UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation into bond strength between EDCC/masonry Yan, Yuan 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_september_yan_yuan.pdf [ 3.43MB ]
Metadata
JSON: 24-1.0303138.json
JSON-LD: 24-1.0303138-ld.json
RDF/XML (Pretty): 24-1.0303138-rdf.xml
RDF/JSON: 24-1.0303138-rdf.json
Turtle: 24-1.0303138-turtle.txt
N-Triples: 24-1.0303138-rdf-ntriples.txt
Original Record: 24-1.0303138-source.json
Full Text
24-1.0303138-fulltext.txt
Citation
24-1.0303138.ris

Full Text

  INVESTIGATION INTO BOND STRENGTH BETWEEN EDCC/MASONRY  by Yuan Yan  B.E. (Civil Engineering), Southeast University, China, 2009  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  May 2016  © Yuan Yan, 2016ii   Abstract  In order to apply Sprayable Eco-Friendly Ductile Cementitious Composites (EDCC) as a thin overlay material for masonry building upgrade, this study aims at understanding one of the key issues of repair: bond strength between old structure and the new repair overlay. Several influencing factors on bond strength were investigated, including repair thickness, fiber addition, substrate properties, curing age and environment. Bond strength was measured by tensile pull-off and friction-transfer tests.  At the conclusion of the research, EDCC was able to achieve satisfactory bond strength provided sufficient penetrability into the substrate, even under field conditions and without curing. Fibers added into EDCC impact bond strength negatively, if they are oriented parallel to the interface as a result of manual casting or if there is a low fractal dimension of the substrate surface. Further, 56 days can be used as the maturity age of bond strength with EDCC overlay. In future applications, penetrability of EDCC overlays can be ensured through sufficient amount of superplasticizer or energy of casting. For example, EDCC with 150mm slump was able to satisfy standard bond strength requirement of concrete in the field, at the age of 45 days. Penetrability of EDCC overlay is of vital importance, since EDCC with low workability (0 slump) can’t achieve requirement of structural repair even under standard curing. To mitigate the negative effect of fiber addition on bond strength, higher substrate roughness and 3D fiber orientation can be of help, through proper surface roughness preparation and the use of spray methods (e.g., shotcrete) instead of hand application.  For further study, it is suggested that measures should be taken to obtain more pure bond strength values for simplification of analysis. Also surface roughness variation and long term properties of interface are worth investigation, once proper substrates are chosen for lab research.       iii   Preface   The thesis is original, unpublished, independent work by the author, Yuan Yan.  This research program is part of the collaborative project: Development of Sustainable Masonry Rehabilitation Technology (SMART) using EDCC. It is based on the material EDCC created by Dr. Nemkumar Banthia, and is funded by Brxton.     iv   Table of contents  Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iii Table of contents .......................................................................................................................... iv List of tables.................................................................................................................................. vi List of figures ............................................................................................................................... vii Acknowledgements ...................................................................................................................... ix Chapter 1: Introduction ............................................................................................................... 1 Chapter 2: Literature review ....................................................................................................... 3 2.1 Repair-substrate bond ................................................................................................... 3 2.2 Bond strength tests...................................................................................................... 12 2.3 Durability of bonded cement-based overlays ............................................................. 16 Chapter 3: Materials, specimens, and methods ....................................................................... 22 3.1 Overlay materials ........................................................................................................ 22 3.2 Substrate specimens .................................................................................................... 24 3.3 Methods ...................................................................................................................... 26 Chapter 4: Characterization of overlay and substrate ............................................................ 30 4.1 Compressive strength of overlay material .................................................................. 30 4.2 Substrate surface roughness characterization ............................................................. 31 Chapter 5: Bond strength under standard curing ................................................................... 36 5.1 Test program ............................................................................................................... 36 5.2 Summary of test results of 56-day curing ................................................................... 36 5.3 Influencing factors from repair material ..................................................................... 40 5.4 Influence of substrate .................................................................................................. 45 Chapter 6: Bond strength under field condition ...................................................................... 49 6.1 Test program ............................................................................................................... 49 6.2 Summary of test results .............................................................................................. 49 Chapter 7: Conclusions and recommendations ....................................................................... 52 Chapter 8: Future research and knowledge gap ...................................................................... 54 v   References .....................................................................................................................................55 vi   List of tables  Table 2.1 Adhesion of mortar A on concrete substrate versus saturation level of concrete substrate and slurry  ...................................................................................................................... 11 Table 2.2 Effect of cyclic freeze-thaw on the shear bond strength of different repair systems  ... 18 Table 2.3 Effect of cyclic temperature change on the shear bond strength of different repair systems  ......................................................................................................................................... 18 Table 3.1 Composition of Lafarge CSA A3000 Type GU normal Portland cement .................... 22 Table 3.2 Chemical analysis of Lafarge Type F fly ash  .............................................................. 22 Table 3.3 Mixture of ELKON densified silica fume .................................................................... 22 Table 3.4 Properties of fibers ........................................................................................................ 23 Table 3.5 Mix design and properties of EDCC............................................................................. 24 Table 3.6 Test program for bond strength evaluation ................................................................... 29 Table 4.1 Compressive strength of plain mortar cylinder............................................................. 30 Table 4.2 Compressive strength of EDCC cylinder...................................................................... 30 Table 4.3 Substrate roughness parameters .................................................................................... 34 Table 4.4 Schematic diagram of roughness parameters  (L: evaluation length; mean line: reference line decided with least squares method; d0(=4.25μm): sampling interval ) ................. 35 Table 5.1 Summary of tension test with 56-day standard curing ................................................. 37 Table 5.2 Summary of torsion test with 56-day standard curing .................................................. 38 Table 5.3 Tension test with or without surface treatment ............................................................. 47 Table 6.1 Tension test result under field condition with 20mm EDCC overlay ........................... 49   vii   List of figures  Figure 2.1 Interfacial Transition Zone between (a) overlay and substrate  (b) aggregate and bulk paste  ............................................................................................................................................... 3 Figure 2.2 Interface identification and analysis (a) SEM image of the interface (b) EDAX Microanalysis result  ....................................................................................................................... 4 Figure 2.3 Factors affecting bond quality between concrete substrate and repair material  ........... 6 Figure 2.4 Shear bond strength in repair (with SSD and air dry substrates) vs. compressive strength of new concrete (with and without silica fume addition)  ................................................. 7 Figure 2.5 Influence of the length of the repair system on the differential shrinkage-induced stress by a linear elastic analysis  .................................................................................................... 8 Figure 2.6 Mechanical tensile and shear bond between substrate and overlay  ........................... 10 Figure 2.7 View of the glass microfiber at the interface  .............................................................. 10 Figure 2.8  Effect of surface conditions on core pull-off failure stress of overlay repair system: S, smooth; N, needlegunned; Cr, cracked; D, dry; W, wet; M0▪4, sand/cement repair mortar with w/c of 0.4; M0▪5, sand/cement repair mortar with w/c of 0.5; MP1, polymer-modified mortar 1; MP2, polymer-modified mortar 2   ............................................................................................... 