Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
Cracking sensitivity of cementitious repair materials: assessments and development of test methods
Mémoire
Maxim Morency
Maîtrise en génie civil
Maître ès sciences (M.Sc.)
Québec, Canada
© Maxim Morency, 2013
iii
Résumé
La détérioration prématurée des réparations en béton est le résultat de divers processus
physico-chimiques et électrochimiques. La fissuration du béton de réparation est une des
causes de dégradation les plus importantes et peut entraîner et accélérer le processus de
corrosion des barres d’armature et ainsi diminuer significativement la durée de vie, non
seulement de la réparation, mais de l’ouvrage dans son ensemble.
Ce travail de recherche avait pour objectif de contribuer à identifier et développer une
méthode d’essai objective et établir une corrélation avec des techniques d’essai indirectes.
L’essai proposé utilise une dalle de béton de référence comportant en surface une cavité à
combler avec le matériau à tester. Cet essai permet de simuler une réparation superficielle
réelle, avec un degré de restriction représentatif et la possibilité de réaliser le vieillissement
dans des conditions d’exposition diverses. Afin d’être en mesure d’apprécier le caractère
représentatif de l’essai de performance et des techniques d’essai indirectes, des réparations
expérimentales de grandeur nature ont aussi été réalisées sur des structures exposées en
conditions extérieures.
v
Abstract
The premature deterioration of concrete repairs in service is a result of a variety of physico-
chemical and electrochemical processes. Among the most serious causes of repair failures
is cracking of the repair. Cracking may result in the reduction of an effective cross-
sectional area of the repaired structure and increase the effective permeability of the
concrete cover, thus promoting corrosion of the reinforcement and further deterioration.
The main objective of this project was to contribute to the development and assessment of a
reliable test method for evaluating the sensitivity to cracking of repair materials. A
performance test was developed and used to establish correlations with existing indirect test
methods (ring test, beam deflexion test, drying shrinkage test, etc.). The performance test
method uses of a reference slab containing a cavity on the top surface to be filled with the
repair material to be tested. The reference test slab, which offers a degree of restraint
comparable to what is found in reality, allows simulating the behavior of the material in
real concrete repair conditions. In order to better evaluate the test methods, experimental
repairs have also been made on existing structures exposed to service conditions.
vii
Table of content
RÉSUMÉ ........................................................................................................................................................ III
ABSTRACT...................................................................................................................................................... V
TABLE OF CONTENT ................................................................................................................................. VII
LIST OF TABLES ........................................................................................................................................... IX
LIST OF FIGURES ......................................................................................................................................... XI
1 INTRODUCTION ................................................................................................................................... 1
1.1 GENERAL .............................................................................................................................................. 1 1.2 CONCRETE REPAIR ................................................................................................................................ 1
1.2.1 Description of a concrete repair ................................................................................................. 1 1.2.2 Compatibility of concrete repairs ............................................................................................... 3
1.3 DETRIMENTAL EFFECTS OF REPAIR CRACKING ...................................................................................... 4 1.4 TESTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIAL .................................... 5 1.5 RESEARCH OBJECTIVES ......................................................................................................................... 6
2 MECHANISMS OF CRACKING OF CEMENTITIOUS REPAIR MATERIALS ......................... 9
2.1 GENERAL .............................................................................................................................................. 9 2.2 DRYING SHRINKAGE ............................................................................................................................ 12
2.2.1 Capillary stress theory .............................................................................................................. 12 2.2.2 Disjoining pressure theory ........................................................................................................ 13 2.2.3 Surface pressure ........................................................................................................................ 13 2.2.4 Factors influencing drying shrinkage ....................................................................................... 13
2.3 CREEP ................................................................................................................................................. 16 2.3.1 General ..................................................................................................................................... 16 2.3.2 Creep mechanisms .................................................................................................................... 18 2.3.3 Factors influencing creep ......................................................................................................... 20
2.4 CONCRETE SENSITIVITY TO CRACKING TESTING METHODS .................................................................. 22 2.4.1 REMR box test method .............................................................................................................. 23 2.4.2 Baenziger block test method ..................................................................................................... 25 2.4.3 Ring test .................................................................................................................................... 25 2.4.4 German angle test ..................................................................................................................... 28 2.4.5 SPS plate test............................................................................................................................. 29 2.4.6 Laval beam deflection test ......................................................................................................... 30 2.4.7 Thoro system test ....................................................................................................................... 31
3 EXPERIMENTAL SET-UPS AND PROCEDURES ......................................................................... 33
3.1 SCOPE OF TEST PROGRAM .................................................................................................................... 33 3.1.1 Phase I ...................................................................................................................................... 33 3.1.2 Phase II ..................................................................................................................................... 33
3.2 DESCRIPTION OF THE REPAIR MATERIALS ............................................................................................ 34 3.2.1 Phase I materials ...................................................................................................................... 34 3.2.2 Phase II materials ..................................................................................................................... 35
3.3 DESCRIPTION OF THE TEST CONDITIONS .............................................................................................. 36 3.4 DESCRIPTION OF THE TEST PROCEDURES ............................................................................................. 36
3.4.1 Mechanical properties .............................................................................................................. 36 3.4.2 Free drying shrinkage ............................................................................................................... 37 3.4.3 Flexural creep test .................................................................................................................... 39 3.4.4 Restrained shrinkage tests ........................................................................................................ 41 3.4.5 Experimental repairs ................................................................................................................ 50
4 RESULTS AND DISCUSSION - PHASE I ......................................................................................... 53
viii
4.1 INTRODUCTION .................................................................................................................................... 53 4.2 CHARACTERIZATION ........................................................................................................................... 53
4.2.1 Fresh material properties .......................................................................................................... 54 4.2.2 Mechanical properties............................................................................................................... 55
4.3 DEFORMATIONAL BEHAVIOR CHARACTERIZATION .............................................................................. 58 4.3.1 Drying shrinkage ....................................................................................................................... 58 4.3.2 Flexural creep test ..................................................................................................................... 61 4.3.3 Laval beam deflection test ......................................................................................................... 63 4.3.4 Vertical Beam Curling test ........................................................................................................ 66 4.3.5 SPS plate test ............................................................................................................................. 66
4.4 CRACKING SENSITIVITY EVALUATION TEST METHODS ......................................................................... 67 4.4.1 General ...................................................................................................................................... 67 4.4.2 Box test and Baenziger block test .............................................................................................. 68 4.4.3 Experimental repair .................................................................................................................. 72 4.4.4 German angle test ..................................................................................................................... 74 4.4.5 Restrained shrinkage test .......................................................................................................... 75
4.5 CONCLUSION AND RECOMMENDATIONS – PHASE I .............................................................................. 79
5 PHASE II................................................................................................................................................ 81
5.1 INTRODUCTION .................................................................................................................................... 81 5.2 PHASE II MATERIALS ........................................................................................................................... 82 5.3 CHARACTERIZATION ........................................................................................................................... 83
5.3.1 Mechanical properties............................................................................................................... 83 5.3.2 Drying shrinkage test (ASTM C 157 Modified) ......................................................................... 88
5.4 EXPERIMENTAL REPAIRS ..................................................................................................................... 90 5.4.1 Baenziger block test .................................................................................................................. 94 5.4.2 Ring test ..................................................................................................................................... 96
5.5 CONCLUSION ..................................................................................................................................... 101
6 CONCLUSION AND RECOMMENDATIONS ............................................................................... 103
6.1 CREEP BOX AND BAENZIGER BLOCK TESTS ....................................................................................... 103 6.2 RING TEST ......................................................................................................................................... 103 6.3 INDIRECT TEST METHOD .................................................................................................................... 104 6.4 RECOMMENDATION FOR FUTURE RESEARCH ...................................................................................... 104
7 BIBLIOGRAPHY................................................................................................................................ 107
8 APPENDIX A ...................................................................................................................................... 113
ix
List of Tables
Table 1: materials mix design ............................................................................................... 35 Table 2: Rate of aggregate addition for extended mortars .................................................... 36
Table 3: Dimensions of the box test specimens .................................................................... 42 Table 4: Box test concrete mixture design ............................................................................ 43 Table 5: Fresh material properties ........................................................................................ 54 Table 6: Three day mechanical properties ............................................................................ 56 Table 7: 28 day mechanical properties ................................................................................. 57
Table 8: Comparison of 28 day and long term results of ASTM C 157 modified test ........ 60 Table 9: Summary of crack evaluation tests ......................................................................... 70
Table 10: German Angle test results ..................................................................................... 75
Table 11: Summary results of the REMR ring test ............................................................... 76 Table 12: ASTM-like test results .......................................................................................... 78 Table 13: Approached calculation of residual stress ............................................................ 80
Table 14: Test carried out for each location ......................................................................... 82 Table 15 : Compressive strenght (ASTM C 39) ................................................................... 84 Table 16 : Modulus of elasticity (ASTM C 469) .................................................................. 86
Table 17 : Splitting tensile strength (ASTM C 496) ............................................................. 87 Table 18 : Overlays cracking summary ................................................................................ 93
Table 19 : Baenziger Block test cracking summary ............................................................. 96 Table 20 : Summary of the ring test results .......................................................................... 98
Table 21: Approach calculation of residual stress .............................................................. 101
xi
List of Figures
Figure 1: Model of a three phases repair system (Emmons and Vaysburd, 1995) ..................... 2 Figure 2: Factors affecting durability of concrete repair (Emmons and Vaysburd 1995) .......... 3
Figure 3: Holistic model of concrete repair failure (Emmons and Vaysburd 2000) ................ 10 Figure 4: Simplified representation of shrinkage-induced stresses and cracks in a
concrete overlay (tension: positive; compression: negative) (After Pigeon and
Bissonnette, 1999) ..................................................................................................... 11 Figure 5: Influence of water to cement ratio and aggregate content on drying shrinkage
(Odman 1968) ........................................................................................................... 14 Figure 6: Effect of aggregate type on drying shrinkage (Troxell et al. 1958) .......................... 15
Figure 7: Effect of relative humidity on creep (Troxell et al. 1958) ........................................ 21
Figure 8: Effect of stress level on creep (Neville 1995) ........................................................... 22 Figure 9: REMR Box test specimen’s geometry (after Emmons et al. 1998) .......................... 24 Figure 10: Schematic representation of the Baenziger block ................................................... 25
Figure 11: German angle test setup (Emmons and Vaysburd 1995) ........................................ 28 Figure 12: Shcematic of SPS plate test ..................................................................................... 29 Figure 13: SPS plate test specimens ......................................................................................... 30
Figure 14: Laval BD Test Specimen ......................................................................................... 30 Figure 15: Thoro System (Emmons and Vaysburd 1995) ........................................................ 31
Figure 16: Standard comparator to measure drying shrinkage strain ....................................... 38 Figure 17: Comparator with mobile plate to measure drying shrinkage strain ........................ 38
Figure 18: Tensile creep system (after Bissonnette 1996) ........................................................ 39 Figure 19: Sketch of the flexural creep test setup ..................................................................... 40
Figure 20: Flexural creep test specimens and deflection measuring device ............................. 40 Figure 21: Box test specimen .................................................................................................... 42 Figure 22: REMR ring test configuration ................................................................................. 45
Figure 23: ASTM-like ring test configuration ........................................................................... 46 Figure 24: Vertical beam curling test configuration ................................................................. 49
Figure 25: AWTTS Pier – US Navy, Port Hueneme, California .............................................. 50 Figure 26: Schematic view of the experimental repair set up ................................................... 51 Figure 27: Surface preparation at the Bureau of Reclamation (BOR) ...................................... 52
Figure 28: Surface preparation at the US Navy ........................................................................ 52
Figure 29: Drying shrinkage (ASTM C157 mod.), laboratory conditions ............................... 59 Figure 30: Drying shrinkage (ASTM C 157 mod.) of repair materials (Laval) versus
REMR project 28-day limit .................................................................................... 60
Figure 31: Drying shrinkage (ASTM C 157 mod.), field conditions – Laval University ........ 61 Figure 32: Flexural creep test results ........................................................................................ 62 Figure 33: Flexural creep test calculations ............................................................................... 63 Figure 34: Beam deflection test results, laboratory conditions ................................................ 64 Figure 35: Beam deflection test calculations, laboratory conditions ........................................ 64
Figure 36: Beam deflection test, field conditions (Laval University) ...................................... 65 Figure 37: Vertical Beam Curling test ...................................................................................... 66 Figure 38: SPS plate test results ............................................................................................... 67
Figure 39: Cracking indices representation .............................................................................. 68 Figure 40: cracking index versus material and test specimen .................................................. 69
xii
Figure 41: Number of transverse crack(s) onto box and Baenziger block tests specimens ..... 70 Figure 42: Pull-off test on box test with Germann Instrument equipment ............................... 71 Figure 43: Phase I pull-Off test results..................................................................................... 71
Figure 44: Depth of crack for P-Mortar ................................................................................... 72 Figure 45: Crack mapping of experimental repair – Concrete mix.......................................... 73 Figure 46: Crack mapping of experimental repair – Mortar mix ............................................. 74 Figure 47: Crack mapping of experimental repair – P-Mortar mix ......................................... 74 Figure 48: Time to first crack in REMR Ring Test .................................................................. 76
Figure 49: Ring test – Concrete ................................................................................................ 77 Figure 50: Ring test – Mortar ................................................................................................... 78 Figure 51: Ring test – P-Mortar ............................................................................................... 78
Figure 52: BOR (left) and US Navy (right) experimental sites ............................................... 81 Figure 53: Paddle type mixer (left) and rotating drum mixer (right) ....................................... 83 Figure 54: Comparison of 28-day compressive strength between BOR and Laval
University (left) ...................................................................................................... 85
Figure 55: Comparison of 28-day modulus of elasticity between BOR and Laval
University (right) .................................................................................................... 85 Figure 56: Comparison of 28-day tensile splitting strength between BOR and Laval
University ............................................................................................................... 88
Figure 57 : Drying shrinkage results (ASTM C 157 modified) – BOR ................................... 88 Figure 58: Relationship between length change and number of cracks in repair – BOR ........ 89
Figure 59: Relationship between free shrinkage and cracking in experimental repairs
(after 9 months) ...................................................................................................... 90
Figure 60 : Cracking in experimental repairs at BOR (9 months after placement) ................. 91 Figure 61 : Cracking in experimental repairs at US Navy (9 months after placement) ........... 92
Figure 62 : Tensile bond strength ............................................................................................. 94 Figure 63 : Mapping of Baenziger blocks showing cracks ...................................................... 95 Figure 64 : ASTM C 1581 Ring test results for material 1M (left) and 1C (right) .................. 96
Figure 65 : ASTM C 1581 Ring test results for material 2M (left) and 2C (right) .................. 97 Figure 66 : ASTM C 1581 Ring test results for material 3M (left) and 3C (right) .................. 97
Figure 67 : ASTM C 1581 Ring test results for material 4M (left) and 4C (right) .................. 98 Figure 68: Relationship between stress development rate in the Ring test and the
number of cracks in the experimental repairs (after 9 months) .............................. 99 Figure 69: Relationship between ASTM C 1581 ring test and cracking of experimental
repairs for mortars (after 9 months) ..................................................................... 100 Figure 70: Relationship between age at cracking in the ring test (ASTM C 1581) and the
Length change test (ASTM C 157 modified) ....................................................... 100 Figure 71: Creep Box test – Concrete, small size .................................................................. 113 Figure 72: Creep Box test – Concrete, medium size .............................................................. 113
Figure 73: Creep Box test – Concrete, large size ................................................................... 114 Figure 74: Baenziger block – Concrete mix ........................................................................... 114 Figure 75: German Angle test, Laval University – Concrete ................................................. 115 Figure 76: German Angle test, USBR – Concrete ................................................................. 115 Figure 77: Creep Box test – Mortar, small size ...................................................................... 115
Figure 78: Creep Box test – Mortar, medium size ................................................................. 116
Figure 79: Creep Box test – Mortar, large size ...................................................................... 117
Figure 80: Crack mapping of Baenziger block – Mortar mix ................................................ 117
xiii
Figure 81: German Angle test, Laval University – Mortar ..................................................... 118 Figure 82: German Angle test, USBR – Mortar ..................................................................... 118 Figure 83: Creep Box test – Polymer Mortar, small size ........................................................ 118
Figure 84: Creep Box test – Polymer Mortar, medium size ................................................... 119 Figure 85: Creep Box test – Polymer Mortar, large size ........................................................ 119 Figure 86: Crack mapping of Baenziger block – Polymer modified mortar mix ................... 120 Figure 87: German Angle test, Laval University – Polymer Modified Mortar ...................... 120 Figure 88: German Angle test, USBR – Polymer Modified Mortar ....................................... 120
xv
To my beloved wife Annie, and my family..
.
xvii
Acknowledgments
First of all, I would like to sincerely thank my research supervisor, Dr. Benoît Bissonnette,
for the constant support, latitude and freedom he offered me throughout the project. I really
appreciated all of the opportunities he gave me such as participation in international
conferences (ACI, RILEM, etc.), internships at the US Bureau of Reclamation and at the
US Navy where I learnt a lot about engineering and research cooperation. I would also like
to thank my co-supervisor, Dr. Alexander M. Vaysburd, who has contributed significantly
to the research project by his comments, thoughts and his guidance during the project.
I would like to thank the CRIB members, who made this project possible from a technical
point of view by their contributions: Pascal Dorion, Julie Conseiller and Denis Lagacé. I
would also like to thank the students that helped me carry out the experimental program:
Kim Lafrance, Carl Fortin, Julie Gingras, Normand Bélair, John Cafarelli, and my friends
François Paradis for his support and Dennis Burns for reviewing English in my thesis. I
apologize to all the others that I might have forgotten, but please consider yourselves
gratefully thanked.
In addition to the CRIB family, I would like to address a special thanks to Steve Reo from
the US Bureau of Reclamation (Denver, CO) for all the technical work he accomplished
with care, accuracy and his dedication to the part of my research project that have been
handled at the Bureau of Reclamation. Also, I would like to give a special thanks to Kurt
VonFay who made possible my internship at the Bureau of Reclamation. Special thanks
also goes to Douglas Burke from the US Navy, who accepted to participate in this project
and made possible my internship at the US Navy site in Port Hueneme, CA, and to his
colleague Jesse McNolty for his technical help.
Finally, I would like to thank my family for their support and a special thanks to my wife,
Annie, who believed in me at all times and provides me with constant support. This was of
great help.
1
1 Introduction
1.1 General
In North America, the nineteenth century gave rise to an outstanding growth of concrete
structures to accommodate for the ever growing population and traffic. Most of these
structures are exposed to various environmental conditions causing concrete degradation
and many of these structures show signs of ageing such as cracking, corrosion of rebar,
delaminating, spalling and other important pathologies which lead to a gradual deterioration
and failure of the structure.
