REVIEW ARTICLE

on

STRUCTURAL JOINTS IN POLYMERIC COMPOSITE MATERIALS

PREPARED FOR

PERMANENT SCIENTIFIC COMMITTEE 56 FOR STRUCTURAL & CONSTRUCTION ENGINEERING

BY

HOSSAM EL-DIN MOHAMMED SALLAM, Ph.D.

January, 2004

STRUCTURAL JOINTS IN POLYMERIC COMPOSITE MATERIALS

BY

HOSSAM EL-DIN MOHAMMED SALLAM, Ph.D.

CONTENTS

1. INTRODUCTION

1

2. MECHANICALLY FASTENED JOINTS 9 2.1 Stress Concentration Factor in Composite Materials

Containing an Open Circular Hole. 11

2.2 Failure Modes in Mechanically Fastened Joints 13 3. BONDED JOINTS 21

3. 1. Stresses in Adhesive Joints 25 3.2 Standard Mechanical Test Procedures 28 3.3. Fracture Mechanics Considerations 30 3.4. Fatigue behavior and Durability of Adhesive lap joints 35 3.5 Prevention of Peeling failure in Adhesive Joints 35 3.6 Adhesive Bonded Dissimilar Materials 37 3.7 Fusion Bonding for Joining Thermoplastic Matrix Composites 40

4. SNAP JOINT IN COMPOSITE STRUCTURES 44 5. HYBRID JOINTS 46

5.1. Hybrid Joints in Plated Beams 47 6. SUMMARY 50

6.1 Mechanical Joints (Bolts or Rivets) 50 6.2 Adhesive Bonding 50 6.3 Hybrid joints 51

7. REFERENCE 52

1. INTRODUCTION

The permanent assembly of individual manufactured components is an important aspect

of fabrication and construction. There are many ways for accomplishing this including the

use of fastenings such as bolts and rivets, the use of adhesives, and by soldering, brazing

and welding [1]. Joint design is dependent upon the nature of materials to be joined as

well as the method of joining. The first cast iron bridge, at Ironbridge, in Shropshire,

England, was erected in 1777-1779 by casting massive half-arches and assembling the

mortise and tenon joints, with fit-on parts and tapered pegs, see Fig. 1-1.

Figure 1-1. The first cast iron bridge in Shropshire.

This mode of construction was the only one known to the designer, and is closely based

on the design of timber structures. The only concession to the new structural material was

in the use of large scale, single piece, curved members. Rolled bars of T- and L-sections

did not become available until after Napoleonic wars and the first I-beams, for light

loading only, were made in Paris in 1847. Heavy I-beams only become available in 1860

after the invention of mild steel. Fastening was originally by slotting a circular cylinder,

cast as an integral of the beam, over a cylindrical column, cast with a retaining lip, but

fish plates and bolting were introduced at early stage [2,3].

1

Steel “nailing” was developed and primarily used for fastening sheet metal to structural

steel for metal roof decks and claddings. Research undertaken at the University of

Toronto has shown that the existing technology of steel nailing also can be used to

connect structural steelwork, particularly hollow structural section steel members [4].

Cold-formed structural members may be joined with bolted connections that are designed

with the aid of applicable national design standards, such as American Iron and Steel

Institute (AISI). Winter [5] categorized the failure of bolted connections into four separate

modes, described as end tear-out, bearing, net-section fracture, and bolt shear. Recently

Rogers and Hancock [6] found AISI specification cannot be used to accurately predict the

failure modes of thin cold-formed sheet-steel bolted connections that are loaded in shear.

It is sometimes necessary to use combination (hybrid) joints in steel construction. For

example, high-strength bolts can be used in combination with welds, or rivets can be used

in combination with bolts. The need for a hybrid joint can rise for a number different

reasons. For example, the load demands on an existing bolted joint may change with time,

necessitating renovation of that joint. If the geometry does not permit the addition of more

bolts, welds can be added to the connection in order to increase its capacity [7]. Bonded

structures can be of two types based on purely adhesive or an adhesive/mechanical

connections. These types of connections include bonded-welded, bonded-riveted, and

bonded-screwed connections [8]. The concept hybrid joint is also used in structural

timber by injecting a resin into the gap between the connector and the wood to improve

the performance of bolted and dowelled joints [9]. On the other hand, steel bolts/screws

were used as shear connectors to joint steel/timber beam to concrete slab [10-12].

Elshafie [12] also used concrete shear key (dovetail) as shear connectors in timber-

concrete composite joints.

It is convenient now to define an adhesive as a polymeric material which, when applied to

surfaces, can join them together and resist separation. The structural members of the joint,

which are joined together by the adhesive, are the adherends, a word first used by de

Bruyne in 1939 [see Ref. 3]. Fracture mechanics has become a very popular tool for

2

characterization of adhesive metal joints in recent years [13]. Furthermore, Conrad et al.

[14] studied the effect of droplet diameter and droplet spacing on mode I fracture

toughness of a discontinuous wood-adhesive bonds. The fracture mechanics concept is

also used to understand the basic dentin adhesion mechanisms [15].

Advanced composite materials and structures have undergone rapid development over the

past four decades [16]. The majority of advanced composite materials are filamentary

with continuous fibers. As such, their behavior and the behavior of structures made from

them are more complicated than that of monolithic materials and their structures.

Advanced composite materials can be divided into classes in various manners. One

simple classification scheme is to separate them according to the reinforcement forms-

particulate-reinforced, fiber-reinforced, or laminar composites. Fiber-reinforced

composites, can be further divided into those containing discontinuous or continuous

fibers. Polymers, ceramics, and metals are all used as matrix materials, depending on the

particular requirements. The matrix holds the fibers together in structural unit and protects

them from external damage, transfers and distributes the applied loads to the fibers, and in

many cases contributes some needed property such as ductility, toughness, or electrical

insulation.

The structure of polymers consists of long molecules with a backbone of carbon atoms

linked by covalent bonds. In non-crystalline or amorphous polymers the molecular chains

have an entirely random orientation and are cross-linked occasionally by a few strong

covalent bonds and numerous but weaker van der Waals bonds. These weaker bonds

break as the temperature reaches a value known as the glass transition temperature, Tg,

characteristic for each polymer. Below Tg the polymer behaves as a linear elastic solid.

Creep becomes increasingly significant as the temperature increases and, above Tg, the

polymer deforms in a viscous manner under load. In crystalline polymers the molecules

are oriented along preferred directions, bringing with them optical and mechanical

anisotropy. Polymers are described as being either thermosets (e.g., epoxy, polyester,

3

phenolic) or thermoplastics (e.g., polyimide, polysulfone, polyetheretherketone,

polyphenylene sulfide). Ceramics, such as glasses, cement & concrete, and engineering

ceramics ( Al2O3, SiC, Si3N4, ZrO2), have a wide range of engineering applications. The

strong ionic and covalent nature of the bonding in most ceramics leads to a stable crystal

structure with a high melting point and high stiffness. Many ceramic materials have very

high elastic moduli and strengths, but the advantages of these properties bestow are often

outweighed by their highly brittle nature, which leads to low and unpredictable failure

stress resulting from the presence of flaws. Metal matrix composites typically comprise a

light metallic alloy matrix, usually based on aluminum, magnesium or titanium alloys.

The major reinforcing elements used in composites are glass, carbon/graphite, organic,

and ceramic. Both metal and ceramic materials have properties closer to those of likely

reinforcements and this leads to a different choice of properties for which these composite

systems are optimized. With polymer matrix composites, it is almost true to say that the

properties of the composite are essentially those of the fibers, with little contribution from

the properties of the matrix.

Fiber Reinforced Polymer (FRP), earlier limited to aerospace structures, are gaining

acceptance in civil engineering applications also. FRP plates, sheets, rebars, and strands

have been increasingly used in the construction industry due to their superior properties.

One of the potential structural applications of FRP composites with concrete or steel

structures is strengthening of RC or steel beams with unidirectional fiber-composite

sheets/plates bonded on their tension faces via the use of epoxy adhesives. Growing

maintenance and durability problems in transportation infrastructure have led researchers

to explore FRP box-girder, I-beam, or other shapes in bridges [17-19]. There are

numerous examples of completed all-composite new bridges and several more are under

construction, see Fig. 1-2. FRP materials could play an important role for the construction

industry of tomorrow.

4

Aberfeldy Footbridge in Scotland Bonds Mill Lifting Bridge near Gloucester

Figure 1-2. Fully composite bridges [20].

Due to the relatively low weight and high strength of FRP girders and decks it has proven

to be efficient to replace old bridges that no longer meet today’s requirements with FRP

alternatives [21-26]. While design concepts vary, several systems employ a modular

system to build up the bridge deck, meaning that composite profiles are transversely

joined to form the required bridge span or occasionally width [22]. Kumar et al. [23]

installed an all-composite bridge deck made up of FRP. The materials used for the

fabrication of this 9.144 m long by 2.743 m wide deck were 76.2 mm pultruded square

hollow glass and carbon FRP (GFRP and CFRP) tubes of varying lengths, Fig. 1-3.

These tubes were bonded using an epoxy adhesive and mechanically fastened using

screws in seven different layers to form the bridge deck with tubes running both

longitudinal and transverse to the traffic direction. The cross-section of the deck was in

the form of four identical I-beams running along the length of the bridge. They [23]

concluded that, all-composite bridge decks made of pultruded glass and carbon tubes are

judged to be a suitable replacement for short span bridges made of conventional materials.

In fact, the better fatigue performance of FRP was one of the main reasons suggested by

design engineers to replace conventional materials in some applications [24].

5

Figure 1-3. Experimental setup for the bridge deck test

One promising structural form is frames, where replacement of steel members is by

standard pultruded composite profiles of similar shapes [27,28]. The two most sensible

production techniques, pultrusion and filament winding, yield products with considerable

anisotropy in their properties. This results in considerable difficulties in making joints.

Drilling for bolts tends to sever the load-carrying fibers, and the lack of isotropy

precludes other load paths. Welding is impossible[20]?!!(is it), and although adhesives

can be used, the design of efficient and durable adhesive joints is not trivial. The most

cost-effective way to manufacture constant-cross-section composite profiles is pultrusion,

since this process is highly automated and uses low-cost forms of raw material. The

nature of the fiber reinforcement lends itself to novel forms. Pultrusions are difficult to

bend into conventional reinforcement shapes without seriously weakening the element at

the bend. So why not make use of the fibers’ flexibility to form the links before adding

resin to stiffen the material? There is research in Germany into the use of a variety of

textile processes, including the double-layer and Raschel knitting techniques (Figure 1-4)

which naturally give three dimensional structures [29]. The performance advantages of

three-dimensionally (3D) reinforced textile composites (as compared to traditional 2D

ones) have been well documented [30]. These advantages include full suppression of

delamination, lower notch sensitivity, and better fatigue and impact properties. 3D woven

6

composites can be easily machined into parts, very similar to metals, without considerable

effect on their properties.

Figure 1-4. Double-knit textile reinforcement [31]

. Just because FRP’s have good potential, they cannot be used in every structure [32-37].