11 Figure 2.9 Different bond strength tests  ...................................................................................... 13 Figure 2.10 Bond strength failure envelope with smooth and rough substrate surface  ............... 13 Figure 2.11 Schematic of setup to test material applied to substrate (ASTM C1583/1583M-13) 15 Figure 2.12 Schematic of failure modes (ASTM C1583/1583M-13) ........................................... 15 Figure 2.13 Sources of misalignment in pull-off test  .................................................................. 16 Figure 2.14 The various parts of a composite system  .................................................................. 19 Figure 2.15 Internal and external causes towards structure deterioration  ................................... 19 Figure 2.16 Implementation of time-dependent effects in structural analysis   ............................ 21 Figure 3.1 Gradation of recycled sand in mortar .......................................................................... 23 Figure 3.2 Dimension of concrete substrate (a) picture from experiment (b) dimension diagram (all units in mm) ............................................................................................................................ 25 Figure 3.3 Dimension of Concrete Masonry Unit ........................................................................ 25 viii   Figure 3.4 Dimension of clay substrate (a) picture from experiment (b) dimensions (all units in mm) ............................................................................................................................................... 26 Figure 3.5 Specimens (a) before overlay casting (b) after overlay casting .................................. 27 Figure 3.6 Pull-off test (a) test schematic  (b) test photograph ..................................................... 28 Figure 3.7 Torsion test (a) test schematic  (b) test photograph ..................................................... 28 Figure 4.1 Working mechanism of MicroTrakII Stand Alone Laser Head LTC-120-20-SA  ..... 31 Figure 4.2 Laser profilometry system ........................................................................................... 32 Figure 4.3 Transmission characteristics of roughness and waviness profiles (1: roughness profile; 2: waviness profile; 1+2: primary profile; X: wavelength; Y: transmission, %; λc: cutoff length separating waviness and roughness; λs: cutoff length suppressing noise of measurement; λf: cutoff length suppressing long wave components)  ...................................................................... 32 Figure 4.4 Sample profile of (a) clay substrate (b) concrete substrate surface (blue curve: primary profile; red curve: waviness; green curve: roughness) ................................................................. 33 Figure 5.1 Test results of 56-day standard curing (a) diagram of failure modes (b) test record of failure modes ................................................................................................................................. 37 Figure 5.2 Spiral crack line in OS failure mode ........................................................................... 39 Figure 5.3 Pattern of crack development due to torsion  .............................................................. 40 Figure 5.4 Individual test results with failure modes .................................................................... 42 Figure 5.5 Wet blasting setup  ...................................................................................................... 45 Figure 5.6 Profile of clay brick substrate before and after wetblasting using laser profilometery (a) filtered profiles of treated clay surface (blue line: primary profile, red line: waviness profile, green line: roughness profile) (b) filtered profiles of original clay surface (c) Abbott curve of treated clay surface (d) Abbott curve of original clay surface ...................................................... 46 Figure 5.7 Tensile bond strength test with 20mm plain mortar overlay ....................................... 48 Figure 6.1 Tension test result with failure mode under field condition ........................................ 50       ix   Acknowledgements  I would like to express my sincere thanks to University of British Columbia, especially the Department of Civil Engineering, for providing this welcoming multicultural environment of research.  Personally, I would like to thank my supervisor Dr. Nemkumar Banthia. His dedication to research quality, open-mindness towards innovation, and respect to everyone in our group have inspired me through this study process. Meanwhile, I have been working closely with Dr. Cristina Zanotti and Salman, who offered me strong support and made possible my academic development.  I want to also offer my gratitude to all the people in our lab of construction materials, especially Mr. Bill Leung, Mr. Harald Schrempp, and Jane.  Finally, I want take the chance to thank my parents. Their selfless love has accompanied me through every stage of my life.   1   Chapter 1: Introduction    The province of British Columbia has a relatively high seismic hazard compared to most of other Canadian provinces. The probabilities of “structurally” damaging earthquakes occurring within the next 50 years in the two largest cities of BC: Vancouver and Victoria, are 12% and 21%, respectively [1], which have drawn the public attention to safety issues of school buildings. Among which unreinforced masonry buildings built around 1920s are at very high risk, due to masonry structure’s vulnerability towards natural disasters.    In 2004, the province started a seismic upgrade program for its at-risk schools, including numerous masonry buildings. With this increasing demand for restoration, a good repair material and robust performance of the repair system are of great importance. Cement-based repair materials, despite their popularity, usually involve the use of large quantities of cement to facilitate ease of application and development of a strong interfacial bond with the substrate. Unfortunately, production of cement is responsible for nearly 10% of global CO2 emissions [2]. One ton of cement production emits 0.97 ton of CO2 into the atmosphere [3] and requires over 4GJ of energy [4]. A primary challenge therefore is to reduce the cement content in repair materials by replacing it with industrial by-products such as fly ash, or by using binders that do not involve the use of Portland cement.    The introduction of Sprayable Eco-Friendly Ductile Cementitious Composites (EDCC) brings the following benefits to the repair system:  reduction of carbon footprint of cementitious repairs by replacing cement with industrial by-products such as fly ash, or other binders, that do not involve the use of Portland cement.  potential of using the spraying process for application.  improved repair performance including cracking resistance, toughness and ductility under static, quasi-static, fatigue, and impact loads by fiber reinforcement.   However there are numerous stringent performance requirements in repair materials and unless care is exercised in selecting and reinforcing repair materials, these may fail in service. Some of the requirements include dimensional stability, transport properties, strong interfacial bond, and high bond toughness.  2     Despite advantages as a repair material stated before, property of EDCC in terms of bond with concrete masonry remains a poorly understood topic.   In the present research, EDCC is used as a thin overlay repair material for masonry substrates. The research purpose is to evaluate bond strength between EDCC and masonry under different curing conditions. At the same time, influence of repair thickness and fiber addition on bond strength was assessed. Both the material EDCC and its mortar counterpart without fiber addition were used as thin repair overlays of 20mm and 30mm thickness, onto concrete and clay masonry substrates. Characterization of repair material and substrate were conducted through cylinder compressive strength test and laser roughness profilometry, respectively. Two series of experimental specimens were prepared. Their bond strengths in tension and torsion were tested after 56-day standard curing, and the bond strength in tension was examined at 28, 45, and 56 days under field condition. Conclusions and recommendations were drawn based on the investigation results on EDCC’s performance regarding its bond strength to masonry with different rheology, repair thicknesses, fiber contents, environmental conditions, development ages, and substrate properties. These outcomes will offer technical guidance on EDCC’s application in seismic retrofit or upgrade.   3   Chapter 2: Literature review  2.1 Repair-substrate bond  2.1.1 Interface in cementitious materials A major concern in repair system is the interface, as it is usually the weakest link in the composite system.  Research shows that interface between old concrete and new concrete develops a similar ITZ (interfacial transition zone) as in concrete, both in microstructural characteristics and morphology, as shown in Figure 2.1 [5]. Basically, they are both more porous and weaker than the bulk matrix. Reasons behind the differences lie in aggregate-cement reaction, wall effect, and bleeding.    (a)                                   (b) Figure 2.1 Interfacial Transition Zone between (a) overlay and substrate [5] (b) aggregate and bulk paste [6] 4   In some other research, the interfacial zone is distinguished based on Ca/Si ratio [7], as analyzed in Figure 2.2 with the help of Scanning Electron Microscope (SEM) and Energy Dispersive X-Ray Analysis (EDAX).   (a)   (b) Figure 2.2 Interface identification and analysis (a) SEM image of the interface (b) EDAX Microanalysis result [7]  5   2.1.2 Characterization of interface  Bond strength tests are widely used to characterize interface mechanical properties, which will be discussed in detail in later sections. However several issues exist with macro-scale bond strength tests, including difficulty of identifying failure location, lack of information on failure process, and limited interface failures among test results [8] [9]. Therefore, nano- indentation was proposed, which can offer detailed information about micromechanical properties of interface. Hardness and elastic modulus obtained from nano-indentation were combined with simulation to predict interfacial mechanical behavior under direct tension. This method is able to investigate local micromechanical properties of interface, instead of generating data from the weakest link of composite repair system [10]. DC electrical resistance measurement was also employed to investigate interface property. When carbon fiber epoxy-matrix composite was used to repair concrete substrate, apparent resistance in the fiber direction was increased by bond degradation due to temperature or stress effects [11].  