Unedited Version 193 Rls

STAR 193RLS
Bonded CementBased Material Overlays

for the Repair, the Lining or the
Strengthening of Slabs or Pavements
This publication has been published by Springer in

2011, ISBN 9789400712386.
This STAR report is available at the following
address:
http://www.springer.com/engineering/civil+engineering/book/9789400712386
Please note that the following PDF file, offered by

RILEM, is only a final draft approved by the
Technical Committee members and the editors.
No correction has been done to this final draft
which is thus crossed out with the mention
"Unedited version" as requested by Springer.
TABLE OF CONTENT
1. INTRODUCTION
(J.-L. Granju, B. Bissonnette, L. Courard)
2. OVERLAY DESIGN PROCESS

(D.W. Fowler, M. Trevio)
Abstract
2.1 Purpose of overlays
2.2 Materials selection
2.2.1 Slabs on grade/pavements
2.2.1.1 Concrete
2.2.1.2 Reinforcement
2.2.2 Structural slabs and decks
2.2.2.1 Concrete
2.3 Joints
2.3.1 Slabs on grade/pavements
2.4 Construction procedures
2.4.1 Steel placement
2.4.2 Environmental effects
2.4.3 Curing
2.5 The BCO process
2.5.1 Project selection
2.5.2 Design
2.5.3 Construction
2.5.4 Quality control/Quality assurance (QA/QC)
2.6 Conclusions
References
3. CONDITION EVALUATION OF THE EXISTING STRUCTURE PRIOR TO

OVERLAY
(L. Courard, M. Trevio, B. Bissonnette)
Abstract
3.1 Introduction
3.2 Assessment of pavement/substrate base
3.2.1 Principles of evaluation
3.2.2 Condition survey of distress
3.2.2.1 Visual examination
3.2.2.2 Structural investigations
3.2.2.3 Interpretation
3.2.3 Test methods for pavement/substrate base
3.2.3.1 Deflection testing
3.2.3.2 Modulus of elasticity
3.2.3.3 Subgrade modulus
3.2.3.4 Concrete flexural strength
3.2.3.5 Load transfer efficiency (LTE)
3.3 Assessment of structural slab/deck/support
3.3.1 Introduction
3.3.3 Structural evaluation
3.3.4 Material evaluation
3.3.5 Test methods and procedures
3.3.5.1 Non destructive testing for the estimation of strength
3.3.5.1.1 Schmidt rebound hammer
3.3.5.1.2 Ultrasonic Pulse Velocity (UPV)
3.3.5.1.3 Probe penetration
3.3.5.1.4 Pullout test
3.3.5.1.5 Breakoff test
3.3.5.1.6 Pulloff test
3.3.5.1.7 Optic fibres
3.3.5.2 Nondestructive tests for flaw detection and condition assessment
3.3.5.2.1 Stress wave propagation methods
3.3.5.2.2 Infrared thermography
3.3.5.2.3 Ground penetrating radar (GPR)
3.3.5.2.4 Electrical and magnetic methods for reinforcement
3.3.5.2.5 Nuclear (radioactive) methods
3.3.6 Methods for the determination of superficial porosity
3.3.6.1 Principles
3.3.6.2 Capillary suction coefficient
3.3.6.3 Mercury intrusion porosimetry
3.3.6.4 Insitu tests
3.3.6.5 Methods for the determination chemical damages
3.4 Feasibility analysis
References
4. BOND
(J. Silfwerbrand, H. Beushausen, L. Courard)
Abstract
4.1 Classification of bond

4.2 Definition of bond strength
4.3 Fundamental bond mechanisms
4.4 Factors affecting bond
4.4.1 Concrete properties
4.4.1.1 Fresh material properties
4.4.1.2 Hardened material properties
4.4.1.3 Overlay strength
4.4.2 Removal of deteriorated concrete
4.4.3 Concrete removal behind rebars, rebar cleaning, and rebar replacement
4.4.4 Cleaning after concrete removal
4.4.5 Surface properties
4.4.6 Surface preparation
4.4.7 Bonding agents
4.4.8 Mechanical devices crossing the interface
4.4.9 Concrete placement
4.4.10 Concrete curing
4.4.11 Short-term bond properties
4.4.12 Long-term bond properties
4.4.13 Traffic vibrations
4.5 Test methods
4.5.1 General
4.5.2 Shear test methods
4.5.3 Comparisons between tensile bond strength and shear bond strength
4.6 Evaluation
4.7 Design strength values
4.8 Performance requirements
References
5. STRUCTURAL BEHAVIOUR
(E. Denari, J. Silfwerbrand, H. Beushausen)
Abstract
5.1 Introduction
5.2 Actions
5.3 Performance of composite structures
5.4 Different forms of restraint and effect of joints
5.5 Mechanical behaviour of composite structures with cementitious materials of
different ages
5.5.1 Overview of existing analytical models
5.5.2 Normal stresses due to differential shrinkage in composite beams with complete
bond
5.5.3 Shear stresses due to differential shrinkage in composite beams
5.5.4 Normal and shear stresses due to differential shrinkage in composite beams with
incomplete bond
5.6 Experimental tests
5.6.1 Swedish tests on mechanically loaded concrete beams
5.6.2 Swedish tests on concrete beams subjected to differential shrinkage
5.6.3 Swedish tests on concrete slabs
5.6.4 Tests at EPFL on composite beams with normal concretes of different ages
5.6.5 Other tests
5.7 On restrained shrinkage set-ups
5.8 Numerical modelling
5.9 Conclusions
5.10 Outlook for future research
References
6. DEBONDING
(A. Turatsinze, H. Beushausen, R. Gagn, J.-L. Granju, J. Silfwerbrand,

R. Walter)
Abstract
6.1 Introduction
6.2 Impact of debonding
6.3 Brief summary of debonding
6.4 Methods for monitoring debonding
6.4.1 Non destructive methods
6.4.2 Semi destructive methods
6.4.3 Laboratory tests
6.5 Debonding mechanism
6.6 Debonding modeling
6.7 Role of reinforcement of overlays
6.7.1 In situ findings
6.7.2 Need to distinguish between first monotonic loading and shrinkage-pre-
cracking plus fatigue
6.7.2.1 First monotonic loading
6.7.2.2 Shrinkage pre-cracking and fatigue loading
6.7.2.3 Conclusion
6.8 Crack propagation and crack opening
6.8.1 First monotonic loading
6.8.2 Shrinkage-pre-cracking and fatigue loading
6.9 Special overlays
6.9.1 Stang and Walters results
6.9.2 Anchoring of the overlay
6.10 Boundaries and joints of the overlays
6.10.1 Boundaries and full depth joints
6.10.2 Sawn joints
6.11 Conclusion
References
7. DESIGN
(M. Trevio, j.-l. Granju, h. Beushausen, a. Chabot, h. Mihashi, j. Silfwerbrand)
Abstract
7.1 Introduction
7.2 Design for sufficient strength
7.2.1 Overview
7.2.1.1 Purpose of the overlay
7.2.1.2 Basic design principle
7.2.2 Design concepts
7.2.2.1 Overall standard deviation
7.2.3 Current overlay design procedures
7.2.3.1 Corps of Engineers
7.2.3.2 Portland Cement Association
7.2.3.3 AASHTO
7.2.4 Reinforcement design of overlaid continuously reinforced concrete pavements
(CRCP)
7.2.5 BCO design procedures in other countries
7.2.6 Summary and conclusions
7. 3 Design to prevent debonding
7.3.1 Introduction
7.3.2 USA recommendations
7.3.3 European requirements
7.3.4 Japanese requirements
7.3.5 Swedish practice
7.3.6 Expected design shear strength
7.4 Conclusion
References
8. PRACTICE AND QUALITY ASSURANCE

(M. Vaysburd, B. Bissonnette, R. Morin)
Abstract
8.1 Scope and definitions

8.2 Importance of substrate surface preparation
8.2.1 General
8.2.2 Safety
8.3 Concrete removal
8.3.1 General
8.3.2 Removal methods
8.3.3 Advantages and limitations of concrete removal methods
8.3.3.1 Basic methods
8.3.3.2 Combining methods
8.4 Substrate surface preparation
8.4.1 General
8.4.2 Pre-overlay repairs
8.4.3 Surface
8.4.3.1 General
8.4.3.2 Concrete surface cleaning
8.4.3.3 Reinforcement cleaning, inspection and repair
8.4.4 Moisture conditioning of the substrate prior to overlay/repair
8.4.4.1 General
8.4.4.2 Optimum water condition of the substrate
8.4.5 Maintenance of the prepared substrate
8.4.6 Quality control of surface preparation
8.4.6.1 Test repair
8.4.6.2 Required strength at the interface
8.5 Application
8.5.1 General
8.5.2 Workmanship
8.5.3 Basic requirement
8.5.4 Repair placement
8.5.5 Consolidation
8.5.6 Surface finishing
8.5.7 Curing
8.6 Quality assurance / construction inspection
8.6.1 Scarification and removal of unsound concrete
8.6.2 Substrate preparation
8.6.3 Placement and consolidation
References
9. MAINTENANCE AND REPAIR OF OVERLAYS

(D.W. Fowler)
Abstract
9.1 Types and causes of distress

9.2 Evaluation of damage
9.3 Repair methods
9.3.1 Cracks
9.3.2 Delaminated concrete
9.3.3 Spalls
9.4 Conclusions
10. CONCLUSION
(B. Bissonnette, L. Courard, J.-L. Granju)
TC 193-RLS: Bonded cement-based material overlays for the repair, the lining or the strengthening of
slabs or pavements
Rparation, re-surfaage ou renforcement des dallages industriels ou des chausses par
un rechargement adhrent base cimentaire
Membership
Name Organization Country Status
B. Bissonnette CRIB - Universit Laval (Qubec, QC) Canada Chairman

H. Beuschausen University of Cape Town (Cape Town) South Africa Member
A. Chabot LCPC Nantes (Nantes) France Associate
L. Courard Universit de Lige (Lige) Belgium Secretary
L. Czarnecki Warsaw University of Technology (Warsaw) Poland Member
E. Dnari EPFL Lausanne (Lausanne) Switzerland Member
R. Gagn University of Sherbrooke (Sherbrooke, QC) Canada Associate
A. Garbacz Warsaw University of Technology (Warsaw) Poland Member
J.-L. Granju LMDC INSA Toulouse (Toulouse) France Honorary chair
D.W. Fowler University of Texas (Austin, TX) U.S.A. Member
H. Mihashi Tohoku University (Sendai) Japan Associate
R. Morin Ville de Montral (Montral, QC) Canada Member
A. Mller CRIB - Universit Laval (Qubec, QC) Canada Member
J. Silfwerbrand Swedish C&C Research Inst. (Stockholm) Sweden Member
M. Trevio Center for Transportation Research, University of Texas (Austin, TX) U.S.A. Associate
A. Turatsinze LMDC INSA Toulouse (Toulouse) France Member
A.M. Vaysburd Vaycon Consulting (Baltimore, MD) U.S.A. Member
R. Walter Technical University of Denmark (Lyngby) Denmark Associate
1. INTRODUCTION
J.-L. Granju (1), B. Bissonnette (2), L. Courard (3)
(1) Laboratoire Matriaux et Durabilit des Constructions (LMDC), UPS-INSA, Toulouse,

France
(2) Centre de Recherche sur les Infrastructures en Bton (CRIB), Universit Laval, Qubec
(QC), Canada
(3) GeMMe Building Materials, ArGEnCo Department, University of Lige, Belgium
1.1 INTRODUCTION
Concrete has been used in construction for more than a century now and a number of existing
structures exhibit distresses and/or lack of loadcarrying capacity. Repair and strengthening of
existing structures are in fact among the biggest challenges civil engineers will have to face in the
years to come. Moreover, the number of concrete structures keeps growing, and therefore repair or
retrofitting needs keep increasing. Present concerns of sustainable development emphasizing
rehabilitation instead of new construction, acts as incentive to this trend.
Among different approaches that can be considered for concrete rehabilitation, bonded overlays are
often the most economical alternative. Overlays are particularly suitable in the case of structures
with large surface areas, where it can be either poured or sprayed. The following types of
applications are of special concern: slabs on grade (for example, industrial floors), pavements, bridge
decks, walls and tunnels. Toppings and linings are also relevant to the same type of problem.
The primary purpose of overlays is to extend the life of the candidate concrete structures, either by
restoring a smooth sound surface and/or the original loadcarrying capacity, or by improving the
loadcarrying capacity by a thickness increase. For a slab or pavement in good condition, a 25 %
increase in thickness can nearly double its stiffness in some cases, resulting in nearly a 50 %
reduction in flexural stress and a significant increase of its service life.
Other reasons for overlaying slabs or concrete pavements include:

y matching the level of an adjacent slab or element;
y replacing deteriorated or contaminated concrete and reinstating the protection of the structure
(especially its reinforcement);
y providing a more durable wearing surface;
y improving the frictional characteristics of the surface for pavements or bridge decks;
y restoring architectural features such as colour or texture.
The potential for overlay applications worldwide sums up to multimillions of square meters.
Nevertheless, the durability of bonded overlay systems still draw concerns in the technical
community because of bond sustainability problems encountered in a number of cases. In such
events, debonding is generally initiated at the boundaries, joints and cracks in the overlay, and can
be aggravated by curling effects in the debonding area. That leads to new cracks and increased
debonding, which accelerate the damage process and soon leads to the need of renewed repairs.
At this time, there is still no accepted design approach or methodology that can warrant the
practitioner a successful outcome of the repair. The available information essentially consists of quite
general recommendations relying on experience and very simplistic design considerations.
In order to progress through understanding and knowledge, the 193RLS RILEM Technical
Committee was created. Its goal was to set the basis for the development of relevant design rules
and technical recommendations for the achievement of durable bonded overlays (for repair or
strengthening).
The primary concern in the scope of work of this TC is the "bond" between the overlay and the
repaired structure (the latter is often referred to as the substrate). Indeed, it is the foremost factor in
a bonded overlay system, provided that a durable repair material is used and that the residual
concrete base is sound, and its durability is synonymous of durability of the whole composite system.
A rational and reliable design requires a thorough understanding of the bond and debonding
mechanisms. The quality of the bond depends on both the substrate preparation and the overlay
placement procedure. In this respect, the quality of workmanship is of major importance. Debonding
generally results from a combination of internal stresses induced by the gravity loads and the
differential drying and thermal length changes between the overlay and the substrate. Irrespective of
the source of internal stresses, the same mechanism governs debonding initiation and propagation.
Two main overlay families must be distinguished: on the one hand, the bridge deck overlays and,
on the other hand, the overlays performed on slabs on ground or on pavements.
This StateoftheArt document is the major deliverable of 193RLS TC. It is intended to provide a
wide and accurate survey of current knowledge and practice, such as to identify rationally what is
really understood, what the shortcomings are, and from that on, what information needs to be
generated to progress towards a real overlay design process. A subsequent task of the members of
this TC will be the writing of Recommendations for reliable and durable repairs, which will be
published in a separate document.
In addition, the State of the Art report highlights inconsistencies in the terminology and definitions
found in the scientific documentation, depending on the country, the sector of the industry, etc. The
Recommendations document will propose some clarifications, based upon the consensus reached
among the TC members.
This State of the Art report is divided into ten chapters.
Chapter 1: Introduction.
Chapter 2: Overlay: decision and construction process.

This chapter first reports, through flow charts, a typical decision process methodology. Then a
review of all relevant aspects of design and construction is proposed: material selection for the
overlay, reinforcement considerations if applicable (rebars, fibre reinforcement), interface with
the substrate, joints, environmental effects, curing of the overlay, and quality assurance/quality
control. Throughout this review, distinctions are made between the main types of overlay
applications (slabs or pavements and bridge decks).
Chapter 3: Condition evaluation of the existing structure prior to overlay.

The first part of this chapter presents a guide to assess the condition of the substrate and to
determine whether a bonded overlay is a relevant solution for the structure under evaluation.
The second part provides a survey and detailed description of available test methods for condition
evaluation of concrete structures.
Chapter 4: Bond.
This chapter first presents an exhaustive overview of all the factors affecting the bond (favorably
or unfavorably), from surface preparation to the use of bonding agents. The available testing
procedures for bond characterization are then discussed. Finally, the existing recommendations to
yield a durable bond are summarized.
Chapter 5: Structural behavior.

This chapter deals with the structural behavior of a composite system made of an existing
substrate and a bonded cementbased overlay. Through a review of analytical models, the
mechanical response of the repair composite is addressed, taking into account gravity loads,
thermal and hygrometric strains, material ageing, as well as boundary conditions such as the
degree of restraint and the presence of joints. Special attention is paid to restrained
deformations, inherent to repaired concrete members. Then, an overview of relevant
experimental work and numerical modeling is presented.
Chapter 6: Debonding.
In this chapter, emphasis is put on debonding, which can lead to the failure of an overlaid system
through partial or total loss of composite action. A summary of field observations and a survey of
existing test methods to detect debonding are first proposed. Debonding mechanisms are then
described, from initiation to propagation, as a result of length changes and/or repeated gravity
loads (fatigue). Further on, proposed modeling approaches are discussed. Supported by
experimental data, modeling is used to demonstrate and help understanding the benefits of
different reinforcement/anchoring solutions. Finally, the influence of joints in an overlaid
structure is reviewed.
Chapter 7: Design.
This chapter summarizes the existing design rules, recommendations and requirements for
bonded concrete overlays, as found in the guidelines of some of the foremost organizations
around the world. The proposed approaches go from very simplistic criteria, such as minimum
bond strength or stress level limitations, to slightly more sophisticated design methods.
Chapter 8: Practice and Quality Assurance.

This chapter sums up the fundamentals of current good field practice, from substrate preparation
to placement and curing of the overlay. It is completed with recommendations for proper quality
control and quality assurance.
Chapter 9: Maintenance and repair of overlays.

Although bonded overlays are an attractive and improving repair alternative, they still carry some
uncertainty and, sometimes, unsuccessful repairs have to be dealt with. This chapter describes
the suitable repair methods or techniques.
Chapter 10: Conclusion.

2. OVERLAY DESIGN PROCESS
D.W. Fowler (1) and M. Trevio (2)
(1) The University of Texas at Austin (TX), U.S.A.
(2) Center for Transportation Research, The University of Texas at Austin, U.S.A.
ABSTRACT
Bonded concrete overlays have been used for nearly 100 year to extend the life of pavements,
concrete slabs, bridge decks or other structural slabs. First, this chapter describes material selection,
for slabs on grade and pavements; joints; and construction procedures including steel placement,
environmental effects, and curing. Secondly, the bonded concrete overlay (BCO) process is
described: the steps required in project selection; design of the BCO; construction, and quality
assurance with flow charts included to provide a graphical overview.
2.1 PURPOSE OF OVERLAYS
Bonded concrete overlays (BCOs) have been used since 1909. The primary purpose of overlays is
to extend the life of a concrete slab or pavement, bridge deck or other structural slab. It has been
shown [1, 2] that as the remaining life of a pavement decreases due to distress, e.g. cracking, spalling
or punchouts, the life can be extended significantly by the use of a bonded concrete overlay. For a
slab or pavement in good condition, a 25 % increase in thickness can nearly double the stiffness,
resulting in nearly a 50 % reduction in flexural stress.
Other reasons for overlaying concrete pavements or slabs include:
provide an improved frictional surface for pavements or bridges;

provide a smooth surface for industrial floors or buildings;
increase the elevation of the top surface to match an adjacent slab;
provide a more durable wearing surface;
repair corrosiondamaged slabs or bridge decks with sound, durable concrete; and
provide architectural features such as color or texture.
2.2 MATERIALS SELECTION
2.2.1 Slabs on grade/Pavements
2.2.1.1 Concrete
The cement content of the BCO concrete must be high enough to ensure that the available paste
is sufficient to achieve bond at the interface, which eliminates the need for a bonding agent, and to
meet strength and permeability requirements. Reduced paste requirements may be met by using
wellgraded aggregates. Some specifications mandate minimum cement content, e.g. 390 kg/m3 for
normal overlay concrete and 490 kg/m3 for dense overlay concrete. Lower levels of cement are
desirable to reduce cost, reduce heat of hydration and reduce shrinkage. Reduced cost and reduced
heat of hydration may also be accomplished by using fly ash as cement replacement; the addition of
fly ash also has the advantages of improving durability and, in many cases, ultimate strength.
The watercement ratio is determined by strength and durability (permeability) requirements.

Generally, a watercement ratio of 0.40 will provide good durability and strength. One specification
requires a watercement ratio of 0.40 for normal overlay concrete and 0.32 for dense overlay
concrete.
Aggregates should be selected for workability; aggregates should also have adequate durability
for the intended application. The paste requirement may be reduced by improving overall aggregate
gradation as suggested by Shilstone [3, 4] and Crouch [5]. Shilstone suggests incorporating an
intermediate aggregate, and this may be particularly helpful when using steel fibers.
Admixtures will often include air entraining agent, which will help workability and improve
freezing and thawing resistance. The use of fly ash may require greater dosages of air entraining
agent to achieve the same percentage of entrained air. High range water reducers are often
specified, and the amount should be based on trial batches. The addition or retarders in hot weather
helps to preserve workability until the concrete can be placed and finished without affecting strength
development.
Steel reinforcing and steel and synthetic fibers have been used as reinforcement. Steel
reinforcing has been used in the form of tied bars and welded mats.
Steel fibers have been used successfully in overlays to control cracks and to minimize drying
shrinkage cracking. Normal steel fiber contents for steel fiber reinforced concrete are normally in
the range of 1 to 2 % by volume. The amount of fibers required to achieve the desired result will
depend on the type of fiber based on its bond characteristics.
Synthetic fibers should be used at a minimum of 0.25 % by volume; however, significant benefits
have been obtained by using up to 1 % volume.
2.2.2.1 Concrete
Overlays for structural slabs and decks are generally used to replace removed concrete since the slab
or deck cannot tolerate additional dead load and still continue to carry the design live load. One of
the main applications of overlays for structural slabs is to replace deteriorated concrete due to
corrosion of the steel reinforcing. The original concrete is removed to a depth of about 25 mm
below the steel, and the overlay is used to reinstate the original surface. Generally, very durable, low
permeability concrete is specified. Guidelines similar to those outlined for pavements are
appropriate for structural slabs. In addition, latex, e.g. styrenebutadiene, is often added (15 % latex
solids by weight of cement) to provide improved bonding, greater flexural strength and decreased
permeability.
Steel bars are normally used for structural slabs and decks. Often the original bars are adequate,
although if significant reduction in cross section due to corrosion has occurred, the affected bars
should be spliced with adequate anchorage length or replaced. In some cases when extensive
replacement of reinforcing is required, corrosion resistant bars, e.g. epoxy coated steel, galvanized,
or stainless steel may be considered for use in highly corrosive environments.
Fibers, steel or synthetic, may be used for crack control and to minimize shrinkage cracking.
2.3 JOINTS
2.3.1 Slabs on grade/Pavements
Jointed slabs and pavements usually require that the joints in the original slab or pavement be
reinstated which is time consuming and costly. Sawing the joints and applying suitable high
elongation joint filler is one possible solution. In some cases it may be possible to clean the joints of
debris and apply joint filler. If it essential to reinstate the dowels at joints to prevent faulting, precast
joint units have been used by removing the concrete for approximately 600 mm on each side of the
joint and installing the units which have dowels in place.
The overlays must be constructed to maintain the integrity of the joints. Saw cutting followed by
application of suitable high elongation joint filler may be the optimum solution.
Joints in structural slabs and decks are expansion joints. In some cases the nosings at the joints of
the original concrete have spalled and deteriorated and must be repaired. The concrete in the
overlay at the joints must have the durability and toughness to resist the impact of vehicular traffic,
where applicable, and one solution is to use very strong, impactresistant concrete headers on either
side of the joints to which the overlay can be butted.
2.4 CONSTRUCTION PROCEDURES
2.4.1 Steel placement
Research has shown that the reinforcing placed on the substrate of a 75mmthick slab with a 75
mm overlay cast on the surface will achieve the same pullout bond strength as reinforcing placed at
middepth of a 150mmthick slab; all bars failed in tension [6]. Placing the reinforcing on the
substrate saves construction time and labor.
2.4.2 Environmental effects
Weather conditions during construction of the overlay can be critical. Hot, dry climates cause the
most problems because of the excessive evaporation of water from the concrete that often results.
High evaporation rates during placements can result in plastic shrinkage cracking. Evaporation rate is
a function of wind velocity, relative humidity, concrete temperature, and air temperature. An
evaporation rate of 1 kg/m2 or higher has been suggested as sufficiently high to cause plastic
shrinkage cracking, but even lower values may cause problems. The evaporation rate can be
estimated from published monographs [7, 8] or calculated from the equation.
The evaporation rate should be monitored through construction by the use of a weather station
which measures the four factors that influence evaporation. If the threshold value of evaporation
rate is approached, action should be taken: discontinue placement of the overlay or provide
measures to reduce evaporation including improved curing methods.
Concrete temperature should be monitored throughout construction. Temperature at placement

has been shown to have a significant effect on performance [9]. Pavements placed in high
temperatures in the summer have been found to have poor performance based on such indicators as
crack spacing and distress occurrence compared to pavements placed in cooler weather. Therefore,
for best performance the BCO placement temperature should be kept relatively low to avoid high
temperatures of hydration. The surface of the substrate should not be permitted to exceed a
temperature of 50 C [9].
Another environmental issue that can be detrimental to the concrete performance is the
temperature differential that occurs in the hours following placement. High ambient temperature
differentials within 24 hours after placement may cause extensive thermal cracking; drops in
temperature from the peak high temperature to the low temperature should not exceed 15 C [9].
The largest drops in temperature in the concrete are usually associated with concrete placements in
the morning in hot weather, which leads to the maximum hydration temperature occurring in
combination with the maximum ambient temperature in the afternoon, followed by a large
reduction in ambient temperature during the night. The rapid cooling of the surface can lead to
thermal contraction strains that exceed the capacity of the concrete at that young age. Weather
forecasts should be consulted to determine if placement conditions will be likely to cause problems.
If any of the adverse environmental conditions described in this section occurs during the
placement of the concrete, placement should be avoided unless the conditions can be offset by such
measures as:
cooling the aggregates or concrete;

special curing methods discussed in a following section; or
use of fly ash as cement replacement to lower the heat of hydration.
2.4.3 Curing
Proper curing procedures are necessary to avoid excessive moisture loss at early ages that can
result in plastic shrinkage and loss in tensile strength capacity at the surface. Curing should begin as
soon after placement and finishing as possible to minimize loss of bleed water. Curing may be
accomplished by:
1) Application of curing compound.

For textured or tined surfaces the spray application should be applied from two directions to
ensure that the entire surface is coated.
2) Application of membrane curing.

Various liquid sealing compounds, e.g. bituminous and paraffinic emulsions, coal tar cutbacks,
pigmented and nonpigmented resin suspensions, or suspensions of wax or nonliquid
protective coating such as sheet plastics or waterproof paper, are used to restrict evaporation
of water.
3) Curing blankets.
A covering of sacks, mattings, burlap, straw, or other suitable paper is placed over the surface
to reduce evaporation and to reduce the temperature reduction at the surface. When used to
reduce evaporation the blankets are generally wetted.
4) Monomolecular film.
Monomolecular films (MMF) are compounds that form a thin monomolecular film to reduce
rapid moisture loss from the concrete surface prior to curing. Another curing method should
be used after the evaporation retardant is sprayed on. Research has shown, however, that the
use of MMF followed by application of curing compound does not consistently provide less
evaporation than curing compound alone.
Table 2.1 summarizes recommended curing procedures for bonded concrete overlays.
Table 2.1: Recommended curing for bonded concrete overlays
Condition Recommendation
Evaporation below 0.5 kg/m2/hr Membrane curing
Evaporation above 0.5 kg/m2/hr but below 1 Membrane curing, plus evaporation retardant or
kg/m2/hr fogging or wet mats, in place for 12 hours
Evaporation over 1 kg/m2/hr Membrane curing, plus wet mat curing or fogging or
other approved methods, in place 36 hours
Temperature drop in next 24 hours less than Membrane curing

15 C below temperature at time of paving
Temperature drop in next 24 hours more than Membrane curing plus wet mats for 36 hours, or
15 C below temperature at time of paving other approved methods
2.5 THE BCO PROCESS
The BCO process is summarized in the flow chart in Fig. 2.1. Each step of the process will be
discussed.
Figure 2.1: BCO development process

2.5.1 Project selection
The project selection involves several decisions as shown in the flow chart in Fig. 2.2.
Figure 2.2: Flowchart of the project selection stage

Figure 2.2 (cont.): Flowchart of the project selection stage
1) Need for rehabilitation, which is based on level of deterioration, age and increase in traffic
loadings.
2) Availability of resources to the owner/agency with the responsibility for maintaining the
highway.
3) Type of rehabilitation: overlay versus nonoverlay.
4) Type of overlay if an overlay is selected. The overlay choices are portland cement concrete
(PCC) and asphalt concrete. For PCC overlays the choice is between bonded and unbonded.
5) Timing and condition. A PCC overlay should be applied after structural failures have occurred
but prior to functional failures at which time the pavement has reached a minimal level of
serviceability. Figure 2.3 shows the pavement serviceability index (PSI) as a function of time
and a hypothetical time at which structural failure and functional failure occurs.
Once the decision is made to construct a BCO, the next step is the design.
Figure 2.3: Manifestation of structural and functional failure along the PSI curve
2.5.2 Design
The design decisions that must be made are:
design period (overlay design life);

traffic analysis; and
remaining life of the original pavement.
The original pavement must be characterized, which usually requires:
coring to determine modulus of elasticity, tensile strength, and thickness; and

deflection testing.
The thickness design is a function of:
structural capacity of existing pavement; and

structural capacity of BCO to fulfill future traffic requirements.
There are several design procedures available for performing the design, but a discussion of the
methods is beyond the scope of this document.
2.5.3 Construction
Some of the constructionrelated considerations are:
1) Materials and aggregates.
Compatibility between the overlay and original pavement concrete is important, and the
coefficient of thermal expansion and modulus of elasticity should be lower than the original
concrete if possible. It has been found that aggregates used in the overlay should have lower
thermal coefficients that can normally be obtained by using limestone materials.
2) Repairs.
The original pavement must be repaired prior to overlay; the overlay will have no more
integrity than the pavement to which it is bonded.
3) Surface preparation.
The surface must be adequately cleaned, preferably by shot blasting using steel shot to a
moderate roughness. The surface must be kept clean prior to placement of overlay concrete.
Bonding agents are not necessary and should be avoided since they provide an additional step
that can cause failure, e.g., a bonding agent that is allowed to cure prior to concrete
placement, becoming a bond breaker [10].
4) Tied PCC shoulders.
Tied PCC shoulders to the BCO provide slab edge support that in turn results in a reduction of
stresses and deflections in the BCO and shoulder due to enhanced load transfer.
5) Critical weather conditions.
Several research studies have shown that the following weather conditions during BCO
construction may have adverse effects on the overlay performance [1, 9, 11, 12]:
a) It has generally been accepted that evaporation rates of ~1 kg/m2/h or more can lead to
plastic shrinkage cracking which may lead to greater probability of delamination. It should
be noted, however, that with the wide range of chemical and mineral admixtures currently
being used the critical evaporation rate to initiate plastic shrinkage cracking might be much
lower than 1.0 kg/m2. Some transportation agencies now require weather stations on
paving construction projects to monitor evaporation rate.
b) Substrate surface temperatures > 50 C at concrete placement may lead to inadequate

bonding.
c) Daily temperature differentials of > 15 C (between the high ambient temperature on the
day of placement and the low night temperature following) may lead to poor bonding.
6) Special.
When evaporation rates approach or exceed the recommended maximum values, special
curing, e.g., evaporation retardants, wet mats or fogging, should be required.
2.5.4 Quality control/Quality assurance (QA/QC)
A comprehensive QA/QC program is required to insure that the BCO provides the owner with a
costeffective pavement and results in a uniform, durable, safe, and low maintenance riding surface
for the users. The program should include some or all of the following items:
statistical sampling samples from each subset of paving should be required;

weather monitoring as described in the previous section relating to construction;
materials tests;
condition surveys of the completed BCO which includes sounding (to locate possible
delamination) and mapping of cracks and other visible defects;
pulloff tests (to determine bond strength);
core tests; and
deflection testing, e.g., falling weight deflectometer.
2.6 CONCLUSIONS
Bonded concrete overlays have many uses for highways, bridges and floors. There are many
factors that must be taken into account in their design: material selection, joints, and construction
procedures.
The use of bonded concrete overlays should be determined based on a rational project selection
process, which considers the need for rehabilitation, availability of resources, type of rehabilitation,
type of overlay and appropriateness of a BCO. The design should be conducted using a rational
procedure to achieve the design life required by the owner. Construction considerations should
include materials to insure compatibility of new and old concrete, repair of existing pavement,
surface preparation, surface cleaning, weather conditions and special curing. A QA/QC program
should be required to insure the quality of the finished BCO.
REFERENCES
[1] Delatte, N.J., Grater, S.F., TrevinoFrias, M, Fowler, D.W., and McCullough, B.F., Partial
Construction Report of a Bonded Concrete Overlay on IH10, El Paso, and Guide for Expedited
BCO Design and Construction Research Report 29115F, Center for Transportation Research,
The University of Texas at Austin (Austin, Texas, 1996).
[2] Van Metzinger, W., Lundy, J.R., McCullough, B.F., and Fowler, D.W., Research Report 12054F
Design and Construction of Bonded Concrete Overlays, The University of Texas at Austin Center
for Transportation Research, Austin, TX, January 1991.
[3] Shilstone, James M., Sr., "Concrete Mixture Optimization," Concrete International, American
Concrete Institute, Farmington Hills, Michigan, June 1990, pages 33 to 39.
[4] Shilstone, J. Sr., and Shilstone, J. Jr., PerformanceBased Concrete Mixtures and Specifications
for Today Making the Best Use of Local Materials to Reduce Costs, Concrete International,
American Concrete Institute, MI, Vol. 24 No. 2, February 2002.
[5] Crouch, L., Sauter, H.; and Williams, J., "92MPa Airentrained HPC," TRBRecord 1698,
Concrete 2000, page 24.
[6] Kailasananthan, K., McCullough, B.F., and Fowler, D.W., A Study of the Effects of Interface
Condition on Thin Bonded PCC Overlays, Research Report 3571, Center for Transportation
Research, The University of Texas at Austin, October 1984.
[7] Kosmatka, S., and Panarese, W., Design and Control of Concrete Mixtures, Portland Cement
Association, Thirteenth Edition, 1992.
[8] Menzel, C., Causes and Prevention of Crack Development in Plastic Concrete, Proceedings,
Portland Cement Association, Annual Meeting, 1954.
[9] Whitney, D., Isis, P., McCullough, B.F. and Fowler, D.W., Research Report 9205 An
Investigation of Various Factors Affecting Bond in Bonded Concrete Overlays, Center for
Transportation Research, The University of Texas at Austin, Austin, TX, June 1992.
[10] Suh, Y.C., Lundy, J.R., McCullough, B.F., and Fowler, D.W., A Summary of Studies of Bonded
Concrete Overlays, Research Report 4575F, Center for Transportation Research, The
University of Texas at Austin (Austin, Texas, 1988).
[11] Teo, K., Fowler, D.W., and McCullough, B.F., Monitoring and Testing of the Bonded Concrete
Overlay on Interstate Highway 610 North in Houston, Texas, Research Report 9203, Center
for Transportation Research, The University of Texas at Austin (Austin, Texas, 1989).
[12] Lundy, J.R., McCullough, B.F., and Fowler, D.W., Delamination of Bonded Concrete Overlays at
Early Ages, Research Report 12052, Center for Transportation Research, The University of
Texas at Austin (Austin, Texas, 1991).
3. CONDITION EVALUATION OF THE EXISTING STRUCTURE PRIOR TO
OVERLAY
L. Courard (1), M. Trevio (2), B. Bissonnette (3)
(2) Center for Transportation Research, The University of Texas at Austin (TX), U.S.A.
(3) Centre de Recherche sur les Infrastructures en Bton (CRIB), Universit Laval, Qubec (QC),
Canada
ABSTRACT
The knowledge and analysis of the causes of degradations is the first step in repairing existing
concrete structures. The assessment is performed by using destructive and non destructive methods.
These allow detecting modification of physical, chemical and/or mechanical properties of concrete:
cracking, deformation, spalling, carbonation, etc., are various causes of distress, with potential
irreversible consequences on concrete structure behaviour.
A description of specific methods for pavements and structural slabs is proposed; it is usually
recommended to combine complementary methods for efficient diagnosis. This chapter wants to
offer an overview of techniques and references that could help the user in selecting the most
appropriate investigation program.
3.1 INTRODUCTION
The proper planning of investigations before carrying out repair works is important if
optimum use is to be made of the test data. a major step needs to be taken from knowing, for
example, that chloride is present in the concrete, to select the best methods for repair. Engineers
require guidance on both the techniques available for the condition assessment of structures and the
methods for data interpretation.
Two major stages are usually recommended for such an operation [1]. The first stage is
based on a rapidscan visual assessment, often including limited sampling in areas obviously
damaged, from which areas can be selected for more detailed investigations. The second stage
contains a detailed diagnostic survey that relies on destructive and non destructive testing
techniques.
Non destructive techniques are more and more utilized not only for the evaluation of the
concrete strength but also for the detection of cracks and delaminations. More sophisticated
techniques have been developed recently, mainly for the study of the concrete at early ages [2]. But
they are easily applicable for the analysis of old structures. Finally, the interest for techniques able to
evaluate concrete permeability is rising, due to the importance of this factor for durability
evaluation.
The sections hereafter will discuss the assessment of pavement and deck/slab concrete
structures. Even if the material is the same, there are differences between these two types of
structures: actions, loading, aggressions, etc., that can lead to differences in the investigations and
the interpretation of the results of the assessment.
3.2 ASSESSMENT OF PAVEMENT/SUBSTRATE BASE
3.2.1 Principles of evaluation
The first stage in the evaluation [3] of an existing pavement is the identification
phase. It involves the following steps:
assessment of the type of road structure with its environmental and loading conditions;
visual examination of the surface layer of the road;
coring to check the thickness and the residual performance of the concrete; and
assessment of the possible causes of distress.
When this identification phase is completed, providing an insight into the possible origins of the
problems, a more quantitative and problemspecific evaluation can be performed [3]. This second
stage is the quantification stage.
Two types of road structures are usually concerned: discontinuous concrete slabs and continuously
reinforced concrete pavement overlays. The evaluation of the road structure is based on the
following activities:
visual examination;
investigations;
analysis of observations; and
diagnostic.
Each step is important in order to have an accurate estimation of the state of the concrete in
the road structure.
3.2.2.1 Visual examination
The general situation of the road must be checked (situation, drainage of water, profile,
environment, etc.) before the observation of the quality of the concrete. The main types of
degradations have been described by the Belgian Road Research Centre and are summarized in Table
3.1 (CRR 1991). The type and the situation of the defaults must be recorded. It is clear that the
visual condition surveys not only involve mapping of the cracks, but also includes a visual survey of
the drainage system. Information is required on the level of the road with respect to that of its
surroundings and to the presence or absence of large slab faulting or other large unevenness.
Table 3.1: Defaults in relation with road concrete degradations (CRR 1991)
Group Type Description

Cracking Transverse crack Rupture line perpendicular to the axis of the road
Longitudinal crack Rupture line parallel to the axis of the road
Corner crack Rupture line between transverse joint and longitudinal edge of
the overlay
Inclined crack Short rupture line inside the slab and with 45 inclination with
road axis
Alligator cracking Mesh cracking
Deformation Pumping (punchout) Water or sludge pumping in the joint when vehicles passage
Stair steps (faulting) Difference of level between the two edges of the joint or the
crack
Settlement Settlement of the edges of the slab, after breaking up
Spalling Pothole Hole with rounded shape and affecting concrete layer
continuity
Scaling Superficial disintegration of the concrete
Joints distress Opening of longitudinal joint Large and irreversible opening of the joint
Degradation (spalling) Explosion of the upper part of the overlay around the joint
Loss of waterproofing Cracking or debonding or pullout of the sealing mass
properties
Creep of sealing mass Overflowing of the sealing mass out of the joint
3.2.2.2 Structural investigations
Structural investigations are needed to give the data necessary for the choice and the design of the
overlay. Analysis of the structure can be done in different ways:
Inspection trenches along the road structure: thickness and aspect of the concrete layer,
composition of the different layers of the road structure, type of subsoil, etc.
Coring: coring can give the same information as obtained from inspection trenches. Cores
taken at cracks can give essential information on the potential for reflective cracking
(orientation, evolution, propagation, etc.). The cores will also permit laboratory testing
investigations to be conducted (see chapter 2.3.).
Falling Weight Deflection measurements (FWD) for the back calculation of E and ES moduli.
These measurements should be taken on sound areas, and preferably between the wheel
tracks [3].
Deflection measurements at joints or cracks to assess the load transfer across the crack or
joint. These measurements should be done together with crack width measurements. They
will be useful to estimate whether or not voids or loss of support have developed due to, for
example, erosion and pumping.
Crack activity measurements to determine in more detail the degree of slab rocking and the
load transfer characteristics across the crack.
A structural analysis every kilometre of road is a minimum (CRR 1991) if the visual inspection
concludes to the homogeneity of the concrete upper layer. If not, it is necessary to select zones
(good and bad) where more investigations are to be done: it can lead to five or more locations every
kilometre of road.
3.2.2.3 Interpretation
Data coming from visual and structural inspections are analysed in order to point out the causes and
to choose the solutions. These data are recorded in a matrix that should include the identification of
the road section and, for each type of defaults (Table 3.1), the percentage of the concrete base
affected or the cumulated length. Structural investigations on site and in the laboratory are also
analysed and introduced in the same type of chart in such a way that zones can be classified and
referred, regarding the intensity of the degradation. Finally, a diagnostic is proposed and potential
causes of defaults and distress are pointed out, leading to the choice of the intervention: localized
repair, overlay, slab replacement, or complete reconstruction of the road structure.
The Belgian Road Research Centre (CRR 1991) has proposed a strategy of rehabilitation,
regarding the existing situation of the concrete roadway structure (Table 3.2).
Table 3.2: Diagnostic and remedial measures (CRR 1991)
Number of vehicles per day