One thing that needs to be considered is the beam-to-column connections. Mosalam [32]

concluded from an evaluation of available information that a majority of connections

research and manufacturers’ information was mimicking ‘steel-like’ details, i.e. they

copied the connection practice of steel frames, and that this wrongly ignored the effect of

difference in mechanical properties. The development of his Universal Connector (UC)

[33] was therefore presented as the first positive step in correcting this deficiency of

current connection design details. It has the conventional leg-angle shape of steel cleats,

but with side ribs to substantially increase its stiffness and strength. Appropriate

placement of the E-glass fiber reinforcement has been achieved by manufacturing the UC

with the resin transfer method. Cleated connections are the most favoured method for

pultruded frames due to the unsuitability of composite materials for conventional steel

endplate or welded joints [34,35]. Mottram and zheng [34,35] have tested both

connections consisting of FRP materials and those of steel. The conclusions from this test

were that the steel option worked better and was recommended in the article.

7

Although many FRP structures can be molded in single pieces to simplify the

manufacturing process and to reduce assembly coast, they cannot be completely exempted

from the joining process. Both mechanical fastening and adhesive bonding are commonly

used in structure assemblies. Each technique has its own advantages and disadvantages in

terms of function and cost. In terms of strength, different techniques also lead to different

damage modes and failure strengths. Based on the high in-plane strengths of FRP

materials, mechanical fastening is commonly used in assembling composite laminates.

However, mechanical fastening can cause high stress concentration due to structural

discontinuity. Adhesive bonding, although free from structural discontinuity, is primarily

based on the shearing strength of the adherend, the adhesive and the bonding between

them. These shearing strengths are not greater than the in-plane strengths and are limited

to the bonding surface, instead of through the thickness. As the joining components

become thicker, a much larger bonding area is required, not only to achieve a higher

joining strength but also to reduce the high stress concentration around the bonding ends.

Since mechanical fastening is actually a three-dimensional technique, it is superior to

two-dimensional adhesive bonding in joining thick composite laminates [36].

The main objective of this review article is to present a sound background regarding the

development of the understanding of the mechanical behavior and the failure modes of

the bolted, adhesive, and hybrid joints in polymeric composite materials. Section 2

presents a review of mechanically fastened joints in FRP including various modes of

failure. In Section 3, the review article discusses the bonded joints in FRP and dissimilar

materials. Section 4 presents a review of snap joints in FRP. Hybrid joints in composite

materials are introduced in section 5.

8

2. MECHANICALLY FASTENED JOINTS

Mechanical joints are used when repeated disassembly and reassembly is

required or when surface preparation is not practical. Mechanical joints require

that bolt or rivet holes are drilled into the composite, that reduced the net cross

sectional area of the structure and introduce localized stress concentration.

These stress concentrations can cause ply delamination since they will include

through thickness tensile and shear stresses. Mechanical joints add weight to

the structure from the added weight of the bolt or rivet. They also pose a risk

for corrosion since the laminate and fastener may comprise dissimilar materials

and moisture can be trapped in the crevices inherent in such joints. However

mechanical joints can be readily inspected before assembly and while in

service. Examples of two typical bolted joints are the single lap joint and

double strap joints as shown in Fig.2-1.

PiPiPi

P D W

e

Pt

Bolted Single Lap

P P

Axial tch

Bolted Double Strap

P D W

e

Pt

Bolted Single Lap

P D W

e

Pt

Bolted Single Lap

P P

Axial tch

Bolted Double Strap

P P

Axial tch

Bolted Double Strap

Figure 2-1. Basic types of mechanical joints.

The single lap joint is the simplest and most weight efficient but the load

results in a moment due to off-set load. The double lap joint eliminate the

moment but adds additional weight from the straps and additional bolt.

9

A circular hole in tensioned FRP plates may be classified into three types as

follows [38]: (1) open-hole tension: the FRP plates were subjected to uniaxial

tension with no constraint imposed on the hole, (2) filled-hole tension: a bolt, φ

= dB, was inserted inside the hole, φ = dH, with/without a clamp-up load. A

washer, φ = dW, was inserted between a bolt head and tail and the FRP plate to

distribute the clamp-up load, as shown in Fig. 2-2. The FRP plates were

subjected to uniaxial tension, and (3) bolted joint, bolt-loaded hole: double-lap

bolted joints were subjected to a uniaxial load with/without clamp-up load.

Dw

W

Load Cell

Rigid Washer

Washer Nut Bolt

Tab Tab

W

D

Specime

Load Cell

Strain Gauge Washer

(b) (a) (c)

Figure 2-2. Geometries of the (a, b) specimens for open- and filled-hole

tension tests, and (c) bolted joint test setup.

In the case of open-hole tension, the mode of failure mainly depends on the

fiber/matrix interface. A weak interface results in longitudinal crack

propagation along the interface, while a strong interface results in transverse

crack propagation across fibers leading to premature composite failure.

However, an interface of intermediate strength leads to optimum composite

performance between these extreme conditions. Based on the specimen

geometry and the interfacial strength of unidirectional FRP, the crack

emanating from notches, such as circular holes, may grow parallel to the

loading and fiber direction, i.e. notch insensitive [39]. A schematic description

of the damage pattern for a composite laminate is shown in Fig. 2-3. In the case

of Fiber Breakage & Matrix Cracking mode of failure, damage is localized to the

10

stress concentration areas before final failure. Edge Delamination mode of

failure may occur in both open- and filled-hole specimens in an early loading

stage. The edge delamination grew throughout the width of the open-hole

specimens. In addition, a Fiber-Matrix Splitting mode occurred along the zero

degree plies (in the direction of the applied load) emanating from the edge of

the hole [38].

Fiber Breakage Matrix Cracking

Fiber-Matrix Splitting

Delamination

Figure 2-3. Typical tensile failure modes in composite laminates containing a circular hole.

2.1 Stress Concentration Factor in Composite Materials Containing an

Open Circular Hole The susceptibility of composite materials due to effects of stress concentrations

such as those caused by notches, holes, etc., is much less than for metals.

Failure in long fiber-reinforced laminated composite structures containing

stress concentration areas, such as circular holes or bolt-loaded holes, has been

one of the technological issues by many researchers during the last decades

[38-42]. The mechanisms of such failures are significantly affected by fiber

orientation, relative strength of the matrix and fiber and the bond strength

between them. Open holes, notches, scratches, inclusions, and so on, all

produce concentrations of strain and stress. Stress concentrations can seriously

weaken brittle materials and can shorten the fatigue life of both ductile and

brittle materials. But a ductile material under static load can redistribute stress

by yielding without fracture. Thus, although strain concentration factor persists,

11

stress concentration factor (SCF) decreases markedly. The theoretical SCF for

an infinite orthotropic plate containing an open circular hole, , can be

calculated from the following relation [43]:

KT∞

12

1112

22

11T G

E υ

EE

2 1K +

−+=∞ (2-1)

Where E11, E22, ν12, and G12 are the elastic constants for an orthotropic plate,

i.e. E11 & E22, are longitudinal and transverse elastic modulus respectively, and

ν12, and G12 are the principal in plane Poisson’s ratio and shear modulus

respectively. It is worth to note that, in the case of isotropic plate, i.e. E = E11 =

E22, and G = G12 = E/2(1+ν), will be equal to the value obtained from the

will known relation = 1+2(D/ρ)

KT∞

KT∞ 0.5 = 3, where D and ρ are the depth and the

radius of the notch respectively. The inverse of isotropic finite width correction

(FWC) factors for open circular hole can be written, according to [44], as

3.

1

)1(2)1(3

. αα

−+−

=

=

∞−

IsoT

T

KK

FWCIso

(2-2)

Where α = dH/w, and KT is the finite SCF. The inverse of orthotropic FWC

factor equals [45]:

( )KK

FWC M K MT

T OrthTIso

∞− ∞

= + − −

.. ( ) ( ) ( )1 60 5 3 1α α 2 (2-3)

&

( )M

FWCIso=

− −−1 8 1 1

2

1

2.

( )α

− (2-4)

Finally, SCF based on net section instead of the whole width, KTn, can be

calculated as follows:

K KTn T= −( )1 α (2-5)

12

2.2 Failure Modes in Mechanically Fastened Joints

There are six basic failure modes in mechanically fastened joints in FRP

[46,38]: net-tension, shear-out, transverse splitting, cleavage, bearing, and pull-

through as shown in Fig. 2-4. The bearing stress, σb is the load, P divided by

the projected transverse cross sectional area of the hole σb = P /(dHt). The

shearout stress is determined by the longitudinal shear surfaces, and is given as

σSO = P /(2et). The net section stress is σN = P /[(W-dH)t]. The transverse

splitting stress is a localized stress normal to the applied load. The gross stress

defined as σ = P /[Wt] is used to rate the effectiveness of the joint. Joint

efficiency is the ratio of the gross section stress at failure to the strength of the

laminate in the gross section. For metals single fastener joints can have

efficiencies as high as 80%. Polymer matrix-fiber composite laminates have

efficiencies generally less than 50% due to their strength anisotropy and

inability to redistribute stress. Net-tension failure is associated with matrix

and fiber tension failures due to stress concentrations. Bearing failure leads to

an elongation of the hole. Shear-out failure can be regarded as a special case

of bearing failure. Shear-out and bearing failures result primarily from the

shear and compression failures of fiber and matrix. Cleave failures are

associated with both an inadequate end distance (e) and too few transverse

plies. Pull-through failure occurs mainly with countersunk fasteners or when

the plate thickness (t) to hole diameter (dH) ratio, t/dH, is sufficiently high to

precipitate failure. Li et al [47] found two others failure modes, i.e., bending

induced cross-section failure and rivet cap penetration failure. Further, they

[47] found that, failure modes may change with increasing loading rate and

rivet rotation decreases with increasing loading rate. Generally, the total energy

absorption of composite riveted joints increases with increasing loading rate,

except for joints used countersunk rivets. It was also shown that rivet rotation is

significant in some joint designs, which could damage the laminate and thus

reduce the joint strength.

13

(b) (d) (e) (c) (a) (f)

P P

Pull-out

P P P

Figure 2-4. Failure Modes (a) Net-Tension, (b) Shear-Out, (c) Transverse splitting,

(d) Cleavage, (e) Bearing, and (f) Pull-out.

Net section failures can be prevented by increasing the ratio of the plate width

to hole diameter, W/dH. Generally, W/dH > 6 is sufficient to prevent net section

failures. Shearout failures can be eliminated the ratio, e/dH to 3 or greater.

Transverse splitting is rare but will occur if there is a high fraction of the fibers

in the load direction such as would be the case in a unidirectional composite.

Bearing failure is the preferred failure mode since the joined members are not

catastrophically separated. Bearing failures are associated with localized hole

damage such as local delamination and matrix crazing. The experimental

observations of the effects of joint geometry, ply-orientation, lay-up, t/dH, dH to

plate width (w) ratio dH/w, e/dH, and clump-up load, i.e. through-thickness

pressure, on the joint behavior were reviewed by Camanho and Matthews [46].

The effect of e/dH on the SCFs of unidirectional CFRP single-bolt tension

joints with α = 0.2, i.e. w/dH = 5, was studied numerically by Sallam [48].