Several problems are left unsolved in the field of concrete repairs nowadays, notably
cracking of the repairs. Although it may seem to be an aesthetic issue at first glance, it
however often turns out to be related to durability. In fact, cracks in concrete structures
become a preferential path for aggressive agents to penetrate into the concrete and attack
steel reinforcement, which leads to further degradation problems.
In order to help engineers, contractors and individuals involved in repair or rehabilitation
processes such as DOT’s, structure owners, etc., a research project was set up to identify
and develop test methods to evaluate cracking sensitivity of repair materials and guide the
engineers in selecting appropriate materials. This project has been named CREEP
(Concrete Repair Engineering Experimental Program) and involves numerous partners such
as the US Government, repair consultants, repair material manufacturers, Laval University
concrete research center and others. The spirit behind this project is clearly to help the
repair industry by diffusing knowledge and provide evaluation tools.
1.2 Concrete repair
1.2.1 Description of a concrete repair
A concrete repair is more complex than an overlay or a screed of concrete resting on a
substrate (old concrete). A concrete repair process consists of first assessing the
2
deterioration of the structure and then the removal of the unsound concrete, an adequate
surface preparation and the application of a proper cementitious repair material, adapted to
the specificity of a repair. The real challenge of a concrete repair lies in a favorable
combination of two different materials having different properties. Emmons and Vaysburd
[1993] represented a repaired system, what was originally a single material, as a new three-
phase composite system. These phases are; the substrate, the transition zone (interface) and
the neoteric material (repair material) (Figure 1).
Figure 1: Model of a three phases repair system (Emmons and Vaysburd, 1995)
The substrate is generally a concrete that aged for many years that has its specific
mechanical properties has also undergone volumetric changes and it is now considered
stable from a volumetric standpoint. This means that there is no more significant volume
change over time due to drying shrinkage. The interface is a thin layer generally constituted
of cement paste that progressively evolves (from the interface to the repair material)
towards a concrete or a mortar. The interface is responsible for the bond between the
substrate and the repair material. The fresh concrete used for the repair is designed upon the
specific needs of the project. It can be a high strength concrete or a concrete especially
designed to resist to chloride penetration etc. Whichever type of concrete chosen, it will
change in volume over time, which will induce stress in the repair material.
REPAIR MATERIAL
TRANSITION ZONE
(INTERFACE)
EXISTING CONCRETE
3
1.2.2 Compatibility of concrete repairs
A repair is considered successful when it is free of problem and excessive deterioration on
a mid-term basis (10-15 years). However this is not always the case and failure of concrete
repairs sometime happens. To be successful and free of crack, a repair material should be
compatible with its substrate. Emmons and Vaysburd [1995] defined the compatibility of
materials in a repair system as a balance of dimensional, chemical and permeability (Figure
2); they also add electrochemical compatibility when steel rebars are part of the repair. In a
balanced system, the repair material would be able to withstand the stresses induces by the
different factors inherent to a repair system.
Figure 2: Factors affecting durability of concrete repair (Emmons and Vaysburd 1995)
Dimensional incompatibility is one of the principal causes of concrete repair failures
[Emmons and Vaysburd 1995]. Typical concrete can sustain a deformation ranging from
100 µm/m to 200 µm/m while the deformation imposed by the drying shrinkage is
generally comprised between 500 µm/m and 1000 µm/m. From these numbers it could look
like that no repair material could resist to cracking. However, a lot of repairs have been
proven durable. The level of the restraint and the relaxation due to creep and other factors
help the repair material to resist to cracking.
Compatibility
of repair
material with
substrate
Dimensional Chemical Permeability Electrochemical
Drying
Thermal
Creep
Re
Modulus of
Elasticyity
4
A successful repair is also tributary of the bond quality of the repair material to the
substrate. A repaired structure leads to a concrete / repair material composite system in
which bond is very important for proper action [Abu-Tair et al. 2000]. When the bond
between a repair material and the substrate is weak or fails, the repair is subject to failure
and will eventually need to be done again.
In addition to cracking of the repair material and the bond between repair material and old
concrete, alkali-silica reaction (ASR) is a common pathology that affects not only original
concrete structures but also concrete repairs. ASR leads to cracking but also to internal
disorders which deteriorate concrete and thus diminish its durability.
Depending in which area repairs are performed, the climate could be a source of problems.
In the northern part of America, there are severe conditions due to freezing and thawing and
the use of deicing salts. Freeze and thaw cycles are an important source of internal stresses
for concrete. If the air void system is not adequate, durability problem will occur.
Another important factor for concrete repair failure which is not directly related to the
repair material itself is poor workmanship. When proceeding to a repair, the quality of
workmanship can easily make the difference between a good and a bad repair. As an
example, if the finishing activities are not properly performed, it can turn out in a poor
surface resistance which led to a rapid surface deterioration. Also if the curing activity is
not done with sufficient care, this will also result in a lack of resistance or cracking. It is
frequent on jobsite to observe important mistake on steps that require precision like
positioning the rebars in the forms. Too many times new reinforcements are bent out of
their real or designed position and this leads to a very thin concrete cover. Also if the
working crews do not pay attention to their work simple activities like the curing will be
forgotten or not properly done and it will result in bad quality repairs.
1.3 Detrimental effects of repair cracking
Everybody agrees that cracking or surface crazing of a repair is aesthetically awful and
diminish the level of confidence for the common people. If this was the only problem with
5
cracking, it would be easy to solve by the application of membranes or painting the surface.
The fact is that cracks in a repair material offer an excellent path for aggressive agents such
as chlorides to reach the steel reinforcements and start the corrosion process. The corrosion
of reinforcement accelerates dramatically the deterioration process. When it corrodes, the
steel increases its volume which causes internal pressures and makes the concrete to crack
even more and finally delaminate it and make it spall.
Cracks in concrete might also be detrimental when free particles fill-in those cracks and
then restrict the movement of concrete due to thermal effects or bending moments. This
would lead in increase of internal pressures which would also accelerate concrete
deterioration. Water easily fills cracks and generates pressures when freezing (climates
where the temperature goes under freezing point). This will also lead to rapid deterioration
and acceleration of cracks progression.
1.4 Testing the sensitivity to cracking of cementitious repair material
When an engineer has to design a repair of a concrete structure, he will have to go through
a series of steps in order to end up with a durable repair at beneficial cost. The engineer will
have to perform an evaluation of the existing structure, assess the actual damages and
decide what will be repaired and how it will be done. Calculation will also be made to
ensure the safety of structural behavior of the new system. The engineer responsible for the
repair design will also make a very important decision: what will be the repair material or
what will be the specifications of this material? Unfortunately most of the engineers would
recommend a concrete mixture for a repair based only on its mechanical properties,
particularly the compressive strength. The main reason that explains why engineers select a
repair material based on the compressive strength is the lack of reliable tools to evaluate the
tendency to cracking of a repair material. Since they do not have the required tools, they
base their decision on what they know and the common statements.
Some researchers have proposed the “Repair like with like” concept which consisted of
repairing a concrete structure with a concrete that has the same properties as the original
one [Pullar-Strecker 1987]. Time has proven this statement not very relevant. Moreover,
6
according to Plum [1990], this concept is illogical. Plum suggests from inspections of
different repairs that have failed that perfect parity of material properties is not a guarantee
of success for a repair. The best approach for selecting a good repair material would be to
evaluate the tendency to cracking of the material. Up to date there is no standardized test
and each company evaluates the sensitivity to cracking of repair material using in-house
tests. Time is now to develop reliable tools in predicting the sensitivity to cracking of a
repair material. If a material can be proven resistant to cracking, better durable repairs will
be performed and it will turn out to be very cost effective.
1.5 Research objectives
For a long time, the only parameter that was used for the selection of a material to be used
to perform a concrete repair was its compressive strength. Experience showed that
compressive strength specification is not sufficient to avoid cracking in a repair. In fact, a
repair system is more complex and involves more parameters than only the compressive
strength. In order to get better crack resistant repairs, it is necessary to have tools to
evaluate the candidate materials for cracking sensitivity. Based on this statement, the
general objectives for this research project are to:
Contribute to the development of
o Performance criteria for cementitious repair materials,
o Tools for evaluating the sensitivity to cracking of repair materials,
o Laboratory test methods.
This project focuses on the performance of cementitious repair materials in concrete surface
repair with the main goal of minimizing cracking. The specific objectives of the project are
the following:
To determine the optimal geometry / size of a test specimen that can be used for
evaluating the sensitivity to cracking of a repair material. The optimal size can be
defined as the smallest size that will be representative of the performance of a repair
material (with regards to cracking) in a real-size repair;
To assess the applicability of some of the existing test methods for evaluation of
cracking tendency of repair materials;
7
To perform in-situ validation of selected tests in the first part of the research
program.
It is not the intent of this program to promote any proprietary repair products. Its sole
purpose is to develop and disseminate knowledge to benefit the North American repair
industry, scientific and technical know-how such as available test data (both field and
laboratory), planned and ongoing studies, and current and new types of repair materials
available. It is vital to remember that many, if not all, of the surface repair durability
problems observed in the field are due to the improper selection and use of repair materials.
Increased knowledge, better test procedures as well as appropriate use of existing tests,
should help to solve this problem.
9
2 Mechanisms of cracking of cementitious repair materials
2.1 General
The premature deterioration and failure of concrete repairs in service is a result of a variety
of physicochemical and electrochemical processes occurring in composite repair systems.
Among the most serious causes of repair failures, cracking in the repair is the most
important one. Cracking may result in the reduction in the effective cross-sectional area of
the repaired structure, and always substantially increases permeability, which leads to
premature corrosion and deterioration. Figure 3 shows an idealized model of repair failure
[Morency et al. 2005].
Many material properties affect the susceptibility of concrete repair to cracking. Drying
shrinkage of repair materials is one of the major mechanisms leading to cracking. The
tensile strains and stresses generated by the restrained shrinkage (Figure 4) can easily
exceed the tensile strength of the repair material, and thus cause cracking and/or debonding
[Pigeon and Bissonnette 1999]. Tensile stresses caused by restrained shrinkage can be, to a
certain degree, relaxed by creep: according to some researchers, it is quite probable that the
satisfactory performance of some superficial repairs is due to this phenomenon.
Also, there are other factors which, to a large degree, affect the cracking tendency of the
repair. Among them are important material properties, such as modulus of elasticity, creep,
and the composite repair system’s characteristics such as degree and uniformity of restraint.
However, there is presently no agreement on the relative influence of each of these
properties and factors on susceptibility of repair to cracking. Some of the properties are
found so interrelated that it is practically impossible to affect one of them without affecting
the second one. Results of the U.S. Army Corps of Engineers (COE) study [Emmons et al.
1998] indicate that the higher stresses are induced by increased drying shrinkage more than
offset any additional stress relaxation caused by increased creep.
10
Figure 3: Holistic model of concrete repair failure (Emmons and Vaysburd 2000)
Repair
REPAIRED CONCRETE STRUCTURE
Repair
Restrained Volume
Changes
Weakened Bond Between
Rebars & Concrete Along
the Repair Perimeter
Exterior Weathering
& Loading Effects
Cracking
Increase in Permeability
Increase in Permeability
along the Perimeter of
the Repair
Penetration of H2O,
CO2, Cl- from Outside
Penetration of H2O, Cl-
from inside
(1) Depassivation of the Steel Reinforcement
(2) Formation of the Rust Products
(1) Accumulation and Expansion of Rust products
(2) Loss of Bond Between Reinforcement and
Repair Material
Expansion, More Cracking,
Enlargement of Existing
Cracks, Spalling
Expansion, Cracking, Spalling of
Existing Concrete Adjacent to the
Repair
REPAIR FAILURE
11
Figure 4: Simplified representation of shrinkage-induced stresses and cracks in a concrete overlay
(tension: positive; compression: negative) (After Pigeon and Bissonnette, 1999)
These difficulties may be solved where the approach of the total strain in a drying repair is
adopted. This approach is adopted in test methods such as the Ring test, German Angle, and
Structural Preservation System plate test (SPS Plate test) [Poston et al. 1998], Laval
University beam deflection and others [Morency et al. 2005].
Though some progress has been achieved lately, it is still difficult to reliably correlate the
results of these tests to actual field performance, and predict with reasonable confidence the
cracking behavior of the repair material in-situ.
In some cases, the deteriorating factor is chloride ions from salt (seawater exposition or de-
icing salts), fatigue failure and even normal wear and tear. In fact, some factors interrelate
and degradation is generally a result of a combination of such factors [Emmons and
Vaysburd 1995]. To reach or go beyond its expected service life, maintenance and repairs
must be performed on the deteriorated structures. Over the last decade, numerous structures
were repaired with cement based materials and in some cases the repair did not last as long
shrinkage-induced
cracking
Old Concrete
?
σ (y)εsh (y)
superficial
drying
free shrinkage
strain distribution
at time t
shrinkage-induced
stress* distribution
at time t
contraction extension comp tension
path?
*Not taking into account external restraints
(ex, slab self-weight, friction with subgrade, etc…)
Overlay
+- +
12
as expected, cracking, spalling and delamination occurred. Most of the time, cracking was
the leading factor in non-durable repairs.
Since cracking is identified as one of the most threatening phenomena affecting durability
of concrete repair, it is of great interest to understand how it happens and how to evaluate
the sensitivity to cracking of a repair material. Generally speaking, a crack in a concrete
repair occurs when the tensile stress is greater than the tensile strength of the material. In a
repair, the tensile stresses can build up due to numerous factors, but drying shrinkage is one
of the most important sources of tensile stress in concrete repairs.
2.2 Drying shrinkage
Shrinkage of cementitious material has different sources such as temperature, moisture
movement, carbonation or wind. The shrinkage due to drying is the most important source
of disturbance for concrete repairs since it is the driving force for stress development; thus
it may lead to cracking of repairs. When submitted to ambient air, which is unsaturated,
cementitious material will lose moisture due to drying. The loss of moisture is accompanied
by a reduction of the specimen’s volume, which is not proportional to the volume of water
evaporated [Neville 1995]. No theory can solely explain entire shrinkage deformation that
occurs in cementitious materials. However, three theories are proposed, and generally
accepted, to explain the mechanisms behind drying shrinkage.
2.2.1 Capillary stress theory
In 1929, Freyssinet brought the capillary stress theory in order to explain the mechanism of
shrinkage due to water loss [Baron and Autery 1982]. According to this theory, the larger
pores in the medium are the first to be emptied. When a pore is emptied, a meniscus is
formed and it induces compressive force into the solid state which contracts the material’s
skeleton. From Kelvin’s law and Laplace’s law, it is said that smaller the pore diameter is,
greater is the force generated by the meniscus. In concrete (or mortar), this is valid upon a
certain limit, which corresponds to a pore diameter of 2.5 nm. Below that limit, it is not
possible to form a meniscus [Young et al. 1986].
13
This theory can explain drying shrinkage for relative humidity over 45%. To explain the
shrinkage phenomenon below this point, other mechanisms are required.
2.2.2 Disjoining pressure theory
In order to complete the explanation of the drying shrinkage mechanisms, Power worked to
demonstrate that the water contained in between two plane surfaces cannot be absorbed
freely if the distance separating these two surfaces is less than 3 nm. The water contained
between those surfaces (areas of hindered adsorption) is compressed and is in equilibrium
with the attraction forces of the CSH [Powers 1968]. In a cementitious medium, when the
capillaries dry out, the free adsorbed water moves into the capillary pores. The water
contained in the areas of hindered adsorption moves toward the capillaries thus relieving
disjoining pressure between solid particles, which leads to shrinkage.
2.2.3 Surface pressure
When a specimen of cementitious material dries, this makes the free surface energy of CSH
to increase and thus increase the hydrostatic pressure on the granular phase of the material.
This results in a contraction of the material which is seen as shrinkage at the material’s
scale.
The surface pressure or Gibbs-Bingham theory is able to explain half of the shrinkage
deformation that happens between 0 and 40% relative humidity [Roper 1966] and
principally on relative humidity levels lower than 10% [Young et al. 1986].
2.2.4 Factors influencing drying shrinkage
2.2.4.1 Cement
Shrinkage mechanisms are directly related to the structure of hardened cement paste. Since
the cement is one of the major components of the paste, it has a strong influence on the
drying shrinkage.
14
The type of cement influences the shrinkage of the paste. High alumina cement will not
have significant change on shrinkage magnitude compared to Portland cement but it
changes the kinetics, which leads to a higher rate of shrinkage. Type III Portland cement
will increase shrinkage magnitude compared to Type I Portland cement. Mineral additives
also increases shrinkage compared to Type I Portland cement [Neville 1995]. With
moderate dosage of fly ash or slag, the shrinkage can be as of 20% higher and can be
increased by 60% for high contents [Brooks and Neville 1992]. Addition of silica fume
tends to increase the long term drying shrinkage [Sellevold 1992].
2.2.4.2 Paste content
Since the shrinkage due to drying occurs in the paste, its volume in a concrete mix will
have a strong influence on the drying shrinkage of a given concrete mix. Numerous studies
show that, for different mixtures using the same constituents, the shrinkage magnitude will
be higher when paste volume is more important [Neville 1995, Bissonnette et al. 1999,
Hansen 1987].
2.2.4.3 Water to cement ratio
Water to cement ratio (w/cm) clearly influences the magnitude of shrinkage on cement
paste. However, this effect diminishes as the aggregate content increases. Figure 5 shows
the effect of the w/cm ratio for different aggregate contents [Ödman 1968].
0
400
800
1200
1600
0.3 0.4 0.5 0.6 0.7 0.8
Water / Cement Ratio
50
70
60
80
Aggregate Contentby Volume - percent:
Shrin
kag
e -
10
-6
Figure 5: Influence of water to cement ratio and aggregate content on drying shrinkage (Ödman 1968)
15
2.2.4.4 Aggregate
The amount and type of aggregate (Figure 6) have an important effect on drying shrinkage
of concrete. The aggregates oppose to the movement of the cement paste, thus reducing the
shrinkage magnitude of concrete. Since aggregates are opposing to the movement of the
paste, the modulus of elasticity, or rigidity of aggregates and their content are the key
parameters.
The size of the larger aggregates does not have a direct influence on shrinkage. However,
when using larger aggregates, it usually needs smaller paste content in the mix, thus
reducing the drying shrinkage [Neville 2000]. Shape and gradation curve of aggregates act
the same way as for the maximum size aggregate [Bissonnette 1996].
Figure 6: Effect of aggregate type on drying shrinkage (Troxell et al. 1958)
2.2.4.5 Admixtures
Air entraining admixture might have limited effects on concrete shrinkage. According to
Neville [Neville 1995], the use of an air entraining admixture does not influence directly
the shrinkage but since it creates air bubbles in the concrete, it tends to reduce the modulus
of elasticity and thus increases the shrinkage.
16
Superplasticizers increase the drying shrinkage but since they allow for lower water to
cement ratio, their effect is then counterbalanced [Neville 1995].