2.1.3 Adhesion  Bonding usually starts with adhesion, the process through which two bodies are brought together and bonded. It is dominated by capillary suction transferring water from repair material to the substrate [12]. In capillary suction, penetration height (lp) of liquid is given by Washburn’s equation [13]: 𝑙𝑝 = √(𝑟𝛾𝐿 cos 𝜃2𝜂) 𝑡 where r is the radius of the pore, γL is the surface free energy of liquid, θ is the contact angle between liquid and solid, η is liquid viscosity, and t is the time of penetration. According to this equation, interface needs a lower contact angle, higher surface free energy and lower viscosity of repair material, and a higher mean pore radius in the substrate to produce higher penetration in repair. But γL should not exceed critical surface energy of the substrate solid to ensure substrate wettability by overlay [14]. Therefore a compromise is needed through surface preparation and careful selection of repair material [12].   6   2.1.4 Factors affecting bond quality Factors influencing bond quality are summarized in Figure 2.3 [15]. Some key factors include: surface roughness and composition of the substrate, mix-design and placement of the repair material, early traffic, fatigue and other environmental conditions [16]. Substrate cleanliness, prewetting, and microcracks are also listed as of high-degree influence [15].    Figure 2.3 Factors affecting bond quality between concrete substrate and repair material [15]  2.1.4.1 Properties of repair material For a fresh repair material, workability and compaction together define the ability of newly applied material to fill cavities and voids in the substrate surface, thus determining the effective contact area between two layers [17].  For the hardened repair material, more factors are taken into account regarding their influences on bond quality. Compressive strength has less significant impact as tensile strength in some cases [15] [17], or even considered to have negative impact on shear bond. As displayed in Figure 2.4, lower w/c ratio of overlay results in higher compressive strength but lower shear bond strength, when substrates in both Saturated Surface Dry (SSD) and air dry conditions are combined with overlay without silica fume (SF) addition [7].  7    Figure 2.4 Shear bond strength in repair (with SSD and air dry substrates) vs. compressive strength of new concrete (with and without silica fume addition) [7]  Influence of overlay tensile strength on shear resistance of interface (τRd) without reinforcement can be explained by the following equation: 𝜏𝑅𝑑 = 𝑐𝑎 × 𝑓𝑐𝑡𝑑 + 𝜇 × 𝜎𝑛 where ca is a constant determined by surface roughness, µ is friction coefficient determined by surface roughness and overlay characteristic strength, σn is the compressive stress due to external normal force, and fctd is design tensile strength of overlay material [18].  Permeability in the newly applied material is another factor. As both very impermeable and permeable overlays will negatively impact bond quality [17], which will also be explained later.  Polymer addition into repair mortars has been demonstrated to be beneficial for bond strength in both short term [19] and long term [20]. These polymers help reduce drying shrinkage and control crack development [17]. It is also suggested to add 10% - 15% silica fume by weight of cement in repair material as it helps reduce large pores and eliminates the growth of calcium hydroxide by pozzolanic reaction [7]. The same benefits were observed when adding fly ash to 8   the material [21] [22]. Expansive agents in new repair materials will increase bond strength too as they restrict drying shrinkage [22].  However silica fume increases menisci capillaries as well, which results in development of drying tensile capillary pressure. Plastic shrinkage cracking in paste with 85% Portland cement and 15% silica fume was found to be several times higher than plastic shrinkage cracking in cement paste without silica fume [23].  In terms of repair patch dimension, generally large repairs tend to crack more easily. As can be seen from Figure 2.5, tensile stress induced by shrinkage increases with repair patch length before the critical length is achieved, which is around 4000mm here. This conclusion is derived by analytical calculation [24]. However, the influence of thickness needs to be further analyzed.    Figure 2.5 Influence of the length of the repair system on the differential shrinkage-induced stress by a linear elastic analysis [24]  9   2.1.4.2 Substrate surface preparation Purpose of substrate surface preparation is to produce sound, clean and suitably roughened surfaces [25].  However, it is still not clear how surface roughness influences bond strength, especially their correlation between each other. There might be a threshold value, beyond which the surface roughness does not enhance bond strength any more. Further investigation is then recommended to verify this theory using shear bond test. Tensile bond strength mainly depends on the vertical anchorage in pores and voids as shown in Figure 2.6 picture (b), rather than on interlocking. Therefore a high interface roughness like that in Figure 2.6 picture (a) usually improves the shear bond strength [17].  A balance between soundness and roughness should be carefully considered. Selection of surface preparation technique should be balanced with minimal aggressiveness to the substrate. It is suggested to use less aggressive treatment for concrete substrate with low compressive strength and more aggressive treatments for higher quality concrete. Substrate surface tensile strength and surface roughness index are considered as variables in some recent studies to give the estimation of bond strength with the accuracy of around 0.5MPa [15]. On the other hand, increase in roughness could induce a high void fraction if microfibers impede the flow of mortar into the surface texture as shown in Figure 2.7, which further confirms the necessity to avoid aggressive surface treatments in some cases. Especially when there is workability issue with repair materials due to microfiber addition, a relatively smooth surface will be more desirable [26].  Figure 2.8 demonstrates this complexity of balancing surface roughness, soundness and fluidity of repair material. Here the roughest surface (denoted as cracked) gave the lowest failure strength in cases of dry surfaces, due to reduced tensile strength of unsound substrates. And with sound and dry surfaces, bond strength was increased substantially when water to cement ratio of repair mortar was increased from 0.4 to 0.5 [27].  10    (a)                                                         (b) Figure 2.6 Mechanical tensile and shear bond between substrate and overlay [17]   Figure 2.7 View of the glass microfiber at the interface [26]  11    Figure 2.8  Effect of surface conditions on core pull-off failure stress of overlay repair system: S, smooth; N, needlegunned; Cr, cracked; D, dry; W, wet; M0▪4, sand/cement repair mortar with w/c of 0.4; M0▪5, sand/cement repair mortar with w/c of 0.5; MP1, polymer-modified mortar 1; MP2, polymer-modified mortar 2  [27]  Surface moisture level is another factor of concern. A too dry surface may result in a heterogeneous and porous zone close to the interface due to water absorption by the substrate [17]. A too wet surface might bring in an interface zone with a high water to cement ratio. As is shown in Table 2.1: the adhesion is relatively weak for low saturation level (≤ 50%); it reaches peak values for saturation levels around 90%, with a decrease of strength above this range [28]. A Saturated Surface Dry (SSD) substrate is therefore often recommended [25][29].  Table 2.1 Adhesion of mortar A on concrete substrate versus saturation level of concrete substrate and slurry [28]  12   2.1.4.3 Bonding agents Generally speaking, bonding agents should be avoided as they induce additional interfaces, not to mention the low strength of some bonding agents like grouts. However, if overlay material has workability issue, bonding agents could help fill the open pores and cavities [17].  2.1.4.4 Placement and curing  Proper placement of repair overlays helps prevent air pockets. Curing after repair is significant for controlling early age shrinkage, such as plastic and autogenous shrinkage. Also since temperature will influence the speed and rate of hydration, the curing system should not be too cold or too hot, as low temperature slows down reaction while relatively high temperature causes fast reaction and thus cracking due to temperature differentials within the concrete or non-uniform hydration [30].  2.2 Bond strength tests 2.2.1  Introduction  Bond strength could be defined as the unit external force necessary to separate the two contacting bodies. It is a representative parameter in the mechanical response of the interface characterized by cohesion and friction. There are four stress combinations in bond strength tests, i.e., combined shear and compression test, direct shear test, combined tension and shear test, and tensile test, with some of them shown in Figure 2.9. Measured bond strength decreases in the order of combined shear and compression test, direct shear test, indirect tension test, and direct tension test [31].  13    Figure 2.9 Different bond strength tests [17]   Figure 2.10 Bond strength failure envelope with smooth and rough substrate surface [32] 14   2.2.2 Combined shear and compression test  Slant shear test is one of the most common bond strength tests, although it doesn’t represent practical stress conditions and has too many influencing factors. The governing mechanism here is the Mohr-Coulomb equation 𝜏𝑛 = 𝐶 + tan(∅)𝜎𝑛  where τn is shear stress along the bond plane, C is cohesion, which equals pure tension strength, and Φ denotes the friction angle, which signifies degree of irregularities of the interface. σn is the normal stress. The relationship between normal stress and shear stress is shown in Figure 2.10.   