SCFs were calculated based on equivalent, shear, and longitudinal stresses. The

calculated SCFs based on equivalent stress are very high due to the high

compressive stress ahead of the loaded-bolt. SCFs decrease with increasing

e/dH up to e/dH equals three, as shown in Fig. 2-5. After that, the effect of e/dH

is insignificant.

14

1

11

21

31

41

0 2 4 6 8 10e/dH

Stre

ss c

once

ntra

tion

fact

or

GPaE11 = 165E22 = 11E33 = 11G12 = 5.3G13 = 5.3G23 = 3.9ν12 = 0.26ν13 = 0.5ν23 = 0.5

σEq.

τXY

σYY

Based on

α = 0.2

Figure 2-5. SCFs of unidirectional CFRP single-bolt tension joints.

Sallam [48] found through an experimental investigation that, the failure mode

of all tested bolt-loaded unidirectional CFRP plates is shear-out, regardless the

values of dH, dB, e, and w or their ratios. This is due to the weakness in the

bond between the fiber and matrix or in the shear strength of the matrix

compared with the tensile strength of the fiber. Load-bolt displacement

responses of all tested bolt-loaded unidirectional CFRP plates are shown in Fig.

2-6. The static failure is defined as the maximum load, Pult, achieved during the

test to failure [49]. As shown in the Fig. 2-6, the stiffness of the joints is not a

function of the values of dH, dB, e, and w, whereas, the ultimate load is

dependent on the contact area alone, i.e. the value of dB. This observation is

acceptable only in the case of shear-out failure in unidirectional composite

plate, i.e. lower stiffness and strength in transverse direction.

0

2

4

0 2 4 6Bolt displacement, mm

Load

, kN

Test # dH dB e w 1 7 6 10 20 2 10 6 11 30 3 10 6 25 30 4 10 6 25 50 5 13 6 17 50 6 13 6 25 50 7 10 8 30 30 8 10 8 30 50 9 13 8 25 50

Figure 2-6. Load-bolt displacement curves of unidirectional CFRP plates.

15

Aktas and Karakuzu [50] found experimentally and numerically that, the

failure mode of the unidirectional composite pinned joint is shear-out except

for the cases of e/dH and w/dH ratios of 4 and 5 respectively. For those values

the mode is bearing mode. When failure caused by the pin continues up to 2–3

mm from the free end of the plate in the direction of loading, the plate tears

immediately. Net-tension mode occurs in the following situations: fiber

orientation ranged between 30o and 60o, w/dH = 2 –3, and e/dH = 4. Although

the failure mode is the same, the propagation of failure is a bit different for

each criterion, because the angle of first failure area is different in each case, as

shown in Fig. 2-7 .

(a) (b) Figure 2-7. Ultimate failure of specimen for (a) E/D, and (b) W/D series.

Furthermore, Karakuzu and his co-workers [51,52] studied the effect of joint

geometry and ply orientation on failure strength and failure mode of composite

laminates with a pin-loaded hole. They concluded that, there is a definite

dependence of bearing strength on stacking sequence and joint geometry. The

lay-up [0/±45]s laminate was found to have higher bearing strength than the

lay-up [90/±45]s laminate. Ultimate strengths are sensitive to w/dH and e/dH

values in a larger range compared to the laminate tensile strength. Increasing

the end distance increased the bearing strength of the joint until a critical end

16

distance was reached; any increase of the end distance beyond that value did

not result in a corresponding increase in the strength of the joint. Pin bearing

strength decreases with decreasing w/dH ratio. As the width of the specimen

decreases, there is a value where the failure changes from the bearing to net-

tension failure mode. Maximum bearing strength is reached when the edge

distance ratio, e/dH, is equal to or greater than 3, and the side distance ratio,

w/dH, is equal to or greater than 4 for the [90/±45]s laminate. For the [0/±45]s

laminate, the critical e/dH and w/dH ratios are 3. In net-tension and shear-out

failure, the reduction in load is both greater and more sudden. But in bearing

failure load doesn’t drop suddenly. Therefore, joints that fail in the bearing

mode are stronger and safer joints although this depends on selected design

rule. The net-tension strength of a single-hole joint is dependent on ply

orientation and specimen width. The effect of a change in the ratio of specimen

width to hole diameter is least for [90/±45]s laminate and most marked for

[0/±45]s laminate. The shear strength of single-hole joints has been shown to

be strongly dependent on the ply orientations within the laminate and the edge

distance. For the [0/±45]s laminate, the maximum shear stress at failure is

reduced to 50% with increasing edge distance to diameter ratio. Whereas for

[90/±45]s the effect of a change in the ratio of edge distance to diameter is of

about 66% loss in strengths. It is evident from the load–displacement curves

that the [(±45)3]S laminate with small w/dH fails in a more sudden fashion than

the [(0/90)3]S laminate. For this reason, the use of mechanically fastened joints

in [(±45)3]S laminates with small w/dH is not recommended. Turvey [53] found

that, both the initial stiffness and the ultimate load depend on the w/dH and e/dH

for pultruded GRP plate. It is worth to mention that, the reinforcement of

pultruded GRP plate is in two forms, viz. E-glass roving (unidirectional fiber

bundles) and Continuous Filament Mat (CFM). The CFM provides the

transverse stiffness and strength. Further, Turvey found that the failure mode

changed from bearing to cleavage to shear-out with decreasing e/dH with

constant w/dH. However, the failure mode changed from bearing to tension

with decreasing w/dH with constant e/dH. In drilling operations, high axial

17

forces are induced locally in the material at the tip of the drill because of the

stationary tool center. In composite materials, the high axial contact force

between the drill and the composite may cause separation of piles at the exit.

This effect can also be observed when sawing with a core drill. At the entrance

of the hole, delamination may occur if the cutting angle is too small and the

cutting edge is wedged between two laminates [54], Fig. 2-8. Zachrisson et al.

[55] developed a new method (KTH-method) which gives defect free holes.

Failure analysis of glass woven roving composites of bolted connections with

drilled and moulded-in holes were examined by Lin & Tsai [56]. They found

that, laminates with moulded-in holes are stronger when the (e/dH =1).

Lamina

Interface (a) (b)

Figure 2-8. Delamination at (a) the exit and (b) the entrance sides of the hole.

The influence of bolt spacing (i.e. pitch distance, row spacing, end distance,

and bolt diameter) and the degree of laminate anisotropy on the bolt load

distribution and failure prediction in multi-fastened composite joint was

investigated numerically by Sergeeva et al.[57] and Chutima and Blackie [58].

Failure loads for two laminate lay-ups [10/60/30] and [25/50/25], with fastener

diameter specified as D= 8 mm and varying spacing, S, are shown in Fig. 2-9.

There is no notable interaction among the stress concentrators for the case of

distantly placed fasteners, and the relationship between bolt spacing and failure

load is approximately linear. As the fastener positions are closer to each other

and the high stress areas near the fastener holes merge, the joint strength starts

to decrease. The results indicate that the optimal bolt spacing may be different

18

for different laminate lay-ups: in the case of the quasi-isotropic lay-up

[25/50/25], the maximum joint strength is achieved at S/D = 3, while for the

[10/60/30] laminate the optimal fastener spacing corresponds to S/D=3.5.

Figure 2-9. Variation of the failure load for different fastener spacing [57].

The effect of edge geometry (i.e. round or rectangular, see Fig. 2-10), stacking

sequence, and clamping force on failure modes in mechanically fastened joints

in FRP was studied by several researchers [59,60] The lateral clamping

pressure suppresses the onset of delamination and continuously suppresses the

propagation of interlaminar cracks. The failure mode changes from a

catastrophic mode to a progressive one. Consequently, the lateral clamping

pressure increases both the delamination and ultimate failure strengths of

bolted joints in composite materials [59].

(a) (b) (c)

Figure 2-10. Types of the failure modes: (A) Net-tension failure mode

(B) Shear-out failure mode (C) Bearing failure mode [60].

19

It was found that, either bolt-to-washer or bolt-to-hole clearance can affect the

initial failure loads but not the ultimate failure loads [61, 62]. Starikov and

Schon [63] found that the amount of load carried by the part with ordinary

holes compared to the part with countersunk holes increased during fatigue

loading. The contact stresses in the cylindrical part of the holes are probably

larger in the plate with countersunk holes than in the part with ordinary holes

since the cylindrical part of the hole is less for countersunk holes than for

ordinary holes. This probably results in more fatigue damage in countersunk

holes than in ordinary holes and the applied load is redistributed to the plate

with ordinary holes. Figure 2-11 shows joints with protruding head bolts and

countersunk head bolts that failed by bolt failure [64]. In all cases of bolt

failure, the bolt failed at the threads. It can be seen that extensive bearing

damage existed at the hole before the bolt failed. The bearing damage at the

shear plane appears more extensive in the countersunk case, which is most

likely due to the bearing load being taken almost entirely by the cylindrical

portion of the hole in the countersunk laminate.

(a) protruding head bolt

(b) countersunk head bolt

Figure 2-11. Final failure of protruding head and countersunk head bolts.

20

3. BONDED JOINTS

The key advantage of adhesive bonded joints over other joint approaches, e.g.

mechanical fasteners (bolting or riveting), is that it enables the development of

large, cost-effective, and highly integrated structures [37]. Adhesively bonded

joints have high structural efficiency and are used extensively to join

composites in advanced aerospace structures. Bonded joints can be between

two composite laminates or between a composite laminate and a metal/concrete

structure. Adhesively bonded joints can distribute load over a much wider area

than mechanical joints. Since no holes are required, the risk of local

delamination is practically eliminated. Compared to bolted joints with weight

of the joint is significantly reduced. On the other hand adhesively bonded joints

cannot be disassembled without destroying the substrate. Some adhesives are

susceptible to degradation by temperature and humidity. The most critical

drawback for safety critical structure such as airframes is their inspectability.

Critical joints may require ultrasonic inspection over their entire area.

Corrosion can be problem in carbon fiber composite to steel/aluminum joints

due to galvanic action. In such cases an intermediate insulating layer can be

used. Some of the most common types of adhesively bonded joints are

illustrated in Fig. 3-1. In addition to the lap and strap joints that are used in

bolted joints additional types are possible with adhesive bonding such as the

stepped joints and scarf joints [3,65,66].

Single-lap joints are efficient at transferring in-plane shear, but they should not

be used for compression loads unless the joint is stabilized because the

eccentricity increases as the load increases in compression. Adhesively bonded

doublers transfer load through the adhesive in the same manner as other

structural joints. The adhesive between the flanges of a stiffener and a skin is

loaded similarly to a bonded doubler case. The single-strap butt joint has two

disadvantages caused by the eccentricity in which the members butt together.

These disadvantages are (1) high bending moment in the splice plate and (2)

high adhesive peel stress, which may lead to either premature failure of the

21

Single lap

Double lap

Double strap

Double scarf

Single scarf

Double stepped lap

Stepped lap

Figure 3-1. Common adhesive joint configurations.

joint or delamination of the composite members. Thick adherends in single-lap

joints should have the overlap ends tapered. This reduces adhesive peel stress

so that the adhesive shear strength can be fully developed. The taper also

reduces the magnitude of the peak in the shear stress distribution, and when

both overlap ends are tapered, the strength of the joint increases.