The shrinkage reducing admixtures (SRA) come from different sources, but they all acts on
the surface tension of water contained in the concrete voids and pore system. In normal
dosage, they can reduce the shrinkage up to 40% [Giroux 2006]. Other studies show that
the alcohol based SRA’s decrease the shrinkage from a range varying from 30 to 50%
[Bissonnette 1996].
2.2.4.6 Specimen size
Bissonnette et al. [1999] showed that the size of the specimen or the surface to volume ratio
does not have a significant effect on ultimate drying shrinkage value. The difference lies in
the kinetics of the shrinkage which will take more time for a larger surface or lower surface
to volume ratio.
2.2.4.7 Ambient conditions
The relative humidity has a direct influence on shrinkage. When the relative humidity of
the surrounding area of the concrete is lower than relative humidity of internal concrete
(almost always de case), loss of moisture occurs and shrinkage happens. The opposite is
also possible when the outside relative humidity is higher than of inside the concrete, it will
swell.
Neville [2000] states that for a saturated concrete (100% R.H.), it will start shrinking if the
relative humidity goes under 94% and will swell if humidity is 100% and when concrete is
exposed to wind.
2.3 Creep
2.3.1 General
The viscoelastic behavior of concrete in tension is particularly important when concrete is
subjected to restrained shrinkage, as is the case in bonded overlays. As already pointed out,
17
during the drying process, the contraction of the overlay is hindered by the generally much
stiffer and hygrometrically stable concrete support. If concrete were purely elastic, the
ultimate shrinkage, which for field concretes can range anywhere from 250×10-6
to
750×10-6
, would inevitably overcome the ultimate elastic strain, which typically lies
between 100×10-6
and 150×10-6
. Durable unreinforced bonded repairs would thus be
impossible to achieve. However, there are many examples of concrete overlays that have
performed very well for many years [Felt, 1956]. This has to be related in some way to
tensile creep.
To allow a rational analysis and a more appropriate design of concrete repairs and, globally,
of all elements where hygrometric and thermal strains are restricted, a better understanding
of the tensile creep behavior of concrete is required [Bissonnette, 1996]. Such an
understanding would also be helpful in the solution of many other practical problems, for
example in the design of water-retaining structures and other types of vessels, or to improve
the prediction of cracking in the tensile zone of reinforced concrete elements, or to provide
a more accurate evaluation of stresses in prestressed elements.
Metha and Monteiro [2006] define creep as a “gradual increase of strain with time under a
given level of sustained stress” and, correspondingly, stress relaxation as the “phenomenon
of gradual decrease in stress with time under a given level of sustained strain”.
According to ACI 209 [2005], creep strain represents the time-dependent increase in strain
under sustained constant load taking place after the initial strain at loading. Experimentally,
it is obtained from the load-induced strain by subtracting the initial strain. The creep strain
may be several times greater than the initial strain.
Creep strain may be subdivided into a drying and a nondrying component, termed drying
and basic creep, respectively.
Basic creep is the time-dependent increase in strain under sustained constant load of a
concrete specimen in which moisture losses or gains are prevented (sealed specimen). It
18
represents the creep at constant moisture content with no moisture movement through the
material, and is consequently independent of the specimen size and shape. To determine
basic creep, it is necessary to measure the deformations of a set of sealed specimens under
constant load and to determine the total strain and, in parallel, deformations of companion
sealed, load-free specimens. The basic creep deformations are determined by subtracting
the strains recorded on the latter from the strains recorded on the loaded specimens.
Drying creep is the additional creep occurring in a specimen exposed to the environment
and allowed to dry. As it is caused by the drying process, drying creep depends on the size
and shape of the specimen. Three sets of specimens are required to determine the drying
creep: a loaded set that is allowed to dry to determine the total strain, a loaded set of sealed
specimens to determine basic creep, and a load-free set at drying to determine the total
shrinkage strain
In most engineering applications, these two types of creep are considered to be additive and
their algebraic sum is generally referred to as the total creep.
2.3.2 Creep mechanisms
Different theories were submitted in order to give explanations of creep phenomenon. None
of them were able to clearly explain all of the aspects of tensile creep of cementitious
materials but they all agree that creep is a paste-scale phenomenon [Bissonnette 1996]. It is
possible that tensile creep is a combination of some of the proposed theories.
2.3.2.1 Water seepage
The water seepage theory [Lynam 1934] suggests that the hydrated cement paste is mainly
formed of a rigid gel, which consists of a solid phase and a viscous phase (water). When
pressure is applied on the viscous phase, water is expelled out of the pores and the viscous
phase is then relieved from initial stress, which is redistributed toward the solid phase.
Later on, Seed and Lea et al. came up with more detailed explanation on water seepage
theory [Bissonnette 1996]. They stated that the application of an external load changes the
19
vapor pressure equilibrium, thus changing pores’ water content. Changing pores’ water
content makes the paste to contract and leads to a change in volume of the concrete.
The water seepage theory can partially explain the creep deformation and experimental
results prove this theory valid. However, some phenomena remain unexplained. One of the
main objections to the water seepage theory is expressed by Neville [1960]. According to
Neville, the water loss of loaded specimen is slightly over water loss of unloaded specimen.
Other works [Neville 1995] states that water may travel from smaller pores to larger pores
without necessity for external migration and having the same effect on shrinkage.
2.3.2.2 Viscous shear
The first theory related to viscous shear came from Thomas in 1937 [Thomas 1937]. In his
theory, Thomas considered two phases which are the viscous phase (paste) and the inert
phase (aggregate). As for any viscous material submitted to load or stress, it starts flowing.
Since Thomas considers the cement paste as the viscous phase in concrete, it also flows
when load is applied: then the load is gradually transferred to the inert phase. As the stress
is transferred to the inert phase, the remaining stress on pastes decreases and thus the rate of
concrete creep diminishes, which is consistent with time dependence of creep observed.
In 1960, Hansen came up with his theory on viscous shear. According to Hansen,
movements happen at contact points (welding points) between gel areas. Ruetz [1968]
proposed a theory similar to Hansen’s in which the movements are strictly based on shear
between gel areas.
The viscous shear theory partially explains deformations due to creep but, as water seepage,
it cannot explain it all. There are experimental results showing that creep deformation is
linked to the previous deformation or deformation history [Illston 1965, from Bissonnette
1996]: creep cannot be only a viscous phenomenon.
20
2.3.2.3 Micro-cracking
Micro-cracking may be another source of creep that can explain up to 10 to 25% of the total
time-dependent creep deformation when concrete is subjected to high levels of stress
[Meyers 1967].
2.3.2.4 Combined theory
None of the creep mechanisms aforementioned can explain solely the total creep
deformations. Some researchers have proposed various theories combining different
mechanisms. ACI Committee 209 [1982] has proposed an explanation combining four
mechanisms. The first mechanism comes from the viscous shear theory and stipulates that
the viscous shear is caused by shearing of the lubricated gel particles. The second
mechanism is a settlement due to water seepage.
2.3.3 Factors influencing creep
2.3.3.1 Cement
Since a creep phenomenon is solely related to the cement paste, it would be expected that
the type of cement used plays an important role on creep intensity. According to Neville
[1995], the type of cement affects creep as it changes the mechanical properties of concrete,
which has an important role on creep strain. As for example, comparing the gain of strength
of concrete made from different cement, the creep results would be sensibly the same
[Neville and Kennington 1960].
2.3.3.2 Paste content and aggregates
As stated previously, a creep phenomenon is almost entirely related to cement paste.
Increasing the volume of paste in a concrete mix will lead to an increase of creep strain
[Bissonnette 1996, Boily 2003]. According to Brooks [1989, from Bissonnette 1996], creep
of concrete is only 10% to 15% of paste’s creep. As for the drying shrinkage, the
aggregates tend to oppose to the paste deformation, thus leading to a lower creep strain.
Since the aggregates have an influence on creep, their properties also do. As for example,
21
for same aggregate content, a mixture with a higher modulus of elasticity will therefore be
the most effective in reducing creep.
2.3.3.3 Admixtures
Water-reducers and superplasticizers (such as lignosulphonate based admixtures) are found
to increase the basic creep and this effect is more pronounced under drying conditions
[Morgan 1975]. Usually, the lignosulfate-based plasticizers tend to increase creep more
than of carboxylic-acid-based plasticizers [Brooks and Neville 1992]
2.3.3.4 Relative Humidity
The relative humidity of the air surrounding the concrete is one of the most important
factors influencing creep strain [Neville 1995]. Figure 7 [Troxell et al. 1958, from Neville
1995] presents experimental results of concrete specimens submitted to different relative
humidity. All specimens were cured in water fog for 28 days and were then loaded and
stored at different relative humidity levels. At 100% R.H., creep strain of specimen on long
term basis is half of the creep strain observed at 70% R.H and 2/5 of the long term value
observed in 50% R.H. conditions.
Figure 7: Effect of relative humidity on creep (Troxell et al. 1958)
2.3.3.5 Stress level
There is a direct proportionality between stress/strength ratio and creep. For stress/strength
ratio up to 50% of the ultimate tensile strength, the relation is linear and applicable for both
22
compressive and tensile creep [Bissonnette 1996]. Figure 8 shows the linearity of relation
between stress level and creep on concrete specimen [Neville 1995]
Figure 8: Effect of stress level on creep (Neville 1995)
There is also a stress level over which the specimen will fail (break apart). Common stress
levels for creep failures are as following [Domone, 1974, Al-Kubaisy et al. 1975]:
85% for sealed specimen (basic creep only),
75% for saturated conditions,
60% for drying conditions.
2.4 Concrete sensitivity to cracking testing methods
There is a broad range of tests with different procedures or specimen size that try to
evaluate the sensitivity to cracking of a repair material. There are two major approaches for
evaluating the sensitivity to cracking of a repair material. The first one is based on the
superposition principle and characterization tests for materials. It consists in separating all
the sources of volumetric deformation and evaluating them with separate tests: the results
are combined to get the net deformation. The other way to proceed, which relies on
material behavior test methods, consists of using test methods that gives the total strain of a
specimen submitted at the same time to all sources of volumetric changes.
23
2.4.1 REMR box test method
This test was developed during the “Repair, Evaluation, Maintenance and Rehabilitation”
(REMR) research project. It consists of a precast concrete slab with a cavity to be filled up
later with the repair material to be tested (Figure 9). The slab represents the concrete
structure and the cavity simulates a concrete surface that had been prepared for repair
[Emmons and Vaysburd 1995]. The depth of the cavity is of 75 mm which is representative
of a typical repair and it allows using coarse aggregates without introducing
disproportionate restraint. To ensure proper restraint, transverse grooves are located in the
bottom of the cavity. Typical concrete (34 MPa and 0.42 w/cm) is used to manufacture the
precast slabs [Emmons et al. 1998]. Before repair material is placed into the cavity, the slab
shall be aged until drying shrinkage is completed, which usually takes up to four months.
After the repair material is cured, it is exposed to drying and the repair is monitored for
cracking.
This test can be carried out in different test conditions. However, the size of the specimens
tends to make it more complex in laboratory conditions with controlled temperature and
relative humidity.
24
Figure 9: REMR Box test specimen’s geometry (after Emmons et al. 1998)
25
2.4.2 Baenziger block test method
The Baenziger Block is a prefabricated non-reinforced concrete slab with a cavity. It is an
in-house test initially developed by Sika. The support slab is 1050 mm by 350 mm and the
thickness of the repair is 30 mm for 500 mm length and then varies from 30 to 60 mm for
the rest of 650 mm. Figure 10 is a schematic view of the Baenziger block geometry
[Gillespie 1999].
Figure 10: Schematic representation of the Baenziger block
The test consists of filling the cavity with the repair material to be tested and monitor it for
cracking and delamination. This test is intended to compare materials together and it can be
carried out in different conditions (field or laboratory conditions). It has the advantage of
being small enough to be easily carried out in a laboratory and is easier to be handled.
2.4.3 Ring test
The ring test method concept is about 70 years old. It has been developed in order to
evaluate the cracking susceptibility of cementitious materials submitted to restrained
shrinkage. The ring test is commonly used for evaluation of cracking potential of a
cementitious repair material because it provides quick information on a comparison basis
for repair materials and is simple to run. The test consists of a steel ring that acts as the
restraining support, around which concrete is poured. When drying, concrete tends to
shrink and the movement is restrained by the steel ring, which generates tensile stresses in
the concrete. When the stress reaches the tensile strength of concrete, a crack appears. In
26
most tests, the moment when the first crack appears is the most important parameter
measured. Also, the evolution of the crack’s width can be monitored over time. In some
cases researchers seal different surfaces to limit drying to specific surfaces like
circumferential or top and bottom [See et al. 2003, Moon and Weiss 2005].
Depending on the relative size of the steel ring and the concrete ring, the test will be more
or less severe. When the steel ring provides too much restraint, the test is as severe that it is
nearly impossible to discriminate the sensitivity to cracking of the tested materials. This
type of ring configuration should not be used.
In 1939, Carlson and Reading [1988] conducted ring test. In their test setup, the steel ring
had an internal diameter of 125 mm and an external diameter of 175 mm. The height of the
specimen was 34 mm. The concrete wall thickness cast around the steel ring was 25 mm.
Drying was only possible from the circumferential surface. The top and bottom surfaces
were sealed. The tests were carried out in three different conditions; 25, 50 and 75 percent
of relative humidity. Carlson and Reading [1988] found the ring test valuable, providing
“useful information on relative crack resistance” and they believe the ring test should be
investigated.
Later on, Couthino [1959] carried out tests on pastes and mortars to evaluate their
resistance to restrained shrinkage. His research’s objective was to evaluate the influence of
the type of cement on the cracking sensitivity of cement based materials. For this study, the
specimens used had a cross-section of 25 x 25 mm. For results interpretation, Couthino
developed his own index system; the quotient of maximum stress. For specimens that
cracked before 90 days of drying, the quotient is calculated at time of cracking. For
specimens that did not cracked after 90 days of drying, the quotient is calculated after 90
days of drying. The quotient of maximum strength is calculated as follow:
Quotient of maximum stress90
90
daysatringedunrestraininstrainshrinkage
ringconcreterestrainedinstress
9090
90
27
The quotient of maximum stress is an indicator of relaxation due to creep. Large values for
the quotient of maximum stress indicate small amounts of relaxation. Conversely,
significant relaxation is represented by smaller values.
Almeida [1990] used the ring test to investigate the crack resistance of high-strength
concretes with compressive strengths between 60 and 100 MPa. The ring specimens had an
external diameter of 810 mm and the concrete cross section was of 80 x 80 mm. These
dimensions were selected in order to eliminate theoretical considerations on stress
variations within the concrete ring. The restraining ring was made of aluminum.
During the test, strain of the aluminum restraining ring was monitored and the time of
concrete failures can be determined from the variation in strain. At the cracking time,
mechanicals properties were determined from companion specimens.
Almeida used the methods proposed by Couthino for determining the quotient of maximum
stress. No simple relationships were found between cracking and the major parameters
typically used such as strength or shrinkage.
In 1996, the Transportation Research Council issued a report [1996] in which studies on
cracking in bridge decks and concrete behaviour in ring test were analyzed. Rapidly after
publication of this report, AASHTO implemented a provisional standard [AASHTO 1999]
that used the ring test for helping to quantify a material propensity for cracking.
In 2001, Attiogbe et al. [2001] reported on studies on cracking potential of concrete and
mortar using ring specimens. Attiogbe’s research unquestionably contributed to the
development of the presently adopted ASTM test method. However, there are no
experimental studies reported on correlation between test results using ring test and in-situ
performance of repair materials. In other words, how reliable is this method for predicting
cracking behaviour of repair materials in real life repairs?
28
2.4.4 German angle test
The German Angle test is a linear restraint test method and it consists of filling a steel angle
with a repair material (Figure 11) and following for cracking. In this test, the steel angle
acts as the restrictive substrate (as in a cementitious repair material to concrete repair),
which represents the substrate in a repair. When the repair material dries out and shrinks,
the movement is restrained by the steel angle and it results in tensile stresses in the repair
material.
Figure 11: German angle test setup (Emmons and Vaysburd 1995)
The surface preparation for the steel angle consists of sandblasting inside to remove any
contaminant such as oil or greasy deposits and rust. It also provides a better grip for the
repair material.
The German Angle test was developed by the Technical Academy, Aachen, Germany
[Emmons and Vaysburd 1995]. This test has been adopted as the Technical Test Regulation
for concrete substitution systems made of cement mortar/concrete with a plastic additive by
the Highway Construction Department of the Federal Ministry for Transport [Emmons and
Vaysburd 1995].
The test should be run for a period of 90 days: any crack should be noted and its opening
measured with a precision of 0.02 mm. The number of cracks, the average and maximum
crack opening, the time of cracking and any bond failure should be reported. The maximum
29
crack opening for an acceptable repair material is set to 0.1 mm. Of course a bond failure
automatically leads into rejecting the candidate repair material.
2.4.5 SPS plate test
The test specimen consists of a 50 by 100 by 1320 mm repair material beam cast over a 1.5
mm thick steel plate (Figure 12 and Figure 13). The specimen is clamped at one end to a
steel channel and the other end is free to move upon drying shrinkage and creep
deformation. The end beam deflection is monitored using a precision caliper.
The SPS Plate test was used to evaluate the net deformation due to restrained shrinkage
[Emmons et al. 1998]. Compared to other characterization tests, the SPS Plate test has the
advantage of evaluating the net deformation of the repair material, taking in account for
differential and restrained shrinkage and creep effect. According to Emmons et al. [1998],
if a material shows a tip deflection less than 0.25 mm at 28 days, it will be relatively
resistant to restrained shrinkage cracking. Emmons et al. [1998] have also found some
correlation between the SPS Plate Test and REMR Ring test method.
[in]
Figure 12: Shcematic of SPS plate test
30
Figure 13: SPS plate test specimens
2.4.6 Laval beam deflection test
The Laval Beam Deflection test [Bissonnette and VonFay 2001] consists in a 50x100x1000
mm repair material beam cast over an epoxy coated steel plate. The test consists of
monitoring the mid-span deflection as a function of time for at least 6 months. Curing of
the specimens is usually of 72 hours but can be adapted to materials requirements. Figure
14 presents the Laval beam deflection test specimen. All surfaces of the beam, except the
top one, are paraffin-wax sealed, which creates a differential drying shrinkage along the
depth of the specimen, which makes it to deflect. This test reproduces the effect of drying
shrinkage on a slab on grade or a concrete repair when the test is used with a steel plate
underneath the beam. This test allows characterizing the moisture gradient and the kinetics
of the drying. More important the drying shrinkage is, more important the mid-span
deflection will be.
Figure 14: Laval BD Test Specimen
31
2.4.7 Thoro system test
The Thoro system products test consists of a small slab (15¾ x 19¼ inches) with a cavity
and exposed rebars as shown in Figure 15 [Emmons and Vaysburd 1995]. It represents a
small typical repair with restraint shrinkage for the repair material. The cavity is filled with
repair material and the slab is monitored for cracking.