2.2.3 Direct shear test  Torsion test is one of the most common direct shear tests. Other pure shear tests include bi-surface shear test and push-out test.     2.2.4 Combined shear and tension  2.2.5 Tensile tests There are both direct and indirect tensile tests. Indirect tension tests include flexural test and splitting test. Figure 2.9 picture A and picture B show direct tensile tests commonly used on the site and in the lab, respectively. However, since most practical stress conditions at repair interfaces are combined shear and tension, pure tensile tests do not represent reality. It is noteworthy that failure stress in these tests represents bond strength only if failure occurs purely in the interface. Otherwise they only produce lower bounds of bond strength.   pull-off Test  There are many guidelines describing pull-off tests due to its popularity coming from low cost, simple procedure and feasibility in the field [33], including ASTM D7522, ASTM D4541, ASTM C1583, ICRI (2004)03739, ACI 503R, and EN 1542. ASTM D4541 describes pull-off test with coating, ASTM 7522 is for FRP-concrete bond, and pull-off test for concrete-overlay strength is measured by ASTM C1583. However ASTM C1583 suffers from disadvantages including small scale, a localized nature of failure, not representing the service condition of most interface (combined tension and shear), and difficulties in result interpretation [33]. Instruments 15   for this method include core drill, core barrel, steel disk, and tensile loading device as shown in Figure 2.11, with its potential failure modes in Figure 2.12.  Core diameter is usually 50mm, but could reach 95mm sometimes [34]. In ASTM C1583, core drilling depth of 10mm below the interface is required, while others require it be 15mm [25]. Clearly, different core drilling depths can influence stress distribution and hence the results [35].   Figure 2.11 Schematic of setup to test material applied to substrate (ASTM C1583/1583M-13)  Figure 2.12 Schematic of failure modes (ASTM C1583/1583M-13) 16   Major challenges in pull-off test are eccentricity of loading and damage during coring [17]. Eccentricity causes stress concentration and scatter of results [31], as it induces a distorted stress field. Misalignment comes from core axis inclination and load inclination as shown in Figure 2.13. Under a certain misalignment angle (around 4 degrees), misalignment does not yield significantly different stress fields or change the failure mode characteristics. An increase of 15% is suggested in pull-off tests to account for misalignment [36].    Figure 2.13 Sources of misalignment in pull-off test [36]   twist-off Test The twist-off test is similar to the pull-off test in setup and procedure. It is easy to apply, therefore also feasible for site applications [30]. Assuming the composite material to be brittle and elastic, governing equation for this test is shown as 𝜏 =16𝑇𝜋𝑑3 where T is the torsion in failure and d is the diameter of the core.   2.3 Durability of bonded cement-based overlays Concrete overlay repair failures are usually manifested in cracking, delamination, and spalling due to imposed tensile stress on the repair beyond its capacity to resist that force [37]. Almost 3/4 of concrete repair failures are attributed to the lack of durability, with the remaining attributed to structural failures [38].  17   To study repair durability, two levels of complexity were proposed [39].  Level one is described as interrelated phenomena causing degradation. This includes mechanical phenomenon like surface abrasion and overlay material placement [39]. Another cause of degradation is the physical phenomenon, including creep, freeze-thaw, shrinkage and thermal cracking [39]. According to Table 2.2, after 300 freeze-thaw cycles shear bond strengths of repair system decreased by different levels in five out six concrete repairs. One explanation for the inferior bond performance of resinous repair systems could be the trapped water vapor flow inside the existing concrete. In the meantime, 200 thermal cycles resulted in different levels of shear bond losses for five out of six repair systems based on Table 2.3, which had to do with thermal compatibility. It is clear from Table 2.3 that strength loss is highest in sand/cement mortar with cement bonding grout, and is lowest in polymer modified mortars like styrene butadiene rubber modified cementitious mortar and fiber reinforced acrylic modified cementitious mortar. The contrast results from cement based material’s large volume changes by shrinkage and extra water added for hydration, and reduced shrinkage of polymer modified material [40]. Chemical phenomenon in this level consists of acid attack, alkali-silica reaction, chloride induced rebar corrosion, carbonation induced rebar corrosion and sulfate attack [39]. Level two of durability study refers to the repair component and system as shown in Figure 2.14, which is typically considered as a chain made of five different parts: steel reinforcement, the old concrete, the new repair layer, the old concrete-steel interface, and the old concrete-repair interface. To summarize these factors, Figure 2.15 concludes on the deterioration process with internal and external causes. And differential shrinkage is regarded a critical influence on durability performance of overlay repaired structure, as stress induced by its differential volume changes have long been recognized as main problem of long term performance [41].    18   Table 2.2 Effect of cyclic freeze-thaw on the shear bond strength of different repair systems [40]   Table 2.3 Effect of cyclic temperature change on the shear bond strength of different repair systems [40]   19    Figure 2.14 The various parts of a composite system [39]   Figure 2.15 Internal and external causes towards structure deterioration [39]  20   An internal factor or concept getting increasing attention in recent years is compatibility between two layers of materials, as the balance in dimension, chemical, electrochemical, and permeability properties between old concrete and repair overlay [37].  Dimension compatibility refers to volume change differences between old concrete and new repair layer, related to drying shrinkage or thermal expansion independent of load and creep or modulus of elasticity depending on external load. Shrinkage is contraction of repair material on the removal of moisture to the environment resulting in tensile stresses in both repair and interface by constraint of substrate [37]. Thermal expansion causes stress development in the interface, and often leads to cracks in the overlay, at the bond line or in the substrate.  These cracks further contribute to durability problems by increasing moisture penetration and subsequent delamination. To reduce the effect of thermal expansion in epoxy resin, which has up to eight times greater coefficient of thermal expansion than concrete, aggregates are sometimes added into the epoxy mortar [42].  Chemical compatibility relates to alkali content, tricalcium aluminate (C3A) content, chloride content, reaction, and their transportation in the repair system. Here overlay should also be able to protect the composite structure from further contamination, but not encapsulate the system resulting in deterioration accumulation [37]. Electrochemical compatibility has to do with reinforced concrete repair and its corrosion cell. Physical and chemical differences between the repair area and surrounding concrete lead to increased corrosion of reinforcing steel in the surrounding zone around the repair. Methods of stopping corrosion include preventing oxygen from reaching cathode, and drying out the cell at any location between cathode and anode [37].  Permeability incompatibility might result in, for example, encapsulation of moisture leading to freeze-thaw damage eventually. Therefore, instead of selecting repair material with lowest permeability, permeability compatibility between repair and substrate should be carefully considered [37].  According to Figure 2.15, interface and compatibility are two material interactions affecting repair durability. Major cause of interface debonding is cracking of the overlay [43]  [44]. There are two origins of debonding: “mechanical origin” and “length-change” origin, where “mechanical origin” is the consequence of flexural straining of structure by applied loads, especially live loads [45] and “length change” results from shrinkage. Peeling effect in both 21   cases occurs after overlay cracking cuts the continuity of the stress transmission within overlay, and causes interface stresses encouraging debonding [46].    Figure 2.16 Implementation of time-dependent effects in structural analysis  [39]   To predict repair performance, Figure 2.16 incorporates time-dependent effects in structure analysis, starting from strains imposed by external loading and combined with effects of shrinkage or external gradients. Material properties are consistently updated to consider their evolution or degradation, and which will in turn influence properties of material or mechanical performance of repair system [39].   High Performance Fiber Reinforced Cementitious Composites (HPFRCC) like Engineered Cementitious Composites (ECC) has been demonstrated to be ideal repair material for structural and durable requirements of repair [38]. Potentially high interfacial shear stress in repair system can be minimized by the large inelastic tensile strain capacity of HPFRCCs [47]. Their other advantages in repair include high ductility in preventing fracture failure, low Young’s modulus in enhancing dimension compatibility with substrate, high delamination resistance [38], freeze-thaw resistance [48], and fatigue load resistance several orders of magnitude higher than a concrete repair patch [49].    22   Chapter 3: Materials, specimens, and methods  3.1 Overlay materials 3.1.1 Cement, fly ash, silica fume Materials in this research include Type GU normal Portland cement with a blaine fineness of 383m2/kg and a composition shown in Table 3.1, Type F fly ash with the chemical composition shown in Table 3.2, and ELKON densified silica fume with the bulk density of 158g/L, whose chemical components are displayed in Table 3.3.   Table 3.1 Composition of Lafarge CSA A3000 Type GU normal Portland cement Compound SiO2 Al2O3 Fe2O3 CaO MgO SO3 LOI @950 LOI @550 % (weight/weight) 20.2 4.7 3.3 63.7 0.7 3.1 2.7 0.9  Table 3.2 Chemical analysis of Lafarge Type F fly ash [50] Compound SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O TiO2 P2O5 LOI % (weight/weight) 55.53 23.24 3.62 10.9 1.22 0.24 2.83 0.76 0.68 0.1 0.54  Table 3.3 Mixture of ELKON densified silica fume Compound SiO2 Al2O3 CaO C Cl - Fe2O3 H2O K2O MgO Na2O P2O5 SO3 % (weight/weight) 95.2 0.16 0.47 1.75 0.10 0.09 0.68 0.59 0.29 0.17 0.11 0.27  3.1.2 Sand Sand used in the mortar was recycled sand with the bulk density of 1592kg/m3; its gradation is given in Figure 3.1.  