Double-lap and double-strap joints are balanced stiffness designs: that is, the

stiffness of a strap or splice sheet should be equal or slightly higher than one-

half of the central sheet stiffness, and the two straps have equal stiffness.

Again, these straps and splice plates should be tapered at the ends to reduce the

adhesive peel stress. A maximum taper tip thickness of 0.5 mm for the splice

22

sheets. Another detail that reduces the adhesive peel stress and the peak shear

stress is locally thickening of the adhesive at the tips of the splice plates and

doublers. The stress condition at the end of the central sheet remains constant

with these modifications, but it is possible to increase the adhesive thickness at

this point by slightly tapering the end of the central sheet. Tension loads on the

joints do not generate adhesive peel stresses at the end of the central sheet

because these are compression stresses.

The stepped-lap bonded composite joint strength is improved when the

adherend stiffness is balanced, that is, a constant stiffness joint. The tip and end

thicknesses should be limited. The length and number of steps are additional

design variables, and the number of steps has the most effect on joint strength.

Step-lap joints have been used to bond composite to metal adhesively where

high magnitude concentrated loads are transferred, and the composite cannot

withstand the high local stresses associated with mechanical fasteners.

Adhesively bonded stepped-lap joints of composite to metal can be found in the

tails of the F-14, F-15, F-16, and F-18 and in the wings of F-18 aircraft. Scarf

joints are usually considered uniformly strained. In practice, however, the tips

of the adherends have a finite thickness and they must be analyzed as an

approximation to a stepped-lap joint with a fine grid over the length of the

overlap. In scarf joints that have adherends with different stiffnesses. there is a

tendency for the thin tip of the stiffer adherend to fail in fatigue [16].

Adhesively bonded composite joints have the following three basic failure

modes:

1. The failure is in the adherend outside of the joint, which fails at 100% of the adherend strength. This is the strongest joint. There is no adhesive failure.

2. The failure is caused by the shear strength of either the adhesive or the

laminate/adhesive interface. This is the next strongest joint.

3. The failure is caused by peel loads either as failure of the adhesive or as delamination of the adherend.

23

For the first failure mode, the strength of the joint is proportional to the

adherend thickness and no bending of the adherend is assumed. The loading

can be tension, compression, or in-plane shear. For the second failure mode, the

shear strength of the bonded joint is proportional to the square root of the

laminate thickness. For the last failure mode, the peel strength is proportional

to the quarter-power of the laminate thickness.

New joint designs were proposed for adhesive bonding of thick multi-layered

composite adherends by Bahei-El-Din and Dvorak [67] to reduce or eliminate

the failure modes associated with delamination and tensile and/or shear failure

of the surface plies that are often observed in lap joints, and provide for a better

stress distribution in the adhesive. In contrast to lap-joint designs, which

transfer in-plane tensile stresses and other loads from the adherends to doubler

plates by out-of-plane shearing of the surface plies, the new joint

configurations transfer most of the load by in-plane shear and normal stresses,

through bonded inserts or interlocking interfaces which have the same

thickness as the laminate adherends, as shown in Fig. 3-2. Doublers will

transfer a calculated percentage of the load.

b=15

bδ=

15 c,g d,h

b,f X1A

l2

3tδ d=9

l1 l=70δ=0.2 3td

Dimensions in mm X3

tA=30

td=3

td=3

Section A-A

X1

a b

fe

A

Figure 3-2. New design with diamond-shaped inserts [67]

24

Avila and Bueno [37] addressed the advantages and disadvantages of the wavy-

lap joint proposed by Zeng and Sun [68], Fig. 3-3, and they modified it. The

results showed an average increase on loading to failure close to 41%. This fact

could be due to the compressive stress field developed inside the wavy-lap

joint. In addition, this stress field distribution can also be the reason for the

adherent delamination observed on the wavy-lap joints. So far, the modiffed

wavy-lap joint seems to lead to stronger joints.

12o

24.6

227.8

8.40

101.6

6.15 6.15 Tab

12o Adhesive

25.40

2.80

Figure 3-3. Wavy-lap joint main dimensions.

3. 1. Stresses in Adhesive Joints

The single lap joint, in which two sheets are joined together with an overlay, is

one of the most common joints encountered in practice. The joint is easy to

make and the results are sensitive to both adhesive quality and adherend

surface preparation. The simplest analysis considers the adherends to be rigid

and the adhesive to deform only in shear. If the width of the joint is b, the

length λ, and the load P, then the shear stress τ is given by: τ = P/(bλ).

However, the adherend tensile stress will decrease linearly to zero over the

joint length from A to B, as shown in Fig. 3-4(a). In Fig. 3-4(b) is shown a

similar joint but in which the adherends are now elastic (i.e. deformable). For

the upper adherend، the tensile stress is a maximum at A and falls to zero at B.

Thus, the tensile strain at A is larger than that at B and this strain must

progressively reduce over the length λ. The converse is true for the lower

adherend. Thus, assuming continuity of the adhesive/adherend interface, the

uniformly sheared parallelograms of adhesive shown in Fig. 3-4(a) become

25

distorted to the shapes given in Fig. 3-4(b). This phenomenon is called

differential shear. Essentially, this is what Volkersen analyzed in 1938 (the

first one proposed a simple shear lag model based on the assumption of one-

dimensional bar-like adherends, i.e. the adhesive deforms only in shear, while

the adherend deforms only in tension) [see Refs. 3 and 69]. The linear elastic

analysis by others and the elasto-plastic analysis (Hart-Smith’s solution)

including the end effects of lap joints are reviewed in Ref. 3. There are many

different forms of structural joint, but most transmit essentially collinear loads

such as the lap joints. More complex configurations exist, such as corner or T-

joints, but these are very difficult to analyze.

(a) L

B A P

P

x

x

τ

B A

L

P

(b)

τ

P

Figure 3-4. Exaggerated deformations in loaded single-lap joint:

(a) with rigid adherends; (b) with elastic adherends.

In general, failure takes place in the adhesive (called cohesive failure) rather

than between the adhesive and the adherend or substrate (called adhesive

failure). Adhesive failure is often due to environmental degradation and cannot

be generally analyzed, although it is possible to treat specific cases, providing

simplifying assumptions are made. Peel loads are the greatest enemy of the

26

designer of bonded joints. Wherever possible, the adhesive should be loaded in

shear so that peel and cleavage stresses are avoided. Increasing the width of lap

or peel joints increases the strength pro rata, whereas increasing the length is

beneficial only for very short overlaps. However, the benefits to be gained from

having a large area of lightly stressed material in the middle of the joint,

especially when creep and fatigue and faulty manufacture need to be taken into

account. Figure 3-5 shows various ways of joining sheet loaded in tension. The

“loading'” rating is indicative of how well the joint will withstand the applied

load and the “cost” is an estimate of the assembly, cutting and purchasing

costs. Great accuracy is not claimed for the numbers quoted as these will vary

with adherend material and thickness and the choice of adhesive.

Figure 3-5. Joining of sheet materials in tension; the higher the loadrating, the stronger is the joint.

10 8

2 2

10 4

12 4

10 2

7 3

15 7

15 1

10 2

1 1

Rating Load Cost

27

3.2 Standard Mechanical Test Procedures

Since adhesives were first used, there has always been a need to define a series

of tests which can be carried out in order to quantify (or at least qualify) their

suitability in technology. Various tests have been proposed and some are still in

use today for structural adhesives which have been developed from the fabric

and wood-working industries. The wide variety of standard test procedures as

listed by the International Standards Organization (ISO), European Standards

(EN), American Society for Testing and Materials (ASTM), British Standards

Institution (BSI) and other organizations are essentially for testing adhesives

and surface treatments rather than joints. Unfortunately, most if not all of these

standard tests consist of joints in which the adhesive stresses are far from

uniform [3]. Let us therefore examine the reasons why we might need to carry

out any form of test on an adhesive, other than the not unworthy cause of

curiosity. The most commonly used test is the single-lap joint illustrated in

Fig. 3-6(a). These dimensions are as specified by ASTM D 1002 (very similar

to ISO 4587) which also specifies the adherend materials. It is recommended

that the specimens be cut from a 177mm wide bonded plate since this gives the

most representative results. The outer strips should be discarded. Note that this

joint is automatically misaligned before it is placed in the testing machine.

Some laboratories bond tabs at the ends to improve alignment, as shown in

Fig. 3-6(b). Even so, under load the joint will bend as shown in Fig. 3-6(c),

giving rise to large transverse peel stresses in the adhesive layer. As known, the

adhesive shear stress is nonuniform owing to differential straining in the

adherend and other factors: this is illustrated schematically in Fig. 3-6(d)

together with the associated adhesive peel (transverse) stresses in Fig. 3-6(e).

Even though it is still recommended in ASTM D 1002 and similar standards

that the results be given as the average shear stress at failure (i.e. load divided

by bond area), it has long been recognized that this average shear stress bears

little relationship to what is actually happening in a joint, especially when

geometric, adhesive and adherend nonlinearities become significant.

28

Figure 3-6. Single-lap joint test piece to ASTM D 1002 (dimensions in mm).

Transverse adhesive stress (acting across the bond-line thickness)

(e)

F F

(c)

Average shear Stress = F / Area

(d)

Adhesive Shear stress

F F

(b) Alignment tabs

F F

101.6 63.5 12.7±0.25

25.4

(a) Single-lap joint

Grib area

1.62±0.125 F

F

29

The advantages of the single-lap test are that it is simple, cheap, uses a standard

tensile testing machine, and there are a lot of data available for comparison. Its

main disadvantage is that the reported average shear stress is not an intrinsic

adhesive property. Curiously, this very disadvantage is also an advantage since

no-one really believes that the average shear stress means anything

fundamental. In other tests with apparently more precisely controlled

conditions, it is common for misleadingly definite values to be quoted. Also,

the complex stress situation in the single-lap test makes it quite representative

of many structural applications and loading situations. This test is widely used,

often abused, but remains one of the most trusted standards. Nevertheless,

there have been many attempts to improve the single-lap test. These include the

laminated assembly (ASTM D 3165) and sometimes wrongly referred to as a

double-lap joint.