Figure 15: Thoro System (Emmons and Vaysburd 1995)
The literature presents a variety of in-house test methods that allow testing of repair
materials. Most of them are mainly used for comparison between materials but do not offer
acceptance criteria for resistance to cracking. Among the test methods for cracking
sensitivity presented in this section, only the Thoro System has not been selected for testing.
33
3 Experimental set-ups and procedures
3.1 Scope of test program
3.1.1 Phase I
Any test method(s) to be developed should cover the determination of the cracking
tendency of a repair material under real-life repair conditions. The procedure is
comparative and not intended to determine possible cracking in a specific type of structure
and specific location of the repair in a structure. Actual cracking in service depends on
several variables including design details, repair methods, degree of restraint, substrate
surface preparation, construction practices, and environmental factors. The method(s)
should be applicable for evaluating the sensitivity of repair materials to cracking, and for
the selection of crack-resistant repair materials.
The scope of Phase I of the project was as follows:
To select the optimum geometry of the experimental repair configuration. The size
and weight of the specimen and the cavity in it has to be sufficiently small to allow
for easy handling in the laboratory, yet being representative of an in-situ repair.
To perform characterization tests on repair materials used in testing the
experimental repair configuration. The basic properties such as strength, shrinkage,
flexural creep, modulus of elasticity were determined.
To perform several standard and non-standard shrinkage and tendency-to-cracking
tests on repair materials.
To evaluate possible correlation between various shrinkage tests and the
experimental repair test (Box test and/or Baenziger Block test).
3.1.2 Phase II
Evaluation of the cracking tendency of a repair material is a complex task. However it is
important to have tests that cover the determination of the cracking sensitivity of a
cementitious repair material. The scope of the Phase II was as follow:
34
To evaluate the possible correlation between selected tests from the Phase I
(Baenziger block test and ASTM Ring test) and real size repair on reinforced
concrete slab
To evaluate the possible correlation between free shrinkage test and in-situ
experimental repair
To perform characterization tests of repair materials used in the experimental
repairs and laboratory tests
To evaluate the possible correlation between tensile pull-off and shear bond test
methods.
3.2 Description of the repair materials
3.2.1 Phase I materials
For the first phase of the research program, three different repair materials were selected.
There were a conventional concrete, a Portland cement mortar and a commercial polymer
modified mortar. These three materials cover a wide range of shrinkage behavior and
should be quite different in cracking resistance.
The materials used in this project were all pre-bagged and the admixtures were already in
the bags as powdered admixtures. When batching, only water was added to the mixture. All
three mixes were supplied by the same company.
3.2.1.1 Concrete mix
The concrete mix design is based on a standard concrete mix commonly used at Laval
University Concrete Research Center (CRIB), which is also used as a repair concrete. It is a
0.43 water to cement ratio and contains Type-I cement, natural sand and 10 mm granite
gravel as coarse aggregates. Air entraining admixture and superplasticizers were included
in the pre-bagged packaging. The mixture composition is presented in Table 1.
35
3.2.1.2 Mortar mix
The mortar mix is basically the concrete mix without the coarse aggregates. Its water to
cement ratio is 0.40 and it is the same cement and sand as the concrete mix. It is designed
to be flowable and easy to pour without segregating or excessive bleeding. This mix is also
commonly used at Laval University laboratory. The mix design is presented in Table 1.
3.2.1.3 Polymer modified mortar mix
A commercially produced material was selected for this study. It is a Portland cement
mortar mix which contains polymer. It is not a fast setting material, but it stiffens rapidly,
so it has to be placed within 20 minutes, but can be shaved and finished within an hour. The
water to cement ratio and mixture composition is not available for this material since it is a
proprietary material.
Table 1: materials mix design
Constituents Concrete
(kg/m3)
Mortar
(kg/m3)
Polymer
mod. Mortar
(kg/m3)
Water 191 270 -
Cement 441 675 -
Sand 837 1340 -
Stone 900 - -
Air entraining admixture (g/kg of cm) 0.26 0.07 -
Superplasticizer (g/kg of cm) 14.5 18.9 -
3.2.2 Phase II materials
The phase II of the study used eight materials; four mortars and four extended mortars
(concretes). The coarse aggregate was added manually when mixing. The rate of addition
of stone is presented in Table 2. All these materials were pre-bagged and supplied by three
36
different concrete material manufacturers. The materials were identified by numbers from 1
to 4 and the suffixes M and C are used for Mortar and Concrete (Extended Mortar)
respectively.
Table 2: Rate of aggregate addition for extended mortars
Material Rate of addition
(kgagg/kgdry mat)
1 0.6
2 1.0
3 1.0
4 0.5
3.3 Description of the test conditions
All the tests performed in laboratory environment were conducted in controlled
environment, which is kept at 23 °C ±2°C and 50 ±2 % relative humidity. For all
specimens, the initial curing (before stripping the forms) was of 24 hours under wet burlap
and polyethylene sheet; after demolding, the cure was made in a fog room at 23 °C and
100 % relative humidity.
In field tests the temperature and relative humidity varied according to the weather
conditions.
3.4 Description of the test procedures
3.4.1 Mechanical properties
The following tests were used to characterize the mechanical properties of the materials
used for repair and testing:
Compressive Strength – American Society for Testing and Materials ASTM C 39
“Standard test method for compressive strength of cylindrical concrete specimen.”
37
Compressive Strength – American Society for Testing and Materials ASTM C 109
“Standard test method for compressive strength hydraulic cement mortars (using 2-
in. or 50-mm cube specimens).”
Modulus of Elasticity – American Society for Testing and Materials ASTM C 469
“Standard test method for static modulus of elasticity and Poisson’s Ratio of
concrete in compression.”
Splitting Tensile Strength – American Society for Testing and Materials ASTM C
496 “Standard test method for splitting tensile strength of cylindrical concrete
specimens.”
For the Phase I of the project, the tests were performed at 3, 28, and 56 days. Three cylinder
specimens were used for each test: 75 by 150 mm (3 by 6 inches) for concrete, and 25 by
50 mm (1 by 2 inches) for mortars. For the Phase II, 100 by 200 mm cylindrical specimens
were used. The tests were performed at 7 and 28 days at Laval University and BOR. At the
US Navy, the tests were performed at 56 days.
3.4.2 Free drying shrinkage
The free drying shrinkage tests were conducted in accordance with the ASTM C 157
modified test method (ACI 364-3R) “Standard test method for length change of hardened
hydraulic cement mortar and concrete”. The modification proposed by ACI 364-3R is
concerning the curing period, which was of three days instead of 28 days as proposed by
the ASTM C 157 standard.
In Phase I, for the concrete mixture, 75 by 7 by 285 mm specimens were cast. For the
mortar mixtures, 25 by 25 by 285 mm specimens were cast. Two types of measuring
comparators were used to obtain the readings. The first one is described by ASTM C 157
(Figure 16) was used at Université Laval for each type of specimens, while it was used only
for mortars at the BOR. The second type is a comparator using a mobile plate (Figure 17).
This type of apparatus was used at the BOR for concrete specimens.
38
Figure 16: Standard comparator to measure drying shrinkage strain
Figure 17: Comparator with mobile plate to measure drying shrinkage strain
In the Phase II, only 75x75x285 mm specimens were used for drying shrinkage tests, either
at Université Laval and the BOR. At Université Laval, the vertical comparator was used
and at the BOR the comparator with the mobile plate was used.
39
3.4.3 Flexural creep test
Bissonnette [1996] developed a system to evaluate the tensile creep of cementitious
materials (Figure 18). The principle of the testing apparatus is to apply a constant tensile
load on a series of three specimens and record the deformation of each specimen as a
function of time. In order to separate the drying shrinkage from the creep, non-loaded
specimens (identical to the tensile creep specimens) are monitored with the same
equipment. Thus, the net creep can be obtained by subtracting the drying shrinkage
deformation from the deformation due to creep. Since the direct tensile creep apparatus is
complex and costly to operate, a new test was developed. The Flexural Creep test is a rather
recent development to evaluate creep potential of a cementitious material. This test is not
standardized yet but follows a strict procedure to ensure reliable results from one batch to
another and even from a concrete mixture to another.
Figure 18: Tensile creep system (after Bissonnette 1996)
The flexural creep test is simple compared to traditional tensile creep evaluation systems. It
consists of a simply supported beam, on which two loads of 20 kg are applied at equal
distances from mid-span, as shown in Figure 19. The mid-span deflection is monitored with
a special device (Figure 20). The measuring tool is the same as the one used for the Beam
Curling test, except that the distance between the seating pins is not the same. For the
Flexural Creep test, the distance between the pins is 762 mm.
40
Figure 19: Sketch of the flexural creep test setup
Figure 20: Flexural creep test specimens and deflection measuring device
To avoid lateral moisture gradient, all the lateral sides of the specimens were sealed with
wax after the specimens were taken out of the molds. Once the specimens were sealed, they
cured in a fog room (100% R.H.) for three days and then the test started. The reading
schedule was tight for the first six hours. When the test started, a reading was taken every
five minutes for the first fifteen minutes. Between fifteen minutes and an hour, the readings
20 kg loads
50
100
1250
1150
800
[mm]
762
Top view
Side view
41
were taken every fifteen minutes. Readings are then taken hourly until the sixth hour. Once
the test started (first six hours) the schedule was the same as drying shrinkage; once a day
for the first week, once a week for the first month and every month after that.
Two pairs of beams were used for this test. The first pair was the loaded specimens, and the
second set was the companion specimens, which were not loaded. When the test started,
three loading and unloading cycles were carried out and the deflection was recorded for
each state (loaded and unloaded) in order to access the instant modulus of elasticity.
3.4.4 Restrained shrinkage tests
Different restrained shrinkage tests were performed during Phase I to evaluate cracking
sensitivity of repair materials. These tests are described in the following sub-sections.
3.4.4.1 Box Test
Three different box sizes were selected for evaluation in this task. The geometry of the
specimens is presented in Figure 21 and Table 3 shows the dimensions of the box test
specimens. The “Large” size precast concrete box with a cavity of 45x1900x75 mm
corresponds to that used in the REMR study [Emmons et al. 1998]. The overall geometry of
the box was revised based on its rigidity analysis. All boxes were required to be sufficiently
rigid to eliminate any deformations caused by handling and/or by stresses generated by the
restrained shrinkage of the repair material. For the three different sizes of box test, the
relative inertia ratio (Irepair / ITotal) is constant and was set to 0.3. The “Medium” size precast
concrete box has a cavity of 375x1400x75 mm and the “Small” size precast concrete box
has a cavity of 300x900x75 mm.
42
Figure 21: Box test specimen
Table 3: Dimensions of the box test specimens
Dimensions Box dimension
Small Medium Large
Lbox (mm) 1 000 1 500 2 000
Wbox (mm) 500 625 750
Hbox (mm) 175 175 175
Lrepair (mm) 900 1 400 1 900
Wrepair (mm) 300 375 450
hrepair (mm) 75 75 75
Volume of repair (liters) 20.3 39.4 64.1
Relative Inertia (Irepair / ITotal) 0.3 0.3 0.3
43
The slabs were manufactured 10 months prior to their filling with a repair material to allow
volumetric stability to take place. A 35 MPa concrete mixture was used to cast the boxes
and the air content ranged from 4 to 7%. The mix design is presented in Table 4. The
precast slab is reinforced with a wiremesh (WWF 4x4) located at 50 mm from the bottom
of the slab.
Table 4: Box test concrete mixture design
Constituents Kg/m³
Water 180
Cement 422
Sand 670
Coarse aggregate (5-14 mm) 1026
Water Reducer (ml/kg of cement) 2.38
Air entraining admixture (ml/kg of cement) 0.25
Surface preparation of the cavity in slab was the same for every slab size (small, medium
and large). The first step was to sandblast the cavity of the slab with grade 24 sand to obtain
a rough surface and provide an appropriate bonding between the repair material and the
substrate. This operation was usually completed after the concrete had reached 28 days of
age. After sandblasting was completed the cavity was cleaned with water (high pressure)
and dried out. Sealer was applied on all surfaces of the slabs. The sealer was a Siloxane
(40% solids) and the purpose of this was to prevent moisture movement in the slab.
In order to avoid any influence of surface preparation and any water transport between the
repair material and the precast slab, an epoxy based bonding agent was used. A thin layer of
water based epoxy was applied on the surface of the cavity a few minutes before pouring
the repair material.
The materials were poured in two equal layers and each layer was vibrated with a 19 mm
vibrating needle. The surface was leveled with a wooden straight edge and trowel finished.
Once the repair materials were cast, they were cured under wet burlap for 72 hours. The
44
repairs were monitored for cracking on a regular basis. Every crack was drawn on
transparencies and the crack openings were also monitored for a period of two years.
3.4.4.2 Baenziger block test
In addition to the “Box test”, the Baenziger block test [Gillespie, 1999] was also used in
this task (Figure 10). The surface preparation is exactly the same as for the Box Test. The
surface of the slab is sandblasted; a Siloxane sealer is applied and epoxy bonding is used
prior to filling the slab with the repair material. The forms used when performing the repair
were quite simple and consisted of a plywood angle strapped with carriage binders. The
concrete was poured in a single layer and vibrated using a 19 mm vibrating needle and
trowel finished. The repair material was cured under wet burlap for a period of 72 hours,
after which it was exposed to drying. Time to crack occurrence was recorded and all cracks
were drawn on transparencies with their opening for two years of exposure to field
conditions.
3.4.4.3 German angle test
The German angle test (Figure 11) was carried out in both field and laboratory conditions.
In order to improve the bond between the mold and the repair material, epoxy was applied a
few minutes before placement of the repair material. The material was cured for three days
and then exposed to drying. The test was monitored for cracking and delamination for a
period of 90 days. Cracks were drawn on a transparency film and crack opening was
recorded.
3.4.4.4 Ring test
For the first phase, two different ring test geometries were selected for evaluation. The first
configuration was the same as used in the REMR Project [Emmons et al. 1998] as shown in
Figure 22. The inner metallic ring is mounted on an acrylic sheet base plate and the external
steel mold was secured onto the inner ring using a flat bar. The repair material was poured
in two equal layers and each layer was vibrated for 10 seconds and the surface was trowel
45
finished. The rings were moist cured for 72 hours and then put in a 23°C and 50% R.H.
room for the duration of the test. Time to first crack and crack opening were monitored.
Figure 22: REMR ring test configuration
The second configuration used was based on the ASTM C 1581 standard. The difference
between the ring test configuration used for this project and the ASTM C 1581 standard is
the steel ring diameter and the time of moist curing. The ring used for this project has an
inner diameter of 278 mm compared to 305 mm for the ASTM test and the moist curing
was of 72 hours in order to be able to compare the results with the REMR ring test
configuration. In both configurations (CREEP project and ASTM C 1581 standard), the
wall thickness of the steel ring is 12.5 mm. The configuration of the ASTM-like ring test is
presented in Figure 23. As for the REMR configuration, the steel ring is secured to a non-
absorbent base plate. However, the system used to secure the mold is different. The mold
was made of high density polyethylene and it was fastened directly to the base plate using
steel angles. Method for placement of material in the mold was the same as for the REMR
ring test configuration. In the ASTM-like ring test, semi-conductive stain gages were
placed on the inner face of the ring in order to monitor the strain of the steel ring due to
pressure exerted by concrete shrinkage. Time to first crack and strain gages were
monitored.
46
Figure 23: ASTM-like ring test configuration
The strain gages allowed for evaluation of the average tensile stress in the concrete at any
moment during the test using the following equation obtained from the mechanic of
materials [See et al. 2003].
610)()( thr
hrEt s
cis
sics
avg
Where:
σavg(t) = Average tensile stress in concrete as a function of time (MPa)
Es = Steel modulus of elasticity (MPa)
ric = Internal radius of concrete (mm)
ris = Internal radius of steel (mm)
hc = Concrete wall thickness (mm)
hs = Steel wall thickness (mm)
εs(t) = Steel strain as a function of time (µm/m)
47
3.4.4.5 SPS plate test
The test configuration consists of a 50x100x1320 mm repair material beam cast over a 1.5
mm thick steel plate and two threaded rods are embedded at one end of the beam. In order
to allow for a good bond between the concrete and the steel plate, the steel plate was lightly
sandblasted and epoxy resin was applied. Before the epoxy has set, sand was evenly
distributed over the epoxy and the excess was wiped off. The beam is clamped over 150
mm from one end, which gives a free cantilever of 1170 mm long. The specimen was
removed from the mold after 24 hours and cured in a fog room 48 hours. At the end of
curing, the specimen is clamped onto a steel channel using the embedded threaded rods as
shown in Figure 13.
The test consists of monitoring the beam tip curling with a precision caliper. The test
started after 72 hours of moist curing. At this time, the specimen is clamped to the steel
channel using the two threaded rods embedded in the material and a steel plate on the top.
The readings were made every day over the first week, and then weekly during the first
month and once a month after that.
3.4.4.6 Laval beam deflection test
The Laval beam deflection test was conducted in order to find relationship between the
Laval Beam Deflection and the SPS Plate test. The specimens were cured in a fog room for
72 hours and then exposed to drying at 23°C and 50% R.H. Readings were made every day
over the first week, and then weekly during the first month and once a month after that. To
monitor the mid-span deflection, a custom made comparator was used (Figure 14). The
difference between the reference bar and the specimen was recorded and calculations were
made to express the results in terms of equivalent tip deflection for a beam that has the
same length than SPS Plate test specimen.
48
The calculations were made as follow:
R
LCL
8
2
and R
LEND
2
2
Where:
CL - Mid-span deflection
END - Beam tip curling
L - Length of span or cantilever arm
R -Radius of curvature
From these two equations: BDCLBDEND 4 and thus the beam tip curling of the Laval
BD test specimen is four times its mid-span deflection.
To compare the SPS Plate test and the Laval BD test, corrections were made for the
different length of the cantilever arm (assuming the beam of a longer span but same
curvature).
2
2
2
2
2
2
BDEND
SPSEND
BDEND
SPSEND
BDEND
SPSEND
L
L
RL
RL
2
2
4BD
SPS
BDCLSPSENDL
L
Where:
END-BD - Beam tip curling for the Laval Beam Deflection test geometry specimen
END-SPS - Beam tip curling for the SPS Plate test geometry specimen
CL-BD - Mid-span deflection of the Laval Beam Deflection specimen
LSPS - Length of cantilever arm in the SPS Plate test specimen
LBD - Span of the Beam Deflection specimen
R - Radius of curvature
49
3.4.4.7 Vertical beam curling test
The vertical beam curling test is similar to the SPS plate test as it consists of measuring the
tip curling of a specimen clamped on one extremity and the other one is free to move. The
differences between the SPS Plate test and the vertical beam curling test is the size of the
specimen and its position. Unlike the SPS Plate test, the vertical beam curling test specimen
is not submitted to gravity load and the curling observed is strictly due to the moisture
gradient. In some way this test could represent a mix of the Laval beam deflection and the
SPS plate test methods.
