform lap
Repair material
Base laminate
• Stepped lap


FIGURE 3. ILLUSTRATION OF THE REPAIR TECHNIQUES USED IN THIS STUDY

For lap repairs two types were considered. One is the double-sided uniform lap, where the
lengths of the layers in the lap are the same. The other one is the double-sided stepped lap,
where the lengths of the repair layers increase, the shortest layer being next to the surface and the
longest on the outside.

The effectiveness of repair was evaluated by subjecting 1-in-wide test specimens, cut out of the
base laminate, to longitudinal tensile loads. The effectiveness of the repair was taken to be the
tensile load at which the specimen fails.
3
Five series of tests were performed to evaluate the effects of the following parameters on the
failure load of the repaired laminate.

1. Type of repair material (i.e., the material used for the repair)
2. Geometry of the repair
3. Moisture content of the specimen and the test temperature
4. Preparation of the surface on which the repair is performed
5. Processing (cure) condition applied during the repair

In addition, analytic techniques are presented for estimating the failure loads of composite
laminates repaired by scarf and by double-sided uniform lap techniques.

This report covers the experimental and analytical research performed for repairs evaluated
under tensile loading. Sandwich test results under compression loads and fracture mechanics
tests characterizing repair materials were reported previously in [25] and [26].

PART I. EXPERIMENTAL RESULTS.
3. EXPERIMENTAL.
Plain weave carbon fiber fabric prepreg plies (with either Hexcel F593, Ciba Geigy R922, or
Ciba Geigy R6376 resin system) were used in the construction of the base laminates (table 1).

TABLE 1. THE BASE MATERIALS, REPAIR MATERIALS, AND CURE CYCLES USED
FOR FABRICATING THE BASE LAMINATE AND FOR MAKING THE REPAIR


Base Material
Cure
Temperature
(°F)
Final Processing
Temperature
(°F)

∆t
1

(min)

∆t
2

(min)

∆t
3

(min)

Pressure
(Psi)
Fabric 3k70 plain weave carbon fiber fabric

Resin

Hexcel F593
Ciba Geigy R922
Ciba Geigy R6376

350

140

120

120

90
45
75
75
Repair Materials (Wet lay-up)
Fabric 3k70 plain weave carbon fiber fabric

Resin

Dexter Hysol EA9396/C2
Dexter Hysol EA9390
Ciba Geigy Epocast 52-A/B

200

70

20
60
220
120

30

14.7
Repair Materials (Prepreg)*
Fabric 3k70 plain weave carbon fiber fabric
Resin Ciba Geigy M20 250 70 30 120 40 14.7

* Dexter Hysol EA 9628 is used as a film adhesive.

The base laminates were 8 in by 4 in for scarf repair and 6 in by 4 in for lap repair (see figure 2).
The lay-up of the base laminate was [(0/45)
2
]
s
for scarf repair and [0]
8
for lap repair. The 0°
4
refers to fabric which has fibers parallel and perpendicular to the longitudinal axis. The 45°
designates plies in which the fibers are t45° from the longitudinal axis. Each base laminate was
laid up by hand, vacuum bagged, and cured in an autoclave by the manufacturer’s cure cycle
(figure 4 and table 1). Same cure cycle was used for the scarf and patch repairs, figure 4. The
thicknesses of the cured base laminates ranged from 0.057 to 0.070 in. The autoclave pressure
(during fabrication of base laminates) or the vacuum (during repair) were applied at time t = 0
and were kept constant during the cure.

Temp. (F)
Time (min)
70 F
∆t
1
∆t
2
∆t
3
Max. cure temp.
Final
processing
temp.


FIGURE 4. THE TEMPERATURE CYCLE USED FOR CURING THE BASE
LAMINATE AND IN MAKING THE REPAIR

After curing, the surfaces of the base laminate, on which a scarf or a lap repair was to be applied,
were sanded with a 120 grit diamond sander. The uniformity of the sanded surface was assessed
by the water-break test, i.e., by spreading a thin layer of water on the surface, and by observing
whether or not the entire surface was wetted [27]. The water was removed by drying the
specimen in an oven. The surface was cleaned by acetone just prior to repair.

Two base laminates were joined (repaired) either by the scarf or by the lap technique, as follows.
The two base laminates were placed side by side (see figure 2) with a 0.5-in gap between the
laminates for scarf repair and a 0.125-in gap for lap repair. The repair was applied with carbon
fiber fabric plies, cut into the desired shape, either in wet lay-up or in prepreg form. For a wet
lay-up repair, the plies were impregnated with a two parts epoxy resin (Dexter Hysol
EA9396/C2, Dexter Hysol EA9390, or Ciba Geigy Epocast 52-A/B, table 1). The prepreg plies
were preimpregnated with Ciba Geigy M20 resin. Each ply (wet lay-up or prepreg) was placed
either into the base laminate (scarf repair) or on the surface of the base laminate (lap repair) (see
figure 5). The 0 degree direction is along the longitudinal axis of the specimen. Zero refers to
plies with 0 and 90 degree fiber orientations. Forty-five refers to plies with +45 and -45 degree
fiber orientations. For the repair with prepreg plies a layer of film adhesive was placed between
the base laminate and the repair material to facilitate bonding between the base laminate and the
repair material (see figure 6).

For the scarf repair, the orientation of each repair ply in the repaired zone was the same as the
corresponding ply orientation of the base laminate. In addition, two 0° external plies were
placed on the surface of the base laminate covering the repaired area (see figure 5). For the lap
5
12.125 in
Repair material
2 in
Fiberglas tab
Base laminate
[0]
8
[0]
0.125 in
• Uniform lap
3
1 in
2
Repair material
Base laminate
16.5 in
[(0/45)
2
]
s
[(0/45)
2
]
s
External plies[0]
0.5 in
0.5 in overlap
2 in
12.125 in
Repair material
2 in
Base laminate
[0]
8
[0]
0.125 in
• Stepped lap
3
Fiberglas tab
0.5 in overlap per ply
• Scarf
Fiberglas tab
0.5 in overlap per ply


FIGURE 5. GEOMETRIES OF THE TEST SPECIMENS

Film adhesive
• Uniform lap
Film adhesive
Prepreg plies
• Stepped lap
Film adhesive
Prepreg plies
• Scarf
Prepreg plies


FIGURE 6. APPLICATION OF THE ADHESIVE LAYER WHEN
REPAIRING WITH PREPREG

repair, the repair plies (with the same 0° orientations as the base laminate) were placed on both
sides of the base laminate, as illustrated in figure 5.

After the repair material was applied, the repair area was vacuum bagged and cured in an oven
under atmospheric pressure (see table 1). After curing, 2-in-long and 0.125-in-thick fiberglas
tabs were attached with room temperature cure Epoxy 907.
6
To make test specimens, the base laminate, repaired in the manner described above, was cut into
1-in-wide strips (see figure 7). The detailed geometries of the specimens are given in figure 5.

Repaired base laminate
Test specimens


FIGURE 7. TEST SPECIMENS CUT OUT OF THE REPAIRED BASE LAMINATE

Each test specimen was subjected to uniaxial tensile loads at a constant displacement rate of
0.05 in/mm. The load at which the specimen failed (failure load) was recorded. The failed
specimens were also inspected visually to establish the mode of failure.

Some of the specimens were tested dry (as prepared), while some were moisturized before
testing either by immersing the specimen in 180°F water for 14 days or by exposing the
specimen to 100 percent humid air at 180°F until 1.1 percent moisture content (weight gain) was
reached (see figure 8). These conditions were adopted because the same conditions are
frequently used in the aircraft industry in testing adhesively bonded joints.



FIGURE 8. SPECIMEN CONDITIONS AND TEST TEMPERATURES

Each specimen was tested either at 70° or 180°F. The temperature was maintained by a chamber
surrounding the specimen. In this chamber the relative humidity was not regulated. Thus, a
small amount (less than 0.007 percent, appendix A) of drying occurred during the 180°F tests,
which lasted about 10 minutes.
70°F
180°F
Water 180°F
Humid air 180°F 100% RH
7
For reference purposes, the tensile properties of undamaged laminates made of the materials used
in constructing the base laminates and in the repair were measured. To this end, 12- by 4-in
([(0/45)
2
]
s
and [0]
8
) plates were prepared in the same manner as the base laminates (see previous
section). Fiberglas tabs were mounted on the plates, and 1-inch-wide, straight-sided tensile
specimens were cut out of the plates. The specimens were subjected to uniaxial tensile loads at a
constant displacement rate of 0.05 in/mm. The load versus strain and the failure load were
recorded. The strain was measured by an extensometer. These tests were performed under the
same conditions as were used for the repaired test specimens (see figure 8). The measured
tensile strengths and longitudinal stiffnesses are given in tables 2 and 3. The values are the
average values of three tests.

TABLE 2. AVERAGE TENSILE PROPERTIES OF SPECIMENS MADE OF MATERIALS
USED FOR FABRICATING THE BASE LAMINATES

Lay-Up

Material*
Specimen Condition/
Test Temperature**
Failure Load
(lbf)
Strength
(ksi)
Stiffness
(Msi)
Dry/70°F 6600 101 8.1
F593 Moist-H
2
O/180°F 6110 93 8.5
Moist-humid air /180°F 6180 95 8.4

Dry/70°F 5890 105 10.8
[0]
8
R922 Moist-H
2
O/180°F 4580 96 10.8
Moist-humid air/180°F 5560 102 10.4

Dry/70°F 6550 107 9.2
R6376 Moist-H
2
O/180°F 6720 107 10.2
Moist-humid air/180°F 6330 101 9.8

Dry/70°F 4780 72 5.9
F593 Moist-H
2
O/180°F 4100 59 5.8
Moist-humid air/180°F 4190 64 6.0

Dry/70°F 4850 90 7.4
[(0/45)
2
]
S
R922 Moist-H
2
O/180°F 4010 69 7.4
Moist-humid air/180°F 4490 83 7.4

Dry/70°F 5850 90 6.5
R6376 Moist-H
2
O/180°F 4730 76 6.9
Moist-humid air/180°F 5070 80 7.3

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see
table 1).
** Specimens were either dry, immersed in 180°F water for 14 days (moist-H
2
O), or exposed to 180°F air at 100%
relative humidity until 1.1% moisture content was reached (moist-humid air). Test temperature was either 70° or
180°F (figure 8).
8
From the data in tables 2 and 3, the properties of unidirectional plies made from the base and
repair materials were deduced. These values, though of not direct concern in this study, are
listed in tables 4 and 5. The data in tables 4 and 5 were normalized to apply to a 0.011-in-thick
ply with a nominal fiber volume content of 63 percent.

TABLE 3. AVERAGE TENSILE PROPERTIES OF SPECIMENS MADE OF
MATERIALS USED FOR REPAIR

Lay-Up

Material*
Specimen Condition/
Test Temperature**
Failure Load
(lbf)
Strength
(ksi)
Stiffness
(Msi)
Dry/70°F 5608 75 7.2

EA9396
Moist-H
2
O/180°F 5581 68 7.7
Dry/70°F 6458 91 7.9
[0]
8

EA9390
Moist-H
2
O/180°F 6039 83 8.7
Dry/70°F 5934 73 7.0

Epocast
Moist-H
2
O/180°F 5635 73 9.7
Dry/70°F 7253 118 9.2

M20
Moist-H
2
O/180°F 7252 106 8.5

Dry/70°F 4623 58 5.0

EA9396
Moist-H
2
O/180°F 3722 47 5.2
Dry/70°F 4150 58 5.5
[(0/45)
2
]
s

EA9390
Moist-H
2
O/180°F 4049 56 6.2
Dry/70°F 4428 51 5.0

Epocast
Moist-H
2
O/180°F 4277 58 5.6
Dry/70°F 6035 95 6.7

M20
Moist-H
2
O/180°F 5101 77 5.9

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this
column (see table 1).
** Specimens were either dry, or immersed in 180°F water for 14 days (moist-H
2
O). Test temperature was
either 70° or 180°F (figure 8).
9
TABLE 4. NORMALIZED (63% FIBER VOLUME) TENSILE PROPERTIES OF
SPECIMENS MADE OF MATERIALS USED FOR FABRICATING THE
BASE LAMINATES

Lay-Up

Material*
Specimen Condition/
Test Temperature**
Strength
(ksi)
Stiffness
(Msi)
Dry/70°F 137 11.0
F593 Moist-H
2
O/180°F 118 10.8
Moist-humid air /180°F 125 11.0

Dry/70°F 175 17.9
[0]
8
R922 Moist-H
2
O/180°F 156 17.6
Moist-humid Air/180°F 166 16.9

Dry/70°F 157 13.5
R6376 Moist-H
2
O/180°F 150 14.2
Moist-humid air/180°F 138 13.5

Dry/70°F 96 7.8
F593 Moist-H
2
O/180°F 74 7.3
Moist-humid air/180°F 84 7.8

Dry/70°F 117 12.1
[(0/45)
2
]
S
R922 Moist-H
2
O/180°F 108 12.5
Moist-humid air/180°F 110 12.3

Dry/70°F 126 9.1
R6376 Moist-H
2
O/180°F 104 9.4
Moist-humid air/180°F 110 10.0

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in
this column (see table 1).
** Specimens were either dry, immersed in 180°F water for 14 days (moist-H
2
O), or exposed to
180°F air at 100% relative humidity until 1.1% moisture content was reached (moist-humid air).
Test temperature was either 70° or 180°F (figure 8).
10
TABLE 5. NORMALIZED (63% FIBER VOLUME) TENSILE PROPERTIES OF
SPECIMENS MADE OF MATERIALS USED FOR REPAIR

Lay-Up

Material*
Specimen Condition/
Test Temperature**
Strength
(ksi)
Stiffness
(Msi)
EA9396 Dry/70°F 88 8.5
Moist-H
2
O/180°F 75 8.2
EA9390 Dry/70°F 108 9.4
[0]
8
Moist-H
2
O/180°F 98 10.4
Epocast Dry/70°F 76 7.3
Moist-H
2
O/180°F 80 10.7
M20 Dry/70°F 162 12.6
Moist-H
2
O/180°F 139 11.2

EA9396 Dry/70°F 64 5.5
Moist-H
2
O/180°F 51 5.7
EA9390 Dry/70°F 71 6.7
[(0/45)
2
]
s
Moist-H
2
O/180°F 67 7.5
Epocast Dry/70°F 61 6.3
Moist-H
2
O/180°F 69 6.7
M20 Dry/70°F 103 9.2
Moist-H
2
O/180°F 101 7.8

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in
this column (see table 1).
** Specimens were either dry, or immersed in 180°F water for 14 days (moist-H
2
O). Test
temperature was either 70° or 180°F (figure 8).

4. TEST RESULTS.
The failure loads of repaired specimens under different conditions are presented in this section.
The failure loads are normalized with respect to failure loads appropriate to the given test. The
actual values of the measured failure loads are included in appendix B. Each data shown in this
section is the average of three to six tests. The spread in the data is indicated by bars.

The materials used for fabricating the base laminates and the repair were given in table 1. The
materials were chosen for this study because they are currently used or being considered for use
in several different commercial aircraft. In the following, for simplicity, both the base laminates
and the repair materials are identified by their resin system. The base laminates are referred to as
11
F593, R922, and R6376, while the repair materials as 9396, 9390, Epocast, and M20 (see
table 1). Note again, that the repair materials using 9396, 9390, and Epocast resins were
prepared by wet lay-up, while the repair material designated as M20 was in prepreg form.

Five series of tests were performed. The results of each five test series are given below.

4.1 TYPE OF REPAIR MATERIAL.
To evaluate the effectiveness of different repair materials, four types of material were used to
repair the base laminates which were made of three different materials. Repairs were made both
by the scarf and the stepped lap techniques. The complete test matrix used in this test series is
given in table 6. Each test was performed with specimens repaired by the scarf and the stepped
lap technique.

TABLE 6. THE TEST MATRIX USED FOR EVALUATING THE EFFECTIVENESS OF
DIFFERENT REPAIR MATERIALS
Specimen Condition/Test Temperature**
Base
Laminate*
Repair
Material*
Dry/70°F Moist-H
2
O/180°F Moist-Humid Air/180°F

F593
9396
9390
Epocast
M20
x
x
x
x
x
x
x
x
x
x
x
x

R922
9396
9390
Epocast
M20
x
x
x
x
x
x
x
x
x
x
x
x

R6376
9396
9390
Epocast
M20
x
x
x
x
x
x
x
x
x
x
x
x

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this
column (see table 1).
** Specimens were either dry, immersed in 180°F water for 14 days (moist-H
2
O), or exposed to 180°F air at
100% relative humidity until 1.1% moisture content was reached (moist-humid air). Test temperature was
either 70° or 180°F (figure 8).

The failure loads of different base laminates repaired with different repair materials (normalized
with respect to the failure load of undamaged laminates) are shown in figures 9 and 10 and table
7 for scarf repair and in figures 11 and 12 and table 8 for stepped lap repair. The results in these
figures are for dry specimens and for specimens moisturized by immersion in water for 14 days.
12
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Prepreg
Wet lay-up
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
R6376
Laminate: F593 R922
Specimen: Dry
Test: 70F
9396 9390 Epocast M20
Repair Material


FIGURE 9. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED
WITH DIFFERENT REPAIR MATERIALS (SCARF REPAIR 70°F)

0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
Prepreg
Wet lay-up
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Laminate: F593 R922
R6376
9396 9390 Epocast M20
Repair Material
2
Specimen: Moist-H O
Test: 180 F


FIGURE 10. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED
WITH DIFFERENT REPAIR MATERIALS (SCARF REPAIR AT 180°F)

13
TABLE 7. THE FAILURE LOADS OF SPECIMENS REPAIRED BY THE
SCARF TECHNIQUE
Specimen Condition/Test Temperature**
Base
Laminate*
Repair
Material*
Dry/70°F Moist-H
2
O/180°F Moist-Humid Air/180°F

9396
F593
R922
R6376
4526
4282
4658
3693
3468
3621
3558
3400
3294

9390
F593
R922
R6376
3695
4168
4222
3705
3746
3735
3404
3293
3124

Epocast
F593
R922
R6376
4175
4064
4281
3408
3279
3514
4178
3388
3580

M20
F593
R922
R6376
4386
4745
5558
2585
3123
2862
3176
3735
2855

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this
column (see table 1).

**Specimens were either dry, immersed in 180°F water for 14 days (moist-H
2
O), or exposed to 180°F air at
100% relative humidity until 1.1% moisture content was reached (moist-humid air). Test temperature was
either 70° or 180°F (figure 8).

0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
Prepreg
Wet lay-up
Laminate: F593 R922
R6376
9396 9390 Epocast M20
Repair Material
Specimen: Dry
Test: 70F


FIGURE 11. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH
DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR AT 70°F)
14
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
Prepreg
Wet lay-up
Laminate: F593 R922
R6376
9396 9390 Epocast M20
Repair Material
2
Specimen: Moist-H O
Test: 180 F


FIGURE 12. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED
WITH DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR AT 180°F)

TABLE 8. THE FAILURE LOADS OF SPECIMENS REPAIRED BY THE
STEPPED LAP TECHNIQUE
Specimen Condition/Test Temperature**
Base
Laminate*
Repair
Material*
Dry/70°F Moist-H
2
O/180°F Moist-Humid Air/180°F

9396
F593
R922
R6376
3360
3547
3302
2958
2533
2520
2624
2311
2096

9390
F593
R922
R6376
3638
3238
3195
2945
2612
2497
2351
2821
2003

Epocast
F593
R922
R6376
3420
3298
3223
2674
2642
2307
2570
2777
2284

M20
F593
R922
R6376
4516
4646
4249
2148
2097
2033
3150
3995
3189

* The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this
column (see table 1).
** Specimens were either dry, or immersed in 180°F water for 14 days (moist-H
2
O), or exposed to 180°F air
at 100% relative humidity until 1.1% moisture content was reached (moist-humid air). Test temperature
was either 70° or 180°F (figure 8).


15
Data were also obtained with specimens with 1.1 percent moisture content achieved by exposure
to 100 percent humid air at 180°F. Both the repaired specimens and the undamaged base
laminates were tested dry at 70°F. Both the repaired specimens and the undamaged base
laminates were immersed in 180°F water for 14 days and were tested at 180°F. Specimen
geometry is given in figure 5. Failure loads of undamaged specimens are given in table 2. The
failure loads of specimens moisturized by water and by humid air showed similar trends.
Therefore, the data for specimens moisturized by humid air are not given here, but are included
in appendix C. The following trends are indicated by the data:

• For the wet lay-up repair (scarf and stepped lap) the type of repair material did not
significantly affect the failure load.

− Failure of these specimens generally occurred due to shear along the base laminate-
repair interface (see figure 13). Apparently, the interfacial shear strengths were
similar for the three wet lay-up repair materials, resulting in similar failure loads,
irrespective on which base material the repair was made.

− The failure loads of specimens repaired by wet lay-up were nearly the same,
irrespective of the base material that was repaired (tables 7 and 8).

• The failure loads were higher with the prepreg repair (scarf and stepped lap) than with the
wet lay-up repair for specimens dry and tested at 70°F.

− For the scarf repair with wet lay-up, failure occurred along the base laminate-
repair interface. (With the prepreg repair, the failure was due to tensile failure of
the base laminate (see figure 13).) This indicates that the failure load (in shear) of
the adhesive film used in the prepreg repair was higher than either the failure load
(in shear) of the wet lay-up interface or the failure load (in tension) of the base
laminate. Thus specimens repaired with prepreg had higher failure loads than
specimens repaired by wet lay-up.

− For stepped lap repair, both with wet lay-up and with prepreg repair, failure
occurred along the base laminate-repair interface (see figure 13). The shear
strength of the adhesive film used in the prepreg repair was higher than the shear
strength of the wet lay-up interface, resulting in the prepreg repair having higher
failure loads than the wet lay-up repair.

• With the prepreg repair (scarf and stepped lap), the failure loads were lower than with
wet lay-up when the specimens were moisturized and tested at 180°F.

− For moisturized specimens tested at 180°F, failure always occurred along the base
laminate-repair interface (see figure 13). Under these conditions the shear
strength of the adhesive film used in the prepreg repair was lower than the shear
strength of the wet lay-up interface. For this reason, specimens repaired by
prepreg had lower failure loads than specimens repaired by wet lay-up.
16
• Scarf
Specimen: Dry
Test: 70F
Specimen: Moist-H
2
O
Test: 180F
Wet lay-up
Wet lay-up Wet lay-up
Wet lay-up
Prepreg
Prepreg
Prepreg
Prepreg
• Stepped lap


FIGURE 13. TYPICAL FAILURE MODES OF SPECIMENS REPAIRED BY
SCARF AND STEPPED LAP TECHNIQUES
(Solid lines indicate failure. Left: dry specimens tested at 70°F, right: moisturized specimens
immersed in 180°F water for 14 days and tested at 180°F.)

The results shown in figures 10, 12, B-1, and B-2 also indicate the effects of specimen moisture
content and test temperature on the failure load. These effects are discussed in section 4.3.

Since the results were similar and consistent for the three wet lay-up repair materials (9396,
9390, and Epocast), the tests presented in the following sections were conducted only with the
9396 repair material.

4.2 GEOMETRY EFFECTS.
The effects of the following geometric factors on the failure load were evaluated (see figure 14):

• For the scarf repair: the scarf angle β and the number of external plies n
e


• For the lap repair: the shape of the repair plies (stepped or uniform), the length of the
repair plies (lap length, d
l
), and the number of repair plies per side
R
l
n .

The effects of these factors on the failure load were evaluated by testing dry as well as
moisturized specimens at 70° and 180°F. The complete test matrix is shown in table 9.
17
• Uniform lap
Lap length, d
l
Repair plies
External plies
• Stepped lap
d
l
Repair plies
Lap length, d
l
(Number per side: n
R
• Scarf
tan(β)=
d
s
H
H
d
s
n
R
l
l
)
(Number per side: n
R
l
)
Base laminate
(Number per side: n
L
l
)
(Number: n
e
)


FIGURE 14. DEFINITION OF THE GEOMETRIC FACTORS INVESTIGATED

TABLE 9. TEST MATRIX USED FOR EVALUATING THE EFFECTS OF DIFFERENT
GEOMETRIES OF THE SCARF AND LAP REPAIRS
(Base Laminate: F593, Repair Material: 9396)

Specimen Condition/
Test Temperature
Repair Type Parameter Dry/70°F Moist-H
2
O/180°F

Scarf
Scarf angle, β (degree)
Number of external plies, n
e

0.72, 1.07, 2.15
0, 1, 2
0.72, 1.07, 2.15
0, 1, 2

Stepped lap
Lap length, d
l
(in)
number of repair plies per side,
R
l
n
0.5, 1, 1.5, 2, 3
1, 2, 3
0.5, 1, 1.5, 2, 3
1, 2, 3

Uniform lap
Lap length, d
l
(in)
Number of plies in the base laminate,
L
l
n
Number of repair plies per side,
R
l
n
0.25, 0.5, 1, 2, 3
8, 32
1, 2, 3, 4
0.25, 0.5, 1, 2, 3
8, 32
1, 2, 3, 4
* Specimens were either dry or immersed in 180°F water for 14 days (Moist-H
2
O). Test temperature was
either 70° or 180°F (figure 8).
18
The results are presented in figures 15 and 16 for the scarf repair and in figures 17 to 21 for the
lap repair. The following general observations can be made on the basis of the data given in
these figures.

• For the scarf repair

− The failure load decreased with increasing scarf angle (see figure 15). Similar
observations were made by Kim, Lee, and Lee [28].

− The addition of external plies on the surface of the specimen resulted in an
increased failure loads (see figure 16). These results are in agreement with those
found by Myhre and Beck [29].

• For the lap repair

− The failure load increased with increasing lap length but only up to a certain
“limiting lap length” d
limit
(see figures 17 to 18). This limiting lap length was
generally between 2 and 3 inches for the specimens used in this investigation.
Beyond the limiting lap length an increase in lap length did not affect the failure
load. Thus, there is no benefit to be gained by making the lap length longer than
the limiting lap length. Similar results were reported by Chan and Sun [30], and
John, Kinloch, and Matthews [31].

− The mode of failure changed with lap length (see figure 19). At short (d
l

0.3d
limit
) and intermediate lap length (0.3d
limit
≤ d
1
≤ d
limit
), failure occurred mostly
along the interface, with some of the repair plies also failing by tension at the
intermediate lap length. At and beyond the limiting lap length (d
l
, ≥ d
limit
) failure
was due to tensile failure of the repair plies. These conclusions seemed to hold
even when the total number of repair plies (2
R
l
n ) was the same as the number of
plies in the base laminate (
L
l
n ). (Note, that for the specimens in section 4.1, the
lap length was 1.5 in, and this corresponds to an intermediate lap length.)

− The failure load increased linearly with the number of plies used in the repair (see
figure 20). This result was valid even when the total number of repair plies (2
L
l
n )
was equal to the number of plies in the base laminate (
R
l
n ) (see figure 20 middle).

− When the lap length was greater than limiting lap length (d
l
≥ d
limit
), the shape of
the repair plies did not markedly affect the failure load (see figure 21). The
failure loads were nearly the same for specimens repaired by stepped and by
uniform repair, provided the following two conditions were simultaneously
satisfied: (a) the number of repair plies were the same and (b) the lap length of
the uniform lap repair was the same as the length of the outside lap of the stepped
repair.