23    Figure 3.1 Gradation of recycled sand in mortar  3.1.3 PVA and PET fibers Properties of fibers used are shown in Table 3.4.   Table 3.4 Properties of fibers Fiber type  Diameter (μm)  Length (mm) Specific Gravity (kg/m3) Tensile Strength (MPa) KuraRay RECS15 PVA fiber 40 8 1300 1600 RECRON 3S PET fiber 40 6 1000 588  3.1.4 Superplasticizer The high range water reducer used was a polycarboxylate-based high range water-reducing admixture ADVA® 195 from Grace.  3.1.5 Mix proportions Mix design and properties of EDCC are presented in Table 3.5. Plain mortar used in the following tests was of the same mix design except no fibers were added.    24   Table 3.5 Mix design and properties of EDCC Consitituents kg/m3 Properties Cement 385.6 W/CM 0.27 Fly ash 771.3 S/CM 0.375 Silica fume 77.1 FA/CM 0.625 Sand 462.8 SF/CM 0.0625 Water 333.2 SP/CM 0.0016207 PVA 13   PET 10   Superplasticizer 2    3.2 Substrate specimens 3.2.1 Concrete and clay substrates Concrete substrate used here composed of two concrete masonry units (CMU) and a 10mm wide mortar joint as presented in Figure 3.2, including locations of drilling cores. CMU has the nominal dimensions of 400mm×100mm×200mm, as depicted in Figure 3.3. Similarly, brick substrates consist of seven clay units of 200mm×100mm×50mm and six mortar joints of 10mm width, as can be seen in Figure 3.4.   25    Figure 3.2 Dimension of concrete substrate (a) picture from experiment (b) dimension diagram (all units in mm)   Figure 3.3 Dimension of Concrete Masonry Unit        80 125 80 120 400 410 200 10   (a) (b) 26    Figure 3.4 Dimension of clay substrate (a) picture from experiment (b) dimensions (all units in mm)  3.2.2 Mortar between masonry units Mortar used between concrete or clay units is a Standard Mortar Type S (ASTM C270, CSA A371), with a minimum 28-day compressive strength of 12.4 MPa, minimum water retention of 75%, and a maximum air content of 12%.   3.3 Methods 3.3.1 Mixing procedures and specimen preparation   A special mixing procedure was developed, in order to obtain a uniform distribution without losing material workability. This procedure was created based on ASTM C192/C192M [51, p. 192] and experimental revision based on compressive strength of cast cylinders. The detailed mixing procedure is shown below:  premix fly ash with silica fume  dilute superplasticizer with 1/4 mixing water  place cement in the mixing bowl, add 1/4 water and mix for one minute   place fly ash and silica fume in the mixing bowl, add 1/4 water and mix for three minutes  place dispersed fiber in the mixing bowl, mix for five minutes while adding 1/4 water  place sand in the mixing bowl, mix for one minute  add 1/4 water with superplasticizer, and mix for two minutes   65 10 515 (a) (b) 27    stop mixer for 30 seconds, while scraping down the material on the mixing blade, mix for another 10 minutes, until fibers are well dispersed.   (a)  (b) Figure 3.5 Specimens (a) before overlay casting (b) after overlay casting  Fresh overlay material was hand applied onto clean and damp substrate surface, as shown in Figure 3.5.  And the repair system was cured in lab condition (23.0±2.0⁰C, RH 98%) for 56 days or put under field condition (20.5⁰C, RH 71% ) for 28, 45, and 56 days, before six cores were drilled in each beam and bond strength test was conducted.  3.3.2 Bond test methods 3.3.2.1 Tension test Pull-off test was used to measure tensile bond strength as shown in Figure 3.6 according to ASTM C1583 [52].  Using the bond test apparatus Proceq DY-216, it was able to conduct the test at a constant loading rate of 0.02MPa/s.   28    (a)                                                     (b) Figure 3.6 Pull-off test (a) test schematic [52] (b) test photograph  3.3.2.2 Torsion test In order to measure shear bond strength, friction-transfer method [53] was selected as shown in Figure 3.7. Through the application of a torque on the overlay, the interface shear strength could be calculated as: 𝜏𝑚𝑎𝑥 =16𝑀𝑡𝜋𝑑3 where Mt is the maximum torque at failure, d (=50mm) is diameter of the drilling core, and 𝜏𝑚𝑎𝑥 is the calculated shear strength.   (a)                                                                        (b) Figure 3.7 Torsion test (a) test schematic [53] (b) test photograph      Substrate Repair overlay Partial core  Gripping device Torque meter 29   3.3.3 Test program Table 3.6 introduces different tests conducted in this research with their corresponding sections, parameters and variables. In all these tests, overlay material was cast onto mature substrate beams by hand application.   Table 3.6 Test program for bond strength evaluation Section 5.2&5.3 5.4.1 5.4.2 6.1&6.2 Curing method Standard1 Standard1 Standard1 Field2 Substrate Concrete &Clay3 Clay4 Concrete &Clay3 Concrete &Clay3 Substrate surface treatment None Wet blasting/None None None Overlay material, thickness (mm) EDCC5&PM, 20 &30 EDCC5, 20 PM, 20 EDCC6, 20 Test method Tension &Torsion Tension Tension Tension Age of overlay at testing (day) 56 28 7&56 28&45&56 Note: 1, moist curing at 23.0±2.0⁰C; 2, curing condition specified in section 6.1; 3, clay type characterized in Chapter 4; 4, clay type different from that of Chapter 4; 5, EDCC as designed in Table 3.5, with slump close to 0 by ASTM C143; 6, EDCC as designed in Table 3.5 except for 4kg/m3 superplasticizer and 150mm slump by ASTM C143   30   Chapter 4: Characterization of overlay and substrate  4.1 Compressive strength of overlay material  The compressive strengths of both EDCC and plain mortar were tested according to ASTM C39/C39M-15a [54] at the ages of three and seven days, with cylinder size of 10cm×20cm. Test results are shown in Table 4.1 and Table 4.2. Notice that the 3-day strength of EDCC is slightly lower than that of plain mortar, while its 7-day strength is 10.6% higher than plain mortar’s 30.1MPa. But the overall average coefficient of variance (COV) from two ages of EDCC (2.95%) is higher than that of plain mortar (2.3%).   Table 4.1 Compressive strength of plain mortar cylinder Cylinder No. Age (day) Compressive strength (MPa) Average compressive strength (MPa) Length (mm) Diameter (mm) Average  density (kg/m3) 1 3 21.8 22.3 201 100.50 1890 2 3 22.8 202 102.00 3 7 29.8 30.1 194 100.50 1890 4 7 30.4 202 101.50  Table 4.2 Compressive strength of EDCC cylinder Cylinder No. Age (day) Compressive strength (MPa) Average compressive strength (MPa) Length (mm) Diameter (mm) Average  density (kg/m3) 1 3 20.8 21.3 200 102.00 1870 2 3 21.2 201 102.00 3 3 22.0 201 102.00 4 7 33.8 33.3 199 99.75 1930 5 7 34.0 202 100.00 6 7 32.2 201 101.30 31    4.2 Substrate surface roughness characterization To understand the characterization of repair substrate, it is necessary to assess its surface profile so as to establish the mechanism of its interaction with overlay. Laser profilometry is chosen here for profile assessment. The device MicroTrakII stand-alone laser head LTC-120-20-SA [55] has been employed for surface scanning, whose testing distance is  120±10mm, resolution is ±2.5µm, and spot size is 100µm. Working mechanism of this laser sensor is shown in Figure 4.1. Displacement speed of the laser head is controlled by Dayton DC speed controller at 0.085mm/s. And the remote control software of MicroTrackII gives real-time distance measurements from substrate surface to the laser head. The entire testing system is shown in Figure 4.2. During the scanning process, each type of surface was scanned eight times in lengths of around 20mm every time. And these eight scanning sequences included both horizontal and vertical scans at around four points on the brick surface.   Figure 4.1 Working mechanism of MicroTrakII Stand Alone Laser Head LTC-120-20-SA [55]  32    Figure 4.2 Laser profilometry system   Figure 4.3 Transmission characteristics of roughness and waviness profiles (1: roughness profile; 2: waviness profile; 1+2: primary profile; X: wavelength; Y: transmission, %; λc: cutoff length separating waviness and roughness; λs: cutoff length suppressing noise of measurement; λf: cutoff length suppressing long wave components) [56]  Laser head Specimen Displacement of laser head LVDT Speed controller Real-time distance display 33   With the help of analysis software MountainsMap, the representative surface profiles of both substrates are shown in Figure 4.4, together with their roughness parameters presented in Table 4.3. From the extraction of raw surfaces to the profiles presented in Figure 4.4 and the parameters calculated in Table 4.3, filtration was conducted. It includes removing nominal tilt or form by least means square fitted line from raw profile, suppressing measurement noise through low pass Gaussian filtration at wavelength of λs, and separating waviness and roughness with branch point wavelength of λc (=2.5mm) as shown in Figure 4.3. Note that the cutoff length of λf  is also presented in Figure 4.3 but not applied here, since it’s rarely used in practice and poorly defined in ISO standards [56] [57] [58].     Figure 4.4 Sample profile of (a) clay substrate (b) concrete substrate surface (blue curve: primary profile; red curve: waviness; green curve: roughness)  Roughness parameters of Table 4.3 are presented to better understand profiles beyond the schematic information of Figure 4.4. Table 4.4 explains each parameter with diagrams. Being close to each other in values of first five parameters, clay and concrete surfaces are obviously different in Kurtosises and fractal dimensions (an index quantifying fractal complexity). With a Kurtosis larger than three with clay and smaller than three with concrete, the profile distinctions can be understood from the Kurtosis diagram in Table 4.4. Also notice that clay has a higher fractal dimension (1.11) than concrete (1.05). Fractal dimensions of this software are calculated by Koch box-counting method:  Length (a) (b) Length Height Height 34   D = log(𝑁(𝑠)) / log(1/𝑠) where D is fractal dimension, N is number of box segments covering profile line, and s is size of boxes [59]. Fractal dimension here is the major influencing factor. The reason lies in the fact that bond resistance consists of adhesive bonding, friction, aggregate interlocking and dowel action (bending resistance of steel connectors) [48]. The parameter Rz of both substrates in Table 4.3 are far lower than the required 1.5mm [18] to produce any aggregate interlocking, and overlay is too light for normal stress of significant friction. Thus fractal dimension here is the controlling parameter for adhesive bonding, since it signifies overall contact area between two layers of materials.  