The specimen geometry and test setup are presented in Figure 24. The specimen is clamped
at one end. At the other end, a dial gage is mounted on a frame attached to the supporting
steel beam. The dial gage measures the specimen’s tip curling.
The specimens were cast and moist cured for 72 hours before tests started. The monitoring
schedule was similar as for the Laval beam deflection test and it lasted for a period of 15
months.
Figure 24: Vertical beam curling test configuration
50
3.4.5 Experimental repairs
3.4.5.1 Phase I experimental repairs
In order to compare the results obtained on sensitivity to cracking tests and small patch
repairs on a concrete structure, experimental repairs were made and monitored for cracking
on the AWTTS Pier at the US Navy – Port Hueneme CA. The AWTTS Pier is a facility
available at the US Navy base located in Port Hueneme, California and it consists of a
series of slabs resting on piles over sea water in the navy base harbor (Figure 25). The slabs
are made of a high strength concrete reinforced with carbon fiber strands and were in
service for over three years prior to be used for experimental repairs. The age of the slabs
allowed them to be stable from a volumetric standpoint.
Figure 25: AWTTS Pier – US Navy, Port Hueneme, California
To simulate small patch repairs on concrete structures, a total of 10 cavities were dug in the
substrate slabs. The size of these small repairs was the same as the size of the cavity on
Medium size box test specimens. For each repair materials used in the Phase I, three
cavities were used and there was one extra cavity.
The cavities were dug with a chipping hammer and the perimeter lines were saw-cut in
order to have sharp and strait edges. The surface preparation was competed by sandblasting
the cavities and cleaning them with high pressure water jet in order to remove any
remaining debris. As for Creep Box test and the Baenziger block test, an epoxy bonding
agent was used. The bonding agent was placed few minutes prior to fill the cavities with the
51
different repair materials. After placement, the materials were cured under wet burlap and
plastic sheets for 72 hours and then exposed to field conditions. The experimental repairs
were monitored on a regular basis for cracking and delamination. All cracks were sketched
and the crack opening was also reported.
3.4.5.2 Phase II experimental repairs
The field testing program included experimental repairs at both the BOR and US Navy
facilities. For each material, one experimental repair was done. The Phase II experimental
repairs consisted of overlays cast on prepared surfaces of reinforced concrete slab. The
overlays were 4.5 m long by 0.38 m wide and the thickness was of 75 mm, which is
representative of a typical repair. Figure 26 show the overlay configuration. The
experimental repairs were performed at two different locations: the US Bureau of
Reclamation (BOR) in Denver, Colorado, and at the US Navy facilities in Port Hueneme,
California.
Figure 26: Schematic view of the experimental repair set up
The surface preparation for repairs at the BOR was accomplished using shotblasting. It
allowed removing the thin layer of paste at the existing concrete surface, lightly exposing
the coarse aggregate (Figure 27). At the US Navy Facilities at Port Hueneme, hydro-jetting
was used, which exposed the aggregate to a larger extent than at the BOR facility (Figure
28).
After the surface preparation was completed, inspection of the concrete surface revealed the
presence of pre-existing cracks on both experimental sites. To diminish the possibility of
crack reflection into the repair and avoid non-uniform moisture absorbtion from the repair
mixture into the substrate, an epoxy bonding compound was applied to the substrate
52
concrete 20 minutes prior to repair material placement. For this purpose, SikaDur HiMod
32® was used.
The material mixtures were placed in one single layer to avoid cold joints and they were
consolidated using a vibrating needle and troweled finished. Each material was cured
according to manufacturer’s instructions.
The monitoring for cracking was conducted every day for the first week and once a week
for the first month. After a month, readings were taken every two months.
Figure 27: Surface preparation at the Bureau of Reclamation (BOR)
Figure 28: Surface preparation at the US Navy
53
4 Results and discussion - Phase I
4.1 Introduction
This chapter presents the results and the discussion of the first phase of the project. These
results were obtained from tests performed on three typical repair materials (concrete,
mortar and polymer modified mortar). These three materials were selected for their
different volumetric behavior. The tests were run in laboratory conditions and/or field
conditions. For each test, the condition is specified.
The presentation of the results debuts with the results of commonly used characterization
tests; fresh material properties (slump, air content and temperature) and mechanical
properties. The mechanical properties include the compressive strength, the splitting tensile
strength and the modulus of elasticity.
The second set of results presented in this section is the characterization of the
deformational behavior of the materials which consist of the evaluation of drying
shrinkage, the tensile creep potential using the flexural creep test method and other non-
standardized test methods.
The last set of results presented for the first phase is related to the restrained shrinkage tests
carried out during the first phase of the project.
4.2 Characterization
The characterization tests measure the common properties one wants to evaluate for a given
concrete or mortar. For fresh concrete properties and mechanical properties, the results are
an average of the results obtained from several batches. No major differences were noted
from one batch to another for the same material.
54
4.2.1 Fresh material properties
The fresh concrete properties consist of the slump test, the air-entrainment test and the
temperature of the concrete at the end of the mixing sequence. The values are an average of
the different batches, but for all batches of the same material the tests results were merely
identical. These results are presented in Table 5.
Table 5: Fresh material properties
Mix
Properties Concrete Mortar P-Mortar
Slump (mm) 200 235 235
Air content (%) 10.5 9.0 6.5
Temperature (°C) 18.5 21.5 20.5
The materials were mixed following the same sequence for each material. Once the mixing
completed, tests on the fresh concrete were performed.
The first test performed was the slump test: all three materials showed a quite high slump
(over 200 mm). Initially, the concrete and mortar mixes were both designed for a 120 to
150 mm slump. The mixture determination and admixtures contents were adjusted at the
manufacturer’s plant. The mixing procedure was not identical to that of Laval University’s
laboratory. To mix the concrete and mortar at the production plant, an electrical drill with
an agitator fixed on its end was used. The mixing energy delivered by such a tool is very
low compared to the planetary motion mixer used and this could explain the difference
between the expected slump and what observed.
High slump is good for workability and it assures a good compaction of the concrete in the
forms. However it could also have side effects such as excessive bleeding, which can lead
to plastic shrinkage cracking. Since the materials were pre-bagged and the admixtures were
already in the bags, it was not possible to modify the mix design. However, for both
concrete and mortar materials used in this study, the high slump did not cause excessive
bleeding; all three mixtures behaved adequately, even for bleeding, segregation and setting
time. The materials were used as is after considering their appropriate behavior.
55
To easily pour the polymer modified mortar and avoid the formation of lumps, the P-
Mortar mix was batched with more water than specified by the manufacturer. This affects
the material behavior, but one must remember that the object of the study is not to evaluate
the materials themselves but rather to evaluate the different sizes of box test specimens.
Similarly, the air content was higher than expected for the concrete and the mortar mixes.
Once again, the admixture used was a powdered type, so it was not possible to modify it
once the material is bagged. The reasons for the higher values of air content are comparable
to that of the slump test. The mixing method is then assessed as responsible for the
difference. Verifications took place with the manufacturer to certify that the specified
amounts of admixture were actually incorporated in the bags. The increased slump also had
an influence on the amount of air entrained into the concrete during the mixing period.
The polymer modified (P-Mortar) is a proprietary material and no problem occurred during
placement.
4.2.2 Mechanical properties
The mechanical properties of the three materials are presented in the following tables.
The results of compressive strength tests after 28 days range from 30 to 45 MPa. The
polymer modified mortar revealed the weakest compressive strength. In the field condition
and most of the lab conditions, the mortar had a higher compressive strength compared to
the concrete mix. Overall, the compressive strengths obtained from the materials mixed at
the BOR were slightly higher than the ones mixed at Laval. Table 6 and Table 7
respectively show the 3 and 28 day properties. Tests were made on 75 x 150 mm
cylindrical specimens. For each test, three specimens were used. The mechanical properties
tests were made in accordance to the corresponding ASTM standards.
The results after 28 days range from 30 to 45 MPa. The polymer modified mortar revealed
the weakest compressive strength. In the field condition and most of the lab conditions, the
mortar had a higher compressive strength compared to the concrete mix. Overall, the
56
compressive strengths obtained from the materials mixed at the BOR were slightly higher
than the ones mixed at Laval.
Table 6: Three day mechanical properties
Test Units Batch Concrete Mortar P-Mortar
Compressive strength
ASTM C 39 (MPa)
Field 1 28.9 29.8 21.7
Lab 1 30.2 30.9 19.0
Lab 2 30.2 21.0 14.6
Lab 3 23.3 33.5 23.7
Lab
USBR
29.5 33.2 23.2
Splitting tensile strength
ASTM C 496 (MPa)
Field 1 3.2 3.1 2.4
Lab 1 3.2 3.2 2.3
Lab 2 3.3 3.5 1.8
Lab 3 2.3 3.0 2.1
Lab
USBR
2.9 2.8 2.2
Modulus of elasticity
ASTM C 469 (GPa)
Field 1 31.8 31.2 16.5
Lab 1 33.5 29.6 16.2
Lab 2 25.0 25.9 12.4
Lab 3 - - -
Lab
USBR
21.2 23.5 11.4
57
Table 7: 28 day mechanical properties
Test Units Batch Concrete Mortar P-Mortar
Compressive strength
ASTM C 39 (MPa)
Field 1 37.6 44.4 36.0
Lab 1 41.1 38.6 32.1
Lab 2 37.2 34.9 23.7
Lab 3 33.3 43.9 33.4
Lab
USBR 41.4 44.9 43.0
Splitting tensile strength
ASTM C 496 (MPa)
Field 1 4.1 3.8 2.5
Lab 1 3.8 3.5 2.1
Lab 2 3.7 4.1 2.7
Lab 3 3.0 3.1 3.6
Lab
USBR 4.0 - -
Modulus of elasticity
ASTM C 469 (GPa)
Field 1 39.9 44.5 18.2
Lab 1 43.9 41.0 18.0
Lab 2 29.2 26.0 17.4
Lab 3 30.1 40.5 18.8
Lab
USBR 24.7 29.9 16.6
The results of the split tensile test revealed values around 4 MPa for the concrete and
mortar mix, and 2.5 MPa for P-Mortar. At the bureau of reclamation, split tensile tests were
performed at 3 and 28 days only for the concrete. The results are close to that obtained at
Laval University.
The moduli of elasticity test results (BOR and Université Laval) are consistent with the
compressive strength test results observed. Concrete and mortar materials show comparable
modulus values, while P-Mortar material had the lower modulus of elasticity.
58
4.3 Deformational behavior characterization
4.3.1 Drying shrinkage
Drying shrinkage test was run in the same fashion for all three materials. Each material
underwent two testing conditions; in the laboratory (23°C and 50% R.H.) and in the field
conditions for the Creep Boxes and the Baenziger Block tests specimens. For the laboratory
conditions, the shrinkage test was run at both locations; Université Laval and the US
Reclamation (BOR).
4.3.1.1 Laboratory conditions
The drying shrinkage tests conducted in a controlled environment were done according to
ASTM C 157 modified standard. The modification to the standard test was done according
to ACI 364-3R. The results of drying shrinkage tests performed on the three repair
materials are presented in Figure 29.
As expected, the materials presented a wide spectrum of drying shrinkage magnitude which
ranges from 1000 µm/m to 3500 µm/m. The material that shrunk less was the concrete
followed by the mortar and finally the P-Mortar material which shown extensive drying
shrinkage. For the concrete and the mortar, considering similar mechanical properties, the
difference of shrinkage magnitude between the concrete and the mortar was noticeable
enough to assure differences in regards to sensitivity to cracking of these two repair
materials.
59
0
500
1000
1500
2000
2500
3000
3500
4000
0 56 112 168 224 280 336
ConcreteMortarPolymer mortar
Sh
rin
ka
ge (
µm
/m)
Elapsed time since initiation of drying (d)
0
500
1000
1500
2000
2500
3000
3500
4000
0 56 112 168 224 280 336
ConcreteMortarPolymer Mortar
Sh
rin
ka
ge (
µm
/m)
Elapsed time since initiation of drying (d)
a) Laval University b) BOR
Figure 29: Drying shrinkage (ASTM C157 mod.), laboratory conditions
Poston et all. [1998] proposed tentative requirements for repair materials. Among the
recommendations made, they suggested that the 28-day shrinkage should be no more than
400 µm/m and the maximum long term shrinkage should not be greater than two times the
28-day shrinkage value when the test is performed in laboratory conditions (23°C and 50%
R.H.). In Figure 30, the 28-day shrinkage is presented side by side with the long term
shrinkage value. Applying Poston’s criterion, all three materials would fail to conform to it.
The material that shows the lowest shrinkage magnitude (Concrete) is almost twice the
threshold limit established by Poston’s criterion. For the ultimate shrinkage criterion, the
situation is different. As shown in Table 8, all materials are satisfactory. However, applying
the criteria, none of these three materials would have been selected because they failed to
comply to the short term criteria.
60
0
500
1000
1500
2000
2500
3000
3500
4000
Concrete Mortar P-Mortar
28 days
Long Term
Dry
ing
sh
rin
kag
e (
µm
/m)
Material
Figure 30: Drying shrinkage (ASTM C 157 mod.) of repair materials (Laval) versus REMR project 28-
day limit
Table 8: Comparison of 28 day and long term results of ASTM C 157 modified test
Material Location 28-day
(µm/m)
Long-Term
(µm/m) LT / 28d
Concrete U. Laval 790 1200 1.5
BOR 625 1000 1.6
Mortar U. Laval 1400 1800 1.3
BOR 1400 1800 1.3
P-Mortar U. Laval 3080 3600 1.2
BOR 2750 3600 1.3
4.3.1.2 Field conditions
The shrinkage test (ASTM C 157 modified) was also carried out in field conditions. The
specimens were stored outside, next to the experimental repair tests (Baenziger block and
Box test). The results presented in Figure 31 show, for all materials, that the ultimate
shrinkage value ranges between 30% and 35% of the ultimate shrinkage obtained in
61
laboratory conditions (relative drying shrinkage magnitude). Comparing the three repair
materials together, it appears that the relative drying shrinkage magnitude is the same as for
the tests carried out in laboratory conditions.
For field conditions, it would not be relevant to compare the shrinkage values obtained to
the 28 day criterion proposed by Poston et all. [1998] because of the drying’s kinetics.
0
400
800
1200
1600
2000
0 56 112 168 224 280 336
Concrete
MortarPolymer Mortar
Dry
ing
sh
rin
ka
ge
(µ
m/m
)
Elapsed time since drying (d)
Figure 31: Drying shrinkage (ASTM C 157 mod.), field conditions – Laval University
4.3.2 Flexural creep test
The tensile creep of the repair material was evaluated through the flexural creep test. This
test was carried out at both Université Laval and BOR laboratories. Figure 32 presents the
mid-span deflection measured in-between the two loads applied on the specimen for tests
carried out at Université Laval and the Bureau of Reclamation respectively.
The material that presents the lowest shrinkage magnitude is also the one which shows less
creep. At the opposite, the material that shows the highest shrinkage magnitude is also the
one that presents the highest creep deformation. The flexural creep tests carried out at
Université Laval show slightly higher deformation compared to what obtained at the
Bureau of Reclamation. However this difference is not significant.
62
Creep potential is usually presented in terms of the creep coefficient (). The creep
coefficient is the ratio of the delayed deflection over the instant deflection due to the
application of loads on the specimen. This calculation allows representing the deformation
due to sustained load compared to the instant elastic deformation due to loading. The
following equation represents the calculation of the creep coefficient:
)(
)()(
0
0
t
tt
Where: Δ(t) = Deflection under loading at time t
Δ(t0) = Instant deflection upon loading
0,00
0,25
0,50
0,75
1,00
1,25
1,50
0 56 112 168 224 280 336
Concrete
MortarP-Mortar
Mid
-sp
an
defl
ecti
on
(m
m)
Elapsed time since initiation of drying (d)
0,00
0,25
0,50
0,75
1,00
1,25
1,50
0 56 112 168 224 280 336
Concrete
MortarP-Mortar
Mid
-sp
an
defl
ecti
on
(m
m)
Elapsed time since initiation of drying (d)
a) Laval University b) BOR
Figure 32: Flexural creep test results
Higher the creep coefficient is, higher the potential of stress relaxation is when concrete is
submitted to restrained shrinkage.
This test was carried out under laboratory conditions (23°C and 50% R.H.) at both Laval
University and Bureau of Reclamation. Figure 33 show the results expressed in terms of
creep coefficient as a function of time.
63
0
2
4
6
8
10
0 56 112 168 224 280 336
ConcreteMortarP-Mortar
Cre
ep
co
eff
icie
nt
()
Elapsed time since drying (d)
0
2
4
6
8
10
0 56 112 168 224 280 336
ConcreteMortar
P-Mortar
Cre
ep
co
eff
icie
nt
()
Elapsed time since drying (d)
a) Laval University b) BOR
Figure 33: Flexural creep test calculations
Long term results from the two laboratories are very similar, but are slightly different at
early age (0-56 days), particularly for the Mortar and the P-Mortar materials. At Laval
University, the increase of the creep coefficient was faster than for the BOR, but this
difference faded out after 56 days.
This test gives information on how the material is able to relax tensile stress developed
under restrained shrinkage condition but the tensile creep potential, solely, cannot be a
performance indicator for selection of a repair material that would resist to cracking. This
test should be used in combination with drying shrinkage (ASTM C 157 and material’s
mechanical properties (Modulus of elasticity and tensile strength) in order to get a residual
stress evaluation and compare the residual stress to the material’s tensile strength.
4.3.3 Laval beam deflection test
The beam deflection test was conducted at two different locations; Laval University and the
Bureau of Reclamation (BOR) in Denver, Colorado. At Laval University the test was
performed in field conditions and at the Bureau of Reclamation the test was done in the
laboratory (controlled environment). For each condition, two graphs are presented; the first
one shows the readings, which correspond to the mid-span deflection and the second
64
presents the results after the calculations to obtain the equivalent end-beam deflection for a
one meter long beam corresponding to the mid-span deflection reading. This calculation
was explained and detailed previously (see section 3.4.4.6). The results are presented in
Figure 34 to Figure 36.