Condition of the
Visual observations Drainage < 4,000 4,000 to > 18,000
substrate
18,000
Isolated cracks affecting less than A or B adapted 1 1 1
10 % of the length of the road A unadapted 1 1&2 1&2
section. B unadapted 1 or 3
Low stair steps (< 3 mm).
No settlement.
Cracks affecting between 10 and 50 A or B adapted 1 or 3 1 or 3 or 4 5
% of the length of the road section. A unadapted 1 or 3 1&2 2&5
Stair steps (< 6 mm). or 2 & 3 or 2
Low settlements. &4
B unadapted 1, 3 or 4 2&4 2&8
or 2 & 5
Cracks affecting more than 50 % of A or B adapted 6 or 7 6, 7 or 8 6, 7 or 8
the length of the road section. A unadapted 6 or 7 2&6 2 & 6 or 2
Stair steps > 6 mm or important or 2 & 7 &7
settlements. B unadapted 2 & 6 or 2 2&6 2&8
&8 or 2 & 8
Note: A = substrate is composed of quality and non polluted materials
B = substrate is composed of low quality and polluted materials
Accordingly, the various possible rehabilitation procedures are classified in eight different
categories (Table 3.3).
Table 3.3: Remedial measures (CRR 1991)
Code Solution proposal

1 Continuing usual maintenance
2 Increasing of drainage efficiency
3 Patch repairs
4 Replacement of deteriorated slabs
5 Thin adhesive overlay
6 Thin non adhesive overlay
7 Replacement of the old overlay
8 Total reconstruction of the road structure
The characterization of the material itself can be conducted in a similar way than the one described
for structural slabs and decks.
3.2.3 Test methods for pavement/substrate base
3.2.3.1 Deflection Testing
Since 1960, the nondestructive deflection testing methods have become more widespread known
and have earned acceptance as a means to evaluate the insitu pavement conditions [4].
The most common device for deflection evaluation is the Falling Weight Deflectometer
(FWD), shown in Fig. 3.1. There are other deflection devices that were used in the past, such as the
Benkelman Beam, Dynaflect and Road Rater.
With the FWD, the loads are applied in several locations, by dropping a fixed weight from a
given vertical height above the pavement; different weights are dropped at each location. The
pavement response is measured as deflections by sensors (geophones) placed at fixed distances from
the load along the pavement.
Figure 3.1: Falling weight deflectometer (FWD)

The deflection testing intervals vary depending on the project characteristics. It is common
practice to conduct FWD measurements approximately every 120 m, but this number is to be taken
only as a general guideline. The length of the project, its overall condition, and the availability of
resources will ultimately determine the frequency and extensiveness of the deflection testing. Two
types of deflection information should be collected: measurements should be taken along
continuous slabs of pavement with no cracks between the deflection sensors, i.e., at the midspan, for
elastic layer moduli evaluation; the second type of measurement should be conducted across
transverse cracks, for load transfer evaluation purposes. For this kind of measurement, it is
recommended to arrange the FWD sensors with respect to the crack as illustrated in Fig. 3.2, in
which sensor number 4 is positioned on the other side of the crack with respect to the remaining
sensors (downside arrangement).
Direction of Measurement
Transverse Crack
Loading Plate
Sensor Number 4 1 2 3 5 6
Figure 3.2: FWD downside loading and sensor arrangement for load transfer measurement
3.2.3.2 Modulus of Elasticity
The modulus of elasticity of a pavement is a measure of its stiffness; it is the most common
way of characterizing its loadcarrying capacity. It is defined as the ratio of stress to strain. For
materials with a linear stressstrain relationship, the modulus is simply the slope of the stressstrain
line, but for nonlinear materials like concrete, the modulus estimation is more complicated.
The nonlinear behavior of concrete makes it obvious that the conventional concept of
modulus of elasticity does not apply to concrete. Therefore, in order to systematically characterize
this property, it is necessary to resort to arbitrary definitions, based on empirical considerations. For
instance, it is possible to define the initial tangent modulus or the modulus tangent to a
predetermined point of the stressstrain curve. The secant modulus is more frequently used in
laboratory tests.
However, the most frequently used method to arrive at moduli of elasticity values for the
pavement structure is from deflection testing. The surface deflection data from FWD collected at the
midspan of slabs, i.e., between cracks, are used to calculate the elastic moduli of the pavement
layers, by a procedure called backcalculation. Normally, only one predetermined loading level of the
FWD is considered, which means that only one of the weight drops is utilized. A load magnitude of
approximately 9,000 pounds is commonly utilized, since it simulates the standard wheel load of an
ESAL at one spot (AASHTO 1993).
Backcalculation is an iterative process that may be tedious and timeconsuming, therefore, it

is recommended to analyze the FWD data by means of computer programs. The goal of
backcalculation is to estimate the pavement material layer stiffnesses, trying to find a set of
parameters that correspond to the best fit of the measured deflection basins and minimizing the
differences between the measured and the calculated deflection bowls. A computer program like
MODULUS [5, 6] may be used for this kind of analysis. The program is intended for flexible
pavements, but it can be successfully applied to PCC pavements as well. This program is designed to
process data collected with FWD using a linear elastic procedure to generate a database of computed
deflection bowls prior to the backcalculation process. The program iterates until the measured and
computed deflections converge.
An alternative method is to backcalculate the elastic moduli using the charts and equation
(AASHTO 1993).
Modulus values of concrete pavement may also be determined in the laboratory by sample
testing. Underlying strata can be tested too, but this is seldom done, as deflection testing constitutes
a more economical option. Many times, laboratory tests are used as a resource to confirm the
certainty of the backcalculated moduli.
When both deflection and core testing are available and the resulting elastic moduli
are compared, it should be noted that the moduli obtained from the core testing will be lower
than the modulus from deflection tests because the stress level applied to the core samples is
higher than the stresses that occur during deflection testing, and because of the non-linear
elastic behavior of concrete. Typically, core specimens are loaded to approximately 45% of
their estimated compressive strength and the secant modulus is calculated, whereas for
deflection tests the stress levels are much lower, resembling an initial tangent modulus, as
illustrated in Fig. 3.3.
tangent
failure
secant
Stress (
()
0.45
failure
Laboratory Testing
Deflection Testing
Deformation (
()
Figure 3.3: Moduli of elasticity for coring and deflection testing

There are several empirical equations that have been developed to estimate the modulus
from other concrete properties, like the expression shown below, where the concrete modulus, Ec, in
psi, is inferred from the compressive strength, fc, in psi (1):
E c = 57,000 f 'c (1)
The same expression in SI units, i.e., with Ec and fc, in Pa, would be (2):
E c = 4,730,000 f 'c (2)
3.2.3.3 Subgrade Modulus
From the backcalculation process, the moduli of all the pavement layers, slab, subbase and
subgrade can be estimated. Nevertheless, the subgrade modulus requires additional considerations.
Several factors affect the stressstrain relationship of soils, such as the moisture content, confining
pressure and density, making it a very complex characteristic. When referring to subgrades, the
elastic modulus of the soil is known as the resilient modulus, MR. In rigid pavement design, besides
the resilient modulus, another concept may be used to characterize the subgrade, known as the
modulus of subgrade reaction, k.
The resilient modulus may be estimated by backcalculation of deflection test results, as

explained above, which may be the easiest way, but it may also be determined in the laboratory
using a triaxial test. The modulus of elasticity is the relationship between stress and strain;
therefore, MR is the ratio of the axial deviator stress, d, and the recoverable axial strain, a (3).
d
MR = (3)
a
The modulus of subgrade reaction is defined as the ratio between an applied pressure and
the ensuing deflection. It is determined by a plateloading test. The plate is 30 in. in diameter and
the load is applied at a predetermined rate until a pressure of 10 psi is reached. The concept of the
modulus of subgrade reaction is illustrated in Fig. 3.4.
tangent

Stress (
() ks =

secant
Deformation (
()
Figure 3.4: Modulus of subgrade reaction

As Fig. 3.4 shows, k varies depending on the stress level used, whether the modulus is considered
tangent or secant. It also varies depending on the moisture conditions; therefore, it is recommended
to analyze it on a seasonal basis. The modulus of subgrade reaction is directly proportional to the
roadbed soil resilient modulus; however, there is no unique correlation between both. An empirical
conversion from MR, in psi, to k, in pci, (AASHTO 1993) is applicable only if the slab is placed directly
on the subgrade, which does not occur very often (4).
MR
k= (4)
19.4
The analogous expression in SI units, i.e., with MR in Pa, and k in N/m3 (5):
MR
k= (5)
0.5
However, the previous equations render values of k that are too large, i.e., it overestimates k.
A chart (AASHTO 1993) is proposed (Fig. 3.5), for the cases in which the pavement slab is placed on
top of a subbase, to determine the composite modulus of subgrade reaction, developed applying the
30in. plate loading onto a twolayer system.
Figure 3.5: AASHTO procedure to estimate the composite modulus of subgrade reaction
3.2.3.4 Concrete Flexural Strength
Only for certain applications of concrete, one of which is PCC pavements, it is necessary to
know its resistance under flexion. Also known as modulus of rupture, the flexural strength of
PCC is evaluated by beam-breaking tests. Concrete beams, 6 in. by 6 in. by 18 in. are cast
and tests are normally performed at 7, 28 and 90 days. The beam specimens are tested in
third point loading, with the break occurring suddenly with the appearance of a single crack.
The value used for design is the mean value of the 28-day test specimens. An alternative way
of estimating the modulus of rupture (6) is by means of an approximate relationship with the
concrete compressive strength [7]:
S 'c = K f 'c (6)
where
Sc = PCC modulus of rupture

K = constant between 8 and 10
fc = PCC compressive strength
3.2.3.5 Load Transfer Efficiency (LTE)
The LTE of a pavement structure refers to its ability to transfer loads across transverse
discontinuities such as joints or cracks. A high LTE value indicates that the pavement structure is
capable of distributing the loads adequately at the discontinuities. The LTE calculations make use of
the deflection data collected across transverse cracks as explained before, and as shown in Fig. 3.2.
In this study, two different procedures are recommended to calculate LTE, which make use of a
deflection ratio. The first procedure is described by the following expression (7) proposed by Teller
[8]:
2WU
LTE = *100% (7)
WU + WL
where LTE is the load transfer efficiency (percentage), WU the deflection on the unloaded
slab and WL the deflection on the adjacent loaded slab.
If the LTE is zero, it means that no load is transferred from the loaded slab to the
adjacent unloaded slab. In the case of perfect load transfer, the load is distributed
completely from the loaded slab to the unloaded adjacent slab (i.e., the deflection is the
same in both slabs) and LTE is equal to 100%.
The second procedure for determining LTE (8) was developed by Darter [8]:
WU
LTE = *100% (8)
WL
where the variables are the same as in Tellers equation.

3.3 ASSESSMENT OF STRUCTURAL SLAB/DECK/SUPPORT
3.3.1 Introduction
All assessments should be made of the defects in the concrete structure, their causes, and of
the ability of the concrete structure to perform its function. The process of assessment should
include, but not be limited to the following (ENV 15049, 4.3):
present condition of the existing concrete structure, including nonvisible and potential
defects;
original design approach;
environment, including exposure to contamination;
conditions during construction (including climatic conditions);
history of the concrete structure;
conditions of use (e.g. loading); and
requirements for the future use of concrete structure.
The results of the completed assessment are only valid at the time that the repair works are
designed and carried out. The nature and causes of defects, including potential combinations of
causes, should be identified and recorded.
In most cases assessment will have been carried out as a separate operation before the start
of the protection or repair, but in all cases it is essential to assess the full extent and the causes of
defects. Assessment may be undertaken in several stages. The purpose of a preliminary stage is to
advise on the immediate safety of the concrete structure, to give an informed opinion on the urgency
of commissioning further surveys, protection or repairs, and to outline what testing should be done
to establish the causes and likely extent of the defects. The purpose of further assessment is as
follows:
to identify cause(s) of defects;

to establish the extent of defects;
to establish whether defects can be expected to spread to parts that are at present unaffected;
to assess the effect of defects on structural capacity; and
to identify all locations where protection or repair may be needed.
Assessment should include testing or other investigation with the aim of revealing hidden
defects and causes of potential defects. The causes and type of defect should be considered during
assessment. Causes may include, but are not limited to those listed below (ENV 15049) and in Table
3.4:
1. Causes of defects due to inadequate structural design

2. Causes of defects due to inadequate construction or materials :
inadequate mix design, insufficient compaction, insufficient mixing, excess water in mix;
insufficient cover;
insufficient or defective waterproofing;
contamination, poor or reactive aggregates; and
inadequate curing and excessive evaporation.
3. Causes of defects revealed during service :
foundation movement, impacted movement joints, overloading; and
impact damage, expansion forces from fires.
4. External environment and agents :
severe climate, atmospheric pollution, chloride, carbon dioxide, aggressive chemicals;
erosion, aggressive groundwater, seismic action; and
stray electric currents.
The most common symptoms of concrete floor damage [9] are fulldepth cracks or, less
frequently, surface cracks. Another type of damage is delamination, displayed in the first stage by
hollow sound while tapping and by breaking off floor panels in the final stage. The damages caused
by chemical corrosion or freeze corrosion can be observed in places subject to aggressive
environment influence. The freeze corrosion frequently occurs in parking decks. Floor defects
include unevenness and lack of slope for drainage, which result in standing water. Concrete floors
defects also include wash boarding and settingoff the joints as well as excessive abrasion and
dusting.
Table 3.4: common causes of defects (env 1504-9:1997)
Defects in concrete Reinforcement corrosion

Mechanical Chemical Physical Corrosive contaminants
At mixing From the environment
Impact Alkali-aggregate Freeze-thaw sodium sodium chloride
and stray currents
Overload reaction Thermal chloride other contaminants

Carbonation
Movement aggressive Salt calcium deicing salts

(e.g. settlement) agents crystallization chloride sea water
Vibration (e.g. sulphates, Shrinkage
Explosion salts, fresh Erosion
water) Wear
Biological Fire
activities
For the repair methods involving the application of mortar and concrete, the european
standard en 1504-10 recommends (table 5) a series of investigations before and/or after
preparation of the concrete substrate.
Table 3.5: Summary of tests and observations for quality control (En 1504-10:2002)
Test
Characteristic Test Method or Observation Standard(s) Frequency Status
No.
1 Delamination Hammer sounding (T) Once before A
application
2 Cleanliness Visual (O) After preparation and A
Wipe test (T) immediately before
application
3 Roughness Visual sand test (O) or ISO 3274 B
Profile Meter (T)
4 Surface tensile Pulloff test (T) EN 1542 B
strength of substrate BS 1881
BS 207
5 Crack movement Mechanical or electrical gages BS 1881 C
(O) BS 206
6 Vibration Accelerometer (O) C
7 Temperature of the Thermometer (O) Throughout application A
substrate
8 Carbonation Phenolphtalein test (T) EN 104865 C
9 Chloride content Site sampling and chemical BS1181124 C
analysis (T) DB5423.78
NTBUILD 208
10 Penetration of other Site sampling and chemical C
contaminants analysis (T)
11 Electrical resistivity Wenner test (T) C
12 Compressive strength Core and crushing test (T) EN 125041 B
Rebound hammer test (T) EN 125042
T : test A : for all intended uses

O : observation B : for certain intended uses where required by the specific or operating
conditions
C : for special applications
Table 3.5 gives a memory-help chart, developed in order to avoid forgetting the main
characteristics and properties of concrete to be involved in diagnosis. This information is
fundamental and will be partially explained and developed into the next sections.
3.3.3 Structural evaluation
It is sometimes necessary to evaluate the residual loading capacity of structural elements in

order to choose the repair technique. Different types of investigations can be realized, in
regards with the local situation of the structure:
tests on slab from the site if the structure is made of many repetitive elements;
loading tests on site, in order to determine residual flexure or axial rigidity;
dynamic evaluation tests, in order to evaluate rigidity from proper frequency measurement;
and
stress relaxation evaluation, with flat jacks and pressiometers into concrete sawed cracks.
It is also possible to evaluate the state of stress inside the concrete structure by means of flat
jacks, double sawing method and hole core stress relaxation [10]. All these techniques have to
be performed in order to attest the global stability of the structure before repair.
Generally speaking, the evolution of the geometry of the structural elements also hidden
needs to be surveyed, and displacements, rotations, and crack growth have to be measured. For the
global movements, geodetic methods, GPS as well as photogrammetry can be useful. In the case of
the cracks, extensometers and gauges can follow their potential activity and give fundamental
information on the cause(s) of defects.
3.3.4 Material evaluation
Material evaluation can be deduced from destructive and non-destructive testing. there is, in
fact, no standard definition for non destructive testing. to some, they are tests that do not alter
the concrete. for some, they are tests that do not impair the function of a structure, in which
case the drilling of cores is considered as a ndt test.
An interesting classification of ndt has been proposed by carino [11]. he proposes to
divide the various methods into two groups (table 3.6): (1) those whose main purpose is to
estimate strength; and (2) those whose main purpose is evaluate conditions other than
strength, i.e. to evaluate integrity.
Table 3.6: Methods for the evaluation of concrete material
In-place tests to estimate strength Non destructive tests for integrity

Rebound hammer Visual inspection
Ultrasonic pulse velocity Stress wave propagation methods
Probe penetration Ground penetrating radar
Pull-out Electrical/magnetic methods
Break-off Nuclear methods
Maturity method Infrared thermography
Visual inspection is one of the most versatile and powerful of the NDT methods, and it is
typically one of the first steps in the evaluation of a concrete structure. Visual inspection can provide
a wealth of information that may lead to positive identification of the cause of observed distress.
However, its effectiveness depends on the knowledge and experience of the investigator. Broad
knowledge in structural engineering, concrete materials, and construction methods is needed to
extract the most information from visual inspection. For example, cracking or spalling of a reinforced
concrete structure or member may occur for many reasons (Table 3.4). But a simple visual
inspection is insufficient to find out detailed information, such as whether reinforcement is
corroding. Cracking of the concrete cover zone may be present without any obvious sign of further
damage [1], such as rust staining or spalling. Equally, rust staining need not be accompanied by
cracking of the cover, and occasionally may be due to other factors than corrosion of the reinforcing
bar.
Useful guides are available to assist less experienced individuals in recognizing different types
of damage and determining the probable cause of the distress (ACI 201.1R,ACI 207.3R, ACI 224.1R,
ACI 362R).
A typical visual investigation might involve the following activities [11]:

perform a walkthrough visual inspection to become familiar with the structure;
gather background documents and information on the design, construction, maintenance, and
operation of the structure;
plan the complete investigation;
perform a detailed visual inspection; and
perform any necessary sampling or inplace tests.
3.3.5 Test methods and procedures
3.3.5.1 Non destructive testing for the estimation of strength
3.3.5.1.1 Schmidt rebound hammer
Due to its simplicity and low cost, the Schmidt rebound hammer is the most widely used non
destructive test device for concrete (related standard test method: ASTM C805 Standard Test
Method for Rebound Number of Hardened Concrete). While the test appears simple, there is no
simple relationship between the rebound number and the strength of concrete. In principle, the
rebound is affected by the movement of the end of the plunger in contact with the concrete. The
more the end of the plunger moves, the lower the rebound is. Thus the rebound number is likely to
be influenced by the elastic stiffness and the strength of the concrete [11]. Since the rebound
number is indicative of the nearsurface properties of the concrete, it may not be indicative of the
bulk concrete in a structural member. The results will be influenced by the moisture conditions, the
carbonation, the surface texture and the orientation of the instrument [11, 12].
In summary, the rebound number method is recognized as a useful tool for performing quick
surveys to assess the uniformity of concrete. However, because of the many factors besides concrete
strength that can affect rebound number, it is not generally recommended where accurate strength
estimates are needed.
3.3.5.1.2 Ultrasonic Pulse Velocity (UP-V)

The Ultrasonic Pulse Velocity method (related standard test method: ASTM C597 Standard
Test Method for Pulse Velocity Through Concrete) is a stress wave propagation method that involves
measuring the travel time, over a known path length, of a pulse of ultrasonic compressional waves
(Fig. 3.6). A timing circuit is used to measure the time it takes for the pulse to travel from the
transmitting to the receiving transducers. The speed of compressional waves in a solid is related to
the elastic constants (modulus of elasticity and Poisson's ratio) and the density. By conducting tests
at various points on a structure, locations with lower quality concrete can be identified by their lower
pulse velocity.
Figure 3.6: Schematic of ultrasonic pulse velocity method
Moisture content, presence of reinforcement parallel to the pulse propagation direction, and
presence of cracks or voids can affect the measured ultrasonic pulse velocity. It should be used for
estimating concrete strength only by experienced individuals. Like the rebound hammer test, the pulse
velocity method is very useful for assessing the uniformity of concrete in a structure. It is often used to
locate those portions of a structure where other tests should be performed or where cores should be
drilled.
3.3.5.1.3 Probe penetration
The probe penetration method (Fig. 3.7) involves using a gun to drive a hardened steel rod,
or probe, into the concrete and measuring the exposed length of the probe (related standard test
method: ASTM C803 Standard Test Method for Penetration Resistance of Hardened Concrete). In
principle, as the strength of the concrete increases, the exposed probe length also increases; that
means that the exposed length can be used to estimate compressive strength.
Figure 3.7: Schematic of conical failure zone during probe penetration [11]
Probe penetration is not strongly affected by nearsurface conditions, and is therefore not as
sensitive as the rebound number method to surface conditions [11].
3.3.5.1.4 Pullout test
The pullout test that is of interest here is the postinstalled version. In this method, an
insert with an enlarged head is sealed in the concrete. The insert and the accompanying conical
fragment of concrete are extracted by using a tension loading device reacting against a bearing ring
that is concentric with the insert. The force required to pull out the insert is an indicator of concrete
strength. Several techniques have been investigated in order to realize the anchorage of the insert
[11].
3.3.5.1.5 Breakoff test
This test (related standard test method: ASTM C1150 Standard Test Method for the BreakOff
Number of Concrete) measures the force required to break off a cylindrical core from the concrete
mass (Fig. 3.8).
Figure 3.8: Schematic of breakoff test [11].
The test subjects the concrete to a slowly applied force and measures a static strength of
concrete. The correlations between breakoff strength and compressive strength have been found to
be nonlinear [13], which is in accordance with the usual practice of relating the modulus of rupture
of concrete to the square root of compressive strength.
3.3.5.1.6 Pulloff test
The pulloff test is commonly used to estimate adhesion of repair systems on concrete structure (Fig.
3.9).
cohesive rupture adhesive rupture cohesive rupture
slurry metallic disc concrete
Figure 3.9: Effects of coring depth and diameter on the value of superficial cohesion of concrete [14]
Also, it can be used to investigate the integrity and cohesion of a concrete surface to be overlaid.
An experimental program has recently been conducted [15] to evaluate the influence of various test
parameters on the measured cohesion of a reference concrete surface: transfer plate thickness and
diameter, core drilling depth, speed of loading, adhesive type and thickness, and number of tests. A
multivariate statistical analysis of the test results has shown that the plate diameter and core depth
are the most significant parameters, presumably with threshold values, and that there exists a
synergetic effect between them (Fig. 3.10). Also, a minimum number of five tests can be defined,
based on the standard deviation and required level of confidence of the results.
5,0
4,5
4,5 Diameter 50 mm
4,0 Diameter 100 mm

Superficial cohesion (MPa)
3,5 3,2 3,1

3,0 2,8
2,5
2,5
2,1
2,0
1,5
1,0
0,5
0,0
Depth 0 mm Depth 15 mm Depth 30 mm
Figure 3.10: Effects of coring depth and diameter on the value of superficial
cohesion of concrete [15]
The results of this test will give not only an evaluation of the cohesion of the superficial layer
of concrete, but also an estimation of the concrete integrity. Presence of cracks parallel to the
surface, induced or not by surface preparation [16], will be easily detected. The related European
Standard procedure (EN15043 Products and Systems for the Protection and Repair of Concrete
Structures Part 3: Structural and NonStructural Repairs) requires a surface cohesion value of at
least 1.0 MPa and 2.0 MPa for nonstructural and structural repairs respectively. In any case, it
should be equal to the recorded substrate tensile strength.
3.3.5.1.7 Optic fibres
If accessible cavities are present, video endoscopes can be used for investigations. A fibre
optic probe consisting in flexible optical fibres, lens and illuminating system is inserted into drilled
hole, cracks, voids, etc. [12]. Observation of invisible parts of the structures can be performed, with
the possibility of multidirectional observation and record by means of a camera.
3.3.5.2 Nondestructive tests for Flaw Detection and Condition Assessment
The other types of NDT methods are those for flaw detection and condition assessment. In
this context the term flaw can include voids, honeycombing, delaminations, cracks, lack of subbase
support, etc. Recent research and development efforts on these methods have far exceeded those
for methods to estimate strength [11]. The research impetus has come primarily from the
transportation industry, since much of the highway infrastructure is need of repair as a result of
natural aging or damage resulting from corrosion of reinforcing steel or deterioration of concrete.
The techniques for flaw detection are based generally on the following simple principle: the
presence of an internal anomaly interferes with the propagation of certain type of waves. By
monitoring the response of the test object when it is subjected to these waves, the presence of the
anomaly can be inferred. The interpretation of the results of these types of NDT methods usually
requires an individual who possesses knowledge both in concrete technology and in the physics
governing the wave propagation.
3.3.5.2.1 Stress wave propagation methods
Tapping an object with a hammer or steel rod (sounding) is one of the oldest forms of non
destructive testing based on stress wave propagation (Fig. 3.11). The disturbance induced by impact
propagates through the solid as three different waves: a Pwave (normal stress), a Swave (shear
stress) and an Rwave (Rayleigh waves). The Pwave and Swave propagate into the solid along
spherical wavefronts. In addition, there is an Rwave that travels away from the disturbance along
the surface. In an infinite isotropic, elastic solid, the Pwave speed is related to the Young's modulus
of elasticity, E, Poisson's ratio , and the density, . The reflected wave R is directly proportional to
the specific acoustic impedance of the material and, consequently, to the modification of the
interface, particularly when it is air (crack, hole, etc).
The method is subjective, as it depends on the experience of the operator, and it is limited to
detecting near surface defects. Despite these inherent limitations, sounding is a useful method for
detecting nearsurface delaminations, and a related standard ASTM test procedure is available
(ASTM D4580).
Figure 3.11: Schematic figure of test using impact to generate stress waves with displacement
wavefront, amplitude spectrum and normalized amplitude spectrum [17].
Figure 3.12 illustrates the disperse generalized Rayleigh surface wave (Rwave) spectrum
analytical process. The received signal is analysed to obtain the dependence of phase velocity on the
frequency [12].
Figure 3.12: Schematic figure of testing configuration for spectral analysis of surface waves.
The principle of ultrasonic pulseecho (UPE) is based on the action of an electromechanical

transducer that generates a short pulse of ultrasonic stress waves that propagates into the object
being inspected [18]. Reflection of the stress pulse occurs at boundaries separating materials with
different densities and elastic properties (these determine the acoustic impedance of a material) [19,
20]. Numerous parameters limit the applicability of the method: the structures cannot be too
complex, the defects not too short or far from the surface. Moreover, the network will sometimes
limit the surface of investigation. The results actually obtained seem more efficient for the detection
of pop outs and voids, or for the global modification of concrete characteristics.
3.3.5.2.2 Infrared thermography
Infrared thermography is a technique for locating nearsurface defects by measuring surface

temperature (related standard test method: ASTM D4788 Standard Test Method for Detecting
Delaminations in Bridge Decks Using Infrared Thermography). The presence of an anomaly having a
lower thermal conductivity than the surrounding material will interfere with the flow of heat and
alter the surface temperature distribution. As a result, the surface temperature will not be uniform.
Thus, by measuring the surface temperature, the presence of the defect can be inferred. In practice,
the surface temperature is measured with an infrared scanner that works in a manner similar to a
video camera. The output of the scanner is a thermographic image of temperature differences (Fig.
3.13).
Figure 3.13: Comparison between the results of the GPR and thermography tests (to locate position
of large voids and inclusions of different materials) [21]
The procedure for performing an infrared thermographic survey to detect delaminations in

bridge decks is described in ASTM D4788 [11]. Infrared thermography is a global inspection method
which needs experts for interpretation of the results [11]. This permits large surface areas to be
scanned in a short period of time, which is an advantage over other methods that have been
discussed.
3.3.5.2.3 Ground Penetrating Radar (GPR)
Radar (acronym for RAdio Detection And Ranging) is analogous to the ultrasonic pulseecho
technique previously discussed, except that pulses of electromagnetic waves (short radio waves or
microwaves) are used instead of stress waves (Fig. 3.14).
The pulse travels through the test object and when it encounters an interface between
dissimilar materials, some of the energy is reflected back toward the antenna as an echo. By
measuring the time from the start of the pulse until the reception of the echo, it is possible to
determine the depth of the interface if the propagation speed through the material is known [11].
This method (ASTM D4748 Standard Test Method for Determining the Thickness of Bound
Pavement Layers using ShortPulse Radar) can lead to the detection of delamination and often for
locating reinforcing bars in the structure [22]. Results seem to depend on moisture and chloride
concentration in concrete.
Figure 3.14: Reflection of radar pulse at interfaces between materials with different relative
dielectric constants and antenna signal caused by arrival of echoes [11]
3.3.5.2.4 Electrical and magnetic methods for reinforcement
Information about the quantity, location, and condition of reinforcement is needed to

evaluate the strength of reinforced concrete members. Magnetic and electrical methods are used to
yield this information about embedded steel reinforcement.
Cover meters
Devices to locate reinforcing bars and estimate the diameter and depth of cover are known as cover
meters, or pachometers. These devices are based upon interactions between the bars and low
frequency electromagnetic fields. Commercial cover meters can be divided into two classes: those
based on the principle of magnetic reluctance and those based on eddy currents. Cover meters are
effective in locating individual bars.
Figure 3.15: Cover measurement on a structure of bar [23]
By using multiple measurements methods, the bar diameter can generally be estimated to
within two adjacent bar sizes if the spacing exceeds certain limits that are also dependent on the
particular meter. But when multiple bars are present within the detection range of the instrument
(Fig. 3.15), the increased signal received from the increased mass of steel invariably causes the
instrument to indicate a cover which is shallower than the true cover [23]. This results in a lower
accuracy of the results (Fig. 3.15), depending on the type of the test device used.
Corrosion activity
Electrical methods are used to evaluate corrosion activity of steel reinforcement. The Half
cell potential method includes a coppercopper sulfate halfcell, connecting wires, and a high
impedance voltmeter. The positive terminal of the voltmeter is attached to the reinforcement and
the negative terminal is attached to the coppercopper sulfate halfcell. A high impedance voltmeter
is used so that very little current flows through the circuit. The halfcell makes electrical contact with
the concrete by means of a porous plug and a sponge that is moistened with a wetting solution (such
as liquid detergent). There are numerous limitations to be considered when planning corrosion rate
testing [24]:
the concrete surface has to be smooth (not cracked, scarred, or uneven);

the concrete surface has to be free of water, impermeable coatings or overlays;
the cover depth has to be less than 100 mm;
the reinforcing steel can not be epoxy coated or galvanized;
the steel to be monitored has to be in direct contact with the concrete;
the reinforcement has not to be cathodically protected;
the reinforced concrete has not to be near areas of stray electric currents or strong magnetic
fields;
the ambient temperature has to be between 5 and 40C;
the concrete surface at the test location must be free of visible moisture; and
test locations must not be closer than 300 mm to discontinuities, such as edges and joints.
3.3.5.2.5 Nuclear (radioactive) methods
Nuclear (or radioactive) methods for non destructive evaluation of concrete involve the use
of highenergy electromagnetic radiation to gain information about the internal structure of the test
object (related standard test methods: ASTM C1040 Standard Test Methods for Density of
Unhardened and Hardened Concrete in Place by Nuclear Methods; BS 1881: Part 205
Recommendations for radiography of Concrete). These involve a source of penetration
electromagnetic radiation and a sensor to measure the intensity of the radiation after it has travelled
through the object. If the sensor is in the form of special photographic film, the technique is called
radiography. If the sensor is an electronic device that converts the incident radiation into electrical
pulses, the techniques is called radiometry.
Radiometric methods
There are two basic radiometric methods that use Xrays and gamma rays in non destructive testing
of concrete. In the transmission method, the amplitude of the radiation passing through a member
is measured. As the radiation passes through a member, the attenuation is dependant on the
density of the material. Direct transmission techniques can be used to detect reinforcement.
However, the main use of the technique is to measure the inplace density, both in fresh and
hardened concrete. In the backscattermethod, a radioactive source is used to supply gamma rays
and a detector close to the source is used to measure the backscattered rays. Backscatter
techniques are particularly suitable for applications where a large number of insitu measurements
are required. Since backscatter measurements are affected by the top 40 to 100 mm, the method is
best suited for measurement of the surface zone of a concrete element. A good example of the use
of this method is the monitoring of the density of bridge deck overlays.
Radiography methods
Radiography provides a radiationbased photograph of the interior of the concrete. From this
photograph, the location of the reinforcement, voids in concrete, or voids in grouting of post
tensioning ducts can be identified. The radiation passing through the test object is attenuated by
different amounts depending on the density and the thickness of the material that is traversed. The
main purpose of these methods is to see into concrete but they require extensive precautions, skilled
personnel and highly sophisticated equipment.
3.3.6 Methods for the determination of superficial porosity
3.3.6.1 Principles
Three major transfer mechanisms may occur in porous media, such as concrete substrate to
be overlaid: diffusion, permeation and capillary suction [25]. Capillary suction, one of the main
phenomena in creating contact, is directly in relation with the superficial porosity of concrete and is
governed by Washburn's equation [26] which gives the change of height penetration of a liquid into a
capillary as the square root of time and radius (9),
r . L . cos
lp = .t (9)
2
where lp is the change of the height penetration of the liquid, L is the surface free energy of the
liquid, r is the radius of the pore, is the contact angle, is the viscosity of the fluid and t is the time.
If we assume a porous material to have a large number of capillaries parallel to each other, with
different diameters and perpendicular to the surface of the liquid, the liquid will be absorbed with a
higher force by the shorter capillaries, according to Laplace [27]), but faster and in large quantities by
the largest pores, according to Washburn. It has been shown that the capillary absorption can be
increased by selecting a liquid characterised by a surface free energy L, just below the critical surface
tension of wetting c [25, 26].
The influence of the pore's entry is also clearly observed and will influence penetration rate
[28]. Evaluation of the porosity of the substrate is consequently important for the appreciation of
the situation. Capillary action test methods as well as more sophisticated analysis can be useful in
this aim.
3.3.6.2 Capillary suction coefficient
The most commonly used test to analyse water transfer at the interface is the capillary suction test
[28, 29]. The capillary suction test is described by several standards around the world: NBN b 14201
(Mortars), DIN 52617, prEN 13057 (Repair products), EN 4805 (Concrete admixtures) differ
essentially by the water level above the bottom surface of concrete specimen and the time when
measurement is taken. Mass change is usually measured after 5, 15, 30 and 45 minutes, as well as
after 2, 6 and 24 hours. Mass is measured on samples wiped off with a damp tissue. From the
capillary suction test it is possible to calculate the rate of water absorbed by the capillaries (Ec %), in
regards with the water absorbed under vacuum (Ev %). One can then calculate the relative
impregnation ratio (10) St, at time ti:
E
St = c . 100 (10)
Ev
This value is characteristic of the absorption rate of the substrate at any time vs. a complete filling of
the porous skeleton (corresponding to accessible porosity of concrete samples). Classical and
standardised tests are, however, not able to characterise the early behaviour of the composite
concrete substrate/repair material because the time interval defined for observations begins too late
and is too large. Adaptation of this test has been realized [28] to continuously record mass change of
concrete samples as of early contact with the liquid phase and evolution over time of liquid capillary
suction (Fig. 3.16).
An impregnation rate calculated between 0 and 20 seconds is an interesting discriminator

parameter for the evaluation of transport kinetics and liquid migration at the interface. Figure 3.11
illustrates the different behaviour of samples prepared by sandblasting and polishing, as well the low
absorption rate of limestone [28]. This test, however, requires that core samples be taken from the
concrete, in order to perform the test in the laboratory. Knowledge of the structural skeleton of the
voids can be completed by Mercury Intrusion Porosimetry.
0,6
0,5
)
g
(
e 0,4 Sandblasted
g
n Polished
a 0,3
h Limestone
c
t
h 0,2
g
i
e
0,1
W
0
0 10 20 30 40 50 60 70
Time (sec)
Figure 3.16: Water capillary suction between 0-60 seconds on sandblasted, polished concrete surface and
limestone [28]
3.3.6.3 Mercury Intrusion Porosimetry
Analysis of Mercury Intrusion Porosimetry (MIP) and nitrogen adsorption isotherms is based
on pore shape, i.e. cylindrical, inkbottle, etc. [30]. They concern lowradius capillaries, i.e. between
2 nm and 75 m, unable to absorb large quantities of water in a short period of time [28]. Only
superficial concrete (5 1 mm) is analysed, as it will be principally influenced by liquid movements.
Mercury Intrusion Porosimetry (2000 bars) gives granulometry gradation of the pores between 7.5
nm and 75 m (Fig. 3.17), as well as evaluation of specific surface, total porous volume and mean
radius.
100
90
) 80
(% 70
e
m
u
l 60
o
v 50
re
o 40
p
e
iv
t 30
a
l
u 20 Polished concrete
m Sandblasted concrete
u
C 10 Limestone rock
0
1 10 100 1 000 10 000
Pore diameter (nm)
Consequently, the potential effect of capillary entry shape on capillary action in repair techniques
needs to be simultaneously analysed by means of MIP and classical absorption tests.
3.3.6.4 Insitu tests
If it is impossible to core samples from the structure, some tests exist and permit to qualify
the porosity directly on the site. These systems can be used to measure the air and water
permeabilities, and the water absorption (sorptivity). In this last case, the volume of water
penetrating into the concrete is recorded at a constant pressure of 0.2 bars. As water introduced by a
syringe is absorbed by capillarity, the pressure inside under the piston tends to decrease; hence it
is maintained constant by pushing the piston through the cylinder (Fig. 3.18). A plot of the quantity
of water absorbed and the square root of time elapsed is linear and the slope of this graph
represents the sorptivity index (m/ min).
Figure 3.18: Insitu permeability test Method operation for water flow tests
The same type of test device is based on a watertight gasket, glued on the concrete substrate (Fig.
3.19). The valves of the housing are opened and the chamber is filled with boiled water. The valves
are closed and the top lid of the housing turned until a desired pressure is achieved (between 0 and
6 bars). The pressure selected is maintained with a micrometer gauge pressing a piston into the
chamber substituting the water penetrating in the concrete. This system can be used for evaluating
the extent of microcracking and porosities of the concrete skin of the finished structure.
Figure 3.19: Water permeation test

3.3.6.5 Methods for the determination chemical damages
The main cause of reinforced concrete deterioration in service is corrosion of the reinforcement, due
to a variety of causes (fig. 3.20). Depending on the processes at work, the following tests are
routinely used in combination: depth of carbonation and cover, and chloride content and halfcell
potential [1].
The depth of carbonation, by using phenolphthalein, is simply assessed either on site or in the
laboratory, followed by cover depth. This will establish the relationship between depth of
carbonation and cover around the structure for a given age, in order to define areas of risks.
Figure 3.20: diagnosis flow chart for corrosion of reinforcement [1]
Chloride ion content can be determined rapidly and inexpensively from the analysis of dust
from drilled samples taken from concrete. However, chloride contents can be highly variable around
a structure. Measurement of halfcell potential can be used to find suspect areas, from which dust
drilling samples can be taken.
In special environments (sewage plants, agricultural, marine and food infrastructures, etc),
other chemical investigations are needed to assess the presence of contaminants in the concrete.
3.4 FEASIBILITY ANALYSIS