19
F
a
i
l
u
r
e

L
o
a
d

(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d

)
Laminate: F593
Repair: 9396
Specimen: Moist-H O
Test: 180 F
Specimen: Dry
Test: 70F
Scarf Angle, β (degree)
2
Data
Fit to data
tan(β)=
H
d
s
d
s
H
0.0
0.5
1.0
1.5
2.0
0.0
0.5
1.0
1.5
0.5 1 1.5 2 2.5


FIGURE 15. THE VARIATION OF FAILURE LOAD WITH SCARF ANGLE

F
a
i
l
u
r
e

L
o
a
d

(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d

)
Specimen: Dry
Test: 70F
Laminate: F593
Repair: 9396
Number of External Plies, n
e
Specimen: Moist-H O
Test: 180F
2
Data
Fit to data
No External Ply 1 Ply 2 Plies
0.0
0.5
1.0
1.5
2.0
0.5
1.0
1.5
0.0


FIGURE 16. THE EFFECT OF THE NUMBER OF EXTERNAL PLIES ON THE
FAILURE LOAD IN SCARF REPAIR

20

0
.
0
1
.
0
2
.
0
3
.
0
4
.
0
L
a
p

L
e
n
g
t
h
,

d
l

(
i
n
)
0
.
0
0
.
5
1
.
0
1
.
5
2
.
0
0
.
0
0
.
5
1
.
0
1
.
5
F a i l u r e L o a d
0
.
0
0
.
5
1
.
0
1
.
5
0
.
0
1
.
0
2
.
0
3
.
0
L
a
p

L
e
n
g
t
h
,

d
l

(
i
n
)
S
p
e
c
i
m
e
n
:

D
r
y
T
e
s
t
:

7
0
F
S
p
e
c
i
m
e
n
:

M
o
i
s
t
-
H

O
T
e
s
t
:

1
8
0
F
2
L
a
m
i
n
a
t
e
:

F
5
9
3

[
0
]
R
e
p
a
i
r
:

9
3
9
6
8
a t l a p l e n g t h d
l
a t l a p l e n g t h d
l
= 3 i n
)
1

r
e
p
a
i
r

p
l
y
2

r
e
p
a
i
r
p
l
i
e
s
3

r
e
p
a
i
r
p
l
i
e
s
F
i
t

t
o

d
a
t
a
D
a
t
a
d
l
(

F
i
t

t
o

d
a
t
a
[
0
]
8 3
2
[
0
]
d
l
0
.
0
0
.
5
1
.
0
1
.
5
2
.
0
0
.
0
0
.
5
1
.
0
S
p
e
c
i
m
e
n
:

M
o
i
s
t
-
H

O
T
e
s
t
:

1
8
0
F
2
0
.
0
0
.
5
1
.
0
F a i l u r e L o a d
L
a
m
i
n
a
t
e
:

F
5
9
3
R
e
p
a
i
r
:


9
3
9
6
S
p
e
c
i
m
e
n
:

D
r
y
T
e
s
t
:

7
0
F
1

r
e
p
a
i
r

p
l
y
a t l a p l e n g t h d
l
a t l a p l e n g t h d
l
= 3 i n
)
2

r
e
p
a
i
r

p
l
i
e
s
3

r
e
p
a
i
r

p
l
i
e
s
0
.
0
1
.
0
2
.
0
3
.
0
4
.
0
L
a
p

L
e
n
g
t
h
,

d
l

(
i
n
)
4

r
e
p
a
i
r

p
l
i
e
s
(
0
.
0
0
.
5
1
.
0
0
.
0
1
.
0
2
.
0
3
.
0
L
a
p

L
e
n
g
t
h
,

d
l

(
i
n
)


F
I
G
U
R
E

1
7
.


T
H
E

E
F
F
E
C
T

O
F

L
A
P

L
E
N
G
T
H

O
N

T
H
E

F
A
I
L
U
R
E

L
O
A
D

I
N

S
T
E
P
P
E
D

L
A
P

R
E
P
A
I
R



F
I
G
U
R
E

1
8
.


T
H
E

E
F
F
E
C
T

O
F

L
A
P

L
E
N
G
T
H

O
N

T
H
E

F
A
I
L
U
R
E

L
O
A
D

I
N

U
N
I
F
O
R
M

L
A
P

R
E
P
A
I
R




21



S
t
e
p
p
e
d

l
a
p


U
n
i
f
o
r
m

l
a
p
F a i l u r e l o a d
~
0
.
3


d
l
i
m
i
t
d
l
i
m
i
t
L
a
p

l
e
n
g
t
h
,


d
l



N
u
m
b
e
r

o
f

R
e
p
a
i
r

P
l
i
e
s
,


n
S
p
e
c
i
m
e
n
:

M
o
i
s
t
-
H

O
T
e
s
t
:

1
8
0

F
2
S
p
e
c
i
m
e
n
:

D
r
y
T
e
s
t
:

7
0
F
R
l
F a i l u r e L o a d ( R e p a i r e d / U n d a m a g e d )
1
2
3
4
0
1
2
3
4
5
L
a
m
i
n
a
t
e
:

F
5
9
3
R
e
p
a
i
r
:

9
3
9
6

d
l
=
3

i
n
D
a
t
a
F
i
t

t
o

d
a
t
a
8

p
l
y
l
a
m
i
n
a
t
e
3
2

p
l
y

l
a
m
i
n
a
t
e
3
2

p
l
y

l
a
m
i
n
a
t
e
8

p
l
y
l
a
m
i
n
a
t
e
8

p
l
y
l
a
m
i
n
a
t
e
8

p
l
y
l
a
m
i
n
a
t
e
d
l
d
l
d
l
0
.
0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
0
.
0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
0
.
0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
0

F
I
G
U
R
E

1
9
.


T
Y
P
I
C
A
L

F
A
I
L
U
R
E

M
O
D
E
S

O
F

S
P
E
C
I
M
E
N
S

R
E
P
A
I
R
E
D

B
Y

S
T
E
P
P
E
D

L
A
P

A
N
D

U
N
I
F
O
R
M

L
A
P

T
E
C
H
N
I
Q
U
E
S

W
I
T
H

V
A
R
I
O
U
S

L
A
P

L
E
N
G
T
H
S

(
S
p
e
c
i
m
e
n

g
e
o
m
e
t
r
i
e
s

a
r
e

g
i
v
e
n

i
n

f
i
g
u
r
e
s

5

a
n
d

1
4
)



F
I
G
U
R
E

2
0
.


T
H
E

E
F
F
E
C
T

O
F

T
H
E

N
U
M
B
E
R

O
F

R
E
P
A
I
R

P
L
I
E
S

O
N

T
H
E

F
A
I
L
U
R
E

L
O
A
D



22
0.0
0.5
1.0
1.5
2.0
0.0
0.5
1.0
1.5
F
a
i
l
u
r
e

L
o
a
d
0 1 2 3 4
Number of Repair Plies, n
Laminate: F593
Repair: 9396
Specimen: Dry
Test: 70F
Specimen: Moist-H O
Test: 180F
2
w
i
t
h

u
n
i
f
o
r
m

l
a
p

r
e
p
a
i
r
)
Lap length, d
l
=3 in
d
l
d
l
w
i
t
h

s
t
e
p
p
e
d

l
a
p


r
e
p
a
i
r
(
R
l
Data
Fit to data


FIGURE 21. COMPARISONS OF THE FAILURE LOADS OF SPECIMENS
REPAIRED BY THE STEPPED LAP AND UNIFORM LAP
REPAIR TECHNIQUES

4.3 EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE.
Two series of tests were performed to evaluate the effects of specimen moisture content and test
temperature on the failure loads of repaired specimens. In one series of tests, the base laminates
were moisturized prior to repair. In the second series, the base laminates were dried prior to
repair, and the repaired specimens were moisturized before being tested.

4.3.1 Base Laminates Premoisturized.
There were two types of premoisturized specimens. The first type of specimen consisted of the
base laminate moisturized prior to repair by exposing the unrepaired base laminate to 100
percent humid air at 180°F. The base laminates were kept in the environmental chamber until
the desired moisture content was reached. The base laminates were then removed from the
environmental chamber, repaired, and tested at 70°F.

The second type of specimen consisted of the base laminate whose moisture content was
established by moisturization followed by drying. In this case the base laminate was moisturized
to either 1.1 percent or 1.5 percent moisture content. When the desired moisture content was
23
reached, the base laminate was removed from the environmental chamber and was dried in an
oven at 180°F. Once the desired moisture content was reached during the drying process, the
base laminate was taken out of the oven, repaired, and tested at 70°F.

With above procedures, the moisture content of the base laminate was established either by
moisturization of the base laminate or by moisturization followed by drying. The test matrix is
given in table 10.

TABLE 10. THE TEST MATRIX USED FOR EVALUATING THE EFFECTS OF THE
MOISTURE CONTENT OF THE BASE LAMINATE PRIOR TO REPAIR
(Base Laminate: F593, Repair Material: 9396)

Specimen Condition/Test Temperature
Repair Type Parameter Moist-Humid Air/70°F* 180°F Dry Air/70°F**
Scarf
Moisture content
(% weight)
0.0 to 1.1
0.0 to 1.5
1.1 to 0.0
1.5 to 0.0
Stepped lap Moisture content
(% weight)
0.0 to 1.1
0.0 to 1.5
1.1 to 0.0
1.5 to 0.0

* Specimens were exposed to 180°F air at 100% relative humidity until either 1.1% or 1.5% moisture
content was reached (moist-humid air). Test temperature was either 70° or 180°F (figure 8).
** Specimens were dried at 180°F after having been moisturized to either 1.1% or 1.5% moisture content.

The failure loads of specimens repaired with premoisturized base laminates are given in
figure 22. In this figure, the failure loads are normalized with respect to the failure load of
the dry specimen. The following major observations can be made on the basis of the data in
figure 22.

• The failure loads of repaired specimens decreased only slightly when the base laminates
were moisturized prior to the repair, provided that the moisture content did not exceed 1.1
percent (see figure 22, top). This conclusion was valid when the base laminate moisture
content was established either by moisturization (solid circle) or by moisturization
followed by drying (open circle).

• The behavior was quite different when the moisture content of the base laminates
exceeded about 1.1 percent (see figure 22, bottom). The same phenomenon was observed
by Robson, et al. [32] in their tests with scarf joints. The failure loads of repaired
specimens made with base laminates which have first been moisturized to 1.5 percent
moisture content then dried (dotted line in figure 22 bottom) were lower than the failure
loads of repaired specimens made with base laminates in which the moisture was
introduced during moisturization (solid line). In fact, the base laminates had to be dried
almost completely to recover the failure loads of specimens made with premoisturized
(but not dried) base laminates. Similar observations were made by Parker [33] on
adhesively bonded single lap joints.

24
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
Laminate: F593
Repair: 9396
Test: 70F
Fit to data
Data, base laminate being moisturized
Data, base laminate being dried at 180 F
0.00
0.25
0.50
0.75
1.00
1.25
1.50
0.0 0.4 0.8 1.2 0.0 0.4 0.8 1.2 1.6
Moisture Content of Base Laminate
Prior to Repair (Percent)
w
i
t
h

p
r
e
m
o
i
s
t
u
r
i
z
e
d

b
a
s
e

l
a
m
i
n
a
t
e
w
i
t
h

d
r
y

b
a
s
e

l
a
m
i
n
a
t
e



F
a
i
l
u
r
e

L
o
a
d

o
f
R
e
p
a
i
r
e
d

S
p
e
c
i
m
e
n
(
(


FIGURE 22. FAILURE LOADS OF REPAIRED SPECIMENS WHEN THE
BASE LAMINATES WERE MOISTURIZED PRIOR TO REPAIR

The above mentioned results have two significant ramifications. First, it is not necessary to dry
the repair area prior to repair as long as the moisture content of the laminates is low (less than 1.0
percent). Second, the repair area should be dried completely (not just partially) when the
laminate moisture content is high (above 1.0 percent).

4.3.2 Specimen Moisture Content and Test Temperature.
Specimens were repaired by scarf and stepped lap techniques using dry base laminates. Each
specimen was then either immersed in 180°F water for 14 days or exposed to 100 percent humid
air at 180°F until 1.1 percent moisture content was reached. The failure loads of the dry and the
moisturized specimens were measured at 70° and 180°F. The complete test matrix is shown in
table 11.

TABLE 11. THE TEST MATRIX USED FOR EVALUATING THE EFFECTS OF
SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE
(Base Laminate: F593, Repair Material: 9396)

Specimen Condition/Test Temperature
Repair
Type
Dry
70°F
Dry
180°F
Moist-H
2
O
70°F
Moist-H
2
O
180°F
Moist-Humid Air/
70°F
Moist-Humid Air/
180°F
Scarf x x x x x x
Stepped
lap
x x x x x x
25
The trend in the data was similar for the specimens moisturized in water and in humid air.
Therefore, here data presented is only for specimens moisturized by exposure to humid air
(see figure 23). The data obtained during moisturization in water are included in appendix D.
The data in figure 23 indicate the following trends.

• The moisture content did not significantly reduce the failure load when the specimens
were tested at 70°F.

• Testing at the elevated temperature (180°F) did not significantly reduce the failure load
when the specimens were tested at dry condition.

• The failure load of the specimen was reduced under hot/wet conditions i.e., when the
specimen was both wet and tested at 180°F. Similar observations were made by Stone
[16] with single lap shear specimens, Mylire, Labor, and Aker [34] with modified single
lap specimens, and John, Kinloch, and Matthews [31] with double lap joints.

Dry & 180F Moist & 70F Moist & 180F
Specimen Condition & Test Temperature
s
p
e
c
i
m
e
n

a
t

i
n
d
i
c
a
t
e
d

c
o
n
d
i
t
i
o
n

d
r
y

s
p
e
c
i
m
e
n

t
e
s
t
e
d

a
t

7
0
F

)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Laminate: F593
Repair: 9396
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
(
F
a
i
l
u
r
e

L
o
a
d


FIGURE 23. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND
TEST TEMPERATURE ON THE FAILURE LOAD

4.4 EFFECT OF SANDING.
Prior to repair the repair surface of the base laminate is usually sanded. To evaluate the
influence of sanding on the failure load of repaired specimens, the repair surfaces were sanded
with different grit diamond sanders (see table 12). The failure loads of specimens repaired using
different grit sanders are shown in figure 24. The data indicate that the grit number of the sander
did not have a major effect on the failure load. For the material and type of repair used in this
study, a sander with grit number about 100 seems appropriate.
26
TABLE 12. TESTS WITH DIFFERENT PREPARATION OF THE REPAIR SURFACE
(Base Laminate: F593, Repair Material: 9396)

Specimen Condition/Test Temperature*
Repair Type Parameter Dry/70°F Moist-H
2
O/180°F
Scarf Grit number of diamond sander, g 60, 120, 400 60, 120, 400
Stepped lap Grit number of diamond sander, g 60, 120, 400 60, 120, 400

* Specimens were either dry, or immersed in 180°F water for 14 days (Moist-H
2
O). Test temperature was
either 70° or 180°F (figure 8).



FIGURE 24. THE FAILURE LOADS OF SPECIMENS SANDED WITH DIFFERENT
GRIT DIAMOND SANDERS

4.5 EFFECT OF CURE CYCLE.
Composite material is generally repaired at an elevated temperature to expedite the chemical
reactions and to cure the resin used in the repair. In addition to the temperature, time is factor in
the cure. For economical reasons, it is desirable to have a low cure temperature and a short cure
time.

The effects of the temperature-time history (referred to as cure cycle) on the effectiveness of the
repair were studied on specimens repaired by scarf and stepped lap techniques. Each specimen
was repaired by surrounding it with a heating blanket. This assembly was enclosed in a vacuum
bag and was kept under vacuum during the repair (see figure 25). The specimen was heated at a
constant rate of 6.5°F/min until the desired cure temperature T
c
was reached (see figure 26
insert). This cure temperature was maintained until the repair material was fully cured.
27
Vacuum
Surface breather
Heating blanket
Insulation
Edge bleeder
Solid FEP parting film
Aluminum plate
Surface bleeder
Microdielectric
sensor
Composite plate
Repaired
part
Solid FEP parting film
Vacuum
bag


FIGURE 25. ILLUSTRATION OF THE TEST SETUP USED IN STUDYING THE EFFECTS
OF CURE CYCLE ON THE REPAIR

0
30
60
90
120
0
30
60
90
2.0
125 150 175 200 225
Cure Temperature, T (F)
Laminate: F593
Repair : 9396
Specimen: Dry
Test: 70F Data
Fit to data
c
Time
T
e
m
p
.
T
c
T
i
m
e

t
o

F
u
l
l

C
u
r
e


(
m
i
n
)


FIGURE 26. THE CURE TIME REQUIRED TO REACH FULL CURE AT
DIFFERENT CURE TEMPERATURES
(Specimen geometries are given in figures 5 and 14)
28
The progress of the cure and the specimen temperature were monitored by a Eumetric System II
Microdielectrometer (Micromet Instruments, Inc.). This instrument provided the temperature
and the ion viscosity of the repair resin, the latter being an indicator of the degree of cure. A
typical output of the instrument is shown in figure 27. Each specimen was cured at a preset cure
temperature T
c
(150°, 175°, or 200°F) until the instrument indicated that full cure was reached.


FIGURE 27. A TYPICAL OUTPUT OF THE MICRODIELECTROMETER

The time required to reach full cure is shown in figure 26. As the heating rate was constant at
6.5°F/min, the time required to reach full cure decreased with increasing cure temperature.

The failure loads of specimens cured at different temperatures are shown in figure 28. The cure
temperature did not seem to affect significantly the failure load for the dry specimens tested at
70°F. It is unknown what would happen if the specimens were tested at 180°F. This limited
testing suggests that it is advantageous to use the highest cure temperature (permissible for the
resin system used in the repair) since this results in shortest cure (and repair) time.


29
0.0
0.5
1.0
1.5
2.0
F
a
i
l
u
r
e

l
o
a
d
125 150 175 200 225
Cure Temperature, T (F)
0.0
0.5
1.0
1.5
2.0
Laminate: F593
Repair: 9396
Data
Fit to data
a
t

c
u
r
e

t
e
m
p
.


T



=
1
5
0

F
a
t

c
u
r
e

t
e
m
p
.


T
)
Time
T
e
m
p
.
T
c
c
c
c
Specimen: Dry
Test: 70F
(


FIGURE 28. FAILURE LOADS OF SPECIMENS REPAIRED AT DIFFERENT
CURE TEMPERATURES
(Specimen geometries are given in figures 5 and 14. Failure loads of the specimens
at cure temperature 150°F are given in appendix C.)

4.6 SUMMARY.
The major findings of the tests are summarized below.

• The type of wet lay-up repair material used in the repair, or the type of parent laminate
material on which the repair is made, does not affect the quality (failure load) of the
repaired part.

• For a scarf repair, there is a gradual decrease in the failure load with increasing scarf
angle. In the present tests the highest failure load occurred at a scarf angle of about 1
degree. Small variations in the scarf angle around this value did not have a significant
effect on the failure load.

• For a lap repair, there is a limiting lap length beyond which the failure load does not
increase.

• If the parent laminate moisture content is low and the prior moisture history of the part is
known, the repair area does not need to be dried prior to repair. However, if the laminate
moisture content is high (above 1.1 percent), the repair area needs to be dried completely
before repair.

30
• The failure loads of repaired specimens are reduced under hot/wet conditions, i.e., when
both the moisture content and the test temperature are high.

• Preparing the repair surfaces by sanding them with diamond sander ranging from 60 to
400 grit number does not significantly affect failure load.

• A repair should be cured at the highest permissible temperature so as to achieve the
shortest cure time.

The aforementioned findings are strictly valid only for the specimens and test conditions used in
this investigation. It is likely that the trends observed here are valid for specimens and test
conditions not covered in the present tests.

PART II. ANALYTICAL MODELS.
5. INTRODUCTION.
Our objective is to develop analytical models which can be used to calculate the failure loads of
laminates repaired by the scarf and lap techniques. In developing the models, previous models
for adhesively bonded joints were consulted because repair patches and bonded joints have
similar features.

Numerous models are available for isotropic materials bonded adhesively by single lap [30, 35-
46], double lap [30, 43, 46-48], and scarf [49] joints.

Fewer models exist for adhesive bonds joining composite materials, and most of these are
restricted to adhesives which may be treated as linearly elastic [28, 38, 40, 50-55]. However, in
many practical situations, the adhesive does not behave in a linearly elastic manner. Nonelastic
behavior of adhesives in bonding composite laminates has been considered for single lap [56],
double lap [57-61], and scarf joints [59, 61, 62] by Hart-Smith and Adams, et al. Of these, the
scarf and double lap joint models are relevant here because of our interest in scarf and double-
sided uniform lap repair. Adams [61] performed stress analysis of scarf and double-lap joints
using finite element methods, but as far as authors can ascertain, their codes are not in the public
domain. Hart-Smith [57-59, 62] did not treat each composite layer separately, but smeared the
properties of the composite laminates which were adhesively bonded.

In this investigation, the Hart-Smith models were adapted to model the scarf and double-sided
lap repair by accounting for each individual layer separately.

6. THE PROBLEM.
Symmetric composite laminates can be repaired either by the uniform lap or by the scarf
technique (see figure 29). Both the laminate and the repair patch consist of fiber reinforced plies
which behave in a linearly elastic manner. The repair may be made by wet lay-up or by prepreg.
For wet lay-up repair there is a thin resin layer, while for prepreg repair there is an adhesive
between the laminate and the repair patch (see figure 30). Both the resin layer and the adhesive
31
Laminate
Repair patch
External plies
(b) Scarf repair without external ply
(c) Scarf repair with external plies
(a) Uniform lap repair
P
P
Laminate
Repair patch
P
P
P
P
Repair patch
Laminate


FIGURE 29. MODELS OF THE UNIFORM LAP AND THE SCARF REPAIRS



FIGURE 30. THE INTERLAYER BETWEEN THE LAMINATE AND THE REPAIR PATCH
WHEN THE REPAIR IS PERFORMED WITH WET LAY-UP (LOWER LEFT) OR
PREPREG (LOWER RIGHT)
32
are treated as an “interlayer” which exhibits elastic-perfectly plastic behavior (see figure 31).
The shear strains at the elastic limit and at plastic failure are γ
ef
and γ
pf
, respectively.

The laminate is subjected to an in-plane tensile load P. The objective is to find the value of this
load (failure load, P = F) which fails the repaired laminate.


FIGURE 31. ILLUSTRATION OF THE SHEAR STRESS-SHEAR STRAIN RELATIONSHIP
OF AN ELASTIC-PERFECTLY PLASTIC INTERLAYER

7. MODEL OF UNIFORM LAP REPAIR.
A model for the double-sided uniform lap repair is illustrated in figure 32. The model for the
stepped lap repair is not given separately because, as was shown in section 4.2, the failure loads
of laminates repaired by uniform and stepped lap techniques are similar. Thus, the model of the
uniform lap repair can be applied to a stepped lap repair.

P
Laminate
Repair patch
Interlayer
Lap length, d
l
P
Number of plies, n
l
R
Number of plies, n
l
L
Thickness, h
i


FIGURE 32. DOUBLE-SIDED UNIFORM LAP REPAIR MODEL

The lay-up of the laminate/repair patch assembly is symmetric. For modeling such a repair it is
sufficient to consider only one half the repaired laminate, as shown in figure 33.
G =
ef
p
γ
τ

33
P
P/2
P/2
Repair patch
Laminate
Interlayer
Repair patch
x
P P
Laminate
Repair patch


FIGURE 33. DOUBLE-SIDED UNIFORM LAP REPAIR TREATED IN THE MODEL

When the repaired laminate is subjected to a tensile load, the laminate may fail in tension, the
repair patch may fail in tension, or the interlayer may fail in shear (see figure 34). In the
following, models are presented which provide the failure loads when failure is due to any of the
three scenarios.

(a) Laminate failure in tension
(b) Repair patch failure in tension
(c) Interlayer failure in shear
F
F
F F
F
F


FIGURE 34. ILLUSTRATION OF THE TYPES OF FAILURE WHICH MAY OCCUR IN A
LAMINATE REPAIRED BY THE DOUBLE-SIDED UNIFORM LAP TECHNIQUE
34
7.1 TENSILE FAILURE OF THE LAMINATE OR REPAIR PATCH.
When the laminate or the repair patch fails in tension (see figure 34), the failure load is
calculated as follows. At tensile failure (of either the laminate or the repair patch) the applied in-
plane tensile load (per unit width) is denoted by F (P = F). At the middle of the repair zone (x =
0, figure 33) the in-plane failure load (per unit width) in the repair plies is F/2 (P/2 = F/2). Since
the model is symmetric, the in-plane strains at failure are [63]

Laminate Repair Patch

( ) F
L
F
L
x 11
α ε · ( ) 2
11
/ F
R
F
R
x
α ε ·
( ) F
L
F
L
y 21
α ε · ( ) 2
21
/ F
R
F
R
y
α ε · (1)
( ) F
L
F
L
s 61
α ε · ( ) 2
61
/ F
R
F
R
s
α ε ·

where ε
x
, ε
y
, and ε
s
are the off-axis in-plane strains (see figure 35). The α
ij
’s are the components
of the compliance matrix. The superscripts L and R refer to the laminate and the repair patches,
respectively.

The on-axis (1, 2, and 6 in figure 35) components of the strains are


F
R , L
s
R , L
y
R , L
x
R , L R , L R , L R , L R , L R , L
R , L R , L R , L R , L
R , L R , L R , L R , L
R , L
R , L
R , L
sin cos sin cos sin cos
sin cos cos sin
sin cos sin cos
¹
¹
¹
'
¹
¹
¹
¹
'
¹

,
`

.
|
− −
− ·
¹
¹
¹
'
¹
¹
¹
¹
'
¹
ε
ε
ε
θ θ θ θ θ θ
θ θ θ θ
θ θ θ θ
ε
ε
ε
2 2
2 2
2 2
6
2
1
2 2
(2)

where θ is the fiber orientation measured with respect to the direction of the applied load.



FIGURE 35. THE ON-AXIS (x, y) AND OFF-AXIS (1, 2) COORDINATE SYSTEMS
35
Failure occurs when, in any of the laminate or repair plies, one of the following maximum strain
failure criteria is met


( )
1
1 1
·
F
L L
L
E
X
ε

( )
1
1 1
·
F
R R
R
E
X
ε


( )
1
2 2
·
F
L L
L
E
X
ε

( )
1
2 2
·
F
R R
R
E
Y
ε
(3)

( )
1
6 6
·
F
L L
L
E
S
ε

( )
1
6 6
·
F
R R
R
E
S
ε


X, Y, and S are the on-axis longitudinal, transverse, and shear strengths of the ply, respectively.
E
1
, E
2
, and E
6
are the on-axis longitudinal, transverse, and shear moduli of the ply, respectively.
Six F values are calculated for each ply from equations 1-3. The lowest value of F resulting
from these calculations is the failure load.

7.2 SHEAR FAILURE.
When the repaired laminate fails due to shear failure of the interlayer (see figure 34(c)), the
failure load is calculated as follows.

At the start of the analysis the repaired laminate has an applied in-plane load P (P ≠ F,
figure 36). At x = 0, the loads on the repair patches (plies) are P/2. For a section dx in length
(x being in the direction of the applied load), balancing the forces for the laminate and each of
the repair patches gives:

0 2 -
dx
dN
L
· τ Laminate
(4)
0
dx
dN
R
· +τ Repair Patch

( )
L
x
L
N N · and ( )
R
x
R
N N · are the in-plane loads (per unit width) in the laminate and each repair
patch, respectively. ( )
zx
τ τ · is the shear stress in the interlayer.

36
Interlayer, I
Laminate, L
Repair patch, R
N
R
N
R
N
R

+ dx
N
R

+ dx
N
L
dx
N
L

+ dx
τ dx
dN
L
dx
dN
R
dx
dN
R
dx
R
R
L
I
I
P
P/2
P/2
x
dx


FIGURE 36. LOADS ON A SECTION dx IN LENGTH OF THE LAMINATE REPAIRED BY
THE DOUBLE-SIDED UNIFORM LAP TECHNIQUE

By neglecting the shear deformations of the laminate and the repair patch, compatibility requires
that the following condition be satisfied [64] (see figure 37)


i
R
x i
L
x
dx)h
dx
d
( 1)dx ( h 1)dx (
γ
γ ε γ ε + + + · + + (5)

This can be simplified to give


i
R
x
L
x
h dx
d ε ε γ −
· (6)

where
L
x
ε and
R
x
ε are the off-axis in-plane strains in the laminate and in the repair patch,
respectively, h
i
is the thickness of the interlayer, and γ (= γ
zx
) is the interlayer shear strain.

37
(1+ε
R
)dx
γ

dx γ +
dx
(1+ε
L
)dx
h
L
R
R
I
I
Interlayer
Laminate
Repair patch
x
x
i


FIGURE 37. DEFORMATIONS OF THE REPAIR PATCH, THE INTERLAYER,
AND THE LAMINATE
(Double-sided uniform lap technique.)

For the symmetric arrangement considered here, the x components of the strains in the laminate
and in the repair patch are [63]


L L
11
L
x
N α ε · Laminate
(7)

R R
11
R
x
N α ε · Repair Patch

where α
11
is the 11 component of the α
ij
compliance matrix. Equations 6 and 7 yield

( )
R R
11
L L
11
i
N N
h dx
d
α α
γ
− ·
1
(8)

By combining equations 4 and 8 obtains

( )
R
11
L
11
i
2
2
2
h dx
d
α α
γ
+ ·
1
(9)

There are four possible scenarios for the behavior of the interlayer: (1) the entire interlayer
behaves in a linearly elastic manner, (2) a perfectly plastic region near the x = 0, (3) a perfectly
plastic region near the x = d
l
end of the interlayer, or (4) a perfectly plastic region near both the
x = 0 and x = d
l
ends of the interlayer (figure 38).