3.3. Fracture Mechanics Considerations

Adhesive joint studies proposed in the literature usually fall into two classes:

those based on stress analysis and those based on fracture mechanics. Fracture

mechanics has become a very popular tool for the characterization of adhesive

joints in recent years. Fracture mechanics tests are routinely conducted by

industry during materials development and have also found extensive

application in fatigue and durability studies over the past 20 years. More

recently, fracture mechanics data has been used to predict the impact failure

response of, for example, the impact wedge peel test and currently fracture

mechanics data are finding application in structural impact studies via the use

of cohesive zone models. The use of Linear Elastic Fracture Mechanics

(LEFM) tests to measure the mode I adhesive fracture energy, GIC; of adhesive

joints dates back to the work of Ripling and co-workers in the 1960s who

developed a mode I test method to measure the toughness of structural bonds

between metallic substrates. This work led to the publication of an ASTM

standard in 1973 [see Ref 70]. Since then, the many developments in the

application of fracture mechanics to, for example, FRP composites has created

30

great potential for the development of a new test protocol for structural

adhesive joints. It was against this background that a technical committee of

the European Structural Integrity Society (ESIS) commenced work on

structural adhesives test methods in 1997. The new protocol accommodated the

use of both metallic and FRP composite substrates. During the course of the

wide ranging program involving ten test laboratories, modifications were made

to both the experimental and analytical procedures compared to the original

ASTM standard of 1973. For example, a new corrected beam theory analysis

for the tapered double cantilever beam was developed and a correction for

system compliance and additional validity checks were built into the

experimental procedure. Following these modifications, the revised protocol

was submitted to the British Standards Institution for consideration as a British

Standard (under the direction of the ‘‘Adhesives Standards Policy Committee

PRI/52’’) and was accepted and subsequently published in 2001. It is intended

that this document should also be published as a European standard. Blackman

et al. [70] described the stages in the development of this new protocol and the

modifications made in the light of the results from the round-robin tests. The

results from the inter-laboratory round-robin highlighted the importance of

correcting for system compliance effects if accurate and reproducible results

are to be obtained. The values of GIC deduced were shown to be independent of

test geometry but dependent upon the substrate material used to make the

joints. Additional studies have shown that the substrate dependence was due to

the cured adhesive in the different joints possessing different values of glass

transition temperature. The existence of pre-bond moisture in the CFRP

substrates and variations in heat-up rate during cure were both shown to affect

the Tg of the cured epoxy-paste adhesive employed in the present work. The

pre-bond moisture effect was however, much more important than the heat-up

rate effect but both would need to be considered when optimizing joints for

toughness with the present adhesive.

31

There are three types of delamination test to obtain pure mode I loading, pure

mode II, and mixed mode I/II loading [71]. For the mode I Double Cantilever

Beam (DCB), Fig. 3-7(a), if it is assumed that the compliance CI is a function

of (a + |∆|)3 where “a” is the crack length and ∆ is a correction parameter, a

first strain energy release rate expression is

)∆2b(a3PδG I +

= (3-1)

where P is the applied load, δ the crack opening displacement, b the specimen

width. ∆ is determined from the plot of C1/3 versus a. Another compliance

expression is the following:

C = k an (3-2)

where n is the slope of the ln(C) versus ln(a) curve. Therefore, the

energy release rate expression becomes

2abnPδGI = (3-3)

For the mode II End Notched Flexure (ENF) test, Fig. 3-7(b), a first energy

release rate formulation comes from beam theory:

f32

2

II Eh16bP9aG = (3-4)

where P is the applied load, Ef is the flexural modulus such that

o3

3

f C4bhLE = (3-5)

32

in which C0 is the compliance obtained experimentally from a three-point

bending test in the elastic range on the uncracked specimen. Another

expression involving � may be used:

)32(29

33

2

aLbPaGII +

=δ (3-6)

If we assume that the compliance determined experimentally can be written as

C = C0 + ma3 (3-7)

the energy release rate is expressed in the form

2bP3maG

22

II = (3-8)

The Mixed Mode Bending (MMB) test, Fig. 3-7(c), is the superposition of the

two previous tests, and it allows any combination of mode I and mode II

loadings to be obtained. It consists of a three point bending test, in which the

load P is applied by means of a lever with a central fulcrum, at a distance c

from this fulcrum. The load bends the MMB specimen at the fulcrum and at the

same time pulls the delamination open, which gives a combination of mode II

(bending) and mode I (crack opening) loadings. The mixed mode ratio I/II can

be modified by changing the lever length c. An MMB test analysis is based on

beam theory. The expressions for mode I and mode II energy release rate

components are

Eh16bP9aG

,EhbP12aG

32

2II

2

II

32

2I

2

I

=

= (3-9)

33

PI and PII are the pure mode loading components:

P,L

LcP P,4L

L3cP III

+

=

−

= (3-10)

where c is the lever length, and ¸ is half of the distance between lower loading

points. The mixed mode ratio is defined by

L/3c ,LcL3c

34

GG 2

II

I ≥

+−

= (3-11)

It is independent of the crack length a and it only depends on the lever length c.

Figure 3-7. Test configuration of three different types of delamination test.

L c P

2h

a

L L

(c) MMB test configuration

(a) DCB test configuration.

δ

a P

P

(b) ENF test configuration.

δ L L

P

a

34

3.4. Fatigue behavior and Durability of Adhesive lap joints

The response of bonded joints to fatigue loading has been extensively

researched in the last years concluding that the failure generally initiates at the

adhesive [72,73] for moderate stress levels. The fatigue life is strongly

influenced by the profile of the edges of the joint. The initiation of small cracks

at the edges points represent the major part of the fatigue life. The static and

fatigue properties of polymeric adhesives can be influenced by centered defects

[73,74] and environmental conditions such as: temperature, moisture and

chemical agents [72]. It has been shown that thermal effects, whether due to

mismatch of the adherends or to adhesive contraction by temperature of cure,

lead to significant changes in the stress state of lap joints. The temperature Tg

of the adhesive should be above the maximum temperature expected in service.

It has been also verified that the strength of joints usually decreases in presence

of humidity and time of exposure. Also the debonding of adhesives caused by

cleavage of cracks increases with the water presence in form of liquid or vapor.

This is the result of the hydrophilic nature of adhesives, which is caused by the

polar groups, needed to confer adhesive properties on polymeric materials.

Therefore chemical degradation of the adhesive, substrate and chemical bonds

across the interface is possible as a result of interaction with water. The water

can enter in the adhesive and then attack it by diffusion through the adhesive

and adherent, and finally transport along the interface and move by capillary

action through cracks in the adhesive. The moisture can not only affect the

adhesive but also the mechanical behavior of the adherend composite. However

no significant effect of moisture was observed on the static strength and fatigue

crack growth of laminate composites.

3.5 Prevention of Peeling failure in Adhesive Joints

Where peel is encountered with lap joints, or even to counter the peel loads

inherent in loading even in the double-lap joint, various techniques can be

used; these are illustrated in Fig. 3-8. Of these, the best is probably the positive

constraint of the joint end by riveting, bolting or spot welding [3]. However,

35

the arrangement shown in Fig. 3-8(d) is also common and is used in the

construction of nearly all motor car doors, the joint being called a clinch or

hem-flange join. Where lap joints are subjected to bending moments a rivet,

bolt or spot weld near the end A, where the joint tends to open in peel, should

be used if possible, as shown in Fig. 3-9(a). At the other end B, the adhesive is

in compression and failure will not occur there. Alternatively, an overhanging

reinforcement plate may be used, as shown in Fig. 3-9(b). This has the

advantages of reducing the bending strain in the upper adherend, thus

strengthening the joint, and transferring the load further into the joint [3].

Rivet, bolt or spot weld

Increase stiffness

(a)

Increase area

(b)

Bead end (if possible)

Peel action

(c) (d)

Figure 3-8. Techniques for combating peel.

36

Reinforcement

(b) A

B

(a) A

B

Figure 3-9. Combating bending loads in lap joints using (a) a rivet, and (b) overhanging reinforcement.

3.6 Adhesive Bonded Dissimilar Materials

The use of adhesive bonded structure is significantly increasing recent years. It

is well known that mechanical properties of the bonded structure depend on the

surface conditions of the substrates. Accordingly, many studies have been

made on the surface treatment of polymer, ceramic, and metallic substrates to

improve the mechanical properties of FRP/metal, FRP/concrete, steel/concrete,

metal/metal and metal/ceramic bonded structures [75-84]. One appealing way

to bond composite parts to steel structures is to mold a steel edge into the

composite, and then weld this assembly to the steel structure via this edge. This

requires that the steel can successfully be molded into the composite during the

manufacturing. The strength of these joints can be increased [76] by using

perforated steel, as patented by Unden and Ridder [77]. The perforations will

decrease the elastic mismatch between the stiff steel and the relatively

compliant FRP, and as an added benefit provide mechanical interlocking.

37

Co-cured joining method [78], which is regarded as an adhesively bonded

joining method, is an efficient joining technique because both the curing and

joining processes for the composite structures can be achieved simultaneously.

The co-cured joining method requires neither an adhesive nor a surface

treatment of the composite adherend because the excess resin, which is

extracted from composite materials during consolidation, accomplishes the co-

cured joining process. Since the adhesive of a co-cured joint is the same

material as the resin of the composite adherend, the analysis and design of the

co-cured joint for composite structures are simpler than those of an adhesively

bonded joint, which uses an additional adhesive.

The manufacturing process of the specimens of co-cured single and double lap

joints of the plate type with steel and composite adherends was introduced by

Shin et al. [78]. They found that, the initial failure mechanism of the co-cured

single lap joint was interfacial failure between steel and composite adherends

and the failure mechanism of the co-cured double lap joint was cohesive failure

by delamination at the first ply of the composite adherend. Out-of-plane tensile

and shear stresses played an important role of the interfacial failure of the co-

cured single lap joint and only an out-of-plane shear stress played a most

important role of the cohesive failure of the co-cured double lap joint.

Dvorak et al. [79] explored a new approach for the joining of thick, woven E-

glass/vinyl-ester composite laminated plates to steel or other composite plates,

with applications in naval ship structures. Adhesive is applied along through-

thickness contoured interfaces, employing tongue-and-groove geometry. They

found that adhesively bonded tongue-and-groove joints between steel and

composite plates loaded in monotonically increasing longitudinal tension are

stronger than conventional strap joints even in relatively thin plates.

Melogranaa et al. [80] studied tongue-and-groove geometries, see Fig. 3-10, in

thin composite laminated plates. Adherends were comprised of 2.7 mm thick

38

stainless steel and 1.6 mm thick T700/SE84HT carbon fiber/epoxy composites.

The adhesive used was the two-component epoxy paste Hysol 9430. They

compared their thin tongue-and-groove joints to conventional single lap joints

manufactured with the same materials and similar dimensions. The tongue-and-

groove specimens with large aspect ratio tongues were stronger than the

conventional single lap joints.

Wgroove

Figure 3-10. The general geometric parameters for a tongue-and-groove joint.

Rmax

L=25.4[1.0]

tg=0.1 [0.039]

w=2

5.4

[1.0

]

Rtip

T700 carbon/epoxy AL–6XN steel

So far, FRP has mostly been used for strengthening concrete structures [21].

Perhaps the single most critical parameter in strengthening reinforced concrete

with externally bonded FRP/steel plate is the bond achieved between the

reinforcing FRP/steel and concrete, which is responsible for the mechanism of

stress transfer and therefore the composite behavior. Debonding of the

FRP/steel plate from the concrete surface can significantly limit the potential

flexural and shear strength enhancement provided by the FRP/steel to the

repaired structure[81-84]. A brief review of this topic was presented by Nehdi

et al. [81] and De Lorenzis et al. [84]. Figure 3-11 summarizes some of the

experimental techniques used in the literature for testing the bond of FRP plates

to concrete. These tests often involve direct tension or bending and therefore

require special handling.

39

Figure 3-11. Various techniques for testing bond of FRP plates to concrete.

3.7 Fusion Bonding for Joining Thermoplastic Matrix Composites

An ideal structure would be designed without joints, since joints are potentially

sources of weakness and additional weight. In practice however, upper limit to

component size is generally determined by the manufacturing processes.

Further requirements for inspection, accessibility, repair and transportation or

assembly mean that load-bearing joints will be part of an engineering structure.