Table 4.3 Substrate roughness parameters Profile  category Parameter Clay  Concrete Average Variance (%)  Average Variance (%) Roughness Rp(mm) 0.14 28.1  0.16 11.6 Rv(mm) 0.14 35.5  0.13 23.8 Rz(mm) 0.28 31.5  0.29 15.3 Ra(mm) 0.05 29.2  0.06 19.0 Rq(mm) 0.07 27.5  0.07 17.5 Rku 3.45 24.7  2.90 16.1 D (Fractal dimension) 1.11 2.2  1.05 4.3        35   Table 4.4 Schematic diagram of roughness parameters [60] (L: evaluation length; mean line: reference line decided with least squares method; d0(=4.25μm): sampling interval ) Parameter Calculation illustration Rp: Maximum Peak Height of the     roughness profile  Rv: Maximum Valley Depth of the roughness profile Rz: Maximum Height of roughness profile Ra=total shaded area/L: Arithmetic Mean Deviation of the roughness profile  𝐑𝒒 = [𝒁𝟏𝟐+𝒁𝟐𝟐+𝒁𝟑𝟐+⋯+𝒁𝑵𝟐𝑵]𝟏/𝟐: Root-Mean-Square (RMS) Deviation of the roughness profile 𝐑𝒌𝒖 =1𝑅𝑞41𝑁∑ 𝑍𝑗4𝑁𝑗=1 : Kurtosis of the roughness profile        Rku<3 Rku=3 Rku>3  36   Chapter 5: Bond strength under standard curing  5.1 Test program Section 5.2 and 5.3 discuss results of the primary set of tests in this study, in order to understand various bond strengths under standard curing and discover influence of different parameters. Section 5.4.1 studies bond strength between 20mm EDCC and clay substrate, with and without substrate roughness treatment. Section 5.4.2 compares bond strengths of different substrates with same overlay, after 7-day and 56-day curing. All test programs are introduced in Section 3.3.3.   5.2 Summary of test results of 56-day curing     The tension and torsion bond strength test results for 56-day curing are given in Table 5.1 and Table 5.2. Figure 5.1 shows the modes of failure and the numbering scheme.  It also shows examples of actual failures in each mode. Purpose of testing bond strength is to ensure enough resistance to interface stresses, which usually arise from overlay shrinkage or external loading (e.g. live loads). These effects were limited here, since overlay was under standard curing without any damage arising from external loading until testing.  Bond strength in structural repair of concrete required by European Standard EN 1504-3 is minimum 2MPa in tensile [61]. According to the present test result, EDCC has achieved 93% of the required strength when cast onto concrete manually with thickness of 20mm. However it’s unfair to comment on the strength value until we consider damage induced by shrinkage and cumulative loading.   37    Figure 5.1 Test results of 56-day standard curing (a) diagram of failure modes (b) test record of failure modes  Table 5.1 Summary of tension test with 56-day standard curing  Substrate Overlay   Tension test Overlay material thickness (mm)  Average failure stress (MPa) No. of test cores STDV1 (MPa) COV2  (%) Failure modes (By area ratio9)  S3 O4 I5 (%) (%) (%) Concrete EDCC7 20  1.86 6 0.28 13.5 83 17 0 Concrete EDCC 30  1.52 6 0.36 12.9 42 42 16 Concrete PM8 20  2.10 6 NA 17.0 100 0 0 Concrete PM 30  1.85 6 0.29 12.9 75 17 8 Clay EDCC 20  1.31 6 0.17 21.5 33 0 67 Clay EDCC 30  1.36 6 0.15 21.0 25 50 25 Clay PM 20  1.62 6 0.09 16.2 92 0 8 Clay PM 30  1.32 6 0.22 19.9 83 17 0 Average      0.22 16.9      S OS I O (a) (b)  38   Table 5.2 Summary of torsion test with 56-day standard curing  Substrate Overlay   Torsion test Overlay material thickness (mm)  Average failure stress (MPa) No. of test cores COV  (%) Failure modes (By area ratio)   S O I OS6  (%) (%) (%) (%) Concrete EDCC 20  3.16 5 19.7 20 0 0 80 Concrete EDCC 30  3.58 4 20.0 25 0 0 75 Concrete PM 20  2.76 4 30.4 0 25 0 75 Concrete PM 30  2.65 4 21.8 0 25 0 75 Clay EDCC 20  1.77 4 40.4 0 0 63 37 Clay EDCC 30  2.62 5 27.4 0 0 20 80 Clay PM 20  2.36 6 43.7 0 0 17 83 Clay PM 30  1.51 4 58.4 0 33 0 67 Average      32.7      Note: 1, Standard Deviation; 2, Coefficient of Variation; 3, Substrate failure;  4, Overlay failure; 5, Interface failure; 6, Overlay-substrate inclined failure; 7, Eco-friendly Ductile Cementitious Composites; 8, Plain mortar; 9, Area ratio of the failure mode among total area of tested cores in the category  5.2.1 Precision of test  ASTM C1583 [52] gives the reference value of 0.29MPa for overall pooled standard deviation in tension (pull-off) test, in which individual standard deviation comes from replicate tests with the same failure mode for each condition. In this research, this value is 0.22MPa, as shown in Table 5.1. Since friction-transfer test for torsion bond strength is newly proposed, no reference value is available to assess its precision in testing bond strength.  Despite recommendation of ASTM C1583, it is considered inappropriate to use standard deviation to analyze precision, due to large differences in averages of different tests. However Coefficient of Variation (COV) is suggested instead, as it takes the average values into account [25]. COV of the present research varies from 12.9% to 21.5% in tension tests, with an overall value of 16.9%. This number is comparable to that implied by ASTM C1583, whose overall COV ranges from 14% to 20% by different equipment. COV of torsion (friction-transfer) test here has a higher average of 32.7%.  39    5.2.2 Significance of failure modes  Large variation in strengths and failure modes of bond test is a common occurrence in most of the researches [25]. Even though only interface failure gives real bond strength, other types of failure stresses offer lower bounds of interface strength which fail in the weakest planes of the partial cores. Overall, true bond strength should be higher than average of test results, since in some cases failure occurs in substrate or overlay before interface.     Figure 5.2 Spiral crack line in OS failure mode It’s easy to understand failure modes of S, O and I as depicted by the diagram in Figure 5.1, which imply substrate, overlay, and interface failure, respectively. However, the failure of OS going across overlay and substrate planes is worth mentioning here. Essentially, cracks in brittle materials like concrete start in tension, regardless of the external load direction. When torsion is applied to the partial core as in the present friction-transfer test, maximum normal stress will be induced at an angle of 45º, as can be explained in Figure 5.3 [30] [62]. As a result, crack in the core’s peripheral surface is a spiral line across both overlay and substrate, as displayed in Figure 5.2. But initiation point of the spiral line remains unknown, and might be either in the overlay or in the substrate.  40    Figure 5.3 Pattern of crack development due to torsion [30]  5.3 Influencing factors from repair material To further analyze test results, average failure strength for the same failure mode in each test condition is plotted in Figure 5.4, together with its variation range (from lowest to highest failure strength measured during tests) of that data set. In this method, each average data point stands for the lower bound of strength by that failure mode, as it was at that stress level the type of failure was initiated. It is also the minimum strength of certain failure mode that is of interest in quality management of repair bond. Beside average strength point of each failure condition in Figure 5.4 is the area ratio of that failure mode among all failures of the condition, i.e., area ratio of the failure mode among total area of tested cores in the category. The reason to use area ratio is that not every partial core failed in the same mode, and some of them had combined failure modes like half interface and half substrate failures. In that case the test core would account for 8.3% (0.5×1/6) interface failure and 8.3% substrate failure for a collection of six cores. Variables studied in the graphs are repair thickness (20mm and 30mm) and fiber addition (EDCC and plain mortar), as introduced before. Together with two test methods - tension and torsion, and two substrates – concrete and clay, 16 different test categories are presented in Figure 5.4. By doing a control variable comparison, effects of repair thickness and fiber addition can be summarized. 41    (a) Tensile test result with concrete substrate    (b) Tensile test result with clay substrate   42    (c) Torsion test result with concrete substrate    (d) Torsion test result with clay substrate  Figure 5.4 Individual test results with failure modes   43   5.3.1 Influence of repair thickness    Two repair thicknesses, 20mm and 30mm were used, with the project targeted at thin overlay repair. Comparing bond strengths and their failure modes between 20mm and 30mm repairs gives an idea of the influence of thickness.  In tensile tests of concrete and clay repair, as shown in Figure 5.4 (a) and (b), 30mm EDCC gives rise to a marginally lower overlay failure stress than that of 20mm. In concrete repair, 30mm EDCC starts to fail at 1.56MPa in average with overlay failure accounting for 42% of all failure modes in terms of area ratio. Whereas 20mm overlay doesn’t produce overlay failure until the stress of 1.83MPa. This trend was similar in that of clay repair: 30mm EDCC starts to fail at 1.55MPa on an average, with a 50% area ratio among all failures of that condition, while 20mm EDCC has no overlay failure even though the applied load goes through similar range as 30mm. Negative influence of increasing EDCC thickness on overlay might result from higher chances of defects or deficiencies in the overlay, if complete compaction was not provided in the present research, as overlay material was cast manually without vibration. However, influence of EDCC thickness on bond strength is not statistically significant from results of tensile tests. In concrete repair of Figure 5.4 (a), 30mm EDCC has pure interface failure at 1.51MPa in average, when 20mm EDCC doesn’t have any interface failure during the test. However from Figure 5.4 (b), clay repair shows similar bond strengths in both thicknesses, which are 1.19MPa in 20mm and 1.18MPa in 30mm. As a result, no clear trend is discovered on influence of repair thickness on bond strength in tensile direction.   In torsion tests, as shown in Figure 5.4 (c), EDCC thickness seems to have no significant influence on overlay or bond strength of concrete repair. According to the graph, torsion tests of concrete repair with 20mm and 30mm EDCC have shown similar failure modes and strengths: both have around 20% substrate failure and 80% OS failure, and average stresses of those failures are close to each other. Torsion test in clay repair, described in Figure 5.4 (d), however shows a different phenomenon. 