38
γ
τ
x = d
l
x = 0
τp
γ
τ
x
τp
Plastic Elastic
Plastic
γ
τ
τp
Elastic Elastic Plastic Plastic
x = d
l
x = 0 x = d
l
x = 0 x
p1
x
p1
x
p1
x
p2
γ
τ
x = d
l
x = 0
τp
Elastic
P
P/2
P/2
Interlayer
x
d
l
Repair patch
Laminate
Repair patch


FIGURE 38. POSSIBLE REGIONS OF THE INTERLAYER UNDER AN APPLIED LOAD P

In the elastic region, the shear strain and shear stress are

γ = γ
e
and τ = Gγ
e
(10)

where G is the shear modulus of the interlayer. From equations 9 and 10 the following
expression for the shear strain can be obtained.

γ = psinh(λx) + qcosh(λx) Elastic Region (11)

where λ is defined as

( )
R
11
L L
11
i
2
N 2
h
G
α α λ + · (12)

In the plastic region the shear strain and the shear stress are (see figures 31 and 38)

γ = γ
p
and τ = τ
p
= Gγ
ef
(13)

where, as shown in figure 31, γ
ef
is constant. Equations 9 and 13 yield

s rx x
ef
p
+ + ·
2
2
2
γ λ
γ Plastic Region (14)

39
In equations 11 and 14, p, q, r, and s are constants which must be determined from the boundary
and continuity conditions.

7.2.1 Boundary and Continuity Conditions.
The following are the boundary and continuity conditions for to equations 11 and 14.

• At the free ends of the laminate and the repair patches, the axial load is zero (see figure
39)

N
L
= 0 at 0 x ·
N
R
= 0 at
l
d x ·

(15)

• At the ends of the laminate and the repair patches where loads are applied (see figure 39),
the load per unit width N
L
in the laminate is equal to the applied load per unit width P,
and the load per unit width N
R
in each repair patch is equal to P/2


2
P
N
R
· at 0 x ·
(16)
P N
L
· at
l
d x ·

By combining equations 8, 15, and 16, these boundary conditions can be expressed as


i
R
h
P
dx
d
2
11
α γ
− · at 0 x · (17)
and

i
L
h
P
dx
d
11
α γ
· at
l
d x ·

(18)

In equations 17 and 18
dx

is either
dx
e dγ
or
dx
p dγ
depending on whether the interlayer is in the
elastic or in the perfectly plastic region at the boundary (see figures 40 through 43).

40
P/2
P/2
Laminate
Repair patch
Repair patch
N
L
=
N
R
= 0
N
L
=
N
R
=
0
N
R
=
P
Laminate
Repair patch
x
x
d
l
γ
x
x
p1
γ
e
γ
p
γ
p
x
p2
γ
p
γ
e
= = γ
ef
d
dx
γ
e
d
dx
γ
p
=
N
R
= 0
at x=0 at x=d
l
P
x=x
p1
x
p2
x=
at
and
Plastic Elastic Plastic
γ
ef
P


FIGURE 39. THE BOUNDARY CONDITIONS FOR THE IN-PLANE LOADS IN
THE LAMINATE AND IN THE REPAIR PATCH AND THE CONTINUITY
CONDITIONS FOR THE SHEAR STRAIN IN THE INTERLAYER


At the locations where the elastic and plastic regions meet (x = x
p1
and x = x
p2
in figure 38) the
shear strains in the elastic and plastic regions are equal (with the value γ
ef
) and are continuous
(see figures 39, 41 through 43). Correspondingly,

( )
ef p e
γ γ γ · ·
at
p1
x x · ,
p2
x x · (19)

dx
d
e
γ
=
dx
d
p
γ


The locations x
p1
and x
p2
are unknown and must be determined from the solutions of the
equations summarized in figures 41 through 43.

41
γ
e
= c
1
+ sinh( λ x) c
2
cosh(λ x)
d
d
a
R
11
x
=
2h
i
P a
L
d
dx
11
=
h
i
P
γ
x
Elastic
γ
ef
0
at x = d
l
at x = 0
γ
e
γ
e
P
P/2
P/2
Interlayer
Repair patch
Laminate
Repair patch
d
l


FIGURE 40. THE EQUATIONS AND BOUNDARY CONDITIONS FOR
CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
ENTIRE INTERLAYER BEHAVES IN A LINEARLY ELASTIC MANNER

γ
p
=
λ

ef
2
x
2
r +s x +
γ
e
=
p + sinh( λ x) q cosh(λ x)
d
d
a
R
11
x
=
2h
i
P
γ
p a
L
d
dx
11
=
h
i
P γ
p
γ
e
= = γ
ef
d
dx
γ
e
d
dx
γ
p
=
γ
x
x
p1
Plastic
Elastic
γ
ef
0
at x = d
l
at x = x
p1
at x = 0
γ
e
P
P/2
P/2
Interlayer
Repair patch
Laminate
Repair patch
d
l


FIGURE 41. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN
THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 END
42
γ
p
=
λ

ef
2
x
2
r +s x + γ
e
=
p + sinh( λ x) q cosh(λ x)
d
d
a
R
11
x
=
2h
i
P
a
L
d
dx
11
=
h
i
P
γ
p
γ
p
γ
e
= = γ
ef
d
dx
γ
e
d
dx
γ
p
=
γ
x
x
p1
Elastic
Plastic
γ
ef
d
l
0
at x = d
l
at x = x
p1
at x = 0
γ
e
P
P/2
P/2
Interlayer
Repair patch
Laminate
Repair patch


FIGURE 42. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = d
l
END

γ
p
=
λ

ef
2
x
2
r
1
+s
1
x +
γ
p
=
λ

ef
2
x
2
r
2
+s
2
x + γ
e
=
p + sinh( λ x) q cosh(λ x)
d
d
a
R
11
x
=
2h
i
P
γ
p a
L
d
dx
11
=
h
i
P
γ
p γ
p
γ
e
= = γ
ef
d
dx
γ
e
d
dx
γ
p
=
γ
p
γ
e
= = γ
ef
d
dx
γ
e
d
dx
γ
p
=
γ
x
x
p1
x
p2
Plastic
Elastic
Plastic
γ
ef
d
l
0
at x = d
l
at x = x
p1
at x = 0 at x = x
p2
P
P/2
P/2
Interlayer
Repair patch
Laminate
Repair patch


FIGURE 43. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 AND x = d
l
ENDS
43
7.2.2 Calculation of the Shear Failure Load.
The applied tensile load F which fails the interlayer in shear is calculated by the following steps
using the equations and boundary and continuity conditions summarized in figure 40 through
figure 43.

a. A load P is applied such that under this load the entire interlayer is in the elastic region
(figure 38, left). The shear strain as a function of location x is calculated by the equation
in figure 40. The load is gradually increased, and at each load the shear strain is
calculated. The procedure is repeated until the shear strain reaches the elastic limit γ
ef

either near the x = 0 or near x = d
1
end of the interlayer.

b. The applied load is gradually increased, and at each load the shear strain as a function of
x is calculated by the equation given either in figure 41 (when the plastic region is near
x = 0) or in figure 42 (when the plastic region is near x = d
l
). The procedure is repeated
until the shear strain is at or above the elastic limit at both the x = 0 and x = d
l
ends of the
interlayer.

c. The applied load is gradually increased, and at each load the shear strain as a function of
location x is calculated by the equations given in figure 43.

At each load in steps b. and c. the shear strain is compared to the plastic failure strain γ
pf
. The
load P at which the shear strain γ
p
, at any point in the interlayer, reaches the plastic failure strain
γ
pf
, is taken to be the failure load (P = F).

8. MODEL OF SCARF REPAIR.
Consider a symmetric composite laminate repaired by the scarf technique (see figure 44). When
such a repaired laminate is subjected to a tensile load, the laminate may fail in tension, the repair
patch may fail in tension, the interlayer may fail in shear, the laminate may fail in tension while
the interlayer fails in shear, and the repair patch may fail in tension while the interlayer fails in
shear (see figure 45). The procedure for calculating the loads at which each of these types of
failure occurs is given below.

44
P
P
Repair patch
Scarf length, d
s
(a) Without external ply
(b) With external plies
Overlap length of external plies
P
P
Laminate
Repair patch
Interlayer
Scarf length, d
s
Overlap length per ply,
d
so
External plies
Number of plies, n
l
L
Number of plies, n
l
R
Thickness, h
i
Number of plies, n
e
d
e
d
e


FIGURE 44. SCARF REPAIRS TREATED IN THE MODEL


(a) Laminate failure in tension
(b) Repair patch failure in tension
F F
(c) Interlayer failure in shear
F F
F F
(d) Laminate / interlayer failure
F F
(e) Repair patch / interlayer failure
F F


FIGURE 45. THE TYPES OF FAILURE WHICH MAY OCCUR IN A
LAMINATE REPAIRED BY THE SCARF TECHNIQUE

45
8.1 INTERLAYER SHEAR STRAIN.
To calculate shear strain in the interlayer, consider half of the repaired laminate, as shown in
figure 46. The shear strain is calculated as follows, idealizing the taper as a series of discrete
steps.

P P
Laminate
Repair patch
P P
Interlayer
k=K
k=1
Laminate
Repair patch
x
x
1
x
2
x
K
k=2
k


FIGURE 46. THE SCARF REPAIR TREATED IN THE MODEL

The repaired laminate is subjected to an applied in-plane load P. The superscript k refers to the
k-th overlap segment. For a section dx in length (x being in the direction of the applied load,
figure 46), a force balance for the laminate and the repair patch gives


( )
0 · − τ
k
L k
dx
N d
Laminate
(20)
0 · + τ
k
R k
dx
) N d(
Repair Patch

( )
L
x
k
L
k
N N · and ( )
R
x
k
R
k
N N · are the in-plane loads (per unit width) inside the laminate and
inside the repair patch for the k-th overlap segment, respectively. ( )
zx
k k
τ τ · is the shear stress in
the interlayer. This is illustrated in figure 47.

46


FIGURE 47. LOADS ON A SECTION dx IN LENGTH OF THE k-th SEGMENT OF
THE LAMINATE REPAIRED BY THE SCARF TECHNIQUE

Neglecting shear deformations of the laminate and the repair patch, compatibility requires that
the following condition be satisfied [64] (figure 48)

( ) ( )
( )
i
k
k R
x
k
i
k L
x
k
h dx
dx
d
dx h dx

,
`

.
|
+ + + · + +
γ
γ ε γ ε 1 1 (21)

This can be simplified to give


( )
i
R
x
k L
x
k k
h dx
d ε ε γ −
· (22)

where
L
x
k
ε and
R
x
k
ε are the off-axis in-plane strains in the laminate and in the repair patch,
respectively, h
i
is the thickness of the interlayer, and
k
γ (=
k
γ
zx
) is the interlayer shear strain.

47
(1+
k
ε
R
) dx
k
γ
d(
k
γ)
dx
dx
(1+
k
ε
L
) dx
h
i
L
R
I
Interlayer
Laminate
Repair patch
k
γ +
x
x


FIGURE 48. DEFORMATIONS OF THE REPAIR PATCH, THE INTERLAYER, AND
THE LAMINATE IN THE k-th SEGMENT OF A LAMINATE REPAIRED BY
THE SCARF TECHNIQUE

At any overlap segment the laminate and the repair patch is not necessarily symmetric. The x
components of the strains in the unsymmetric laminate and in the unsymmetric repair patch
are [63]


L k L k L
x
k
N
11
α ε · Laminate
(23)

R k R k R
x
k
N
11
α ε · Repair Patch

where
k
α
11
is the 11 component of the laminate or repair patch compliance matrix in the k-th
overlap segment (see figure 49).

Equations 22 and 23 yield


( )
( )
R k R k L k L k
i
k
N N
h dx
d
11 11
1
α α
γ
− · (24)

By combining equations 20 and 24 one obtains


( )
( ) τ α α
γ
k R L k
i
k
h dx
d
11 11
2
2
1
+ · (25)
48


FIGURE 49. THE COMPLIANCE MATRICES IN THE LAMINATE AND
THE REPAIR PATCH

There are four possible scenarios for the behavior of the interlayer (see figure 50). The entire
interlayer behaves in a linearly elastic manner, there is a perfectly plastic region either near the
x = 0 or near the x = x
K
end of the repair patch, or there are perfectly plastic regions near both the
x = 0 and x = x
K
ends of the repair patch (x
K
is defined in figure 46).

Interlayer
Laminate
Repair patch
P
P
Elastic Elastic Elastic Elastic Plastic Plastic Plastic Plastic
x=0 x=x
K


FIGURE 50. POSSIBLE REGIONS OF THE INTERLAYER UNDER AN APPLIED LOAD P
(From left to the right: interlayer is linearly elastic; interlayer is perfectly plastic near
the x = 0 end; interlayer is perfectly plastic near the x = x
K
end; and interlayer is
perfectly plastic near the x = 0 and x = x
K
ends. (Scarf technique))

In the elastic region, the shear strain and shear stress are


e
k k
γ γ · and
e
k k
G γ τ · (26)

where G is the shear modulus of the interlayer. From equations 25 and 26 the following
expression for the shear strain is obtained

( ) ( ) x cosh q x sinh p
k k k k
e
k
λ λ γ + · Elastic Region (27)
where
k
λ is defined as
( ) ( )
R k L k
i
k
h
G
11 11
2
α α λ + · (28)
49
In the plastic region the shear strain and shear stress are (see figure 31)


p
k k
γ γ · and
ef p
k k
Gγ τ τ · · (29)

where, as shown in figure 31, γ
ef
is constant. Equations 25 and 29 yield


( )
s rx x
k k
ef
k
p
k
+ + ·
2
2
2
γ λ
γ Plastic Region (30)

In equations 27 and 30,
k
p,
k
q,
k
r, and
k
s

are constants which must be determined from the
boundary and continuity conditions.

8.1.1 Boundary and Continuity Conditions.
The terminations of the external plies are modeled as illustrated in figure 51. Then, the boundary
and continuity conditions corresponding to equations 27 and 30 are as follows.

• At the inside edge of the repair patch (x = 0) the axial load (per unit width) is zero in the
laminate, and is equal to the applied load (per unit width) P in the repair patch
(see figure 52). At the outside edge of the repair (x = x
K
) the axial load (per unit width) is
zero in the repair patch, and is equal to the applied load (per unit width) P in the laminate


1
N
L
= 0
at x = 0 (31)

l
N
R
= P

K
N
L
= P
at x = x
K
(32)

K
N
R
= 0

By combining equations 24, 31, and 32, these boundary conditions can be expressed as


( )
i
R
h
P
dx
d
11
1 1
α γ
− · at x=0, (33)
and

( )

,
`

.
|
·
i
L k K
h
P
dx
d
11
α γ

at x = x
K
(34)

In equations 33 and 34
k
γ is either
k
γ
e
or
k
γ
p
depending on whether the interlayer is in the
elastic or in the perfectly plastic region at the boundary.

• At the edge of each overlap segment (at x = x
K
, figure 53) the shear strain in the
interlayer is continuous and the in-plane loads are equal and opposite on the left and right
sides of the segment. It is assumed that the load is transmitted only by continuous layers
50
(for example in figure 53 load is transmitted by the bottom four layers in the laminate and
the top three layers in the repair). Accordingly, at the edge of each segment the following
continuity conditions are applied.


k+1
γ =
k
γ (35)

k+1
N
L

k
N
L
at x = x
k
(36)

k+1
N
R

k
N
R
(37)

By using equations 24, 36, and 37, the above continuity conditions for the k-th segment
can be expressed in terms of
k
γ as follows


k+1
γ =
k
γ (38)


( ) ( )
* P *
dx
d
dx
d
k k
k k
+ ·
+
α
γ γ
1
at x = x
k
(39)

where
k
α
*
and
k
P
*
are defined as


R k L k
R k L k
k
* a
11 11
11
1
11
1
α α
α α
+
+
·
+ +
(40)

,
`

.
|
+

·
+ +
R k L k
L k R k R k L k
i
k
h
P
* P
11 11
11
1
11 11
1
11
α α
α α α α
(41)

In equations 35, 38, and 39 γ = γ
e
when interlayer is in the elastic region, and γ = γ
p
when
interlayer is in the plastic region.

• At the locations where the elastic and plastic regions meet (x = x
p1
and x = x
p2
in
figure 54), the shear strains in the elastic and plastic regions are equal (with the
value γ
ef
) and are continuous (see figures 54 and 56 through 58). Correspondingly,

( )
ef p
k
e
k
γ γ γ · ·
at x = x
p1,
x = x
p2
(42)

( )
( )
dx
d
dx
d
p
k
e
k
γ
γ
·

The locations x
p1
and x
p2
are unknown, and must be determined from the solutions of the
equations summarized in figures 55 through 58.

51
Laminate Laminate
Laminate
No external ply 1 external ply 2 external plies
k=K
k=K
k=K
d
e
d
av
=1.5d
e
d
e
d
e
d
e
x=x
K
x=x
K
x=x
K


FIGURE 51. MODEL OF THE BOUNDARY AT THE OUTER EDGE (x = x
K
) OF
THE REPAIR PATCH

Interlayer
Laminate
Repair patch
at x=x
K
at x=0
K
N
R
=0
K
N
L
=P
1
N
L
=0
1
N
R
=P
P
P


FIGURE 52. THE BOUNDARY CONDITIONS FOR THE IN-PLANE LOADS
IN THE LAMINATE AND IN THE REPAIR PATCH
(Scarf technique)

52
Interlayer
k
N
R
=
k+1
N
R
k
N
L
=
k+1
N
L
k+1
γ
k
γ
=
Interlayer
Laminate
Repair patch
P
P
x=x
k
k
k+1
γ · γ
e if elastic
γ · γ
p if perfect plastic
k
k+1


FIGURE 53. THE CONTINUITY CONDITIONS AT THE EDGE OF THE k-th SEGMENT
(Scarf technique)



FIGURE 54. THE CONTINUITY CONDITIONS AT THE INTERLAYER BETWEEN
THE ELASTIC AND PERFECTLY PLASTIC REGIONS
53
P P
x=0 x=x
K
x
k
Elastic
k
γ
e
=
k
p + sinh(
k
λ
x)
k
q cosh(
k
λ
x)
d(
dx
1
γ
e
)
1
α
R
11
P(
h
i
)
=
at x = 0
K
α
L
11
h
i
P( )
dx
=
d(
K
γ
e
)
at x = x
K
at x = x
k
k+1
γ
e
k
γ
e =
=
k
α
*
d(
dx
k+1
γ
e
)
d(
dx
k
γ
e
)
+
k
P
*
k=1
k=K


FIGURE 55. THE EQUATIONS AND BOUNDARY CONDITIONS FOR CALCULATING
THE SHEAR STRAIN IN THE INTERLAYER WHEN THE ENTIRE INTERLAYER
BEHAVES IN A LINEARLY ELASTIC MANNER

Elastic
k
γ
e
=
k
p + sinh(
k
λ
x)
k
q cosh(
k
λ
x)
at x = 0 at x = x
K
at x = x
k
k
γ
p
=
k
λ

ef
2
x
2 k
r +
k
s
x +
at x = x
p1
x
p1
Plastic
d(
dx
1
γ
p
)
=
k
γ
p
k
γ
e
= = γ
ef
=
dx
=
d(
K
γ
p
)
dx
d(
k
γ
p
)
dx
d(
k
γ
e
)
=
k+1
γ
k
γ
=
k
α
*
d(
dx
k+1
γ) d(
dx
k
γ)
+
k
P
*
γ · γ
e if elastic
γ · γ
p
if perfect plastic
1
α
R
11
P(
h
i
)
K
α
L
11
h
i
P( )
P P
x=0 x=x
K
x
k
k=1
k=K


FIGURE 56. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 END
54
Elastic
k
γ
e
=
k
c
1
+ sinh(
k
λx)
k
c
2
cosh(
k
λx)
at x = 0
at x = x
K
at x = x
k
k
γ
p
=
k
λ

ef
2
x
2 k
c
3
+
k
c
4
x +
k
γ
p
k
γ
e
= = γ
ef
=
at x = x
p1
x
p1
Plastic
d(
dx
1
γ
e
)
= dx
=
d(
K
γ
p
)
dx
d(
k
γ
p
)
dx
d(
k
γ
e
)
=
k+1
γ
k
γ
=
k
α
*
d(
dx
k+1
γ) d(
dx
k
γ)
+
k
P
*
γ · γ
e if elastic
γ · γ
p
if perfect plastic
K
α
L
11
h
i
P( )
1
α
R
11
P(
h
i
)
P P
x=0 x=x
K
x
k
k=1
k=K


FIGURE 57. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = x
K
END

Elastic
k
γ
e
=
k
c
3
+ sinh(
k
λx)
k
c
4
cosh(
k
λx)
at x = 0 at x = x
K
at x = x
k
k
γ
p
=
k
λ

ef
2
x
2 k
c
1
+
k
c
2
x +
at x = x
p1
x
p1
Plastic
x
p2
Plastic
and x = x
p2
k
γ
p
=
k
λ

ef
2
x
2 k
c
5
+
k
c
6
x +
d(
dx
1
γ
p
)
=
k
γ
p
k
γ
e
= = γ
ef
=
dx
d(
k
γ
p
)
dx
d(
k
γ
e
)
dx
=
d(
K
γ
p
)
=
k+1
γ
k
γ
=
k
α
*
d(
dx
k+1
γ) d(
dx
k
γ)
+
k
P
*
γ · γ
e if elastic
γ · γ
p
if perfect plastic
1
α
R
11
P(
h
i
)
K
α
L
11
h
i
P( )
P P
x=0 x=x
K
x
k
k=1
k=K


FIGURE 58. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS
FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE
INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 AND x = x
K
ENDS
55
8.1.2 Calculation of the Interlayer Shear Strain.
The interlayer shear strain is calculated by the following steps using the equations and boundary
and continuity conditions summarized in figures 55 through 58.

a. Values for the applied load P and the interlayer shear strain
1
γ at x = 0 are assumed such
that the shear strain both at x = 0 and at x = x
K
(i.e.,
1
γ and
K
γ) are in the elastic region
(see figure 50, left). The shear strain as a function of location x is calculated by the
equations in figure 55. At the x = x
K
boundary the x derivative of the shear strains is
calculated and compared to the value of
i
L
11
K
h
P α


( )
i
L
11
K k
h
P
dx
d α γ
>
·
<
at x = x
K
(43)

If the left-hand side of this equation is less or higher than the right-hand side, a new value
of
1
γ is assigned. These calculations are repeated until the boundary condition at x = x
K
is
satisfied, i.e., until the left-hand side of the equation becomes equal to the right-hand
side. The applied load is then gradually increased, and at each load the calculation is
repeated until the shear strain reaches the elastic limit γ
ef
either near the x = 0 or near x =
x
K
end of the repair patch.

b. The applied load is gradually increased. At each load the shear strain as a function of x is
calculated by the procedure in step a. using the equations given either in figure 56 (when
the plastic region is near x = 0) or in figure 57 (when the plastic region is near
x = x
K
). The process is repeated until the shear strain is at or above the elastic limit at
both the x = 0 and x = x
K
ends of the interlayer.

c. The applied load is gradually increased in small increments, and at each load the shear
strain as a function of x is calculated by the procedure in step a. using the equations given
in figure 58.

8.2 FAILURE LOAD.
The applied tensile load at which the interlayer fails in shear is determined at each load step
(steps b. and c. in section 8.1.2) by comparing the shear strains to the plastic failure strain γ
pf
.
The load P at which the shear strain
k
γ
p
at any point in the interlayer reaches the plastic failure
strain γ
pf
, is taken to be the interlayer shear failure load.

To determine the applied tensile load under which either the laminate or the repair patch fails,
the in-plane loads in the laminate and the repair patch are calculated as functions of axial
position x
dx 0 (x) N
x
0
k L k

+ · τ Laminate
(44)
dx P (x) N
x
0
k R k

− · τ Repair Patch
56
k
τ is the shear stress having the value of
k
τ = Gγ
e
in the elastic region and
k
τ =
k
τ
p
= Gγ
ef
in the
plastic region.

The off-axis in-plane strains as a function of x in the laminate and in the repair patch are

Laminate Repair Patch

( ) ( ) x N x
L k L k L
x
k
11
α ε · ( ) ( ) x N x
R k R k R
x
k
11
α ε ·
( ) ( ) x N x
L k L k L
y
k
21
α ε · ( ) ( ) x N x
R k R k R
y
k
21
α ε · (45)
( ) ( ) x N x
L k L k L
s
k
61
α ε · ( ) ( ) x N x
R k R k R
s
k
61
α ε ·

where the α
ij
’s are the components of the compliance matrix [63]. The on-axis in-plane strains
of each ply in the laminate or in the repair patch in k-th overlap segment are obtained by
transformation


¹
¹
¹
'
¹
¹
¹
¹
'
¹

,
`

.
|
− −
− ·
¹
¹
¹
'
¹
¹
¹
¹
'
¹
R , L
s
R , L
y
R , L
x
k
R , L R , L R , L R , L R , L R , L
R , L R , L R , L R , L
R , L R , L R , L R , L
k
R , L
R , L
R , L
k
sin cos sin cos sin cos
sin cos cos sin
sin cos sin cos
ε
ε
ε
θ θ θ θ θ θ
θ θ θ θ
θ θ θ θ
ε
ε
ε
2 2
2 2
2 2
6
2
1
2 2
(46)

Failure is calculated by the maximum strain failure criteria

Laminate Repair Patch


L k
L k L k
L k
p
E
X
·
1 1
ε

R k
R k R k
R k
p
E
X
·
1 1
ε


L k
L k L k
L k
p
E
Y
·
2 2
ε

R k
R k R k
R k
p
E
Y
·
2 2
ε
(47)

L k
L k L k
L k
p
E
S
·
6 6
ε

R k
R k R k
R k
p
E
S
·
6 6
ε


1 <
R , L k
p Failure
1 ≥
R , L k
p No Failure (48)

k
X,
k
Y, and
k
S are the on-axis longitudinal, transverse, and shear strengths,
k
E
1
,
k
E
2
, and
k
E
6
are the
on-axis longitudinal, transverse, and shear moduli of the ply in k-th segment and p are loads.

The above failure criteria is evaluated in each laminate and repair ply at every overlap segment.

57
Failure occurs when under the applied load p is less than unity in any ply.

• If failure occurs at the x = 0 end of the repaired laminate, the failure is considered to have
occurred in the repair patch (see figure 45, b).

• If failure occurs at the x = x
K
end of the repaired laminate, the failure is considered to
have occurred in the laminate (see figure 45, a).

• If failure occurs in the laminate or the repair patch at one of the overlap segments, then
failure is considered to have occurred due to a combination of laminate (or repair patch)
failure and interlayer failure (see figure 45, d and e).

9. RESULTS.
Two computer programs were written in MATLAB to generate numerical results for the two
models described in the previous two chapters. These codes are designated as “RepairL” and
“RepairS” and apply, respectively, to the double-sided uniform lap repair and the scarf repair.
The input parameters required by the programs are listed in tables 13 and 14. The properties of
the interlayer were obtained by matching the model to one data point. The properties selected by
such a “backcalculation” procedure were then used in all subsequent calculations. The programs
provide the tensile failure loads in the laminate and in the repair patch and the shear failure load
in the interlayer. The lowest of these three is the failure load of the repaired composite.

Failure loads generated by the model and the corresponding computer codes were compared to
data generated with repaired specimens specified in figures 5 and 14. For the uniform lap repair,
comparisons were made for 8- and 32-ply laminates repaired with either 1, 2, or 3 repair plies.
The repair lap length varied from 0 to 3 inches. For scarf repair, comparisons were made for
different scarf angles (β = 1.07, 1.43, and 2.15 degree) and for different number of external plies
(n
e
= 0, 1, and 2). The material properties and parameters used in the calculations are given in
tables 13 and 14.


58
TABLE 13. THE INPUT PARAMETERS REQUIRED BY RepairL AND THE NUMERICAL
VALUES USED IN THE PRESENT CALCULATIONS FOR 3k70 PLAIN WEAVE FABRIC
IMPREGNATED WITH 9396 RESIN
Specimen Condition/Test Temperature
Category Parameter
Dry/70°F Moist(H
2
O)/180°F
Laminate*
Longitudinal modulus,
L
E
1
(Msi)
[a]

8.1 8.5

Transverse modulus,
L
E
2
(Msi)
[a]

8.1 8.5

Shear modulus,
L
E
6
(Msi)
[b]

0.48 0.45

Longitudinal Poisson’s ratio,
[ ] b L
1
υ
0.045 0.045

Transverse Poisson’s ratio,
[ ] b L
2
υ
0.045 0.045
Longitudinal tensile strength, X
L
(ksi)
[a]
101 93
Transverse tensile strength, Y
L
(ksi)
[a]
101 93
Shear strength, S
L
(ksi)
[b]
5.0 4.5
Ply thickness, h
L
(in)
[a]
0.0094 0.0094
Orientation of each ply, θ
L
(degree) 0 0
Number of plies, n
l
L
8, 32 8, 32
Repair Patch* Longitudinal modulus, E
1
R
(Msi)
[a]
7.2 7.5
Transverse modulus, E
2
R
(Msi)
[a]
7.2 7.5
Shear modulus, E
6
R
(Msi)
[b]
0.48 0.45
Longitudinal Poisson’s ratio, ν
1
R[b]
0.045 0.045
Transverse Poisson’s ratio, ν
2
R[b]
0.045 0.045
Longitudinal tensile strength, X
R
(ksi)
[a]
75 68
Transverse tensile strength, Y
R
(ksi)
[a]
75 68
Shear strength, S
R
(ksi)
[b]
5.0 4.5
Ply thickness, h
R
(in) 0.0094 0.0094
Orientation of each ply, θ
R
(degree) 0 0
Number of plies per side, n
l
R
1,2,3 1,2,3
Maximum lap length, d
l
(in) 0 to 3.0 0 to 3.0
Interlayer Shear modulus, G (ksi)]
[c]
200 100
Shear strength, τ
p
(ksi)]
[c]
2.0 1.6
Maximum shear strain, γ
pf
[c]
0.28 0.30
Thickness, h
i
(in)
[c]
0.008 0.008

* Parameters listed are for each ply.
[a] from tables 2 and 3.
[b] from Naik [65]
[c] by backcalculation.