It is known that the values measured from NDT, which characterize only one particular physical
property of the concrete, are influenced in different ways by the constituents of concrete and curing
conditions. For this reason [31], it is important to combine different testing methods (fig. 3.21).
Figure 3.21: Chart of the non destructive test methods [32]
In regards with the properties that have to be determined, it is often interesting to first
assess all the structure, in such a way that it is possible afterwards to point out zones where more
investigations are needed. The NDT methods are particularly interesting in this aim: for example, UP
V and rebound hammer are usually combined. Results of tests on surface of the concrete member
are combined with those concerning its whole thickness. The RILEM TC 43 CND [32] has provided
recommendations about the essential criteria for the establishment of beneficial combinations (ibid.
p.43) and examples are given about potential combinations of two or three tests (ibid. p 4748).
For sure, the strength relationship should be developed experimentally before using a test
method to estimate inplace strength, by comparison with destructive tests on cores from different
locations. Finally, the importance of having qualified operators cannot be overemphasized because
NDT are indirect methods by which the property or characteristic of primary interest is inferred by
measuring other properties or characteristics. A lack of understanding of the underlying principles
and the interferences associated with these methods [11] could lead to incorrect assessment of
concrete structures.
REFERENCES
[1] Robery, P.C. (1995) Investigation methods utilizing combined NDT techniques, Construction
Repair 9 (6) 1116.
[2] Ballivy, G. et Rhazi, J. (2000) Progrs rcents dans le domaine de lauscultation. Auscultation
et Instrumentation des Ouvrages de Gnie Civil , Facult des Sciences Appliques de Bthune.
[3] Molenaar, A.A.A. and Potter, J. (1997) Assessment of the reflection crack potential,
Prevention of Reflective Cracking in Pavements RILEM REPORT 18, Chapter 3 (E. and F.N. Spon,
London) 2841.
[4] Huang, Y. H. (1993), Pavement Analysis and Design, PrenticeHall, Englewood Cliffs, NJ.
[5] Uzan, J., T. Scullion, C. H. Michalak, M. Paredes, and R. L. Lytton, A Microcomputer Based
Procedure for Backcalculating Layer Moduli From FWD Data, Research Report 11231, Texas
Transportation Institute, Texas A&M University, July 1988.
[6] Scullion, T., and C. Michalak (1991), Modulus 4.0: Users Manual, Research Report 11234,
Texas Transportation Institute, Texas A&M University, January 1991.
[7] Yoder, E. J., and M. W. Witczak, Principles of Pavement Design, John Wiley & Sons, Inc., New
York, NY, 1975.
[8] Wei, C., and B. F. McCullough, Development of Load Transfer Coefficients for Use with the
AASHTO Guide for Design of Rigid Pavements Based on Field Measurements, Research Report
11693, Center for Transportation Research, The University of Texas at Austin, February 1992.
[9] Czarnecki, L. (2003) Mechanisms of Industrial Floors Deterioration Causes, Results and
Preventive Means, Fifth International Colloquium Industrial Floors 03, Esslingen, Germany
(Technische Akademie Esslingen, Ed. P. Seidler) 6572.
[10] Plumier, A. (1994) Pathologie et rparations structurelles des constructions Partie 2 :
Caractrisation dune structure existante , Dpartement Mcanique des Matriaux et des
Structures, Facult des Sciences Appliques, Universit de Lige (notes de cours), 38 p.
[11] Carino, N.J. (2003) Non destructive test methods to evaluate concrete structures. Sixth
CANMET/ACI International Conference on the Durability of Concrete, Special Seminar,
Thessaloniki, Greece, 75 p.
[12] Garbacz, A. and Garboczi, E. (2003) Ultrasonic evaluation methods applicable to polymer
concrete composites, United States Department of Commerce, Report no. NISTIR 6975,
Gaithersburg, MD.
[13] Barker, M.G. and Ramirez, J.A. (1988) Determination of Concrete Strengths with BreakOff
Tester, ACI Materials Journal, 85 (4), 221228.
[14] Courard L. (2005) Adhesion of repair systems to concrete: influence of interfacial typology and
transport phenomena. Magazine of Concrete Research, 57, 110.
[15] Courard, L. et Bissonnette, B (2004) Essai driv de lessai dadhrence pour la caractrisation
de la cohsion superficielle des supports en bton dans les travaux de rparation : analyse des
paramtres dessai , Materials and Structures, 37, 342350.
[16] Bissonnette, B., Courard, L. Vayburd, A. and Belair, N. (2006) Concrete removal techniques :
influence on residual cracking and bond strength , Concrete International, Dec.
[17] Pratt, D. and Sansalone, M. (1992) Impactecho Signal Interpretation Using Artificial
Intelligence, ACI Materials Journal, 89 (2), 178187.
[18] Cheng, C. and Sansalone, M. (1993) The impactecho response of concrete plates containing
delaminations: numerical, experimental and field studies, Mater. Struct., 26 (159), 274285.
[19] Garbacz, A., Czarnecki, L. and Clifton, J.R. (1999) Nondestructive methods to assess adhesion
between polymer composites and Portland cement concrete, Proceedings of the Second
International Rilem Symposium on Adhesion between Polymers and Concrete, Dresden,
Germany (ISAP99,Ed. M. Puterman and Y. Ohama, Rilem Publications), 467474.
[20] Dondonn, E. et Toussaint, P. (2002) Synthse sur lutilisation de limpactecho par la
Direction de lExpertise des Structures du Ministre Wallon de lEquipement et des
Transports , Bulletin des laboratoires des Ponts et Chausses, Numro spcial : Evaluations
non destructives pour le gnie civil, 239, 5162.
[21] Binda, L., Saisi, A. and Tiraboschi, C. (2000) Investigation procedures for the diagnosis of
historic masonries, Construction and Building Materials, 14, 199233.
[22] Rhazi, J., Laurens, S. and Ballivy, G. (2000) Insights on the GPR Non Destructive Testing
Method of Bridge Decks , Auscultation et Instrumentation des Ouvrages de Gnie Civil,
Facult des Sciences Appliques de Bthune.
[23] Alldred, J.C. (1995) Quantifying losses in cover meter accuracy due to congestion of
reinforcement, Construction Repair, 9 (1), 4147.
[24] Cady , P.D. and Gannon, E.J. (1992) Condition Evaluation of Concrete Bridges to
Reinforcement Corrosion. Volume 8: Procedure Manual, SHRPS/FR92110, National
Research Council, Wash. DC, 124 pp.
[25] Courard, L. (2000) Parametric Study for the Creation of the Interface between Concrete and
Repairs Products, Mater. Struct., 33 (225), 6572.
[26] Kinloch, A.J. (1987). Adhesion and adhesives: science and technology, (Chapman and Hall,
London) 441 p.
[27] Jouenne, C.A. (1980). Trait de cramique et matriaux minraux . Chapitre XII: capillarit et
tension superficielle, (Ed. Septima, Paris), 513575.
[28] Courard, L. and Degeimbre, R. (2003) A capillary action test for the investigation of adhesion
in repair technology, Canadian Journal of Civil Engineering, 30(6), 11011110.
[29] Justnes, H. (1995). Capillary suction of water by polymer cement mortars, Proceedings of the
RILEM Symposium on Properties and Test Methods for ConcretePolymer Composites,
Oostende (ICPIC, Ed. D. Van Gemert), 2937.
[30] Lecloux, A. (1981). Texture of catalysts. Catalysis: Science and Technology, (Ed. J.R. Anderson
and M. Boudart, Springer Verlag, Berlin), 171227.
[31] Teodoru, G.Y.M. and Mommens, A. (1991) Nondestructive testing in the quality control of
buildings : why, what and how ? Proc.fo the Int. Symp. On Quality Control of Concrete
structures, Gent (E &FN Spon, Eds. L. Taerwe and H. Lambotte), 367376.
[32] Facaoaru, R.C.L. (1993) Rilem Draft Recommandation for insitu concrete strength
determination by combined nondestructive methods, 43CND Committee, Mater. Struct., 26
(155), 4349.
Standards
AASHTO Guide for Design of Pavement Structures, American Association of State Highway and
Transportation Officials, 1993.
American Concrete Institute. ACI Committee 201.1R, Guide for Making a Condition Survey of
Concrete in Service.
American Concrete Institute. ACI Committee 207.3R, Practices for Evaluation of Concrete in Existing
Massive Structures for Service Conditions.
American Concrete Institute. ACI Committee 224.1R. Causes, Evaluation and Repair of Cracks in
Concrete Structures.
American Concrete Institute. ACI Committee 228.1R (1989), In-Place Methods for Determination of
Strength of Concrete.
American Concrete Institute. ACI Committee 228.1R (1995), In-Place Methods to Estimate Concrete
Strength.
American Concrete Institute. ACI Committee 228.2R, Non Destructive Test Methods for Evaluation
of Concrete in Structures.
American Concrete Institute. ACI Committee 362R, Sate-of-the-Art Report on Parking Structures
American Concrete Institute. ACI Committee 437 R, Strength Evaluation of Existing Concrete
Buildings.
American Concrete Institute ( surface roughness)
American Concrete Institute. ACI Committee 546. ACI Manual of Concrete Practice - Part 4:
Concrete Repair Guide, 1998, 41 p.
Belgian Standard NBN B 14-201 (1973) Test on mortars - Capillary suction test. Institut Belge de
Normalisation, Bruxelles.
Belgian Standard NBN B 15-225 (1984) Essais des btons - Mesure de la duret de surface - Indice
sclromtrique. Institut Belge de Normalisation, Bruxelles.
Belgian Standard NBN B 15-229 (1991) Essais des btons - Essais non destructifs - Mesure de la
vitesse du son. Institut Belge de Normalisation, Bruxelles.
Belgian Guidelines. G0007 (2002) Agreement and Certification Guidelines for Hydraulic Binder-
Based Repair Mortars. Belgian Union for Technical Agreement in Construction (UBAtc), Sector
Civil Engineering, Walloon Ministry of Equipment and Transportation, Lige, Belgium.
British Standard. BS 1881 :Part 207 (1992) Recommendations for the assessment of concrete strength
near-to-surface tests. British Standards Institution, London.
Code de bonne pratique pour le renforcement des chausses laide du bton de ciment (1991),
Recommandations CRR R63/91, Bruxelles.
Code de bonne pratique pour le dimensionnement des chausses revtement en bton de ciment
(1985), Recommandations CRR R57/85, Bruxelles.
German Standard DIN 52617 (1987) Determination of the coefficient of Building materials.
Bestimmung des Wasseraufnamekoeffizienten von Baustoffen, Deutsches Institut fr Normung,
Berlin.
European Standard ENV1504-9 (1997) Products and systems for the protection and repair of concrete
structures Definitions, requirements, quality control and evaluation of conformity Part 9:
General principles for the use of products and systems. Comit Europen de Normalisation,
Bruxelles.
European Standard prEN1504-10 (2002) Products and systems for the protection and repair of
concrete structures Definitions, requirements, quality control and evaluation of conformity
Part 10: Site application of products and systems and quality control of the works. Comit
Europen de Normalisation, Bruxelles.
European Standard prEN 1766 (1995) Produits et systmes pour la protection et la rparation des
structures en bton - Mthodes d'essais - Btons de rfrence pour essais, Comit Europen de
Normalisation, Bruxelles.
European Standard prEN 13036-1 (2000) Surface characteristics. Test methods - Part 1: Measurement
of pavement surface macrotexture depth using a volumetric patch technique. European
Committee for Standardisation, Brussels.
European Standard prEN 13057 (2000) Products and systems for the protection and the repair of
concrete structures - Test methods - Measurement of capillary absorption. European Committee
for Standardisation, Brussels.
European Standard EN 480-5 (1996) Admixtures for concrete, mortar and grout - Test methods - Part
5: Determination of capillary absorption. European Committee for Standardisation, Brussels.
4. BOND
J. Silfwerbrand (1), H. Beushausen (2), L. Courard (3)
(1) Swedish Cement and Concrete Research Institute (CBI), SE-100 44 Stockholm, Sweden
(2) University of Cape Town, Department of Civil Engineering, Cape Town, South Africa
ABSTRACT
Good bond is a key factor for providing monolithic action in bonded concrete overlays. This chapter
starts with three theoretical sections classifying bond in three groups (complete, uncertain, and
poor), defining bond strength, and describing the fundamental bond mechanisms. The main part of
this chapter covers a thorough description in chronological order of how the 13 most important
factors affect bond. All events from removal of deteriorated concrete to concrete placing and curing
and longterm exposure are investigated. There exist a number of different methods to determine
bond strength and the most frequent ones are briefly described in a separate section. It is shown that
there is an evident relationship between the two most frequently used families of test methods, i.e.,
between methods determining bond strength in tension and bond strength in shear. Furthermore, it
is shown that it is possible to provide durable bond in concrete overlays if all operations for concrete
removal, surface cleaning, concrete placing, and curing are conducted meticulously. The chapter is
ended by two sections devoted to design strength values in various international codes and
performance requirements that can be used for quality control.
Classification of Bond
Bond may be classified into three groups: (i) complete bond, (ii) uncertain bond, and (iii) poor bond
or debonding (Figure 4.1). Complete bond is a prerequisite for structural interaction and the
possibilities to consider at twolayer system as a monolithic one. Uncertain bond is the most difficult
type. Since the overlay is bonded to the substrate in certain areas, a high restraint is developed. This
restraint may together with overlay shrinkage lead to the development of wide cracks perpendicular
to and between two areas of bonding. In cases with 100 percent debonding, the overlay may be
considered as a slab on (stiff) grade.
Perfect bond
Debonding
Figure 4.1: Classification of bond into three groups: complete bond (left), uncertain bond (centre), and
debonding (right)
Definition of Bond Strength

The aim of a concrete repair is to restore the load carrying capacity and the stiffness of a concrete
structure or member. Consequently, monolithic action is the final objective. A prerequisite for
monolithic action is sufficient bond between the remaining concrete and the newcast overlay. The
characteristics of adhesion, or bond, can be perceived from two different angles [1]: firstly, the
conditions and kinetics of joining two materials, taking into account different bond mechanisms; and
secondly, the quantitative measure of the magnitude of adhesion, usually expressed in stress or
energy required to separate the two materials. For concrete overlays, bond strength is usually
defined as the tensile strength perpendicular to the interface plane. Bond strength in shear may also
be considered, given its inherent role and intrinsic relationship with the tensile behaviour [2], but
tensile bond strength is much easier to measures (see 4.5 below).
In practice the bond strength is determined by coring and pulloff tests. A failure stress can be
computed by dividing the maximum force by the area of the cross section. It is important to note
that the bond strength solely equals this failure stress if the failure occurs completely at the
interface. In all other cases (failures completely or partly in substrate or overlay), the failure stress is
only a lower bound of the bond strength.
Fundamental Bond Mechanisms

First, adhesion has to be clearly defined because of the duality of the term [3]: on one hand,
adhesion is understood as a process through which two bodies are brought together and attached
bonded to each other, in such a way that external force or thermal motion is required to break the
bond. On the other hand, we can examine the process of breaking a bond between bodies that are
already in contact. In this case, as a quantitative measure of the intensity of adhesion, we can take
the force or the energy necessary to separate the two bodies.
Adhesion has therefore two different aspects, according to whether our interest is mainly (1) in the
conditions and the kinetics of contact or (2) in the separation process. The intensity of adhesion will
depend not only on the energy that is used to create the contact, but also on the interaction existing
in the interface zone [4]. The mechanism of adhesion can be classified in two phases (Figure 4.2):
Adhesion
Specific adhesion Mechanical adhesion
Physico-chemical interactions
Thermodynamic approach
Chemical bonding
Figure 4.2: Principles of the theory of adhesion
Each individual model of adhesion can contribute to an explanation of the phenomenon, but is
unable by itself to give a comprehensive explanation [5].
Bond mechanisms all depend on the true surface area, as opposed to the geometric surface area,
and the contact surface area, also termed effective surface area (Figure 4.3). The effectiveness of
mechanical adhesion [6] is explained by the fact that the liquid will penetrate through the roughness
of the substrate and, after hardening, will induce cohesion by interlocking effect. Special preparation
technique can increase the true surface area [7] and contribute to the increase of potential
interaction sites. It ought to be mentioned that the true surface area is difficult to measure, not least
since it is dependent on the fineness of the measurement scale; the mathematicians call this
phenomenon fractals.
Geometrical surface True surface Effective surface
Figure 4.3: Geometrical, true, and effective surface areas between substrate and overlay
When the materials are in contact, the effective area, that means the surface where contact really
exists, will be a fundamental parameter to be taken into account to explain the adhesion process.
This is the result of the wetting procedure of the solid body by the liquid phase. The wetting
procedure can be explained as follows [5]: the surface energies of the solid and the liquid interact
each other and a change of the energy conditions occurs due to surface decrease of liquid/vapour
and solid/vapour interfaces while a new interface (liquid/solid) is created. In this point of view, the
contact angle is an interesting representation of this phenomenon: the lower the contact angle, the
better is the spreading on the surface and the more effective will be the intermolecular interactions
at the interface. The third step is the development of these interactions van der Waals forces
between the different phases. Their effect is based on the formation of electric fields of different
intensities, depending on the presence of permanent or induced dipole bonds, or only dispersion
bonds. They can be attributed to two different effects [8]:
(a) dispersion forces arising from internal electron motions which are independent of dipole
moments;
(b) polar forces arising from the orientation of permanent electric dipoles and the induction effect
of permanent dipoles on polarisable molecules.
Hydrogen bond forces can also be seen as a special case of dipole interactions: their range of actions
extends further than that of the other secondary forces. At a higher level of energy, chemical bonds
can occur when there is the development of covalent or ionic bond. This is the case if bonding
agents are intentionally used, particularly at the interface polymer/mineral substrate (e.g., silane
family products). Pigeon et al. [9] stated that the transition zone that is formed when new concrete
is cast against old concrete is very similar to bond between aggregates and cement paste.
Mechanical adhesion in tension differs significantly from that in shear. For example, a high interface
roughness usually improves the shear bond strength, whereas the tensile mechanical bond strength
primarily depends on vertical anchorage in pores and voids (Figure 4.4). In practice, the difference is
most obvious after debonding, if any.
=0
=0
Figure 4.4: Mechanical tensile and shear bond between substrate and overlay
4.4 FACTORS AFFECTING BOND
4.4.1 Concrete properties
Materials for bonded overlays include:
Portland cement mortar and concrete;

Silica fume mortar and concrete;
Polymer-modified mortar and concrete;
Steel fibre reinforced concrete; and
Pre-packaged repair mortar and concrete.
Of course, the selection of the overlay material influences the material properties of the overlay and,
hence, also the bond. However, this chapter is focused on bond, whereas Chapter 2 is devoted to the
overlay and the overlay materials. Consequently, this subsection is limited to a general discussion on
the influence of fresh and hardened material properties on the bond.
4.4.1.1 Fresh material properties

The fresh repair material properties are very important both for early age bond strength
development and bond durability. The workability and compaction of the freshly placed overlay
influence the ability to fill open cavities and voids on the substrate concrete surface, thus
determining the effective contact area between the two composites. Concrete of conventional
fluidity and workability is generally sufficient in this respect. Small repair patches are commonly
carried out with premixed, relatively stiff mortars, which are applied with a trowel. Selfconsolidating
materials (with high workability) are expected to lead to higher effective contact area and then to
higher bond strength.
4.4.1.2 Hardened material properties

The compressive strength of the repair material usually does not influence the bond strength
significantly. However, tensile strength is important as it affects crack development and, therefore,
the formation of boundary conditions that may support the initiation of debonding. Delatte et al.
[10] found that an increase in early age concrete strength increased both tensile and shear bond
strength significantly.
Overlay permeability may influence bond durability, for example very impermeable overlays result in
stresses at the interface when moisture from the substrate cannot migrate through the overlay [11].
The addition of polymers to cementitious repair mortars was found to result in better bond
characteristics on specimens subjected to extensive temperature cycles [12]. Chen et al. (1995)
measured a significant increase in shear bond strength with the addition of short carbon fibres to
repair mortar. They attributed the effect to the decrease in drying shrinkage and the resulting
decrease in interface stress. Granju [13] states that fibres enhance bond durability through the
control of crack development.
Repair patch dimensions, such as area and thickness, can affect bond durability due to the influence
on stresses resulting from differential movement between substrate and overlay. In general, large
repairs tend to crack more easily than smaller areas. According to Banthia and Bindiganavile [14],
thin repairs are more likely to debond than thicker repairs. On the contrary, Laurence et al. [15] have
concluded from their measurements that bond strength was not influenced by the repair thickness.
Equal bond strength does not necessarily mean that the likelihood of a bond failure is equal. The
failure likelihood is dependent on both bond strength and bond stress, and the thickness is likely to
influence the bond stress, e.g., by thickness dependent shrinkage stresses.
4.4.1.3 Overlay strength

Pigeon and Saucier [9] state that the interface between old and new concrete is very similar to bond
between aggregates and cement paste. According to their research, a wall effect exists between
overlay and substrate, resulting in a transition zone that creates a layer of weakness (Figure 4.5).
Cement grains
Interfacial transition zone
of the overlay Aggregates
Hardened
cement paste Interface
of the substrate
Figure 4.5: Transition zone between substrate and overlay, according to Pigeon and Saucier [9]
Van Mier [16] has summarized existing knowledge on interfaces between aggregates and cement
matrix. The bond mechanisms between aggregate and cement paste depend largely on the porosity
of the aggregate. Generally a thin layer of calcium hydroxide (CH) forms at the physical boundary
between aggregate and cement matrix, followed by a relatively open layer containing oriented CH
crystals, ettringite, and calcium silicate hydrate (CSH). This socalled contact or transition layer has a
high porosity. Van Mier explains this high porosity with absorption of mixing water at the surface of
aggregate particles, which increases the effective w/c ratio. According to his research, fracture
surfaces generally exist not directly at the physical boundary between aggregate and matrix but
rather slightly remote from the interface in the porous transition zone.
These mechanisms have not yet fully been investigated in relation to interface between concretes of
different ages but may be useful for the characterization of fundamental bond properties in
composite members. Misra et al [17] found a relation between air permeability at the interface and
bond strength between substrate and overlay, which could be linked to the effects described above.
Beushausen [18] found that interface failure in shear bond tests commonly occurs inside the
overlay very close to the interface, which was observed both on shortterm specimens and
specimens tested after more than two years. Interface shear bond strength was closely related to
compressive overlay strength, with the ratio between the two being time and overlay strength
independent at approximately 0.10 to 0.12 (compare 4.4.11). This supports the theory that an
interfacial transition zone exists as illustrated in Figure 4.5, representing a zone of weakness. On
wellprepared substrates, overlay strength is therefore one of the decisive factors for bond strength
and bond durability.
Removal of deteriorated concrete

The method of concrete removal has a major influence on both the surface of the remaining
concrete and the properties of the uppermost layer of the remaining concrete. Some removal
methods leave a rough and sound surface that promotes a good bond, while others may even
introduce microcracks to a certain depth in the remaining concrete. Some removal methods can
only remove a thin layer of concrete, while others have the ability to remove significant depth. Some
of the most frequently used methods for removing concrete are summarised in Table 4.1. Additional
methods, e.g., other handheld tools, springaction hammers, concrete crushers, drills and saws, non
explosive demolition agents, and jetflame cutting are described by ACI Committee 555. According to
Silfwerbrand (1990), only waterjetting (or hydrodemolition) has the ability to remove concrete
selectively, i.e., remove bad, damaged, and cracked concrete while leaving strong, sound, and
undamaged concrete.
Table 4.1: Methods of concrete removal

Removal Principle Concrete Important advantages Important disadvantages
method behaviour removal
capability
Action depth
(mm)
Sandblasting Blasting with No No microcracking. Not selective, leaves considerable
sands. sand.
Scrabbling Pneumatically Little No microcracking, no dust. Not selective.
driven bits impact (6)
the surface.
Shotblasting Blasting with steel Little No microcracking, no dust. Not selective.
balls. (12)
Grinding Grinding with Little Removes uneven parts. Dust development, not selective.
(planning) rotating lamella. (12)
Flame Thermal lance No Effective against pollutions The reinforcement may be damaged,
cleaning and painting, useful in smoke and gas development, safety
industrial and nuclear considerations limit use, not
facilities. selective.
Milling (scari Longitudinal Yes Suitable for large volume Microcracking is likely,
fying) tracks are (75) work, good bond if reinforcement may be damaged,
introduced by followed by water flushing. dust development, noisy, not
rotating metal selective.
lamellas.
Pneumatic Compressedair Yes Simple and flexible use, Microcracking, damages

(jack) operated large ones are effective. reinforcement, poor working
hammers chipping environment, slow production rate,
(chipping), not selective.
handheld
or boom
mounted
Explosive Controlled Yes Effective for large removal Difficult to limit to solely damaged
blasting blasting using volumes. concrete, safety and environmental
small, densely regulations limit use, not selective.
spaced blasting
charges.
Water High pressure Yes Effective (especially on Water handling, removal in frost
jetting water jet from a horizontal surfaces), degrees, costs for establishment.
(hydro unit with a selective, does not
demolition) movable nozzle damage reinforcement or
concrete, improved
working environment.
The surface of the remaining concrete needs to be free from microcracking; otherwise, the top layer
of the remaining concrete will constitute a zone of weakness. The amount of microcracking is
governed by the selected method of concrete removal. In principle, mechanical methods (hammers)
are likely to introduce microcracking, whereas waterjetting has been shown to be a lenient but
efficient method (Silfwerbrand, 1990). A comparison between surfaces treated by waterjetting and
pneumatic hammers, respectively, is shown in Table 4.2. Because of microcracking, the bond
strength was considerably lower in the case where the surface was treated with pneumatic
hammers. Field tests, however, have shown that the bond strength can reach satisfactory values, if
mechanical removal is followed by highpressure water cleaning. The field tests also included the
removal of a rutted top part of a concrete pavement by scarifying and placement of a new concrete
overlay. Average bond strength of 2.3 MPa was obtained [19]. Talbot et al. [20] and Carter et al. [21]
found that sandblasting subsequent to the use of heavy mechanical methods could remove the
damaged concrete and provide a sound interface. Wells et al. [22] and Warner et al. [23] achieved
good bond strength on surfaces that were sandblasted without prior roughening.
Table 4.2: Laboratory pull-off tests: influence of microcracking [24]

Interface treatment Presence of All tests Interface failures
microcracking
Number of Average Number of Average
cores failure cores failure
stress stress
(MPa) (MPa)
Waterjetting No 16 1.86 1 2.23
Pneumatic hammers Yes 16 1.10 5 0.94
Concrete removal behind rebars, rebar cleaning, and rebar replacement

According to Silfwerbrand [25], the interface between substrate and shotcrete should not coincide
with the reinforcement plane because it may affect bond durability under cyclic loading. According to
the bridge maintenance regulations of the Swedish National Road Administration [26], the entire
reinforcement bar has to be completely surrounded by new concrete if more than 30 percent of the
circumference has been exposed through concrete removal operations. The distance between bar
and substrate has to be equal to the maximum aggregate size of the overlay concrete plus 5 mm.
Vaysburd et al. [27] give a similar recommendation: concrete removal around the bar shall continue
to provide a minimum 19 mm clear space between the rebar and the surrounding concrete or 6 mm
larger than the maximum size aggregate in the repair material. Vaysburd et al. also discuss other
reasons than bond for exposing the rebar. The most obvious one is to clean the rebar from corrosion
products. The other is to remove the concrete surrounding the corroded rebar, concrete that can be
highly stressed due to the corrosion products volume increase.
Exposed reinforcement bars should be thoroughly cleaned from dust and corrosion products. The
Swedish National Road Administration [26] states that corroded reinforcement shall be replaced by
new with equal area. New bars thoroughly anchored should replace reinforcement bars that loosen
or have lost bond to the substrate concrete.
Cleaning after concrete removal

The single most important factor influencing bond is the surface cleanliness. A surface which is
contaminated at the time of overlay placement will produce poor bond characteristics. In the first
Swedish bridges repaired with waterjetting and bonded overlays in 1984 and 1985, coring showed
poor bond at several locations. In most cases, insufficient cleaning was the reason for this. Loose
particles were found in the interface between old and new concrete [24]. In order to warrant
cleanliness, the waterjetted surface needs to be cleaned twice. The first cleaning has to be done
shortly after waterjetting to prevent loose concrete particles, such as exposed unhydrated cement
surfaces, to bond to the surface. The second cleaning has to be carried out prior to overlay
placement to make sure that the surface is free from sand, oil, dust or other particles with origin
from the environment or the construction works. The best methods are hosing down with high
pressure water and the use of vacuum cleaners.
Surface properties
Interface roughness depends to a large extent on the method of substrate surface preparation.
Mechanical methods of concrete removal normally leave the substrate surface much rougher than
blast methods. The magnitude of surface roughness for concrete repairs is commonly measured in
mm. The most widespread test method is the sand area method (Figure 4.6) [28], in which sand of
known volume V is spread over the concrete surface to form a circle until all sand has settled in the
surface cavities. The roughness Rt can be calculated from the diameter d of the circle, using the
following equation:
4 V [mm 3 ]
Rt [mm] = (4.1)
(d [mm])2
d
Rt
Figure 4.6: Determination of surface roughness using the sand area method (Kaufmann 1971)
Other methods are used in Belgium and by the International Concrete Repair Institute (ICRI). Still
another surface roughness method using a sawtooth curve (Figure 4.7) was introduced by
Silfwerbrand [29]. More accurate surface profiles can be measured electronically using touchpins [7]
or laser techniques [30].
10 mm
2a
Figure 4.7: The surface profile is transferred to a saw-tooth curve with double amplitude 2a. The distance
between consecutive measuring points is 10 mm (Silfwerbrand, 1987, 1990)
The surface roughness used to be considered to have a major influence on the bond between old and
new concrete. Bond tests have, however, shown that surface roughness only has a minor influence
on the bond (Table 4.3). In a test series, bond to rough, waterjetted surfaces was compared with
bond to smooth, sandblasted surfaces [24]. The average bond strength was approximately equal, but
interface failures were more frequent on the sandblasted surface. The conclusion is that there might
be a threshold value. If the surface roughness is higher than the threshold value, further
improvement of the roughness does not seem to enhance bond strength. According to theses tests,
this threshold value ought to be close to the surface roughness of typical sandblasted surfaces.
However, further investigations using interface shear bond tests are recommended for the
verification of this theory for different bond mechanisms.
Table 4.3: Laboratory pull-off tests: influence of surface roughness [24]

Interface Surface All tests Interface failures
treatment roughness*
Double Number of Average Number of Average

amplitude cores failure cores failure
(mm) stress stress
(MPa) (MPa)
Waterjetting 7.7 16 1.86 1 2.23
Sandblasting 0.4 8 2.38 3 1.73
* Surface roughness was measured on surface profiles transferred to a sawtooth curve with two
characterising parameters: double amplitude and wavelength (see Figure 4.7).
The existence of a threshold value is evidenced by tests by Takuwa et al. [31]. They
studied the influence of surface preparation of the substrate concrete and defined increase of
area as a surface roughness parameter. They compared manual treatment, shot blasting
treatment and water-jetting with no treatment (Figure 4.8).
Mainz and Zilch [32] achieved high bond strengths on waterjetted surfaces with a roughness of > 1
mm. Similarly, Tschegg et al. [33] compared roughness of 1.75 mm to 0.65 mm on waterjetted
surfaces and found better bond characteristics for the rougher interface.
The laitance is a layer of weak and nondurable material containing cement and fines from the
aggregates, brought by bleed water to the top of the concrete. If the laitance is not removed prior to
overlay placement, it will lower the bond markedly. If the new overlay replaces old concrete, the
laitance will be removed together with the deteriorated concrete. However, if the overlay is placed
on unaltered parts, the removal of the laitance must not be forgotten. Usually, sandblasting will be
sufficient to remove the laitance.
The old weathered and carbonated concrete surface can be removed using waterjetting, thus
exposing a fresh and uncontaminated surface. However, long periods of time between substrate
resmoval and overlay placement may result in new carbonation of the exposed surface, especially in
dry environments. Test results by Gulyas, Wirthlin, and Champa [34] show that carbonation may
decrease the bond strength significantly. Block and Porth [35], on the contrary, found that substrate
carbonation does not affect pulloff bond strength.
The substrate temperature at the time of overlay placing was found to have a significant effect on
shear bond strength development. Cold substrate (4 oC) results in lower initial bond strength but
higher longterm bond strength, compared to substrates of higher temperature (21 or 38 C) [36].
2
(N/mm )
Average bond strength (MPa)
Av
er
ag
Manual treatment
e
Blasting treatment
bo
Water-jet method
n
No treatment
str
en
gt
Relative increase in area
Figure 4.8: The relationship between the increase of surface roughness (described as increase of substrate area in
percent) and bond strength [31]
One line in two perpendicular directions of the treated surface was measured every 0.1 mm using a
laser device. The area increase was determined by transferring the obtained zigzag lines. Note that
an area increase also was measured for the nontreated surface (indicated with a star).
Surface preparation
According to Vaysburd et al. [37], the process of concrete surface preparation for repair is the
process by which sound, clean and suitably roughened surface is produced on concrete substrates.
The workmanship is of uppermost importance. Besides removal of unsound concrete and all foreign
materials that may disturb bond development, the process also covers the opening of the substrate
concrete pore structure.
The influence of surface moisture on the bond between old and new concrete has been investigated
in many studies. A too dry surface may absorb water from the fresh concrete overlay giving risk of a
heterogeneous, porous zone close to the interface. If the surface is too wet, an overlay zone with a
high watercement ratio will develop close to the interface, which will lead to a local reduction in
overlay strength. Free water at the surface may destroy the bond completely. Zhu [38, 39] has found
experimental signs of optimal moisture, but the moisture influence on the bond was so small that it
was difficult to discern between moisture influence and scatter of test results. Li et al. [40] measured
the bond strength of repaired specimens after freezing and thawing cycles and found that different
repair materials correspond to different optimum interface moisture conditions at the time of
casting. There is widespread agreement to promote the "saturated substrate with dry surface" as
one of the best compromises [41].
Bonding agents
Bonding agents, e.g., Portland cement grout, latex modified Portland cement grout, and epoxy resins,
are sometimes used to improve bond [42]. However, bonding agents cannot compensate for bad
substrate surface preparation and may act as a bond breaker when used inappropriately [9, 11]. The
chapter authors opinion is that bonding agents should normally be avoided. The use of bonding
agents leads to two interfaces and thus to the creation of two possible planes of weakness instead of
one. Besides, a grout often has a high watercement ratio leading to a low strength and the risk of a
cohesive failure within the bonding agent itself. On the other hand, grouts may have an ability to
assimilate loose particles on an insufficiently cleaned surface. This assimilation may increase the
bond strength for this specific case. Bonding agents may improve bond strength for certain materials,
especially stiff repair mortars that cannot properly fill open pores and cavities.
Mechanical devices crossing the interface

In cases of poor or uncertain bond or in areas with specific demands (e.g., high shear or tensile
stresses, serious consequences in the case of failure), it might be necessary to strengthen the shear
and tension capacity by installing some kind of reinforcement crossing the interface. It is mandatory
that the reinforcement be sufficiently anchored in both remaining concrete and in the overlay. Two
Swedish solutions are given in Figure 4.9. It is important to note that the reinforcement units do not
work until the bond has broken. The reason is that the reinforcement unit has to be strained before
it carries more than a negligible part of the load. This is similar to ordinary reinforcement that mainly
carries load after concrete cracking.
> 300 mm
> 220 mm > 300 mm
a) b)
Figure 4.9: Example of dowel bars crossing the interface a) Nut, plate, and stud M20 in injected hole = 30
mm. b) Bent reinforcement bar = 16 mm in injected hole = 25 mm.
Concrete placement
The compaction is important to obtain a dense and homogeneous overlay as well as a good and
uniform bond. Especially, the compaction is important in overlays on rough surfaces to prevent the
development of air pockets in the valleys of the surface texture. Air pockets were found in some
cores taken from repairs in the early eighties [24]. The Swedish National Road Administration [26]
recommends the use of vibration pokers and vibration platforms.
Concrete curing
Early age shrinkage may result in overlay cracking. Cracks may initiate debonding due to the
formation of boundary conditions (free edges). Curing is therefore one of the most important factors
in reducing earlyage overlay and interface stresses. It prevents moisture loss and thereby reduces
early age shrinkage, leading to higher tensile strength at the onset of shrinkage (Figure 4.10).
Simultaneously, other advantages are gained: reduced risk of plastic cracking, higher strength,
improved durability, and better wear resistance. Paulsson and Silfwerbrand [25] recommend a
minimum of five days water curing on bridge deck overlays. Exposure to direct sunlight was found to
have a detrimental effect on shear bond strength even under proper curing conditions which
included wet burlap and plastic [10]. Li et al. [40] tested different repair materials and found that, for
some materials, wet curing, and for others, dry curing results in better bond strength. Schrader [11],
on the contrary, states that curing mainly affects the surface of a concrete repair, but has little
influence on the material or bond properties at a depth of more than 25 mm.
0,7
0,6 no water curing
Relative stress
0,5
0,4
0,3
0,2 7 days of water curing
0,1
0,0
0 100 200 300 400 500
Time (d)
Figure 4.10: Relationship between relative stress (equals ratio between maximum tensile stress and available
tensile strength) and time after casting - computational result according to Silfwerbrand [29]
Short-term bond properties

The development of early age bond strength is important for the structures ability to withstand
interface stresses induced by early differential movement between substrate and overlay. For
pavement and bridge deck overlays high early bond strength is usually required due to traffic and live
loads. According to Delatte et al. [10, 36], bond strength develops rapidly after placement, similar to
concrete compressive strength development. In their studies they suggest a concrete maturity
approach, which characterises bond strength development in relation to the concretes rate of
hydration rather than its age. Laboratory tests carried out by Silfwerbrand [42] have shown that the
bond strength development is rapid (Figure 4.11). At an age of 24 hours, the obtained bond strength
exceeded 1 MPa. The bond strength development was found to be faster than the development of
compressive strength of the concrete overlay. The tests were carried out indoors. At time of testing,
the old concrete had compressive cube strength of 54 MPa. The overlay had 28day compressive
cube strength of 63 MPa. The surface of the old concrete was waterjetted prior to overlay
placement.
3.0
2.5
Failure stress (MPa) 2.0
1.5
1.0
slab 1
slab 2
0.5
0.0
0 10 20 30 40 50 60 70 80 90 100
Time (d)
Figure 4.11: Tensile bond strength development according to Silfwerbrands test results [42]
Beushausen [18] measured the development of bond strength with time, using a modified version of
the direct shear test proposed by FIP [43] (compare 4.5). Test specimen size was 150150x50 mm
with both substrate and overlay having a depth of 75 mm. The interface was sandblasted with an
average roughness of 0.7 mm. All specimens were fully water cured at 23C. Different overlay
strengths were used (28day design strengths 10, 20, 30, and 50 MPa), the substrate compressive
strength being 48 MPa. Shear bond strength development was closely related to the development of
compressive overlay strength, with a relatively constant ratio between the two of approximately 0.10
to 0.12 over the whole test period for all specimens. Shear bond strength development is shown in
Figure 4.12. The test results represent the mean value of 4 to sax6 specimens after exclusion of
outliers.
50 MPa 30 MPa
20 MPa 10 MPa
3.5
Shear bond strength [MPa]
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0 5 10 15
Time [d]
Figure 4.12: Interface shear bond strength development for overlays of different compressive strengths
For fully bonded overlays it appears that the rate of bond strength development can be related to
that of mechanical overlay strength. This facilitates the design of interface shear strength for
structural overlays.
Carter et al. [21] state that bond strength develops more fully in the centre of an overlay as the
boundaries are especially subjected to cyclic stresses related to differential temperature and
moisture content.
Long-term bond properties