20mm EDCC repair starts to fail in interface at an average stress of 1.49MPa, with two out of four test cores purely failing in the interface. In the meantime, 30mm EDCC starts to fail in interface at 2.82MPa averagely, and its interface failures are all combined with OS failures for each test core.  44    In conclusion, increasing repair thickness of EDCC seems to have negative influence on overlay quality in tensile direction. But its effect on overlay in torsion or bond strength is not significant.  5.3.2 Influence of fiber addition As introduced before, difference between EDCC and plain mortar is that EDCC has 2% fiber addition while plain mortar uses the same mortar as EDCC without fiber. Comparing test results from EDCC and plain mortar repairs, with all the other conditions remaining the same, will reveal influence of fiber addition on repair quality. From the results of tensile tests in Figure 5.4 (a) and (b), fiber addition mostly reduced the lower bound of bond strength between repair overlay and substrates, especially when overlay is 30mm thick. In concrete repair (Figure 5.4 (a)), interface with 30mm EDCC starts to fail at 1.51MPa on an average, while plain mortar starts at 1.71MPa. Similarly, interface with 30mm EDCC in clay repair starts to fail at 1.18MPa, while that of plain mortar doesn’t have interface failure around the stress range, according to results of Figure 5.4 (b). No comparisons in the context of fiber addition in 20mm repair with concrete substrate can be made, as neither EDCC nor plain mortar had interface failure. In 20mm repair of clay, EDCC and plain mortar have close lower bounds of interface strengths, which are 1.19MPa and 1.16MPa, respectively, according to Figure 5.4 (b). In summary, fiber addition plays mostly a negative role in tensile bond strength based on present results. This might be caused by weak bond strength between fiber and matrix, unfavorable fiber orientation which is likely parallel with interface plane due to manual casting process, or more drilling damage to EDCC bond due to its higher toughness than plain mortar. In other cases fiber may be able to strengthen the interface [63], but it is not happening here as no significant overlay shrinkage happened under standard curing and both substrate surfaces have fractal dimensions no higher than 1.11. Influence of fiber addition on overlay tensile strength is not obvious. Fibers seem to worsen overlay quality in concrete repair, as EDCC have lower overlay failure stresses than plain mortar in both 20mm and 30mm repairs based on Figure 5.4 (a). But there are opposite trends in clay repair, as 30mm EDCC has overlay failure stress starting at 1.55MPa while 30mm plain mortar has a lower value of 1.32MPa. Unlike interface of the present case where substrate surface is relatively smooth and fibers are not able to penetrate and help increase bond strength, fibers in overlay have the potential to improve its tensile 45   strength when orientated relatively vertically. But since the failure stress of overlay depends largely on strength of the weakest plane, different trends are shown due to random fiber orientations.    According to Figure 5.4 (d), EDCC with both thicknesses gave rise to lower interface failure stresses than plain mortar. 20mm EDCC starts to have bond failure at 1.49MPa with that of plain mortar being at 1.84MPa, and 30mm EDCC has bond failure at 2.48MPa, when 30mm plain mortar shows none interface failure. Reasons behind this trend might be similar as that in tensile tests. Theoretically EDCC should demonstrate higher strength in shear than plain mortar, with strengthening effect of fibers perpendicular to 45º crack direction. But it’s hard to draw that conclusion based on the current result, since calculating ratio of overlay failures in OS failure modes is difficult and can be inaccurate.   5.4 Influence of substrate  With further tensile bond tests on brick repair systems with before and after surface treatment, correlation between fractal dimension and bond strength was verified.   5.4.1 Surface roughness treatment on clay bricks  Figure 5.5 Wet blasting setup [64] For the sake of convenience, wet blasting was chosen as the substrate surface treatment method, as illustrated in Figure 5.5. A Karcher 3000psi commercial pressure washer was used with glass beads, where sandblasting was simulated with water pressure instead of air pressure. The purpose of this treatment is to produce a rough surface, without damaging mortar joints or 46   causing microcracks in bricks. The sample surface was pressure washed for five minutes, as mortar joints started to crack beyond this limit.          Figure 5.6 Profile of clay brick substrate before and after wetblasting using laser profilometery (a) filtered profiles of treated clay surface (blue line: primary profile, red line: waviness profile, green line: roughness profile) (b) filtered profiles of original clay surface (c) Abbott curve of treated clay surface (d) Abbott curve of original clay surface                                                                                   (c) Length Length Density  Length Density Height Height Height Height (a) (b) (d) 47   Table 5.3 Tension test with or without surface treatment Substrate Overlay material Treatment Bond strength (MPa) STDV (MPa) COV  (%) Fractal  dimension  Failure mode S O I (%) (%) (%) Clay EDCC None 1.78 0.44 25 1.36 0 0 100 Wet  blasting 1.47 0.38 26 1.31 0 25 75  Overlay with the thickness of 20mm was applied to clay beams (please note: this is not identical to clay used in previous 56-day standard curing tests) with and without wet blasting, in order to study the effect of roughness profile on bond strength. Tensile bond strength test was conducted on the composite specimens after 28 days of standard curing. Results are shown in Figure 5.6 and Table 5.3, together with fractal dimensions measured by laser profilometry. According to Figure 5.6 (a) and (b), treated clay substrate is for most part a smoother surface than that of untreated surface, with a relatively uniform distribution of hills and valleys along one scanned line in untreated surface. This is further verified by the Abbot curves in Figure 5.6 (c) and (d), which denote cumulative probability densities of the surface profiles’ heights. As shown in Table 5.3, wet blasting wasn’t able to increase the tensile bond strength; neither did it help increase complexity of surface roughness indicated by fractal dimension. Bond strength consists of adhesive bonding, friction, aggregate interlocking and dowel action [18]. Among these mechanisms only adhesive bonding worked here, as overlay material was too light for significant friction and aggregate interlocking is of little effect in tensile bond. Adhesive bonding is positively correlated to surface contact area, which can be represented by fractal dimension [65]. Here with same overlay and substrate material, lower fractal dimension brought in by wet blasting caused lower bond strength. Further treatment is not suggested, since mortar joint starts cracking afterwards. As a result of this trial, surface treatment was not applied in other tests of this research.  48   5.4.2 Differences between concrete and clay substrates  Figure 5.4 shows that with the same overlay, the lower bound of bond strength in clay is lower than that of concrete in most of the cases. This could be further demonstrated by Figure 5.7, giving tensile bond strengths between plain mortar and different substrates of different ages.   Figure 5.7 Tensile bond strength test with 20mm plain mortar overlay  As introduced in Section 0, concrete surface has a lower fractal dimension of 1.05, comparing to the value of 1.11 for clay surface. Since only adhesive bonding works here for interface bonding as explained in Section 5.4.1, there might be differences in unit area adhesions to overlay or water absorptions between these two substrates. To further demonstrate this, a closer look into the interface or water absorption ability should be conducted in the future.  Furthermore, it is shown in Figure 5.7 that tensile bond strength between clay and plain mortar develops from 0.49MPa in seven days to 1.16MPa in 56 days, which means 7-day bond strength is around 42% of the mature bond strength under standard curing.    49   Chapter 6: Bond strength under field condition  6.1 Test program This set of specimens was prepared with 20mm EDCC manually cast onto clay and concrete substrates, and cured under field condition for 28, 45, and 56 days. Superplasticizer added to EDCC was 4kg/m3, i.e. double the amount used in other tests of this study, bringing in a more workable overlay. Average field temperatures for 28-day, 45-day, and 56-day curing were 20.8ºC, 20.4ºC, and 20.3ºC, respectively. Average relative humidity for field condition was 71.13%.  At the end of curing period, tensile bond tests were conducted on these composite beams with pull-off method.   6.2 Summary of test results  Table 6.1 Tension test result under field condition with 20mm EDCC overlay Substrate Age (days) Average failure stress (MPa) STDV (MPa) COV  (%) Failure mode S  (%) O  (%) I  (%) Concrete 28 1.06 0.30 41.3 53 0 47 45 2.17 0.50 25.9 33 24 43 56 1.88 0.26 14.8 67 7 27 Clay 28 1.18 0.13 9.9 53 0 47 45 1.69 0.36 33.7 88 0 12 56 1.95 0.47 23.9 100 0 0 Average   0.33 24.9     6.2.1 Precision of test  According to Table 6.1, the present tests have an overall standard deviation of 0.33MPa and COV of 24.9%, comparing to the values of 0.22MPa and 16.9% from standard curing tests. Even though data from field condition has higher average COV than that of standard curing, it is 50   understandable since temperature and relative humidity of the field vary a lot within the period of curing time.   Figure 6.1 Tension test result with failure mode under field condition  6.2.2 Bond strength development under field condition  In Figure 6.1, each data point stands for average strength from same failure mode of each test condition, which represents the lower bound strength of that repair plane as introduced in previous standard curing tests. Interface failure, as shown in 28-day and 45-day tests on both substrates, brought lowest average strength among all failure modes. Based on the result of Figure 6.1, there are some interface failures in 28-day tests of both concrete and clay repairs. And bond failure strength of clay was higher than that of concrete. This is opposite to previous standard curing tests, when overlay with lower slump was used. It could result from different water absorptions between concrete and clay substrates with different workability in overlay, or different fractal dimensions of two substrate surfaces.  