59
TABLE 14. THE INPUT PARAMETERS REQUIRED BY RepairS AND
THE NUMERICAL VALUES USED IN THE PRESENT CALCULATIONS FOR
3k70 PLAIN WEAVE FABRIC IMPREGNATED WITH 9396 RESIN
Specimen Condition/Test Temperature
Category Parameter
Dry/70°F Moist(H
2
O)/180°F
Laminate*
Longitudinal modulus,
L
E
1
(Msi)
[a]

8.1 8.5

Transverse modulus,
L
E
2
(Msi)
[a]

8.1 8.5

Shear modulus,
L
E
6
(Msi)
[b]

0.48 0.45

Longitudinal Poisson’s ratio,
[ ] b L
1
υ
0.045 0.045

Transverse Poisson’s ratio,
[ ] b L
2
υ
0.045 0.045
Longitudinal tensile strength, X
L
(ksi)
[a]
101 93
Transverse tensile strength, Y
L
(ksi)
[a]
101 93
Shear strength, S
L
(ksi)
[b]
5.0 4.5
Ply thickness, h
L
(in) 0.0094 0.0094
Orientation of each θ
L
(degree) [(0/45)
2
]
s
[(0/45)
2
]
s

Number of plies, n
s
L
8 8
Scarf angle, β (degree) 1 to 2.5 1 to 2.5
Repair Patch* Longitudinal modulus, E
1
R
(Msi)
[a]
7.2 7.5
Transverse modulus, E
2
R
(Msi)
[a]
7.2 7.5
Shear modulus, E
6
R
(Msi)
[b]
0.48 0.45
Longitudinal Poisson’s ratio, ν
1
R[b]
0.045 0.045
Transverse Poisson’s ratio, ν
2
R[b]
0.045 0.045
Longitudinal tensile strength, X
R
(ksi)
[a]
75 68
Transverse tensile strength, Y
R
(ksi)
[a]
75 68
Shear strength, S
R
(ksi)
[b]
5.0 4.5
Ply thickness, h
R
(in) 0.0094 0.0094
Orientation of each ply, θ
R
(degree) [(0/45)
2
]
s
[(0/45)
2
]
s

Orientation of external ply, θ
e
(degree) 0 0
Number of plies, n
s
R
8 8
Number of external plies, n
e
0,1,2 0,1,2
Overlap length of external ply, d
e
(in) 0.25 to 0.375 0.25 to 0.375
Interlayer Shear modulus, G (ksi)]
[c]
200 100
Shear strength, τ
p
(ksi)]
[c]
2.0 1.6
Maximum shear strain, γ
pf
[c]
0.28 0.30
Thickness, h
i
(in)
[c]
0.008 0.008

* Parameters listed are for each ply.
[a] from tables 2 and 3.
[b] from Naik [65].
[c] by backcalculation.

60
The failure loads calculated by the model are compared to the data in figures 59 through 62. The
calculated and measured failure loads agree well, both for dry specimens tested at 70°F and for
moisturized specimens tested at 180°F. These agreements lend support to the validities of the
models, keeping in mind that the first failure prediction at each environment was used to obtain
interlayer properties which were subsequently used for all other predictions. Thus, one can
conclude that the trends are very well predicted by the model but not necessarily the absolute
values of the failure load.

2000
4000
6000
8000
F
a
i
l
u
r
e

L
o
a
d

(
l
b
f
)
2000
4000
6000
8000
2000
4000
6000
8000
0
1.0 2.0 3.0 4.0
Lap length, d
l
(in)
0.0 1.0 2.0 3.0 4.0
Laminate: F593
Repair: 9396
Data
Model
Specimen: Dry
Test: 70F
Specimen: Moist (H O)
Test: 180 F
2
1 repair ply
2 repair plies
3 repair plies
d
l
8
[0]


FIGURE 59. COMPARISONS OF THE CALCULATED (MODEL) AND THE
MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF
LAP LENGTH FOR 8-PLY LAMINATE


61
2
0
0
0
4
0
0
0
6
0
0
0
8
0
0
0
2
0
0
0
4
0
0
0
6
0
0
0
8
0
0
0 0
2
0
0
0
4
0
0
0
6
0
0
0
8
0
0
0
1
.
0
2
.
0
3
.
0
4
.
0
L
a
p

l
e
n
g
t
h
,

d
l


(
i
n
)
0
.
0
1
.
0
2
.
0
3
.
0
4
.
0
1

r
e
p
a
i
r

p
l
y
2

r
e
p
a
i
r

p
l
i
e
s
3

r
e
p
a
i
r

p
l
i
e
s
S
p
e
c
i
m
e
n
:

D
r
y
T
e
s
t
:

7
0
F
S
p
e
c
i
m
e
n
:

M
o
i
s
t

(
H

O
)
T
e
s
t
:

1
8
0

F
2
L
a
m
i
n
a
t
e
:

F
5
9
3
R
e
p
a
i
r
:

9
3
9
6
D
a
t
a
M
o
d
e
l
d
l
F a i l u r e L o a d ( l b f )
[
0
]
3
2


0
1
0
0
0
2
0
0
0
3
0
0
0
4
0
0
0
5
0
0
0
6
0
0
0
F a i l u r e l o a d ( l b f )
1
1
.
5
2
S
c
a
r
f

A
n
g
l
e
,
β
(
d
e
g
r
e
e
)
0
1
0
0
0
2
0
0
0
3
0
0
0
4
0
0
0
5
0
0
0
6
0
0
0
7
0
0
0
S
p
e
c
i
m
e
n
:

D
r
y
T
e
s
t
:

7
0
F
D
a
t
a
M
o
d
e
l
S
p
e
c
i
m
e
n
:
M
o
i
s
t
-
H

O
T
e
s
t
:

1
8
0

F
2
L
a
m
i
n
a
t
e
:

F
5
9
3
R
e
p
a
i
r
:

9
3
9
6
t
a
n
(
β
)
=
H d
s
d
s
H


F
I
G
U
R
E

6
0
.


C
O
M
P
A
R
I
S
O
N

O
F

T
H
E

C
A
L
C
U
L
A
T
E
D

(
M
O
D
E
L
)

A
N
D

T
H
E

M
E
A
S
U
R
E
D

(
D
A
T
A
)

F
A
I
L
U
R
E


L
O
A
D
S

A
S

A

F
U
N
C
T
I
O
N

O
F

L
A
P

L
E
N
G
T
H

F
O
R


3
2
-
P
L
Y

L
A
M
I
N
A
T
E

F
I
G
U
R
E

6
1
.


C
O
M
P
A
R
I
S
O
N
S

O
F

T
H
E

C
A
L
C
U
L
A
T
E
D

(
M
O
D
E
L
)

A
N
D

T
H
E

M
E
A
S
U
R
E
D

(
D
A
T
A
)

F
A
I
L
U
R
E


L
O
A
D
S

A
S

A

F
U
N
C
T
I
O
N

O
F

S
C
A
R
F

A
N
G
L
E


62
0
1000
2000
3000
4000
5000
6000
7000
0
1000
2000
3000
4000
5000
6000
F
a
i
l
u
r
e

l
o
a
d

(
l
b
f
)
No Extra Ply 1 Ply 2 Plies
Number of Extra Plies, n
Specimen: Dry
Test: 70F
Laminate: F593
Repair: 9396
Specimen: Moist-H O
Test: 180 F
2
Data
Model
Scarf angle,
α=1.07 degree
e


FIGURE 62. COMPARISON OF THE CALCULATED (MODEL) AND THE
MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF
EXTRA PLIES

10. CONCLUDING REMARKS.
The models and the corresponding computer codes described in the previous chapters are
applicable to one-dimensional specimens. Nonetheless, the results of these models serve as
useful guides in the design of actual repairs by providing information on the various parameters
affecting the failure load. In the present forms the models are for composite laminates.
However, the models may be applied to the repair of isotropic plates by replacing the properties
of the composite with the corresponding properties of an isotropic material.




Scarf angle,
β = 1.07 degree
Specimen: Dry
Test: 70° °° °F
Specimen: Moist-H
2
O
Test: 180° °° °F
63
11. REFERENCES.
1. Boeing 777 Structural Repair Manual. Boeing Commercial Aircraft Co., 1995.

2. Boeing 737 Structural Repair Manual. Boeing Commercial Aircraft Co., 1991.

3. Repair Procedures for Composite Structures, D-53900. Boeing Commercial Aircraft
Co., 1993.

4. E. F. Chesmar, “Metal Bond and Composite Repairs: Similarities and Differences,”
Composites ’96 Manufacturing and Tooling Conference, Society of Manufacturing
Engineers, pp. 281-290, 1996.

5. R. F. Wegman and T. R. Tullos, “Adhesive Bonded Structural Repair, Part III-Repair of
Composite, Honeycomb Cored, and Solid Cored Structures,” SAMPE Journal, Vol. 29,
No. 6, pp. 8-12, 1993.

6. F. Lee, S. Brinkerhoff, and S. McKinney, “Low Energy Cured Composite Repair
System,” Composite Repairs, SAMPE Monograph No. 2 (C. L. Hamermesh, ed.),
pp. 24-35, 1991.

7. M. J. Cichon, “Repair Adhesives: Development Criteria for Field Level Conditions,”
Composite Repairs, SAMPE Monograph No.2 (C. L. Hamermesh, ed.), pp. 36-50, 1991.

8. R. S. Riefier and W. D. Steinmetz, “Development of Rapid Cure Adhesive for Naval
Aircraft Field Repair Applications,” Composite Repairs, SAMPE Monograph No. 2 (C. L.
Hamermesh, ed.), pp. 51-62, 1991.

9. E. C. Clark and K. D. Cressy, “Field Repair Compounds for Thermoset and
Thermoplastic Composites,” Composite Repairs, SAMPE Monograph No. 2
(C. Hamermesh, ed.), pp. 63-70, 1991.

10. C. L. Ong and S. B. Shen, “Repair of F-104 Aircraft Nosedome by Composite Patching,”
Theoretical and Applied Fracture Mechanics, Vol. 15, pp. 75-83, 1991.

11. S. R. Hall, M. D. Raizenne, and D. L. Simpson, “A Proposed Composite Repair
Methodology for Primary Structure,” Composites, Vol. 20, pp. 479-483, 1988.

12. J. W. Deaton, “Repair of Advanced Composites Commercial Aircraft Structures,”
Engineered Materials Handbook, ASM International, Vol. 3, pp. 829-839, 1987.

13. S. H. Myhre, “Advanced Composite Repair-Recent Development and Some Problems,”
Composite Repairs, SAMPE Monograph No. 1 (H. Brown, ed.), pp. 14-25, 1985.

64
14. R. H. Stone, “Development of Repair Procedures for Graphite/Epoxy Structures on
Commercial Transports,” Composite Repairs, SAMPE Monograph No. 1 (H. Brown, ed.),
pp. 26-39, 1985.

15. J. F. Knauss and R. H. Stone, “Demonstration of Repairability and Repair Quality on
Graphite/Epoxy Structural Supplements,” Composite Repairs, SAMPE Monograph No. 1
(H. Brown, ed.), pp. 40-51, 1985.

16. R. H. Stone, “Field-Level Repair Materials and Processes,” Composite Repairs, SAMPE
Monograph No. 1 (H. Brown, ed.), pp. 87-99, 1985.

17. A. S. Falcone, “Field Repairs for the AH-1 Composite Main Rotor Blade,” Composite
Repairs, SAMPE Monograph No. 1 (H. Brown, ed.), pp. 100-109, 1985.

18. J. R. Scott and C. K. Mashiba, “Improved Resins for Wet Layup Repair of Advanced
Composite Structure,” Composite Repairs, SAMPE Monograph No. 1 (H. Brown, ed.),
pp. 110-116, 1985.

19. J. Mahon and J. Candella, “Development of Large-Area Damage Repair Procedures for
the F-14A Horizontal Stabilizer, Phase I-Repair Development,” Composite Repairs,
SAMPE Monograph No. 1 (H. Brown, ed.), pp. 184-194, 1985.

20. R. H. Stone, Repair Techniques for Graphite/Epoxy Structures for Commercial Transport
Applications. NASA Contractor Report, CR 159056, 1983.

21. Advanced Composite Repair Guide, Nor 82-60, Northrop Corporation, 1982.

22. S. H. Myhre and J. D. Labor, “Repair of Advanced Composite Structures,” Journal of
Aircraft, Vol. 18, pp. 546-552, 1980.

23. S. H. Myhre and R. W. Kiger, “Problems and Options in Advanced Composite Repair,”
Fibrous Composites in Structural Design, Plenum Press, pp. 359-380, 1981.

24. L. O. Bardygula, “CACRC: Progress and Plans,” Composites ’96 Manufacturing and
Tooling Conference, Society of Manufacturing Engineers, pp. 229-239, 1996.

25. P. Shyprykevich, C. S. Springer, and S. H. Ahn, “Standardization of Composite Repairs
for Commercial Transport Aircraft,” Composite Repair of Aircraft Structures Symposium,
pp. 1-24, 1995.

26. S. H. Ahn, G. S. Springer, and P. Shyprykevich, “Composite Repair with Wet Lay-up and
Prepreg,” Composites ’96 Manufacturing and Tooling Conference, Society of
Manufacturing Engineers, pp. 211-226, 1996.

65
27. L. J. Hart-Smith, R. W. Ochsner, and R. L. Radecky, “Surface Preparation of Composites
for Adhesive-Bonded Repair,” Engineered Materials Handbook, ASM International,
Vol. 3, pp. 840-844, 1987.

28. H. S. Kim, S. J. Lee, and D. G. Lee, “Development of a Strength Model for the Cocured
Stepped Lap Joints Under Tensile Loading,” Composite Structures, Vol. 32, pp. 593-600,
1995.

29. S. H. Myhre and C. E. Beck, “Repair Concepts for Advanced Composite Structures,”
Journal of Aircraft, Vol. 16, pp. 720-728, 1979.

30. W. S. Chan and C. T. Sun, “Interfacial Stresses and Strength of Lap Joints,” 21st
Conference of Structures, Structural Dynamics, and Materials, AIAA/ASME/ASCE/AHS,
pp. 656-663, 1980.

31. S. J. John, A. J. Kinloch, and F. L. Matthews, “Measuring and Predicting the Durability
of Bonded Carbon Fiber/Epoxy Composite Joints,” Composites, Vol. 22, pp. 121-126,
1991.

32. J. E. Robson, F. L. Matthews, and A. J. Kinloch, “The Bonded Repair of Fibre
Composite: Effect of Composite Moisture Content,” Composite Science and Technology,
Vol. 52, pp. 235-246, 1994.

33. B. M. Parker, “The Effect of Composite Prebond Moisture on Adhesive-Bonded CFRP-
CFRP Joints,” Composites, Vol. 14, pp. 226-232, 1983.

34. S. H. Mylire, J. D. Labor, and S. C. Aker, “Moisture Problems in Advanced Composite
Structural Repair,” Composites, Vol. 13, pp. 289-299, 1982.

35. O. Volkersen, “Die Nietkraftverteilung in Zugbeanspruchten Nietverbindungen mit
Konstanten Laschenquerschnitten,” Luftfahrtforschung, Vol. 15, pp. 41-47, 1938.

36. M. Goland and E. Reissner, “The Stresses in Cemented Joints,” Journal of Applied
Mechanics, Vol.11, pp. A17-A27, 1944.

37. J. Pirvics, “Two Dimensional Displacement-Stress Distributions in Adhesive Bonded
Composite Structures,” Journal of Adhesion, Vol. 6, pp. 207-228, 1974.

38. D. J. Aliman, “A Theory for Elastic Stresses in Adhesive Bonded Lap Joints,” Quarterly
Journal of Applied Mathematics, Vol. XXX, Pt. 4, pp. 415-436, 1977.

39. I. U. Ojalvo and H. L. Eidinoff, “Bond Thickness Effects Upon Stresses in Single-Lap
Adhesive Joints,” AIAA Journal, Vol. 16, pp. 204-211, 1978.

40. W. J. Renton and J. R. Vinson, “The Efficient Design of Adhesive Bonded Joints,”
Journal of Adhesion, Vol.7, pp.175-193, 1975.
66
41. F. Thamm, “Stress Distribution in Lap Joints With Partially Thinned Adherends,”
Journal of Adhesion, Vol. 7, pp. 301-309, 1976.

42. G. R. Wooley and D. R. Carver, “Stress Concentration Factors for Bonded Lap Joints,”
Journal of Aircraft, Vol. 8, pp. 817-820, 1971.

43. R. D. Adams and N. A. Peppiatt, “Stress Analysis of Adhesive-Bonded Lap Joints,”
Journal of Strain Analysis, Vol. 9, pp. 185-196, 1974.

44. D. A. Bigwood and A. D. Crocombe, “Elastic Analysis and Engineering Design
Formulae for Bonded Joints,” International Journal of Adhesion and Adhesives, Vol. 9,
pp. 229-242, 1989.

45. D. A. Bigwood and A. D. Crocombe, “Nonlinear Adhesive Bonded Joints Design
Analyses,” International Journal of Adhesion and Adhesives, Vol. 10, pp. 31-41, 1990.

46. A. D. Crocombe, “Global Yielding as a Failure Criterion for Bonded Joints,”
International Journal of Adhesion and Adhesives, Vol. 9, pp. 145-153, 1989.

47. F. Szépe, “Strength of Adhesive-Bonded Lap Joints With Respect to Change of
Temperature and Fatigue,” Experimental Mechanics, Vol. 6, pp. 280-286, 1966.

48. R. D. Adams, J. Coppendale, and N. A. Peppiatt, “Failure Analysis of Aluminuium-
Aluminuium Bonded Joints,” Adhesion 2 (K. W. Allen, ed.), Applied Science Publisher,
pp. 105-120, 1978.

49. J. L. Lubkin, “A Theory of Adhesive Scarf Joints,” Journal of Applied Mechanics,
Vol. 24, pp. 255-260, 1957.

50. W. J. Renton and J. R. Vinson, “On The Behavior of Bonded Joints in Composite
Material Structures,” Engineering Fracture Mechanics, Vol. 7, pp. 41-60, 1975.

51. W. J. Renton and J. R. Vinson, “Analysis of Adhesively Bonded Joints Between Panels
of Composite Materials,” Journal of Applied Mechanics, Vol. 44, pp. 101-106, 1977.

52. R. M. Baker and F. Hatt, “Analysis of Bonded Joints in Vehicular Structures,” AIAA
Journal, Vol. 11, pp. 1650-1654, 1973.

53. G. C. Grimes, “Stress Distribution in Adhesive Bonded Lap Joints,” SAE Trans. 710107,
pp. 370-378, 1971.

54. F. Erdogan and M. Ratwani, “Stress Distribution in Bonded Joints,” Journal of
Composite Materials, Vol. 5, pp. 379-393, 1971.

55. C. L. Johnson, “Effect of Ply Staking Sequence on Stress in a Scarf Joint,” AIAA Journal,
Vol. 27, pp. 79-86, 1989.
67/68
56. L. J. Hart-Smith, Adhesive-Bonded Single-Lap Joints. NASA Contract Report,
CR 112236,1973.

57. L. J. Hart-Smith, Adhesive-Bonded Double-Lap Joints. NASA Contract Report,
CR 112235, 1973.

58. L. J. Hart-Smith, Non-Classical Adhesive-Bonded Joints in Practical Aerospace
Construction, NASA Contract Report, CR 112238,1973.

59. L. J. Hart-Smith, “Further Developments in the Design and Analysis of Adhesive-Bonded
Structural Joints,” Joining of Composite Materials, ASTM STP 749 (K. T. Kedward, ed.),
American Society for Testing and Materials, pp. 3-31, 1981.

60. R. D. Adams, R. W. Atkins, J. A. Harris, and A. J. Kinloch, “Stress Analysis and Failure
Properties of Carbon-Fiber-Reinforced-Plastic/Steel Double-Lap Joints,” Journal of
Adhesion, Vol. 20, pp. 29-53, 1986.

61. R. D. Adams, “Theoretical Stress Analysis of Adhesively Bonded Joints,” Joining Fibre-
Reinforced Plastics, (F. L. Matthews, ed.), Elsevier Applied Science Publisher, pp. 185-
226, 1987.

62. L. J. Hart-Smith, Adhesive-Bonded Scarf and Stepped-Lap Joints, NASA Contract
Report, CR 112237, 1973.

63. S. W. Tsai and H. T. Hahn, Introduction to Composite Materials, Technomic Publishing
Company, 1980.

64. T. H. G. Megson, Aircraft Structures for Engineering Students, 2nd ed., Halsted Press,
pp. 405-406, 1990.

65. N. J. Naik, Woven Fabric Composites, Technomic Publishing Co., 1994.



A-1
APPENDIX A÷MOISTURE LOSS OF SPECIMEN DURING TEST

In this appendix, the percent change of the moisture content of the specimen, tested at 180°F, is
given. The measured moisture content as a function of time is shown in figure A-1 for laminates
moisturized at 180°F. The diffusivity D is obtained from the expression [A-1]


2
1 2
1 2
2
4

,
`

.
|

,
`

.
|
·
t t
M M
M
h
D
m
π (A-1)
= 6 x 10
-10
in
2
/sec

where M
m
is the maximum moisture content (% weight gain) ( M
m
= 1.8 percent) and h is the
specimen thickness (h = 0.075 in). M
1
and M
2
are the moisture contents at time t
1
and t
2
,
respectively. The change in the moisture content as a function of time is [A-1]

M = G
M
(M
m
- M
i
) + M
i
(A-2)

where M is the moisture content, M
i
is the initial moisture content, and G
M
is

,
`

.
|

,
`

.
|
·
75 0.
2
M
h
Dt
7.3 - exp - 1 G

(A-3)

The initial moisture content was M
i
= 1.1 percent, and M
m
= 0.0 percent for drying. The duration
of the test was 600 seconds. With these values, equations A-2 and A-3 give

M = 1.093 percent (A-4)

The change in moisture content is

M
i
- M = 1.1 percent - 1.093 percent (A-5)

= 0.007 percent (A-6)


A-2
0.0
0.5
1.0
1.5
2.0
M
o
i
s
t
u
r
e

C
o
n
t
e
n
t

o
f

L
a
m
i
n
a
t
e

M


(
P
e
r
c
e
n
t
)
0.0
0.5
1.0
1.5
0 500 1000
sec
0 500 1000 1500
in 180 F H O
Data
Fit to data
to 180 F humid air
Laminate: F593 [0]
8
Laminate: F593 [(0/45) ]
2
Laminate Immersed
2
Laminate exposed
sec
s


FIGURE A-1. MOISTURE CONTENTS OF THE BASE LAMINATES IMMERSED IN
WATER AT 180°F (LEFT), OR EXPOSED TO 100 PERCENT HUMID
AIR AT 180°F (RIGHT)

REFERENCE.

A-1. H. Shen and G. S. Springer, “Moisture Absorption and Desorption of Composite
Materials,” Journal of Composite Materials, Vol. 10, pp. 2-20, 1976.



B-1/B-2
APPENDIX B÷FAILURE LOADS OF SPECIMENS EXPOSED TO HUMID AIR

The failure loads of the specimens repaired by the scarf (figure B-1) and stepped lap (figure B-2)
techniques are included in this appendix.


0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
Prepreg
Wet lay-up
R922 Laminate: F593
R6376
Specimen: Moist-Humid Air
Test: 180 F
9396 9390 Epocast M20
Repair Material


FIGURE B-1. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH
DIFFERENT REPAIR MATERIALS (SCARF REPAIR)

0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
F
a
i
l
u
r
e

L
o
a
d



(

R
e
p
a
i
r
e
d

/

U
n
d
a
m
a
g
e
d
)
R6376
Laminate: F593 R922
9396 9390 Epocast M20
Repair Material
Prepreg
Wet lay-up
Specimen: Moist-Humid Air
Test: 180 F


FIGURE B-2. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED
WITH DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR)
C-1
APPENDIX C÷TEST RESULTS

The failure loads used in normalizing the data are summarized in tables C-1 through C-5.

TABLE C-1. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE
DATA IN FIGURES 17, 18, AND 21

Specimen Condition/
Test Temperature

Repair Type


Parameter
Dry/70°F Moist(H
2
O)/180°F
Stepped lap 1 repair ply (d
l
= 3 in) 1583 1295
2 repair ply (d
l
= 3 in) 3893 2394
3 repair ply (d
l
= 3 in) 4289 3224
Uniform lap 8 laminate, 1 repair plies (d
l
= 3 in) 1601 1290
8 laminate, 2 repair plies d
l
= 3 in) 3134 2521
8 laminate, 3 repair plies (d
l
= 3 in) 4368 3168
8 laminate, 4 repair plies (d
l
= 3 in) 5646 3858
32 laminate, 1 repair plies (d
l
= 3 in) 1710 1238
32 laminate, 2 repair plies (d
l
= 3 in) 2997 2502
32 laminate, 3 repair plies (d
l
= 3 in) 4055 3256


TABLE C-2. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE
DATA IN FIGURES 22, 23, AND D-1


Repair Type
Specimen Condition/
Test Temperature
Failure Load
(lbs)
Scarf Dry/70°F 4526
Stepped lap Dry/70°F 3360

TABLE C-3. THE FAILURE LOADS (in lbf) OF SPECIMENS WITH GRIT
NUMBER g = 60 USED IN NORMALIZING THE DATA IN FIGURE 24


Repair Type
Specimen Condition/
Test Temperature
Failure Load
(lbs)
Scarf Dry/70°F 4475
Stepped lap Dry/70°F 3278
Scarf Moist (H
2
O)/180°F 3536
Stepped lap Moist (H
2
O)/180°F 3012

C-2
TABLE C-4. THE FAILURE LOADS (in lbf) OF SPECIMENS WITH CURE
TEMPERATURE T
c
= 150°F USED IN NORMALIZING THE DATA IN FIGURE 28


Repair Type
Specimen Condition/
Test Temperature
Failure Load
(lbs)
Scarf Dry/70°F 4553
Stepped lap Dry/70°F 3365

TABLE C-5. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE
DATA IN FIGURE D-2


Repair Type
Specimen Condition/
Test Temperature
Failure Load
(lbs)
Scarf Dry/70°F 4541
Stepped lap Dry/70°F 4516



D-1/D-2
APPENDIX D÷FAILURE LOADS OF SPECIMENS UNDER MOISTURE AND
TEMPERATURE

The failure loads of the specimens repaired with wet lay-up (figure D-l) and with prepreg
(figure D-2) materials are included in this appendix.

Dry & 180F Moist & 70F Moist & 180F
Specimen Condition & Test Temperature
s
p
e
c
i
m
e
n

a
t

i
n
d
i
c
a
t
e
d

c
o
n
d
i
t
i
o
n

d
r
y

s
p
e
c
i
m
e
n

t
e
s
t
e
d

a
t

7
0
F

)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Laminate: F593
Repair: 9396
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
(
F
a
i
l
u
r
e

L
o
a
d


FIGURE D-1. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST
TEMPERATURE ON THE FAILURE LOAD

Dry & 180F Moist & 70F Moist & 180F
Specimen Condition & Test Temperature
s
p
e
c
i
m
e
n

a
t

i
n
d
i
c
a
t
e
d

c
o
n
d
i
t
i
o
n

d
r
y

s
p
e
c
i
m
e
n

t
e
s
t
e
d

a
t

7
0
F

)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Laminate: F593
Repair: M20
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
(
F
a
i
l
u
r
e

L
o
a
d


FIGURE D-2. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST
TEMPERATURE ON THE FAILURE LOAD. REPAIR MATERIAL M20 (PREPREG).

NOTICE This document is disseminated under the sponsorship of the U.S. Department of Transportation in the interest of information exchange. The United States Government assumes no liability for the contents or use thereof. The United States Government does not endorse products or manufacturers. Trade or manufacturer's names appear herein solely because they are considered essential to the objective of this report. This document does not constitute FAA certification policy. Consult your local FAA aircraft certification office as to its use.