Differential movement between substrate and overlay, such as differential shrinkage and thermal
expansion and contraction, is commonly the main factor affecting longterm bond properties. It
ought to be emphasized that high initial bond strength does not necessarily lead to bond durability.
Proper substrate preparation, material selection and curing procedures should, however, normally
result in good longterm bond properties. Repaired beams and columns with wellbonded overlays
were shown to have structural capacities similar to that of monolithic members [29, 44]. Carter et al.
[21] state that well designed bridge deck overlays can be expected to provide more than 30 years of
service life if they are placed and cured correctly. Paulsson and Silfwerbrand [45] tested repaired
concrete bridge decks, which had been overlaid with steel fibre reinforced concrete and found a
slight increase in bond strength after a period of 9 years. Specimens subjected to 300 freezing and
thawing cycles were found to have similar bond strength as air cured specimens [40]. Talbot et al.
[20] investigated the influence of different interface textures and concluded that smooth surfaces as
well as sandblasted surfaces experienced a significant loss of bond strength with time. However,
surfaces which were roughened mechanically and subsequently sandblasted had good bond
durability. Debonding on vertical or overhead concrete repair patches usually leads to spalling and
hence to failure of the repair. Delaminated bridge deck overlays, on the contrary, may be rebonded
using epoxy injection [46].
Traffic vibrations
Normally, trafficinduced vibrations do not harm the development of bond between old and fresh
concrete. Independent researchers have even found that continuous and limited vibrations may
increase both overlay strength and bond strength, see, e.g., Silfwerbrand [42]. Heavy vibrations
starting a few hours after overlay placement should, however, be avoided. Manning [47] stated that
the best way of preventing heavy vibrations is to maintain a smooth riding surface and a smooth
transition at the expansion joints of the bridge.
Test Methods
4.5.1 General
The results and interpretation of interface bond tests depend to a large extent on the test method
used. Common bond test methods include interface shear, torsion and tension tests and a wide
range of possible test setups have been developed for laboratory testing. Interface shear strength
values obtained by different test methods may differ substantially as test results depend on
specimen size, test setup, loading rate, etc. Li et al. [40] investigated the size effect in bond tests and
concluded that smaller specimen sizes led to larger bond strength in prism splitting tests. A
comparison of test results obtained with different test methods, therefore, is problematic.
Most cracks in concrete develop under combined modes, of which the mixed mode 1 (tension) and 2
(inplane shear) is the most common. According to Van Mier et al. [48], most interface shear tests
will initiate in mode 1, and true mode 2 failures are very difficult to obtain. Robins and Austin [49]
proposed a bond failure envelope for concrete repairs based on the Mohr circle and related it to
interface shear and tension test results.
Many test methods have been developed to determine interface bond strength between concrete
substrate and overlay (Figure 4.13).
In a direct tension test, the specimen is pulled apart by loads applied perpendicular to the bonded
interface. The most common test method is the pulloff test, which can be carried out insitu (Figure
4.13 A). The pulloff test encounters several problems as the test results may be strongly influenced
by eccentricity in the load application and damage during coring [36]. Laboratory tensile strength
tests (Figure 4.13 B) are of course possible. They give especially reliable results when the test is
performed with the lowest possible eccentricity and no hinge in the loading chain. After debonding
initiation at the weaker point of the interface, hinges allow debonding propagation by flexure,
resulting in lower measured strength and higher scatter of the results.
A B C
D E F G
H J K L
M N O
Figure 4.13: Various test methods to determine interface bond strength between
concrete substrate and overlay
Tensile bond strength is hard to quantify on properly placed overlays, yet easy to quantify in cases
where the tensile interface strength is less than the tensile material strength. The results of pulloff
tests, therefore, often only give indication of the lower bound of the interface bond strength. The
main stress situation at the interface between substrate and overlay is that of a mixed mode of
tension and shear (compare section 2.3) and bond characterisation by sole tensile strength seems
problematic. Weber [50] indicates the usefulness of pulloff bond tests at locations where tensile
interface stresses are predominant, for example at the boundaries of a repaired member. Vaysburd
and McDonald [37] have evaluated the pulloff test method through 257 cores on 77 experimental
repairs. Their main results could be summarized in the following conclusions:
results obtained from pulloff tests were variable;

variation in exposure conditions did not appear to have a significant effect on failure mode or
bond strength;
only two of three pulloff test devices investigated ought to be used to evaluate bond strength;
load must be perpendicular to the repair surface, attached centrally to the core and increased
gradually with a specified speed (Vaysburd and McDonald [37] recommend 0.02 0.05 MPa/s,
Silfwerbrand [29] has used 0.064 MPa/s); and
the depth of the coredrilling into the substrate below the interface should be a minimum of 25
mm or onehalf of the core diameter, whichever is larger.
Silfwerbrand [2] developed an insitu test for torsional bond strength, which can be equated to
interface shear strength (Figure 4.13 C). Common shear bond tests apply a force parallel to the
interface (Figure 4.13 D). Modified shear bond methods have been presented by Pigeon and Saucier
[9] (Figure 4.13 E) and ElRakib et al. [51] (Figure 4.13 F). The pushout specimen [52] (Figure 4.13 G)
has the disadvantage of having two interfaces, which does not represent site conditions, thus making
the test very impractical The slant shear test (BS 6319) (Figure 4.13 H) measures bond strength under
a combination of shear and compression. Several researchers [53, 54, 10] indicated shortcomings of
this test with respect to unrealistic loading conditions and the vast number of parameters which may
affect the test results. The guillotine test (Figure 4.13 J) can be used both on cores and prisms.
Delatte et al. [10], however, reported difficulties with the alignment of the loading head for use on
cores.
4.5.2 Shear test methods
The disadvantage of most common shear test methods is the occurrence of an interface bending
moment due to force eccentricity. The FIP [43] developed a test method for interface shear strength,
in which the interface is theoretically subjected to pure shear forces (Figure 4.13.K). Robins and
Austin [49] developed different patch tests, which measure interface shear and tensile bond strength
under structural loading (Figures 4.13 L and M). The wedge splitting test device [33] characterises
bond by fracture mechanical parameters such as crack opening and specific fracture energy, and
tensile interface strength in bending (Figure 4.13 N). Li et al. [40] measured interface tensile strength
with a prism splitting test device (Figure 4.13 O).
In cases where destructive insitu bond testing is not appropriate, chain dragging or tapping the
surface with a hammer may be used to determine locations of debonding [21]. Lacombe et al. [55]
analysed bond characteristics of different repair materials visually and by a scanning electron
microscope. They evaluated the quality of the bond by observation of overlay compaction, cracks
and voids at the interface.
All shear test methods described above have the disadvantage in common that the test specimens
have to be prepared in the laboratory. In some cases, the specimens even have to be cast in the
laboratory. Consequently, a test method that can be conducted insitu is desired. The Department of
Structural Engineering, KTH, has developed such a method [2, 29]. The test specimen is of the same
dimensions as the one used insitu for pull off test. The difference is that a torsional moment is
applied to the core instead of a tensile force (Figure 4.13 C).
The apparatus for determining the shear bond strength consists of a stand, three steel bars, a
moment converter (25 times amplification), a torsion gauge, rotational and spherical bearings, a
cylindrical steel plate, and a data acquisition unit holding the maximum value (Figure 4.14). The stand
is fixed to the concrete surface with three expander bolts. The steel plate is glued to the top surface
of the drilled core that is still attached to the old concrete. The three bars serve as a moment
converter. The bearings prevent the development of normal forces and shear forces, and are fixed to
three holes in the steel plate. A torsional moment is applied to the top of the bars by using the
bending moments. During the test, a torsional moment T is successively increased until failure.
Assuming linear elastic behaviour, the relationship between shear stress and torsional moment can
be computed by the following equation:
16 T
= (4.2)
3
where, is the core diameter. For purely plastic materials, the relationship between shear stress and
torsional moment has the following expression:
12 T
= (4.3)
3
i.e., for a constant torsional moment, the shear stress for a plastic material is 25 percent less than
corresponding value for an elastic value. Materials having other constitutive laws ought to have
shear stress torsional moment relationships somewhere between these two extremes. Thus, the
error using either of the two expressions should be limited. Concrete is a brittle material and,
consequently, the linear elastic relationship (Eq. (4.2)) has been applied. Numerical analyses of the
stresses and strains within the test specimen would be desirable, but have not been included in this
investigation.
The shear stress equals the shear strength max when the torsional moment T reaches the failure
moment Tmax. Only in cases with complete interface failures, the shear strength is equal to the shear
bond strength. For other failure modes, the shear strength constitutes a lower bound of the shear
bond strength. At KTH, the core diameter has been 100 mm and the stress velocity approximately 0.1
MPa/s.
Possible sources of error in addition to the one due to simplifications when transferring torsional
moment to shear strength are normal forces that may occur simultaneously when applying the
torsional moment and calibration errors when transferring the gauge signal to a moment value. As
mentioned above, the bearings have been designed to minimize the influence of normal forces. A
rough estimation limits the sum of these latter errors to ten percent of the measured value. This is of
about the same magnitude as for pulloff tests measured insitu. It should be noted that shear and
tensile strengths of concrete both have a rather large inherited scatter. In order to use this test
method for standardised laboratory tests, further studies and further development of the test
apparatus are needed.
4.5.3 Comparisons between tensile bond strength and shear bond strength
Relating interface shear and tension tests is questionable as both bond mechanisms have
substantially different characteristics (compare section 2.2.1). However, Silfwerbrand [2] and Delatte
et al. [10] indicated a correlation between the two test methods. The latter measured a mean ratio
(shear bond divided by tension bond) of 2.04.
Measurements show that the shear bond strength is considerably higher than the tensile bond
strength. Analysing the test results available [2], the ratio between average torsional shear stress and
average tensile stress varies between 1.9 and 3.1. For waterjetted surfaces, the obtained shear bond
strength exceeds 3 MPa. In Japan, Sato (1989) has found a relationship between shear strength and
tensile strength slightly lower than 1.5 by tests on specimens subjected to solely tension, solely shear
or combinations of tension and shear (Figure 4.15).
Figure 4.14: Torsion test apparatus (Silfwerbrand, 1987)

Non-jointed
JJointed (test A)
Jointed (test B)
Shear stress (MPa)

Computed
Tensile stress (MPa)
Figure 4.15: The relationship between tensile stress and shear stress at failure according to
test results by Sato [56]
Evaluation
In the mid 1980s, more than 250 Swedish concrete bridge decks were repaired by means of waterjet
removal and a newcast, bonded concrete overlay. A few weeks after repair, the Department of
Structural Engineering, Royal Institute of Technology, tested the bond on more than 20 bridges [24,
29]. The average failure tensile stress was above 1.5 MPa. The bond strength might have been even
higher, since only a small percentage of the tested cores failed at the interface. If the failure occurs
elsewhere, the failure stress only gives a lower bound of the bond strength.
Between 1995 and 1999, a major research project was carried out at the Department of Structural
Engineering to study durable repairs [25, 45, 57, 58]. Seven bridges repaired in the mid 1980s were
selected. Tests on concrete strength, bond strength, and freezethaw resistance were carried out as
well as crack surveys and measurements on chloride penetration.
In previous sections, the most important factors influencing bond have been discussed. They are
connected to measures taken prior to, during, and shortly after overlay placement. The longterm
performance of the bond may, however, also be influenced by traffic loads and environmental
factors. The following factors were considered to be most important: (i) the number of heavy
vehicles, (ii) the stiffness of the bridge, in turn dependent on the structural system, (iii) the use of de
icing salts, and (iv) the number of freezethaw cycles. Within the bridge, the bond may be different in
zones with positive moment and negative moment, respectively. The seven bridges chosen have
different traffic volumes, different structural systems, locations in different climate zones, and
different needs for deicing salts (Table 4.4). Consequently, all factors were covered in the
investigation, even though all could not be studied separately.
Table 4.4: Bridges investigated (Silfwerbrand & Paulsson, 1998)
bridge deck repair
heavy vehicles
Climate zone
Percentage of
construction
deicing salt
daily traffic
Average
Year of
Year of
Use of
Bridge Structural system
Bjurholm Steel girders 5 1966 1985 1 440 12 occasional
Mlsund Concrete arch 2 1924 1986 210 10 occasional
Skellefte Steel girders 4 1960 1987 25 000 8 moderate to

intensive
Sdertlje Flat slab 2 1966 1989 ~10 000 10 intensive
Ume Steel girders 4 1949 1987 23 000 20 moderate to

intensive
Vrena Steel girders 2 1937 1987 170 1015 occasional
verboda Concrete arch 2 1946 1986 7 600 10 intensive
Note: Climate zone 2 means an annually frost amount of 300600 daydegree of frost, zone 4: 9001200, and zone 5: 12001500 (300
daydegree of frost = 30 days of 10 C or 60 days of 5 C).
Bond strength was measured on cores drilled from the concrete bridge decks. The cores had a
nominal diameter of 100 mm. At time of repair, the pulloff tests were carried out in situ. In 1995,
the pulloff tests were carried out in the laboratory some time after drilling. Between drilling and
testing, the cores were stored in sealed plastic bags. Before testing, the cores were sawn
perpendicularly to the cylinder axis. Steel plates were glued to the end surfaces. The cores were
placed in a testing machine. A loading rate of 500 N/s (0.064 MPa/s) was chosen. Failure stresses and
failure modes were registered. Most failures occurred in the old concrete. All tests are described in
the test report [25, 57]. The results are summarised in Table 4.5.
Table 4.5: Field tests on bond at time of repair and ten years afterwards (Silfwerbrand & Paulsson, 1998)
Bridge Number of cores tested Failure stress (MPa) Failure stress

at time of repair in 1995 at time of repair in 1995 increase (%)
Bjurholm 9 8 1.96 1.99 2
Mlsund 6 8 1.71 2.17 27
Skellefte 3 9 1.43 1.82 27
Sdertlje ? 7 > 1.5 2.05 ~ 35
Ume 4 9 1.56 1.61 3
Vrena 8 9 1.49 1.56 5
verboda 12 9 2.09 2.18 3
On all bridge decks investigated, the failure stress had increased slightly during the time after repair.
That means, that neither traffic loading, nor deicing salts and frostthaw cycles had deteriorated the
bond between old and new concrete. The conclusion is that a good and durable bond can be
obtained providing all works with concrete removal, surface cleaning and preparation, overlay
placement, and curing are carried out correctly and carefully.
Design Strength Values

Several codes deal with shear resistance at the interface between two concretes. Usually, the design
shear resistance consists of three parts: (i) the strength of the interface itself, (ii) a term dependent
on external compression forces across the interface, and (iii) a term dependent on the reinforcement
crossing the interface. In the following comparison, only the first part has been included. It should be
noted that higher shear resistance values may be used when verified by extensive testing.
According to the Swedish handbook for concrete structures [59], the design shear resistance at a
waterjetted, properly cleaned interface is equal to 0.4 MPa.
The American code ACI 31899 [60] deals with composite structures in Chapter 17. Waterjetted
surfaces are not dealt with. For clean surfaces, free of laitance, and intentionally roughened, the
design shear resistance is limited to 0.55 MPa.
According to CEBFIP MC 90 [61], the design shear resistance at the interface is proportional to the
compressive strength fccd, if the surface is rough or indented. The coefficient of proportionality is
equal to 0.06. Hence, if fccd = 25 MPa, e.g., the design shear resistance is 1.5 MPa. The code does not
cover waterjetted surfaces.
The draft of Eurocode 2 [62] states that the design shear resistance at the interface is proportional to
the design tensile strength fctd of the weakest concrete. The coefficient of proportionality is 0.45 for
rough surfaces and 0.5 for indented surfaces. If fctd = 1.6 MPa, e.g., the design shear strength is 0.72
MPa at a rough interface. The draft does not deal with waterjetted surfaces.
Comparisons between the design strength and the shear strength obtained for waterjetted surfaces
show a very large discrepancy. The code writers ought to discern between waterjetted and other
rough surfaces and increase the design shear strength for waterjetted interfaces considerably.
Another, maybe easier, way to deal with the problem is that the designer only considers shear
stresses due to shear forces and ignores shear stresses due to differential shrinkage. Most of the
latter stresses will disappear with time due to creep, and the likelihood that maximum shear force
and maximum differential shrinkage stress occur simultaneously is very low. The maximum shear due
to traditional shear forces (e.g., due to traffic, wind or snow) appears only once during the lifespan of
60 to 120 years, while shear stresses due to differential shrinkage peaks after some months. By
comparing stresses caused by only one type of loading (e.g., traffic) with a considerably reduced
strength, the result will still be far on the conservative side.
Performance Requirements
Standard EN 15043 [63] states that the tensile bond strength should equal or exceed 2.0 MPa on
structural repairs and 1.0 MPa on nonstructural ones. Considering obtained insitu measurements
and the fact that all deteriorated old concrete usually cannot be removed, these requirements may
be fairly difficult to fulfil. The Swedish National Road Administration with a long experience on
concrete bridge repair, has used and is still using [64] the following requirements:
m f v + 1.4 s (4.4)
x 0.8 f v (4.5)
where, fv is the required tensile bond strength equalling 1.0 MPa, m and s are the average and the
standard deviation (s = 0.36 MPa) of the measuring values, respectively, and x is a single measuring
value.
REFERENCES
Cited references
[1] Courard, L. (1999), How to Analyse Thermodynamic Properties of Solids and Liquids in
Relation with Adhesion? Proceedings, 2nd International RILEM Symposium ISAP 99, pp.9-19.
[2] Silfwerbrand, J., (2003), Shear Bond Strength in Repaired Concrete Structures. Materials &
Structures, Vol. 36, July 2003, pp. 419-424.
[3] Derjagin, B.V., Krotova, N.A., & Smilga, V.P. (1978), Adhesion of Solids. Studies in Soviet
Science: Physical Sciences, Plenum Publishing Corporation, New York, 455 pp.
[4] Courard, L. (2000), Parametric Study for the Creation of the Interface Between Concrete and
Repair Products, Materials and Structures, Vol. 33, January-February 2000, pp. 65-72.
[5] Fiebrich, M.H. (1994b), Influence of the Surface Roughness on the Adhesion between
Concrete and Gunite Mortars Overlays, Proceedings, 2nd Bolomey Workshop, Adherence of
Young and old Concrete, Unterengstringen, Switzerland (Aedification Verlag ,Ed. F.Wittmann),
pp. 107-114.
[6] Kinloch, A.J. (1987), Adhesion and Adhesives: Science and Technology. Chapman and Hall,
London, 441 pp.
[7] Courard, L. (1998), Parametric Definition of Sandblasted and Polished Concrete Surfaces,
Proceedings, IX ICPIC 98, Bologna, Italy, September 14-18th, 1998, pp.771-778.
[8] Fiebrich, M. (1994a), Grundlagen der Adhsionskunde, Deutscher Ausschuss fr Stahlbeton,
Heft 334, Beuth Verlag, Berlin 1994, pp. 75-90.
[9] Pigeon, M., and Saucier, F. (1992), Durability of Repaired Concrete Structures Proceedings,
International Symposium on Advances in Concrete Technology, Athens, 11-12 May 1992, pp.
741-773.
[10] Delatte, N.J., Williamson, M.S., and Fowler, D.W. (2000a), Bond Strength Development of
High-Early-Strength Bonded Concrete Overlays, ACI Materials Journal, no.97-M27, March-
April 2000, pp. 201-207.
[11] Schrader, E.K. (1992), Mistakes, Misconceptions, and Controversial Issues Concerning
Concrete and Concrete Repairs, Part 1, 2, and 3, Concrete International, September, October,
& November 1992.
[12] Atzeni, C., Massidda, L., and Sanna, U. (1993), Dimensional Variations, Capillary Absorption
and Freeze-Thaw Resistance of Repair Mortars Admixed with Polymers, Cement and Concrete
Research, Vol. 23, pp.301-308.
[13] Granju, J.L. (1996), Thin Bonded Overlays--About the Role of Fibre Reinforcement on the
Limitation of Their Debonding, Advanced Cement Based Materials, Vol. 4, pp.21-27.
[14] Bindiganavile, V, and Banthia, N. (2001), Repairing with Hybrid-Fiber-Reinforced Concrete,
Concrete International, June 2001, pp. 29-32.
[15] Laurence, O., Bissonnette, B., Pigeon, M., & Rossi, P. (2000), Effect of Steel Macro Fibres on
Cracking of Thin Concrete Repairs. Proceedings, 5th International RILEM Symposium on
Fibre-Reinforced Concretes (BEFIB 2000), Lyon, France, pp. 213-222.
[16] Van Mier, J.G.M. (1997), Fracture Processes of Concrete Published by CRC Press Inc, USA,
1997.
[17] Misra, A., Cleland, D. J., and Basheer, P. A. M. (2001), Effect of Different Substrate and
Overlay Concretes on Bond Strength and Interfacial Permeability, Concrete Science and
Engineering, Vol.3, No.10, 2001, pp.73-77.
[18] Beushausen, H. (2005), Long-Term Performance of Bonded Concrete Overlays Subjected to
Differential Shrinkage. Dissertation, University of Cape Town unpublished
[19] Silfwerbrand, J., & Petersson, . (1993), Thin Concrete Inlays on Old Concrete Roads,
Proceedings, 5th International Conference on Concrete Pavement Design & Rehabilitation,
Purdue University, West Lafayette, Indiana, USA, April 1993, Vol. 2, pp. 255-260.
[20] Talbot, C., Pigeon, M., Beaupr, and Morgan, D.R. (1994), Influence of Surface Preparation on
Long-Term Bonding of Shotcrete, ACI Materials Journal, Title no. 91-M56, Vol. 91, No. 6,
November-December 1994, pp. 560-566.
[21] Carter, P., Gurjar, s., and Wong, J. (2002), Debonding of Highway Bridge Deck Overlays,
Concrete International, July 2002, pp. 51-58.
[22] Wells, J.A., Stark, R.D., and Polyzois, D. (1999), Getting better bond in concrete overlays,
Concrete International, March 1999, pp.49-52.
[23] Warner, J., Bhuyan, S., Smoak, W.G., Hindo, K.R., and Sprinkel, M.M. (1998), Surface
Preparation for Overlays, Concrete International, May 1998, pp. 43-46.
[24] Silfwerbrand, J. (1990), Improving Concrete Bond in Repaired Bridge Decks, Concrete
International, Vol. 12, No. 9, September 1990, pp. 61-66.
[25] Silfwerbrand, J., and Paulsson, J. (1998), The Swedish Experience: Better Bonding of Bridge
Deck Overlays, Concrete International, Vol. 20, No. 10, pp. 56-61.
[26] Swedish National Road Administration, (2002), Technical Regulations for Bridges
Maintenance, Borlnge, Sweden, 2002, 90 pp. (In Swedish).
[27] Vaysburd, A.M., Sabnis, G.M., & McDonald, J.E. (2001), Interfacial Bond and Surface
Preparation in Concrete Repair. ICJ, Vol. 75, No. 1, pp. 27-33.
[28] Kaufmann, N. (1971), Das Sandflchenverfahren, Strassenbau Technik , 24, no.3, 1971, pp.
31-50.
[29] Silfwerbrand, J. (1987), Effects of Differential Shrinkage, Creep, and Properties of the Contact
Surface on the Strength of Composite Slabs of Old and New Concrete, Bulletin No. 147, Dept.
of Structural Mechanics and Engineering, Royal Institute of Technology, Stockholm, 131 pp. (In
Swedish).
[30] Schfer, H.G., Block, K., and Drell, R. (1996), Oberflchenrauheit und Haftverbund,
Deutscher Ausschuss fr Stahlbeton DafStb, Heft 456, Berlin 1996, pp. 75-94.
[31] Takuwa, I., Shitou, K., Kamihigashi, Y., Nakashima, H., & Yoshida, A., (2000): The
Application of Water-jet Technology to Surface Preparation of Concrete Structures, Journal of
Jet Flow Engineering, Vol. 17, No. 1, pp. 29-40.
[32] Mainz, J. and Zilch, K. (1998), Schubtragfhigkeit von Betonergnzungen an nachtrglich
aufgerauhten Betonoberflchen bei Sanierungs- und Ertchtigungsmassnahmen, Research
report, Technical University Munich, Germany, February 1998.
[33] Tschegg, E.K., Ingruber, M., Surberg, C.H., & Mnger, F. (2000), Factors Influencing Fracture
Behavior of Old-New Concrete Bonds, ACI Materials Journal, Vol. 97, No. 4, pp. 447-453.
[34] Gulyas, R.J., Wirthlin, G.J., and Champa, J.T. (1995), Evaluation of Keyway Grout Test
Methods for Precast Concrete Bridges, PCI Journal, Vol. 40, No. 1, Jan.-Feb. 1995, pp. 44-57.
[35] Block, K. and Porth, M. (1989), Spritzbeton auf carbonatisiertem Beton Haftzugfestigkeit bei
nachtrglichem Aufspritzen, Beton 7/89, pp.299-302.
[36] Delatte, N.J., Wade, D.M., and Fowler, D.W. (2000b), Laboratory and Field Testing of
Concrete Bond Development for Expedited Bonded Concrete Overlays, ACI Materials Journal,
no.97-M33, May-June 2000, pp.272-280.
[37] Vaysburd, A.M. and McDonald, J.E., (1999), An Evaluation of Equipment and Procedures for
Tensile Bond Testing of Concrete Repairs. Technical Report REMR-CS-61, US Army Corps of
Engineers, Waterways Experiment Station, Vicksburg, Mississippi USA, 75 pp.
[38] Zhu, Y., (1991), Evaluation of Bond Strength between New and Old Concrete by Means of
Fracture Mechanics Method, Bulletin No. 157, Dept. of Structural Mechanics and Engineering,
Royal Institute of Technology, Stockholm, 102 pp.
[39] Zhu, Y., (1992), Effect of Surface Moisture Condition on Bond Strength between New and Old
Concrete, Bulletin No. 159, Dept. of Structural Mechanics and Engineering, Royal Institute of
Technology, Stockholm, 27 pp.
[40] Li, S.E., Geissert, D.G., Frantz, G.C., and Stephens, J.E. (1999), Freeze-Thaw Bond Durability
of Rapid-Setting Concrete Repair Materials, ACI Materials Journal, Title no. 96-M31,
March/April 1999, pp.241-249.
[41] Saucier, F. and Pigeon, M. (1991), Durability of New-to Old Concrete Bondings. Proceedings
of ACI Int. Conf. on Evaluation and Rehabilitation of Concrete Structures and Innovations in
Design, Hong Kong, December, 1991 (ACI SP-128) pp. 689-705.
[42] ACI Committee 546, (1980), Guide for Repair of Concrete Bridge Superstructures, (ACI
546.1 R-80), American Concrete Institute, Detroit, 20 pp.
[42] Silfwerbrand, J. (1992), The Influence of Traffic-Induced Vibrations on the Bond between Old
and New Concrete, Bulletin No. 158, Dept. of Structural Mechanics and Engineering, Royal
Institute of Technology, Stockholm, 78 pp.
[43] FIP Federation Internationale de la Prcontrainte (1978), Shear at the Interface of Precast and
In-Situ Concrete, Technical report, August 1978
[44] Souza, R.F.F. de, and Silva Appleton, J.A. da (2003), Assessing the Structural Performance of
Repaired Reinforced Concrete Members, Proceedings, ICPCM, Cairo, Egypt, February 2003.
[45] Paulsson, J., & Silfwerbrand, J. (1998), Durability of Repaired Bridge Deck Overlays - Effects
of Deicing Salt and Freeze-Thaw Cycles," Concrete International, Vol. 20, No. 2, February
1998, pp. 76-82.
[46] Smithson, L.D., and Whiting, J.E. (1992), Rebonding Delaminated Bridge Deck Overlays,
Concrete Repair Digest, June/July 1992, pp.100-101.
[47] Manning, D.G. (1981), Effects of Traffic-Induced Vibrations on Bridge-Deck Repairs,
NCHRP Synthesis No. 86, Transportation Research Board, Washington D.C., 40 pp.
[48] Van Mier, J.G.M., Nooru-Mohammed, M.B., and Timmers, G. (1991), Experimental Study of
Shear Fracture and Aggregate Interlock in Cement-Based Composites, Heron, Vol. 36, No. 4,
Delft, Netherlands, 1991, pp.8-30.
[49] Robins, P.J. and Austin, S.A. (1995), A Unified Failure Envelope from the Evaluation of
Concrete Repair Bond Test, Magazine of Concrete Research, Vol. 47, No. 170, Mar., pp 57-68.
[50] Weber, M. (1971), Mechanischer Verbund zwischen Beton verschiedenen Alters mittels
Kunststoffen, Dissertation, Fakultt fr Bauingenieur- und Vermessungswesen der Universitt
Karlsruhe, Karlsruhe, 1971.
[51] El-Rakib, T.M., Farahat, A.M., El-Degwy, W.M., and Shaheen, H.H., Shear Transfer
Parameters at the Interface Between Old and New Concrete, Proceedings, ICPCM, Cairo,
Egypt, February 2003.
[52] Chen, P., Fu, X., and Chung, D.D.L. (1995), Improving the Bonding Between Old and New
Concrete by Adding Carbon Fibres to the New Concrete, Cement and Concrete Research, Vol.
25, No. 3, 1995, pp. 491-496.
[53] Emberson, N.K. and Mays, G.C. (1990), Significance of Property Mismatch in the Patch
Repair of Structural Concrete. Part 1: Properties of Repair Systems, Magazine of Concrete
Research, Vol. 42, No. 152, Sept., pp.147-160.
[54] Austin, S., Robins, P., and Pan, Y. (1999), Shear Bond Testing of Concrete Repairs, Cement
and Concrete Research, Vol. 29, pp. 1067-1076.
[55] Lacombe, P., Beaupr, D., and Pouliot, N. (1999), Rheology and Bonding Characteristics of
Self-Levelling Concrete As A Repair Material, Materials and Structures, Vol. 32, October
1999, pp. 593-600.
[56] Sato, R. (1998), Recent Technology of Concrete Pavement en Japan. Proceedings, S.
Nagataki Symposium on Vision of Concrete: 21st Century. Part of the 4th CANMET/ ACI/ JCI
International Symposium on Advances in Concrete Technology, Tokushima, Japan, Vol. 1998.6,
pp. 71-85.
[57] Paulsson, J. (1997), Effects of Repairs on the Remaining Life of Concrete Bridge Decks,
Bulletin No. 27 (Licentiate Thesis), Dept. of Structural Engineering, Royal Institute of
Technology, Stockholm, Sweden, 238 pp.
[58] Paulsson-Tralla, J. (1999), Service Life of Repaired Concrete Bridge Decks, Bulletin No. 50
(Ph.D. Thesis), Dept. of Structural Engineering, Royal Institute of Technology, Stockholm,
Sweden, 244 pp.
[59] Swedish National Board of Housing, Building and Planning and AB Svensk Byggtjnst (1994),
BBK 94. Concrete Structures, Part 1 Structural Design. Handbook. (Karlskrona and
Stockholm, Sweden). 185 pp. (Only available in Swedish).
[60] American Concrete Institute, (1999), ACI 318 (1999). Building Code Requirements for
Structural Concrete (ACI 318-99) and Commentary (ACI 318R-99). (Farmington Hills,
Michigan, USA). 391 pp.
[61] Comit Euro-International du Bton & Federation Internationale de la Precontrainte (1993),
CEB-FIP Model Code 1990. (Thomas Telford Services Ltd, London, 1993). 437 pp.
[62] European Committee for Standardization (2001), Eurocode 2. Design of Concrete Structures.
2nd draft. (Brussels, Belgium, January 2001).
[63] CEN (Comit Europen de Normalisation) (2001), prEN 1504-3, Products and Systems for the
Protection and Repair of Concrete Structures Definitions, Requirements, Quality Control and
Evalution of Conformity Part 3: Structural and Non-Structural Repair, Brussels, Belgium,
March 2001.
[64] Swedish National Road Administration (2004), General Technical Regulations for Bridges,
Publication No. 2004:56, Borlnge, Sweden, 2004. (In Swedish).
Other references
ACI Committee 555 (2002), Removal and Reuse of Hardened Concrete, (ACI 555 R-01), ACI
Materials Journal, Vol. 99, No. 3, May-June 2002, pp. 300-325.
Bernard, O. (2000), Comportement long terme des elements de structure forms de btons dges
diffrents . Thesis No. 2283. (Department of Civil Engineering, EPFL, Lausanne, Switzerland).
189 pp.
British Standards, BS 6319 (1984), Part 4, Slant Shear Test Method for Evaluating Bond Strength of
Epoxy Systems, British Standard Institution BSI, London, 1984.
Chausson, H. 1997. Durabilit des rechargements minces en bton : relation entre leur dcollement,
leur fissuration et leur renforcement par des fibres. PhD thesis (in French), Universit Paul
Sabatier, Toulouse, France : 198 pp.
Fdration Internationale de la Prcontrainte (1982), Shear at the interface of precast and in-situ
concrete. Guide to good practice. (Wexham Springs, Slough). 31 pp.
Ingvarsson, H., and Eriksson, B., Hydrodemolition for Bridge Repairs, Nordisk Betong (Stockholm),
No. 2-3, 1988, pp. 49-54.
Jonasson, J.-E., (1977) Computer programs for non-linear analyses of concrete in the view of
shrinkage, creep and temperature. Research report No. 7:77. (Swedish Cement and Concrete
Research Institute, Stockholm, Sweden). 161 pp. (Only available in Swedish).
Julio, E.N.B.S., Branco, F.A.B. Silva, V.D. 2004. Concrete-to-concrete Bond Strength. Influence of
the Roughness of the Substrate Surface. Construction and Building Materials: 18 675-681.
Silfwerbrand, J. (1987), Theoretical and experimental study of strength and behaviour of repaired
concrete bridge decks summary, Bulletin No. 149. (Dept. of Structural Mechanics and
Engineering, Royal Institute of Technology, Stockholm, Sweden). 16 pp.
Silfwerbrand, J. (1988), Concrete Repair with Shotcrete. Tests on Beams under Static and Fatigue
Load, Bulletin No. 153, Dept. of Structural Mechanics and Engineering, Royal Institute of
Technology, Stockholm, 77 pp. (In Swedish).
Silfwerbrand, J. (1989), Concrete Repair with Shotcrete, Proceedings, IABSE Symposium on
Durability of Structures, Lisbon, September 1989, pp. 785-790.
Silfwerbrand, J. (1997), Concrete overlays, Report No. 10, 3rd Edition 10. (Chair of Structural
Mechanics and Engineering, Dept. of Structural Engineering, Royal Institute of Technology,
Stockholm, Sweden). 65 pp. (Only available in Swedish).
Silfwerbrand, J. (1997), Stresses and Strains in Composite Concrete Beams Subjected to Differential
Shrinkage, ACI Structural Journal. 94 (4), pp. 347-353.
Silfwerbrand, J., & Sundquist, H. (1999), Drift, underhll och reparation av konstbyggnader.
Rapport nr 53, brobyggnad, institutionen fr byggkonstruktion, KTH, Stockholm, 262 s.
Strmdahl, C., (2000), The history of hydrodemolition, Concrete Engineering International. 4 (8),
pp. 32-36.
5. STRUCTURAL BEHAVIOUR
E. Denari (1), J. Silfwerbrand (2), H. Beushausen (3)
(1) MCS-IS-ENAC, Ecole Polytechnique Fdrale de Lausanne, Switzerland

(2) Swedish Cement and Concrete Research Institute (CBI), SE-100 44 Stockholm, Sweden
(3) University of Cape Town, Dept. of Civil Engineering, Rondebosch, South Africa
ABSTRACT
Bonded cement-based material overlays and their substrates constitute a hybrid or composite
structural system. The interaction of these two material layers (with different ages), with each
other, with the external boundary conditions (foundations, supports) and possible joints, and
under loading, defines the structural behaviour of this composite system.
The main actions governing this structural behaviour are (1) the differential deformations of the two
layers due to autogenous shrinkage, thermomechanical effects due to cement hydration and/or
external climatic influences and drying shrinkage, (2) settlements, and/or (3) imposed forces (dead
loads, traffic loads). These actions give rise to stresses which depend on the stiffness of the substrate
with respect to the new layer, and that eventually may result in failure either by transverse crack
propagation, by debonding or both.
Composite structures formed of building materials of different kinds and ages are very common:
slabs on grade, steelconcrete, woodconcrete, concreteconcrete, concrete repairs, cementbased
overlays, etc. However, until now, the causes of distress in these composite structures and their
design were mostly addressed by empirical approaches. This document shows how a common
approach might treat all these applications on the unique basis of the mechanical description of the
behaviour of composite structural members under restrained shrinkage.
5.1 Introduction
Bonded cementbased material overlays and their substrates constitute a hybrid or composite
structural system. The interaction under loading of these two material layers (with different ages),
with each other and with the external boundary conditions (foundations, supports) and possible
joints defines the structural behaviour of this composite system.
Composite structures formed of building materials of different kinds and ages are very common:
slabs on grade, steelconcrete, woodconcrete, concreteconcrete, concrete repairs, cementbased
overlays, etc. However, until now, the causes of distress in several of these composite structures and
their design were mostly addressed by means of empirical approaches. This document shows how a
common approach originally proposed by Silfwerbrand [1, 2] can treat all these applications on the
unique basis of the mechanical description of the behaviour of composite structural members under
restrained shrinkage.
Composite structural members are subjected to actions such as: (1) the differential deformations of
the two layers due to autogenous shrinkage, thermomechanical effects due to cement hydration
and/or external climatic influences and drying shrinkage, (2) settlements, and/or (3) imposed forces
(dead loads, traffic loads). These actions give rise to stresses which depend on the stiffness of the
substrate with respect to the new layer, and that eventually may result in failure either by transverse
crack propagation, by debonding or both. The early age effects such as the thermomechanical
effects are difficult to approximate with simplified approaches, due to the rapidly changing material
properties.
There is a complex interaction between the structural behaviour and the modes of failure. If de
bonding can be prevented, the hybrid structure behaves in a monolithic way and structural failure
will happen by transverse cracking. Otherwise, if debonding occurs, near cracks or joints, complex
failure patterns can be obtained.
In the case of restrained deformations of the overlay, a simplified approach is to summarize the
parameters related to the structure using the notion of degree of restraint of the overlay. Maximum
restraint is obtained when all degrees of freedom (flexural and axial) of the overlay are blocked, i.e.
the worst case scenario. The degree of restraint represents the loading level of the overlay with
respect to maximum restraint. This approach has been investigated by several authors with the
hypothesis of elastic materials. In the most general case, the structural response of a hybrid element
can only be determined by a finite element simulation taking into consideration all relevant
boundary conditions and material properties. The viscoelastic behaviour of the overlay, and if
applicable, of the substrate, significantly contributes to the stress relaxation under imposed
deformations. It can be taken into consideration by comprehensive numerical analyses, or by means
of simplified analytical models.
Both the simplified approach based on the degree of restraint and the advanced numerical
simulations can be used to predict the structural performance of composite systems, under imposed
deformations of the overlay, with the aim to select appropriate combinations of materials,
thicknesses, reinforcements and joints.
5.2 Actions
Note: The word "Actions" has to be understood here in a broad sense encompassing imposed forces
and dead loads, as well as imposed displacements or deformations.
At early age and during their service life, overlays and composite members are subjected to various
kinds of actions such as:
Imposed deformations autogenous effects at early age, drying shrinkage, thermo-

mechanical effects associated with hydration of binders (temperature gradients) at early
age or to temperature gradients over the long term (see chapter 3).
Imposed displacements settlements, seismic loads.
Imposed forces braking forces of vehicles, moving (fatigue) or gravity loads
permanent loads in storage facilities (see chapter 7).
The aim of this chapter is to provide an understanding of the mechanical behaviour in the case of
restrained deformations of the overlay, rather than provide a comprehensive design method. A more
detailed description of the deformations of overlay materials is given in Chapter 3 of this report. The
overall actions to be considered for design as well as usual design methods are discussed in Chapter
7, for which this chapter can be seen as background.
5.3 Performance of composite structures

Composite structural components are formed of layers of different properties and age connected in a
more or less monolithic way. The following applications belong to this category:
steel-concrete composite constructions;
partial replacement of existing concrete superstructures (kerbs on bridges);
overlays on old concrete substrates (beams, slabs, columns);
overlays on precast slabs; and
pavements.
In the case of overlays, the overall performance of a composite system after the casting of a new
layer on an existing substrate must be evaluated in the following terms:
the protective function of the new layer and its serviceability; and
the load-bearing capacity of the hybrid component and its behaviour at serviceability
and ultimate (limit) state.
Figure 5.1 after Bernard [3], shows the evolution of the performance of a typical hybrid structural
component formed of a new layer cast on an existing substrate. At early age, the partial restraint of
the deformations of the new layer gives rise to eigenstresses that decrease with time due to the
viscoelastic behaviour of the materials. The value of these eigenstresses can be considered as a
"penalty" imposed on the intrinsic performance of the new layer, in terms of a reduction of the
tensile stress that can be supported by this layer under service loads. This drop of the initial
performance of the new layer must be taken into consideration in the calculation of the mechanical
behaviour of the composite system. The mechanical behaviour of a composite structural member
under loads can be calculated by means of simple well known theories from continuum mechanics
and strength of materials. However the resistance to be considered must take into consideration the
history of the various layers at early age or due to shrinkage by means of eigenstresses. Finally,
eigenstresses tend to vanish with time as a consequence of the decreasing shrinkage rate of the
overlay and viscoelastic response of the materials. The time to decrease the maximum eigenstresses
by a factor 2 is however counted in years rather than in month.
Figure 5.1: Performance of composite structures [3]
5.4 Different forms of restraint and effect of joints
Restraint can be defined as all mechanical effects that counteract the deformations of a given body.
At the micro level of heterogeneous materials such as concretes, the aggregates act as a
restraint to the deformations of the cement paste under all kinds of shrinkage. The effect
of this "internal restraint" on the onset of stresses and microcracks was numerically
demonstrated among others by Sadouki [4]. The same applies to Steel Fibre Reinforced
Concretes, for which the fibres may significantly decrease the apparent shrinkage.
The frictional forces on the subgrade of slabs on grade are another major source of
restraint extensively studied in the literature. This effect is commonly described by
interface models with adapted Mohr-Coulomb models.
The flexural and axial stiffness of the substrate on which an overlay is applied act as a
restraint. The combined axial and flexural stiffness of the composite system formed of
substrate and overlay have to be considered to determine the degree of restraint of the
overlay.
The restraint by dowels is the main effect governing the structural behaviour of steel-
concrete composite beams, as well as any other combination of different layers of
different materials, linked by connectors.
Reinforcement bars and in some cases formworks also act as a restraint to the
deformations of cementitious materials at early age and long term, according to [3].
Finally, the static system of a structural member also provides a restraint which acts on
different degrees of freedom.
On the other hand, joints (artificial or cracks) in a composite structure act as a local release of the
restraint, with varying consequences on the overall behaviour, positive or negative. The combination
of the various kinds of restraint and of the joints defines the kinematic system.
5.5 Mechanical behaviour of composite structures with cementitious materials of
different ages
5.5.1 Overview of existing analytical models
In what follows, the focus will be put on the description of the mechanical response of composite
structural members under differential shrinkage. The existing analytical models can be classified on
the basis of three main features:
consideration of the flexural degree of freedom (effect of curvature);

consideration of partial debonding;
consideration of the viscoelastic behaviour of overlay and substrate.
There are 2 major internationally accepted analytical theories for the modelling of bonded concrete
overlays. These are the theories presented by Birkeland [5], which does not consider debonding, and
the one from Silfwerbrand [1, 6] which does. The models presented by AlonsoJunghanns [7],
Silfwerbrand [1,2], Bernard [3], Denari et al. [8], and Hartl [9, 10] are essentially adaptations and
modifications of Birkelands theory. The prestress analogy (introduction of concentrated forces at the
free ends of a composite member to model the effect of shrinkage) which is applied in the models
based on Birkelands theory, is convenient to use, especially because it is easy to understand and
easy to apply.
Beushausen [11] questioned both: (1) existing analytical models based on the prestress analogy, in
their way to describe the introduction of forces in a composite member subjected to differential
shrinkage of an overlay, and (2) the use of the Bernoulli principle in the derivation of most models.
This author presented a semiempirical analytical model for the analysis of restrained overlay
shrinkage stresses based on localised, i.e. nonlinear strain and stress conditions inside the composite
member. Extensive strain measurements were realized on composite specimens with various
geometries and aspect ratios (curvature free or prevented), with different surface preparations of
the substrate. As expected, for deep composite elements with aspect ratio l/h < 1, Bernoullis
principle of plane sections remaining plane after being stressed does not apply, Beushausen et al.
[1214].
On a general basis, one must keep in mind that shrinkage of an overlay corresponds to an imposed
displacement, not an imposed force or imposed distributed forces. It is clear that the displacement
due to shrinkage is partly restrained on all the length of the interface with the substrate. This
however does not imply that the resulting reaction of the substrate leads to constant shear forces
along the interface. At the contrary, for "slender geometries" with l/h superior to 5, following what
was shown by Jonasson [15] and Haardt [16], a simple finite element simulation of a drying overlay
on a substrate leads to concentrated shear forces in the interface, at the extremities of the member,
similar to the predictions of the "prestress analogy", which turns out to be well adapted to simply
represent a composite member with shrinkage of the overlay.
In what follows, the two main approaches will be referred to as:
perfect bond by Silfwerbrand [1, 2], further generalized by Bernard [3] and modified by Denari
et al. [8].
partial debonding, from Silfwerbrand [1, 17].
5.5.2 Normal stresses due to differential shrinkage in composite beams with complete
bond
Stresses due to restrained movements can principally be computed by the product of three factors
according to the following equation:
Stress = stiffness free strain degree of restraint