4.3.3.1 Laboratory conditions
0
1
2
3
4
5
6
7
8
0 56 112 168 224 280 336
Concrete
Mortar
Polymer Mortar
Mid
-Sp
an
defl
exio
n (
mm
)
Elapsed time since drying (d)
0
1
2
3
4
5
6
7
8
0 56 112 168 224 280 336
Concrete
Mortar
Polymer Mortar
Mid
-Sp
an
defl
exio
n (
mm
)
Elapsed time since drying (d)
a) Laval University b) BOR
Figure 34: Beam deflection test results, laboratory conditions
0
10
20
30
40
0 56 112 168 224 280 336
Concrete
Mortar
P-Mortar
En
d b
ea
m d
efl
ex
ion
(m
m/m
)
Elapsed time since drying (d)
0
10
20
30
40
0 56 112 168 224 280 336
Concrete
Mortar
P-Mortar
En
d b
eam
defl
exio
n (
mm
/m)
Elapsed time since drying (d)
a) Laval University b) BOR
Figure 35: Beam deflection test calculations, laboratory conditions
65
4.3.3.2 Field conditions
0,0
0,25
0,50
0,75
1,0
1,3
1,5
1,8
2,0
0 56 112 168 224 280 336
Concrete
Mortar
P-Mortar
Mid
-Sp
an
de
flex
ion
(m
m)
Elapsed time since drying (d)
0
10
20
30
40
0 56 112 168 224 280 336
Concrete
Mortar
P-Mortar
En
d b
ea
m d
efl
ex
ion
(m
m/m
)
Elapsed time since drying (d)
a) Readings b) Calculations
Figure 36: Beam deflection test, field conditions (Laval University)
The results from the Laval Beam Deflection test in laboratory conditions exhibit the same
trend than other drying shrinkage related tests. The P-Mortar material shows a lot more
deformation than other materials tested while the Mortar material shows a slightly higher
deflection.
In field conditions the Mortar material is the one which shows a smaller deflection while
the P-Mortar and the Concrete material show higher deflection. After 336 days, the P-
Mortar material tends to match the trend where this material usually shows much more
deformation than other tested materials.
The Laval beam deflection test allows comparing materials but is not suitable for a
theoretical analysis of the sensitivity to cracking. Of course it can give an idea of the
material deformation due to drying. As for some other tests, the Laval Beam Deflection test
cannot be used solely to predict cracking sensitivity of a repair material because it only
takes into account for the deformation side of the phenomena and it neglects the relaxation
due to creep phenomenon and the material’s tensile strength.
66
4.3.4 Vertical Beam Curling test
Figure 37 presents a graph of the actual end-beam curling and shows the end-beam curling
for a one meter long equivalent beam. This was done in order to be able to compare results
from Vertical Beam Curling test and the SPS Plate test results.
The P-Mortar material exhibits more than twice the end beam deflection compared to the
Mortar and the Concrete materials. The Mortar end-beam curling is slightly higher than for
Concrete material. This correlates with other volumetric tests conducted in this project.
0
4
8
12
16
20
24
28
32
0 112 224 336 448
Concrete
Mortar
P-Mortar
En
d-B
ea
m d
efl
ex
ion
(m
m)
Elapsed time since drying (d)
0
10
20
30
40
0 56 112 168 224 280 336
Concrete
Mortar
P-Mortar
En
d-B
ea
m d
efl
exio
n (
mm
/m)
Elapsed time since drying (d)
a) Readings b) Calculations
Figure 37: Vertical Beam Curling test
The Vertical Beam Curling test is a combination of the SPS plate test method and the Laval
beam deflection test. As for the Laval beam deflection test, the vertical beam curling test
does not take into account for any imposed or gravitational load. The only load applied to
the system is due shrinkage. Also, the relaxation due to creep is not taken into account in
this test.
4.3.5 SPS plate test
This test was run only at the Bureau of Reclamation with the intent to compare it with the
beam deflection and vertical beam deflection tests, both performed at Laval University.
Figure 38 presents the results for the three materials. The curve for the polymer mortar
stops earlier than for the two other materials because the specimens broke off. Figure shows
67
the end beam deflection (a). This gives the exact value of the deflection. Due to the
different geometries of existing tests for curling tests (Laval BD, SPS Plate test and others),
the reading values provide comparative information on different materials, but it does not
allow for direct comparison of the test methods. In order to compare the tests, calculations
were made to transform readings and obtain the corresponding end beam deflection for a
one meter long beam, which is presented in the graph on the right in Figure 38.
0
10
20
30
40
50
0 56 112 168 224 280 336
Concrete
MortarPolymer MortarE
nd
be
am
de
fle
xio
n (
mm
)
Elapsed time since drying (d)
0
10
20
30
40
0 56 112 168 224 280 336
Concrete
MortarP-Mortar
En
d b
ea
m d
efl
ex
ion
(m
m/m
)
Elapsed time since drying (d)
a) Readings b) Calculations
Figure 38: SPS plate test results
4.4 Cracking sensitivity evaluation test methods
4.4.1 General
Results from sensitivity to cracking tests are summarized and grouped in Table 9 and the
crack mappings are presented in Appendix A. The monitoring was done every day for the
first week and no cracks appeared on any repaired slabs. Monitoring took place once a
week for the following month after which the monitoring had to stop until spring due to the
snow cover. When snow started to cover the slabs, no crack was visible, even when
moisturizing the concrete surface. As soon as the snow melted (late April) crack mapping
was done. A visual inspection of the slabs was made twice a month and the mapping has
been updated three times during the summer. The results presented here are the
observations after two years of exposure in field condition.
The crack mappings and tables of crack intensity calculated for each material are shown in
Figure 40 and 42. The cracking intensity is evaluated via two different calculation methods.
68
The first one is the cracking area per square meter of surface. This is obtained by the
addition of the crack areas divided by the surface of the repair material to get a surface of
cracking per square meter of repair material. The crack area is obtained by multiplying the
length of a crack by its average opening, as illustrated in Figure 39. The second cracking
index used is the length of crack per square meter of repair and it is obtained by adding the
length of each crack and dividing the result by the surface of the repair. The number of
cracks that cross the repair from one side to the other is also reported.
Figure 39: Cracking indices representation
All crack sensitivity tests carried out in this study ranked the performance of the three
materials in the same order. The Concrete performed better than other materials and the
Mortar was better that the P-Mortar. The concrete material did not show any crack on the
various sizes of Box test and Baenziger block test, except for one specimen of medium size
Box test specimen that show one single crack. The crack was hairline and a few centimeters
long. The only test in which the concrete material cracked is the German Angle test, in both
laboratory and field conditions.
4.4.2 Box test and Baenziger block test
One of the objectives of the Phase I of this study was to select the optimal size and/or
geometry of test specimen for sensitivity to cracking test method. As stated previously,
three repair materials were used in this phase of the study and they were selected based
upon their difference of volumetric and mechanical properties. As expected, the repair
materials used had a different sensitivity to cracking and the results obtained reflect this in
w1
Li w2
69
the various sizes of the Box test and the Baenziger block test. According to Figure 40,
increasing the size of the box test specimen leads to higher cracking index values (surface
and linear) for Mortar and P-Mortar materials. The Concrete material was not sensitive to
the size of the box test specimen. None of the specimens showed cracks for the Concrete
material. Based on the behavior of the Mortar and the P-Mortar materials, the small box test
specimen can be discarded as sensitivity to cracking evaluation test method. Cracking
indexes are lower than for other sizes and thus interpretation of cracking index is more
complex and less significant.
0
200
400
600
800
1000
1200
1400
1600
Concrete Mortar P-Mortar
Small Box
Medium BoxLarge BoxBaenziger Block
Su
rfa
ce c
rack
ing
in
dex (
mm
²/m
²)
Materials
0
1 104
2 104
3 104
4 104
5 104
Concrete Mortar P-Mortar
Small
MediumLargeBaenziger Block
Lin
ear
cra
ckin
g i
nd
ex (
mm
/m²)
Materials
a) Surface cracking index b) Linear cracking index
Figure 40: cracking index versus material and test specimen
The Baenziger block test specimen shows higher cracking indexes (surface and linear) than
the small box test specimen, but also has a different geometry. These two indexes (surface
and linear) are close for the Baenziger block and medium and large box test specimens. The
number of transverse cracks also increases with the size of the box test specimen
(Figure 41). The Baenziger block test specimen shows less transverse crack than the
medium and large size of box specimens but more than the smaller size of box test
specimen, which confirms the choice of discarding the small box specimen. Considering
that the Baenziger block test gives similar results than the medium and large box test
specimens, the Baenziger block is preferred because it is more convenient for use in
laboratory and it is easier to manipulate due to its smaller size.
70
0
20
40
60
80
100
Concrete Mortar P-Mortar
Small Box
Medium BoxLarge BoxBaenziger Block
Nu
mb
er
of
tra
nsvers
e c
rack
Materials
Figure 41: Number of transverse crack(s) onto box and Baenziger block tests specimens
Table 9: Summary of crack evaluation tests
Material
Test Specimen
Cracking density (mm²/m²) Crack length (mm/m²) Number of cracks
Concr. Mortar P-Mortar Concr. Mortar P-Mortar Concr. Mortar P-Mortar
Baenziger
1 0 163 1394 0 4217 43067 0 14 75
2 0 203 1187 0 5500 37056 0 23 54
3 0 198 1536 0 6970 49111 0 27 88
AVG 0 188 1372 0 5562 43078
Creep Box
Small
1 0 133 744 0 2570 17678 0 9 28
2 0 108 1354 0 1493 34400 0 2 56
3 0 3 1373 0 63 37000 0 2 59
AVG 0 81 1157 0 1375 29693
Creep Box
Medium
1 9 56 1238 367 808 31656 8 2 62
2 0 209 1477 0 4733 35544 0 35 71
3 0 91 1500 0 1898 59178 0 17 114
AVG 3 118 1405 122 2480 42126
Creep Box
Large
1 0 212 1660 0 2882 43933 0 28 74
2 0 154 1263 0 2717 33700 0 22 63
3 0 121 1743 0 2343 61111 0 22 130
AVG 0 162 1556 0 2647 46248
71
4.4.2.1 Bond strength evaluation
The Concrete material performed very well on the box test and the Baenziger block test. It
was therefore necessary to make sure that there was no bond failure between the repair
material and the substrate. To verify this point, a series of pull-off tests were conducted
(Figure 42). The bond was found to be adequate for all materials, ranging from 1.9 MPa to
2.6 MPa (Figure 43). Therefore, the resistance to cracking is not resulting from a bond
failure between the repair material and the substrate. The slabs were also sounded with a
hammer and no delamination was found in any repair.
Figure 42: Pull-off test on box test with Germann Instrument equipment
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Concrete Mortar P-mortar
1.89
2.63
1.97
Te
ns
ile
str
en
gth
(M
Pa)
Material
Figure 43: Phase I pull-Off test results
72
4.4.2.2 Depth of cracking
The polymer-modified repair material exhibited very severe surface cracking in all
specimens’ geometries, and therefore it was difficult to determine with a reasonable degree
of certainty whether it was cracking or a surface phenomenon called crazing. To clarify this
point, longitudinal saw-cuts were made on a small box test and Baenziger block test
specimens showing that the depth of the cracks for most of the cracks is between 1 to 3
centimeters (Figure 44).
a) Small box test b) Baenziger block test
Figure 44: Depth of crack for P-Mortar
4.4.3 Experimental repair
Figure 45 - Figure 47 show crack mappings of experimental repairs after two years of
exposure in field conditions on the seacoast in Port Hueneme, CA. Like in the tests
performed at Laval University and BOR, the Concrete material was the most resistant to
cracking followed by the Mortar and the P-Mortar materials, respectively. After two years
of exposure, the repairs made with concrete did not show any crack that spreads from side
to side of the repair; one of the three specimens did not show any crack at all. In order to
make sure that the absence of cracking was not the result of a bond failure, all repairs were
inspected with a hammer. No delamination was found on any repair.
From these experimental repairs, it is clear that the Box Test (medium size) and the
Baenziger block test are representative of real repair situation and this observation makes
73
these test attractive for further investigation, especially the Baenziger Block because of its
convenient size.
Figure 45: Crack mapping of experimental repair – Concrete mix
74
Figure 46: Crack mapping of experimental repair – Mortar mix
Figure 47: Crack mapping of experimental repair – P-Mortar mix
4.4.4 German angle test
The mappings of the German angle test were done the same way as for the other cracking
tests (slab tests). For each material, there are two testing conditions; field conditions
(University Laval) and laboratory conditions (BOR). The surface of material exposed to
drying on the German Angle test is 1000 mm by 107 mm.
The German angle test results after 90 days of exposure are presented in Table 10. As for
the Baenziger block test and the box test, the results presented consists of the cracking
density in mm²/m², the crack length in mm/m² and the number of transverse crack. A crack
mapping of the German Angle test specimens is also presented in Appendix A.
75
The German angle test show extensive cracking for all materials tested; for each material,
cracking occurred within the first month of drying. This test is severe and is not applicable
for repair material evaluation.
Table 10: German Angle test results
Material
Test Specimen
Cracking density (mm²/m²) Crack length (mm/m²) Number of cracks
Concr. Mortar P-Mortar Concr. Mortar P-Mortar Concr. Mortar P-Morar
G. Angle
Field cond.
U.Laval
1 100 450 852 2000 8000 15000 2 8 15
2 50 375 1196 1000 8000 18000 1 8 18
3 0 613 792 0 9000 12000 0 9 12
AVG 50 479 946 1000 8333 15000
G. Angle
Lab cond.
BOR
1 1 10 24 1292 7000 16000 11 9 16
2 3 12 24 3792 11000 16000 16 11 16
3 1 14 23 1050 13000 17000 7 13 17
AVG 2 12 24 2045 10333 16333
4.4.5 Restrained shrinkage test
4.4.5.1 REMR ring test
This test was performed at two different locations (Université Laval and BOR) and test was
carried out following the same procedure is described in section 3.4.4.4. For each specimen,
the number of crack(s), average crack opening in millimeters and the time of cracking are
reported.
The REMR configuration ring test results are presented in Table 11 and summarized in
Figure 48 This ring test configuration is very severe and it is not very representative of a
real repair. According to Emmons and Vaysburd [1995] ring test criteria, all materials
tested in this research program would have failed this test. The Concrete, which cracked
within 14 days in the REMR ring test, performed well on the Box test and the Baenziger
Block test. Moreover, the Concrete did not show cracking on the experimental repair
performed at the US Navy. Comparing the Mortar and the P-Mortar results, one can
observe that the time to first crack is almost the same while these materials did not behave
76
similarly in the Box test, Baenziger Block test and experimental repairs. Based on these
observations, it clearly appears that the REMR ring test is severe and does not allow for
good comparison between repair materials since the time of first crack is similar for all
materials tested. However, for the crack opening, the concrete performed better than other
materials tested.
Table 11: Summary results of the REMR ring test
Location Characteristics
Concrete Mortar P-Mortar
Sample Sample Sample
1 2 3 1 2 3 1 2 3
Univ.
Laval
Crack(s) 1 5 - 5 1 - 2 2 1
Opening (mm) 0.90 0.080 - 0.21 1.5 - 2 -* 3.5
First crack (d) 6 10 - 3 5 - -* 2 3
USBR
Crack(s) 2 4 5 2 2 4 4 9 6
Opening (mm) 0.40 0.27 0.04 1.00 1.15 0.32 3.68**
3.71**
3.49**
First crack (d) 6 6 9 2 2 5 4 2 2
* The concrete ring broke off and fell apart when taken out of the shelf for inspection
** Only one major crack and the other were hairlines
0
2
4
6
8
10
Concrete Mortar P-Mortar
LavalBOR
Tim
e t
o f
irs
t cra
ck
(d
)
Material
Figure 48: Time to first crack in REMR Ring Test
77
4.4.5.2 ASTM-like ring test
The ASTM-like ring test method generates more quantitative results than the REMR ring
test method since the ring is equipped with strain gages on the inner surface of the steel
ring.
Figure 49 to Figure 51 present the average stress developed in the repair material. On these
graphs it is also possible to pinpoint the exact time of crack occurrence, which is
characterized by a gradual drop or a sudden drop in the average stress.
Table 12 summarizes the results of ASTM-like Ring test in terms of time to first crack.
Compared to the REMR setup, the ASTM-like configuration gives a longer time to first
crack and more widely spread than results obtained from the REMR ring test.
a) Specimen 1 b) Specimen 2
Figure 49: Ring test – Concrete
78
a) Specimen 1 b) Specimen 2
Figure 50: Ring test – Mortar
a) Specimen 1 b) Specimen 2
Figure 51: Ring test – P-Mortar
Table 12: ASTM-like test results
Material Time to first crack (days)
Ring 1 Ring 2
Concrete 35 17
Mortar 3 4
P-Mortar 7 7
79
4.5 Conclusion and recommendations – Phase I
Based on the results of the Phase I of this study, the following conclusions and
recommendation are drawn:
The Baenziger block is the most optimal specimen configuration to evaluate the
sensitivity to cracking of repair materials, and therefore it is recommended for
further studies in Phase II of the project.
The results of the ring test configuration [COE, March 1999] employed in this
program did not allow for direct correlation with experimental repairs, with
adequate degree of confidence. Therefore it is recommended for further studies in
the Phase II to use the ASTM ring test configuration (ASTM C 1581-04 “Standard
test method for determining age of cracking and induced tensile stress
characteristics of mortar and concrete under restrained shrinkage).
Free shrinkage test should be carried out in Phase II. Further tests are necessary to
establish whether or not maximum ultimate shrinkage can be set as an acceptance
criterion for repair material.
Since correlation between SPS Plate test and Laval beam deflection test has been
found on only two materials out of three, further studies are recommended for
Phase II.
The sensitivity to cracking of repair material test program anticipated in Phase II
should anticipate placements of experimental in-situ actual field repairs for
correlation with test methods employed.
The approached calculation for residual stress in a repair material is presented in Table 13.
The residual stress takes into account the 28-day mechanical properties (tensile strength and
modulus of elasticity), long term drying shrinkage and creep potential. Based on the
80
mechanics of materials, the following equation was used to calculate the long term residual
stress into the repair material.
1
Eres
Where:
σres = Residual stress (MPa)
E = Modulus of elasticity (MPa)
= Tensile creep potential
Since the box test and the Baenziger block test were carried out in field conditions, the
mechanical properties and drying shrinkage values of field conditions tests were used.
However, the tensile creep potential was estimated from the laboratory test results. Since
drying shrinkage and creep are interrelated, the proportion of drying shrinkage of field
condition versus laboratory condition was used for evaluation of field condition tensile
creep potential. This ratio ranged between 30% and 35% depending of the material
considered.
Table 13: Approached calculation of residual stress
Shrinkage
(Lab.)
Shrinkage
(in-situ)
Modulus
of elast.
Creep
(Lab.)
Creep
(in-situ)
Residual
stress
Tensile
strength
x10-6
x10-6
MPa MPa MPa
Concrete 930 350 34.4 6.5 2.5 2.0 3.5
Mortar 1800 600 35.8 7.8 2.6 6.0 3.6
P-Mortar 3540 1125 18.1 7.3 2.3 6.2 2.8
According to residual stress calculation, only Concrete material would resist to cracking on
a long term basis, which was verified with testing on box test and Baenziger block test. On
experimental repairs, there were hairline cracks on the Concrete material but the
environment was not the same and these cracks are very fine and none of the cracks
observed are going across the repair width.