This report is available at the Federal Aviation Administration William J. Hughes Technical Center's Full-Text Technical Reports page: actlibrary.tc.faa.gov in Adobe Acrobat portable document format (PDF).

Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

DOT/FAA/AR-00/46
4. Title and Subtitle 5. Report Date

REPAIR OF COMPOSITE LAMINATES

December 2000
6. Performing Organization Code

7. Author(s)

8. Performing Organization Report No.

Sung-Hoon Ahn and George S. Springer
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)

Department of Aeronautics and Astronautics Stanford University Stanford, CA 94305
12. Sponsoring Agency Name and Address

11. Contract or Grant No.

13. Type of Report and Period Covered

U.S. Department of Transportation Federal Aviation Administration Office of Aviation Research Washington, DC 20591
15. Supplementary Notes

Final Report
14. Sponsoring Agency Code

ACE-110

The FAA William J. Hughes Technical Center COTR was Mr. Peter Shyprykevich.
16. Abstract

The effectiveness of repair of damaged fiber reinforced composite laminates was investigated; the effectiveness of repair being assessed by the tensile failure load of the repaired laminate. First, tests were conducted measuring the failure loads of laminates repaired either by the scarf, the uniform lap, or the stepped lap technique. Data were generated with the following parameters having been varied: type of material of the damaged laminate, type of repair material, scarf angle and number of external plies in scarf repair, length and number of repair plies in uniform and stepped lap repair, moisture content of the laminate prior to repair, moisture content of the laminate after repair, test temperature, roughness of grinding tool used in preparing the repair surface, and the temperature applied during the cure of the repair patch. The aforementioned parameters were varied over wide ranges, and provided systematic sets of data which shed light on the influence of each of these parameters on the effectiveness of the repair. Second, models were developed for calculating the failure loads of composite laminates repaired by scarf and uniform lap techniques. The models take into account anisotropy of each ply in the laminate and in the repair ply and nonelastic behavior of the adhesive or resin interlayer between the laminate and the repair patch. On the basis of the models, two computer codes were written for generating numerical values of the failure loads. The first code is for calculating the failure loads of laminates repaired by the uniform lap technique, the second for calculating the failure loads of laminates repaired by the scarf technique. Failure loads calculated by the models and the corresponding computer codes were compared with data and good agreement was found between the results of the analysis and the tests.

17. Key Words

18. Distribution Statement

Bonded repairs, Composites, Analysis methods, Parameter studies
19. Security Classif. (of this report)

This document is available to the public through the National Technical Information Service (NTIS) Springfield, Virginia 22161.
21. No. of Pages 22. Price

20. Security Classif. (of this page)

Unclassified
Form DOT F1700.7
(8-72)

Unclassified
Reproduction of completed page authorized

85

TABLE OF CONTENTS Page EXECUTIVE SUMMARY 1. 2. INTRODUCTION THE PROBLEM xi 1 1 3 3 10 11 16 22 22 24 25 26 29 30 30 30 32 34 35 39 43 43 45 49

PART I. EXPERIMENTAL RESULTS 3. 4. EXPERIMENTAL TEST RESULTS 4.1 4.2 4.3 Type of Repair Material Geometry Effects Effects of Specimen Moisture Content and Test Temperature 4.3.1 Base Laminates Premoisturized 4.3.2 Specimen Moisture Content and Test Temperature 4.4 4.5 4.6 Effect of Sanding Effect of Cure Cycle Summary

PART II. ANALYTICAL MODELS 5. 6. 7. INTRODUCTION THE PROBLEM MODEL OF UNIFORM LAP REPAIR 7.1 7.2 Tensile Failure of the Laminate or Repair Patch Shear Failure 7.2.1 Boundary and Continuity Conditions 7.2.2 Calculation of the Shear Failure Load 8. MODEL OF SCARF REPAIR 8.1 Interlayer Shear Strain 8.1.1 Boundary and Continuity Conditions

iii

2 9.8.2 Calculation of the Interlayer Shear Strain 8. 11.1. Failure Load 55 55 57 62 63 RESULTS CONCLUDING REMARKS REFERENCES APPENDICES AMoisture Loss of Specimen During Test BFailure Loads of Specimens Exposed to Humid Air CTest Results DFailure Loads of Specimens Under Moisture and Temperature iv . 10.

LIST OF FIGURES Figure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Illustrations of the Scarf and Lap Repair Techniques The Base Laminates Prior to Repair Illustration of the Repair Techniques Used in This Study The Temperature Cycle Used for Curing the Base Laminate and in Making the Repair Geometries of the Test Specimens Application of the Adhesive Layer When Repairing With Prepreg Test Specimens Cut Out of the Repaired Base Laminate Specimen Conditions and Test Temperatures Failure Loads of Different Base Laminates Repaired With Different Repair Materials (Scarf Repair at 70°F) Failure Loads of Different Base Laminates Repaired With Different Repair Materials (Scarf Repair at 180°F) Failure Loads of Different Base Laminates Repaired With Different Repair Materials (Stepped Lap Repair at 70°F) Failure Loads of Different Base Laminates Repaired With Different Repair Materials (Stepped Lap Repair at 180°F) Page 1 2 2 4 5 5 6 6 12 12 13 14 Typical Failure Modes of Specimens Repaired by Scarf and Stepped Lap Techniques 16 Definition of the Geometric Factors Investigated The Variation of Failure Load With Scarf Angle The Effect of the Number of External Plies on the Failure Load in Scarf Repair The Effect of Lap Length on the Failure Load in Stepped Lap Repair The Effect of Lap Length on the Failure Load in Uniform Lap Repair Typical Failure Modes of Specimens Repaired by Stepped Lap and Uniform Lap Techniques With Various Lap Lengths 17 19 19 20 20 21 v .

y) and Off-Axis (1. 2) Coordinate Systems Loads on a Section dx in Length of the Laminate Repaired by the Double-Sided Uniform Lap Technique Deformations of the Repair Patch. the Interlayer.20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 The Effect of the Number of Repair Plies on the Failure Load Comparisons of the Failure Loads of Specimens Repaired by the Stepped Lap and Uniform Lap Repair Techniques Failure Loads of Repaired Specimens When the Base Laminates Were Moisturized Prior to Repair The Effects of Specimen Moisture Content and Test Temperature on the Failure Load The Failure Loads of Specimens Sanded With Different Grit Diamond Sanders Illustration of the Test Setup Used in Studying the Effects of Cure Cycle on the Repair The Cure Time Required to Reach Full Cure at Different Cure Temperatures A Typical Output of the Microdielectrometer Failure Loads of Specimens Repaired at Different Cure Temperatures Models of the Uniform Lap and the Scarf Repairs The Interlayer Between the Laminate and the Repair Patch When the Repair is Performed With Wet Lay-Up (Lower Left) or Prepreg (Lower Right) Illustration of the Shear Stress-Shear Strain Relationship of an Elastic-Perfectly Plastic Interlayer Double-Sided Uniform Lap Repair Model Double-Sided Uniform Lap Repair Treated in the Model Illustration of the Types of Failure Which May Occur in a Laminate Repaired by the Double-Sided Uniform Lap Technique The On-Axis (x. and the Laminate Possible Regions of the Interlayer Under an Applied Load P The Boundary Conditions for the In-Plane Loads in the Laminate and in the Repair Patch and the Continuity Conditions for the Shear Strain in the Interlayer 21 22 24 25 26 27 27 28 29 31 31 32 32 33 33 34 36 37 38 40 vi .

and the Laminate in the k-th Segment of a Laminate Repaired by the Scarf Technique The Compliance Matrices in the Laminate and the Repair Patch Possible Regions of the Interlayer Under an Applied Load P Model of the Boundary at the Outer Edge (x = xK) of the Repair Patch The Boundary Conditions for the In-Plane Loads in the Laminate and in the Repair Patch The Continuity Conditions at the Edge of the k-th Segment The Continuity Conditions at the Interlayer Between the Elastic and Perfectly Plastic Regions The Equations and Boundary Conditions for Calculating the Shear Strain in the Interlayer When the Entire Interlayer Behaves in a Linearly Elastic Manner The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 End The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = xK End 41 41 42 42 44 44 45 46 47 48 48 51 51 52 52 53 53 54 44 45 46 47 48 49 50 51 52 53 54 55 56 57 vii .40 41 42 43 The Equations and Boundary Conditions for Calculating the Shear Strain in the Interlayer When the Entire Interlayer Behaves in a Linearly Elastic Manner The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 End The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = dl End The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 and x = dl Ends Scarf Repairs Treated in the Model The Types of Failure Which May Occur in a Laminate Repaired by the Scarf Technique The Scarf Repair Treated in the Model Loads on a Section dx in Length of the k-th Segment of the Laminate Repaired by the Scarf Technique Deformations of the Repair Patch. the Interlayer.

58 The Equations and Boundary and Continuity Conditions for Calculating the Shear Strain in the Interlayer When the Interlayer is Perfectly Plastic Near the x = 0 and x = xK Ends Comparisons of the Calculated (Model) and the Measured (Data) Failure Loads as a Function of Lap Length for 8-Ply Laminate Comparison of the Calculated (Model) and the Measured (Data) Failure Loads as a Function of Lap Length for 32-Ply Laminate Comparisons of the Calculated (Model) and the Measured (Data) Failure Loads as a Function of Scarf Angle Comparison of the Calculated (Model) and the Measured (Data) Failure Loads as a Function of Extra Plies 54 60 61 61 62 59 60 61 62 LIST OF TABLES Table 1 2 3 4 5 6 7 8 9 10 Page The Base Materials. and Cure Cycles Used for Fabricating the Base Laminate and for Making the Repair Average Tensile Properties of Specimens Made of Materials Used for Fabricating the Base Laminates Average Tensile Properties of Specimens Made of Materials Used for Repair Normalized (63% Fiber Volume) Tensile Properties of Specimens Made of Materials Used for Fabricating the Base Laminates 3 7 8 9 Normalized (63% Fiber Volume) Tensile Properties of Specimens Made of Materials Used for Repair 10 The Test Matrix Used for Evaluating the Effectiveness of Different Repair Materials The Failure Loads of Specimens Repaired by the Scarf Technique The Failure Loads of Specimens Repaired by the Stepped Lap Technique Test Matrix Used for Evaluating the Effects of Different Geometries of the Scarf and Lap Repairs The Test Matrix Used for Evaluating the Effects of the Moisture Content of the Base Laminate Prior to Repair 11 13 14 17 23 viii . Repair Materials.

11 12 13 14 The Test Matrix Used for Evaluating the Effects of Specimen Moisture Content and Test Temperature Tests With Different Preparation of the Repair Surface The Input Parameters Required by RepairL and the Numerical Values Used in the Present Calculations for 3k70 Plain Weave Fabric Impregnated With 9396 Resin The Input Parameters Required by RepairS and the Numerical Values Used in the Present Calculations for 3k70 Plain Weave Fabric Impregnated With 9396 Resin 24 26 58 59 ix/x .

• The type of wet lay-up repair material used in the repair or the type of parent laminate material on which the repair is made does not affect the quality (failure load) of the repaired part. In the present tests the highest failure load occurred at a scarf angle of about 1 degree. there is a limiting lap length beyond which the failure load does not increase. the moisture content of the repaired area. On the basis of the xi . Data were generated with the parameters having been varied. it is becoming important to develop repair techniques for structures made of such composites. Repair should be cured at the highest permissible temperature so as to achieve the shortest cure time. models were developed for calculating the failure loads of composite laminates repaired by scarf and uniform lap techniques. and the processing conditions used during repair affect the strength of the repaired part. or the stepped lap technique under tensile loading. The models take into account anisotropy of each ply in the laminate and in the repair ply and nonelastic behavior of the adhesive or resin interlayer between the laminate and the repair patch. if the laminate moisture content is high (above 1. i. Preparing the repair surfaces by sanding them with diamond sander ranging from 60 to 400 grit number does not significantly affect failure load. The failure loads of repaired specimens are reduced under hot/wet conditions. Specifically.e. In addition. • • • • • • In conjunction with experimental work.EXECUTIVE SUMMARY As fiber reinforced organic matrix composites are gaining wide acceptance. the preparation of the surface prior to repair. the objective was to obtain a systematic set of data which indicates how the type of repair material. the repair area does not need to be dried prior to repair. For a lap repair. The present work studied the repair of composite laminates in a manner which shed light on significant factors influencing the effectiveness of a repair. However. For a scarf repair..1 percent). the geometry of the repair. there is a gradual decrease in the failure load with increasing scarf angle. analytic techniques were developed which can be used to estimate the strengths of composite laminates after they have been repaired. Tests were conducted measuring the failure loads of laminates repaired either by the scarf. the uniform lap. the repair area needs to be dried completely before repair. when both the moisture content and the test temperature are high. The aforementioned parameters were varied over wide ranges and provided systematic sets of data with the following findings. If the parent laminate moisture content is low and the prior moisture history of the part is known.

xii .models. the second for calculating the failure loads of laminates repaired by the scarf technique. two computer codes were written for generating numerical values of the failure loads. Failure loads calculated by the models and the corresponding computer codes were compared with data and good agreement was found between the results of the analysis and the tests. The first code is for calculating the failure loads of laminates repaired by the uniform lap technique.

and these are referred to as scarf and lap (see figure 1). and Society of Automotive Engineers (SAE) [24]. In the scarf technique the repair material is inserted into the laminate in place of the material removed due to the damage. The major goal of the CACRC is to develop standard repair procedures for composite structures. several parameters relevant to the effectiveness of repair need to be thoroughly investigated. the geometry of the repair. The damage is then repaired by removing some of the material around the damaged area and by applying a repair patch. It would be advantageous to develop repair techniques which can be employed in a wide range of applications. Most of these techniques were developed for repairing specific parts with specific materials in the field [1-23]. In the lap technique the repair material is applied either on one or on both sides of the laminate over the damaged area. Owing to the importance of the problem. The specific problems investigated are described in next section. THE PROBLEM. the Commercial Aircraft Composite Repair Committee (CACRC) was formed under the aegis of Air Transport Association (ATA). With this objective in mind.1. To develop the procedures. In addition. and the processing conditions used during repair affect the strength of the repaired part. analytic techniques are presented which can be used to estimate the strengths of composite laminates after they have been repaired. as illustrated in figure 1. International Air Transport Association (IATA). ILLUSTRATIONS OF THE SCARF AND LAP REPAIR TECHNIQUES There are two main repair techniques. the moisture content of the repaired area. and can not readily be generalized to different repair applications. several techniques have been proposed for repairing composite laminates. the objective of the present work was to study the repair of composite laminates to shed light on significant factors influencing the effectiveness of a repair. Specifically. the preparation of the surface prior to repair. 2. it is becoming important to develop repair techniques for structures made of such composites. Therefore. INTRODUCTION. the objective was to obtain a systematic set of data which indicates how the type of repair material. Damaged laminate Scarf Double sided lap Single sided lap FIGURE 1. When composite laminates are damaged the damage frequently occurs only in a certain region of the laminate. 1 . As fiber reinforced organic matrix composites are gaining wide acceptance.

One is the double-sided uniform lap.5 in 4 in 0. to longitudinal tensile loads. where the lengths of the repair layers increase. cut out of the base laminate. Bottom: laminates used for lap repair. The effectiveness of repair was evaluated by subjecting 1-in-wide test specimens. the shortest layer being next to the surface and the longest on the outside.125 in Lay-up: [0]8 6 in FIGURE 2. 2 . The other one is the double-sided stepped lap.5 in Lay-up: [(0/45)2]s 8 in • Lap 4 in 0. • Scarf 3. The effectiveness of the repair was taken to be the tensile load at which the specimen fails. THE BASE LAMINATES PRIOR TO REPAIR (Top: laminates used for scarf repair.In this investigation.) • Scarf Repair material Base laminate • Stepped lap Repair material Base laminate • Uniform lap Repair material Base laminate FIGURE 3. where the lengths of the layers in the lap are the same. see figure 2) either by the scarf or by the lap technique (see figure 3). ILLUSTRATION OF THE REPAIR TECHNIQUES USED IN THIS STUDY For lap repairs two types were considered. a repair was simulated by joining two laminates (referred to as the base laminates.

3.7 * Dexter Hysol EA 9628 is used as a film adhesive. THE BASE MATERIALS. Sandwich test results under compression loads and fracture mechanics tests characterizing repair materials were reported previously in [25] and [26]. The base laminates were 8 in by 4 in for scarf repair and 6 in by 4 in for lap repair (see figure 2). EXPERIMENTAL. analytic techniques are presented for estimating the failure loads of composite laminates repaired by scarf and by double-sided uniform lap techniques. 4. 5. EXPERIMENTAL RESULTS. 2. This report covers the experimental and analytical research performed for repairs evaluated under tensile loading.e. 1. Ciba Geigy R922. TABLE 1. Plain weave carbon fiber fabric prepreg plies (with either Hexcel F593. PART I.7 Repair Materials (Prepreg)* Fabric Resin 3k70 plain weave carbon fiber fabric Ciba Geigy M20 250 70 30 120 40 14. AND CURE CYCLES USED FOR FABRICATING THE BASE LAMINATE AND FOR MAKING THE REPAIR Cure Temperature (°F) Final Processing Temperature (°F) ∆t1 (min) ∆t2 (min) ∆t3 (min) Pressure (Psi) Base Material Fabric Resin 3k70 plain weave carbon fiber fabric Hexcel F593 Ciba Geigy R922 Ciba Geigy R6376 350 140 120 120 90 45 75 75 Repair Materials (Wet lay-up) Fabric Resin 3k70 plain weave carbon fiber fabric Dexter Hysol EA9396/C2 Dexter Hysol EA9390 Ciba Geigy Epocast 52-A/B 200 70 20 60 220 120 30 14. REPAIR MATERIALS. or Ciba Geigy R6376 resin system) were used in the construction of the base laminates (table 1). The 0° 3 . the material used for the repair) Geometry of the repair Moisture content of the specimen and the test temperature Preparation of the surface on which the repair is performed Processing (cure) condition applied during the repair In addition..Five series of tests were performed to evaluate the effects of the following parameters on the failure load of the repaired laminate. Type of repair material (i. The lay-up of the base laminate was [(0/45)2]s for scarf repair and [0]8 for lap repair. 3.

and cured in an autoclave by the manufacturer’s cure cycle (figure 4 and table 1). cure temp. The thicknesses of the cured base laminates ranged from 0. For the lap 4 . two 0° external plies were placed on the surface of the base laminate covering the repaired area (see figure 5). Each base laminate was laid up by hand. either in wet lay-up or in prepreg form. In addition. The repair was applied with carbon fiber fabric plies. Same cure cycle was used for the scarf and patch repairs. and by observing whether or not the entire surface was wetted [27].. The autoclave pressure (during fabrication of base laminates) or the vacuum (during repair) were applied at time t = 0 and were kept constant during the cure. Forty-five refers to plies with +45 and -45 degree fiber orientations. vacuum bagged. cut into the desired shape. table 1). on which a scarf or a lap repair was to be applied. For a wet lay-up repair.e. the surfaces of the base laminate. The 0 degree direction is along the longitudinal axis of the specimen. The two base laminates were placed side by side (see figure 2) with a 0. For the scarf repair. Zero refers to plies with 0 and 90 degree fiber orientations. The surface was cleaned by acetone just prior to repair. figure 4. The uniformity of the sanded surface was assessed by the water-break test.refers to fabric which has fibers parallel and perpendicular to the longitudinal axis. The prepreg plies were preimpregnated with Ciba Geigy M20 resin.5-in gap between the laminates for scarf repair and a 0. Two base laminates were joined (repaired) either by the scarf or by the lap technique. The water was removed by drying the specimen in an oven. were sanded with a 120 grit diamond sander. 70 F ∆t1 ∆t2 ∆t3 Final processing temp.057 to 0. Time (min) FIGURE 4.070 in. i. The 45° designates plies in which the fibers are ±45° from the longitudinal axis. Dexter Hysol EA9390. the plies were impregnated with a two parts epoxy resin (Dexter Hysol EA9396/C2. (F) Max. Temp. as follows. Each ply (wet lay-up or prepreg) was placed either into the base laminate (scarf repair) or on the surface of the base laminate (lap repair) (see figure 5). by spreading a thin layer of water on the surface. or Ciba Geigy Epocast 52-A/B. THE TEMPERATURE CYCLE USED FOR CURING THE BASE LAMINATE AND IN MAKING THE REPAIR After curing. the orientation of each repair ply in the repaired zone was the same as the corresponding ply orientation of the base laminate. For the repair with prepreg plies a layer of film adhesive was placed between the base laminate and the repair material to facilitate bonding between the base laminate and the repair material (see figure 6).125-in gap for lap repair.

5 . After the repair material was applied.5 in overlap Repair material [(0/45)2 ] s Base laminate [(0/45) 2 ] s 2 in 0.5 in 16.125 in FIGURE 5. as illustrated in figure 5.125 in 0. After curing.125 in 0.5 in overlap per ply Repair material [0] 3 Base laminate [0] 8 2 in 12. 2-in-long and 0.125 in • Uniform lap Fiberglas tab 1 in Repair material [0] 3 Base laminate [0] 8 2 in 12. the repair plies (with the same 0° orientations as the base laminate) were placed on both sides of the base laminate.5 in • Stepped lap Fiberglas tab 0.125-in-thick fiberglas tabs were attached with room temperature cure Epoxy 907.• Scarf External plies[0] Fiberglas tab 2 0. the repair area was vacuum bagged and cured in an oven under atmospheric pressure (see table 1).5 in overlap per ply 0. APPLICATION OF THE ADHESIVE LAYER WHEN REPAIRING WITH PREPREG repair. GEOMETRIES OF THE TEST SPECIMENS • Scarf Prepreg plies Film adhesive • Stepped lap Prepreg plies Film adhesive • Uniform lap Prepreg plies Film adhesive FIGURE 6.

while some were moisturized before testing either by immersing the specimen in 180°F water for 14 days or by exposing the specimen to 100 percent humid air at 180°F until 1. Repaired base laminate Test specimens FIGURE 7. Thus. In this chamber the relative humidity was not regulated. a small amount (less than 0. was cut into 1-in-wide strips (see figure 7). 6 . The detailed geometries of the specimens are given in figure 5.05 in/mm.007 percent. These conditions were adopted because the same conditions are frequently used in the aircraft industry in testing adhesively bonded joints. the base laminate. appendix A) of drying occurred during the 180°F tests. The failed specimens were also inspected visually to establish the mode of failure. 70°F Water 180°F 180°F Humid air 180°F 100% RH FIGURE 8. SPECIMEN CONDITIONS AND TEST TEMPERATURES Each specimen was tested either at 70° or 180°F. The load at which the specimen failed (failure load) was recorded.1 percent moisture content (weight gain) was reached (see figure 8). Some of the specimens were tested dry (as prepared). which lasted about 10 minutes. repaired in the manner described above.To make test specimens. The temperature was maintained by a chamber surrounding the specimen. TEST SPECIMENS CUT OUT OF THE REPAIRED BASE LAMINATE Each test specimen was subjected to uniaxial tensile loads at a constant displacement rate of 0.

4 7. the tensile properties of undamaged laminates made of the materials used in constructing the base laminates and in the repair were measured.8 10.2 10. To this end. straight-sided tensile specimens were cut out of the plates. 7 . TABLE 2.4 10.1% moisture content was reached (moist-humid air).9 5.4 6.8 6.0 7.1 8. and 1-inch-wide. AVERAGE TENSILE PROPERTIES OF SPECIMENS MADE OF MATERIALS USED FOR FABRICATING THE BASE LAMINATES Specimen Condition/ Test Temperature** Dry/70°F Moist-H2O/180°F Moist-humid air /180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Failure Load (lbf) 6600 6110 6180 5890 4580 5560 6550 6720 6330 4780 4100 4190 4850 4010 4490 5850 4730 5070 Strength (ksi) 101 93 95 105 96 102 107 107 101 72 59 64 90 69 83 90 76 80 Stiffness (Msi) 8. 12.2 9.05 in/mm.3 Lay-Up Material* F593 [0]8 R922 R6376 F593 [(0/45)2]S R922 R6376 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1).4 7. Fiberglas tabs were mounted on the plates. or exposed to 180°F air at 100% relative humidity until 1.8 5. immersed in 180°F water for 14 days (moist-H2O).by 4-in ([(0/45)2]s and [0]8) plates were prepared in the same manner as the base laminates (see previous section).8 10.9 7. The values are the average values of three tests. The specimens were subjected to uniaxial tensile loads at a constant displacement rate of 0.5 8.4 9.For reference purposes. These tests were performed under the same conditions as were used for the repaired test specimens (see figure 8). ** Specimens were either dry.5 6. The measured tensile strengths and longitudinal stiffnesses are given in tables 2 and 3. Test temperature was either 70° or 180°F (figure 8). The load versus strain and the failure load were recorded. The strain was measured by an extensometer.

9 Lay-Up Material* EA9396 EA9390 [0]8 Epocast M20 EA9396 EA9390 [(0/45)2]s Epocast M20 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1). These values.2 5.0 9.7 7.7 7. though of not direct concern in this study. The data in tables 4 and 5 were normalized to apply to a 0.7 9.011-in-thick ply with a nominal fiber volume content of 63 percent.2 5. 8 . AVERAGE TENSILE PROPERTIES OF SPECIMENS MADE OF MATERIALS USED FOR REPAIR Specimen Condition/ Test Temperature** Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Failure Load (lbf) 5608 5581 6458 6039 5934 5635 7253 7252 4623 3722 4150 4049 4428 4277 6035 5101 Strength (ksi) 75 68 91 83 73 73 118 106 58 47 58 56 51 58 95 77 Stiffness (Msi) 7.2 8.0 5. Test temperature was either 70° or 180°F (figure 8).7 5. are listed in tables 4 and 5.9 8. or immersed in 180°F water for 14 days (moist-H2O). the properties of unidirectional plies made from the base and repair materials were deduced.6 6.From the data in tables 2 and 3.5 5.0 5. ** Specimens were either dry. TABLE 3.5 6.2 7.

9 13.0 17.6 16.5 12. or exposed to 180°F air at 100% relative humidity until 1.0 10.TABLE 4.1% moisture content was reached (moist-humid air).8 11. immersed in 180°F water for 14 days (moist-H2O).0 Lay-Up Material* F593 [0]8 R922 R6376 F593 [(0/45)2]S R922 R6376 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1). ** Specimens were either dry.1 9.1 12.2 13.8 12.5 14.4 10.3 9.8 7. NORMALIZED (63% FIBER VOLUME) TENSILE PROPERTIES OF SPECIMENS MADE OF MATERIALS USED FOR FABRICATING THE BASE LAMINATES Specimen Condition/ Test Temperature** Dry/70°F Moist-H2O/180°F Moist-humid air /180°F Dry/70°F Moist-H2O/180°F Moist-humid Air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Dry/70°F Moist-H2O/180°F Moist-humid air/180°F Strength (ksi) 137 118 125 175 156 166 157 150 138 96 74 84 117 108 110 126 104 110 Stiffness (Msi) 11. Test temperature was either 70° or 180°F (figure 8).5 7. 9 .3 7.9 17.

6 11. temperature was either 70° or 180°F (figure 8).2 5.7 12. The actual values of the measured failure loads are included in appendix B.7 9.3 6.8 Lay-Up Material* EA9396 EA9390 [0]8 Epocast M20 EA9396 EA9390 [(0/45)2]s Epocast M20 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1). In the following.7 7.TABLE 5. Each data shown in this section is the average of three to six tests. TEST RESULTS. The materials used for fabricating the base laminates and the repair were given in table 1. The materials were chosen for this study because they are currently used or being considered for use in several different commercial aircraft. NORMALIZED (63% FIBER VOLUME) TENSILE PROPERTIES OF SPECIMENS MADE OF MATERIALS USED FOR REPAIR Specimen Condition/ Test Temperature** Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Dry/70°F Moist-H2O/180°F Strength (ksi) 88 75 108 98 76 80 162 139 64 51 71 67 61 69 103 101 Stiffness (Msi) 8. for simplicity. The base laminates are referred to as 10 . The failure loads of repaired specimens under different conditions are presented in this section.2 7.5 8.5 5.7 6.5 6. Test 4.2 9.4 7.4 10. both the base laminates and the repair materials are identified by their resin system. The spread in the data is indicated by bars. The failure loads are normalized with respect to failure loads appropriate to the given test. or immersed in 180°F water for 14 days (moist-H2O).3 10. ** Specimens were either dry.

while the repair materials as 9396. Repairs were made both by the scarf and the stepped lap techniques. The results of each five test series are given below. TABLE 6. ** Specimens were either dry. 9390. Test temperature was either 70° or 180°F (figure 8). THE TEST MATRIX USED FOR EVALUATING THE EFFECTIVENESS OF DIFFERENT REPAIR MATERIALS Base Laminate* F593 Repair Material* 9396 9390 Epocast M20 9396 9390 Epocast M20 9396 9390 Epocast M20 Specimen Condition/Test Temperature** Dry/70°F x x x x x x x x x x x x Moist-H2O/180°F x x x x x x x x x x x x Moist-Humid Air/180°F x x x x x x x x x x x x R922 R6376 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1). Note again. or exposed to 180°F air at 100% relative humidity until 1.1% moisture content was reached (moist-humid air). 11 . The failure loads of different base laminates repaired with different repair materials (normalized with respect to the failure load of undamaged laminates) are shown in figures 9 and 10 and table 7 for scarf repair and in figures 11 and 12 and table 8 for stepped lap repair. and R6376. and M20 (see table 1). 9390.F593. four types of material were used to repair the base laminates which were made of three different materials. Five series of tests were performed. 4. that the repair materials using 9396.1 TYPE OF REPAIR MATERIAL. The complete test matrix used in this test series is given in table 6. R922. Each test was performed with specimens repaired by the scarf and the stepped lap technique. To evaluate the effectiveness of different repair materials. Epocast. while the repair material designated as M20 was in prepreg form. immersed in 180°F water for 14 days (moist-H2O). and Epocast resins were prepared by wet lay-up. The results in these figures are for dry specimens and for specimens moisturized by immersion in water for 14 days.