Consequently, all three factors are equally important. The stiffness is dependent on modulus of
elasticity but also on creep or relaxation. The free strain is the strain that a completely free member
would develop due to thermal or moisture changes, shrinkage, or any other internal or external
source causing volumetric change of the member material. The degree of restraint defines the
conditions of restraint as the ratio between the actual stress rest taking into consideration the
effective stiffness of the composite structure and the stress full that would occur in a totally
restrained composite structure:
rest
= (1)
full
Restraints can be associated to all degrees of freedom of a structure. For a composite beam, two
degrees of freedom can be realized, one axial, one flexural. Many structural engineers are not aware
of the fact that complete bond between overlay and substrate does not necessarily cause complete
restraint in the repaired concrete structure. The reason is that the stiffness of the remaining part of
the old structure is not infinite. The striving of the overlay to contract is only partly prevented by the
remaining part of the old structure. The absence of a complete restraint leads to substantial stress
reductions. Combined with creep these reductions will limit the maximum tensile stress below the
tensile strength and, hence, explain the absence of shrinkage cracking. The easiest way to explain
this is by studying a composite beam exposed to differential shrinkage. This analysis was done first by
Silfwerbrand [1] for rectangular layers of equal width, and generalized by Bernard [3] to distinguish
the contribution of the various degrees of freedom, for arbitrary cross sections of the two layers, as
shown on Figure 5.2, for a statically determinate beam, with new,2 [MPa]: tensile stress in the new
layer at the interface, []: degree of restraint, free []: mean shrinkage strain in the new layer, cog:
centre of gravity of the composite section, hnew, cognew, Anew, Snew, Inew, resp. height, centre of gravity,
area, static moment and inertia of the new section (overlay), hold, cogold, Aold, Sold, Iold, resp. height,
centre of gravity, area, static moment and inertia of the old section (substrate), and n=hnew/hold,
m=Enew/Eold, a=Anew/Aold, i=Inew/Iold, Icomp= inertia of the composite section.
Figure 5.2: Eigenstresses in a statically determinate composite member, combination of axial (N) and flexural
(M,2) release effects (adapted from [3])
The degree of restraint is calculated under the following hypotheses: linearelastic material
behaviour, Poissons ratio = 0, crosssection of the new layer is a rectangle, the crosssection of the
substrate can be of any shape, and plane sections remain plain (hypothesis of Bernoulli), perfect
bond between new layer and substrate. As a consequence, the calculations only apply for the case of
slender composite beams (l/h < 5). For deep elements such as walls on slabs, with l/h < 5, similar
principles can be applied. The calculation of the strains and stresses has however to be adapted.
The principle of the analysis (prestress analogy) consists in determining the tensile force Nt that is
necessary to compensate the free shrinkage deformation free in the new layer. The tensile force is
balanced in the composite member by a compressive force Nc and a bending moment Mc acting at
the centre of gravity (cog) of the composite section. The stress state in the composite element is
determined by the superposition of the resulting effects of Nt, Nc and Mc on the composite cross
section. The resulting stress is the sum of the stress in the case of a total restraint, corresponding to
the effort Nt, plus the relaxing effects due to the axial (Nc stress N) and flexural (Mc stress M,2)
degrees of freedom. The axial release effect inducing the stress N is constant throughout the overlay
thickness. This is not the case of the flexural release effect inducing the stress M which is maximum
at the top of the overlay and has its smallest value in the overlay nearby the interface. In order to
estimate the maximum overall stress in the overlay, the flexural release effect is evaluated at its
location of minimal value, i.e.: nearby the interface, and is noted M,2.
In his original analysis, Bernard [3] proposed to define the degree of restraint as the sum of "1" plus
the axial factor N plus the flexural factor M: both factors are negative in their individual expressions.
In the following, this approach is slightly modified, to define the degree of restraint as "1" minus the
axial release minus the flexural release, (Equation (2)), to associate the notion of degree of restraint
to the effect of release of the stresses for each degree of freedom available, in a more
straightforward way, [8].
new ,2 full + N + M ,2
= = = 1 N M (2)
full full
The individual expressions of the release factors are adapted from [3]:
N ma 1 1 (3)
N = = = =
full ma + 1 1 + 1/ ma 1 + Eold . Aold
Enew . Anew
M ,2 N t (cog new cog ) 1
M = =
full W2 Enew . free. (4)
Anew (cog new cog ) [ m ( hold cog )]
=
[ I old + Aold ( cog cog old ) + m ( I new + Anew ( cog new cog ) )]
2 2
With W2: resisting moment of the composite section, at the level of the interface, Icomp: inertia of the
composite section, y2, lever arm between cog and interface (location of stress new,2)
Sold + m S new
cog = Centre of gravity (cog) of the composite section (5)
Aold + m Anew
[ I + Aold ( cog cog old ) + m ( I new + Anew ( cog new cog ) )]
2 2
W2 = I comp / y2 = old (6)

m ( hold cog )
It is worth mentioning that the release factor N associated to the axial degree of freedom
corresponds to the approach proposed by AlonsoJunghans [7].
The graphical representation of equations (2) to (6) is shown in Figure 5.3 for m = 0.71 (Enew = 25
GPa and Eold = 35 GPa) and rectangular sections of similar width for the old and new layers (bold=bnew).
The axial release N increases in a monotonic way when the thickness of the new layer increases. The
flexural release M first increases, reaches a maximum and then decreases down to 0 when the
centre of gravity (COG) of the composite cross section enters the new layer, to become then
negative. Two representations are shown, one a), with the ratio of the layer thicknesses n, and the
other b) with the ratio between the thickness of the new layer and the total thickness of the
composite section as xaxis.
The range of most cases encountered in practice corresponds to the domain shown on Figure 5.3a
with parameter n on the xaxis. For the chosen set of parameters, the global restraint varies in a
significant way for values of n smaller than 0.3. For values of n larger than 0.3, the overall degree of
restraint is almost constant and equal to 0.5 to 0.6.
a) b)
Figure 5.3: Axial and flexural effects on restraint as a function of a), n = hnew/hold , and b), = hnew/(hold+hnew),
with m = 0.71 and bold = bnew, after [8]
Silfwerbrand [1, 2] generalized this method for various combinations of materials, as shown
on Figure 5.4a, and to different boundary conditions, [17]. When the flexural degree of
freedom is almost or fully blocked (for example in the middle spans of hyper static multiple
span beams), only axial release acts and the degree of restraint is given by = 1-N, as
illustrated in Figure 5.4b.
a) b)
Figure 5.4: a) Degree of restraint as a function of relative thickness and relative stiffness, after [1, 2], b) Effect
of statical boundary conditions with m = 0.71 and bold = bnew
Concrete viscoelasticity has a beneficial effect on the developed stresses due to differential
shrinkage, in terms of relaxation. An engineering approach to estimate this effect is to replace the
modulus of elasticity Enew of the overlay with a fictitious modulus of elasticity Enew* = Enew/(1 + new)
and Eold with Eold* = Eold/(1 + old) where, new and old are the creep coefficients of the overlay and
base concrete, respectively. The maximum tensile stress in the overlay may be computed by the
following equation (7) according to Silfwerbrand [2, 18].
max =
(
m * (1 ) m * (1 ) + 2 ( 3 + )
3
) Enew E
free = new free
(7)
(
m * + ( m * 1) m * (1 )
4 4
) 1 + new 1 + new
with, m* = Eold = Eold /(1 + old ) = Eold 1 + new 1 + new since Enew Eold and old 0.
*
Enew* Enew /(1 + new ) Enew 1 + old
Since the development of drying shrinkage starts as soon as the curing of the newcast overlay is
ended, also the stresses due to differential shrinkage start to develop at an early stage. This means
that also the creep starts at an early stage, before the concrete is mature. This will in turn lead to a
large amount of creep. Creep coefficients of 5 to 6 have been measured, Silfwerbrand [2, 18]. The
beneficial influence of creep is shown in the following example of computation of stresses due to
differential shrinkage in a repaired concrete beam.
Geometrical and material data: = 2/7 ( 0.286), Enew = Eold = 35 GPa, free = 0.45 mm/m.
Case a) Full restraint = 1 full = Enew free = 15.8 MPa.
Case b) Beam theory with neglected creep m = 1, = 0.452 and full = 7.12 MPa.
Case c) Beam theory with creep coefficients 1 = 4 and 2 = 0 m = 0.2, = 0.733, and full = 2.31
MPa.
The maximum normal stress diminishes from 16 to 2.3 MPa. The latter value is of the same
magnitude as the tensile strength of concrete. It explains why some overlays are crackfree while
cracks are visible in others. The creep factors assumed in this example are theoretical estimates to
illustrate the possible effect of relaxation of stresses due to viscoelastic effects.
One must however also mention that the resistance of cementitious materials subjected to long term
tensile actions with low loading rates such as shrinkage is significantly lower than the resistance
obtained from quasi static tensile tests. This effect has to be considered when evaluating the risk of
cracking at early age or long term in bonded overlays.
Beushausen et al. [19] estimated the relaxation in overlays from experimental tests on composite
concrete/concrete members. Strain measurements on free and restrained materials and time of
cracking were used to determine the relaxation at a given time. Values of 40 to 50 % were found.
More generally speaking, these authors propose an incremental method to determine the relaxation
factor as a function of time.
A more precise analytical approach of the effect of linear viscoleasticity on the relaxation of stresses
in composite members can be found in the multilayer theory from Huet [20], adapted from the
elastic case from Siestrunck et al. [21].
Finally, the viscoelastic response depends on the stress level above certain limits. There is no general
agreement in the literature on the threshold value above which viscoelastic response would be
significantly influenced by the stress level. Gustch et al. [22], Horimoto et al. [23] tend to show that
the relaxation under tension is not significantly influenced by the stress level. Denari et al. [24]
showed that relaxation starts to deviate from a linear response at load levels around 50 % in Wedge
Splitting Tests. However, for this specific testing geometry with a strong strain gradient, the material
enters the softening domain at the notch tip for load levels close to 50 %. At this level, the deviation
from nonlinearity corresponds to the beginning of activation of the tensile strain softening
(equivalent load level 100 % under uniaxial tension).
Under creep, it is well known that deviation from a linear viscoelastic response start to be clear
above 40 to 50 % under compression and tension.
On another hand, for high degrees of restraint, stresses can very quickly be in the range of the quasi
static tensile strength. At those levels of loading, linear viscoelastic models do not apply anymore and
any substantial increase in the viscoelastic response is likely to be due to ongoing damage.
Reliable relaxation tests under tension, at various load levels pre peak are extremely difficult to
perform and constitute an interesting and necessary challenge for future research.
5.5.3 Shear stresses due to differential shrinkage in composite beams

The simple beam model (described in 5.5.2) assuming complete bond between overlay and base
(substrate) is easy to use and gives decent estimations of normal stresses in the interior parts of
overlay and base. However, in reality horizontal cracking leading to debonding sometimes is
observed especially in vicinity to the vertical borders and joints of the repaired structure. In order to
estimate the probability for debonding or explain its presence if it already exists, we need to be able
to estimate shear stresses along the interface between overlay and substrate. Jonasson [15] has
developed a simple model that also has been used by FIP [25]. Jonasson made computer
investigations and found that the shear stress has its maximum magnitude at the boundary of the
structure and diminishes approximately linearly to zero at a distance approximately equal to three
times the thickness of the overlay (Figure 5.5).
Figure 5.5: Shear stress distribution according to [15]
By utilizing the symbols presented above, the maximum shear stress max is given by the following
equilibrium equation:
2 F
max = (8)
3 bh
Where h and b are the thickness and width of the overlay, respectively, and F is the resultant to the
normal stresses in the overlay, i.e.
F = ( z ) b dz (9)
h
For the simply supported or free composite beam (Fig. 5.3), the maximum shear stress could be
computed by the following equation (if creep is considered):
max =
(
2 m 1 4 + 6 2 3 3 + ( m 1)(1 ) 4 E
1 cs
)

( )
(10)
3 m + (m 1) m(1 )
4 4 1 + 1
5.5.4 Normal and shear stresses due to differential shrinkage in composite beams with
incomplete bond
A shortcoming with the simple beam model described in 5.3 is that it leads to constant normal
stresses along the beam in cases with symmetrical boundary conditions. Since the shear stress is the
derivative of the normal stress with respect to the length coordinate, constant normal stresses result
in zero shear stresses. In reality, shear stresses appear in the vicinity to vertical boundaries, vertical
joints, and vertical cracks, if any.
Figure 5.6: Composite beam with incomplete bond between overlay and base (substrate) [1, 6]
Silfwerbrand [1, 6] has developed an engineering beam model considering incomplete bond between
overlay and base. By using this model, it is possible not only to compute normal stresses and strains
along the beam, but also shear stresses (x) and horizontal slip u(x) along the interface. This model
uses a linear relationship between shear stress and horizontal slip between overlay and base, i.e.,
(x) = Ku(x), where K is a constant of proportionality.
The shear stresses can be computed by the following equation [6]:
( x) =
(
m (1 ) 1 3 (1 ) + ( m 1)(1 )3 ) E1 h
cs L
sinh x (11)
(
m + (m 1) m(1 )
4 4
) 1 + 1 L cosh L / 2
with, (L) 2 =
(
m + (m 1) m(1 )4 4 ) 1 + 1 KL2
( ) (12)
m (1 ) 1 3 (1 ) + (m 1)(1 )3 E1 h
The engineering problem is to estimate the coefficient K and, hence, the governing nondimensional
product L. Direct tests are difficult since the horizontal slip u is difficult to measure and likely to be
dependent on loading rate. The interesting rate is given by the shrinkage velocity and such slow tests
are if not practically impossible at least economically impossible. However, the effect of varying values
of the product L could easily be studied (Figure 5.7), with z*=z coordinate at interface.
Shear stresses increase with increasing L value and normal stresses approach the constant distribution
given by the simple beam model with complete bond (5.3).
a) b)
Figure 5.7: a) Maximum relative normal stresses (= ), and b) relative shear stresses along a simply supported
composite beam due to differential shrinkage. m = 1, = 2/7, 1 = 2 = 0, after [1, 6]
Normal stresses can be seen as the reflection of the strains and vice versa. The normal stress
vanishes at the beam edges and has its maximum value in the interior parts of the beam. The strain
(that contrary to stress is easy to measure) has, consequently, its maximum variation at the edges.
This relationship has been verified by strain measurements in two 6 and 8 m long repaired concrete
beams subjected to 14 months of differential shrinkage [1]. By considering creep effects and
comparing the measuring results with computed strain curves for varying L values, it was concluded
that L = 54 gave the best fit to the 8 m long beam (Figure 5.8). This value has in turn been used to
estimate the shear stresses in this beam (Figure 5.9).
The maximum shear stress is estimated to 1.6 MPa, i.e., probably below the shear strength of
concrete. This computation result is an explanation to the fact that no cracks were observed in the
test beams [1].
The computation result also shows that the simple shear stress computation model [15] described in
5.3 gives a fairly good estimation of both maximum shear value and its corresponding triangular
shape.
Figure 5.8: Measured strain values and computed ones, from [6]
Figure 5.9: Computed shear stresses for an example based on geometrical and material data for the 8 m long
beam [6, 18]. Solid line = simple model proposed by Jonasson [15], dotted line = model for incomplete bond [6]
5.6 Experimental tests
Various authors have performed tests on composite structural members. However, very few have
produced results that support or discard models.
5.6.1 Swedish tests on mechanically loaded concrete beams

Silfwerbrand [26] carried out tests on simply supported composite concrete beams subjected to a
static load at mid span (Figure 5.10). The substrate concrete was sawn from a 70 year old bridge
(Skurubron). It had a compressive strength of 85 MPa (measured on drilled cores). The substrate
concrete was chipped with pneumatic hammers prior to overlay casting. The measured compressive
cube strength of the overlay was 60 MPa. For comparison, additional test beams made of newcast
concrete were included in the investigation. Bonding agents were used in two cases. The beams were
tested upside down producing tension in the overlay. As shown in Table 5.1, the composite beams
with hammered interfaces developed equally high failure loads and obtained same failure mode as
the homogeneous beam. Obviously, the bond between substrate and overlay was sufficient.
Contrary, a smooth, steel grinded interface does not provide sufficient bond to promote composite
action. In both beams with this interface, a premature interface failure occurred. After the interface
failure, the (reinforced) bottom part of the beam was able to carry an increasing load, but not to the
same level as the other beams.
Table 5.1: Tests on mechanically loaded concrete beams [26]

Interface Static tests
Preparation Bonding Other treatment Failure load (kN) Failure mode
agent
Pneumatic hammer No 130 Shear
Pneumatic hammer Epoxy 150 Shear
adhesive
Pneumatic hammer No Vacuum treated 130 Shear
overlay
None (homogeneous) No 130 Shear
Steel grinded No 40 & 70* Interface & Shear
Steel grinded Mortar 17 & 69* Interface & Shear
* The composite beam first cracked at the interface; subsequently, the load could be increased markedly to
shear failure in the bottom (reinforced) part of the beam.
Figure 5.10: Test specimen and loading case in beam tests [26]. (unit: mm)
5.6.2 Swedish tests on concrete beams subjected to differential shrinkage
Differential shrinkage is considered to be an important loading case for concrete overlays. In order to
study strain development and cracking, if any, four composite beams of varying length were cast and
stored outdoors in a tent for 14 months (Silfwerbrand, 1986). The thickness ratio between overlay
and total beam height was in all cases 2/7. Two beams were supported on soft airbags providing free
curvature whereas the other two were supported on three solid supports preventing curvature
(Figure 5.11). Old concrete columns constituted the substrate concrete. Drilled cores obtained a
compressive strength of 62 MPa. The substrate concrete was chipped with pneumatic hammers prior
to overlay casting. The surface was prewetted during a couple of days but superficially dry at overlay
placement. No bonding agents were used. At the end of the tests, the concrete used for the overlays
obtained compressive cube strength of 65 MPa. Maximum free shrinkage (measured on
400100100 mm standard prisms) was 0.45 mm/m. Despite the magnitude of the free shrinkage
and the bond providing restraint, no cracks were observed, neither in the overlays perpendicular to
the beam length nor in the interface. Strain measurements supported composite action. After
terminated tests, a bond strength averaging 1.9 MPa was obtained through pulloff tests.
Figure 5.11: Tests on composite beams [1] and composite slabs [18] subjected to differential shrinkage
5.6.3 Swedish tests on concrete slabs

In mid 1980s, the waterjet technique was introduced in Sweden. In order to compare it with other
methods of concrete removal and substrate preparation, slab tests were carried out at KTH in
Stockholm, Sweden [18]. The preparation methods investigated included waterjet technique,
pneumatic hammers, and sandblasting. The substrate surface was prewetted during three days but
superficially dry at overlay placement. No bonding agent was used. The tests consisted of five
composite slabs (Figure 5.11) and two homogeneous ones. The ratio between overlay thickness and
total height was 1/3. The substrate concrete was seven months old at time of overlay placement.
Differential shrinkage was investigated during six months under indoor conditions. No visible
cracking occurred. Displacement measurements supported composite behaviour. At terminated
tests, the compressive cube strength exceeded 50 MPa for both substrate and overlay concrete. The
free shrinkage (measured on 400100100mm prisms) of the overlay concrete was 0.6 mm/m.
Simultaneously, the shrinkage of the old concrete increased with less than 0.05 mm/m.
In a second phase of the tests (Figure 5.12a), the five composite and the two homogeneous slabs
were simply supported at the four corners and loaded to failure with a centrically placed static load.
The base layer was reinforced with 13 8 mm, s = 150 mm in both directions. The obtained load
displacement curves show a similar behaviour for all slabs. All slabs (including homogeneous ones)
obtained approximately equally high ultimate loads between 83 and 98 kN (Figure 5.12b). Finally, the
bond strength was investigated by pulloff tests (see 4.4).
a) b)
Figure 5.12: Composite slab loading experiments: a) experimental layout;

b) load-displacement curves [18]
5.6.4 Tests at EPFL on composite beams with normal concretes of different ages
Bernard [3] tested 14 composite beams with substrate dimensions length/width/thickness of

5400/500/170 mm, hydrojetted with removal of 20 mm of concrete, and new concrete layers
between 70 to 170 mm, with various reinforcement ratios, with reinforcement bars or steel fibres.
The beams, placed in a controlled environment (20 C, 30 % RH) were instrumented with external
LVDT's and thermocouples and optical deformation sensors embedded in the substrate and the new
layer. The substrates were between 70 and 237 days old when the overlays were cast. The new
concretes composition was 300 kg/m3 CEM I 52.5 R, W/C = 0.50 without steel fibres and 350 kg/m3
CEM I 52.5 R and W/C = 0.5, with 80 kg/m3 steel fibres. The deformational behaviour of the
composite members was monitored at early age for 3 to 6 days, and at longer term, up to 270 days,
under creep loading. The optical deformation sensors measured total displacements in the central
part of the beam (distance 2000 mm). The thermocouples gave the temperature at the same level.
The measurements were started when the new concrete was cast and stopped after 100 hours. The
main results from this study can be summarized as follows:
the time lag between the end of chemical swelling and the peak of temperature in the
upper part of the new layer and the evolution of strains are well predicted only if the
autogenous shrinkage is taken into account in the numerical modelling;
the autogenous shrinkage is the main phenomenon which induces curvature and tensile
stresses at early age in the new layer of tested hybrid elements;
in the conditions of these tests, the long term residual stresses due to autogenous
shrinkage attain approximately 50 % of the maximum tensile stress at early age; and
the reinforcement in the new layer induces a significant restraint of the overall
deformations of the composite member; for high reinforcement ratios of the new layer,
this restraint can lead to cracking.
5.6.5 Other tests

Finally, it is also worth mentioning the works by Chausson [27] that used a steel substrate with a
grinded surface to simulate a concrete substrate. This method has significant advantages in terms of
reutilization of the substrates for several tests.
5.7 On restrained shrinkage set-ups

Another way to experimentally investigate the effect of restraint on the development of stresses and
the risk of cracking, without having to test costly large size structural members, is given by so called
"restrained shrinkage testing devices". These devices have been known for many years now, starting
with socalled "cracking frames", [28], where the restraint is passive. Active systems were proposed
by Kovler [29]. Finally, so called TemperatureStress Testing Machines TSTM are systems with active
restraint by means of closedlooptesting machines as well as cooling circuits placed around the
specimens, to impose temperature conditions at will [3036].
Such systems can be used either in isothermal conditions to isolate fundamental material properties
such as autogenous shrinkage, or to test overlay materials in realistic conditions of restraint and
curing, thus providing and "experimental simulation tool".
In a similar way, to induce a high restraint with compact dimensions, Martinola [37] used
instrumented ring geometry, inspired from the "Bolomey" ring shrinkage test.
5.8 Numerical modelling

The numerical modelling of the structural behaviour of bonded concrete overlays and more generally
of composite concreteconcrete structural members has been realized in various ways in the past,
with either general purpose finite element packages, specifically oriented packages for multilayer
systems, or original numerical solutions of mathematical formulations of the problem. Basically, any
finite element package is able to model multi layer systems. The main differences come from the
ability of the software to deal with more advanced material properties such as:
time-dependent evolution of the mechanical properties of the overlay and substrate

maturation;
comprehensive simulation of the physical processes such as drying, release of heat of
hydration, thermal transport;
viscoelasticity and ageing;
non-linear crack propagation (smeared crack models or discrete crack models);
interface between two layers;
reinforcing bars.
It is impossible to quote all modelling works on hybrid systems however, among the most recent
ones, one can mention the following:
Bernard [3, 38, 39] used the finite element package MLS (Multi-Layer Systems) from
FEMMASSE [40], with a comprehensive description of thermo-hygro-mechanical processes
in multi-layer systems, at early age and long term, including effects of maturation,
viscoelasticity, reinforcement in the overlay and non-linear fracture mechanics (smeared crack
model) for the bulk of the materials and for interfaces. The significant role of autogenous
shrinkage at early age on the structural response of hybrid elements with concretes and the
consequences of the restraining effect of reinforcement bars in the new layer were
demonstrated. The results include an extensive parametric study on the effect of the overlay
thickness and reinforcement ratio on the risk of delamination. Habel [41, 42] used the same
package with material models extended to the case of tensile hardening fibre reinforced
composites to model composite beams with Ultra-High Performance Fibre Reinforced
Concrete (UHPFRC) overlays.
Martinola [37, 43] used FEM package DIANA [44] for thermo-hygro-mechanical
calculations with a strong emphasis on moisture transport and self-desiccation, in order to
study the risk of cracking and delamination of composite structural systems. Realistic
interfacial constitutive laws were used with parameters from experimental tests. Based on
comprehensive experimental tests, the hygro-mechanical coupling coefficient was
implemented in the code as a function of the moisture content, and the major influence of this
variation was demonstrated. A parametric study was performed and recommendations for the
design of bonded concrete overlays were given.
Laurence [45] used software CESAR-LCPC to calculate the evolution of stresses at early
age and long term in bonded concrete overlays. The modules TEXO, MEXO and HEXO of
the finite element package were used to calculate the stresses induced by thermo-chemo-
mechanical effects at early age, taking maturation into consideration, for linear elastic
materials. In a second step, the effect of drying shrinkage, including water absorption by the
substrate, was introduced. The results showed the very significant influence of autogenous
shrinkage as well as of the absorption of water by the substrate on the development of stresses
in the overlay.
Finite elements computations of composite structures with bonded overlays subjected to drying have
been performed by Tran et al. [46] using the FEM code CAST3M They take into account the effects of
the shrinkage and of the loads applied on an overlaid structure. Shrinkage is computed from the
water loss associated, 1) to the hydration process, 2) to the water exchanges with the surrounding
(mainly evaporation). The results faithfully represent the experimental relationship between
measured shrinkage, applied loads and debonding. Apart from the boundaries, the Bernoulli principle
appears as valid.
Fron [47] developed a mechanical model to determine the stresses in multilayer systems due to
shrinkage, taking into consideration ground friction and chemothermohygro mechanical couplings
in the materials. The water loss was calculated from diffusion laws by finite differences. The
calculation of deformations and stresses in the multilayer system relies on the assumptions of 1) a
perfectly bond between the layers, and 2) the respect of the NavierBernoulli principle. This model
has been validated by comparative finite element simulations and successfully applied to the cases of
industrial floors with Fibre Reinforced Concretes and repair layers.
HIPERPAV, McCullough et al., [48, 49] Rasmussen et al. [50, 51], although not a finite element code,
takes into consideration thermohygromechanical phenomena by means of semiempirical
mathematical formulas from codes or recommendations, to predict the risk of cracking in overlays.
The physical processes involved (thermal and moisture transport) are based on semiempirical
analytical solutions for current geometries.
The model from Rostasy et al. [52 is also numerical-analytical but includes very
comprehensive descriptions of the physical and mechanical phenomena acting at early age
and long term in composite structures. Although more dedicated to massive structures, walls
on slabs for instance, this model is also applicable to overlays on substrates.
5.9 Conclusions
Composite structural members are very common. They are however still often designed
on an empirical basis.
The notion of degree of restraint, associated to the degrees of freedom of a composite
structural member provides a common background for the analysis of various practical
applications among which are cement based overlays on existing concrete substrates,
under restrained shrinkage.
This notion initially derived for linear elastic overlay materials case can be extended to
viscoelastic materials.
This approach enables one to predict in a realistic way the stresses induced by restrained
deformations and access the risks of cracking in the new layers.
It can be further extended to three dimensional structures such as slabs, or to include the
restraining effect of reinforcement bars in the substrate or in the new layer.
For current cases of application (overlays on flexible substrates), in the linear-elastic
case, the degree of restraint varies between 0.4 and 0.8 and can be calculated from
simple geometrical and material parameters.
Shear stresses at the interface of composite members can be accurately predicted, taking
into consideration partial debonding.
5.10 Outlook for future research
Experimental tests to determine accurately the strain profiles/distributions in various

cases of member slenderness and combinations of materials.
Extension to 3 D (case of slabs) of analytical models taking into consideration the
degree of restraint
Determination of design values of the time/rate dependent tensile strength.
Implementation of non-linear viscoelastic models adapted for calculations with high
stresses high degrees of restraint.
Relaxation tests under tensile loading at various pre-peak load levels, with temperature
control of the specimen.
REFERENCES
[1] Silfwerbrand, J., 'Differential Shrinkage in Composite Concrete Beams of Old Concrete and a
New-Cast Concrete Overlay'. Bulletin No. 144, Dept. of Structural Mechanics and Engineering,
Royal Institute of Technology, Stockholm, Sweden, 149 pp. (1986) (In Swedish)
[2] Silfwerbrand, J., 'Differential Shrinkage in Normal and High Strength Concrete Overlays'.
Nordic Concrete Research, No.19, (1996) 55-68.
[3] Bernard, O., Comportement long terme des lments de structure forms de btons dges
diffrents, Doctoral thesis, Swiss Federal Institute of Technology n2283, Lausanne,
Switzerland (2000) (in French)
[4] Sadouki H., 'Simulation et analyse numrique du comportement mcanique de structures
composites', Doctoral thesis, Swiss Federal Institute of Technology n676, Lausanne,
Switzerland (1987) (in French)
[5] Birkeland H.W., 'Differential shrinkage in composite beams', Journal of the American Concrete
Institute, (May 1960), 1123-1136.
[6] Silfwerbrand, J., 'Stresses and Strains in Composite Concrete Beams Subjected to Differential
Shrinkage', ACI Structural Journal, 94 (4), (1997) 347-353.
[7] Alonso Junghans M.T., 'Zur Risssicherheit zementgebundener dehnungsbehinderter Schichten
unter Bercksichtigung von Dauereinflssen', PhD thesis, Universitt Hamburg-Harburg,
Shaker Verlag, Aachen, Germany (1997) (in German).
[8] Denari E., Silfwerbrand J., 'Structural behaviour of bonded concrete overlays', in Proceedings
International RILEM Workshop on Bonded Concrete Overlays, June 7-8, 2004, Stockholm,
Sweden, edited by Swedish Cement and Concrete Research Institute (CBI), RILEM PRO 43,
(2004).
[9] Hartl G, 'Kraftverlauf in Beschichtungen', Zement und Beton, 28. Jahrgang, heft 2, Germany,
(1983) 45-51 (in German).
[10] Hartl G., 'Materialtechnologische Beurteilung von Verstrkungsmassnahmen', Beton und
Stahlbetonbau 95 (12), (2000) 707-712 (in German).
[11] Beushausen H., 'Long-term performance of bonded concrete overlays subjected to differential
shrinkage', PhD Thesis, University of Capetown, South Africa (2005).
[12] Beushausen H., Alexander M., 'Bonded concrete overlays subjected to differential shrinkage
an analytical model based on localized strain and stress', in Proceedings International RILEM
Workshop on Bonded Concrete Overlays, June 7-8, 2004, Stockholm, Sweden, edited by
Swedish Cement and Concrete Research Institute (CBI), RILEM PRO 43 (2004).
[13] Beushausen H., Alexander M., 'Spannungen durch Verformungsbehinderung in gebundenen
Aufbetonen', Beton und Stahlbetonbau 101 (6), (2006) 394-401 (in German).
[14] Beushausen H., Alexander M., 'Localised strain and stress in bonded concrete overlays
subjected to differential shrinkage', Materials and Structures, 40 (2) (2007), 189-199.
[15] Jonasson, J.-E., 'Computer Program for Non-linear Computations in Concrete with Regard to
Shrinkage, Creep, and Temperature'. CBI Report No. 7:77, Swedish Cement and Concrete
Research Institute, Stockholm, 161 pp. (1977) (In Swedish).
[16] Haardt P., 'Zementgebundene und kunststoffvergtete Beschichtungen auf Beton', Dissertation,
Heft 13, TU Karlsruhe (1991) (in German).
[17] Silfwerbrand, J., 'Concrete Overlays', Report No. 10, 3rd Edition, Chair of Structural Mechanics
and Engineering, Dept. of Structural Engineering, Royal Institute of Technology, Stockholm,
Sweden, 65 pp. (1997) (In Swedish).
[18] Silfwerbrand, J., 'Effects of Differential Shrinkage, Creep and Properties of the Contact Surface
on the Strength of Composite Concrete Slabs of Old and New Concrete'. Bulletin No. 147, Dept.
of Structural Mechanics and Engineering, Royal Institute of Technology, Stockholm, Sweden,
131 pp. (1987) (In Swedish).
[19] Beushausen H., Alexander M., 'Failure mechanisms and tensile relaxation of bonded concrete
overlays subjected to differential shrinkage', Cement and Concrete Research, 36, (2006) 1908-
1914.
[20] Huet C., 'Adaptation d'un algorithme de Bazant au calcul des multilames visco-lastiques
vieillissants', Materials and Structures, 13, (74), (1980) 91-98 (in French).
[21] Siestrunck R., Lamer P., Huet C., Alviset L., 'Action de l'humidit sur la cramique envisage
dans le cadre de l'association bton-cramique', Communication C.T.T.B. au symposium
Rilem/C.I.B., Helsinki (1965) (in French).
[22] Gutsch A. , Rostsy F.S., 'Young concrete under high tensile stresses-creep relaxation and
cracking', in: R. Springenschmidt (Ed.), Proceedings: RILEM Symposium Thermal Cracking
in Concrete at Early Ages, Chapman & Hall, London, (1995), 111116.
[23] Horimoto H., Koyanagi W., 'Estimation of stress relaxation in concrete at early ages', in: R.
Springenschmidt (Ed.), Proceedings: RILEM Symposium Thermal Cracking in Concrete at
Early Ages, Chapman & Hall, London, (1995), 95102.
[24] Denari E., Ccot C., Huet C., 'Characterization of Creep and Crack Growth Interactions in the
Fracture Behaviour of Concrete', Cement and Concrete Research, 36 (3), (2006), 571-575.
[25] FIP. Federation Internationale de la Precontrainte, 'Structural Effects of Time-Dependent
Behaviour of Concrete', Guide to good practice, FIP, Wexham Springs (1982).
[26] Silfwerbrand, J., 'Composite Action between Partially Chipped Concrete Bridge Deck and
Overlay. Beam Tests', Bulletin No. 142, Dept. of Structural Mechanics and Engineering, Royal
Institute of Technology, Stockholm, Sweden, 72 pp. (1984) (In Swedish).
[27] Chausson H., 'La durabilit des rechargements minces adhrents en bton renforc de fibres
mtalliques', Thse de doctorat de l'universit Paul Sabatier / LMDC N ordre: 2708,
Toulouse, France, (1997) (in French).
[28] RILEM TC-119-TCE, 'Avoidance of Thermal cracking in concrete at early ages TCE3: testing
of the cracking tendency of concrete at early ages using the cracking frame test', Materials and
Structures 30, (1997), 461-464.
[29] Kovler K., 'Testing system for determining the mechanical behaviour of early age concrete
under restrained and free shrinkage', Materials and Structures, 27, (1994) 324-330.
[30] Bjontegaard O., 'Thermal dilatation and autogenous deformation as driving forces to self-
induced stresses in high performance concrete', Doctoral thesis, Trondheim, Norway (1999).
[31] Pigeon M., Toma G., Delagrave A., Bissonnette B., Marchand J., Prince J.C., 'Equipement for
the analysis of the behaviour of concrete under restrained shrinkage at early ages', Magazine of
Concrete Research, 52 (4), (2000) 497-302.
[32] Altoubat S.A. and David A. Lange D.A., 'Tensile basic creep: measurements and behavior at
early age', ACI Materials Journal, 95 (5), (2001) 386-393.
[33] Bentur A., Kovler K., 'Evaluation of early-age cracking characteristics in cementitious systems',
Materials and Structures, 36, (2003) 183-190.
[34] Charron J-P., 'Contribution l'tude du comportement au jeune ge des matriaux cimentaires
en conditions des dformations libre et restreinte', PhD thesis, University Laval, Quebec,
Canada, (2003) (in French).
[35] Sule M., 'Effect of Reinforcement on Early-Age Cracking in High Strength Concrete', PhD
Thesis, TU Delft, (2003).
[36] Kamen A., Denari E., Brhwiler E., 'Mechanical Behaviour of Ultra High Performance Fibre
Reinforced Concretes (UHPFRC) at early age, and under restraint', Proceedings CONCREEP 7,
Eds. G. Pijaudier-Cabot, B. Grard, P. Acker, Herms Publishing, (2005) 591-596.
[37] Martinola G., 'Rissbildung und Ablsung zementgebundener Beschichtungen auf Beton'.
Dissertation. ETH Zrich, Switzerland, Nr. 13520 (2000) (in German).
[38] Bernard O., Brhwiler E., 'The influence of reinforcement in the new layer on hygral cracking in
composite structural elements', Materials and Structures, 36, (2003), 118-126.
[39] Bernard O., Brhwiler E., 'Influence of autogenous shrinkage on early age behaviour of
structural elements consisting of concretes of different ages', Materials and Structures, 35,
(2002), 550-556.
[40] Roelfstra P. E., Salet A. M., Kuiks J. E., 'Defining and application of stress-analysis-based
temperature difference limits to prevent early-age cracking in concrete structures', Proceedings
n25 of the International RILEM Symposium: Thermal cracking in concrete at early age,
Mnich, (1994) 273-280.
[41] Habel K., 'Structural Behaviour of Elements Combining Ultra-High Performance Fibre
Reinforced Concretes (UHPFRC) and Reinforced Concrete', Doctoral thesis n 3036, Ecole
Polytechnique Fdrale de Lausanne, Lausanne, Switzerland, (2004).
[42] Habel K., Denari E., Brhwiler E., 'Time Dependent Behaviour of Elements Combining
UHPFRC and Concrete', Materials and Structures, 39 (2006) 557-569.
[43] Martinola G., Sadouki H., Wittmann F.H., 'Numerical model for minimizing the risk of damage
in repair system', ASCE Journal of Materials in Civil Engineering, (2001) 121-129.
[44] DIANA, 'User's manual-release 6.1', Diana Analysis BV, Delft, Niederlanden (1996).
[45] Laurence O., 'La fissuration due au retrait restreint dans les rparations minces en bton: apports
combins de l'exprimentation et de la modlisation', thse de doctorat de lcole Nationale des
Ponts et Chausses, Paris, France (2001) (in French).
[46] Tran, Q.T, Toumi A., Granju J-L., 'Experimental and numerical investigation of the debonding
interface between an old concrete and an overlay', Materials and Structures, 39 (3), (2006) 379-
389.
[47] Fron C., 'Etude des mcanismes de generation de contraintes et de fissuration par retrait gn
dans les structures base de matriaux cimentaires', thse de doctorat de l'institut national des
sciences appliqus de Lyon, ref. 2002 ISAL 0025 (2002) (in French).
[48] McCullough, B.F. and Rasmussen, Robert Otto, 'Fast-Track Paving: Concrete Temperature
Control and Traffic Opening Criteria for Bonded Concrete Overlays Volume I: Final Report',
Federal Highway Administration Report FHWA-RD-98-167, Washington, USA (1999).
[49] McCullough, B.F. and Rasmussen, Robert Otto, 'Fast-Track Paving: Concrete Temperature
Control and Traffic Opening Criteria for Bonded Concrete Overlays Volume II: HIPERPAV
User's Manual', Federal Highway Administration Report FHWA-RD-98-168, Washington,
USA (1999).
[50] Rasmussen, Robert Otto, et. al., 'Fast-Track Paving: Concrete Temperature Control and Traffic
Opening Criteria for Bonded Concrete Overlays Volume III: Addendum to the HIPERPAV
User's Manual', Federal Highway Administration Report FHWA-RD-99-200, Washington
(1999).
[51] Rasmussen, R. O., B. F. McCullough, and J. Weissmann, 'Development of a Bonded Concrete
Overlay Computer-Aided Design System', Research Report 2911-1, Centre for Transportation
Research, The University of Texas at Austin, (1995).
[52] Rostasy F.S., Kraus M., Budelmann H., 'Planungswerkzeug zur Kontrolle der frhen
Rissbildung in massigen Betonbauteilen Teil 5: Behinderung und Zwang', Bautechnik 79 (11),
(2002) 778-789 (in German).
6. DEBONDING
A. Turatsinze (1), H. Beushausen (2), R. Gagn (3), J.-L. Granju (1), J. Silfwerbrand (4), R. Walter (5)