81
5 Phase II
5.1 Introduction
The main objective of the Phase I was to evaluate the performance of cementitious repair
materials in experimental repairs placed in geometrically different cavities of prefabricated
reinforced concrete slabs and select the optimum cavity geometry and slab configuration
for further studies of sensitivity to cracking of repair materials. As a result of the Phase I
study, the “Baenziger Block” was judged as the optimal specimen configuration and was
recommended for further studies in Phase II of the project.
Three locations were selected to conduct field and laboratory testing for the second phase
of this project. The BOR (Denver, CO), the US Navy facilities (Port Hueneme, CA) and
Laval University (Quebec, Canada) were selected. At both the BOR and US Navy facilities,
experimental field repairs were performed, and laboratory tests were conducted at Laval
University. Figure 52 presents the two experimental sites and Table 14 presents the
different tests performed at each location.
Figure 52: BOR (left) and US Navy (right) experimental sites
82
Table 14: Test carried out for each location
Location
Test performed BOR US Navy
Laval
University
Compressive strength (ASTM C 39)
Tensile Splitting test (ASTM C 496)
Modulus of Elasticity (ASTM C 469)
Free drying shrinkage (ASTM C 157
Modified ACI 364-3R)
Flexural Creep test
Ring test (ASTM C 1581)
Baenziger block test
Experimental repair
Pull off test
5.2 Phase II materials
Four proprietary cementitious repair materials (number 1 to 4) from three manufacturers
were used as mortars (M) and extended mortars (C). To extend the mortars, 3/8 inch gravel
provided by each manufacturer was added to the mixture. The amount of aggregate addition
was specified by each manufacturer. The repair materials used were pre-mixed and bagged
at the respective manufacturer’s plants and sent to each location.
Since materials used were proprietary materials, their composition was unknown. After
testing started, it was found that materials 1M and 1C were magnesium phosphate cement
based fast-setting materials. Materials 2M, 2C, 3M and 3C were Portland cement based
materials and had typical setting times. Materials 4M and 4C exhibited shorter setting times
than for usual Portland cement based materials. A problem occurred at the US Navy
location for material 4C and the results for this material at the US Navy location were not
used, nor was the experimental repair monitored.
83
The mortars were mixed in a mortar mixer (paddle type) and the extended mortars were
mixed in a concrete mixer (rotating drum mixer) as shown in Figure 53.
Figure 53: Paddle type mixer (left) and rotating drum mixer (right)
5.3 Characterization
5.3.1 Mechanical properties
Mechanical properties test results are presented in Table 15 to Table 17 and Figure 54 to
Figure 56. Test results presented are the average of the results on three specimens. As
shown in Figure 54, materials 1C, 3C and 4C show significant difference in results between
BOR and Université Laval. The materials came from the same production batch and the
aggregate (added manually) came from the same source. The same types of mixers were
used at both Université Laval and BOR for mixing activities. The only difference between
materials at Université Laval and the BOR is the mixing conditions. At the BOR, the
materials were mixed in field conditions and the weather was hot and sunny with medium
wind. At Université Laval, the materials were mixed in laboratory conditions. This might
explains the difference in mechanical properties.
84
Table 15 : Compressive strenght (ASTM C 39)
Material Age (d) USBR
PSI (MPa)
US Navy
PSI (MPa)
Laval
University
PSI (MPa)
1M
3 4077 (28.1) - 3130 (21.6)
28 5273 (36.4) - 4220 (29.1)
56 - 4701 (32.4) -
1C
3 5017 (34.6) - 3393 (23.4)
28 6447 (44.5) - 4205 (29.0)
56 - 6178 (42.6) -
2M
3 3550 (24.5) - 3364 (23.2)
28 4843 (33.4) - 4886 (33.7)
56 - 6637 (45.8) -
2C
3 3740 (25.8) - 2958 (20.4)
28 5043 (34.8) - 3900 (26.9)
56 - 5353 (36.9) -
3M
3 2680 (18.5) - 2856 (19.7)
28 4067 (28.0) - 4524 (31.2)
56 - 5846 (40.3) -
3C
3 2960 (20.4) - 2480 (17.1)
28 5043 (34.8) - 3378 (23.3)
56 - 5933 (40.9) -
4M
3 4847 (33.4) - 5031 (34.7)
28 5870 (40.5) - 6264 (43.2)
56 - 4410 (30.4) -
4C
3 4033 (27.8) - 5365 (37.0)
28 4870 (33.6) - 6870 (47.4)
56 - 1773 (12.2) -
85
0
10
20
30
40
50
1M 1C 2M 2C 3M 3C 4M 4C
BOR
U. LavalS
tre
ng
th (
MP
a)
Material
0
10
20
30
40
50
1M 1C 2M 2C 3M 3C 4M 4C
BOR
U. Laval
Mo
du
lus
(G
Pa
)
Material
Figure 54: Comparison of 28-day compressive strength between BOR and Laval University (left)
Figure 55: Comparison of 28-day modulus of elasticity between BOR and Laval University (right)
86
Table 16 presents results from modulus of elasticity test performed at the BOR and Laval
University.
Table 16 : Modulus of elasticity (ASTM C 469)
Material Age (d) USBR
x106 PSI (GPa)
Laval University
x106 PSI (GPa)
1M 3 3.84 (26.5) 4.04 (27.8)
28 4.09 (28.2) 4.28 (29.6)
1C 3 5.36 (36.9) 5.16 (35.6)
28 6.05 (41.7) 5.86 (40.4)
2M 3 2.93 (20.2) 3.33 (23.0)
28 3.57 (24.6) 3.88 (26.8)
2C 3 3.27 (22.5) 2.94 (20.2)
28 3.39 (23.4) 3.35 (23.1)
3M 3 2.40 (16.6) 2.64 (18.2)
28 2.84 (19.6) 3.39 (23.4)
3C 3 2.67 (18.4) 2.75 (18.9)
28 2.97 (20.5) 3.41 (23.5)
4M 3 3.39 (23.4) 4.07 (28.1)
28 3.49 (24.0) 4.47 (30.8)
4C 3 3.41 (23.5) 3.83 (26.4)
28 N/A 4.61 (31.8)
87
Splitting tensile strength results for tests carried out at the BOR and Laval University are
presented in Table 17.
Table 17 : Splitting tensile strength (ASTM C 496)
Material Age (d) USBR
PSI (MPa)
Laval
University
PSI (MPa)
1M 3 457 (3.1) 218 (1.5)
28 560 (3.9) 323 (2.2)
1C 3 540 (3.7) 267 (1.8)
28 640 (4.4) 348 (2.4)
2M 3 450 (3.1) 371 (2.6)
28 500 (3.4) 463 (3.2)
2C 3 423 (2.9) 312 (2.2)
28 590 (4.1) 374 (2.6)
3M 3 283 (2.0) 384 (2.6)
28 413 (2.9) 490 (3.4)
3C 3 363 (2.5) 241 (1.7)
28 550 (3.8) 332 (2.3)
4M 3 540 (3.7) 441 (3.0)
28 657 (4.5) 473 (3.3)
4C 3 468 (3.2) 405 (2.8)
28 570 (3.9) 523 (3.6)
88
0
1
2
3
4
5
1M 1C 2M 2C 3M 3C 4M 4C
BOR
U. Laval
Mo
du
lus (
GP
a)
Material
Figure 56: Comparison of 28-day tensile splitting strength between BOR and Laval University
5.3.2 Drying shrinkage test (ASTM C 157 Modified)
The results of free drying shrinkage test conducted at the BOR and Laval University are
presented in Figure 57.
0
200
400
600
800
1000
1200
1400
0 28 56 84 112 140 168
1M
1C
2M
2C
3M
3C
4M
4C
Sh
rin
ka
ge
(µ
m/m
)
Age of material (d)
0
200
400
600
800
1000
1200
1400
0 28 56 84 112 140 168
1M
1C
2M
2C
3M
3C
4M
4C
Sh
rin
ka
ge
(µ
m/m
)
Age of material (d)
a) Laval University b) BOR
Figure 57 : Drying shrinkage results (ASTM C 157 modified) – BOR
89
Free shrinkage tests were conducted at BOR and Laval University. The results were similar
for most materials but slight differences were found in the case of materials 3C, 4M and
4C.
The ultimate free-drying shrinkage values range from 100 µm/m to 1200 µm/m. Materials
1M and 1C exhibited particularly low magnitude shrinkage compared to other materials
tested. Since materials 1M and 1C are magnesium phosphate cement based materials, most
of the deformation occurs within the first day, which is not covered by the free shrinkage
test method used in this study. It is therefore normal to see a low magnitude of shrinkage.
For materials 2M, 3M and 4M and 4C, there is a relationship between the length change
and number of cracks in the repair (Figure 58). However, this relationship becomes less
obvious for rapid set materials 1M and 1C.
0
5
10
15
20
25
30
35
0 200 400 600 800 1000 1200
2M
2C
3M
3C
4M
4C
Nu
mb
er
of
cra
ck(s
) o
n o
ve
rla
y
Length change - ASTM C157 modified (µm/m)
Figure 58: Relationship between length change and number of cracks in repair – BOR
90
The correlation between the length change (free drying shringage) test results and the
number of cracks in repair is much stronger for mortar materials than for extended mortars
(Figure 59)
0
5
10
15
20
25
30
35
0 200 400 600 800 1000 1200
2M
3M
4M
Nu
mb
er
of
cra
ck(s
) in
rep
air
Length change - ASTM C157 modified (µm/m)
0
5
10
15
20
25
30
35
0 200 400 600 800 1000 1200
2C
3C
4C
Nu
mb
er
of
cra
ck(s
) in
rep
air
Length change - ASTM C157 modified (µm/m)
a) Mortars (BOR) b) Extended Mortars (BOR)
Figure 59: Relationship between free shrinkage and cracking in experimental repairs (after 9 months)
The test results and analysis lead to the conclusion that the free drying shrinkage test,
ASTM C 157 modified, can be used to predict cracking behavior of mortar types
cementitious materials and it is not reliable for predicting the performance of extended
mortars and concretes.
5.4 Experimental repairs
The crack patterns at the end of the monitoring periods are presented in Figure 60 and
Figure 61 for BOR and US Navy respectively. In these figures, the localized debonding of
the repairs are shown by the shaded areas. Debonding occurred only for materials 1M and
4M at the BOR site.
Table 18 presents the number of cracks for each experimental repair.
91
Figure 60 : Cracking in experimental repairs at BOR (9 months after placement)
4C
4M
3C
3M
2C
2M
1C
1M
92
Figure 61 : Cracking in experimental repairs at US Navy (9 months after placement)
4M
3C
3M
2C
2M
1C
1M
93
Table 18 : Overlays cracking summary
Location
Material
BOR
Number of crack(s)
US Navy
Number of crack(s)
1M 4 9
1C 12 8
2M 31 12
2C 0 1
3M 45 16
3C 0 4
4M 22 8
4C 8 -
As expected, mortars 2M, 3M and 4M exhibited more severe cracking than their extended
counterpart. The only exception was found with mortar 1M at the BOR experimental site.
This can be attributed to large areas of debonding of the 1M repair, which significantly
reduced the restrained tensile stresses responsible for cracking.
Under otherwise equal conditions, all extended mortars – 1C, 2C, 3C and 4C exhibited
substantially improved characteristics with regards to cracking sensitivity.
Materials 2C and 3C were found to be the most crack-resistant: no cracking occurred at the
BOR testing site, and only one crack was recorded in material 2C at the US Navy location.
Materials 1M and 1C exhibited very early age cracking with several cracks recorded after 4
days at both locations. This can be explained by the high temperature of hydration
generated in this rapid-setting material. Discrepancies between the observed test results for
material 4C at the BOR and US Navy test sites were most likely caused by variations in
material compositions. In fact, the 4C material at the BOR site had very different
workability characteristics, as well as different mechanical properties from the 4C material
at the Navy site.
94
In order to verify whether the absence of crack in some of the experimental repairs was not
due to bond failure, pull-off tests were carried out on Materials 2C, 3C and 4C. Test results
are presented in Figure 62 and show that tensile bond was more than satisfactory (> 2 MPa)
for all tested materials at the BOR site. At the US Navy site, materials 2C and 3C still
exhibited adequate bond strength (> 2 MPa), but material 4C developed a relatively low
bond to the substrate (about 0.9 MPa).
0
0.5
1
1.5
2
2.5
2C 3C 4C
BOR
US Navy
Bo
nd
str
en
gth
(M
Pa
)
Material
Figure 62 : Tensile bond strength
In addition to the bond tests performed on selected materials, each experimental repair was
sounded with a hammer in order to check for any bond failure between the repair material
and the substrate. At the BOR site, material 1M exhibited significant debonding, as
idendtified by shaded areas in Figure 60. Material 4M also exhibited small zones of
debonding at both ends of the repair (Figure 61). All other experimental repairs were free of
debonding areas.
5.4.1 Baenziger block test
The results of the Baenziger block test are presented in Figure 63 and the Table 19
summarizes the results. The Baenziger Block tests were only conducted at the BOR site,
95
where they were exposed to the environment next to the experimental repairs. Cracks were
not observed for material 1M and 1C at the end of the 9-month monitoring period. After the
same period, the experimental repairs with materials 1M and 1C exhibited substantial
cracking. For materials 2M and 3M, 15 and 50 cracks were recorded for the Baenziger
blocks respectively. The experimental repairs with these materials accounted for 31 and 45
cracks. Baenziger block tests and experimental repairs with materials 2C and 3C remained
crack-free. Materials 4M and 4C did not crack in Baenziger block tests, but cracked
moderately in the experimental repairs.
Figure 63 : Mapping of Baenziger blocks showing cracks
4C 4M
3C 3M
2C 2
M
1C 1M
96
Table 19 : Baenziger Block test cracking summary
Material Number of crack(s)
1M 0
1C 0
2M 15
2C 0
3M 50+
3C 0
4M 0
4C 0
5.4.2 Ring test
Ring test results are presented from Figure 64 to Figure 67. In these figures, the tensile
stress is presented in graphs as a function of time. Table 20 summarizes the information
related to the Ring test.
-7
-6
-5
-4
-3
-2
-1
0
1
0 56 112 168 224 280
Ring 1
Ring 2
Str
ess
(M
Pa
)
Elapsed time after initiation of drying (d)
-7
-6
-5
-4
-3
-2
-1
0
1
0 56 112 168 224 280
Ring 1
Ring 2
Str
ess
(M
Pa
)
Elapsed time after initiation of drying (d)
Figure 64 : ASTM C 1581 Ring test results for material 1M (left) and 1C (right)
97
-7
-6
-5
-4
-3
-2
-1
0
1
0 1 2 3 4 5 6 7
Ring 1Ring 2
Str
ess (
MP
a)
Elapsed time after initiation of drying (d)
-7
-6
-5
-4
-3
-2
-1
0
1
0 28 56 84 112 140 168
Ring 1Ring 2
Str
ess (
MP
a)
Elapsed time after initiation of drying (d)
Figure 65 : ASTM C 1581 Ring test results for material 2M (left) and 2C (right)
-7
-6
-5
-4
-3
-2
-1
0
1
0 1 2 3 4 5 6 7
3M-1
3M-2
Str
ess (
MP
a)
Elapsed time after initiation of drying (d)
-7
-6
-5
-4
-3
-2
-1
0
1
0 7 14 21 28 35
Ring 1Ring 2
Str
ess
(M
Pa
)
Elapsed time after initiation of drying (d)
Figure 66 : ASTM C 1581 Ring test results for material 3M (left) and 3C (right)
98
-7
-6
-5
-4
-3
-2
-1
0
1
0 28 56 84 112 140 168 196
Ring 1
Ring 2
Str
ess (
MP
a)
Elapsed time after initiation of drying (d)
-7
-6
-5
-4
-3
-2
-1
0
1
0 7 14 21 28 35 42 49 56
Ring 1
Ring 2
Str
ess
(M
Pa)
Elapsed time after initiation of drying (d)
Figure 67 : ASTM C 1581 Ring test results for material 4M (left) and 4C (right)
Table 20 : Summary of the ring test results
Material Specimen Time of first
crack (d)
Number of
crack(s)
Crack opening (mm)
A B C
2M 2M-1 7 1 2.0 - -
2M-2 8 1 21 - -
2C 2C-1 44 0 0.2 - -
2C-2 164 1 0.1 - -
3M 3M-1 5 1 1.7 - -
3M-2 5 2 1.5 0.75 -
3C 3C-1 26 1 1.5 - -
3C-2 36 1 1.1 - -
4M 4M-1 56 1 0.05 - -
4M-2 94 1 0.05 - -
4C 4C-1 No Crack 0 - - -
4C-2 No Crack 0 - - -
Figure 68 shows the relationship between the number of cracks in the experimental repair
and the stress development rate in the ASTM C 1581 ring test for the BOR and the US
Navy locations, respectively. A strong correlation was found between the cracking behavior
99
of repair mortars 2M, 3M and 4M and the stress development rate in the ring test
(Figure 69). However, the age at cracking in the ring test did not correlate well with the
number of cracks in the repair slabs for extended mortar repair materials 2C, 3C and 4C.
Materials 2C and 3C did not show crack in the repair slabs but cracked in the ring test.
Material 4C did not crack in the ring test but exhibited cracking in the repair slab. The
correlation between the mortar materials test results in the ring test and cracking in
experimental repair slabs is consistent with the relationship found between the
ASTM C 157 modified free shrinkage test and the cracking behavior in repair slabs. This
means that for mortar type materials, there is a good relationship between the ring test
results and the ASTM C 157 modified test results (see Figure 70).
0
10
20
30
40
50
0 0.2 0.4 0.6 0.8 1
2M
2C
3M
3C
4M
4C
Nu
mb
er
of
cra
ck(s
) in
rep
air
Stress Development Rate in Ring Test (MPa/d)
0
10
20
30
40
50
0 0.2 0.4 0.6 0.8 1
2M
2C
3M
3C
4M
Nu
mb
er
of
cra
ck(s
) in
rep
air
Stress Development Rate in Ring Test (MPa/d)
a) BOR b) US Navy
Figure 68: Relationship between stress development rate in the Ring test and the number of cracks in
the experimental repairs (after 9 months)
Using the materials properties from shrinkage, modulus of elasticity and creep, it is
possible to approach calculation for residual stress in the repair material. The residual stress
in repair material and its tensile strength can be put in ratio and compared to the intensity
(number of cracks) of cracking observed on experimental repairs. Table 21 presents the
properties and residual stress over strength ratios for materials 1M, 1C, 2M, 2C, 3M and
3C. The materials that cracked more in the experimental repairs are the one that presents
100
the highest stress/strength ratio. When the ratio is equal or higher than 1 (residual stress
higher than material’s tensile strength, the cracking in experimental repair is severe.