8 0.0 9396 9390 Epocast M20 R6376 Specimen: Moist-H2O Test: 180 F Wet lay-up Prepreg Repair Material FIGURE 10.0 0.2 0.4 1.6 0.2 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 1.0 0.2 1. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (SCARF REPAIR AT 180°F) 12 .4 0.4 1.2 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 1.6 0.0 1.4 0.0 0.2 0.4 1.0 9396 9390 Epocast M20 R6376 Specimen: Dry Test: 70F Wet lay-up Prepreg Repair Material FIGURE 9.2 0.6 0.4 1. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (SCARF REPAIR 70°F) 1.1.8 0.8 0.2 1.6 0.0 1.4 0.0 0.4 0.2 0.8 0.

2 0.0 0.TABLE 7.4 0.4 1.6 0.6 0.2 0.0 9396 9390 Epocast M20 R6376 Specimen: Dry Test: 70F Wet lay-up Prepreg Repair Material FIGURE 11.0 1.4 1.8 0.1% moisture content was reached (moist-humid air). FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR AT 70°F) 13 . immersed in 180°F water for 14 days (moist-H2O).2 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 1.0 0. or exposed to 180°F air at 100% relative humidity until 1. Test temperature was either 70° or 180°F (figure 8).4 0. 1. **Specimens were either dry. THE FAILURE LOADS OF SPECIMENS REPAIRED BY THE SCARF TECHNIQUE Base Laminate* 9396 Repair Material* F593 R922 R6376 F593 R922 R6376 F593 R922 R6376 F593 R922 R6376 Specimen Condition/Test Temperature** Dry/70°F 4526 4282 4658 3695 4168 4222 4175 4064 4281 4386 4745 5558 Moist-H2O/180°F 3693 3468 3621 3705 3746 3735 3408 3279 3514 2585 3123 2862 Moist-Humid Air/180°F 3558 3400 3294 3404 3293 3124 4178 3388 3580 3176 3735 2855 9390 Epocast M20 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1).2 1.8 0.

or exposed to 180°F air at 100% relative humidity until 1.4 0. THE FAILURE LOADS OF SPECIMENS REPAIRED BY THE STEPPED LAP TECHNIQUE Base Laminate* 9396 Repair Material* F593 R922 R6376 F593 R922 R6376 F593 R922 R6376 F593 R922 R6376 Specimen Condition/Test Temperature** Dry/70°F 3360 3547 3302 3638 3238 3195 3420 3298 3223 4516 4646 4249 Moist-H2O/180°F 2958 2533 2520 2945 2612 2497 2674 2642 2307 2148 2097 2033 Moist-Humid Air/180°F 2624 2311 2096 2351 2821 2003 2570 2777 2284 3150 3995 3189 9390 Epocast M20 * The material was plain weave carbon fiber fabric impregnated with the type of resin indicated in this column (see table 1).2 0. or immersed in 180°F water for 14 days (moist-H2O).6 0.0 0. ** Specimens were either dry.8 0.1.0 0. 14 .2 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 1.2 0.4 1. Test temperature was either 70° or 180°F (figure 8).6 0.2 1.0 9396 9390 Epocast M20 R6376 Specimen: Moist-H2O Test: 180 F Wet lay-up Prepreg Repair Material FIGURE 12.0 1. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR AT 180°F) TABLE 8.1% moisture content was reached (moist-humid air).4 1.8 0.4 0.

) This indicates that the failure load (in shear) of the adhesive film used in the prepreg repair was higher than either the failure load (in shear) of the wet lay-up interface or the failure load (in tension) of the base laminate. For this reason.1 percent moisture content achieved by exposure to 100 percent humid air at 180°F. − • The failure loads were higher with the prepreg repair (scarf and stepped lap) than with the wet lay-up repair for specimens dry and tested at 70°F. the failure was due to tensile failure of the base laminate (see figure 13). specimens repaired by prepreg had lower failure loads than specimens repaired by wet lay-up. Both the repaired specimens and the undamaged base laminates were immersed in 180°F water for 14 days and were tested at 180°F. failure occurred along the base laminate-repair interface (see figure 13). 15 . Under these conditions the shear strength of the adhesive film used in the prepreg repair was lower than the shear strength of the wet lay-up interface. − Failure of these specimens generally occurred due to shear along the base laminaterepair interface (see figure 13). both with wet lay-up and with prepreg repair. The failure loads of specimens repaired by wet lay-up were nearly the same.Data were also obtained with specimens with 1. Specimen geometry is given in figure 5. but are included in appendix C. Failure loads of undamaged specimens are given in table 2. Therefore. Thus specimens repaired with prepreg had higher failure loads than specimens repaired by wet lay-up. failure occurred along the base laminaterepair interface. resulting in the prepreg repair having higher failure loads than the wet lay-up repair. resulting in similar failure loads. The shear strength of the adhesive film used in the prepreg repair was higher than the shear strength of the wet lay-up interface. The failure loads of specimens moisturized by water and by humid air showed similar trends. The following trends are indicated by the data: • For the wet lay-up repair (scarf and stepped lap) the type of repair material did not significantly affect the failure load. irrespective on which base material the repair was made. the failure loads were lower than with wet lay-up when the specimens were moisturized and tested at 180°F. Apparently. the data for specimens moisturized by humid air are not given here. − For moisturized specimens tested at 180°F. − • With the prepreg repair (scarf and stepped lap). (With the prepreg repair. failure always occurred along the base laminate-repair interface (see figure 13). Both the repaired specimens and the undamaged base laminates were tested dry at 70°F. irrespective of the base material that was repaired (tables 7 and 8). − For the scarf repair with wet lay-up. For stepped lap repair. the interfacial shear strengths were similar for the three wet lay-up repair materials.

16 . and Epocast). the tests presented in the following sections were conducted only with the 9396 repair material. B-1. These effects are discussed in section 4.) The results shown in figures 10. Since the results were similar and consistent for the three wet lay-up repair materials (9396. 12. TYPICAL FAILURE MODES OF SPECIMENS REPAIRED BY SCARF AND STEPPED LAP TECHNIQUES (Solid lines indicate failure. the length of the repair plies (lap length.3. and B-2 also indicate the effects of specimen moisture content and test temperature on the failure load. 4. dl). 9390. The complete test matrix is shown in table 9. and the number of repair plies per side nlR . right: moisturized specimens immersed in 180°F water for 14 days and tested at 180°F.Specimen: Dry Test: 70F • Scarf Specimen: Moist-H2O Test: 180F Wet lay-up Wet lay-up Prepreg Prepreg • Stepped lap Wet lay-up Wet lay-up Prepreg Prepreg FIGURE 13. Left: dry specimens tested at 70°F. The effects of the following geometric factors on the failure load were evaluated (see figure 14): • • For the scarf repair: the scarf angle β and the number of external plies ne For the lap repair: the shape of the repair plies (stepped or uniform). The effects of these factors on the failure load were evaluated by testing dry as well as moisturized specimens at 70° and 180°F.2 GEOMETRY EFFECTS.

3. 1.07.72. dl • Uniform lap Repair plies (Number per side: nR ) l Lap length. 1. Test temperature was either 70° or 180°F (figure 8). 1. 3 1. 17 . 4 nlR nlL 0. 1. 2. 2 Moist-H2O/180°F 0. 3 8. Repair Material: 9396) Specimen Condition/ Test Temperature Repair Type Scarf Stepped lap Parameter Scarf angle. β (degree) Number of external plies.72. Lap length. 1. 2.25. 3 8. 2. 32 1. 2. 1. 2 0.5. 3 0. 2. 2. nlR Dry/70°F 0. 2. 1. 1.15 0.5.5. dl (in) Uniform lap Number of plies in the base laminate. DEFINITION OF THE GEOMETRIC FACTORS INVESTIGATED TABLE 9. 2. dl Base laminate (Number per side: nLl ) FIGURE 14. 4 * Specimens were either dry or immersed in 180°F water for 14 days (Moist-H2O). 0. 2. Number of repair plies per side. 3 1. 2. 32 1.5. 1.• Scarf External plies (Number: ne ) tan(β)= H ds ds H • Stepped lap dl nR l Repair plies (Number per side: nR ) l Lap length. 0.15 0.5. TEST MATRIX USED FOR EVALUATING THE EFFECTS OF DIFFERENT GEOMETRIES OF THE SCARF AND LAP REPAIRS (Base Laminate: F593. 3 0.25. ne Lap length.07. dl (in) number of repair plies per side.5. 3. 1.

3dlimit) and intermediate lap length (0. failure occurred mostly along the interface. and John. • For the lap repair − The failure load increased with increasing lap length but only up to a certain “limiting lap length” dlimit (see figures 17 to 18). with some of the repair plies also failing by tension at the intermediate lap length. and this corresponds to an intermediate lap length. These conclusions seemed to hold even when the total number of repair plies (2 nlR ) was the same as the number of plies in the base laminate ( nlL ). and Matthews [31].3dlimit ≤ d1 ≤ dlimit). that for the specimens in section 4.1. Kinloch. The addition of external plies on the surface of the specimen resulted in an increased failure loads (see figure 16). At and beyond the limiting lap length (dl. there is no benefit to be gained by making the lap length longer than the limiting lap length. This result was valid even when the total number of repair plies (2 nlL ) was equal to the number of plies in the base laminate ( nlR ) (see figure 20 middle). the shape of the repair plies did not markedly affect the failure load (see figure 21).The results are presented in figures 15 and 16 for the scarf repair and in figures 17 to 21 for the lap repair. The mode of failure changed with lap length (see figure 19). − 18 .5 in. Thus. Beyond the limiting lap length an increase in lap length did not affect the failure load. − When the lap length was greater than limiting lap length (dl ≥ dlimit). and Lee [28]. The following general observations can be made on the basis of the data given in these figures. These results are in agreement with those found by Myhre and Beck [29]. ≥ dlimit) failure was due to tensile failure of the repair plies. This limiting lap length was generally between 2 and 3 inches for the specimens used in this investigation.) − The failure load increased linearly with the number of plies used in the repair (see figure 20). • For the scarf repair − − The failure load decreased with increasing scarf angle (see figure 15). provided the following two conditions were simultaneously satisfied: (a) the number of repair plies were the same and (b) the lap length of the uniform lap repair was the same as the length of the outside lap of the stepped repair. (Note. Lee. Similar results were reported by Chan and Sun [30]. the lap length was 1. The failure loads were nearly the same for specimens repaired by stepped and by uniform repair. Similar observations were made by Kim. At short (dl ≤ 0.

THE EFFECT OF THE NUMBER OF EXTERNAL PLIES ON THE FAILURE LOAD IN SCARF REPAIR 19 .2.5 Failure Load ( Repaired / Undamaged ) 1.5 Specimen: Dry Test: 70F 0. THE VARIATION OF FAILURE LOAD WITH SCARF ANGLE 2.5 Scarf Angle.0 0.0 1.0 Specimen: Moist-H2O Test: 180F 1.0 Laminate: F593 Repair: 9396 1.0 ds H 0.5 tan(β)= H ds 1.0 No External Ply 1 Ply 2 Plies Number of External Plies.5 1 1.0 0.5 Failure Load ( Repaired / Undamaged ) Data Fit to data 1.5 1.5 0. ne FIGURE 16.5 Specimen: Moist-H2O Test: 180 F 0.5 2 2.0 Laminate: F593 Repair: 9396 1.0 Data Fit to data 0.5 Specimen: Dry Test: 70F 0.0 0. β (degree) FIGURE 15.

0 1.0 0. dl (in) FIGURE 18.0 1.5 at d ( at lap lap length=3 in ) length d 1.0 0.0 2.0 0. dl (in) Lap Length.5 1 repair ply Fit to data l [0] 8 [0] 32 Fit to data 0.5 1.0 1.5 Laminate: F593 [0] 8 Repair: 9396 dl 1.0 4.5 20 3 repair plies 2.0 Failure Load 1.0 3.0 2 repair plies ( 0.0 0.0 1.5 1 repair ply 0.0 1.0 3.5 0.0 at lap length dl at lap length dl=3 in 1.0 0.0 Data l 0. dl (in) Lap Length.0 0.0 4 repair plies 0.0 1.0 3 repair plies 2 repair plies 1.0 2. dl (in) Lap Length.0 1.0 Laminate: F593 Repair: 9396 Specimen: Dry Test: 70F Specimen: Moist-H2O Test: 180F 2.0 3.0 4.0 0.5 0.0 0.5 dl ) 1. THE EFFECT OF LAP LENGTH ON THE FAILURE LOAD IN STEPPED LAP REPAIR .0 0.0 Lap Length.Specimen: Dry Test: 70F Specimen: Moist-H2O Test: 180F 2.0 2.5 0.0 Failure Load 1.5 0. THE EFFECT OF LAP LENGTH ON THE FAILURE LOAD IN UNIFORM LAP REPAIR FIGURE 17.0 3.

0 0.Specimen: Dry Test: 70F Laminate: F593 Repair: 9396 dl Specimen: Moist-H2O Test: 180 F dl=3 in 8 ply laminate 8 ply laminate Failure load 1. nR l FIGURE 19.6 0.8 0.4 0.8 0.8 0.6 0.4 0.4 0.2 1.0 0 1 • Stepped lap dl 32 ply laminate 2 3 4 0 1 32 ply laminate 2 3 4 5 Number of Repair Plies.2 0. dl 1.0 0.3 dlimit dlimit Lap length.0 8 ply laminate 8 ply laminate • Uniform lap Failure Load ( Repaired / Undamaged) 21 1. THE EFFECT OF THE NUMBER OF REPAIR PLIES ON THE FAILURE LOAD .0 0. TYPICAL FAILURE MODES OF SPECIMENS REPAIRED BY STEPPED LAP AND UNIFORM LAP TECHNIQUES WITH VARIOUS LAP LENGTHS (Specimen geometries are given in figures 5 and 14) FIGURE 20.0 Data Fit to data dl ~0.2 0.2 0.6 0.

nR l FIGURE 21. COMPARISONS OF THE FAILURE LOADS OF SPECIMENS REPAIRED BY THE STEPPED LAP AND UNIFORM LAP REPAIR TECHNIQUES 4.5 Specimen: Moist-H2 O Test: 180F 0.0 0.0 1.5 Specimen: Dry Test: 70F 0. The base laminates were then removed from the environmental chamber.5 Failure Load 1. The base laminates were kept in the environmental chamber until the desired moisture content was reached. The second type of specimen consisted of the base laminate whose moisture content was established by moisturization followed by drying. and the repaired specimens were moisturized before being tested. 4.0 0.5 with uniform lap ( with stepped lap repair ) repair 1. In one series of tests.1 Base Laminates Premoisturized.0 dl Laminate: F593 Repair: 9396 Lap length.0 0 1 2 3 4 Number of Repair Plies.3. and tested at 70°F. There were two types of premoisturized specimens. The first type of specimen consisted of the base laminate moisturized prior to repair by exposing the unrepaired base laminate to 100 percent humid air at 180°F.2.3 EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE.1 percent or 1. dl =3 in dl Data Fit to data 1. In the second series. repaired. When the desired moisture content was 22 . Two series of tests were performed to evaluate the effects of specimen moisture content and test temperature on the failure loads of repaired specimens. the base laminates were dried prior to repair.5 percent moisture content. In this case the base laminate was moisturized to either 1. the base laminates were moisturized prior to repair.

• 23 . et al. The same phenomenon was observed by Robson. The behavior was quite different when the moisture content of the base laminates exceeded about 1. Test temperature was either 70° or 180°F (figure 8).1 percent (see figure 22.0 to 1. With above procedures. [32] in their tests with scarf joints. the base laminate was removed from the environmental chamber and was dried in an oven at 180°F. The failure loads of specimens repaired with premoisturized base laminates are given in figure 22.5% moisture content. This conclusion was valid when the base laminate moisture content was established either by moisturization (solid circle) or by moisturization followed by drying (open circle).0 to 1. the base laminates had to be dried almost completely to recover the failure loads of specimens made with premoisturized (but not dried) base laminates.1 percent (see figure 22.0 1. Once the desired moisture content was reached during the drying process. provided that the moisture content did not exceed 1. the moisture content of the base laminate was established either by moisturization of the base laminate or by moisturization followed by drying. In this figure.5 180°F Dry Air/70°F** 1. The failure loads of repaired specimens made with base laminates which have first been moisturized to 1.1 0. Repair Material: 9396) Specimen Condition/Test Temperature Repair Type Scarf Stepped lap Parameter Moisture content (% weight) Moisture content (% weight) Moist-Humid Air/70°F* 0. In fact. and tested at 70°F. TABLE 10.reached.5 0. The test matrix is given in table 10.5 to 0.5 percent moisture content then dried (dotted line in figure 22 bottom) were lower than the failure loads of repaired specimens made with base laminates in which the moisture was introduced during moisturization (solid line). The following major observations can be made on the basis of the data in figure 22.0 * Specimens were exposed to 180°F air at 100% relative humidity until either 1.0 1.1 to 0. bottom).0 1. • The failure loads of repaired specimens decreased only slightly when the base laminates were moisturized prior to the repair. ** Specimens were dried at 180°F after having been moisturized to either 1.0 to 1.1% or 1.5% moisture content was reached (moist-humid air).5 to 0.1 0. the base laminate was taken out of the oven. repaired.0 to 1.1% or 1. the failure loads are normalized with respect to the failure load of the dry specimen. top). Similar observations were made by Parker [33] on adhesively bonded single lap joints. THE TEST MATRIX USED FOR EVALUATING THE EFFECTS OF THE MOISTURE CONTENT OF THE BASE LAMINATE PRIOR TO REPAIR (Base Laminate: F593.1 to 0.

8 1. base laminate being moisturized Data.50 1.75 1. it is not necessary to dry the repair area prior to repair as long as the moisture content of the laminates is low (less than 1.2 0.50 0.0 0.0 percent).25 0. Repair Material: 9396) Specimen Condition/Test Temperature Repair Type Scarf Stepped lap Dry 70°F x x Dry 180°F x x Moist-H2O 70°F x x Moist-H2O 180°F x x Moist-Humid Air/ 70°F x x Moist-Humid Air/ 180°F x x 24 . TABLE 11. THE TEST MATRIX USED FOR EVALUATING THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE (Base Laminate: F593.50 0.0 percent).8 1. First. The complete test matrix is shown in table 11. FAILURE LOADS OF REPAIRED SPECIMENS WHEN THE BASE LAMINATES WERE MOISTURIZED PRIOR TO REPAIR The above mentioned results have two significant ramifications.4 0. Each specimen was then either immersed in 180°F water for 14 days or exposed to 100 percent humid air at 180°F until 1.0 0.2 Specimen Moisture Content and Test Temperature.6 Fit to data Data.2 1.1 percent moisture content was reached.50 1. the repair area should be dried completely (not just partially) when the laminate moisture content is high (above 1.25 1. 4. base laminate being dried at 180 F Laminate: F593 Repair: 9396 Test: 70F Failure Load of Repaired Specimen Moisture Content of Base Laminate Prior to Repair (Percent) FIGURE 22.00 0. Second.3.00 0.75 0. The failure loads of the dry and the moisturized specimens were measured at 70° and 180°F.00 0.base laminate ( with premoisturizedlaminate ( with dry base 1.4 0. Specimens were repaired by scarf and stepped lap techniques using dry base laminates.25 0.00 1.75 0.25 1.

The trend in the data was similar for the specimens moisturized in water and in humid air. Therefore, here data presented is only for specimens moisturized by exposure to humid air (see figure 23). The data obtained during moisturization in water are included in appendix D. The data in figure 23 indicate the following trends. • • • The moisture content did not significantly reduce the failure load when the specimens were tested at 70°F. Testing at the elevated temperature (180°F) did not significantly reduce the failure load when the specimens were tested at dry condition. The failure load of the specimen was reduced under hot/wet conditions i.e., when the specimen was both wet and tested at 180°F. Similar observations were made by Stone [16] with single lap shear specimens, Mylire, Labor, and Aker [34] with modified single lap specimens, and John, Kinloch, and Matthews [31] with double lap joints.
1.6 1.4 Laminate: F593 Repair: 9396

)
1.2

specimen at indicated condition dry specimen tested at 70F

1.0 0.8 0.6 0.4 0.2 0.0
1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

Failure Load

(

Moist & 70F

Dry & 180F

Moist & 180F

Specimen Condition & Test Temperature

FIGURE 23. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE ON THE FAILURE LOAD 4.4 EFFECT OF SANDING. Prior to repair the repair surface of the base laminate is usually sanded. To evaluate the influence of sanding on the failure load of repaired specimens, the repair surfaces were sanded with different grit diamond sanders (see table 12). The failure loads of specimens repaired using different grit sanders are shown in figure 24. The data indicate that the grit number of the sander did not have a major effect on the failure load. For the material and type of repair used in this study, a sander with grit number about 100 seems appropriate.

25

TABLE 12. TESTS WITH DIFFERENT PREPARATION OF THE REPAIR SURFACE (Base Laminate: F593, Repair Material: 9396)
Specimen Condition/Test Temperature* Dry/70°F Moist-H2O/180°F 60, 120, 400 60, 120, 400 60, 120, 400 60, 120, 400

Repair Type Scarf Stepped lap

Parameter Grit number of diamond sander, g Grit number of diamond sander, g

* Specimens were either dry, or immersed in 180°F water for 14 days (Moist-H2O). Test temperature was either 70° or 180°F (figure 8).

FIGURE 24. THE FAILURE LOADS OF SPECIMENS SANDED WITH DIFFERENT GRIT DIAMOND SANDERS 4.5 EFFECT OF CURE CYCLE. Composite material is generally repaired at an elevated temperature to expedite the chemical reactions and to cure the resin used in the repair. In addition to the temperature, time is factor in the cure. For economical reasons, it is desirable to have a low cure temperature and a short cure time. The effects of the temperature-time history (referred to as cure cycle) on the effectiveness of the repair were studied on specimens repaired by scarf and stepped lap techniques. Each specimen was repaired by surrounding it with a heating blanket. This assembly was enclosed in a vacuum bag and was kept under vacuum during the repair (see figure 25). The specimen was heated at a constant rate of 6.5°F/min until the desired cure temperature Tc was reached (see figure 26 insert). This cure temperature was maintained until the repair material was fully cured. 26

Insulation Heating blanket Surface breather Vacuum bag Surface bleeder Aluminum plate Solid FEP parting film Vacuum Edge bleeder Solid FEP parting film Composite plate

Microdielectric sensor

Repaired part

FIGURE 25. ILLUSTRATION OF THE TEST SETUP USED IN STUDYING THE EFFECTS OF CURE CYCLE ON THE REPAIR
120 Laminate: F593 Repair : 9396 90

60

Time to Full Cure (min)

30

Data Fit to data

Specimen: Dry Test: 70F

0 2.0

90

60
Temp.

30

Tc

Time 0 125 150 175 200 225

Cure Temperature, Tc (F)

FIGURE 26. THE CURE TIME REQUIRED TO REACH FULL CURE AT DIFFERENT CURE TEMPERATURES (Specimen geometries are given in figures 5 and 14)

27

The progress of the cure and the specimen temperature were monitored by a Eumetric System II Microdielectrometer (Micromet Instruments, Inc.). This instrument provided the temperature and the ion viscosity of the repair resin, the latter being an indicator of the degree of cure. A typical output of the instrument is shown in figure 27. Each specimen was cured at a preset cure temperature Tc (150°, 175°, or 200°F) until the instrument indicated that full cure was reached.

FIGURE 27. A TYPICAL OUTPUT OF THE MICRODIELECTROMETER The time required to reach full cure is shown in figure 26. As the heating rate was constant at 6.5°F/min, the time required to reach full cure decreased with increasing cure temperature. The failure loads of specimens cured at different temperatures are shown in figure 28. The cure temperature did not seem to affect significantly the failure load for the dry specimens tested at 70°F. It is unknown what would happen if the specimens were tested at 180°F. This limited testing suggests that it is advantageous to use the highest cure temperature (permissible for the resin system used in the repair) since this results in shortest cure (and repair) time.

28

Tc =150 F ) 1. or the type of parent laminate material on which the repair is made.5 at cure temp. Tc at cure temp. 0. For a scarf repair. • • • 29 . if the laminate moisture content is high (above 1. there is a limiting lap length beyond which the failure load does not increase. does not affect the quality (failure load) of the repaired part.6 SUMMARY. • The type of wet lay-up repair material used in the repair. For a lap repair. Small variations in the scarf angle around this value did not have a significant effect on the failure load. The major findings of the tests are summarized below. FAILURE LOADS OF SPECIMENS REPAIRED AT DIFFERENT CURE TEMPERATURES (Specimen geometries are given in figures 5 and 14. the repair area needs to be dried completely before repair.5 Data Fit to data 0. Tc (F) FIGURE 28.5 Failure load ( 1.5 Tc 0.) 4. the repair area does not need to be dried prior to repair.0 125 Time 150 175 200 225 Cure Temperature. Failure loads of the specimens at cure temperature 150°F are given in appendix C.0 1. If the parent laminate moisture content is low and the prior moisture history of the part is known.0 Temp. However.0 Specimen: Dry Test: 70F 0.0 2.2.0 Laminate: F593 Repair: 9396 1. there is a gradual decrease in the failure load with increasing scarf angle.1 percent). In the present tests the highest failure load occurred at a scarf angle of about 1 degree.

Both the resin layer and the adhesive 30 . ANALYTICAL MODELS. i. the Hart-Smith models were adapted to model the scarf and double-sided lap repair by accounting for each individual layer separately. the adhesive does not behave in a linearly elastic manner. double lap [30. Hart-Smith [57-59. INTRODUCTION. 50-55]. 40. their codes are not in the public domain. It is likely that the trends observed here are valid for specimens and test conditions not covered in the present tests. when both the moisture content and the test temperature are high. while for prepreg repair there is an adhesive between the laminate and the repair patch (see figure 30). 61. For wet lay-up repair there is a thin resin layer. Fewer models exist for adhesive bonds joining composite materials. A repair should be cured at the highest permissible temperature so as to achieve the shortest cure time. in many practical situations. 43. 3546].e. the scarf and double lap joint models are relevant here because of our interest in scarf and doublesided uniform lap repair. and most of these are restricted to adhesives which may be treated as linearly elastic [28. PART II. 5. et al. Preparing the repair surfaces by sanding them with diamond sander ranging from 60 to 400 grit number does not significantly affect failure load. 62] by Hart-Smith and Adams. 46-48]. Nonelastic behavior of adhesives in bonding composite laminates has been considered for single lap [56]. The repair may be made by wet lay-up or by prepreg. In developing the models. Numerous models are available for isotropic materials bonded adhesively by single lap [30. Our objective is to develop analytical models which can be used to calculate the failure loads of laminates repaired by the scarf and lap techniques. In this investigation. double lap [57-61]. 62] did not treat each composite layer separately. However.. and scarf [49] joints. Of these. 38. Symmetric composite laminates can be repaired either by the uniform lap or by the scarf technique (see figure 29). The aforementioned findings are strictly valid only for the specimens and test conditions used in this investigation. but smeared the properties of the composite laminates which were adhesively bonded. THE PROBLEM. previous models for adhesively bonded joints were consulted because repair patches and bonded joints have similar features. but as far as authors can ascertain. Both the laminate and the repair patch consist of fiber reinforced plies which behave in a linearly elastic manner. Adams [61] performed stress analysis of scarf and double-lap joints using finite element methods.• • • The failure loads of repaired specimens are reduced under hot/wet conditions. and scarf joints [59. 6.