France
(3) Dpartement de Gnie Civil, Facult de Gnie, Sherbrooke (QC), Canada
(4) Swedish Cement and Concrete Research Institute (CBI), Stockholm, Sweden
(5) Department of Civil engineering, Technical University of Denmark (DTU), Lyngby,
Denmark
ABSTRACT
This chapter focuses on the causes and the mechanisms of debonding of cement-based
material overlays. Additionally, it describes how debonding affects durability of the
composite structure. Methods for monitoring are described. The debonding mechanism is
discussed, and debonding modeling is described. The role of reinforcement in debonding is
important and is discussed. Crack propagation and crack opening affects debonding and is
described. There are special overlays including those bonded to steel and overlays that are
anchored to the substrate with metal anchors and they are discussed. The effect of boundaries
and joints is discussed.
6.1 Introduction
This chapter focuses on the causes and the mechanisms of debonding of cement-based
material overlays. Additionally, it describes how debonding affects durability of the
composite structure. As described by Carter et al. [1], the term debonding implies, that a bond
had previously developed between the repair layer and the substrate. The bond strength is one
of the main parameter governing debonding. The development of bond depends on several
factors, see, e.g., Silfwerbrand [2], Silfwerbrand et al. [3], Saucier et al. [4], Langlois et al.
[5], Delatte et al. [6], Tschegg et al. [7]. These factors are not discussed herein (it is dealt with
in chapter 4) but will remain implicit in this chapter.
6.2 Impact Of Debonding

Substrateoverlay composites should work monolithically, and the bond at the interface ensures
the continuity of deformations between the base and the overlay. In these conditions, the durability
of a repair relies not only on the durability of the material used, but also on the durability of its bond
with the base.
6.3 Brief Summary Of Debonding

A holistic model presented by Vaysburd et al. [8] and Emmons et al. [9] shows that whatever the
original cause of the repair failure is, cracking is involved. According to Carter et al. [1], there are
many possible causes of debonding initiation also for properly placed overlays. They include adverse
ambient weather conditions during construction, improper wet curing, excessive vehicle impact due
to wear surface roughness, formation of cracks allowing water access and the subsequent frost,
vibrations during overlay construction and minor bonding defect. In a mechanical respect, it is well
accepted that there are two classes of debonding causes: the effects of the external mechanical
loading and the effects of the different length changes of the overlay and of its substrate.
Additionally it is known that debonding initiates preferentially from discontinuities of the overlay:
boundaries, cracks and joints. Here, the straining of the overlay induces concentrations of builtin
stresses that act to debond. Although, the debonding mechanism is related to a mixed mode cracking
[1011], mode I cracking dominates at debonding initiation. The vulnerability of boundary locations,
joints and cracks, is inherent in their deficiency in stress transfers. Such a weakness cannot be
corrected for the boundary locations but it can be attenuated for cracks and partial depth sawed
joints thanks to overlay reinforcement. Reinforcement can be applied using traditional
reinforcements such as steel fabrics and steel bars or by fibres in the ready mix concrete. Laboratory
tests, field trials and numerical simulations lead to the conclusion that these various types of
reinforcement efficiently improve durability of cementbased overlay (see section 6.7).
6.4 Methods For Monitoring Debonding

There are several methods to detect debonding both for in situ and laboratory evaluations. In
both cases non destructive and semi destructive methods can be implemented. These methods are
briefly described below.
6.4.1 Non destructive methods

Visual inspection is mainly a laboratory method because the eye must have access to the
side edge of the overlaid structure. The method is subjective, and the help of magnifying
devices significantly improves accuracy. A video microscope is especially efficient.
Chain drag The method is rudimentary but gives a qualitative assessment in good
agreement with results from more sophisticated methods [12-14]. It consists of dragging
a chain across the overlay surface while listening for a change in acoustic response. It is
a rapid and inexpensive method.
Infrared thermography [13-15]. Delaminated areas appear as areas of different
temperature. This technique offers some advantages:
no direct contact required between the camera and the object under investigation;
monitoring of quite large areas with an accuracy which can reach 0.08C; and
the required equipment is lightweight and can be easily transported.
Drawbacks are :
emissivity varies from material to material and with the brightness of different objects
within an area;
high sensitivity to environmental conditions; and
the results are influenced by the time of the day and whether the object is subjected to
heating or to cooling.
Electromagnetic methods. Typically Ground Penetrating Radar (GPR): it is a
standardised test (ASTM D6087) and a widely documented method to detect
delaminations [13].
Acoustic methods: acoustic emission, ultrasonic pulse velocity and impact-echo [14, 16-
18]. These methods are generally reputed to be more expensive and time consuming. In
return, they are quantitative methods.
6.4.2 Semi destructive methods

The widest spread ones are:
the pull-off test (presented in chapter 4); and
coring and visual plus mechanical investigation of the core.
6.4.3 Laboratory tests

In addition to the above listed methods, in laboratory conditions, accessibility of the boundary
locations provides additional information. Typically LVDTs or strain gages spanning the interface are
placed and will be activated when debonding reaches their location [19]. Based on their low cost and
high reliability, LVDT sensors are often recommended [20]. Farhat [21] successfully used a set of
breaking electrical conductive links spanning the overlaysubstrate interface. Dynamic measurements
giving the structures eigen frequencies have also been used. Increasing cracking would lead to
decreasing stiffness and frequencies [22].
6.5 Debonding Mechanism

Debonding of an overlay is considered to result from one or the superposition of the two causes
sketched in Figure 6.1, respectively, designated as "due to flexural effects" (Figure 6.1a) and "due to
differential length change" (Figure 6.1b). In situ, these two causes of debonding superimpose their
effects.
Cracking and
debonding
Wheel
Overlay
Substrate
Soil
(a) Debonding due to flexural effects, consequence of

the flexural straining of the structure by the applied
loads, especially by live loads. The critical zone is apart
from the applied loads, where the overlay is in
tension. Debonding is possible because of the cracking
of the overlay.
Cracking and debonding
Overlay
Base structure
(b) Debonding due to differential length change,

induced by the effects of the different length changes
of the overlay and of the base, because of shrinkage or
temperature changes. Debonding is a consequence of
the cracking of the overlay. At each side of the crack,
the peeling effect induces tensile stresses
perpendicular to the interface.
Figure 6.1: The two fundamental causes of debonding [23]: a) due to flexural effects, b) due to
differential length change
In both cases, debonding results from a full depth separation of the overlay, e.g. a crack, disabling
the continuity of stress transmission within the volume of the overlay. Emmons et al. [9] and Bernard
[24] also pointed out the prominent role of cracking in the debonding mechanism.
At the crack location, sharp builtin peak stresses develop at the interface, which subsequently
causes debonding [2528]. Reinforcement of the overlay, which enables force to be transmitted
through the crack, decreases the intensity of the mechanical discontinuity at the crack location.
Consequently, it decreases the builtin peak stresses at the interface. It is through this process that
reinforcement of the overlay, notably fibre reinforcement, acts to improve the durability of the bond
to the substrate.
This is schematically shown in Figure 6.2 [29] in the example of a debonding due to loading
stresses (where d is the distance from the loading point). It compares the case of: 1) an uncracked
overlay, 2) a neat cut through the overlay simulating a crack in a nonreinforced overlay (no force is
transmitted through the crack), 3) a lower elastic modulus section (onethird of the modulus of the
repair material) inserted in the overlay to simulate an overlay section containing a fibre reinforced
crack. In the case of fibre reinforcement, the peak stresses acting to debond are reduced by 75 to
80%.
Figure 6.2: Stress peaks induced at the interface by a crack and effect of fibre reinforcement, example of a
debonding of mechanical origin [29]
In the case of steel reinforcement bars or welded fabric the same mechanism is involved. Steel
bars and welded fabric can transmit much higher forces than fibre reinforcement, but they can only
be used with thick overlays with respect to cover depth requirements. Another advantage of fibres is
that, unlike steel bars, they present a reduced risk of corrosion [30].
Whatever the origin of debonding is, mechanical or length changes, the basic cause of
developed tensile stress, perpendicular to the interface, is the socalled peeling effect. It is the
consequence of the unbalanced shear stresses at a cut of the bonded overlay (cracks, sawcut joint
or slab edge). It is easily illustrated by the peeling moment as presented in Figure 6.3. The peeling
moment increases with the edge overlay thickness.
In the case of a cementbased overlay, because of its significant thickness (often more than 20
mm), the peaks of shear stress and of tensile stress are often of the same order of magnitude.
Knowing that the tensile strength of an interface between cementbased materials is at the most
50% of its shear strength [19, 3133], it is demonstrated that debonding is always initiated by tension
perpendicular to the interface [19, 29, 3435].
In the case of debonding caused by length changes, a large amount of information is available,
notably on the effect of restrained shrinkage [3639]. Fowler et al. [26] presented thermallyinduced
stresses at the interface with a polymer concrete used as a material repair. It was demonstrated that
debonding could occur due to length change alone, and the model confirmed that debonding
preferentially starts at the overlay boundaries.
Peeling moment
M=e *Nx Nx
xy xy
Figure 6.3: Peeling effect [23]
Curling (Figure 6.4) due to non uniform shrinkage through the thickness of the overlay [4042]
increases the builtin tensile stress perpendicular to the interface. In the case of mechanical origin
debonding, the trend of the overlay to not follow the curvature of the substrate, designated in the
following by "geometrical effect", (Figure 6.1a) is also an aggravating factor.
Figure 6.4: Curling effect at cracks, saw-cut joint and slab edge [40]
6.6 Debonding Modeling
Numerical modeling offers exciting tools to investigate debonding mechanism. Debonding
is modeled as a crack propagating along the substrate-overlay interface. It is now known that
the strain softening (generally designated in the following as interlocking) is essential in
modeling crack propagation. For the same reasons debonding modeling has to integrate the
softening behaviour of the interface. Moreover, it must be taken into account that the
debonding mechanism is characterized by mixed mode, mode I plus mode II. Recently Walter
et al. [11] and Sabathier et al. [10, 43] proposed an improved model. They consider
interaction between residual interfacial shear and tensile stresses. Breakage of bond in pure
tension will necessarily, at some point, affect the shear interlocking, and vice versa. Tran et al
[44] completed this modeling. he shows that in the case of debonding of mechanical origin
(bending of the overlaid structure with the overlay on the tension side) mode II can be
neglected. The same authors also presented an interesting numerical analysis of factors
influencing the overlay debonding taking account of shrinkage strains, calculated from the
hydration development and water exchanges between the overlay and its surroundings, and
including the fatigue damage of interlocking (in the cracks and along the debonded
interfaces). In addition, effects of Youngs modulus and of tensile strength of the overlay
material, effects of type of the overlay reinforcing fibres (slipping or high bond fibres) and
the influence of tensile strength of overlay-substrate interface were investigated. In particular
Tran et al [45] demonstrated that a low modulus of elasticity and high strength, two
characteristics mutually exclusive, are suitable for durable thin cement-based overlays.
6.7 Role Of Reinforcement Of Overlays

The reinforcement can be provided by steel bars or welded fabric or by fibres. Although all
these types of reinforcement act in similar ways, the literature mainly reports cases of fibre
reinforcements.
6.7.1 In situ findings

One of the first and most convincing attempt to use fibre reinforced bonded overlays was made in
Greene County (Iowa, USA) to repair concrete road pavements [46]. It illustrated the beneficial effect
of steel fibre reinforcement of the overlay on the durability of the repair. The fibres were common
steel wire fibres. The fibre content was very high, 90 kg/m3. Paulsson et al. [47] reported a similar
case, with success, where 75 kg/m3 of "enlarged end" steel fibres were used as reinforcement.
Another typical example is the Belgian experience in resurfacing existing roads with concrete,
which has been in use for about 20 years. Improved incorporation methods allowed for subsequent
lower fibre content in thin bonded repairs, less than 150 mm thick [48].
Other results from various places confirmed the positive effect of metal fibre reinforcement of
the overlay even at much smaller dosage, from 25 to 40 kg/m3. A significant example is the case of
the Motorway 40 in Montreal, Canada. It has three lanes in each direction withstanding a traffic
intensity of 30,000 vehicles per day. In 1986, an experimental repair was performed with the goal to
compare the efficiency of different types of bonded overlays. On a total length of 134 meters, each
investigated overlay system covered a section of about 20 meters in length. This experimental repair
is interesting because it is thoroughly documented. Its conception, the preparation of the substrate,
and the placing of the overlay were supervised by a team from Sherbrooke University (Quebec,
Canada) which, further, ensured the monitoring of its behaviour over a 12year period [12, 15, 49
50]. The different types of overlay systems accounted for and compared included plain concrete (100
mm thick) and others (75 mm thick) reinforced with 22 to 34 kg/m3 of steel fibres. Three types of
steel fibres were tested: hooked ended, and undulated, of circular or rectangular cross section, all 50
mm long. The effect of attachment of the overlay to the base pavement by steel nails was also
investigated. These were 37.5mm long and previously driven to half of their length into the old
pavement. No nail was driven in the central lane; the two other lanes were nailed, one with a nail
300 mm long, and the other with a nail 450 mm long. The interface was prepared by sand jetting to
remove oil stains, other impurities and all lose material. In the case of the plain concrete overlay, 25
mm of the old concrete pavement was removed by scarification. All the overlays were bonded with a
thin cement grout (even if as it is discussed in chapter 4 the actual influence of a such bonding agent
is still a controversial topic). A sketch of the different test sections is shown in Figure 6.5.
The repair materials were concrete with a maximum aggregate size of 20 mm and a water-
cement ratio of 0.38 (160/420). The 28-day compressive strengths of the mixes are given in
Table 6.1.
Cracking detection was achieved by visual inspection. For improved detection, observations were
carried out on a slightly moist repair surface. The crack growth, quantified by the total length of
observed cracks per unit length of lane, is presented in Figure 6.6. In the plain concrete overlays a
sharp rate of crack growth was observed associated with a quick deterioration of the pavement:
cracks induced breaking in many sections and led to debonding of the overlay, and within less than
two years the overlay had to be replaced. On the other hand, in the fibre reinforced sections, the
crack growth was much slower and almost vanished after the first year. No significant difference was
seen between the different types and dosages (from 22 to 34 kg/m3) of steel fibres. After twelve
years of use, the fibre reinforced repairs were still sound, with limited cracking and no debonding.
The current cracks remained narrow and limited in numbers. The majority of cracks were hairline
cracks and developed to less than half the depth of the overlay. However, the reflection of the cracks
or other discontinuities of the substrate structure could not be prevented. Their opening was
reduced when fibre reinforcement was used. As illustrated in Figure 6.7, the presence of connecting
nails had an insignificant effect on the crack growth.
Nails Without
Without Nails
every 45 cm nails every 30 cm
Repair thickness = 100 mm

Scarified depth
25 mm
PLAIN CONCRETE
FIBRES C
FIBRES A
FIBRES B
FIBRES A
Repair thickness = 75 mm
FIBRES B Sandblasted,
FIBRES A
Limit between different construction conditions
Joints
Sensor for crack opening measurement
Figure 6.5: In situ experimental rehabilitation - details of different construction

conditions [12, 15, 49-50]
Table 6.1: Fibres used and 28-day compressive strengths of the mixes
Type of fibre Steel fibres
A B A B C
Dosage (kg/m3) 22 24 34 34 34 0
28day compressive 53.2 45.8 54.4 49.2 28.9 46.4

strength fc (MPa)
Fibre A: length: 50 mm, shape: hooked end, section: circular, diameter: 0.50 mm
Fibre B: length: 53.52 mm, shape: undulated, section: rectangular, equivalent diameter: 0.96
mm
Fibre C: length: 60 mm, shape: undulated, section: circular, diameter: 0.1 mm
Figure 6.6: Crack growth versus time, effect of fibre reinforcement [12, 15, 49-50]
3
Crack growth (m/m)

Nails every 450 mm
2
Without nails
1
Nails every 300 mm
0
0 2 4 6 8 10 12 14
Age (months)
Figure 6.7: Crack growth versus time, effect of connecting nails : case of a concrete reinforced with 22 kg/m3
[12, 15, 49-50]
For the fibre dosages used (less than 40 kg/m3) shrinkage and thermal length changes were not
significant. Consequently, the explanation of the beneficial effect of the fibre reinforcement of the
overlay cannot be found in lower length changes between the overlay and the substrate. The
beneficial effect of the fibres actually must come from their ability to transmit significant through
crack stresses and to restrain cracking.
6.7.2 Need to distinguish between first monotonic loading and shrinkage-pre-cracking

plus fatigue
6.7.2.1 First monotonic loading
Most of the laboratory tests are related to first monotonic loading (or first monotonic straining)
cases.
A thorough investigation carried out in Toulouse (France) over ten years demonstrated the
correlation between cracking and debonding and confirmed the beneficial effect of fibre
reinforcement of the overlay[21, 4445, 5154, 55]. The case of debonding of mechanical origin was
approached by threepoint flexure tests with the overlay on the tension face.
To limit scatter of results, the concrete substrates were replaced by steel hollow shapes designed
and treated to simulate concrete substrates. Their inertias were similar to that of concrete substrates
of same sizes. The interface with the overlay was treated in order to have the same characteristics as
a concretetoconcrete interface: the surface of the base was milled to a rough finish, and cycles of
casting, curing and debonding of a cementbased overlay were repeated until the metal was covered
by a thin, uniform adhesive layer of mortar (for more details see [55]). The interface bond shear
strength was in the order of twice its tensile strength. Actually, with a 50 MPa 28day compressive
strength mortar as overlay, measured bond tensile strength and bond shear strength were about
0.5 MPa and 1 MPa, respectively. The test samples were all 100 mm wide with a flexural span of 430
mm. The height of the substrates varied from 50 to 150 mm and the thickness of the overlays (tensile
side) from 20 to 60 mm. The curvature of the structure was characterised by the deflection measured
at midspan. Depending on the test series, debonding initiation and propagation were detected by
three LVDTs or by five electrical strain gages or a set of breaking electrical conductive links spanning
the overlaysubstrate interface. Visual observation with a magnifying device (X25 at least) completed
the information.
Two types of fibres were investigated: usual hooked steel wire fibres and flexible ribbonshaped
amorphous metal fibres, both types were 30 mm in length. The first ones were designed to slip inside
the concrete matrix and not to break and were designated as slipping fibres; they provided a long but
relatively low postcrack strength plateau. The second ones, characterized by a high bond with the
concrete matrix, provided a high but short postcrack strength plateau.
Such tests, illustrated in Figure 6.8, demonstrated the beneficial effect of a fibre reinforcement of
the overlay. As long as there is no debonding, the devices spanning the interface measures
compression. When debonding occurs, tension is indicated, rightwards on the plot of figure 6.8. It
can be seen that, without fibres, debonding occurs at the same deflection as the first crack and
instantly propagates over all the investigated length along the interface. With 35 kg/m3 of high bond
metal fibres, the first crack initiated debonding but there was no propagation. A second crack, which
opened in the vicinity of LVDT 2, initiated another debonding at this location, but still with no
significant sign of propagation. Obviously, the fibre reinforcement of the overlay had a beneficial
effect.
In the case of a first monotonic loading, 35 kg/m3 of high bond fibres were necessary to bring a
significant delay of debonding. With slipping steel fibres, 80 kg/m3 were necessary for a similar result.
Figure 6.8: Debonding of mechanical origin: an example of test results.
Debonding is spotted by the signals of 3 LVDTs. The diagram indicates the evolution of
signals of the transducers 1, 2 and 3 with beam deflection, debonding is evidenced by the
rightwards move of the plot [52, 54].
More detailed information is provided by the modeling of these tests, by Sabathier [43]. A
sample of the results presented in Figure 6.9 clearly illustrates that fibre reinforcement by
35 kg/m3 of high bond metal fibres of the overlay significantly slows down the rate of
debonding propagation. But the beginning of the curves shows that, in the present case of a
first monotonic loading, this beneficial effect is not clearly visible on the debonding initiation
nor on the very first phase of debonding propagation.
100
80
Debonded length (mm)

60
40
20 W ithout fibers
W ith fibers
0
0 50 100 150 200 250 300 350
Mid-span deflection (m)
Figure 6.9: Effect of a fibre reinforcement of 35 kg/m3 of high bond metal fibres on debonding initiation and
propagation: a 20-mm thick overlay on a 50-mm deep substrate, modelling by Sabathier (2003b) of Chausson's
tests (1997)
Bernard [24] and Habel et al. [56] carried out a study focusing on the case of hybrid reinforced
concrete elements. It combined experiments and modeling. For this purpose, the thickness of the
overlays exceeded 100 mm. Given a concrete substrate 150 mm thick, different overlay thicknesses
were investigated in the range of 70 to 170 mm. The focus was on debonding initiation, rather than
the extent of debonding propagation. Concerning the role of fibre reinforcement, the results
presented in Figure 6.10 are in agreement with those presented above. Reinforcement with steel
bars or welded fabric concentrates higher crackbridging forces, thus a more significant effect is
present.
From these results, the authors concluded that it is better to improve the bond tensile strength or
the overlay tensile strength rather than to add fibres. Moreover, they demonstrated that
reinforcement content required in the new layer to avoid initiation of debonding is a function of the
ratio of thickness of the new to old concrete. With a high ratio, it is more difficult to control
debonding, and the interfacial adherence strength must be high to ensure monolithic behaviour.
These sets of investigations in first monotonic loading condition confirm that 80 kg/m3 of hooked
ended steel fibres have a small effect on the structural behaviour and have a minor effect to delay
debonding initiation.
where: Thickness of the new/old concrete : hnew/hold
Tensile strength of the new/old concrete : fct,new/ fct,old
Tensile strength perpendicular to the newold concrete interface: fct, int
Content of reinforcement in repair layer:
Fibre reinforcement 80 kg/m3 hooked end steel fibres
Diameter 10 mm with a spacing of the rebars = 150 mm
Diameter 14 mm with a spacing of the rebars = 80 mm
Figure 6.10: Required properties of the interface to avoid delamination: area below the curves indicates
partial or total delamination will occur [56]
6.7.2.2 Shrinkage-pre-cracking and fatigue loading

In accordance with some results from first monotonic loading presented in previous sections, a
minimum amount of 80 kg/m3 of common steel fibres is needed to observe an improvement, which
is in the range of three times more than the 22 to 27 kg/m3, which demonstrated a total efficiency in
the Montreal repair.
Farhat et al. [57], Turatsinze et al. [58], and the analysis of the last results of Sabathier and
Turatsinze [10, 55] explain this very large difference by a series of hypothesis on the cumulative
effects of restrained shrinkage, inducing early cracking, and of fatigue, damaging the concrete
interlock in the cracks of the overlay. Indeed, in real slabs on soil or pavements the maximum
expected curvature placing the overlay in tension is much less than the curvature necessary to
initiate cracking in the case of a first mechanical monotonic loading, and it is the effect of restrained
shrinkage which makes cracking possible, designated as "shrinkageprecracking". In addition, the
fatigue type loading and vibrations caused by the vehicular wheel loads accelerates cracking
initiation. At this low level of curvature, a low quantity of fibres can have a visible effect. After cracks
have been initiated, still at this low level of curvature, fatigue causes their gradual propagation by
gradual damaging of the interlocking in the cracks. Considering interlocking by two components, the
one provided by the cementitious matrix and the one brought by the fibres bridging the crack, the
portion coming from the fibre reinforcement is much less affected by fatigue than the one coming
from the cementitious matrix [5960]. This asymmetry in the cracking, in the case of fatigue,
indicates that the global influence of the fibre reinforcement is more visible. That is schematically
illustrated in Figure 6.11.
Farhat et al. [57] and Turatsinze et al. [58] provided an experimental confirmation of this
hypothesis. The experimental setup, similar to the one presented in Figure 6.8 is detailed in Figure
6.12. It was completed using a prenotch when casting located at midspan of the overlay. Due to
shrinkage, this notch initiated a crack in the overlay, at zero curvature, before the time of testing (at
the 7th day). Moreover, the notch located the crack and made possible to fit a LVDT sensor to
monitor the crack opening.
The metal bases were 150 mm high and the overlay 60 mm thick. The samples were tested in
sinusoidal fatigue loading at a frequency of 5 Hz between 0.1 kN and 9 kN. The maximum load (9 kN)
corresponded to the lowest load necessary to propagate cracking from the tip of the notch when the
overlay was protected against drying shrinkage. The associated curvature was significantly lower
than the one imposed during the "first monotonic loading" type tests, performed on notched
specimens. Three types of mortar overlays were investigated. Mix M0 without fibre, mix M1
reinforced with 40 kg/m3 of 30 mm long hooked steel fibres (slipping fibres) and mix M2 reinforced
with 20 kg/m3 of 30 mm long high bond metal fibres. The main mechanical characteristics of the
mixtures are given in Table 2, in terms of fc and ft, the compressive and tensile strength values
respectively and of r, wc, wl that characterise the postcrack loadbearing capacity of the mixture as
indicated in Figure 6.13. These last three parameters are obtained from a tensile test on notched
specimen.
Case of a "first monotonic loading": After fatigue: the interlocking in the cementitious
debonding occured early and even 30 kg/m3 matrix is greatly damaged (it vanishes at thinner crack
of fibres was not observed to have any effect. opening), the crack bridging capacity of the fibre
reinforcement is affected but still remains significant.
In such conditions, 30 kg/m3 of fibres is effective in
delaying debonding.
(a): without fibres
(b): with high fibre dosage (80 kg/m3) of slipping steel fibres
(c): with usual fibre dosage ( 30 kg/m3) of slipping steel fibres
wdi: crack opening corresponding to the debonding initiation
Figure 6.11: Schematic illustration of the effect of fatigue on the influence of a fibre reinforcement on
debonding initiation
Figure 6.12: Experimental setup for "shrinkageprecracking and fatigue" tests, details of the baserepair interface
and of the repair notch [58]
Table 6.2: Main mechanical characteristics of the repair materials
M0 M1 M2
fc (MPa) 60 55 51
ft (MPa) 3.5 3.4 3.4
r (MPa) 1.3 1.6
wc(mm) several 0.2
wl (mm) several 1.0
Figure 6.13: Post-crack load-bearing capacity versus crack opening
The results are presented in Figure 6.14. They confirm that, in "shrinkageprecracking plus
fatigue" condition, fibre reinforcement significantly less than 80 kg/m3 of slipping fibres can have a
significant effect to delay debonding initiation and propagation. They also indicate that, although
their short postcrack strength plateau, the high bond fibres remain fully efficient: 20 kg/m3 of metal
high bond fibres are more efficient than 40 kg/m3 of metal slipping fibres.
Debonding propagation (mm)
Figure 6.14: Case of "shrinkage-pre-cracking plus fatigue", debonding propagation versus cycles number [21,
55, 57].
Gagn et al. [61], Lemieux et al. [62] and Guindon [63] generated additional experimental data that
confirm and clarify some of the debonding mechanisms previously described. They notably
confirmed that, at load level representative of actual concrete slabs, debonding usually does not
start during the first monotonic loading but gradually develops under cyclic loading. It should be
noted the effect of the interlocking damaging by fatigue. Their experimental program was designed
to provide a better understanding of the structural behaviour and durability of thin bonded concrete
overlays. An additional purpose was to provide technical recommendations in order to improve the
techniques performance in practice. It involved testing 15 simply supported reinforcedconcrete slab
panels (3.3 m 1.0 m 0.2 m) designed to support actual service loads. Slab design, dimensions and
steel reinforcement were designed to accurately simulate the characteristics of typical panels in
bridge decks.
Some slab panels were hydrodemolished to a depth of 20 mm to simulate shallow

deterioration (nonexposed reinforcing steel) while others were hydrodemolished to a depth of 95
mm to simulate deep degradations (exposed reinforcing steel). Five types of concretes were used
to produce the overlays: 35 MPa, 35 MPa + 5 % Latex, 35 MPa + steel fibres, 50 MPa, and 50 MPa +
steel fibres.
All fibrous concretes contain 40 kg/m3 of 60mm hooked steel fibres. All slabs were subjected to
cycling loading (up to 500,000 cycles) and monitored for load, displacement, flexural cracking and
debonding (from observation of lateral faces). Figure 6.15 shows that flexural tests were designed to
simulate either zones of positive (repair in compression) or negative moments (repair in tension).
Overall, no debonding was observed with the overlay located in the compression zone [62]. This
absence of damage was mainly responsible for the good structural behaviour of this category of slab
panels, regardless of the configuration type assessed. Results strongly suggest that it is possible, in
the case of an overlay located in compression zones, to develop and maintain monolithic structural
action in the substrate/overlay composite.
Some debonding occured for overlays located in tension zones. For this loading mode,
debonding was influenced by interface location (above or below reinforcement), fibres, type
of concrete and number of cycling loading. When debonding occurred, it usually did not start
during the first monotonic loading but gradually developed under cycling loading. Abrupt
debonding (first loading) only occurred with the two non-fibre-reinforced high-strength
concrete overlays.
Figure 6.15: Overlays location and loading mode [62]
Overall, for overlays located in tension zones, the results indicated that, associated to their
cracking, the debonding trend increased with the curvature imposed to the overlaid structure.
Once the flexural cracks reached the interface, tensile stresses developed at the interface
between the two concrete layers because of peeling and because the overlay resisted the
curvature. The behaviour was similar to that of a plane rigid concrete plate stuck to a slightly
curved surface; it was the geometrical effect. In case of debonding of mechanical origin,
peeling and geometrical effect are tightly linked. Tensile stresses developed perpendicular to
the plane of the interface along plate edges. As the curvature of the support increased, the
tensile stresses generated at the edges of the plate increased (Figure 6.16).
These tensile stresses were probably not large enough to initiate debonding under the first
monotonic loading. However, the results showed that the level of stress under cyclic loading was
sufficient to generate fatigue debonding and edge curling due to differential curvature between the
substrate and the overlay. According to these mechanisms, thick or highly rigid overlay concretes will
form rigid blocks between flexural cracks. Each block strongly resists the curvature imparted by the
substrate slab and, therefore, develops high tensile stresses perpendicular to the plane of interface
along block edges. This mechanism was experimentally confirmed by the sudden debonding of high
strength nonfibrereinforced overlays during the first monotonic loading of the slabs [6263].
The use of relatively low fibre content (40 kg/m3 anchored steel fibres) contributed to reduce the
risk of interface debonding. Experimental results indicated that overlays containing fibres had a
higher number of finer flexural cracks [63]. The bridging action of fibres and formation of smaller
blocks in the fibrereinforced overlays contributed to reduce the magnitude of tensile stresses
perpendicular to the interface.
Erreur ! Objet incorpor incorrect.
Figure 6.16: Evolution of tensile stress imposed by the curvature of the substrate [63]
Interface debonding was reduced for thick overlays containing reinforcing bars. The action of the
reinforcing bars seems to have more uniformly distributed tensile stresses at the interface. The
blocks are thus smaller and, consequently, interface debonding is less severe [62].
In general, the results showed that the use of a bonded concrete overlay in a tension zone does
not compromise the structural capacity of a reinforced concrete slab when the repair is carried out
with appropriate techniques. Damage at the overlay/substrate interface has little effect on the
structural capacity of the repaired slab, however, because the reinforcing steel under tension
provides greater resistance to the generated tensile stresses [62].
6.7.2.3 Conclusion
Shrinkage precracking plus fatigue, representative of actual insitu conditions, drastically change
the levels of the relevant parameters involved in debonding initiation and propagation. Any reliable
attempt of behaviour prediction must take that into account.
First, shrinkage (or more generally differential length changes of the overlay and the substrate)
induces builtin stresses that make cracking occur at very low, even nil, curvature levels. Second,
fatigue progressively damages the available interlocking in the crack and along the debonded area of
interface. Over the long term, only the interlocking provided by a reinforcement of the overlay will
be significant, and consequently it will then play a major role.
On this basis, for a realistic calculation including fatigue effects, a conservative solution would be
to neglect all interlocking excepted that provided by overlay reinforcement. Figure 6.17 presents the
results of Sabathier [43], modeling Chausson's tests [54]. They indicate that the calculated crack
opening at debonding initiation is less than 10 m and that, for a debonded length of 180 mm (90
mm each side of the initiating crack), the crack opening remains thinner than 120 m.
6.8 Crack Propagation And Crack Opening

Fibres and other reinforcements act to delay debonding by their ability to restrain crack opening.
What crack opening width is the limit to prevent debonding?
6.8.1 First monotonic loading
In this straining condition, the cracks propagate quickly through the overlay, and their opening at
debonding initiation is very small. Meanwhile, it is difficult to get an accurate experimental
measurement. To investigate this initial phase, modeling is of great help. To investigate the
subsequent phase, debonding propagation, experimental measurements and modeling are
complementary.
Figure 6.17 presents the results of Sabathier [43], modeling Chausson's tests [54]. They indicate
that the calculated crack opening width at debonding initiation is less than 10 m and that, for a
debonded length of 180 mm (90 mm each side of the initiating crack), the crack opening remains
thinner than 120 m.
Tests carried out by Farhat [21, 57, 58], with the test fitting already described in subsection 6.7.2
but in monotonic loading, brought a complementary experimental information. They were different
from the above calculation by the shrinkageprecracking and a lower fibre dosage (20 kg/m3). The
consequence of precracking was larger crack openings. As indicated in figure 6.18, crack openings
ranged from 0.05 mm at debonding initiation to about 0.2 mm for a total debonded length of 160
mm (80 mm each side of the initiating crack). It confirms that to prevent or limit debonding, the
reinforcement must act as soon as the cracks appear and that beyond a 0.2mm crack opening it is
too late.
Figure 6.17: Crack opening versus debonding propagation: modeling of Chausson's tests [54] by
Sabathier [43] (a 20-mm thick overlay on a 50-mm deep substrate; with fibres: 35 kg/m3 of high
bond metal fibres)
Figure 6.18: Crack opening versus debonded length each side of the initiating point: case of
monotonic loading after shrinkage pre-cracking (M0: plain mortar, M1: reinforced with 40 kg/m3
of slipping fibres and M2: reinforced with 20 kg/m3 of high bond fibres) [21, 58]
6.8.2 Shrinkage-pre-cracking and fatigue loading

The monitoring of the Montreal motorway repair concluded that, in the case of the fibre
reinforced repairs, debonding was prevented and, except the cracks reflected from the substrate,
only very thin hairline cracks were observed. Moreover, most of them had not propagated through
the full depth of the overlay.
This trend is confirmed by the tests performed by Farhat [21], Turatsinze et al. [55] and Farhat et
al. [57] already presented in section 6.7.2. Their results demonstrated that, in insitu condition
(shrinkageprecracking and fatigue at a low curvature level), the crack actually propagated gradually
through the overlay. That is illustrated in Figure 6.19. The results showed a first phase of practically
instantaneous propagation until a depth related to the maximum imposed curvature. Afterwards, in
the case of fibre reinforcement, the crack propagation continued gradually, caused by the gradual
damaging of the interlocking along the opencrack length. In these tests, more than 60,000 cycles
were necessary for the crack to reach the interface. In that sense, once again, the high bond fibres
provided the best service. On the other hand, without reinforcement the crack propagated very
quickly.
Figure 6.19: Shrinkage-pre-cracking and fatigue tests: crack propagation through the depth of the overlay
versus the number of cycles [21, 55, 57].
The evolution of the measured crack width versus the debonded length is presented in Figure
6.20. The crack opening for a 160mm debonded length (80 mm each side of the initial crack) after
two millions cycles remained limited. It did not exceed 0.2 mm. In the case of high bond metal fibres
it was even less than 0.15 mm. The crack opening at debonding initiation, about 0.05 mm, reflected
the effect of shrinkage precracking.
Figure 6.20: Crack opening versus debonded length each side of the initiating point: case of shrinkage-pre-
cracking plus fatigue loading. M0: plain mortar, M1: reinforced with 40 kg/m3 of slipping fibres and M2:
reinforced with 20 kg/m3 of high bond fibres [21, 55, 57].
All these results converge to indicate that the prevention of debonding initiation or propagation
must be achieved as soon as a crack opens and certainly before a 0.2mm crack opening. The
European Standard 15043 "Products and systems for the protection and repair of concrete
structures Structural bonding" also confirms this point. It systematically accompanies the limit
bond strength requirements by the statement " with maximum average crack width 0.05 mm
with no crack 0.1 mm ".
6.9 Special Overlays

6.9.1 Stang and Walters results [11, 64-65]
Stang and Walter developed and investigated a cementbased overlay stiffening a steel structure,
typically an orthotropic steel bridge deck. Their studies included experiments and modeling that
emphasized similarities and differences with more conventional overlay systems.
The specimens were made of a steel plate, 8 mm thick, overlaid with a bonded layer, 40 mm thick,
of selfcompacting fibre reinforced cementbased material. The bond was ensured by pouring the
selfcompacting cementbased material onto the sand blasted steel plate. Since crack formation in
the overlay and associated debonding under negative bending was of particular concern, a test set
up was used, simulating a part of a stiffening overlay cast on a steel bridge deck, loaded in negative
bending, using a composite beam turned upside down (Figure 6.21).
Figure 6.21: Schematically experimental setup simulating a cement-based overlays (in this case FRC) cast on a
steel plate [65]
Two main classes of fibre reinforced materials were tested: strain softening and hardening.
Furthermore, in the strain softening class, two materials were tested, ordinary Fibre Reinforced
Concrete (FRC) and Fibre Reinforced DENSIT (FRD). The two strain softening materials differ in their
different stresscrack opening relationships (Figure 6.22a) determined in uniaxial tension using
notched specimens according to the RILEM standard. The FRD material is stronger than FRC and
supports, experimentally, the hypothesis that debonding is closely related to formation of bending
cracks in the overlay. Furthermore, a special strain hardening material, an Engineered Cementitious
Composite (ECC), was tested [66]. Figure 6.22b shows its response in uniaxial tension. Note that due
to the strain hardening characteristics of the material and the distributed nature of the crack pattern,
the uniaxial response in tension is presented as stress vs. strain. The motivation for using this
material was to determine how debonding is affected in the case where no macro bending cracks are
formed in the overlay. As for this material, when exposed to uniaxial tension and also in negative
bending, multiple cracking was formed in a strain range from 0 3 %.
Figure 6.22: (a) Uniaxial response for Fibre reinforced DENSIT (FRD) and Fibre reinforced concrete (FRC);
(b) Uniaxial response for ECC (Walter et al. 2004)
The results from the bending tests with composite beams are presented in a load vs. deflection
diagram (Figure 6.23). A second yaxis couples debonding of the overlay to the overall response. The
study supports the fact that debonding is initiated from a defect or macro crack in the overlay.
Furthermore it was shown that when no macrocrack formed in the overlay, which is the ECC case,
until a deflection of about 15 mm debonding was not initiated. In the case of FRD, debonding was
initiated for a higher load level due to the higher tensile strength and resistance to crack
propagation; however debonding was initiated for the same amount of deflection.
Figure 6.23: Experimental results for three fibre reinforced cement-based materials [65]
All cases have been modelled numerically using finite element software [65]. The studies
concluded that the interfacial interlocking effect, mixed normal and shear stresses, has to be taken
into account.
The proportions of the investigated specimens can explain the difference with the other results,
especially results not reported here related to crack opening vs. debonding initiation and
propagation. The overlay was several times thicker than the substrate as opposed to conventional
overlay systems. Consequently, the overlaysubstrate interface was on the compression side of the
composite specimen. Initially, the whole specimen behaved as a fibrereinforced concrete beam with
a steel skin on its compression side. Debonding was expected late when a macro crack propagated
deeply enough through the overlay, initiating tensile stresses perpendicular to the interface.
6.9.2 Anchoring of the overlay