This is particularly so in the manufacturing of thermoplastic composites (TPC)

for which high melt resin viscosity and constraints imposed by the continuous

reinforcement limit the production to relatively simple geometry components

which must be joined together to produce large, complex structures. The

extensive experience available from the thermoplastic polymer (TP) industry

showed that to make large or complex parts, the most cost-effective method

often involves molding two or more parts and joining them together.

Thermosetting (TS) adhesive bonding is inherently preferable to mechanical

40

fastening because of the continuous connection avoiding large stress

concentrations induced at each discrete fastener hole. However, there are a

wide range of contaminants present on substrate surfaces [85]. These need to

be eliminated prior to bonding by using a surface preparation treatment and

increase the surface roughness (improving mechanical interlocking and

increasing bonding surface area). Extensive surface preparation and long curing

times make adhesive bonding labor intensive. Typical surface treatments used

for adhesive bonding are generally hard to control and affect directly the

strength and durability of bonded joints.

Fusion bonding, or welding, is a long established technology in the

thermoplastic industry where the efficiency of the welded joint can approach

the bulk properties of the adherends [85]. Although welding may induce

residual stresses if performed without adequate control, it eliminates the stress

concentrations created by holes required for mechanical fasteners and so does

TS adhesive bonding. In addition, welding reduces processing times and

surface preparation requirements. However, the high content in carbon fiber

(CF) reinforcement in TPCs, resulting in high thermal and electrical

conductivity, imposes difficulties such as uneven heating, delamination and

distortion of the laminates. These problems become more difficult when

bonding large components. In addition, as fiber volume fraction increases, the

amount of resin available to melt and reconsolidate into a fused joint is reduced

and this can affect the welding quality.

Fusion bonding techniques have often been classified according to the

technology used for introducing heat. In this review article, fusion bonding

techniques were classified into four classes, as shown in Fig. 3-12, namely bulk

heating (co-consolidation, hot melt adhesives, dual resin bonding), frictional

heating (spin welding, vibration welding, ultrasonic welding), electromagnetic

heating (induction welding, microwave heating, dielectric heating, resistance

welding) and two-stage techniques (hot plate welding, hot gas welding, radiant

41

welding). Bulk heating techniques such as autoclaving, compression molding

or diaphragm forming are available for performing co-consolidation.

Co-consolidation is an ideal joining method as no weight is added to the final

structure, no foreign material is introduced at the bondline, essentially no

surface preparation is required and the bond strength is potentially equal to that

of the parent laminate. However the entire part is brought to the melt

temperature, and this generally implies the need for complex tooling to

maintain pressure on the entire part and to prevent de-consolidation. Hot melt

thermoplastic adhesive films may be inserted at the bondline to improve filling

of parts mismatch. Inserting of an amorphous polymer interlayer proved to

reduce the scatter of strength, which widens the processing window.

The dual resin bonding, or amorphous bonding, involves co-molding an

amorphous TP film to a semi-crystalline TP matrix laminate prior to bonding as

in the Thermabond® process, in which a PEI film is coated onto APC-2

laminates. During the joining step, the amorphous PEI film can be fused at a

temperature above its glass transition temperature below the melting

temperature of the semi-crystalline PEEK polymer avoiding any deterioration

of the bonded structure.

In two-stage techniques the heating device needs to be removed from between

the substrate surfaces between the stages of heating and forging. This aspect

involves limitations on size of the component since the whole welding surface

must be heated in a single step. Heating times are normally long as they rely on

the low thermal conduction of heat through the polymer. Between the heating

and forging steps, surface temperature drops and the region experiencing the

maximum temperature is located below the skin of the laminate. The high

pressure required to consolidate the bondline may cause warpage/flow in the

higher temperature inner region.

42

Spin welding and vibration welding have been extensively used in the plastics

industry but are less appropriate to joining TPCs as the motion of the substrates

relative to one another may cause deterioration of the microstructure, such as

fiber breaking. The process was however investigated for joining APC-2 and

GF/PP systems.

Microwave and dielectric welding are available for joining thermoplastics but

the fact that heating occurs volumetrically and that multi-layer composites are

excellent shields in the microwave range make these techniques poorly suitable

to welding of TPCs particularly when they are reinforced by CFs. The three

most promising fusion bonding techniques are ultrasonic welding, induction

welding and resistance welding. In these techniques, only the welding interface

is brought to the melt temperature, minimizing the impact on the rest of the

structure. Welding times are very short. Large-scale welding may be performed

through sequential or scanning approaches, and on-line monitoring of the

consolidation is possible.

Bulk Heating

Co-consolidation Hot-melt Adhesives Dual Resin Bonding

Frictional Heating Electromagnetic Heating

Hot Plate Welding Hot Gas Welding Radiant Welding

Infrared Welding Focused Infrared Welding Laser Welding Solar Energy

Two-stage Techniques

Spin Welding Vibration Welding Ultrasonic Welding

Induction Welding Microwave Heating Dielectric Heating Resistance Welding

Fusion Bonding

Figure 3-12. Fusion bonding techniques.

43

4. SNAP JOINT IN COMPOSITE STRUCTURES

In the majority of the composite structural components, both bolted and/or

adhesive bonded joint was used. Most of the details are similar to those for

metal joints. It was shown from extensive testing on bolted composite joints

that failure always occurs in a catastrophic manner due to high stress

concentration developed at the bolt locations. Due to the inherent low bearing

and interlaminar shear strengths of composites, these stress concentrations

threaten the downfall of every piece of the composite structure [86-87]. The

optimum composite joint design is the one capable of distributing stresses over

a wide area rather than to concentrate them at a point. Adhesively bonded joints

can satisfy these requirements, however, most of the adhesives are brittle, and

brittle failure is unavoidable. This was the motivation of developing what is

called the SNAP joint, Fig. 4-1.

Snap Joint Concept Hardware for fasternerless snap joint

B A

Figure 4-1. Snap joint.

The snap joint technology developed by W. Brandt Goldworthy & Associates,

Inc. [86]. The concept is based on similar joining technology used for

connecting wooden parts. Also, this technique is very similar to techniques

which were used a decade or so again for plastic. Figure 4-1 shows a pultruded

structural composite member (A) with one end shaped as a fir-tree, and

therefore has a large load bearing area. In this figure, part (A) has been snapped

into another structural shape (B). From Fig. 4-1 one can see that, the later shape

has been designed to combine its structural shape with functionality that allows

44

for the engagement of the load-bearing surface of member (A). It is possible to

"snap" joint both parts together since part (A) has been cut for a short distance

along it length to provide enough lateral flexibility to move out of the way

when entering part (B). In order to make this joining concept successful, the

fiber architecture of part (A) must be designed in such a way that the load

bearing surfaces have higher interlaminar shear strength capacity. Also, it can

be noticed from the figure, that a circular hole was introduced at the end of the

horizontal slot of part (A) to inhibit the crack propagation along the length of

the pultruded member. The applications of this technology in composite

structures will have benefits as follow:

• The structures are easy to assembly.

• Installation of structure members become faster.

• Installation needs smaller number of labor and equipment.

• Since it use composite materials, its weight is less than traditional

structures.

The first prototype or "Demonstration project" using this joining method was in

designing and construction of three Transmission Tower Structures, Fig.4-2(a),

near Los Angeles by W. Brandt Goldworthy & Associates, Inc. and Ebert

Composites Corporation [86]. Through California Department of

Transportation, they have proposed the design and the construction of a truss

structure to carry highway singes, Fig. 4-2(b).

The snap joining technique is considered to be one of the optimum techniques

to join composite structural members. However, it has a major limitation, and

can only be used in specific applications. That is, this method can only be used

to transmit axial loads, which make it ideal for truss-type structures. However,

in my opinion, this method should NOT be recommended when out-of-plane

loads or any shear loads are introduced since the connection is not design to

carry any major bending moments. Under flexural loads, it is expected that the

joint will be very flexible, and the artificial cracks introduced to members will

45

propagate and a complete failure will occur even under moderate service

flexural and/or shear loading.

(a) Composite Transmission Tower (b) Truss Joint

Figure 4-1. Snap joint projects.

5. HYBRID JOINTS

In an attempt to improve the joint strength of composite materials, a hybrid of

adhesive and bolted joints has also been explored [88], Fig. 5-1. Hybrid joints

failed at a higher load than the bolted joints and with the proper clamping

torque reached the same failure load as the adhesive joints. Furthermore, unlike

the adhesive joints, hybrid joints failed in two steps, first by initiation of fiber

tear (akin to delamination in laminated continuous fiber composites) at one of

the lap ends and then by tensile failure across the bolt hole. This led to a

slightly higher overall elongation at failure for specimens with the hybrid

joints. Failure in fatigue also started by fiber tear and when the fiber tear

progressed to the bolted area, a combination of half-net-tension failure and

splitting (cleavage failure) occurred. In both static as well fatigue, failure was

initiated by fiber tear and the round washers with their edges located slightly

away from the lap ends were not effective in preventing fiber tear. Fu and

Mallick [88] found that, hybrid joints give better static as well as fatigue

46

performance than adhesive joints in structural reaction injection molded

composites when fiber tear, the primary failure mode in adhesive joints, is

either prevented or delayed by the presence of clamping. Their finite element

analyses proved that the presence of the lateral clamping pressure can

significantly decrease the maximum peel stress at the interface, which helps in

achieving improved joint performance.

Figure 5-1. Sketch of an adhesive/bolted (hybrid) joint.

5.1. Hybrid Joints in Plated Beams

The main drawback of the RC member strengthened by externally bonded

plates is the peeling failure [89,90], Fig. 5-2. Several researchers suggested

analytical models to predict the critical tensile and shear stresses at the ends of

glued plate at the bottom face of the RC beams. This means that, the peeling

crack may grow under mode I, opening mode, or mixed mode, opening and

sliding modes. The anchorage applied at the ends of externally bonded plate

may take several forms, including bolted anchorage systems, bonded angle

sections to provide anchorage of the reinforcing plate to both sides of the beam,

and trapping the plate under the beam supports, Fig. 5-3.

Figure 5-2. Cover separation failure mode.

47

Adhesive-filled bore

CFRP plate Mild steel bolt

Anchorage block

a: Plate ends under supports

CFRP plate

Support roller

Bearing pad

b: Plate ends bolted to beam

Figure 5-3. Plate end anchorage systems adopted by

Garden and Hollaway [see Ref. 90]

Saba [89] suggested two different techniques to preclude the peeling failure at

the ends of steel plates glued to the soffits of reinforced concrete, RC, beams.

In the first technique, concrete cover was replaced by grout to enhance the

resistance of substrate to crack initiation and propagation. In the second one,

permanent compressive forces at the ends of bonded plate were inserted using

different plate end anchorage systems, i.e. end anchorage by bolts or by

clamps, Fig. 5-4.

Grout 25 cm

2 m

5 cm35 cm

70 cm 80 cm80 cm

P/2 P/2

P/2 P/2

(a)

(b)

15 mm St. 6 Studs 14 mm

2 m

70 cm 80 cm80 cm

P/2 P/2

P/2 P/2

25 cm

Figure 5-4. Plated beam with (a) partial replacement of concrete cover, and (b) plate end anchorage by steel clamps

48

Schnerch et al. [91] used a steel angle to clamp the fiber to the steel base plate

before the pot life of the epoxy had been reached as shown in Fig. 5-5.