Repair patch P Laminate (a) Uniform lap repair P Repair patch P Laminate (b) Scarf repair without external ply P Repair patch P External plies Laminate P (c) Scarf repair with external plies FIGURE 29. MODELS OF THE UNIFORM LAP AND THE SCARF REPAIRS FIGURE 30. THE INTERLAYER BETWEEN THE LAMINATE AND THE REPAIR PATCH WHEN THE REPAIR IS PERFORMED WITH WET LAY-UP (LOWER LEFT) OR PREPREG (LOWER RIGHT) 31 .

as shown in figure 33. A model for the double-sided uniform lap repair is illustrated in figure 32.2. the model of the uniform lap repair can be applied to a stepped lap repair. ILLUSTRATION OF THE SHEAR STRESS-SHEAR STRAIN RELATIONSHIP OF AN ELASTIC-PERFECTLY PLASTIC INTERLAYER 7. the failure loads of laminates repaired by uniform and stepped lap techniques are similar. The laminate is subjected to an in-plane tensile load P. For modeling such a repair it is sufficient to consider only one half the repaired laminate. The model for the stepped lap repair is not given separately because. n L l Repair patch Number of plies.are treated as an “interlayer” which exhibits elastic-perfectly plastic behavior (see figure 31). hi Lap length. Thus. The objective is to find the value of this load (failure load. respectively. as was shown in section 4. G= τp γ ef FIGURE 31. Laminate Number of plies. DOUBLE-SIDED UNIFORM LAP REPAIR MODEL The lay-up of the laminate/repair patch assembly is symmetric. P = F) which fails the repaired laminate. The shear strains at the elastic limit and at plastic failure are γef and γpf. 32 . MODEL OF UNIFORM LAP REPAIR. n l R P P Interlayer Thickness. dl FIGURE 32.

models are presented which provide the failure loads when failure is due to any of the three scenarios.Repair patch P Laminate P Interlayer P/2 Repair patch Laminate P/2 Repair patch x P FIGURE 33. the repair patch may fail in tension. DOUBLE-SIDED UNIFORM LAP REPAIR TREATED IN THE MODEL When the repaired laminate is subjected to a tensile load. or the interlayer may fail in shear (see figure 34). In the following. F F (a) Laminate failure in tension F F (b) Repair patch failure in tension F F (c) Interlayer failure in shear FIGURE 34. ILLUSTRATION OF THE TYPES OF FAILURE WHICH MAY OCCUR IN A LAMINATE REPAIRED BY THE DOUBLE-SIDED UNIFORM LAP TECHNIQUE 33 . the laminate may fail in tension.

R   − 2 cosθ L . 2. the failure load is calculated as follows. THE ON-AXIS (x.R  6   sin 2 θ L . figure 33) the in-plane failure load (per unit width) in the repair plies is F/2 (P/2 = F/2).R  cos 2 θ L . Since the model is symmetric.R 2 cosθ L .R − sin 2 θ L .R cos 2 θ L .R cosθ L .1 TENSILE FAILURE OF THE LAMINATE OR REPAIR PATCH. The αij’s are the components of the compliance matrix.R ε xL . When the laminate or the repair patch fails in tension (see figure 34). and 6 in figure 35) components of the strains are ε1L . At tensile failure (of either the laminate or the repair patch) the applied inplane tensile load (per unit width) is denoted by F (P = F).R ε sL .R   L .R    (2) F where θ is the fiber orientation measured with respect to the direction of the applied load. At the middle of the repair zone (x = 0.R sinθ L . the in-plane strains at failure are [63] Laminate Repair Patch (ε ) (ε ) (ε ) L x F L y F L = α11F L = α 21 F L = α 61F L s F (ε ) (ε ) (ε ) R x F R y F R = α11F / 2 R = α 21 F / 2 R = α 61F / 2 (1) R s F where εx. The superscripts L and R refer to the laminate and the repair patches. FIGURE 35.R sinθ L .R ε 2  =  ε L .R sinθ L . The on-axis (1. respectively.7.R   L  − cosθ L .R  sin 2 θ L .R sinθ L . and εs are the off-axis in-plane strains (see figure 35).R ε y . 2) COORDINATE SYSTEMS 34 . y) AND OFF-AXIS (1. εy.R   cos 2 θ L .

respectively. transverse. and shear moduli of the ply.2 SHEAR FAILURE. balancing the forces for the laminate and each of the repair patches gives: dN L . ( ) ( ) 35 . 7. the loads on the repair patches (plies) are P/2. τ (= τ zx ) is the shear stress in the interlayer. respectively. E2.2τ = 0 dx dN +τ = 0 dx R Laminate (4) Repair Patch L R N L = N x and N R = N x are the in-plane loads (per unit width) in the laminate and each repair patch. and S are the on-axis longitudinal. transverse. E1. one of the following maximum strain failure criteria is met XL =1 E1L (ε1L )F XL =1 L E2L (ε 2 )F SL =1 L E6L (ε 6 )F XR =1 E1R (ε1R )F YR =1 R E2R (ε 2 )F SR =1 E6R (ε 6R )F (3) X. At the start of the analysis the repaired laminate has an applied in-plane load P (P ≠ F. Y. When the repaired laminate fails due to shear failure of the interlayer (see figure 34(c)). Six F values are calculated for each ply from equations 1-3. and shear strengths of the ply. the failure load is calculated as follows. and E6 are the on-axis longitudinal. figure 36). For a section dx in length (x being in the direction of the applied load). The lowest value of F resulting from these calculations is the failure load.Failure occurs when. in any of the laminate or repair plies. At x = 0. respectively.

hi is the thickness of the interlayer. I P Laminate. L NR R τ dx I NR + dNR dx dx NL L NL + dNL dx dx I NR R NR + dNR dx dx dx FIGURE 36. and γ (= γzx) is the interlayer shear strain. R P/2 P/2 x dx Interlayer. compatibility requires that the following condition be satisfied [64] (see figure 37) ( ε xL + 1)dx + γhi = ( ε xR + 1)dx + ( γ + This can be simplified to give dγ ε xL − ε xR = dx hi (6) dγ dx)hi dx (5) where ε xL and ε xR are the off-axis in-plane strains in the laminate and in the repair patch. respectively. 36 . LOADS ON A SECTION dx IN LENGTH OF THE LAMINATE REPAIRED BY THE DOUBLE-SIDED UNIFORM LAP TECHNIQUE By neglecting the shear deformations of the laminate and the repair patch.Repair patch.

or (4) a perfectly plastic region near both the x = 0 and x = dl ends of the interlayer (figure 38).(1+εR)dx x Repair patch γ+ hi γ I dγ dx dx R Interlayer L Laminate (1+εL)dx x I R FIGURE 37. AND THE LAMINATE (Double-sided uniform lap technique. (2) a perfectly plastic region near the x = 0. THE INTERLAYER. the x components of the strains in the laminate and in the repair patch are [63] L ε xL = α11 N L Laminate (7) Repair Patch ε =α N R x R 11 R where α11 is the 11 component of the αij compliance matrix. 37 . (3) a perfectly plastic region near the x = dl end of the interlayer. DEFORMATIONS OF THE REPAIR PATCH. Equations 6 and 7 yield dγ 1 L L R = (α 11 N − α11 N R ) dx hi By combining equations 4 and 8 obtains d 2γ 1 L R = (2α 11 + α 11 ) 2 dx hi (9) (8) There are four possible scenarios for the behavior of the interlayer: (1) the entire interlayer behaves in a linearly elastic manner.) For the symmetric arrangement considered here.

Equations 9 and 13 yield γp = λ2γ ef 2 x 2 + rx + s Plastic Region (14) 38 . the shear strain and shear stress are γ = γe and τ = Gγe (10) where G is the shear modulus of the interlayer. expression for the shear strain can be obtained.dl P/2 Repair patch Laminate P/2 x Repair patch Interlayer P τ τp Elastic τ τp Plastic Elastic τ τp Elastic Plastic τ τp Plastic Elastic Plastic γ γ γ γ x x=0 x = dl x=0 xp1 x = dl x=0 xp1 x = dl x=0 xp1 xp2 x = dl FIGURE 38. as shown in figure 31. POSSIBLE REGIONS OF THE INTERLAYER UNDER AN APPLIED LOAD P In the elastic region. From equations 9 and 10 the following γ = psinh(λx) + qcosh(λx) where λ is defined as Elastic Region (11) λ2 = G L R (2α11 N L +α11 ) hi (12) In the plastic region the shear strain and the shear stress are (see figures 31 and 38) γ = γp and τ = τ p = Gγ ef (13) where. γ ef is constant.

• At the free ends of the laminate and the repair patches. the load per unit width NL in the laminate is equal to the applied load per unit width P. p.1 Boundary and Continuity Conditions. 7. these boundary conditions can be expressed as dγ dx and L dγ Pα11 = dx hi =− R Pα 11 2hi at x = 0 (17) at x = dl (18) In equations 17 and 18 dγ dγe dγp is either or depending on whether the interlayer is in the dx dx dx elastic or in the perfectly plastic region at the boundary (see figures 40 through 43). q. 15. r.In equations 11 and 14. and 16.2. The following are the boundary and continuity conditions for to equations 11 and 14. and s are constants which must be determined from the boundary and continuity conditions. the axial load is zero (see figure 39) NL = 0 NR = 0 • at x = 0 at x = dl (15) At the ends of the laminate and the repair patches where loads are applied (see figure 39). 39 . and the load per unit width NR in each repair patch is equal to P/2 NR = L P 2 at x = 0 (16) at x = dl N =P By combining equations 8.

x = x p2 dγ p dγ e = dx dx The locations xp1 and xp2 are unknown and must be determined from the solutions of the equations summarized in figures 41 through 43. (19) 40 . Correspondingly. THE BOUNDARY CONDITIONS FOR THE IN-PLANE LOADS IN THE LAMINATE AND IN THE REPAIR PATCH AND THE CONTINUITY CONDITIONS FOR THE SHEAR STRAIN IN THE INTERLAYER At the locations where the elastic and plastic regions meet (x = xp1 and x = xp2 in figure 38) the shear strains in the elastic and plastic regions are equal (with the value γef) and are continuous (see figures 39.Repair patch P Laminate x P at x=0 NR = P/2 NL = 0 NR = P/2 x γ γef Plastic dl Repair patch Laminate Repair patch at x=dl NR = 0 NL = P NR = 0 γp γe Elastic γp Plastic xp1 γe = γp = γef d γe d γp dx = dx xp2 x at x=xp1 and x= xp2 FIGURE 39. 41 through 43). γ e = γ p = (γ ef ) at x = x p1 .

Interlayer P/2 Repair patch Laminate P/2 γ γef Elastic P Repair patch 0 dl x γe = c1 sinh( λ x) + c2 cosh( λ x) at x = 0 d γe = dx P aR 11 2hi at x = dl P aL dγe 11 = dx hi FIGURE 40. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 END 41 . THE EQUATIONS AND BOUNDARY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE ENTIRE INTERLAYER BEHAVES IN A LINEARLY ELASTIC MANNER Interlayer P/2 Repair patch Laminate P/2 Repair patch P γ γef Plastic Elastic 0 xp1 dl x 2γ γp = λ ef x2 + r x + s 2 γe = p sinh( λ x) + q cosh( λ x) at x = 0 d γp = dx P aR 11 2hi at x = xp1 γe = γp = γef d γe d γp dx = dx at x = dl P aL dγe 11 = dx hi FIGURE 41.

THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = dl END Interlayer P/2 Repair patch Laminate P/2 γ γef Plastic Elastic Plastic P Repair patch 0 xp1 xp2 x dl 2γ γp = λ ef x2 + r1 x + s1 2 γe = p sinh( λ x) + q cosh( λ x) 2 γp = λ γef x2 + r2 x + s2 2 at x = 0 d γp = dx P 2hi aR 11 at x = xp1 γe = γp = γef d γe d γp dx = dx at x = xp2 γe = γp = γef d γe d γp dx = dx at x = dl P aL dγp 11 = dx hi FIGURE 43. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 AND x = dl ENDS 42 .Interlayer P/2 Repair patch Laminate P/2 γ γef Elastic Plastic P Repair patch 0 xp1 dl x γe = p sinh( λ x) + q cosh( λ x) 2γ γp = λ ef x2 + r x + s 2 at x = 0 dγe = dx P aR 11 2hi at x = xp1 γe = γp = γef d γe d γp dx = dx at x = dl P aL dγp 11 = dx hi FIGURE 42.

and the repair patch may fail in tension while the interlayer fails in shear (see figure 45). and at each load the shear strain as a function of x is calculated by the equation given either in figure 41 (when the plastic region is near x = 0) or in figure 42 (when the plastic region is near x = dl). The load is gradually increased. at any point in the interlayer. When such a repaired laminate is subjected to a tensile load. left).2 Calculation of the Shear Failure Load. the shear strain is compared to the plastic failure strain γpf. and at each load the shear strain as a function of location x is calculated by the equations given in figure 43. the laminate may fail in tension while the interlayer fails in shear. The procedure is repeated until the shear strain is at or above the elastic limit at both the x = 0 and x = dl ends of the interlayer. The applied load is gradually increased. the interlayer may fail in shear. 43 . c. is taken to be the failure load (P = F). The procedure for calculating the loads at which each of these types of failure occurs is given below. At each load in steps b. A load P is applied such that under this load the entire interlayer is in the elastic region (figure 38. reaches the plastic failure strain γpf. 8. and c.2. The shear strain as a function of location x is calculated by the equation in figure 40. Consider a symmetric composite laminate repaired by the scarf technique (see figure 44). and at each load the shear strain is calculated. the laminate may fail in tension. a. The procedure is repeated until the shear strain reaches the elastic limit γef either near the x = 0 or near x = d1 end of the interlayer. The load P at which the shear strain γp. MODEL OF SCARF REPAIR. The applied load is gradually increased. b.7. The applied tensile load F which fails the interlayer in shear is calculated by the following steps using the equations and boundary and continuity conditions summarized in figure 40 through figure 43. the repair patch may fail in tension.

ne Overlap length of external plies de de P Scarf length. n l Interlayer Thickness. ds P Laminate L Number of plies. dso P Scarf length.Repair patch Number of plies. ds P (b) With external plies FIGURE 44. nR l Overlap length per ply. SCARF REPAIRS TREATED IN THE MODEL F (a) Laminate failure in tension F F (d) Laminate / interlayer failure F F (b) Repair patch failure in tension F F (e) Repair patch / interlayer failure F F (c) Interlayer failure in shear F FIGURE 45. hi (a) Without external ply Repair patch External plies Number of plies. THE TYPES OF FAILURE WHICH MAY OCCUR IN A LAMINATE REPAIRED BY THE SCARF TECHNIQUE 44 .

as shown in figure 46. THE SCARF REPAIR TREATED IN THE MODEL The repaired laminate is subjected to an applied in-plane load P. For a section dx in length (x being in the direction of the applied load. The superscript k refers to the k-th overlap segment. 45 .1 INTERLAYER SHEAR STRAIN. consider half of the repaired laminate. figure 46).8. Repair patch P P Laminate Interlayer Repair patch k=1 k=2 k P k=K P Laminate x1 x2 xK x FIGURE 46. This is illustrated in figure 47. The shear strain is calculated as follows. τ (= τ ) is the shear stress in k ( ) N R k R x k k zx the interlayer. respectively. idealizing the taper as a series of discrete steps. To calculate shear strain in the interlayer. a force balance for the laminate and the repair patch gives d ( N )− k L k dx τ =0 Laminate (20) d( N ) k + τ =0 dx k L N = k N x and L k R Repair Patch (= N ) are the in-plane loads (per unit width) inside the laminate and inside the repair patch for the k-th overlap segment.

46 .FIGURE 47. respectively. hi is the thickness of the interlayer. compatibility requires that the following condition be satisfied [64] (figure 48) (ε k L x + 1)dx + γh = ( ε k k i R x  + 1)dx +   d (k γ )  γ+ dx hi  dx   k (21) This can be simplified to give d (k γ ) k ε xL − k ε xR = dx hi (22) where k ε xL and k ε xR are the off-axis in-plane strains in the laminate and in the repair patch. and kγ (= kγzx) is the interlayer shear strain. LOADS ON A SECTION dx IN LENGTH OF THE k-th SEGMENT OF THE LAMINATE REPAIRED BY THE SCARF TECHNIQUE Neglecting shear deformations of the laminate and the repair patch.

AND THE LAMINATE IN THE k-th SEGMENT OF A LAMINATE REPAIRED BY THE SCARF TECHNIQUE At any overlap segment the laminate and the repair patch is not necessarily symmetric. The x components of the strains in the unsymmetric laminate and in the unsymmetric repair patch are [63] k L ε xL = k α11 k N L Laminate (23) Repair Patch k ε =α R x k Rk 11 N R where kα11 is the 11 component of the laminate or repair patch compliance matrix in the k-th overlap segment (see figure 49).(1+ kεR) dx x Repair patch kγ R + hi d( kγ) dx dx kγ I Interlayer L Laminate L (1+ kεx) dx FIGURE 48. Equations 22 and 23 yield d (k γ ) 1 k L k L k R k R = ( α11 N − α11 N ) dx hi By combining equations 20 and 24 one obtains d 2 (k γ ) 1 k L R = ( α11 + α11 )kτ 2 dx hi (25) (24) 47 . DEFORMATIONS OF THE REPAIR PATCH. THE INTERLAYER.

there is a perfectly plastic region either near the x = 0 or near the x = xK end of the repair patch. THE COMPLIANCE MATRICES IN THE LAMINATE AND THE REPAIR PATCH There are four possible scenarios for the behavior of the interlayer (see figure 50). (Scarf technique)) In the elastic region. or there are perfectly plastic regions near both the x = 0 and x = xK ends of the repair patch (xK is defined in figure 46). interlayer is perfectly plastic near the x = 0 end. Repair patch Interlayer P x=0 x=xK P Laminate Elastic Plastic Elastic Elastic Plastic Plastic Elastic Plastic FIGURE 50. the shear strain and shear stress are k γ =k γ e and k τ =G k γ e (26) where G is the shear modulus of the interlayer. The entire interlayer behaves in a linearly elastic manner. POSSIBLE REGIONS OF THE INTERLAYER UNDER AN APPLIED LOAD P (From left to the right: interlayer is linearly elastic. From equations 25 and 26 the following expression for the shear strain is obtained k γ e = k p sinh k λx + k q cosh k λx ( ) ( ) Elastic Region (27) where kλ is defined as ( λ) k 2 = G k L k R ( α11 + α11 ) hi 48 (28) . interlayer is perfectly plastic near the x = xK end.FIGURE 49. and interlayer is perfectly plastic near the x = 0 and x = xK ends.

1. 31. the boundary and continuity conditions corresponding to equations 27 and 30 are as follows. figure 53) the shear strain in the interlayer is continuous and the in-plane loads are equal and opposite on the left and right sides of the segment. γef is constant. and ks are constants which must be determined from the boundary and continuity conditions. these boundary conditions can be expressed as R d (1γ ) P1α11 =− dx hi at x=0. Then. It is assumed that the load is transmitted only by continuous layers 49 . and is equal to the applied load (per unit width) P in the laminate 1 NL = 0 at x = 0 (31) l R N =P NL = P N =0 R K K at x = xK (32) By combining equations 24. At the outside edge of the repair (x = xK) the axial load (per unit width) is zero in the repair patch. (33) and L d (K γ )  P kα11  =  h   dx  i  at x = xK (34) In equations 33 and 34 kγ is either kγe or kγp depending on whether the interlayer is in the elastic or in the perfectly plastic region at the boundary. and is equal to the applied load (per unit width) P in the repair patch (see figure 52).1 Boundary and Continuity Conditions. as shown in figure 31.In the plastic region the shear strain and shear stress are (see figure 31) k γ =k γ p and k τ = k τ p = Gγ ef (29) where. kq. • At the inside edge of the repair patch (x = 0) the axial load (per unit width) is zero in the laminate. 8. kr. The terminations of the external plies are modeled as illustrated in figure 51. kp. and 32. • At the edge of each overlap segment (at x = xK. Equations 25 and 29 yield k k γp = (λ )2 γ ef 2 x 2 + k rx + k s Plastic Region (30) In equations 27 and 30.

and γ = γp when interlayer is in the plastic region. and 39 γ = γe when interlayer is in the elastic region. the above continuity conditions for the k-th segment can be expressed in terms of kγ as follows k+1 γ = kγ k k k (38) at x = xk (39) d ( γ ) = d ( γ ) α *+ P * k +1 dx dx where kα* and kP* are defined as k a* = k +1 L R α 11 + k +1α 11 k L R α 11 + k α 11 (40) k L R R L P  kα11 k +1α11 − k α11 k +1α11    P* =  k L k R  α11 + α11 hi   (41) In equations 35. the shear strains in the elastic and plastic regions are equal (with the value γef) and are continuous (see figures 54 and 56 through 58). 38. k γ e = k γ p = (γ ef ) at x = xp1. and 37. 36. x = xp2 (42) d kγ p d kγ e = dx dx ( ) ( ) The locations xp1 and xp2 are unknown. at the edge of each segment the following continuity conditions are applied. • At the locations where the elastic and plastic regions meet (x = xp1 and x = xp2 in figure 54). 50 .(for example in figure 53 load is transmitted by the bottom four layers in the laminate and the top three layers in the repair). and must be determined from the solutions of the equations summarized in figures 55 through 58. Correspondingly. Accordingly. k+1 k+1 γ = kγ at x = xk (35) (36) (37) NL ≅ kNL k+1 R N ≅ kNR By using equations 24.

5de de k=K k=K k=K x=xK x=xK x=xK FIGURE 51. MODEL OF THE BOUNDARY AT THE OUTER EDGE (x = xK) OF THE REPAIR PATCH Interlayer P Repair patch P Laminate at x=0 1NR=P 1NL=0 at x=xK KNR=0 KNL=P FIGURE 52.No external ply 1 external ply 2 external plies de de de Laminate Laminate Laminate dav=1. THE BOUNDARY CONDITIONS FOR THE IN-PLANE LOADS IN THE LAMINATE AND IN THE REPAIR PATCH (Scarf technique) 51 .

THE CONTINUITY CONDITIONS AT THE EDGE OF THE k-th SEGMENT (Scarf technique) FIGURE 54.Repair patch P k k+1 Interlayer P Laminate x=xk Interlayer k k+1 kNR = k+1NR kNL = k+1NL kγ = k+1γ γ = γe if elastic γ = γp if perfect plastic FIGURE 53. THE CONTINUITY CONDITIONS AT THE INTERLAYER BETWEEN THE ELASTIC AND PERFECTLY PLASTIC REGIONS 52 .

THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 END 53 . THE EQUATIONS AND BOUNDARY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE ENTIRE INTERLAYER BEHAVES IN A LINEARLY ELASTIC MANNER k=K k=1 x=0 xp1 xk x=xK P P Plastic kλ2γef kγ p= Elastic 2 x2 + kr x + ks kγ e = kp sinh( kλ x) + kq cosh( kλ x) at x = 0 d( 1γp) = dx P( 1αR ) 11 hi at x = xp1 kγ k e = γp = γef at x = xk k+1γ k = γ at x = xK d( Kγp) P( KαL ) 11 = hi dx d( kγe) d( kγp) dx = dx d( k+1γ) d( kγ) k * k * α +P = dx dx γ = γe γ = γp if elastic if perfect plastic FIGURE 56.P k=K k=1 x=0 xk x=xK P Elastic kγ e = kp sinh( kλ x) + kq cosh( kλ x) at x = 0 d( 1γe) P( 1αR ) 11 = hi dx at x = xk kγ e at x = xK d( Kγe) P( KαL ) 11 = dx hi = k+1γe d( k+1γe) d( kγe) k * k * α +P = dx dx FIGURE 55.

P k=K k=1 x=0 xk xp1 x=xK P Elastic Plastic kγ e = kc1sinh( kλ x) + kc2cosh( kλ x) kλ2γef kγ p= 2 x2 + kc3 x + kc4 at x = 0 d( 1γe) = dx at x = xp1 kγ k e = γp = γef at x = xk k+1γ k = γ at x = xK d( Kγp) P( KαL ) 11 = dx hi P( 1αR ) 11 hi d( kγe) d( kγp) dx = dx d( k+1γ) d( kγ) k * k * α +P = dx dx γ = γe γ = γp if elastic if perfect plastic FIGURE 57. THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = 0 AND x = xK ENDS 54 . THE EQUATIONS AND BOUNDARY AND CONTINUITY CONDITIONS FOR CALCULATING THE SHEAR STRAIN IN THE INTERLAYER WHEN THE INTERLAYER IS PERFECTLY PLASTIC NEAR THE x = xK END P k=K k=1 x=0 xp1 xk xp2 x=xK P Plastic kλ2γef kγ p= Elastic Plastic kλ2γef kγ p= 2 x2 + kc1 x + kc2 kγ e = kc3sinh( kλ x) + kc4cosh( kλ x) 2 x2 + kc5 x + kc6 at x = 0 d( 1γp) = dx P( 1αR ) 11 hi at x = xp1 and x = xp2 kγ k e = γp = γef at x = xk k+1γ k = γ at x = xK d( Kγp) P( KαL ) 11 = hi dx d( kγe) d( kγp) dx = dx d( k+1γ) d( kγ) k * k * α +P = dx dx γ = γe γ = γp if elastic if perfect plastic FIGURE 58.

8. the in-plane loads in the laminate and the repair patch are calculated as functions of axial position x k N L (x) = 0 + ∫0x k τdx N (x) = P − R xk ∫0 τdx Laminate (44) Repair Patch k 55 . left). is taken to be the interlayer shear failure load.1. i.2 FAILURE LOAD. The applied load is gradually increased in small increments.. and at each load the calculation is repeated until the shear strain reaches the elastic limit γef either near the x = 0 or near x = xK end of the repair patch. The process is repeated until the shear strain is at or above the elastic limit at both the x = 0 and x = xK ends of the interlayer. The applied tensile load at which the interlayer fails in shear is determined at each load step (steps b.1.. The interlayer shear strain is calculated by the following steps using the equations and boundary and continuity conditions summarized in figures 55 through 58. in section 8. The applied load is gradually increased. a. At each load the shear strain as a function of x is calculated by the procedure in step a. using the equations given in figure 58. The applied load is then gradually increased. The load P at which the shear strain kγp at any point in the interlayer reaches the plastic failure strain γpf . The shear strain as a function of location x is calculated by the equations in figure 55. a new value of 1γ is assigned. and at each load the shear strain as a function of x is calculated by the procedure in step a. At the x = xK boundary the x derivative of the shear strains is L P K α 11 calculated and compared to the value of hi L d k γ < P K α 11 = dx > hi ( ) at x = xK (43) If the left-hand side of this equation is less or higher than the right-hand side.2) by comparing the shear strains to the plastic failure strain γpf.8. These calculations are repeated until the boundary condition at x = xK is satisfied. Values for the applied load P and the interlayer shear strain 1γ at x = 0 are assumed such that the shear strain both at x = 0 and at x = xK (i. c. 1γ and Kγ) are in the elastic region (see figure 50.e.e. using the equations given either in figure 56 (when the plastic region is near x = 0) or in figure 57 (when the plastic region is near x = xK). b.2 Calculation of the Interlayer Shear Strain. To determine the applied tensile load under which either the laminate or the repair patch fails. and c. until the left-hand side of the equation becomes equal to the right-hand side.

R  ε s  (46) Failure is calculated by the maximum strain failure criteria Laminate XL k L =p k Lk L E1 ε1 YL =k p L k Lk L E2 ε 2 SL =k p L k Lk L E6 ε 6 k k k k Repair Patch XR k R =p k Rk R E1 ε1 YR =k p R k Rk R E2 ε 2 SR =k p R k Rk R E6 ε 6 p L . and shear moduli of the ply in k-th segment and p are loads.R sinθ L . k The off-axis in-plane strains as a function of x in the laminate and in the repair patch are Laminate k L ε xL (x )= k α11 k N L (x ) L L ε y (x )= k α 21 k N L (x ) L ε sL (x )= k α 61 k N L (x ) k Repair Patch R ε xR (x )= k α11 k N R (x ) R R ε y (x )= k α 21 k N R (x ) R ε sR (x )= k α 61 k N R (x ) k k (45) k k where the αij’s are the components of the compliance matrix [63]. k 56 . transverse.R 2 cosθ L .R  sin 2 θ L . transverse.R − sin 2 θ L .R sinθ L . The above failure criteria is evaluated in each laminate and repair ply at every overlap segment. and shear strengths. and kS are the on-axis longitudinal.R cos 2 θ L . kE2.R  cos 2 θ L .R  sin 2 θ L .R  ε y   L .R   k ε xL .τ is the shear stress having the value of kτ = Gγe in the elastic region and kτ = kτp = Gγef in the plastic region.R ≥ 1 Failure No Failure (48) k k k (47) k X. The on-axis in-plane strains of each ply in the laminate or in the repair patch in k-th overlap segment are obtained by transformation k ε1L .R   L . and kE6 are the on-axis longitudinal.R sinθ L .kY.R   − 2 cosθ L .R sinθ L .R   − cosθ L . kE1.R   6  k  cos 2 θ L .R cosθ L .R < 1 p L .R   L .R  ε 2 = ε L .