0
10
20
30
40
50
0 0.2 0.4 0.6 0.8 1
2M
3M
4M
R2= 0.99884
Nu
mb
er
of
cra
ck(s
) in
rep
air
Stress Development Rate in Ring Test (MPa/d)
y = a*((1+((b/x)^c))^(-c/2))
ErrorValue
0,141270,51402a
13,55912,883b
0,671411,5181c
NA0,18891Chisq
NA0,8864R
0
10
20
30
40
50
0 0.2 0.4 0.6 0.8 1
2M
3M
4M
R2= 0.97496
Nu
mb
er
of
cra
ck(s
) in
rep
air
Stress Development Rate in Ring Test (MPa/d)
y = a*((1+((b/x)^c))^(-c/2))
ErrorValue
0,141270,51402a
13,55912,883b
0,671411,5181c
NA0,18891Chisq
NA0,8864R
a) BOR b) US Navy
Figure 69: Relationship between ASTM C 1581 ring test and cracking of experimental repairs for
mortars (after 9 months)
0
200
400
600
800
1000
1200
0 0.2 0.4 0.6 0.8 1
2M
2C
3M
3C
4M
4C
Len
gth
Ch
an
ge (
µm
/m)
Stress Development Rate (MPa/d)
Figure 70: Relationship between age at cracking in the ring test (ASTM C 1581) and the Length change
test (ASTM C 157 modified)
101
Table 21: Approach calculation of residual stress
Property
Material
Shrinkage
50% R.H.
Modulus
of elast.
Creep
50% R.H.
Residual
stress
Tensile
strength
Stress
ratio
Crack in repair
x10-6
MPa MPa MPa /fst
US
BOR
US
Navy
1M 110 29.6 1.8 1.2 3.9 0.31 4 9
1C 100 40.4 1.3 1.8 4.4 0.41 12 8
2M 860 26.8 5.7 3.4 3.4 1.00 31 12
2C 580 23.1 4 2.7 4.1 0.66 0 1
3M 1225 23.4 5.8 4.2 2.9 1.45 45 16
3C 800 23.5 4.1 3.7 3.8 0.97 0 4
5.5 Conclusion
Based on results of the testing program performed during Phase II of this study, the
following conclusions can be drawn:
A good correlation was found between the ring test (ASTM C 1581) and cracking
behavior of experimental repairs performed utilizing repair mortars. No such
correlation was found for ASTM C 1581 and experimental repair performed
utilizing mortars extended with coarse aggregate
A good correlation was found between ASTM C 157 modified and cracking
behavior of experimental repairs performed with mortars. As for ASTM C 1581
ring test, there was no correlation between ASTM C 157 modified and behavior of
experimental repairs made with mortars extended with coarse aggregate
A good correlation was found between ASTM C 1581 ring test results and
ASTM C 157 (modified) test results for repair mortars and mortars extended with
coarse aggregate..
For the ASTM C 1581 ring test, the stress development rate was found to be the
most reliable indicator of experimental repair cracking and should be used instead
of time to first crack.
102
For proprietary repair materials, geometry of experimental repairs and exposure
conditions utilized in this program, no reliable correlation was found between
ASTM C 1581 ring test results and cracking behavior of experimental repairs
performed with mortars extended with coarse aggregate.
For the residual stress to tensile strength ratio, there is a correlation between the
ratio and the number of cracks observed on the experimental repairs.
Further studies are recommended to verify the applicability of the ASTM C 1581 ring test
for predicting the cracking behavior of concretes and mortars extended with coarse
aggregate in repair projects.
103
6 Conclusion and recommendations
6.1 Creep Box and Baenziger block tests
The objective of the first phase was to identify the optimal size/geometry of a cavity slab
test specimen intended to evaluate the cracking sensitivity of a repair material. Four
different test specimens were investigated: the Small box, the Medium box, the Large box,
and the Baenziger block test specimens. The Small box test specimen was found to provide
insufficient restraint in order to characterize adequately the cracking sensitivity of the
various repair materials tested and was therefore rejected. Conversely, the Medium box,
Large box, and Baenziger block all yielded significant cracking and overall, the various sets
of data generated with the three different test specimens compared fairly well. The
Baenziger block, the smallest of the three different test specimens, stood out as the optimal
geometry, since it appeared to provide a comparable level of restraint as the Medium box
and Large box. Therefore, the Baenziger block test was selected for further studies in phase
II.
In the second phase the Baenziger block test did not performed as expected, or at least as it
did in the first phase. At the end of this research project, the Baenziger block remains the
recommended test geometry, although it will require a more comprehensive testing
program to fully validate this test method and fine tune the procedure. It will also be
necessary to verify the applicability of the Baenziger block test to non-cementitious repair
materials.
6.2 Ring test
In the first phase of the project, the ring test method used did not give satisfactory results.
However in the second phase, the freshly standardized ring test method (ASTM C 1584)
was used with good correlation between the stress rate development and the cracking on
experimental repairs and to a certain extent to the cracking activity on the Baenziger block
test. Therefore, the ASTM ring test method can be used as a tool to evaluate the sensitivity
104
to cracking of repair materials on a comparison basis. However, more investigation should
be made to establish performance criteria for the ASTM ring test method.
6.3 Indirect test method
The most common test performed to evaluate the dimensional behavior of cementitious
based material is the drying shrinkage test (ASTM C 157 modified by ACI 364-3R), which
does not take into account for very early age shrinkage. This test is interesting but takes
into account only for the shrinkage and cannot, solely, predict the sensitivity to cracking of
a repair material. When combined with other characterization tests (flexural creep, beam
deflexion test, tensile strength, modulus of elasticity, etc.) it can be possible to identify the
most crack resistant material among a selection. To evaluate the sensitivity to cracking, it is
better to have tests that combine the global behavior (stress and relaxation) or a
combination of tests that represents the whole dimensional balance component and the
mechanical properties of the material.
6.4 Recommendation for future research
This project was a contribution toward the development of a test method that will help
engineers to better evaluate the sensitivity to cracking of repair materials. However, there is
more work left to be done. There are two promising test method evaluated in this research
project; the ring test method and the Baenziger block test method.
The ring test method is now standardized as a method, but no criteria have been developed
so far. Recent research project [Modjabi, 2009] developed sensitivity to cracking index
based on the ring test results. More research in this field would help setting the performance
criteria for a material tested with the ring test.
The Baenziger block test method can be useful because the test can easily be performed in
various exposure conditions that allow to evaluate a material in real condition of exposure
to environment. However, this test method is more adapted to mortar than extended mortar
or concrete. With a repair thickness varying from 30 to 60 mm, it is impossible to test
105
material that have 14 mm aggregate in it and it is barely suitable for concrete containing 10
mm aggregate. It is recommended to modify slightly the design of the slab in order to have
a minimum thickness of 50 mm at any point and run series of test with a broad selection of
cementitious based materials.
Another interesting avenue for research would be to develop a numeric model that gathers
results from various characterization tests such as drying shrinkage, tensile creep
properties, mechanical properties, etc. in order to see the evolution of the deformation
compatibility of the material to be tested.
107
7 Bibliography
ABU-TAIR, A.I., LAVERY, D., NADJAI, A., RIGDEN, S.R., AHMED, T.M.A., “A new
method for evaluating the surface roughness of concrete cut for repair or
strengthening”, Construction and Building Materials”, 14, 2000, pp. 171-176.
ACI COMMITTEE 209, “Prediction of Creep, Shrinkage and Temperature Effects in
Concrete Structures”, American Concrete Institute (ACI-209-82), Detroit, MI, 1982,
108 pages.
ACI COMMITTEE 209, “Report on Factors Affecting Shrinkage and Creep of Hardened
Concrete (ACI 209.1R-05),” American Concrete Institute, Farmington Hills, MI,
2005, 12 pp.
Al-KUBAISY, M.A. and YOUNG, A.G., “Failure of Concrete under Sustained Tension”,
Magazine of Concrete Research, 27 (92), 1975, pp. 171-178.
BARON, J. and SAUTEREY, R., Le béton hydraulique, Paris, Presses de l'École nationale
des ponts et chaussées (1982), 560 pages.
BISSONNETTE, B. and PIGEON, M., "Tensile creep at early ages of ordinary, silica fume
and fiber reinforced concretes", Cement and Concrete Research, Vol. 25,
no. 5 (July 1995), pp. 1075-1085.
BISSONNETTE, B., "Le fluage en traction: un aspect important de la problématique des
réparations en béton", Thèse de doctorat, Université Laval (1996), 279 pages.
BISSONNETTE, B., PIERRE, P. and PIGEON, M., "Influence of key parameters on
drying shrinkage of cementitious materials", Cement and Concrete Research, Vol.
29, no. 10 (October 1999), pp. 1655-1662.
BISSONNETTE, B., VONFAY, K., “Tools for evaluating cracking sensitivity of repair
materials”, Improving the performance of repaired concrete structures, 3rd
Workshop, Quebec City, 2001.
BOILY, D., “Fluage en traction des bétons de réparation”, Mémoire de maîtrise, Université
Laval, 2003, 143 pages.
108
BROOKS, J.J., « Influence of mix proportions, Plasticizers and Superplasticizers, on Creep
and Drying Shrinkage of Concrete”, Magazine of Concrete Research, 41 (148),
1989, pp. 463-466.
BROOKS, J.J., Neville, A.M., “Creep and shrinkage as affected by admixtures and cement
replacement materials, Creep and Shrinkage of Concrete: Effect of materials and
environment, ACI SP-135, 1992, pp. 19-36
CARLSON, R.W., and READING, T.J., “Model Study of Shrinkage Cracking in Concrete
Building Walls”. ACI Structural Journal, Vol. 85, No.4, Jul. Aug.1988 pp. 395-404.
COUTHINO, A. de S., “The Influence of the Type of Cement on its Cracking Tendency”.
RILEM News Series No.5, December 1959.
DOMONE, P.L.,”Uniaxial Tensile Creep and Failure of Concrete”, Magazine of Concrete
Research, 26 (88), 1974, pp. 144-152.
EMMONS, P., VAYSBURD, A., “Factors affecting the durability of concrete repair: the
contractor’s viewpoint", Construction and Building Materials, Vol. 8, no. 1 (1994),
pp. 5-16.
EMMONS, P., VAYSBURD, A., “Performance Criteria for Concrete Repair Materials,
Phase I", Technical Report REMR-CD-47, US Army Corps of Engineers, April
1995, 113 pages.
EMMONS, P., VAYSBURD, A., POSTON, R., McDONALD, J.E., “Performance Criteria
for Concrete Repair Materials, Phase II – Field studies", Technical Report REMR-
CD-60, US Army Corps of Engineers, September 1998, 67 pages.
EMMONS, P., VAYSBURD, A., “How to make today’s repairs durable for tomorrow -
Corrosion protection in concrete repair”, Construction and Building Materials, Vol. 14,
no. 1 (2000), pp. 189-197.
FELT, E.J., “Resurfacing and Patching Concrete Pavement with Bonded Concrete,”
Proceedings of the Highway Research Board, 1956, pp. 444-469.
GIROUX, P., “Efficacité des adjuvants réducteurs de retrait et influence sur les propriétés
de base des bétons ordinaires et projetés”, Mémoire de maîtrise, Université Laval,
2006 (In press).
109
GILLESPIE, T., “New method to predict performance of concrete repair materials in the
field”, IRCI Meeting, March 1999.
HANSEN, W., "Drying shrinkage mechanisms in Portland cement paste", Journal of the
American Ceramic Society, Vol. 70, no. 5 (May, 1987), pp. 323-328.
ILLSTON, J.M., (1965), « The Components of Strain in Concrete under Sustained
Compressive Stress », Magazine of Concrete Research, 17 (50), pp. 21-28.
JESSOP, E.L., WARD, M.A. and NEVILLE, A.M., “Influence of water reducing and set
retarding admixtures on creep of lightweight aggregate concrete”, RILEM
International Symposium on Admixtures for Mortar and Concrete, Brussels,
Belgium, 1967, pp. 35-46.
LYNAM, C.G., “Growth and Movement in Portland Cement Paste”, Oxford University
Press, London, 1934, 139 pages.
METHA P.K. AND MONTEIRO P.J., “Concrete: Microstructure, Properties and Material”,
Department of Civil and Environmental Engineering, Univeristy of California at
Berkeley, 659p., McGraw-Hill, 2006
MEYERS, B.L., (1967), “Time-Dependent Strains and Microcracking of Plain Concrete”,
Thèse de doctorat, Cornell University, 86 p. (from Neville et al. 1983).
MOON, J. and WEISS, J., “Estimating residual stress in the restrained ring test under
circumferential drying” Cement and Concrete Composites, 2005, Article in Press.
MORENCY, M., VAYSBURD, A., VON FAY, K., BISSONNETTE, B., “Development of
a test method to evaluate cracking tendency of repair materials, Phase I Report”,
Research Report 2004-1, US Bureau of Reclamation, Denver, CO, March 2005.
MORENCY, M., VAYSBURD, A., VON FAY, K., BISSONNETTE, B., “Development of
a test method to evaluate cracking tendency of repair materials, Field study, Phase II
Report”, Research Report 2004-1, US Bureau of Reclamation, Denver, CO, May
2009.
MORGAN, D. R., “Discussion of the paper “Volume change on first drying of hydrated
Portland cement with and without admixtures by Feldman and Swenson”, Cement
and Concrete Research, Vol. 5, pp 533-538. 1975.
110
NEVILLE, A.M., “Recovery of Creep and Observations on the Mechanism of Creep of
Concrete”, Applied Science Research, Vol.9, 1960, pp. 71-84.
NEVILLE, A.M., KENNINGTON, H.W., “Creep of Aluminous Cement Concrete”,
Proceedings of the 4th
International Symposium on the Chemistry of Cement,
Washington DC, 1960, pp. 703-708.
NEVILLE, A.M., DILGER, W. and BROOKS, J.J., “Creep of Plain and Structural
Concrete”, Construction Press, Londres, 1983, 361 pages.
NEVILLE, A., “Properties of concrete” 4th
Edition, Longman Group Limited, Essex,
England, 1995, 844 pages.
NEVILLE, A. Propriétés des bétons, Paris, Éditions Eyrolles (2000), 804 pages.
ÖDMAN, S.T.A., “Effects of variations in volume, surface area exposed to drying and
composition of concrete on shrinkage”, International Colloquium on the Shrinkage
of Hydraulic Cement Concretes I, RILEM/CEMBUREAU, Madrid, Spain, 1968, 20
pages. (From Neville 1995)
POSTON, R., KESNER, K., EMMONS, P., VAYSBURD, A., “Performance Criteria for
Concrete Repair Materials, Phase II – Laboratory results", Technical Report REMR-
CS-57, US Army Corps of Engineers, April 1998, 76 pages.
PIGEON, M., and BISSONNETTE, B., "Tensile Creep and Cracking Potential of Thin
Concrete Repairs", Concrete International, Vol. 21, no. 11 (Novembre 1999), pp.
31-35.
PULLAR-STRECKER, P., “Corrosion damaged concrete assessment and repair", CIRIA,
London, 1987.
PLUM, D.R., “The behavior of polymer materials in concrete repair and factors influencing
selection", The structural engineer, United Kingdom,Vol. 68 No. 17, 1990 pp. 337-
346.
POWERS, T.C., “Mechanisms of shrinkage and reversible creep of hardened cement paste”
The structure of concrete and its Behaviour under load, Conference proceedings,
Cement and Concrete Association, London, England 1965, pp. 319-344.
111
ROPER, H., “Dimensional change and water sorption studies of ciment paste”, Symposium
on structure of Portland cement paste, Highway Research Board, Special report 90,
Washington DC, 1966, pp. 74-83.
RUETZ, W., (1968), « A Hypothesis for the Creep of Hardened Cement Paste and the
Influence of Simultaneous Shrinkage », The Structure of Concrete and its Behaviour
under Load, Comptes rendus de conference (1965), Londres, pp. 365-387.
SEE, H., ATTIOGBE, E. and MILTENBERGER, M., 2003, “Shrinkage cracking
characteristics of concrete using ring specimens, ACI Materials Journal, May-June
2003, pp. 239-245.
SELLEVOLD, E.J., “Shrinkage of concrete: Effect of binder composition and aggregate
volume fraction from 0 to 60%, Nordic Concrete Research, 1992, Publication No.11,
pp. 139-152.
THOMAS, F.G., (1937), « Creep of Concrete under Load », International Association of
Testing Materials, Comptes renduss de conference, Londres, pp. 292-294 (from
Neville et al. 1983).
TROXELL, G.E., RAPHAEL, J.M. and DAVIS, R.E., “Long-Time Creep and Shrinkage
Tests of Plain and Reinforced Concrete”, ASTM Proceedings 58, 1958, pp. 1101-
1120.
YOUNG, J.-F., et al., “Physical mechanisms and their mathematical description”, Creep
and Shrinkage of Concrete: Mathematical Modeling, Preprint of the fourth RILEM
International Symposium, Evanston, IL, 1986, pp. 44-78.
113
8 Appendix A
Figure 71: Creep Box test – Concrete, small size
Figure 72: Creep Box test – Concrete, medium size
L=1400mm, b=375mm
L=1400mm, b=375mm
L=1400mm b=375mm
L=900mm, b=300mm
L=900mm, b=300mm
L=900mm, b=300mm
114
Figure 73: Creep Box test – Concrete, large size
Figure 74: Baenziger block – Concrete mix
L=1050mm, b=300mm
L=1050mm, b=300mm
L=1050mm, b=300mm
L=1900mm, b=450mm
L=1900mm, b=450mm
L=1900mm, b=450mm
115
Figure 75: German Angle test, Laval University – Concrete
Figure 76: German Angle test, USBR – Concrete
Figure 77: Creep Box test – Mortar, small size
L=900mm, b=300mm
L=900mm, b=300mm
L=900mm, b=300mm
116
Figure 78: Creep Box test – Mortar, medium size
L=1400mm, b=375mm
L=1400mm, b=375mm
L=1400mm, b=375mm
117
Figure 79: Creep Box test – Mortar, large size
Figure 80: Crack mapping of Baenziger block – Mortar mix
L=1050mm, b=300mm
L=1050mm, b=300mm
L=1050mm, b=300mm
L=1900mm, b=450mm
L=1900mm, b=450mm
L=1900mm, b=450mm
118
Figure 81: German Angle test, Laval University – Mortar
Figure 82: German Angle test, USBR – Mortar
Figure 83: Creep Box test – Polymer Mortar, small size
L=900mm, b=300mm
L=900mm, b=300mm
L=900mm, b=300mm
119
Figure 84: Creep Box test – Polymer Mortar, medium size
Figure 85: Creep Box test – Polymer Mortar, large size
L=1900mm, b=450mm
L=1900mm, b=450mm
L=1900mm, b=450mm
L=1400mm, b=375mm
L=1400mm, b=375mm
L=1400mm, b=375mm
120
Figure 86: Crack mapping of Baenziger block – Polymer modified mortar mix
Figure 87: German Angle test, Laval University – Polymer Modified Mortar
Figure 88: German Angle test, USBR – Polymer Modified Mortar
L=1050mm, b=300mm
L=1050mm, b=300mm
L=1050mm, b=300mm