Figure 5-5. Steel Angles used to mechanically anchor CFRP plys.

Figure 5-6 shows the upgrading techniques, which used hybrid joint concepet,

for reinforced concrete beam–column joints proposed by El-Amoury and

Ghobarah [92].

FRP sheets

Steel angle

U-shaped steel plates Steel plates

Steel plate

FRP sheets

Steel rods

Figure 5-6. Retrofitting schemes

49

6. SUMMARY 6.1 Mechanical Joints (Bolts or Rivets)

Mechanical joints are used when repeated disassembly and reassembly is required or

when surface preparation is not practical. Unfortunately, well-established joining

technologies for metallic structures are not directly applicable to composites for several

reasons, e.g.:

• Stress concentration created by the presence of holes and cut-outs, which is worsened by the lack of plasticity limiting stress redistribution.

• Delamination originating from the localized wear occurring during drilling. • Differential thermal expansion of fasteners relative to composite. • Water intrusion between fasteners and composite. •Possible galvanic corrosion at fastened joints. •Additional weight of fastening system.

Evolution of Mechanical Joints In mechanical joining, there seems to be a distinct shift toward integral attachment using

features either designed into or formed into the parts to be joined:

Snap-fit integral features are

beginning to appear in composites:

The use of so-called “hook-and-loop” attachments (likeDuPont VelcroTMand 3M Dual-LocksTM) will find greater and more-

diverse application:

Tog-L-Loc®

A generic hook-and-loop concept representing DuPon's VelcroTM

Punch side

Cross Section

Die side "Button"

6.2 Adhesive Bonding

This joining technique is inherently preferable to mechanical fastening as the continuous

connection avoids large stress concentration induced at each discrete fastener hole. The

50

presence of a wide range of contaminants on substrate surfaces (release agents and

bagging materials, fluorocarbon release sprays and films, machining oils, fingerprints,

etc.) makes however necessary a preliminary surface preparation treatment, which may

also serve to:

• Improve wetting of low energy surfaces. • Chemically modify the surfaces (introducing polar groups or coupling agents). • Increase the surface roughness.

Adhesive Bonding Benefits

• It is still possible to make joints not only between two composite laminates, but also between a composite laminate and a metal, or a concrete, or a wood structure.

• Loads are distributed over a much wider area than in the case of mechanical joints. • Since no holes are required, the risk of local delamination is practically eliminated. • Significantly reduced weight of the joint. • Possibility to make a wide range of configurations (e.g. stepped or scarfed joints).

Adhesive Bonding Drawbacks

• Adhesively bonded joints cannot be disassembled without destroying the substrate. • Some adhesives are susceptible to degradation by temperature and humidity. • Poor inspectability(critical joints may require ultrasonic inspection over their entire

area). • Corrosion can be a problem due to galvanic action. • Typical surface treatments are generally hard to control in an industrial

environment and directly affect the strength and durability of bonded joints.

6.3 Hybrid joints

Hybrid joints give better static as well as fatigue performance than adhesive joints in

composites when fiber tear, the primary failure mode in adhesive joints, is either

prevented or delayed by the presence of clamping. The presence of the lateral clamping

pressure can significantly decrease the maximum peel stress at the interface, which helps

in achieving improved joint performance. The square washer representing full clamping

to the edges of the overlap area, gives a better performance compared to round washer,

representing partial clamping.

51

7. REFERENCE 1- Jackson, N. and Dhir, R. K. ed. (1996). Civil Engineering Materials, 5th ed., Macmillan

Press LTD, London. 2- Hamilton, S. B. (1958). “Building and civil engineering construction”, in A History of

Technology, Vol. IV (Singer, C., Holmyard, E.J., Hall, A.R., and Williams, T.L. eds), Clarendon, Oxford, Chapter 15.

3- Adams, R. D., Comyn, J., and Wake, W. C. (1997). Structural Adhesive joints in Engineering, 2nd ed., Chapman $ Hall, London.

4- Kosteski, N., Packer, J. A., and Lecce, M. (2000). “Nailed tubular connections under fatigue loading”, J. Struct. Engrg., ASCE, Vol. 126, No. 11, pp 1258-1267.

5- Winter, G. (1956). “Tests on bolted connections in light gage steel”, J. Struct. Div., ASCE, Vol. 82, No. 2, pp 920-1-920-25.

6- Rogres, C. A. and Hancock, G. J. (2000). “Failure modes of bolted-sheet-steel connections loaded in shear”, J. Struct. Engrg., ASCE, Vol. 126, No. 3, pp 288-296.

7- Manuel, T. J. and Kulak, G. L. (2000). “Strength of joints that combine bolts and welds”, J. Struct. Engrg., ASCE, Vol. 126, No. 3, pp 279-287.

8- Al-Samhan, A. and Darwish, S. M. H. (2003). “Strength prediction of weld-bonded joints”, Int. J. Adhesion & Adhesives, Vol. 23, pp 23-28.

9- Davis, T. J. and Claisse, P. A. (2001). “Resin-injected dowel joints in glulam and structural timber composites”, Construct. Build. Mater., Vol. 15, pp 157-167.

10- Liew, J. K. R., Teo, T. H., and Shanmugam, N. E. (2004). “Composite joints subjected to reversal of loading-Part 1: experimental study”, J. Construct. Steel Res., Vol. 60, pp 221-246.

11- Sebastian, W. M. and McConnel, R. E. (2000). “Nonlinear FE analysis of steel-concrete composite structures”, J. Struct. Engrg., ASCE, Vol. 126, No. 6, pp 662-674.

12- Elshafie, H. (2003). “Using FRP to repair and strengthen timber beams”, 10th Int. Coll. Struct. & Geotec. Engng., Ain-shams Univ., Cairo, Egypt, Vol. 5, pp E03mat27-1 - E03mat27-15.

13- ASTM D4533-99 (2000). “Standard test method for fracture strength in cleavage of adhesives in bonded metal joints”. In Annual Book of ASTM Standards, Section 15(06), pp 226-232.

14- Conrad, M. P. C., Smith, G. D., and Fernlund, G. (2003). “Fracture of discontinuous wood-adhesive bonds”, Int. J. Adhesion & Adhesives, Vol. 23, pp 39-47.

15- Armstrong, S. R., Keller, J. C., and Boyer, D. B. (2001). “Mode of failure in the dentine-adhesive resin-resin composite bonded joint as determined by strength based (mTBS) and fracture-based (CNSB) mechanical testing”, Dental Materials, Vol. 17, pp 201-210.

16- Hoa, S. V. ed. (1995). Computer-Aided Design of Polymer-Matrix Composite Structures, Marcel Dekker, Inc., Hong Kong.

17- Upadyay, A. and Kalyanaraman, V. (2003). “Simplified analysis of FRP box-girders”, Compos. Struct., Vol. 59, pp 217-225.

18- Gan, L.-H., Ye, L., and Mai, Y.-W. (1999). “Simulation of mechanical performance of pultruded I-beams with various flange-web conjunctions”, Composites Part B, Vol. 30, pp 423-429.

19- Gilchrist, M. D., Kinloch, A. J., and Matthews, F. L. (1996). “Mechanical performance of carbon-fiber- and glass-fiber-reinforced epoxy I-beams: I. Mechanical behaviour”, Compos. Sci. & Technol., Vol. 56, pp 37-53

20- Burgoyne, C. J. (1999). “Advanced composites in civil engineering”, Struct. Eng. Int., Vol. 9, pp 267-273.

52

21- Nordin, H. and Taljsten, B. (2004). “Testing of hybrid FRP composite beams in bending”, Composites Part B, Vol. 35, pp 27-33.

22- Zetterberg, T., Astrom, B. T., Backlund, J., and Burman, M. (2001). “On design of joints between composite profiles for bridge deck applications”, Compos. Struct., Vol. 51, pp 83-91.

23- Kumar, P., Chandrashekhara, K., and Nanni, A. (2004). “Structural performance of a FRP bridge deck”, Construct. Build. Mater., Vol. 18, pp 35-47.

24- Nagaraj V. and Gangarao, H.V. (1998). “Fatigue behavior and connection efficiency of pultruded GFRP beams”, J. Compos. Construct, ASCE, Vol. 2, pp 57-65.

25- Qian, P., Davalos, J. F., and Brown, B. (2000). “A systematic analysis and design approach for single-span FRP deck/stringer bridges”, Composites Part B, Vol. 31, pp 593-609.

26- Alampalli, S., O’Connor, J., and Yannotti, A. P. (2002). “Fiber reinforced polymer composites for superstructure of a short-span rural bridge”, Compos. Struct., Vol. 58, pp 21-27.

27- EXTREN fiberglass structural shapes design manual (1989) Strongwell, Bristol, Va. 28- Fiberline design manual for structural profiles in composite materials (1995). Fiberline

Composites A/S, Kolding, Denmark. 29- Wörner, J.-D. and Deusser S.(1997). “Surface near textile reinforcement for steel

reinforced Concrete”, Paper No. 421, TechTextil Symposium, Frankfurt. 30- Mohamed, M. H. and Bogdanovich, A. (2003). “3D woven composites for civil

engineering applications”, 2nd Int. Workshop on Struct. Compos. for Infrast. Applic., T. Rizk ed., 16-18 Dec. 2003, Cairo, Egypt, CD-Room.

31- Head, P. R. (1994). “The world's first advanced composite road bridge”, Short & medium Span Bridges, Canada.

32- Mosallam, A. S. (1994). “Connections for pultruded composites: a review and evaluation”, Proc. Of the ASCE Conf., San Diego, CA, USA, pp 1001-1017.

33- Mosallam, A. S. (1995). “Connection and reinforcement design for pultruded fiber-reinforced plastic (PFRP) composite structures”, J. Reinf. Plastics Compos., Vol. 14, pp 752-784.

34- Mottram, J. and Zheng, Y. (1999). “Further tests of beam-to-column connections for pultruded frames: Flange-cleated”, J. Compos Construct, ASCE, Vol. 3, pp 108-116.

35- Mottram, J. T. and Zheng, Y. (1999). “Further tests on beam-to-column connections for pultruded frames: Web-cleated”, J. Compos. Construct, ASCE, Vol. 3, pp 3-11.

36- Liu, D. and Hou, L. (2003). “Three-dimentional size effects in composite pin joints”, Experimental Mechanics, Vol. 43, No. 2, pp.115-123.

37- Avila, A. F. and Bueno, P. (2004). “An experimental and numerical study on adhesive joints for composites”, Compos. Struct., In Press.

38- Y. Yan, W.-D. Wen, F.-K. Chang, and P. Shyprykevich, (1999), “Experimental study on clamping effects on the tensile strength of composite plates with a bolt-filled hole”, Composites A, Vol. 30, pp. 1215-1229.

39- H.E.M. Sallam and H. Nayeb-Hashemi, (1996), “Effect of specimen geometry on fatigue crack growth behavior of metal matrix composite”, Recent Advances in Solid/Structures and Application of Metallic Materials, PVP-Vol. 342/MD-Vol. 72, pp. 17-26, ASME 1996.