In the continuation of the tests by Gagn et al. [61], Guindon [63] and by Lemieux et al. [62]
presented in 6.7.2, Benzerzour [67] used the same experimental approach as that used by Guindon
[63] and Lemieux et al. [62] to analyze the effect of mechanical anchorages on the performance of
overlays located in a tension zone. Anchors were commercial steel rods having a length of 114 mm
and a diameter of 9.5 mm. They were fixed to the substrate by inserting them into drilled holes (84
mm depth) partially filled with a special epoxy resin.
A reference reinforced concrete slab (3 x 3 x 0.18 m) was repaired with a 40mm bonded overlay
(nonanchored) made with a 48 MPa airentrained concrete. Two other slabs were overlaid with the
same type of concrete (thickness of 40 mm) but with different anchor configurations. One
configuration consisted of a uniform pattern of anchors equally spaced every 0.2 m in the
longitudinal direction and every 0.13 m in the transverse direction. The other configuration was
similar except that the longitudinal spacing was reduced to 0.1 m in a 1mlong central segment of
the slab.
All the slabs were subjected to cycling loading according to the procedure used by Guindon [63]
and Lemieux [62]. For the reference slab, fatigue failure occurred after 490 000 loading cycles. During
the test, several debonding cracks appeared on the lateral faces of the slab in the midspan section
[67]. No sign of debonding was found for either type of anchored overlays. The flexural cracking
patterns were the same for the three slabs, indicating that no preferential cracking was induced by
the anchor pattern. These initial results suggested that the use of mechanical anchors is technically
possible and could help reduce the risk of debonding. However, more research is needed to optimize
the anchor pattern and to assess the cost/benefit ratio of this approach more precisely.
A Swedish experience of overlay anchoring is also discussed in chap 4, section 4.4.8.
6.10 Boundaries And Joints Of The Overlays
6.10.1 Boundaries and full depth joints

Boundaries and full depth joints are the most critical zones because lack of reinforcement to
compensate for the associated discontinuity. Consequently, they must be placed at locations of
lowest stressing.
Debonding is the consequence of builtin stresses coming from the cumulative effects of length
changes, essentially shrinkage, and negative moment putting the overlay in tension.
To have the least stress at the boundaries or joints, they must be located where there is no
negative moment, i.e. exactly with the top of the joints of the substrate.
Else, if the overlay is continuous over a joint of the substrate, there is a great risk that the
unavoidable moves of the substrates joint induces a reflected crack in the overlay. Consequently, as
far as possible the edge and joints pattern of the overlay should reflect the one of the substrate.
The through cracks of the substrate, if not carefully reinforced, will behave as joint.
6.10.2 Sawn joints

If, as is recommended, the overlay is fibrereinforced, the depth of the joint must be the least
possible in order to preserve the maximum force transmission capacity through this weakened
section. A reasonable depth is at least a quarter and no more than a third of the overlay thickness.
If necessary such fibre reinforced joints do not have to always match joints of the substrate. If the
overlay is reinforced by steel bars or welded fabric, the saw cut must preserve a sufficient cover of
the steel.
6.11 Conclusion
Debonding initiates from a mechanical discontinuity (e.g. a crack, a joint or an edge) in an overlay
tensioned by the combined effects of the mechanical loading of the structure and of the differential
length changes between the overlay and its substrate. From the discontinuity, debonding is always
initiated by tension perpendicular to the interface. The main origin of the debonding tensile stress
perpendicular to the interface is the peeling effect. Secondarily, the geometrical effect resulting from
the curvature of the structure placing the overlay in tension (Figure 6.1) and the curling trend of the
overlay because of its non uniform shrinkage add to the effects. Consequently, the relevant
characteristic of the interface to prevent debonding is its tensile strength perpendicular to the
interface.
Reinforcement of the overlay by fibres, steel bars, or welded fabric sustains the ability to maintain
a partial continuity of throughcrack force transmission. This feature acts positively to delay or
prevent debonding.
When modeling the debonding mechanism, an essential feature is to take into account the
interlocking effect, which exists between the two faces of the crack and of the debonded interface.
Tests or modeling in "first monotonic loading or straining" condition are efficient to investigate
the debonding mechanism. However they fail to predict behaviour and durability insitu conditions.
In these last conditions only "shrinkageprecracking and fatigue" investigations are relevant. The
reason is that, on the one hand, precracking debonding may occur at lower curvature level. On the
other hand, fatigue significantly alters interlocking.
The beneficial effect of fibre reinforcement is almost inconsequential in the case of first
monotonic loading. On the other hand, it has a major effect in the case of "shrinkageprecracking
and fatigue". Moreover, debonding has to be prevented in the early phase of cracking. When the
crack width reaches 0.2 mm, it is too late to prevent debonding. Fibre reinforced cementbased
overlays are highly recommended, and high bond fibres provide the best performance.
REFERENCES
[1] Carter, A., Gurjar, S. and Wong, J., Debonding of highway bridge deck overlays, Concrete
international, 24 (7) (2002) 51-58.
[2] Silfwerbrand, J., Improving concrete bond in repaired bridge decks, Concrete international, 12
(9) (1990) 61-66.
[3] Silfwerbrand, J. and Paulsson, J., Swedish experience: Better bonding of bridge deck overlays,
Concrete international, 20 (10) (1998) 56-61.
[4] Saucier, F. and Pigeon, M. , Durability of new-to old concrete bondings, Proc. Of Aci Int.
Conf. On evaluation and rehabilitation of concrete structures and innovations in design, Hong
Kong, December ,1991 (ACI SP-128) 689-705.
[5] Langlois, M., Pigeon, M., Bissonnette B. and Allard D., Durability of pavement repairs: a field
experiment, Concrete international, 16 (8) (1994) 39-43.
[6] Delatte, N.J, Wade, D. M and Fowler, D. W. , Laboratory and field testing of concrete bond
development for expedited bonded concrete overlays, ACI Materials journal, 97 (3) (2000)
272-280.
[7] Tschegg, K.E, Igruber, M., Srberg, C.H; and Mnger F., Factors influencing fracture behavior
of old-new concrete bonds, ACI materials journal, 97 (4) (2000) 447-453.
[8] Vaysburd, A.M. and Emmons, P.H., How to make todays repairs durable for tomorrow-
corrosion protection in concrete repair, Construction and building materials 14 (2000) 189-197
[9] Emmons, P. H., Vaysburd, A.M., Czarnecki, L., Durability of repair materials : current
practice and challenges, Proceedings of an International Symposium? Warsaw 2000, (Brittle
Matrix Composites 6) 263-274.
[10] Sabathier, V., Granju, J-L., Turatsinze, A. and Bissonnette B., Repair by cement-based thin
overlays interlocking at the interface and modeling of debonding, in Industrial floors'03,
Proceedings of an International colloquium, Esslingen, January, 2003 (Editor: P. Seidler) 621-
626.
[11] Walter, R., Stang, H., Olesen, J.F. and Gimsing, N.J., Debonding of FRC composite deck
bridge, in Brittle matrix composites 7, Proceedings of an International Symposium, Warsaw,
October 2003 (Woodhead publishing limited) 191-200.
[12] Lupien, C., Chanvillard, G. Atcin, P-C. and Gagn, R., Rhabilitation d'une chausse par
resurfaage mince adhrent en bton renforc de fibres d'acier, Proceedings of AIPCR, comit
C-7, Montral, Canada (1995) 246-250.
[13] Scott, M., Rezaizadeh, A., Delahaza, A., Santos, C.G., Moore, M., Graybeal, B. and Washer G.,
A comparison of non destructive evaluation methods for bridge deck assessment, NDT & E
International 36 (4) (2003) 245-255.
[14] Clark, M. R., McCann, D. M. and Forde M. C., Application of infrared thermography to the
non-destructive testing of concrete and masonry bridges, NDT & E International, 36 (4) (2003)
265-275.
[15] Lupien, C., Chanvillard, G. Atcin, P-C. and Gagn R. , Rhabilitation d'une chausse en bton
avec une chape mince en bton renforc de fibres d'acier, In les techniques de transport au
service de la qualit de vie,Expos des communications du 25congrs annuel de l'AQTR,
Montral, Avril, 1990, 108-122.
[16] Bungey, J.H., Sub-surface radar testing of concrete: a review, Construction and Building
Materials 18 (2004) 18.
[17] Dalhuisen, D.H. , Stroeven, P., Moczko, A.T. and Peng, X., Modern Testing Methods for "In-
Situ" Non-Destructive Examination of Concrete Structures, In: Proceedings of The
NOCMAT/3-Vietnam 3rd International Conference of Non-Conventional Materials and
Technologies, Hanoi, Vietnam, March 2002 (K. Ghavami and Nguyen Tien Dich eds., ISBN 80-
902992-1-0) 253-259.
[18] Moczko, A.T., Pszonka, A. and Stroeven, P., Acoustic Emission as a useful tool for reflecting
cracking behaviour of concrete composites, Proceedings of International Symposium Non-
DestructiveTesting in Civil Engineering (NDT-CE), Berlin, September 1995 (Edited by G.
Schickert, H. Wiggenhauser ISBN 3-931 381-02-1) 805-812.
[19] Chausson, H. and Granju J.-L., Optimized design of fiber reinforcement thin bonded overlays",
in Brittle matrix composites 5, Proceedings of an International Symposium, Warsaw, October
1997 (Woodhead publishing limited) 133-142.
[20] Chowdhury, M.R. and Ray, J.C., Accelerometers for bridge load testing, NDT & E
International 36 (4) (2003) 237-244.
[21] Farhat, H., Durabilit des rparations en bton de fibres: effets du retrait et de la fatique, PhD
thesis (in French), Universit Paul Sabatier, Toulouse France, 1999, 178 pages.
[22] Basu, B. (2005), Identification of stiffness degradation in structures using wavelet analysis,
Construction and building material 19 (9) (2005) 713-721.
[23] Granju, J-L., Debonding of thin cement-based overlays, Journal of Materials in Civil
Engineering 13 (2) (2001) 114-120.
[24] Bernard, O., Comportement long terme des lments de structure forms de btons d'ges
diffrents, PhD thesis n 2283, Ecole Polythechnique Fdrale de Lausanne, Suisse, 2000, 189
pages.
[25] Bigwood, D.A., Grocombe, A.D., "Elastic analysis and engineering design formulae for bonded
joints." International Journal of Adhesion and adhesive, 9, (1989) 229-242.
[26] Fowler, D.W., Wheat, D.L., Choi, D.U. and Zalatimo, J., Stresses in PC overlays due to thermal
changes, in Industrial floors'03, Proceedings of an International colloquium, Esslingen,
January, 2003 (Editor: P. Seidler) 29-36.
[27] Naciri, T., Ehrlacher, A. and Chabot, A., Interlaminar stress analysis with a new multiparticle
modelisation of multilayered materials (M4), Composites sciences and technology 58 (3)
(1998) 337-343.
[28] Caron, J.F., Diaz Diaz, A., Carreira, R.P., Chabot, A. and Ehrlacher, A., Multi-particle
modelling for prediction of delamination in multi-layered materials, Composites sciences and
technology , 66 (6) (2006) 755-765.
[29] Granju, J.-L., Thin bonded overlays : about the role of fiber reinforcement on the limitation of
their debonding, Advanced Cement Based Materials 4 (1) (1997) 21-27.
[30] Balouch, S. U. and Granju, J.-L., Corrosion of different types of steel fibres in SFRC and
testing of corrosion inhibitors", in 'Infrastructure regeneration and rehabilitation Improving the
quality of life through better construction. A vision for the next millennium', Proceedings of an
International Conference, Sheffield, June-July 1999 (Sheffield Academic Press, 1999) 735-747.
[31] Silfwerbrand, J., Shear Bond Strength in Repaired Concrete Structures, Materials and
Structures, 36 (260) (2003) 419-424.
[32] Sabathier, V., Rechargements minces adhrents base cimentaire renforcs de fibres
mtalliques. Conditions de leur durabilit, modlisation et calcul, PhD thesis, Universit
Toulouse III, 2004, 190 pages.
[33] Julio, E.N.B.S, Branco, F.A.B. and Silva, V.D., Concrete-to-concrete bond strength. Influence
of the roughness of substrate surface, Construction and building Materials, 18 (9) (2004) 675-
681.
[34] Turatsinze, A., Farhat, H. and Granju, J-L. (2000), Durability of metal-fibre reinforced concrete
repairs: drying shrinkage effects, Proceedings of an International Symposium, Warsaw,
October 2000 (Woodhead publishing limited) 296-305.
[35] Granju, J.L., Sabathier, V., Turatsinze, A. and Toumi, A., Interface Between an Old Concrete
and a Bonded Overlay: Debonding Mechanism, Interface Science Journal, 12 (4) (2004) 381-
388.
[36] Grzybowski, M. and Shah, P.S., Shrinkage cracking of fiber reinforced concrete, ACI
materials journal 87 (2) (1990) 138-148.
[37] Marosszeky, M., Stress performance in Concrete Repairs, Proceedings of a RILEM
International Conference On rehabilitation of concrete structures, Melbourne, 1992, 467-474.
[38] Saucier, F. and Pigeon, M., Testing of superficial repairs for sidewalks in Canada, concrete
international 18 (5) (1996) 39-43.
[39] Banthia, N., Yan, C, and Mindess, S., Restrained shrinkage cracking in fiber reinforced
concrete : a novel test technique, Cement and concrete research 26 (1) (1996) 9-14.
[40] Mailvaganam, N., Springfield, J., Repette, W. and Taylor, D., Curling of concrete slabs on
grade, Construction Technology Update (44) (2000) 1-6.
[41] Suprenant, B.A and Malisch, R.W., Repairing curled slabs, Concrete construction 9 (1999)
58-65.
[42] Suprenant, B.A., A look at the curling mechanism and the effect of moisture and shrinkage
gradients on the amount of curling, Concrete international 24 (3) (2002) 56-61.
[43] Sabathier, V., Granju, J-L., Bissonnette, B., Turatsinze, A. and Tamtsia, B., Cement-based thin
bonded overlays : numerical study of the influence of a bond defect, in Brittle matrix
composites 7, Proceedings of an International Symposium, Warsaw, October 2003 (Woodhead
publishing limited) 181-189.
[44] Tran, Q.T., Toumi, A. and Turatsinze, A., Durability of an overlay-old concrete interface: the
role of a metal fibre reinforcement, in Brittle Matix Composites 8, Proceedings of an
International Symposium, Warsaw, October 2006 (Woodhead publishing limited) 409-419.
[45] Tran, Q.T., Toumi, A. and Turatsinze, A., Thin bonded cement-based overlays: numerical
analysis of factors influencing their debonding under fatigue loading, Materials and Structures
41 (5) (2008) 951-967.
[46] Betterton, R.M., Knutson, M.J. and Marks, V.J., Fibrous portland cement concrete averlay
research in Green County, Iowa Transportation research record, N 1040, TRB, National
Research Council, Washington DC, (1985) 1-7.
[47] Paulsson, J. and Silfwerbrand, J., Durability of repaired bridge deck overlays, Concrete
international 20 (2) (1998) 76-82.
[48] Verhoeven, K., Thin overlays of steel fiber reinforced concrete and continuously reinforced
concrete, state of the art in Belgium, Proceedings of the 4th International Conference on
Concrete Pavement Design and Rehabilitation, Purdue University, West Lafayette, IN, April,
1989, 205-219.
[49] Chanvillard, G.; Aitcin, P.C. and Lupien, C., Field evaluation of steel-fibre reinforced concrete
overlays with bonding mechanism, Transportation Research Record 1226, TRB, Washington
(1990) 48-56.
[50] Chanvillard, G., Aitcin and P.C., Thin bonded overlays of fiber-reinforced concrete as a method
of rehabilitation of concrete roads, Canadian Journal of Civil Engineering 17 (4) (1990) 521-
527.
[51] Belaghmas, A., Fissuration et dcollement d'une couche de bton adhrente un support, DEA
memory, LMDC, Gnie Civil INSA-UPS, 1993, Toulouse.
[52] Granju, J.-L. and Chausson, H., Serviceability of fiber reinforced thin overlays relation between
cracking and debonding, in ConChem, Proceedings of an International exhibition &
Conference, Brussels, November, 1995 (Verlag fr chemische industrie) 133-142.
[53] Granju, J.-L. and Chausson, H., Fiber reinforced thin bonded overlays: the mechanism of their
debonding in relation with their cracking, in Concrete repair, rehabilitation and protection,
Proceedings of an International Congress , Dundee, June, 1996 (E. & FN. SPON) 583-590.
[54] Chausson, H., Durabilit des rechargements minces en bton : relation entre leur dcollement,
leur fissuration et leur renforcement par des fibres, PhD thesis (in French), Universit Paul
Sabatier, Toulouse, France, 1997, 198 pages.
[55] Turatsinze, A., Granju, J.L., Sabathier, V. and Farhat, H., Durability of bonded cement-based
overlays: effect of metal fibre reinforcement, Materials and Structures 38 (277) (2005) 321-327.
[56] Habel, K., Brhwiler, E. and Bernard, O., The numerical investigation of delamination in
hybrid reinforced concrete elements Proceedings of International PhD Symposium in Civil
Engineering, Vienna, October, 2000 (Ed. by K. Bergmeister) 221-228.
[57] Farhat, H., Turatsinze, A. and Granju, J-.L., Durabilit des rechargements minces adhrents
soumis la fatigue mcanique, Proceedings (cdroom) of 14ime Congrs Franais de Mcanique,
Toulouse, Aot-Septembre, 1999.
[58] Turatsinze, A., Farhat, H. and Granju, J.-L., Influence of autogenous cracking on the durability
of repairs by cement-based overlays reinforced with metal fibres Materials and Structures 36
(264) (2003) 673-677.
[59] Zhang, J., Stang, H. and Li, V.C. , Crack bridging model for fibre reinforced concrete under
fatigue tension, International Journal of Fatigue 23 (8) (2001) 655-670.
[60] Rossi, P. (sous la direction de), Le dveloppement industriel des btons de fibres mtalliques,
conclusions et reconclusions, (Presse de l'cole nationale des Ponts et Chausses, 2002).
[61] Gagn, R., Bissonnette, B., Lachemi, M.and Lemieux, M., Vzina, Analyse du comportement
de resurfaages adhrents utiliss pour rparer des dalles en bton arm, 9e colloque sur la
progression de la recherche qubcoise sur les ouvrages d'art, Qubec, Mai, 11 pages.
[62] Lemieux, M., Gagn, R., Bissonnette, B. and Lachemi, M., Behavior of overlaid reinforced-
concrete slab panels under cyclic loading Effect of interface location and overlay thickness,
ACI Structural J. 102(3) (2005) 454-461.
[63] Guindon, M.-A., tude du comportement des resurfaages adhrents mcanismes
dendommagement et influence des paramtres de conception, Mmoire de matrise, Universit
de Sherbrooke, Dpartement de gnie civil.
[64] Walter, R., Stang., H, Gimsing, N. J. and Olesen, J. F., High performance Composite Decks
using SCSFRC, Proc. Fourth int. Workshop on high performance fiber reinforced cementitious
composites, Ann Arbor, June, 2003, 495-504
[65] Walter, R., Li, V.C. and Stang, H.: Comparison of FRC and ECC in a Composite Bridge Deck,
in: 5th International PhD Symposium in Civil Engineering, June 2004, Delft, The Netherlands,
pp. 477-484.
[66] Li, V.C.,: Advances in ECC Research, ACI Special Publication on Concrete: Material Science
to Applications, SP 206-23, 373-400.
[67] Benzerzour, M., tude exprimentale et numrique du renforcement des tabliers de pont en
bton arm par des resurfaages adhrents, PhD thesis (in French), Universit de Sherbrooke,
Dpartement de gnie civil, Canada.
7. DESIGN
M. Trevio (1), J.L. Granju (2), H. Beushausen (3), A. Chabot (4), H. Mihashi (5), J. Silfwerbrand (6)
(1) Center for Transportation Research, The University of Texas at Austin (TX), U.S.A.
(2) Laboratoire Matriaux et Durabilit des Constructions (LMDC), UPSINSA, Toulouse, France
(4) Laboratoire Central des Ponts et Chausses, Nantes, France
(5) Department of Architecture and Building Science, School of Engineering, Tohoku University,
Sendai, Japan
(6) Swedish Cement and Concrete Research Institute (CBI), SE100 44 Stockholm, Sweden
ABSTRACT
This chapter reviews the design methods typically use of designing Bonded Concrete Overlays (BCO).
Two levels of design are discussed: the design for strength and the design to resist debonding. For
the design for strength various methods are discussed: U.S. Army Corps of Engineers, Portland
Cement Association, American Association of Highway and Transportation Officials, reinforcement
design of overlaid continuously reinforced concrete pavements and procedures from other countries.
For design to resist debonding, methods from the U.S., Sweden, Japan, other countries in Europe are
described.
7.1 INTRODUCTION
The objective of this chapter is to review the design methods, codes or recommendations
currently in use for Bonded Concrete Overlays (BCO) for repairs, such as pavement rehabilitation.
Two levels of design must be distinguished.
The design for sufficient strength of the overlaid structure. For BCO, it assumes a perfect bond
between the overlay and the substrate. If that is the case, the design procedures are similar to
those of monolithic structures. Also, depending on the code, they may include the evaluation
of the structure to be overlaid and of its contribution to the strength in the final overlaid
structure.
The design to prevent debonding. Its role is to ensure a monolithic behaviour of the overlaid
structure (required in the first level above) and its durability. The calculations are applicable to
composite structures, and the behaviour of the overlaysubstrate interface is a prevalent
parameter.
7.2 DESIGN FOR SUFFICIENT STRENGTH

If a BCO project is represented as a series of stages, as illustrated in Fig. 7.1, the
design would correspond to the second phase of the BCO process, following the project
selection and preceding the construction. Once a BCO has proven an appropriate
rehabilitation choice for the pavement in question, during the project selection stage, the next
step is to propose a thickness design for the overlay and a reinforcement design. The
thickness design is based upon the condition of the existing pavement, the purpose of the
overlay, the projected design life for the rehabilitated structure, the historic and projected
traffic data, and the material properties of the existing pavement, as well as those of the new
overlay.
Project Selection
Design Quality Control
Construction
Quality Assurance
Figure 7.1: Design stage as part of the BCO process

The following discussion is focused on the common pavement BCO design
methodology utilized in the United States. However, the procedures described herein
are also applicable to other cases, such as bridge decks or slabs on grade.
Among the methods presented, the AASHTO procedure distinguishes itself by
the implementation of an interesting concept, the "Condition Factor" (CF) which
accounts for an eventual weakening (due to fatigue) of the overlaid substrate.
This section is comprised of two subsections. The first covers a discussion of
some of the most utilized BCO thickness design methods. The second presents
reinforcement design procedure for BCOs on continuously reinforced concrete
pavements (CRCP).
7.2.1 Overview
The overlay design is not very different from the design of a new pavement.
Many of the concepts involved in the design of a new pavement apply to a BCO
design as well.
7.2.1.1 Purpose of the overlay

The rehabilitation is aimed to remedy a specific situation, which the designer
knows from the project selection phase. The need for rehabilitation arises as the
pavement experiences a decrease in serviceability; this occurs as a consequence of
load applications, age, and traffic increases. The interaction of these factors is
manifested as structural and functional failures. When the overlay is being placed for
the purpose of structural improvement, the required thickness of the overlay is a
function of two components: the structural capacity of the existing pavement under
current conditions, and the structural capacity necessary to fulfill future traffic
demands.
However, if the pavement deterioration is a consequence of non-loading
factors, i.e., the purpose of the overlay is only functional improvement, the design
equations will render a minimal or zero thickness, in which case, a minimum
constructible BCO thickness will be enough to address the functional deficiency.
BCOs as thin as one inch ( 25 mm) have been used successfully on sound pavements
[1].
7.2.1.2 Basic design principle

The foremost structural characteristic that differentiates a BCO from other
rehabilitation concepts is that, by definition, the overlay behaves as a single unit in
conjunction with the existing pavement. Therefore, the structural capacity remaining
in the existing substrate is fully utilized. As such, it is accounted for in the design
equations, which contributes to reduce the thickness of the overlay required. This is
only attainable if the bond between overlay and substrate is achieved and maintained.
Therefore, if the purpose of the BCO is to remedy structural deficiencies, the
design is based on a simple equation: the BCO is designed by determining the
additional thickness of concrete needed to carry the anticipated traffic. Thus, the
equation is as follows:
DBCO = Df D (7.1)
where DBCO is the overlay thickness, Df is the required thickness to carry the future
traffic if the pavement were constructed new, and D is the effective existing pavement
slab thickness. The design methods vary in the way in which the existing pavement
contribution, Df, is determined and the way the original thickness is affected by the
pavement condition.
7.2.2 Design concepts

A great deal of information and decisions necessary for new construction
projects are also required in rehabilitation projects such as BCOs. Many of the
following concepts apply to both new construction and rehabilitation projects.
7.2.2.1 Overall standard deviation

Since there is variability inherent in traffic predictions, material properties,
quality control, and environmental conditions during construction, it is reasonable to
use a probabilistic approach in the design, rather than a deterministic one. The overall
standard deviation accounts for all the sources of uncertainty involved in the overlay
design, like the material properties and traffic data, adding flexibility to the design. If
the information corresponding to the materials characterization and traffic is deemed
accurate, the data will have a small standard deviation, rendering a thinner overlay.
On the other hand, if the information were not collected properly, or is not entirely
available or trustworthy, the data will have a large standard deviation resulting in a
thicker overlay. The AASHTO Guide [2] recommends a standard deviation of 0.39
for PCC overlays.
7.2.3 Current overlay design procedures

There are several BCO design procedures available. The basic concepts of
some of the most common procedures are discussed below.
7.2.3.1 Corps of Engineers

The U.S. Corps of Engineers procedure was originally devised for the design of
PCC overlays over PCC airfield runways and taxiways. It was developed using full-
scale accelerated test tracks [3].
This method uses a version of the basic design equation (7.1), modified by
empirical coefficients. In this method, the required thickness of the overlay is the
difference between the thickness required for a new pavement and the thickness of the
existing slab. In Equation 7.1, instead of the thickness of the existing slab, the
thickness considered is the effective thickness. Three models were developed,
namely, for the bonded, partially bonded and unbonded cases, represented by
Equations 7.2, 7.3 and 7.4, respectively. Even though the scope of this chapter covers
only the bonded overlay case, the other two are presented as a reference to provide a
better understanding of the methodology.
ho = hn he (7.2)
1.4 1.4
ho = 1.4 hn Che (7.3)
2 2
ho = hn Che (7.4)
where:
ho = required overlay thickness
hn = required theoretical thickness for the design loading, for a new pavement
he = existing pavement thickness, and
C = condition correction coefficient
The values for C are determined by visual inspection, and range between 1 (for a
pavement in good structural condition) to 0.35 (for a pavement with severe structural
defects).
As can be inferred by these equations, it is assumed that, for a BCO, the
condition correction coefficient equals one, which means that the existing slab is in
good structural condition or will be upgraded to this condition. Besides this
coefficient, the other difference among these equations is the exponent of the
thicknesses, which is related to the bonding characteristics of each type of overlay.
For a BCO, the value of the exponent is 1, given that the BCO and the substrate will
behave monolithically. In a similar fashion, it is equal to 1.4 and 2 in Equations 7.3
and 7.4, for the partially bonded and unbonded overlays, respectively.
In this method, it is implied that the existing concrete has suffered no fatigue
due to traffic or other factors, and it is as strong as the concrete in a new pavement,
which contradicts the fatigue damage concept explained above and the idea of
remaining life. Furthermore, it assumes that the failure mechanism of the overlaid
pavement is the same as that of a new pavement.
Failure is defined as the load application level at which cracking or structural
breakup first appears, which does not apply properly to highway concrete overlays,
where cracking is an inherent occurrence of portland cement concrete pavements.
7.2.3.2 Portand Cement Association

The PCA methodology consists in designing an overlay system that is
structurally equivalent to a new full-depth pavement placed on the same subbase and
subgrade. Unlike the Corps of Engineers procedure, it uses an evaluation of the
existing pavement by means of condition surveys, deflection tests, and in-situ sample
testing, to take its condition into consideration in the design.
The design basis is the analysis of the stresses at the edge of the pavement [4],
i.e., the stress at the bottom of the slab along the lateral edge of the slab plate. The
model equates the edge stress at the bottom of the new full-depth pavement (n) with
that of the overlaid system at the bottom of the existing pavement (e), as shown in
Fig. 7.2.
BCO
New
CRCP Existing
n e CRCP
Subbase Subbase
Subgrade Subgrade
Figure 7.2: Edge stresses for new and overlaid pavement for PCA method design equivalency
Because the new full-depth slab and the existing concrete will have different
moduli of rupture, Sc, the equivalency is based on the stress ratio to the modulus of
rupture. If the stress ratio for the overlaid system is the same than that of the new
pavement, both pavements will be structurally equivalent:
n e
(7.5)
Scn Sce
where:
n = critical stress edge in the new pavement
Scn = modulus of rupture of the new concrete
e = critical stress edge in the existing pavement
Sce = modulus of rupture of the existing concrete
In developing this method, a finite element program was used to create a

design chart in which the critical tensile stresses due to edge loading in both new
pavement and the BCO structure are related to the modulus of rupture of the existing
concrete, for which three different ranges of moduli are considered.
For the BCO design, the first step consists of calculating the thickness of the
new full-depth pavement, and this can be accomplished by using the PCA design
method or other PCC design method. With this thickness and the design chart, the
combined thickness of BCO and existing pavement is computed, and the BCO
thickness is determined by subtracting the existing slab thickness from this value. The
maximum BCO thickness recommended is 5 in (125 mm). When the required
thickness exceeds this value the use of an unbonded overlay is more advisable.
In this method, the fatigue consideration is dependent on the procedure used to
arrive at the new full-depth pavement thickness. If the Portland Cement Association
method is used, then it is assumed that the pavement can withstand an infinite number
of applications, as long as those occur under the stress limit established by the method,
which is based on plain concrete tests.
7.2.3.3 AASHTO
The AASHTO method, outlined in detail in the 1993 AASHTO Guide [2], is
based upon the AASHO Road Test, the basic design equation (7.1), the remaining life
concept, and thus, fatigue, and the idea of serviceability.
It is mostly an empirical method, since the design equations for the method were
derived from regression analyses performed on the Road Test data, but it includes a
mechanistic part, in the determination of stresses and strains. Like the PCA method,
the AASHTO procedure advocates conducting a comprehensive evaluation of the
existing pavement conditions, and applying the results as input design parameters for
the BCO.
The thickness design equation is:
Dol = Df Deff (7.6)
where:
Dol = required thickness of BCO
Df = slab thickness to carry future traffic
Deff = effective thickness of existing slab, calculated by applying a condition factor, CF,
to the existing slab thickness, D, as in the following expression:
Deff = CFxD (7.7)
The value of CF can be determined in two ways, either by the use of remaining life or
by means of the condition survey.
The value of CF ranges from 0 to 1. When the pavement condition is
satisfactory, it will take the value of one, or close to one, which means that the
condition of the pavement does not affect the effective thickness. However, as the
condition of the slab is more deteriorated, the value of the condition factor decreases.
Guidelines for selecting values for CF appear in AASHTO 1993.
The slab thickness to carry future traffic, Df, is calculated as if it were a
new pavement design.
As a reference to the reader, the AASHTO design equation for rigid pavements is provided
below:
PSI
log10
4.5 1.5
log10 W18 = Z R xSO + 7.35x log10 (D + 1) 0.06 +
1.624 x10 7
1+
(D + 1)8.46

(7.8)

Sc ' xCd x(D 0.75 1.132)
+ (4.22 0.32 pt )x log10

215.63xJ D 0.75 18.42
0.25

Ec

k
where
W18 = predicted number of 18-kip ESAL applications
ZR = standard normal deviate
SO = overall standard deviation of rigid pavement
D = thickness of pavement slab, in.
PSI = difference between initial serviceability, po, and terminal serviceability index, pt
Sc = PCC modulus of rupture, psi

J = load transfer coefficient
Cd = drainage coefficient
Ec = PCC modulus of elasticity, psi
k = modulus of subgrade reaction, pci
The first term (ZRSO) corresponds to the reliability. The remaining terms on the first line of the
equation are the empirical part of the procedure, derived from the data gathered at the AASHO Road
Test, while the second line, related to stress computations, is the theoretical part, which was added
to account for changes in strength and stresses owing to physical constants (e.g., Ec, k) occurring in
conditions other than those that existed during the road test.
7.2.4 Reinforcement design of overlaid continuously reinforced concrete pavements

(CRCP)
This section presents the design of the overlay reinforcement. The purpose of
the steel reinforcement in a CRCP is to hold cracked concrete together and to maintain
load transfer; it does not increase the pavement structural capacity. It is recommended
that the BCO be reinforced in a similar fashion as the existing pavement, i.e., similar
steel percentage, bar sizes and spacings, unless there is a significant deficiency in the
original pavement reinforcement design.
There are three main design parameters that have to be satisfied in the design
of longitudinal reinforcement for continuously reinforced concrete (CRC) pavements:
crack width at freezing temperature;
maximum steel stress, at the minimum temperature expected to occur during the design
life of the pavement; and
cumulative percentage of transverse cracks spaced below three feet.
The most widespread longitudinal steel design procedure for CRCP was developed by
Noble [5]; a computerized system, known as the CRCP computer program
incorporated the design method, which was adopted by the AASHTO Guide.
The CRCP program has the capability to predict the time history of crack
spacing, crack width, steel and concrete stress for a range of material properties,
environmental conditions, and layer thicknesses. In the steel design process, the first
step would be to run the program with the existing pavement properties, including the
existing steel percentage and past environmental conditions, to try to replicate the
existing crack spacing. In the next step, the new pavement structure, with the overlay
included, is analyzed with the program, varying the percentage of steel. The output of
the program will include the combination of the reflective cracking of the existing
pavement plus the new BCO cracking.
The transverse reinforcement can be resolved with the following expression,
which relates the percent of transverse steel, Ps, with the slab length, L (ft), the steel
working stress, fs (psi), and the friction factor, F:
LF
Ps = x100 (7.9)
2fs
7.2.5 BCO design procedures in other countries
The European standard prEN 15043 (2001), dedicated to repair systems and works (see
7.2.3) ignores the "design for sufficient strength" of BCO.
In France, there is no specific code for BCO design. The design code or recommendations for
new slabs or pavements are used as a reference. The design, therefore, assumes a complete bond.
There are also some design computer programs especially developed for pavement design that can
be used for this purpose. Among those, the program "ALIZE", developed by LCPC (Laboratoire
Central des Ponts et Chausses, France), for multilayered pavements; implemented for asphalt
based pavements, works also for cementbased pavements.
7.2.6 Summary and conclusions

In this chapter, a historical overview of the development of the most widely
used design procedures was presented, along with their general equations and input
parameters.
The current BCO design procedures have conceptual differences and
limitations; therefore, there is not an absolute solution to BCO thickness design. One
of the main differences across design methods is the failure criterion. The Corps of
Engineers procedure was developed for airports and, as such, it has the limitation that
its failure criterion, based on the appearance of structural cracking, is not applicable to
highway pavements. It assumes that the existing concrete has had no fatigue damage
and will behave just as new concrete. Hence, it does not use any evaluation of the
existing pavement. The AASHTO method has its failure criterion based upon
serviceability, whereas the Portland Cement Association method is based on a stress
limit established for plain concrete.
7.3 Design to Prevent Debonding

7.3.1 Introduction
In Europe, North America or Japan, up to now, strictly speaking, there are no methods or
codes that address the overlay design to prevent debonding. North American, European and
Japanese standards or recommendations focus on the quality of the overlaying work (subject of
chapters 2, 3, 4 and 8 of this State of the Art), expecting that it should result in durable repairs. The
"design" aspect is limited to very crude requirements or recommendations of limit values, usually
strengths, to achieve or not to exceed. The specific case of reinforced overlays, notably by fibres, is
not dealt with.
7.3.2 USA recommendations

It seems that the only available recommendations related to bonded overlays are from the
American Concrete Pavement Association (ACPA), [6]. Although published in 1990, no updates are
available yet.
About bond, it states: "Achieving bond is a key to longterm life extension. A (shear) bond
strength of 200 pound per square inch (psi) is sufficient to withstand shearing forces and ensure
bond is maintained. This number was determined in the laboratory [7]. At bond shear strength of
around 200 psi the laminated or bonded beams were nearly as strong as the monolithic ones."
Translated into MPa strength, it means that a 1.4 MPa shear strength is expected to be enough to
ensure a durable bond. Knowing that in the case of a bond between cementbased materials, the
tensile strength of the bond is around half of its shear strength, it can be assumed that according to
ACPA, for a durable bonded overlay, it is sufficient to achieve a 0.7 MPa tensile strength of the bond.
7.3.3. European requirements

Currently, in Europe, a standard is available for repair works [8]. The ten parts listed in Table
7.1 comprise the standard. Despite the fact that the standard is currently in the developing stages,
all ten parts are already available, some as finished European Standards (EN) and others as
provisional standards (prEN). In its field of relevance, it rules the national standards.
Table 7.1: Content of European Standard EN 1504: "Products and systems for the protection and
repair of concrete structures"
Reference Title
15041 General scope and definitions
15042 Surface protection systems
15043 Structural and non structural repair
15044 Structural bonding
15045 Concrete injection
15046 Grouting to anchor reinforcement or to fill external voids
15047 Reinforcement corrosion prevention
15048 Quality control and evaluation of conformity
15049 General principles for the use of products and systems
150410 Application of products and systems and quality control of the
work
Several types of overlaying work are considered, which are:
Type 3.1: concrete restoration by applying mortar by hand;

Type 3.2: concrete restoration by recasting with concrete;
Type 3.3: concrete restoration by spraying with concrete or mortar;
Type 4.4: structural strengthening by adding mortar or concrete;
Type 7.1: restoring passivity by increasing cover to reinforcement with additional cementitious
mortar or concrete; and
Type 7.2: restoring passivity by replacing contaminated or carbonated concrete.
In the text of the Standard, all the requirements are collected in two tables distinguishing the cases
of "structural" and "nonstructural" overlaying systems and can be summarized in Table 7.2.
Table 7.2: European Standard prEN 45043: "design" requirements for overlaying works
Property Test method Requirement Comments

Structural Non structural
Compressive EN 12190 30 MPa 20 MPa
strength
Elastic modulus prEN13412 20 GPa

(only for type 4
overlaying)
Bond tensile EN 1542 2.0 MPa 1.5 MPa Mean value.

strength (pulloff test)
No single value less than 75 %
of minimum requirement
Bond tensile 8401 2.0 MPa 1.5 MPa After 50 cycles, with maximum
strength after permissible average crack
freezethaw test width 0.05 mm with no crack
(if exposure 0.1 mm and no delamination
conditions require it)
thundershower width 0.05 mm with no crack
test 0.1 mm and no delamination
(if exposure
strength after dry permissible average crack
thermal cycling test width 0.05 mm with no crack
(if exposure 0.1 mm and no delamination
Bond tensile 8164 2.0 MPa 1.5 MPa After test, with maximum
restrained width 0.05 mm with no crack
shrinkage/swelling 0.1 mm and no delamination
test
(only for overlaying
types 3.1, 3.2 and 4.4)
It results that:
for "structural" applications, achieve a tensile bond strength 2.0 MPa;

for "non structural" applications, achieve a tensile bond strength 1.5 MPa;
in all cases the crack width must remain limited; on the test specimens, no crack 1 mm is
accepted.
Note that South Africa follows the British codes, then the European code.
7.3.4 Japanese requirements
The Japan Highway Research Foundation has published BCO guidelines in its Design and
Execution Manual for Bonded Concrete Overlays [9] for repairing bridge decks, which is the only
commonly accepted design manual for concrete overlays in Japan. It is comprised of six chapters as
shown in Table 7.3.
Table 7.3: Contents of the design and execution manual from Japan
Reference Title
Chapter 1 General Principles
Chapter 2 Planning
Chapter 3 Materials
Chapter 4 Design
Chapter 5 Execution
Chapter 6 Quality Control
The design principles are presented in Chapter 4. Mixture proportions for the overlay concrete
specify the use of ultra-rapid-hardening cement to get the compressive strength of 24 MPa at the
established age (usually 3 hours). This condition gives a higher value for the strength than about 40
MPa when the road is opened for the traffic service.
The minimum thickness required for the overlay is 50 mm, which was determined by taking into
account the maximum aggregate size, shrinkage, and the quality of the workmanship. The standard
thickness of the treated surface of the substrate concrete is specified as 10 mm. Shot blast methods are
recommended for the surface treatment since even shot blasted surfaces can achieve a bond strength of
1 MPa. Although the high performance of the water jet method has been acknowledged in recent
studies [10], shot blast methods are still most commonly used in Japan because of their cost. Shear
reinforcement is not necessary for overlay concrete. This is based on results of recent research
projects related to the interface bond. According to the experimental results of the research projects, a
tensile bond strength of 1 MPa between overlay and substrate is sufficient for up to three times higher
loads than the design value.
The minimum section area to be overlaid should cover at least the area between expansion joints in
the longitudinal direction and both edges for the transverse direction.
7.3.5 Swedish practice
The Swedish National Road Administration [11], which possesses a long experience on
concrete bridge repair, has used and is still using in 2004 the following requirements for BCOs.
The required tensile bond strength is fv = 1.0 MPa. This requirement is satisfied if:
m f v + 1.4 s (7.10)
x 0.8 f v (7.11)
where, m and s are the average and the standard deviation (s 0.36 MPa) of the measuring
values, respectively, and x is a single measuring value.
7.3.6 Expected design shear strength

In this section, the following notation is used:
Expected design shear bond strength = fbd,

with f = strength, b = bond, = shear, d = design value.
Expected design tensile bond strength = fbtd; knowing that the bond tensile strength is about
half the bond shear strength, it can be assumed that fbtd fbd .
Design concrete compressive strength = fccd.
Design concrete tensile strength = fctd.
Overlapping with the above strength requirements, different codes or recommendations
provide a value that could be considered as a "maximum design shear stress value". It seems that, for
carefully constructed overlays, it is the maximum acceptable shear stress calculated with the
assumption of a monolithic structure (notably ignoring the builtin stresses deriving from the
composite nature of the actual overlaid structure). Because a wide range of behaviour of the actual
overlaid structures is ignored, the proposed maximum acceptable shear stresses are very
conservative.
According to CEBFIP MC 90 (1993), the expected design shear strength at the interface is
proportional to the compressive strength fccd. If the surface is rough or indented, the coefficient of
proportionality is equal to 0.06. Hence,
fccd = 25 MPa fbd = 1.5 MPa fbtd 0.8 MPa,
fccd = 45 MPa fbd = 2.7 MPa fbtd 1.4 MPa.