The lowest of these three is the failure load of the repaired composite. The repair lap length varied from 0 to 3 inches. For scarf repair. 1. 9. comparisons were made for 8. If failure occurs in the laminate or the repair patch at one of the overlap segments. and 2). the failure is considered to have occurred in the repair patch (see figure 45. d and e). respectively.and 32-ply laminates repaired with either 1. The properties selected by such a “backcalculation” procedure were then used in all subsequent calculations. the failure is considered to have occurred in the laminate (see figure 45. a). The input parameters required by the programs are listed in tables 13 and 14. If failure occurs at the x = xK end of the repaired laminate.15 degree) and for different number of external plies (ne = 0. RESULTS. and 2. Two computer programs were written in MATLAB to generate numerical results for the two models described in the previous two chapters. 57 . or 3 repair plies. Failure loads generated by the model and the corresponding computer codes were compared to data generated with repaired specimens specified in figures 5 and 14. comparisons were made for different scarf angles (β = 1. These codes are designated as “RepairL” and “RepairS” and apply. b). 1. The properties of the interlayer were obtained by matching the model to one data point. to the double-sided uniform lap repair and the scarf repair. The programs provide the tensile failure loads in the laminate and in the repair patch and the shear failure load in the interlayer. 2. For the uniform lap repair.07.Failure occurs when under the applied load p is less than unity in any ply. • • • If failure occurs at the x = 0 end of the repaired laminate. then failure is considered to have occurred due to a combination of laminate (or repair patch) failure and interlayer failure (see figure 45.43. The material properties and parameters used in the calculations are given in tables 13 and 14.

θL (degree) Number of plies.48 0.28 0.045 0.1 8.TABLE 13. [b] from Naik [65] [c] by backcalculation. nlL Longitudinal modulus.5 0. 58 . ν2R[b] Longitudinal tensile strength.045 93 93 4.045 75 75 5.45 0. υ1L[b ] Transverse Poisson’s ratio.045 0. dl (in) Shear modulus. γpf[c] Thickness. hi (in)[c] * Parameters listed are for each ply. YL (ksi)[a] Shear strength.045 0.5 7. XL (ksi)[a] Transverse tensile strength. ν1R[b] Transverse Poisson’s ratio.008 L Longitudinal Poisson’s ratio. 32 7. hL (in)[a] Orientation of each ply. YR (ksi)[a] Shear strength.3 0 to 3.0094 0 8. E 6 (Msi)[b] Specimen Condition/Test Temperature Dry/70°F Moist(H2O)/180°F 8.5 0.48 0.0094 0 1.0094 0 8.5 0. G (ksi)][c] Shear strength.0 0.0 100 1. [a] from tables 2 and 3. SL (ksi)[b] Ply thickness. THE INPUT PARAMETERS REQUIRED BY RepairL AND THE NUMERICAL VALUES USED IN THE PRESENT CALCULATIONS FOR 3k70 PLAIN WEAVE FABRIC IMPREGNATED WITH 9396 RESIN Category Laminate* Parameter Longitudinal modulus.0 200 2. XR (ksi)[a] Transverse tensile strength.0 0. nlR Maximum lap length.5 0. υ L[b ] 2 Repair Patch* Interlayer Longitudinal tensile strength.2 7. E1R (Msi)[a] Transverse modulus.30 0. 32 7.045 101 101 5.2. θR (degree) Number of plies per side.0 0.1 0.2. E6R (Msi)[b] Longitudinal Poisson’s ratio. E1L (Msi)[a] Transverse modulus.0094 0 1. SR (ksi)[b] Ply thickness.045 68 68 4.045 0.6 0. E2 (Msi)[a] L Shear modulus.5 8. E2R (Msi)[a] Shear modulus.2 0.45 0.008 8. hR (in) Orientation of each ply. τp (ksi)][c] Maximum shear strain.3 0 to 3.

ne Overlap length of external ply.045 93 93 4. E6R (Msi)[b] Longitudinal Poisson’s ratio. SR (ksi)[b] Ply thickness.5 0. θR (degree) Orientation of external ply.5 7.0 0.TABLE 14.0 0.1. [c] by backcalculation.2 0.0094 [(0/45)2]s 0 8 0. τp (ksi)][c] Maximum shear strain. 59 . G (ksi)][c] Shear strength. YL (ksi)[a] Shear strength.045 0. hi (in)[c] * Parameters listed are for each ply.045 0.1 0.008 8.5 0.0094 [(0/45)2]s 0 8 0.0094 [(0/45)2]s 8 1 to 2. γpf[c] Thickness.0 0.48 0.48 0. SL (ksi)[b] Ply thickness.5 0.45 0.375 100 1. THE INPUT PARAMETERS REQUIRED BY RepairS AND THE NUMERICAL VALUES USED IN THE PRESENT CALCULATIONS FOR 3k70 PLAIN WEAVE FABRIC IMPREGNATED WITH 9396 RESIN Category Laminate* Parameter Longitudinal modulus. [b] from Naik [65].5 7.25 to 0.1. ν2R[b] Longitudinal tensile strength.28 0. XL (ksi)[a] Transverse tensile strength.045 0.045 0. E1L (Msi)[a] Transverse modulus. υ1L[b ] Transverse Poisson’s ratio. hR (in) Orientation of each ply. θe (degree) Number of plies.375 200 2.5 0.30 0.5 8.2 0. ν1R[b] Transverse Poisson’s ratio.045 75 75 5. nsL Scarf angle. [a] from tables 2 and 3. de (in) Shear modulus. E2R (Msi)[a] Shear modulus. E2 (Msi)[a] L Shear modulus.5 7.0094 [(0/45)2]s 8 1 to 2. hL (in) Orientation of each θL (degree) Number of plies.008 L Longitudinal Poisson’s ratio.1 8. nsR Number of external plies. β (degree) Longitudinal modulus.25 to 0.45 0. υ L[b ] 2 Repair Patch* Interlayer Longitudinal tensile strength.045 68 68 4. E1R (Msi)[a] Transverse modulus.045 101 101 5. E 6 (Msi)[b] Specimen Condition/Test Temperature Dry/70°F Moist(H2O)/180°F 8.6 0. YR (ksi)[a] Shear strength.2 7. XR (ksi)[a] Transverse tensile strength.2 0.

8000 6000 4000 Specimen: Dry Test: 70F Laminate: F593 [0] 8 Repair: 9396 Specimen: Moist (H2O) Test: 180 F dl 1 repair ply Data Model 2000 Failure Load (lbf) 8000 6000 4000 2000 2 repair plies 8000 6000 4000 2000 0 0. COMPARISONS OF THE CALCULATED (MODEL) AND THE MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF LAP LENGTH FOR 8-PLY LAMINATE 60 .0 2.0 2.The failure loads calculated by the model are compared to the data in figures 59 through 62. both for dry specimens tested at 70°F and for moisturized specimens tested at 180°F.0 3.0 3. The calculated and measured failure loads agree well.0 4.0 1.0 3 repair plies Lap length. These agreements lend support to the validities of the models. dl (in) FIGURE 59. Thus.0 4. keeping in mind that the first failure prediction at each environment was used to obtain interlayer properties which were subsequently used for all other predictions. one can conclude that the trends are very well predicted by the model but not necessarily the absolute values of the failure load.0 1.

COMPARISON OF THE CALCULATED (MODEL) AND THE MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF LAP LENGTH FOR 32-PLY LAMINATE FIGURE 61.0 1.0 3.0 2000 2 repair plies H tan(β)= H ds ds 61 2. β (degree) FIGURE 60.Specimen: Dry Test: 70F 6000 5000 dl Specimen: Moist (H2O) Test: 180 F 7000 Laminate: F593 Repair: 9396 8000 6000 Laminate: F593 [0] 32 Repair: 9396 4000 4000 Data Model 1 repair ply 2000 3000 2000 1000 0 Data Model Specimen: Dry Test: 70F 8000 6000 4000 Failure Load (lbf) Failure load (lbf) 6000 5000 4000 3000 2000 1000 4.0 2.0 8000 6000 4000 2000 3 repair plies Specimen :Moist-H2O Test: 180 F 0 1 1.0 1.0 3.0 4. dl (in) Scarf Angle.0 Lap length.5 2 0 0. COMPARISONS OF THE CALCULATED (MODEL) AND THE MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF SCARF ANGLE .

07degree Laminate: F593 Repair: 9396 Failure load (lbf) 1000 0 6000 5000 4000 3000 2000 1000 0 No Extra Ply Specimen: Moist-H22 Specimen: Moist-H O Test: 180° Test: 180°F 1 Ply 2 Plies Number of Extra Plies. Scarf angle. However. Nonetheless.07 degree β = 1. In the present forms the models are for composite laminates. CONCLUDING REMARKS. 62 . the models may be applied to the repair of isotropic plates by replacing the properties of the composite with the corresponding properties of an isotropic material. n e FIGURE 62.7000 6000 5000 4000 3000 2000 Data Model Specimen: Dry Specimen: Dry Test: 70F F 70° ° Scarf angle. the results of these models serve as useful guides in the design of actual repairs by providing information on the various parameters affecting the failure load. α=1. COMPARISON OF THE CALCULATED (MODEL) AND THE MEASURED (DATA) FAILURE LOADS AS A FUNCTION OF EXTRA PLIES 10. The models and the corresponding computer codes described in the previous chapters are applicable to one-dimensional specimens.

and Solid Cored Structures.” Composite Repairs. 1995. R. Tullos. 8-12. L. 63-70. Boeing 737 Structural Repair Manual. Boeing Commercial Aircraft Co. M.” Composite Repairs. “Development of Rapid Cure Adhesive for Naval Aircraft Field Repair Applications. Simpson. 12. ed. C. McKinney. “Low Energy Cured Composite Repair System. ed. 1988. and D. W. 2 (C. “Metal Bond and Composite Repairs: Similarities and Differences. pp.” SAMPE Journal. Boeing Commercial Aircraft Co. 36-50. 75-83. M. “Repair of Advanced Composites Commercial Aircraft Structures. 29. C. Ong and S. F. ed. 1993. Vol. Deaton. Chesmar. “Repair Adhesives: Development Criteria for Field Level Conditions. 1985. SAMPE Monograph No. 1987. S. L. pp.” Theoretical and Applied Fracture Mechanics. ASM International. L.). 1991. No. 2 (C.).. 14-25.. 1991. 829-839. 5. “Repair of F-104 Aircraft Nosedome by Composite Patching.. Hall.” Engineered Materials Handbook.” Composite Repairs. L. Hamermesh. Steinmetz. Hamermesh. F. pp. Myhre. “Advanced Composite Repair-Recent Development and Some Problems. H. 1 (H. 6. pp. Raizenne. Clark and K. Cressy. D. 63 . and S. 11. pp. 2. S. “Adhesive Bonded Structural Repair. 4. Wegman and T. Shen. Brinkerhoff. Society of Manufacturing Engineers. 6. 1991. SAMPE Monograph No. “Field Repair Compounds for Thermoset and Thermoplastic Composites. Vol. 1991. SAMPE Monograph No.” Composite Repairs. Hamermesh. 51-62. Brown. pp.” Composites. B. 479-483. Hamermesh. D. pp. pp.2 (C. 3. Vol. SAMPE Monograph No. S. ed. Boeing Commercial Aircraft Co.). 2 (C. Vol. 3. 24-35. R. SAMPE Monograph No. 1991. R. 1993.” Composite Repairs. 1996. 9. Riefier and W. E. 15. Part III-Repair of Composite. Honeycomb Cored. D-53900. Lee. pp. “A Proposed Composite Repair Methodology for Primary Structure.11. L. D. Repair Procedures for Composite Structures. Cichon. ed. 10.). 20. 7.” Composites ’96 Manufacturing and Tooling Conference. J. J. Boeing 777 Structural Repair Manual. REFERENCES. 8.). 281-290. R. 1. pp. E. F. S. 1991. 13.

ed. Vol.” Composite Repair of Aircraft Structures Symposium. 26-39. 18.” Composite Repairs. pp. “Composite Repair with Wet Lay-up and Prepreg. “Repair of Advanced Composite Structures.” Composites ’96 Manufacturing and Tooling Conference.” Journal of Aircraft.” Composite Repairs. P. Kiger. 1985. D. “Field-Level Repair Materials and Processes.). and S. pp. Stone.). H.). H. 19. 184-194. Brown. “Improved Resins for Wet Layup Repair of Advanced Composite Structure. 1 (H. Society of Manufacturing Engineers. 1985. Stone. Repair Techniques for Graphite/Epoxy Structures for Commercial Transport Applications.” Composite Repairs. R. 359-380. Ahn. C. “Development of Large-Area Damage Repair Procedures for the F-14A Horizontal Stabilizer. CR 159056. 1 (H. G. 22. Scott and C. 1982. Myhre and J. 1981. “Development of Repair Procedures for Graphite/Epoxy Structures on Commercial Transports. 1 (H. Northrop Corporation. 1985. 1983. “Problems and Options in Advanced Composite Repair. H. ed. and P. 211-226. H. R.). 24. H. Springer. 23. Stone. 229-239. R. S. S. Bardygula. R. “CACRC: Progress and Plans. 1985. 15.” Composite Repairs. H. pp. Brown. SAMPE Monograph No. 1985. 1 (H. 40-51. 21. 110-116. J. Falcone. A. K. 20. S. W. pp. “Demonstration of Repairability and Repair Quality on Graphite/Epoxy Structural Supplements. 16. H. SAMPE Monograph No. L. J. SAMPE Monograph No. pp. 17. Mahon and J.” Fibrous Composites in Structural Design. 25. Labor. 1996. Springer. 26. 1980. O. 64 . Stone. H. NASA Contractor Report. 100-109. 1-24. S.” Composite Repairs. J. SAMPE Monograph No. “Field Repairs for the AH-1 Composite Main Rotor Blade. “Standardization of Composite Repairs for Commercial Transport Aircraft. pp. pp. Brown. F. pp. Candella.). Myhre and R. Shyprykevich. 1995. ed. Nor 82-60. Shyprykevich. 1 (H. pp.14. Knauss and R.” Composite Repairs. Brown.).” Composites ’96 Manufacturing and Tooling Conference. Advanced Composite Repair Guide. Mashiba. Phase I-Repair Development. S. 18. SAMPE Monograph No. Brown. Society of Manufacturing Engineers. SAMPE Monograph No. Plenum Press. ed. ed. 1996. Ahn. pp. 546-552. ed. S. 1 (H. Brown. pp. 87-99. 1985.

415-436.” Composites. L. G. and A.27. A17-A27. pp.” 21st Conference of Structures. Lee. Eidinoff. “The Stresses in Cemented Joints. 656-663. Chan and C. 33.” Quarterly Journal of Applied Mathematics. 207-228. 226-232. ASM International.” Journal of Adhesion. Vol. “Interfacial Stresses and Strength of Lap Joints. U. Aliman. Vinson. and R. Lee. S. 13. S. Aker. Kim. 1980. L. pp. W.” Composites. Kinloch. “Two Dimensional Displacement-Stress Distributions in Adhesive Bonded Composite Structures. S. O. H. 1938. W. Reissner. Vol. 720-728. 121-126. Radecky. Matthews. Volkersen. pp. pp. Vol. Kinloch.7. L. Pirvics. Parker. 16.” Composites. “The Bonded Repair of Fibre Composite: Effect of Composite Moisture Content. XXX. A. R. I. “Repair Concepts for Advanced Composite Structures. J. R. 1995. 1987. 30. 32. C. S. Matthews. “Moisture Problems in Advanced Composite Structural Repair. D. J. and F.11. 1944. 40. L. 1991. 1975. and S. Mylire. pp. Ojalvo and H. “Development of a Strength Model for the Cocured Stepped Lap Joints Under Tensile Loading. S. 204-211. 840-844. 593-600. 38. and Materials. 1977. Vol. 1979. “The Efficient Design of Adhesive Bonded Joints. Vol. pp. “Measuring and Predicting the Durability of Bonded Carbon Fiber/Epoxy Composite Joints. Vol. J.” Engineered Materials Handbook. H. pp. pp. 6. Labor. pp.” Journal of Applied Mechanics. “Bond Thickness Effects Upon Stresses in Single-Lap Adhesive Joints. J. E.” AIAA Journal. T. J. Goland and E. “Surface Preparation of Composites for Adhesive-Bonded Repair. “The Effect of Composite Prebond Moisture on Adhesive-Bonded CFRPCFRP Joints. L. 4. Structural Dynamics. 15. 31. 1983. 1978. Vol.” Luftfahrtforschung. 29. Pt.175-193. S.” Journal of Aircraft. J. Robson. Hart-Smith. Renton and J. pp. H. pp. 36. and D. 1974. F. J. 41-47. Vol. John. Vol. pp.” Composite Science and Technology. 1994. 35.” Composite Structures. Sun. M. 22. 235-246. 16. 32. 289-299. J. W.” Journal of Adhesion. 3. 14. pp. Vol. 39. 34. 37. J. D. “Die Nietkraftverteilung in Zugbeanspruchten Nietverbindungen mit Konstanten Laschenquerschnitten. 52. Vol. pp. “A Theory for Elastic Stresses in Adhesive Bonded Lap Joints. 28. Ochsner. 65 . B. M. Vol. Vol. J. AIAA/ASME/ASCE/AHS. Myhre and C. E. Beck. 1982.

“Stress Concentration Factors for Bonded Lap Joints. Peppiatt. J. pp. Hatt. 55. “Stress Distribution in Lap Joints With Partially Thinned Adherends. Vinson. 9. 27. Carver.” International Journal of Adhesion and Adhesives.” International Journal of Adhesion and Adhesives. 54. “Nonlinear Adhesive Bonded Joints Design Analyses. Grimes.” Adhesion 2 (K. J. 101-106. D. Vol. Lubkin. 817-820. Crocombe. 1971. pp. L. 52. 49. A. Adams. 1989. Vol. D. C. 5. D. 9.” SAE Trans. 66 . 7. 6. 1975. 1974. 710107. pp. pp. Vol. C.” Journal of Applied Mechanics. Bigwood and A. Vol.41. 46. “A Theory of Adhesive Scarf Joints. Vol. A. 145-153. “Effect of Ply Staking Sequence on Stress in a Scarf Joint. “Strength of Adhesive-Bonded Lap Joints With Respect to Change of Temperature and Fatigue. R. Vol. R. Johnson. 1989. Erdogan and M. “Analysis of Adhesively Bonded Joints Between Panels of Composite Materials. Vol. J.” Experimental Mechanics. A. 53. pp. Applied Science Publisher. “Stress Distribution in Adhesive Bonded Lap Joints. 370-378. Allen. R. 1957. 1973.” Journal of Composite Materials. pp. “Global Yielding as a Failure Criterion for Bonded Joints. Szépe. Adams and N. 280-286. D. A. 1977. M. 44. Vol. D. Crocombe. 42. 41-60. 48. “Elastic Analysis and Engineering Design Formulae for Bonded Joints. 50. 10.” Journal of Adhesion. 44. Wooley and D. Peppiatt. Vol. 229-242. Vol. 1650-1654. 185-196. Crocombe. W.” Journal of Strain Analysis. Vol. pp. F. “On The Behavior of Bonded Joints in Composite Material Structures.” AIAA Journal. 1971. D. 105-120. pp. 1990. D. 79-86. “Analysis of Bonded Joints in Vehicular Structures. Thamm. Ratwani. 1978. F. Baker and F. W. 7. “Stress Distribution in Bonded Joints. A.” Journal of Applied Mechanics. Renton and J. “Stress Analysis of Adhesive-Bonded Lap Joints. 47. pp. ed. 8. R. 1976. 9. Bigwood and A. 24. 11. pp. pp.” International Journal of Adhesion and Adhesives. 31-41. W. 1966.” Engineering Fracture Mechanics.” AIAA Journal. F. 301-309. and N. 45. R. pp. Vol. pp. J. Renton and J. R. pp. 1989. G. pp.). 379-393. Vol. G.” Journal of Aircraft. Coppendale. R. 51. L. Vinson. 1971. 43. “Failure Analysis of AluminuiumAluminuium Bonded Joints. 255-260.

60. 58. A. 1994. J. Matthews.). Technomic Publishing Co. pp. 1973. (F. Hart-Smith. Vol. Kinloch. L. Atkins. pp. 20. CR 112236. Tsai and H. pp. 62. Adhesive-Bonded Single-Lap Joints.. Hart-Smith. NASA Contract Report. Technomic Publishing Company. Megson. Naik. W. J. “Stress Analysis and Failure Properties of Carbon-Fiber-Reinforced-Plastic/Steel Double-Lap Joints. CR 112238. “Further Developments in the Design and Analysis of Adhesive-Bonded Structural Joints.). 67/68 . D. 1980. Aircraft Structures for Engineering Students. ASTM STP 749 (K. 1987. Hart-Smith. R. 185226. 63. Harris. Introduction to Composite Materials. Kedward. Halsted Press. Hahn. L. 2nd ed.” Journal of Adhesion. 1973. S. H. NASA Contract Report. CR 112237. Hart-Smith. Woven Fabric Composites.. American Society for Testing and Materials. 64. Hart-Smith. 57. Elsevier Applied Science Publisher. 59. 1981. ed. Adhesive-Bonded Double-Lap Joints. 29-53. J. CR 112235. R. 1986. L. 65. L. Adams. J. Adams. and A. J. Adhesive-Bonded Scarf and Stepped-Lap Joints. J. T. 1990. D. L. W. R. pp. L.1973. T. NASA Contract Report. J. J. 3-31.56. 405-406. N.” Joining of Composite Materials. 61. ed. NASA Contract Report. “Theoretical Stress Analysis of Adhesively Bonded Joints. G.1973.” Joining FibreReinforced Plastics. T. Non-Classical Adhesive-Bonded Joints in Practical Aerospace Construction.

3  Dt   = 1 . The change in the moisture content as a function of time is [A-1] M = GM(Mm .75    .093 percent = 0.7. The diffusivity D is obtained from the expression [A-1]  h  D = π  4M   m   2  M 2 − M1     t − t  1   2 2 (A-1) = 6 x 10-10 in2/sec where Mm is the maximum moisture content (% weight gain) ( Mm = 1. With these values. The measured moisture content as a function of time is shown in figure A-1 for laminates moisturized at 180°F.1 percent. respectively.1.Mi) + Mi where M is the moisture content.exp  2  h     (A-2) (A-3) The initial moisture content was Mi = 1. the percent change of the moisture content of the specimen. equations A-2 and A-3 give M = 1.0 percent for drying. is given.007 percent (A-5) (A-6) (A-4) A-1 . M1 and M2 are the moisture contents at time t1 and t2.M = 1. and GM is GM 0.093 percent The change in moisture content is Mi . The duration of the test was 600 seconds.1 percent . and Mm = 0. tested at 180°F.APPENDIX AMOISTURE LOSS OF SPECIMEN DURING TEST In this appendix. Mi is the initial moisture content.8 percent) and h is the specimen thickness (h = 0.075 in).

5 0. 2-20. Springer. 1976.0 0. Vol.0 Laminate: F593 1. Shen and G. H. A-1. pp.0 0 500 1000 0 500 1000 1500 sec sec FIGURE A-1. “Moisture Absorption and Desorption of Composite Materials.5 [0] 8 1.” Journal of Composite Materials.0 Laminate: F593 1. A-2 . S.0 0.5 Data Fit to data 0.5 [(0/45) 2] s Laminate exposed to 180 F humid air Moisture Content of Laminate M (Percent) 1. 10.Laminate Immersed in 180 F H2O 2. MOISTURE CONTENTS OF THE BASE LAMINATES IMMERSED IN WATER AT 180°F (LEFT). OR EXPOSED TO 100 PERCENT HUMID AIR AT 180°F (RIGHT) REFERENCE.

0 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 0.6 0.2 1.0 R6376 Specimen: Moist-Humid Air Test: 180 F Wet lay-up Prepreg 9396 9390 Epocast M20 Repair Material FIGURE B-2.2 0.8 0. FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (STEPPED LAP REPAIR) B-1/B-2 . FAILURE LOADS OF DIFFERENT BASE LAMINATES REPAIRED WITH DIFFERENT REPAIR MATERIALS (SCARF REPAIR) 1.0 Laminate: F593 R922 Failure Load ( Repaired / Undamaged) 0.0 0.8 0.8 0.2 0.6 0.2 1.2 1.APPENDIX BFAILURE LOADS OF SPECIMENS EXPOSED TO HUMID AIR The failure loads of the specimens repaired by the scarf (figure B-1) and stepped lap (figure B-2) techniques are included in this appendix.4 1. 1.0 0.8 0.6 0.0 1.2 0.4 0.4 0.0 R6376 Specimen: Moist-Humid Air Test: 180 F Wet lay-up Prepreg 9396 9390 Epocast M20 Repair Material FIGURE B-1.2 1.4 0.6 0.2 0.0 1.4 1.4 0.

AND 21 Specimen Condition/ Test Temperature Repair Type Stepped lap Parameter 1 repair ply (dl = 3 in) 2 repair ply (dl = 3 in) 3 repair ply (dl = 3 in) 8 laminate. 18. TABLE C-1. 1 repair plies (dl = 3 in) 8 laminate. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE DATA IN FIGURES 17. 2 repair plies dl = 3 in) 8 laminate. THE FAILURE LOADS (in lbf) OF SPECIMENS WITH GRIT NUMBER g = 60 USED IN NORMALIZING THE DATA IN FIGURE 24 Specimen Condition/ Test Temperature Dry/70°F Dry/70°F Moist (H2O)/180°F Moist (H2O)/180°F Failure Load (lbs) 4475 3278 3536 3012 Repair Type Scarf Stepped lap Scarf Stepped lap C-1 . 3 repair plies (dl = 3 in) Dry/70°F 1583 3893 4289 1601 3134 4368 5646 1710 2997 4055 Moist(H2O)/180°F 1295 2394 3224 1290 2521 3168 3858 1238 2502 3256 Uniform lap TABLE C-2. 1 repair plies (dl = 3 in) 32 laminate. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE DATA IN FIGURES 22. 4 repair plies (dl = 3 in) 32 laminate. 3 repair plies (dl = 3 in) 8 laminate. AND D-1 Specimen Condition/ Test Temperature Dry/70°F Dry/70°F Failure Load (lbs) 4526 3360 Repair Type Scarf Stepped lap TABLE C-3.APPENDIX CTEST RESULTS The failure loads used in normalizing the data are summarized in tables C-1 through C-5. 2 repair plies (dl = 3 in) 32 laminate. 23.

TABLE C-4. THE FAILURE LOADS (in lbf) USED IN NORMALIZING THE DATA IN FIGURE D-2 Specimen Condition/ Test Temperature Dry/70°F Dry/70°F Failure Load (lbs) 4541 4516 Repair Type Scarf Stepped lap C-2 . THE FAILURE LOADS (in lbf) OF SPECIMENS WITH CURE TEMPERATURE Tc = 150°F USED IN NORMALIZING THE DATA IN FIGURE 28 Specimen Condition/ Test Temperature Dry/70°F Dry/70°F Failure Load (lbs) 4553 3365 Repair Type Scarf Stepped lap TABLE C-5.

6 0.6 0.2 specimen at indicated condition dry specimen tested at 70F 1.2 0.4 0.8 0.0 Moist & 70F Dry & 180F Moist & 180F Failure Load ( Specimen Condition & Test Temperature FIGURE D-2.4 Laminate: F593 Repair: 9396 ) 1.2 specimen at indicated condition dry specimen tested at 70F 1.APPENDIX DFAILURE LOADS OF SPECIMENS UNDER MOISTURE AND TEMPERATURE The failure loads of the specimens repaired with wet lay-up (figure D-l) and with prepreg (figure D-2) materials are included in this appendix.2 0.8 0.0 1. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE ON THE FAILURE LOAD. THE EFFECTS OF SPECIMEN MOISTURE CONTENT AND TEST TEMPERATURE ON THE FAILURE LOAD 1.6 1.4 0.8 0. REPAIR MATERIAL M20 (PREPREG).2 0.2 0.4 1. D-1/D-2 .2 1.4 0.2 1.4 Laminate: F593 Repair: M20 ) 1.4 0.0 Moist & 70F Dry & 180F Moist & 180F Failure Load ( Specimen Condition & Test Temperature FIGURE D-1.6 1. 1.0 1.8 0.6 0.0 0.6 0.0 0.0 0.4 1.0 0.