40- R. L. Pinckney, (1973), “Helicopter rotor blades”, ASTM STP 524, pp. 108-133. 41- M. J. Salkind, (1972), “Fatigue of composites”, ASTM STP 497, pp. 143-169. 42- Hassan, N. K., Mohamedien, M., and Rizkalla, S. H. (1997). “Rational model for

multibolted connections for GFRP members”, J. Compos. Construct, ASCE, Vol. 1, No. 2, pp 71-78.

53

43- N. Tsangapakis, J. M. Slepetz, and J. Nunes, (1988), “Static and fatigue notch strength in alumina fiber reinforced aluminum plates with a circular hole”, J. Compos. Mater., Vol. 22, pp. 386-393.

44- R. E. Peterson, (1974), “Stress Concentration Factors”, Chap. 3, Wiley, New York, USA. 45- C. Tan and R. Y. Kim, (1990), “Strain and stress concentrations in composites laminates

containing a hole”, Experimental Mechanics, Vol. 30, pp. 345-351. 46- P. P. Camanho and F. L. Matthews, (1997), “Stress analysis and strength prediction of

mechanically fastened joints in FRP: a review”, Composites A, Vol. 28A, pp. 529-547. 47- Li, Q.M., Mines, R.A.W., and Birch, R.S. (2001). “Static and dynamic behaviour of

composite riveted joints in tension”, Int. J. Mechanical Sci., Vol. 43, pp 1591-1610. 48- Sallam, H.E.M. (2003). “Effect of a circular hole on static and fatigue strength of

unidirectional carbon-fiber/epoxy plate” Tenth Int. Colloquium on Struct. and Geotec. Engng., 10th ICSGE, Vol.5, E03MAT18-1, Ain Shams Univ., Cairo, Egypt.

49- H.-Sh. Chen, (2001), “The static and fatigue strength of bolted joints in composites with hygrothermal cycling”, Composite Structures, Vol. 52, pp. 295-306.

50- Aktas¸ A. and Karakuzu, R. (1999). “Failure Analysis of Two-dimensional Carbon-Epoxy Composite Plate Pinned Joint”, Mechanics of Composite Materials and Structures, Vol. 6, pp-347-361.

51- Okutan, B. and Karakuzu, R. (2003). “The strength of pinned joints in laminated composites“, Compos. Sci. & Technol., Vol. 63, pp 893–905.

52- Icten, B., Okutan, B., and Karakuzu, R. (2003). “Failure strength of woven glass fiber-epoxy composites pinned joints”, J. Compos. Mater., Vol. 37, No. 15, pp 1337-1350.

53- G. J. Turvey, (1998), “Single-bolt tension joint tests on pultruded GRP plate: effects of tension direction relative to pultrusion direction”, Composite Struct., Vol. 42, pp. 341-351.

54- Persson, E., Eriksson, I., and Zachrisson, L. (1997). “Effects of hole machining defects on strength and fatigue life of composite laminates”, Composites Part A, Vol. 28A, pp 141-151.

55- Zachrisson, L., Eriksson, I., and Backlund, J. (1994). “Method and Tool for Machining a Hole in a Fiber-reinforced Composite Material”, Swedish Patent No. 500933, October 1994.

56- Lin, H. J. and Tsai, C. C. (1995). “Failure analysis of bolted connections of composites with drilled and moulded-in hole”, Compos. Struct., Vol. 30, pp 159-168.

57- Sergeeva, B., Madencia, E., Amburb, D.R. (2000). “Influence of bolt spacing and degree of anisotropy in single-lap joints”, Computers and Structures, Vol. 76, pp 89-103.

58- Chutima, S. and Blackie, A. P. (1996). “Effect of pitch distance, row spacing, end distance and bolt diameter on multi-fastened composite joints”, Composites Part A, Vol. 27A, pp 105-119.

59- H.-J. Park, (2001), “Effects of stacking sequence and clamping force on the bearing strengths of mechanically fastened joints in composite laminates”, Composite Structures, Vol. 52, pp. 213-221.

60- Choi, J.-H and Chun, Y.-J. (2003). “Failure load prediction of mechanically fastened composite joints”, J. Compos. Mater., Vol. 37, No. 24, pp 2163-2177.

61- F. L. Di Scalea, G. L. Gloud, and F. Cappello, (1998), “A study on the effects of clearance and interference fits in a pin-loaded cross-ply FGRP laminate”, J. Compos. Mater., Vol. 32, pp. 783-802.

62- Tong, L. (2000). “Bearing failure of composite bolted joints with non-uniform bolt-to-washer clearance”, Composites Part A, Vol. 31, pp 609-615.

63- Starikov, R. and Schon, J. (2002). “Fatigue resistance of composite joints with countersunk composite and metal fasteners”, Int. J. Fatigue, Vol. 24, pp 39–47.

54

64- McCarthy, M.A., Lawlor, V.P., Stanley, W.F., and McCarthy, C.T. (2002). “Bolt-hole clearance effects and strength criteria in single-bolt, single-lap, composite bolted joints”, Compos. Sci. & Technol., Vol. 62, pp 1415–1431.

65- Mortensen, F. and Thomsen, O. T. (1997). “Simplified linear and non-linear analysis of stepped and scarfed adhesive-bonded lap-joints between composite laminates”, Composite Structures, Vol. 38, pp. 281-294.

66- Ganesh, V. K. and Choo, T. S. (2002). “Modulus graded composite adherends for single-lap bonded joints”, J. Compos. Mater., Vol. 36, No. 14, pp 1757-1767.

67- Bahei-El-Din, Y. A. and Dvorak, G. J. (2001). “New designs of adhesive joints for thick composite laminates”, Compos. Sci. & Technol., Vol. 61, pp 19-40.

68- Zeng, Q. and Sun, C. T. (2001). “Novel design of bonded lap joint”, AIAA J., Vol. 39, pp 1991–1996.

69- Tsai, M. Y., Oplinger, D. W., and Morton, J. (1998). “Improved theoretical solutions for adhesive lap joints”, lnt. J. Solids Structures, Vol. 35, No. 12, pp 1163-1185.

70- Blackman, B.R.K., Kinloch, A.J., Paraschi, M., and Teo, W.S. (2003). “Measuring the mode I adhesive fracture energy, GIC, of structural adhesive joints: the results of an international round-robin”, Int. J. Adhesion & Adhesives, Vol. 23, pp 293-309.

71- Ducept, F., Davies, P., and Gamby, D. (2000). “Mixed mode failure criteria for a glass/epoxy composite and an adhesively bonded composite/composite joint”, Int. J. Adhesion & Adhesives, Vol. 20, pp 233-244.

72- Ferreira, J. A. M., Reis, P.N., Costa, J.D.M., and Richardson, M.O.W. (2002). “Fatigue behaviour of composite adhesive lap joints” Compos. Sci. & Technol., Vol. 62, pp 1373-1379.

73- Vaziri, A., Hamidzadeh, H. R., and Nayeb-Hashemi, H. (2001). “Dynamic response of adhesively bonded single-lap joints with a void subjected to harmonic peeling loads”, Proc. Instn. Mech. Engrs., ImechE, Vol. 215 Part K, pp 199-206.

74- Berry, N. G. and d’Almeida, J. R. M. (2002). “The influence of circular centered defects on the performance of carbon–epoxy single lap joints”, Polymer Testing, Vol. 21, pp 373–379.

75- Rhee, K. Y., Lee, S. G., Choi, N. S., and Park, S. J. (2003). “Treatment of CFRP by IAR method and its effect on the fracture behavior of adhesive bonded CFRP/aluminum composites”, Materials Science and Engineering A, Vol. A357, pp 270-/276.

76- Melogranaa, J. D. and Grenestedt, J. L. (2002). “Improving joints between composites and steel using perforations“, Composites Part A, Vol. 33, pp 1253-11261.

77- Unden H. and Ridder, S-O. (1985). “Load-Introducing Armature as Component Part of a Laminated Structural Element”, US Patent No. 4,673,606; 13 February 1985.

78- Shin, K. C., Lim, J. O., and Lee, J. J. (2003). “The manufacturing process of co-cured single and double lap joints and evaluation of the load-bearing capacities of co-cured joints”, J. of Materials Processing Technology, Vol. 138, pp 89–96.

79- Dvorak, G. J., Zhang, J., and Canyurt, O. (2001). “Adhesive tongue-and-groove joints for thick composite laminates”, Compos. Sci. & Technol., Vol. 61, pp 1123-1142.

80- Melogranaa, J. D., Grenestedt, J. L., and Maroun, W. J. (2003). “Adhesive tongue-and-groove joints between thin carbon fiber laminates and steel “, Composites Part A, Vol. 34, pp 119-124.

81- Nehdi, M., El Damatty, A., and Rahimi, R. (2003). “Investigation on lap-joint behaviour of GFRP plates bonded to silica fume and rice husk ash concrete”, Int. J. Adhesion & Adhesives, Vol. 23, pp 323-333.

82- Täljsten, B. (1997). “Defining anchor lengths of steel and CFRP plates bonded to concrete”, Int. J. Adhesion & Adhesives, Vol. 17, pp 319-327.

55

83- Chen, J. F. and Teng J. G. (2001). “Anchorage strength models for FRP and steel plates bonded to concrete”, J. Struct. Engrg., ASCE, Vol. 127, No. 7, pp 784-791.

84- De Lorenzis, L., Miller, B., and Nanni, A. (2001). “Bond of fiber-reinforced polymer laminates to concrete”, ACI Mater J., Vol. 98, No. 3, pp 256–264.

85- Ageorges, C., Ye, L., and Hou, M. (2001). " Advances in fusion bonding techniques for joining thermoplastic matrix composites: a review", Composites Part A, Vol. 32, pp 839-857.

86- http://www.new-technologies.org/ECT/Civil/snap.htm. 87- Meitinger, T. and Pfeiffer, F. (1995). "Modeling and simulation of the assembly of snap

joints", IEEE International Symposium on Assembly and Task Planning, August 10-11, Pittsburgh, Pennsylvania, USA, p. 0015.

88- Fu, M. and Mallick, P.K. (2001). "Fatigue of hybrid (adhesive/bolted) joints in SRIM composites", Int. J. Adhesion & Adhesives, Vol. 21, pp 145-159.

89- Saba A. M. (2001). "Crack Arresters Techniques in Concrete Structures". Ph. D. Thesis, Engineering. Materials Department, Zagazig University, Zagazig, Egypt, 2001.

90- Hollaway L. C. and Lemming M. B. (1999). Strengthening of Reinforced Concrete Structures: Using Externally-Bonded FRP Composites in Structural and Civil Engineering, CRC press, Woodehead Pub. Limited, Cambridge, UK.

91- Schnerch, D., Stanford, K., Lanier, B., and Rizkalla, S. (2003). "Use of high modulus carbon fiber reinforced polymers (CFRP) for strengthening steel structures", 2nd Int. Workshop on Struct. Compos. for Infrast. Applic., T. Rizk ed., 16-18 Dec. 2003, Cairo, Egypt, CD-Room.

92- El-Amoury, T. and Ghobarah, A. (2002). "Seismic rehabilitation of beam–column joint using GFRP sheets", Engineering Structures, Vol. 24, pp 1397–1407.

56