The draft of Eurocode 2 (2001) states that the design shear resistance at the interface is
proportional to the design tensile strength fctd of the weakest concrete. The coefficient of
proportionality is 0.45 for rough surfaces and 0.5 for indented surfaces. Hence, at a rough interface,
fccd = 25 MPa fctd = 2.1 MPa fbd = 0.95 MPa fbtd 0.5 MPa,
fccd = 45 MPa fctd = 3.3 MPa fbd = 1.5 MPa fbtd 0.8 MPa.
The American code ACI 31899 (1999) deals with composite structures in Chapter 17. For
clean surfaces, free of laitance, and intentionally roughened, the maximum design shear resistance is
limited to
fbd = 0.55 MPa fbtd 0.3 MPa.
The Swedish handbook for concrete structures (1994), for waterjetted properly cleaned
interface proposes
fbd = 0.4 MPa fbtd 0.24 MPa.
None of these values is realistic. Indeed, average measured bond shear strength are in the
vicinity of 3 MPa (see Chapter 4). The codes of all countries derive the design strength from the
measured values through a relationship of the kind of the one presented in 7.2.5 above. With the
common scatter of bond strength measurements, it results in fbd 0.8 average measured strength.
Hence, the realistic expectable value of fbd 2.4 MPa fbtd 1.2 MPa.
Only the value proposed by CEBFIP is in the realistic range. All the other proposed values are
too low.
In the case of the ACI 31899 and Swedish proposals, the proposed values are almost one
order of magnitude too low. In these cases, it can be supposed that the recommended values
correspond, for carefully constructed overlays, to the maximum acceptable shear stress calculated
with the assumption of a monolithic structure and ignoring the builtin stresses deriving from the
composite nature of the actual overlaid structure. Then, the very conservative values could
compensate for ignoring a wide range of behaviour of the actual overlaid structure.
7.4 Conclusion
As already underlined, the existing "design" recommendations or requirements are very

limited. Moreover, the few available minimum strength values for the bond are not in agreement.
The proposed or required tensile strength ranges from about 0.75 MPa in the USA to 1.5 MPa or 2.0
MPa in Europe. Moreover, major parameters such as, on the one hand the relative thickness of the
overlay and the substrate, on the second hand a reinforcement of the overlay, notably by fibres, are
not accounted for.
This limited knowledge highlights the need for the development of relevant design methods.
References
[1] Barenberg, E. J., Rehabilitation of Concrete Pavement by Using Portland Cement Concrete
Overlays, Transportation Research Record 814, (Washington, DC, USA, 1981).
[2] AASHTO, Guide for Design of Pavement Structures, American Association of State Highway and
Transportation Officials, (USA, 1993).
[3] Army, Army Airfield and Heliport Rigid and Overlay Pavement Design, Department of the Army
Technical Manual, TM 58233, (USA, 1968).
[4] Tayabji, S. D., and Okamoto, P. A., Thickness Design of Concrete Resurfacing, Third
International Conference on Concrete Pavement Design and Rehabilitation, Purdue University,
(West Lafayette, Indiana, USA, 1985).
[5] Noble, C. S., McCullough, B. F., and Ma, J. C. M., Nomographs for the Design of CRCP Steel
Reinforcement, Research Report 17716, Center for Highway Research, The University of
Texas at Austin, (USA, August 1979).
[6] ACPA Technical bulletin TB 007P, American Concrete Pavement Association, (Skokie, Illinois,
USA 1990).
[7] Felt E. J., Resurfacing and patching concrete pavement with bonded concrete, Highway
Research Board, Proceedings Volume 35 (Washington, DC, USA, 1956).
[8] prEN 15043 Products and systems for the protection and repair of concrete structures
Structural and non structural repair, European Committee of Standardization (CEN), (Brussels,
Belgium, 2001).
[9] Design and Execution Manual for Bonded Concrete Overlays, Japan Highway Research
Foundation, (Japan, 1995).
[10] Takuwa, I., Shitou, K., Kamihigashi, Y., Nakashima, H. and Yoshida, A., The Application of
Waterjet Technology to Surface Preparation of Concrete Structure, J. of Jet Flow
Engineering,17(1), pp. 2940 (2000).
[11] Swedish National Road Administration, General Technical Regulations for Bridges,
Publication No. 2004:56 (in Swedish), (Borlnge, Sweden, 2004).
8. PRACTICE AND QUALITY ASSURANCE
M. Vaysburd (1), B. Bissonnette (2), R. Morin (3),
(1) Vaycon Consulting, Baltimore (MD), U.S.A

(2) Centre de Recherche sur les Infrastructures en Bton (CRIB), Universit Laval, Qubec (QC),
Canada
(3) Laboratory of the City of Montreal, Montreal (QC), Canada
ABSTRACT
This chapter discusses construction practice and quality assurance including the importance of
substrate surface preparation. The steps in applying a bonded concrete overlay including concrete
removal, substrate surface preparation, and application methods for applying the overlay are re
described. Methods of quality assurance and construction inspection are discussed.
8.1. Scope And Definitions

This chapter reviews observations of good practices and quality control/quality assurance in the
implementation of concrete repair projects involving cementbased overlays.
8.2. Importance of Substrate Surface Preparation.
8.2.1. General
The importance of surface preparation cannot be overstated. Proper attention to surface
preparation is essential for a durable repair. Regardless of the cost, complexity and quality of the
repair material and application method selected, the care with which deteriorated concrete is
removed and concrete and reinforcement surface are prepared will often determine whether a
repair project will be successful.
The process of preparation for repair is the process by which sound, clean, and suitably roughened
surfaces are produced on concrete substrates. This process includes the removal of unsound and, if
necessary, sound concrete and bond inhibiting foreign materials from the concrete and
reinforcement surfaces, opening the concrete pore structure, reinforcement damage verification and
repair, if necessary.
8.2.2. Safety
Before starting removal of existing concrete, the effect of the removal on the structural
integrity has to be reviewed. In cases where removal of deteriorated concrete and/or severely
corroded reinforcing steel can affect the load carrying capacity of the structure or its elements,
a temporary shoring system should be provided to relive the loads from the structure or its
member being repaired. Caution needs to be exercised in order that the safety of the structure
is not jeopardized by repair activities.
The areas where concrete to be removed have to be examined to determine if there are
electrical conduits, utility lines, or other embedments in the concrete which may be damaged
during removal.
All necessary precautions shall be taken to ensure that dust or falling debris do not
constitute a hazard to personnel, equipment, the structure, its occupants and the general
public. Effective means of clearing dust and debris away from the working area be
continuously implemented.
8.3. Concrete Removal
8.3.1. General
Concrete removal on flat horizontal surfaces such as bridge decks, marine piers, etc. differs
substantially from concrete removal on other components. The flat horizontal surfaces provide easy
access and permit use of highproduction mechanized methods. Cleanup and containment of debris
are also easier on the horizontal surface.
Other structural components have smaller surfaces that frequently are vertical or overhead and
difficult to reach. Difficulties with access and debris containment preclude the use of heavy
equipment and make concrete removal less productive and more expensive. The need for
scaffolding to provide access for workers and equipment adds to the difficulty and safety hazards
involved with the work.
8.3.2. Removal Methods

Concrete removal is addressed in detail in Chapter 4 and in ACI 546R [1].
Removal tasks have to be performed in a manner that ensures that the remaining concrete
and reinforcing steel retains its structural integrity. Equipment used to perform the work must
not overload the structure, and care must be taken to remove so much concrete that the
member being repaired or rehabilitated is critically weakened. Any impact forces used to
remove damaged concrete should be applied in a manner that minimizes cracking in the
residual concrete and minimizes damage to the bond between the remaining concrete and
steel. Methods to remove rust and chlorides from the steel should also minimize damage and
loss to the remaining steel.
8.3.3. Advantages and Limitations of Concrete Removal Methods

The study of comparative abilities showed that each concrete removal method had very specific
strengths and weaknesses with regard to work characteristics, quality, and production, as follows [2].
8.3.3.1. Basic methods
1) Pneumatic breakers. They are the most flexible and most labor-intensive. They can be
used for all sizes and shapes of area, to all depths, and on all bridge structural elements;
however, they cause surface microcracking of the substrate, and production rates are
low.
2) Milling machines. They are the most inflexible. They can only be used to remove large
areas of surface and/or cover concrete on decks and are not labor-intensive but capital-
intensive. If properly used, they cause little surface microcracking of the substrate and
have high production rates.
3) Hydrodemolition. This technology lies between pneumatic breakers and milling
machines in terms of flexibility. Surface, cover, matrix, and core concrete can be
removed; it creates no microcrack and is harmless to the steel reinforcements (bars or
welded mesh) . Economies are only realized if work is done on large horizontal areas
such as decks. Hydrodemolition also has the advantage of selective removal of poor
concrete that might be left in place by the breaker method.
8.3.3.2. Combining methods

Much can be done by combining the strengths and weaknesses of the methods as determined below.
1) Milling and Breakers

Combining these two technologies on surfaces with relatively minor deterioration in
cover and matrix concrete results in the following sequence of operations:
a) Contaminated and deteriorated concrete is milled to a level conservatively above the
reinforcing steel.
b) Areas inaccessible to the milling machine are removed using pneumatic breakers.
c) Areas of deterioration remaining below the reinforcing steel are determined by
sounding and removed to the required depth using pneumatic breakers.
d) All exposed steel is sandblasted to remove loose material and rust immediately before
repairing damaged areas and overlaying the deck.
b) Hydrodemolition and Breakers
Combining these two technologies on bridges with relatively large areas of
deterioration in cover and matrix concrete results in the following sequence of
operations:
a) The level of contamination and deterioration in the deck is determined using half-cell
potential measurement, chloride sampling, and/or sounding. Large areas that require
removal are delineated, or a decision is taken to hydrodemolish the whole deck and
rely on the selective removal capability of the technique to identify areas of above or
below average deterioration.
b) Contaminated and deteriorated cover and matrix concrete is removed using
hydrodemolition equipment calibrated to achieve the desired results.
c) Areas inaccessible to the hydrodemolition equipment and areas of particularly hard
concrete are removed using pneumatic breakers.
d) All exposed steel is sandblasted to remove loose material and rust immediately prior
to repairing damaged areas and overlaying the deck.
c) Milling, Hydrodemolition and Breakers
The efficiency of milling large areas of cover concrete can be used to advantage by
using a milling machine to remove all the material above the reinforcing steel. This
greatly increases the productivity of the hydrodemolition that follows. The reduction in
volume of concrete to be removed by hydrodemolition produces a corresponding
reduction in the quantity of wastewater produced, and this greatly improves the
environmental and safety impacts. Breakers are used to remove the concrete from areas
inaccessible to the milling and hydrodemolition machines.
8.4. Substrate Surface Preparation
8.4.1. General
Overlay surfaces of piers, parking structures, bridges, ramps and loading docks may be exposed to
abrasion, rapid temperature changes, salts, moisture, oil, heavy wheel loads, snowplow blades,
shrinkage induced shear stresses, live load shear stresses caused by vehicles turning, and impact
stresses caused by roughness of the riding surface. The ability of the overlay to survive these
conditions is highly dependent on the sound bond of the overlay to the substrate and, therefore, on
the adequate surface preparation.
This section addresses the final concrete and reinforcement surface preparation steps taken after the
removal of deteriorated concrete necessary to receive the repair materials. The appropriate
preparation of the concrete surface depends on the preceding operations and on the type of repair
being undertaken.
Concrete removal methods may leave the surface to receive the repair material too smooth,
to rough, too irregular, and without open pores. In these cases, procedures specifically
intended for final surface cleaning are necessary. Also, microcracking of the concrete surface
is common when impact tools are used which may weaken the bond between the existing
concrete and the repair. In this case, a less aggressive method of surface preparation such as
abrasive or waterblasting is necessary.
8.4.2. Pre-overlay repairs

Surface preparation for overlays shall include the preoverlay repair of defects such as deterioration,
corrosion of reinforcement, honeycombed areas, delaminations, small and large holes, sharp
protrusions, and cracks. Preoverlay repair shall consist of all the repair procedures required to bring
the substrate to a condition suitable for overlay.
Damaged sections shall be removed with tools that will not further damage adjacent
areas, including the reinforcing steel.
Careful attention shall be given to the repair of cracks in the concrete substrate.
The cause of movement in the substrate shall be evaluated and addressed if possible, or
the movement shall be accommodated by the overlay.
The repair materials shall be cement-based and compatible with overlay concrete and
substrate concrete being repaired.
Cure the repairs to be overlaid a minimum of 48 hours or until specified strength has
been achieved before beginning final surface preparation procedures. In certain cases,
the substrate repairs may be cast monolithically with the overlay material.
Surface preparation for overlaying shall include the entire surface to be overlaid.
Pre-overlay repairs shall be performed as recommended by ACI 546R [1].

8.4.3. Surface
8.4.3.1. General
Surfaces of existing concrete and exposed reinforcing steel which will be in contact with repair
material must be thoroughly cleaned of all bondinhibiting materials. The concrete surface must be
free of microcracking bruising, dust, and must have an open pore system. All exposed reinforcing
steel is to be clean of all grease, dirt, cement mortar, injurious rust, and traces of chlorides from pits
in the bars. Injurious rust includes all scale, loose rust deposits, or all rust not firmly bonded to the
steel.
When impact tools are used for concrete removal fracturing and microcracking (bruising) of the
concrete surface occurs, which weakens the bond between the existing concrete and the repair.
Therefore, a less aggressive method of concrete cleaning such as abrasive or waterblasting shall be
employed.
8.4.3.2. Concrete Surface Cleaning

a) General
First stage cleaning operations shall be commenced in a repair area after all necessary
concrete removal has been completed. Concrete surface usually has laitance, partially
loosened chips of concrete and bruised concrete layer, removed by blasting. The
concrete removal process leaves the resulting surface with varying degrees of
microcracks and fractures, commonly referred to as bruising. Bruising creates a zone
of weakness that will affect the bond of the repair material to the substrate. Removal
objects the concrete substrate to a wide range of dynamic loads and the resulting
bruising will depend on the method used and the quality of the concrete. The depth of
the bruised layer varies, but is typically on the order of 3.0 mm.
b) Available techniques
Cleaning techniques consist of removing thin layers of surface concrete using
abrasive equipment such as sandblasters, shotblasters, or high-pressure water blasters
including:
Sand Blasting Sand blasting is the most commonly used method of cleaning concrete
and reinforcing steel.
Shotblasting Shotblasting equipment has the capability to remove
finite amounts of sound or unsound concrete. The shot erodes the concrete from the
surface. The shot rebounds with the pulverized concrete and is vacuumed into the
body of the shotblasting machine.
High-pressure water blasting (with or without abrasive) High-
pressure water blasting with abrasives is a cleaning system using a stream of water at
high pressure with an abrasive such as aluminum oxide or garnet introduced into the
stream. This equipment has the capability of removing dirt or other foreign particles
as well as concrete laitance thereby exposing the fine aggregate.
8.4.3.3. Reinforcement cleaning, Inspection and Repair

a) General
The initial cleaning of exposed reinforcement is usually achieved during the concrete
surface cleaning procedures using blasting techniques.
After the initial cleaning, reinforcing steel should be carefully inspected to determine
whether the steel shall be simply cleaned or repaired.
b) Corrosion Protection
When epoxy-coated steel reinforcement is exposed in the repair area it should be
recoated with an epoxy coating. Special care must be exercised in recoating operations
to achieve defect-free full surface coverage. Uncoated spots may result in severe
corrosion in repair areas.
When uncoated reinforcing steel is exposed in the repair area, application of a protective
coating should not be done, because it may cause corrosion in areas immediately adjacent to a
repair.
A commonly observed phenomenon in concrete repair is increased corrosion activity in

existing concrete areas immediately adjacent to a repair. This effect has been referred to as
the anodic ring effect, the incipient anode effect, and the halo effect. The repair of corrosion
affected concrete usually addresses the areas where the corrosion activity and related damage
is worst sometimes referred to as the hot spot. In many cases, it is likely that the
conditions in surrounding areas are such that corrosion can occur; however, corrosion activity
in these areas has been reduced or dormant due to the active corrosion occurring at the hot
spot. This reduction is a form of cathodic protection provided by the hot spot. When the
repair is completed, the hot spot has been removed, and the adjacent areas are no longer
cathodically protected. In such case the repair area can contribute to corrosion in the
adjacent, nonrepaired areas. The result is a new hot spot which may require additional
repair in three to five years.
8.4.4. Moisture conditioning of the substrate prior to overlay/repair

8.4.4.1. General
Factors that influence the formation of a bond between repair material and prepared substrate
include: the properties of the substrate concrete and its surface, the properties of the repair
material, absorption, adhesion, and environmental conditions. Several of these factors are critically
dependent on moisture condition of the substrate prior to application of repair materials.
The moisture condition of the substrate will determine the rate of movement of water from the
repair mortar to substrate concrete due to the moisture imbalance between the two layers. Both the
surface moisture condition and the moisture distribution inside the substrate are important. During
the process of water movement two things will occur: penetration of water from the repair mortar in
the capillaries of substrate concrete, and hydration of cement paste in the repair.
The optimum moisture condition will vary from substrate to substrate in otherwise equal conditions
because the performance of the bond depends on the way the substrate will affect the direction and
rate of water movement between phases of the composite repair system.
8.4.4.2. Optimum water condition of the substrate

The optimum water condition of a concrete substrate for a particular cementbased repair material
can be determined by preliminary testing using different moisture surface conditioning:
Saturated Surface Dry (SSD);

Saturated Surface Wet (SSW);
Unsaturated Surface Dry (USD); and
Unsaturated Surface Wet (USW).
In cases when such testing cannot be performed, SSD moisture conditioning should be applied.
Under this condition the substrate looks damp but contains no free water on the surface. The surface
absorbed all the moisture possible but does not contribute water to the repair material mixture.
8.4.5. Maintenance of the prepared substrate

After the substrate has been prepared, it should be maintained in a clean condition and
protected from damage until the repair/overlay material is placed.
In hot climates shade should be provided if practically possible to keep the substrate
cool, thereby reducing rapid hydration or hardening of repair material. In wintertime,
necessary steps should be taken to provide sufficient insulation and/or heat to prevent
the repair area from being covered with snow, ice, or snowmelt water.
Prepared areas should be protected from repair activities in adjacent areas. Mud, debris,
cement, dust, etc., when deposited on a prepared surface, may act as a bond breaker if
not cleaned up before the repair material is placed.
8.4.6. Quality control of surface preparation

8.4.6.1. Test repair
To provide assurance that the surface preparation procedures, materials, installation procedures,
and curing will provide the specified repair, the test repair should (it is my understanding that since
shall is a specification word, we would use should instead) be installed at the same thickness and
with the same material and techniques, equipment, personnel, timing, sequence of operations, and
curing period that will be used for the repair project.
The surface of the designated test repair area shallx be cleaned, and the repair material shall be
applied, cured, and tested. A test result shallx be the average of three tests on a test area of not less
than 6 m.
8.4.6.2. Required strength at the interface

The bond strength shall be tested using the direct tension test. In US practice, it is recommended
that bond strength shall be not less than 0.7 MPa. For other countries, the above minimum required
tensile strength of the bond and of the substrate may be different (Section 7.2.5). If the failure occurs
in the repair material or existing substrate, the bond strength is known to exceed the cohesive
strength of the system.
The direct tension test may also be used to determine the soundness of the existing concrete surface
before the application of the repair. Tested tensile adhesion strength shall be greater than or equal
to 0.7 MPa unless otherwise specified. If the rupture strength is less than 0.7 MPa, the data shall be
examined, and the surface preparation procedures shall be modified.
If the strength of the existing substrate concrete is inadequate to achieve the specified repair
requirements, the data shall be examined by the engineer and the specified requirements may
require reevaluation.
8.5. Application
8.5.1. General
Experience has repeatedly demonstrated that no step in a repair application can be omitted or
carelessly performed without detriment to the durability. Inadequate workmanship, procedures, or
materials result in inferior repairs, which will prematurely fail.
8.5.2. Workmanship
It is important that both foremen and workmen be fully instructed concerning procedural details of
repairing concrete and reasons for them. Constant vigilance must be exercised to assure
maintenance of the necessary standards of workmanship. They should also be apprised of the more
critical aspects of repairing concrete
8.5.3. Basic requirement

The materials used for repairs shall be as recommended by ACI 546.3R [3].
The basic requirement in repair material mixture handling is both quality and uniformity of the
material have to be preserved.
Adequate placement capacity should be provided so that material can be kept workable and free of
cold joints during the placement. Adequate and consistent supply of repair material during the
placement.
The placement process and equipment should be arranged to deliver the material to its final position
without segregation. The equipment should be adequately and properly arranged so that placing can
proceed without undue delays and manpower should be sufficient to ensure the proper and timely
placing, consolidating, finishing and curing of the repair.
8.5.4. Repair placement

The repair placement should not commence when there is a chance of freezing temperature
occurring, unless adequate facilities for coldweather protection have been provided.
Temperature of the repair material mixture should be between 10o and 29oC, and it should not vary
more than approximately 5oC from the temperature of the substrate against which it is placed.
8.5.5. Consolidation
Cementitious repair materials should be consolidated by systematic vibration during placement.
Failure to do so, and to do it properly, may result in poor durability, and rapid deterioration. Voids
between the repair and substrate can cause total debonding. Voids between embedded reinforcing
steel and repair material can cause early corrosion.
The vibration of the repair should be performed by interior vibrators or vibrating screeds or both in
combination. However, for relatively small repair areas, application of vibrating screeds is not
recommended.
Vibrators should not be used to move repair material laterally. They should be inserted and
withdrawn at 15 to 30 degrees from the vertical, so that they quickly penetrate the layer and are
withdrawn slowly to remove entrapped air. Vibrators should be placed at close intervals using a
systematic pattern to ensure that all repair material is adequately consolidated. The mixture is
adequately consolidated when it stops settling, air bubbles no longer emerge, and a smooth layer of
mortar appears at the surface.
There is little likelihood of overvibration when the slump of the repair mixture is as low as is
practicable for placement using vibration. When overvibration occurs, the surface not only appears
very wet, but it actually consists of a layer of mortar containing little coarse aggregate. When
overvibration is indicated, the slump, and not the amount of vibration, should be reduced. Efforts to
avoid overvibration often result in inadequate vibration. Experience indicates that objectionable
results are much more likely to be obtained from undervibration than from overvibration.
Considerably more vibration is sometimes required to satisfactorily reduce the amount of entrapped
air and the number of surface bubbles than is necessary to eliminate voids.
Revibration is beneficial rather than detrimental, provided the material is again brought to a plastic
condition. Revibration could well be more widely practiced to eliminate settlement cracks and the
internal effects of bleeding and also as an aid in making tight concrete repairs.
8.5.6. Surface finishing

a) Surface finishing (beyond bull floating) should not be initiated before initial set nor
before bleed water has disappeared from the repair material surface. Any finishing of
the surface (floating or troweling), while water is rising to the surface will force the
bleed water into the repair.
Before initial set, the material is not stiff enough to hold a texture nor stiff enough to
support the weight of a finisher or finishing machine. Furthermore, bleeding also
controls the timing of finishing operations. Bleed water rises to the surface of freshly
cast material because of the settling of the denser solid particles in response to gravity
and accumulates on the surface until it evaporates or is removed by the contractor.
Bleed water is evident by the sheen on the surface of freshly cast repair material.
Finishing the concrete surface before settlement and bleeding has ended can trap the
residual bleed water below a densified surface layer, resulting in a weakened zone just
below the surface. Finishing before the bleed water fully disappears remixes
accumulated bleed water back into the concrete surface, thus increasing w/cm and
decreasing strength and durability in this critical near-surface region. Remixing bleed
water can also decrease air content at the surface, further reducing durability. Proper
finishing should not start until bleeding has ceased and the bleed water has disappeared
or has been removed. In most cases, the concrete surface is drying while it is being
finished.
b) When required, the repair should match the existing concrete surface in texture and
color.
c) To obtain a durable surface of the repair, proper procedures should be carefully
followed. Following consolidation, the operations of screeding, floating, and troweling
should be performed in such a manner that the repair material will be worked and
manipulated as little as possible to produce the desired result.
Overmanipulation of the concrete brings excessive fines and water to the surface, which
lessens the quality of the finished surface, causing checking, crazing, and dusting. For
the same reason, each step in the finishing operation, from the first floating to the final
floating or troweling, should be delayed as long as possible while still working toward
the desired surface smoothness. Free water is not as likely to appear and accumulate
between finishing operations if proper mixture proportions and consistency are used. If
free water does accumulate, however, it should be removed by blotting with mats,
draining, or pulling off with a loop of hose so that the surface loses its water sheen
before the next finishing operation is performed. Under no circumstances should any
finishing tool be used in an area before accumulated water has been removed, nor
should neat cement or mixtures of sand and cement be worked into the surface to dry
such areas.
8.5.7. Curing
Good curing is vital to producing a satisfactory repair. Good curing not only increases durability and
wear resistance, but reduces early drying shrinkage cracking. Curing must be started early. Proper
curing is even more important for overlays than for ordinary concrete because of the potential for
rapid early drying of the relatively thin repair since they have large surface areas in relation to their
volumes and moisture can be lost quickly.
8.6. Quality Assurance/Construction Inspection

The following should be considered the minimum inspection requirements to ensure a quality
overlay. More than one inspector should be present when the overlay is being placed.
8.6.1. Scarification and Removal of Unsound Concrete

Following the completion of the milling, inspection for delineation and repair of unsound areas
should be conducted as described in this chapter.
8.6.2. Substrate Preparation

In assessing substrate preparation, the inspector should determine the following:
(a) Ensure that the surface is free from all latencies, organic residues, and debris.
(b) Ensure that the substrate is in a saturated surface dry condition and free from standing
water.
(c) Ensure that the screed rails are properly adjusted to provide the minimum overlay thickness
over the entire area to be overlaid, and that the finishing equipment is clean and functioning
properly.
8.6.3. Placement and Consolidation

The inspector should account for the following before and during each placement as applicable:
a) Before placement, a trial batch should be prepared to ensure that the specified slump and air
content can be obtained with the job materials and mixture.
b) Ensure that sufficient amount of material is present to allow continuous placement. Slump
tests, total aircontent measurements, and a minimum of six compressivestrength specimens
should be prepared from each batch. Slump tests should be performed five minutes after
mixing.
c) Monitor the air temperature, material temperature, relative humidity, and wind velocity.
Overlays shall not be placed under the following conditions:
Air temperature is below 10 C;

Evaporation rate exceeds 0.49 kg/m/hr as determined by ACI 305R
Overlays shall be placed at night when the daytime air temperature is above 29 C.
d) Check the levelness requirements of the finished overlay.
e) Curing
The site inspector should monitor the overlay during the curing period to ensure that it is
maintained at a temperature in excess of 7 C and uniformly saturated; membraneforming
curing compounds can be used with double application.
f) Job Acceptance
Following the curing period, the entire overlay should be sounded with a drag chain and hammer
to ensure bond. The surface of the overlay should be inspected for cracking. All unsound areas
should be repaired by the contractor. Finally, job is accepted or declined based on the results
of tensile pulloff tests
The key requirement of a successful repair is an adequate bond between the repair material
and existing concrete substrate, which remains intact throughout its service life. At the present
time, practical answers to the problems of bond may depend only on a short-term bond testing
rather than long-term performance. An initially achieved adequate bond is only an indication
of conformance with the specified parameters. There is no well-defined relationship between
initial bond strength and the longevity of a repair. Longevity is influenced by many factors
including substrate surface preparation and texture, shrinkage of the repair material, and
service conditions.
References
[1] American Concrete Institute (ACI) (2004), ACI 546R-04, Concrete Repair Guide.
[2] Strategic Highway Research Program Report SHRP-S-360, Washington, DC, 1993, 268
pp.
[3] American Concrete Institute (ACI) (2006), ACI 546.3R-06, Guide for the Selection of
Materials for the Repair of Concrete.
9.0 MAINTENANCE AND REPAIR OF OVERLAYS
D.W. Fowler
The University of Texas at Austin (TX), U.S.A.
ABSTRACT
In the event that Bonded Concrete Overlays (BCO) are subjected to distress mechanisms, they may
require repair. The types and causes of distress are discussed. Methods for evaluating the extent of
distress are discussed, and methods for repairing cracks, spalls and delaminations are described.
9.1. TYPES AND CAUSES OF DISTRESS
Bonded concrete overlays are designed to strengthen and/or rehabilitate existing slabs and
pavements and are often a part of the repair process. But overlays sometimes exhibit distress and
must be repaired. The most common types of distress are:
1) cracking;
2) delamination; and
3) spalling.
Cracking is caused by one or more of:
1) differential contraction due to thermal stresses or shrinkage;

2) inadequate curing;
3) reflective cracking from substrate;
4) chemical reactions;
5) freezing and thawing;
6) stresses due to external loading; and
7) differential movement of soils or supports.
Cracking often leads to delamination since the stresses due to differential expansion and
contraction are the highest at boundaries, e.g. cracks, joints or edges. If the normal or axial stresses
at the interface exceed the bond strength between the substrate and the overlay, the overlay will
begin delaminating or debonding at the cracks, joints and/or edges. Delamination is also due to
inadequate surface preparation, surface contamination prior to installation of overlays, or premature
curing of some bonding agents, e.g. cement slurries or latex slurries.
Spalling can occur at joints due to improper construction or inadequate or nonmaintained

joint filler. In some cases spalling can occur in the interior areas of overlays due to inadequate
curing, expansive aggregate or corrosion of reinforcing.
9.2 EVALUATION OF DAMAGE
It is important that a condition survey be conducted to determine the extent and causes of damage.
The condition survey should:
1) map the location of cracks;

2) mark the location and extent of delamination:
a) sounding is the most commonly used technique using a chain drag or a dropped steel rod in
a regular pattern on the overlay surface area.
b) infrared thermography;
c) ground penetrating radar;
d) acoustic methods: impactecho and ultrasonic pulse velocity;
e) coring/pulloff tests for determining bond strength; and
3) map the location of spalls and any other distresses.
9.3 REPAIR METHODS
9.3.1 Cracks
If delamination has not occurred at the cracks, the cracks may be repaired by one of several
methods:
1) For nonmoving cracks, epoxy injection or a topically applied high molecular methacrylate
monomer or a low viscosity epoxy resin especially formulated for crack repair can be used.
Epoxy injection is quite expensive and is very time consuming, but provides the best structural
repair and is usually the best solution if some moisture is present (the cracks should be as dry
as possible for the best results.) The topically applied materials are much less costly to apply
and are best for extensive cracking in the overlay.
2) For moving cracks which will probably be encountered in bonded concrete overlays only at
joints in the substrate where a proper joint was not formed or cut in the overlay or other
special cases, an elastomeric joint filler which can accommodate the anticipated joint
movement should be used.
9.3.2 Delaminated concrete
Repairing delaminated overlays requires either rebonding the overlay to the substrate or removing
and replacing the overlay in the debonded zone. Rebonding the overlay may be possible using one of
several methods:
1) Epoxy injection by drilling access holes through the overlay at regular intervals and injecting
epoxy to provide a bonded interface. This repair is usually possible only if the repair is made
before the interface becomes contaminated or wet and adequate bonding is compromised.
Drying the interface using high pressure air may be helpful if moisture is present.
2) Mechanical bonding using connectors such as steel rods placed and epoxied into drilled holes
extending well into the substrate is a possible solution.
Removing and replacing may be the least expensive and most effective repair for many cases.
If extensive cracking and delamination has occurred, complete removal and replacement may be
required. Prior to repair, however, it is essential to determine the cause of the failure by a thorough
forensic analysis to insure that the failure is not repeated. If the complete removal is determined to
not be required, the delaminated overlay should be removed beyond the point of delamination by
careful saw cutting and careful removal by lightweight pneumatic hammers with care being taken
not to damage the remaining sound overlay or substrate concrete. If reinforcing steel or fabric was
used in the overlay, it will probably be necessary allow the reinforcing extend into the new concrete
overlay a distance to develop the strength of the steel.
The surface preparation, steel placement, concrete mixing and placing, finishing/texturing
and curing should follow the procedures for new overlays.
9.3.3 Spalls
The cause of the spalling should be determined prior to repair. The damaged concrete
should be carefully removed by chipping hammer or other suitable methods. If corrosion of steel has
occurred the concrete should be removed to a depth of at least 20 mm below the bar and the
corrosion scale should be completely removed. If expansive aggregate is determined to be the cause
the aggregate should be carefully chipped out.
For small repairs, latexmodified mortar or polymer mortar may be the best materials. For
larger repairs, latexmodified concrete, polymer concrete or portland cement concrete made with 6
to 10mm coarse aggregate is preferred.
9.4 Conclusions
Bonded concrete overlays occasionally suffer distress in the form of cracks, spalls and
delamination. The extent of distress should first be determined. If the distress requires repair,
method of repair that are typically used of other types of concrete distress are suitable for BCOs.
10. CONCLUSION
B. Bissonnette (1), L. Courard (2), J.-L. Granju (3)
(1) Centre de Recherche sur les Infrastructures en Bton (CRIB), Universit Laval, Qubec
(QC), Canada
France
This State of the Art addresses the relevant issues and challenges in view of achieving
durable concrete overlays on a systematical basis.
Various aspects of the current practice have been revisited and validated. Among them, those
relating to condition assessment, surface preparation and joint location owe to be reminded.
y In order to select the appropriate repair technique and material, reliable condition assessment
is essential. Significant developments in nondestructive techniques recently have helped
improving the assessment of the structure to be repaired and establishing a comprehensive
diagnosis. Nevertheless, the "old methods" starting with visual inspection remain
necessary.
y Preparation and cleanliness of the surface to be overlaid is again emphasized to be of
paramount importance. Preparation includes the removal of all deteriorated, polluted or
corroded material. The removal process must expose sound material with unobstructed
porosity, without inducing bruising or microcracking. Then, in order not to jeopardize the
quality of the bond, thorough cleaning of the surfaces is required. In most cases, flushing with
highpressure water is satisfactory.
y Locating the overlay joints is also a question of quite significant importance. Most
generally, the best approach is still to locate them exactly above the existing joints or
discontinuities in the substrate.
y Obviously, attention shall not be paid strictly to the abovementioned steps, as a tight control of
all the phases of the repair process is critical. Education and training of the workmanship and
enforcement of a quality control process are keys to successful overlay applications.
Besides, the State of the Art reviews theoretical and practical considerations relating to bond
enhancement, trying along the way to resolve some issues that have been much debated.
y Optimum bond is associated with maximum "effective contact area" (see 4.3) at the
interface. An overlay material exhibiting a rheological behavior such as to maximize the
filling of the smaller asperities of the substrate surface is thus advisable. In this respect,
self-compacting overlaying materials can be expected to yield excellent bond.
y The relevance of using of bonding agents is seriously questioned, again. Bonding agents
cannot compensate for inadequate surface preparation and may even sometimes act as
bond breakers when used improperly. In all cases, grouts as bonding agents should be
avoided.
y Placing a rheologically adequate overlay material on a sound and clean SSD ("saturated
surface dry") surface without any bonding agent yields the most consistently successful
end results. The expected tensile bond strength then typically ranges within 1.5 and 2.5
MPa, and beyond.
The State of the Art delivers groundbreaking information, notably with respect to the
understanding of debonding in overlays and a practical reinforcing approach to prevent it.
y Based on experimental evidence as well as on numerical simulations, it can be concluded that

overlay debonding is almost inevitably initiated from a discontinuity, crack, joint or boundary,
where the most critical interfacial tensile and shear stresses arise.
y Fiber reinforcement is beneficial in controlling crack opening and propagation in these areas. In
addition, it efficiently counteracts the aggravating influence of shrinkage cracking upon the
overlay fatigue behavior.
y Fatigue is progressively exhausting the natural interlocking effect in the cracks and along the
debonded interface area, thus altering the overall debonding process. At some point, only the
interlocking effect provided by the presence of reinforcement (such as fibers), if any, becomes
available. On that basis, a conservative calculation to evaluate fatigue effects in fiber
reinforced overlays is proposed, allowing only for the reinforcementinduced interlocking.
y The available design rules for overlay bond are rather basic and empirical. In fact, the required
minimum bond strength values prescribed in the various consulted codes vary by a factor of up
to 3. Moreover, the codes ignore fundamental parameters such as the relative thickness and
elastic modulus of the substrate and the overlay, and overlay reinforcement.
More comprehensive design rules taking into consideration the abovementioned issues are
needed. Some are currently under development.
The key practical issues highlighted in this State of the Art are at the heart of the final task of this
TC's work, which is the publication of "Recommendations" for durable bonded overlays.

Unedited Version 193 Rls

Uploaded by

Copyright:

Available Formats

Unedited Version 193 Rls

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unedited Version 193 Rls

Uploaded by

Copyright:

Available Formats

STAR 193RLS

Bonded CementBased Material Overlays

This publication has been published by Springer in

Please note that the following PDF file, offered by

2. OVERLAY DESIGN PROCESS

3. CONDITION EVALUATION OF THE EXISTING STRUCTURE PRIOR TO

4.1 Classification of bond

(A. Turatsinze, H. Beushausen, R. Gagn, J.-L. Granju, J. Silfwerbrand,

8. PRACTICE AND QUALITY ASSURANCE

8.1 Scope and definitions

9. MAINTENANCE AND REPAIR OF OVERLAYS

9.1 Types and causes of distress

B. Bissonnette CRIB - Universit Laval (Qubec, QC) Canada Chairman

(1) Laboratoire Matriaux et Durabilit des Constructions (LMDC), UPS-INSA, Toulouse,

Other reasons for overlaying slabs or concrete pavements include:

Chapter 2: Overlay: decision and construction process.

Chapter 3: Condition evaluation of the existing structure prior to overlay.

Chapter 5: Structural behavior.

Chapter 8: Practice and Quality Assurance.

Chapter 9: Maintenance and repair of overlays.

Chapter 10: Conclusion.

(1) The University of Texas at Austin (TX), U.S.A.

2.1 PURPOSE OF OVERLAYS

Other reasons for overlaying concrete pavements or slabs include:

provide an improved frictional surface for pavements or bridges;

2.2 MATERIALS SELECTION

2.2.1 Slabs on grade/Pavements

The watercement ratio is determined by strength and durability (permeability) requirements.

2.2.2 Structural slabs and decks

2.3.1 Slabs on grade/Pavements

2.3.2 Structural slabs and decks

2.4 CONSTRUCTION PROCEDURES

2.4.1 Steel placement

2.4.2 Environmental effects

Concrete temperature should be monitored throughout construction. Temperature at placement

cooling the aggregates or concrete;

1) Application of curing compound.

2) Application of membrane curing.

Evaporation below 0.5 kg/m2/hr Membrane curing

Temperature drop in next 24 hours less than Membrane curing

2.5 THE BCO PROCESS

Figure 2.1: BCO development process

Figure 2.2: Flowchart of the project selection stage

The design decisions that must be made are:

design period (overlay design life);

The original pavement must be characterized, which usually requires:

coring to determine modulus of elasticity, tensile strength, and thickness; and

structural capacity of existing pavement; and

Some of the constructionrelated considerations are:

1) Materials and aggregates.

4) Tied PCC shoulders.

5) Critical weather conditions.

b) Substrate surface temperatures > 50 C at concrete placement may lead to inadequate

2.5.4 Quality control/Quality assurance (QA/QC)

statistical sampling samples from each subset of paving should be required;

3.2 ASSESSMENT OF PAVEMENT/SUBSTRATE BASE

3.2.1 Principles of evaluation

3.2.2 Condition survey of distress