Some interface failures occurred at each test age of concrete masonry. Strength development is represented by the trend line of Figure 6.1.  From 28 days to 45 days, the minimum bond strength between EDCC and concrete increased by 156%, from 0.74MPa to 51   1.90MPa. Increase rate of the same interface slowed down from 45 days to 56 days, which was 11.5%. On the 56th day, the strength development almost came to a plateau. This is in accord with the previous assumption that bond strength between EDCC and substrates achieves maturity at around the age of 56 days. The continuous increase of bond strength, despite shrinkage under field condition from 28 days to 56 days, could be explained by pozzolanic reaction of 60% fly ash replacement. Development trend in bond strength with clay substrate shows there was an increase in bond strength beyond 28 days.  Adhesion was going up with time and interface failures were reduced with time. 52   Chapter 7: Conclusions and recommendations   The objective of research performed here is to investigate static, short term bond strength between EDCC and masonry, and offer recommendations for seismic upgrade with EDCC as an overlay in terms of the quality of its bond with the substrate. As stated before, EDCC is environmental friendly, and excels in both strength and toughness as a 2% fiber reinforced, fly ash based mortar. To achieve the intended performance of EDCC, its bond strength to existing masonry substrates should be ensured so that the upgraded composite system will act as an integrated piece. In combination with standards of structural repair, the following conclusions and recommendations based on present research results will further help with the ultimate goal of applying EDCC in seismic upgrade as a thin overlay.  1. EDCC is able to achieve satisfactory bond strength provided enough penetrability (the extent to which overlay penetrate) into substrate, even under field condition without curing. Fibers added into EDCC may impact the bond strength negatively under standard curing, if they become predominantly oriented parallel to the surface as was found to occur with manual casting methods and low fractal dimension of substrate surface.  2. The maturity age of bond strength with EDCC overlay can be assumed to be 56 days. Substantial bond strength increase between 28 and 56 days was observed for both standard-cured and field-cured EDCC on both concrete and clay brick surfaces.  This is believed to be due to continued hydration of the high fly ash content and its effect on the adhesive bond.  3. In the first part of this study, EDCC was used with 2kg/m3 superplasticizer (0 slump by ASTM C143) and the composite system was cured under standard condition. Numerically, EDCC achieved highest tensile bond strength for concrete masonry units with the thickness of 20mm. This value is 93% of that required by EN 1504-3 for structural repair. Additionally, 30mm EDCC showed lower tensile strength in overlay than 20mm EDCC, probably resulting from higher demand of compaction with thicker overlay.  In the meantime, bond strength decreased in both tension and torsion tests as fibers were added into overlay with EDCC. Several factors might have contributed to this effect, including weak bond strength between fiber and matrix, potentially unfavorable fiber orientation which is in parallel with 53   interface plane due to manual casting process and low fractal dimension of substrate, or more drilling damage to EDCC bond considering its higher toughness than plain mortar. 4. The case is different when EDCC was applied with 4kg/m3 superplasticizer (slump is 150mm by ASTM C143), and kept under field condition without curing in the second series of tests. EDCC with 20mm thickness acquired bond strength as high as 2.17MPa in concrete repair, and 1.95MPa in clay repair. Both of the values are higher than that of the first part in this study, because of higher slump obtained through doubling the superplasticizer content. Especially with concrete repair, EDCC was able to meet the standard requirement of 2.0MPa by EN 1504-3. The importance of sufficient rheology for overlay to penetrate into substrate is therefore self-evident. Additionally, according to test results under field condition, it is verified that bond strength development with EDCC matures in about 56 days after casting, which can be explained by pozzolanic reaction within the overlay material.   Based on the above conclusions, recommendations are made for applying thin EDCC overlay in seismic upgrade: 1. Penetrability of EDCC overlay should be ensured, through sufficient amount of superplasticizer or energy of casting (e.g. shotcrete). For example, EDCC with 150mm slump is able to satisfy standard bond strength requirement of concrete in the field, at the age of 45 days. Penetrability of EDCC overlay is of vital importance, since EDCC with low workability (0 slump) can’t achieve requirement of structural repair even under standard curing. 2. Overlay thickness should not yet be a factor in consideration of bond strength, but rather an influencing factor in overlay quality if compaction is an issue. After all, overlay thickness needs to be chosen based on structural need and economical consideration, combined with its effect on deformation related to shrinkage, if possible. 3. To mitigate negative effect of fiber addition on bond strength, higher substrate roughness and 3D fiber orientation can help, through surface roughness preparation and shotcreting.   54   Chapter 8: Future research and knowledge gap  Failure mode is always an issue in bond strength test. Since interface failure does not always happen, it’s difficult to assess real bond strength of the repair system. Future investigation could use high strength substrates instead and test the bond in an earlier age (if strength development can be reasonably predicted), obtain the real bond strength, and make variable analysis simpler.  Since the critical factor for EDCC to achieve satisfactory bond strength is its penetrability into substrate, further experiment can be done on EDCC repair by using higher amount of superplasticizer (> 2kg/m3) or better compaction methods (table vibration, shotcrete, etc), in combination with proper curing of field condition. EDCC is supposed to show great bond strength capability and completely satisfy standard requirement with this adjustment. Actually, spraying EDCC is already under investigation by some researchers. Problem related to this process is to successfully spray this fiber reinforced material with a portable sprayer, for the technology to be applied in industry at a larger scale. And as usual, rheology design of the spray process needs to be done with any kind of newly developed shotcrete material.    Finally, the long term interface property with EDCC should also be studied, including the change in strength or chemical composition after freeze-thaw cycles or other deterioration emanating from deleterious environment.           55   References [1] T. Onur and M. R. Seemann, “Probabilities of significant earthquake shaking in communities across British Columbia: implications for emergency management,” in Proceedings of the 13th World Conference in Earthquake Engineering, Vancouver, BC, Canada, Paper, 2004, vol. 1065. [2] D. Gielen, F. Boshell, and D. Saygin, “Climate and energy challenges for materials science,” Nat. Mater., vol. 15, no. 2, pp. 117–120, Jan. 2016. [3] E. Taşçı, B. Yılmaz, and I. Işık, “The characterization of blended cement hydration with thermal analysis,” J. Aust. Ceram. Soc. Vol., vol. 52, no. 1, pp. 47–55, 2016. [4] A. Atmaca and R. YumrutaA, “Analysis of the parameters affecting energy consumption of a rotary kiln in cement industry,” Appl. Therm. Eng., vol. 66, no. 1–2, p. 435, May 2014. [5] M. Pigeon and F. Saucier, “Durability of repaired concrete structures,” in Proceedings, International Symposium on Advances in Concrete Technology, 1992, pp. 741–773. [6] P. K. Mehta and P. J. M. Monteiro, Concrete: structure, properties, and materials. Prentice Hall, 1993. [7] H.-C. Shin and Z. Wan, “Interfacial properties between new and old concretes,” in Second international conference on sustainable construction materials and technologies, 2010. [8] H. Beushausen and M. G. Alexander, “Bond strength development between concretes of different ages,” Mag. Concr. Res., vol. 60, no. 1, pp. 65–74, 2008. [9] M. Luković, B. Šavija, H. Dong, E. Schlangen, and G. Ye, “Micromechanical study of the interface properties in concrete repair systems,” J. Adv. Concr. Technol., vol. 12, no. 9, pp. 320–339, 2014. [10] B. Š. Mladena Lukoviü, “Micromechanical study of the interface properties of concrete repair systems,” 2013. [11] Z. Mei and D. D. L. Chung, “Effects of temperature and stress on the interface between concrete and its carbon fiber epoxy-matrix composite retrofit, studied by electrical resistance measurement,” Cem. Concr. Res., vol. 30, no. 5, pp. 799–802, 2000. [12] L. Courard and R. Degeimbre, “A capillary action test for the investigation of adhesion in repair technology,” Can. J. Civ. Eng., vol. 30, no. 6, pp. 1101–1110, 2003. [13] E. W. Washburn, “The dynamics of capillary flow,” Phys. Rev., vol. 17, no. 3, p. 273, 1921. [14] L. Courard, “Evaluation of thermodynamic properties of concrete substrates and cement slurries modified with admixtures,” Mater. Struct., vol. 35, no. 3, pp. 149–155, 2002. [15] L. Courard, T. Piotrowski, and A. Garbacz, “Near-to-surface properties affecting bond strength in concrete repair,” Cem. Concr. Compos., vol. 46, pp. 73–80, 2014. [16] A. Garbacz, L. Courard, and B. Bissonnette, “A surface engineering approach applicable to concrete repair engineering,” Bull. Pol. Acad. Sci. Tech. Sci., vol. 61, no. 1, pp. 73–84, 2013. 56   [17] B. Bissonnette, L. Courard, D. W. Fowler, and J.-L. Granju, Bonded Cement-based Material Overlays for the Repair, the Lining Or the Strengthening of Slabs Or Pavements: State-of-the-art Report of the RILEM Technical Committee 193-RLS, vol. 3. Springer Science & Business Media, 2011. [18] N. Randl, “Design recommendations for interface shear transfer in fib Model Code 2010,” Struct. Concr., vol. 14, no. 3, pp. 230–241, 2013. [19] T. M. El-Rakib, A. M. Farahat, W. M. El-Degwy, and H. H. Shaheen, “Shear transfer parameters at the interface between old and new concrete,” Proc. ICPCM, 2003. [20] C. Atzeni, L. Massidda, and U. Sanna, “Dimensional variations, capillary absorption and freeze-thaw resistance of repair mortars admixed with polymers,” Cem. Concr. Res., vol. 23, no. 2, pp. 301–308, 1993. [21] G. Xiong, J. Liu, G. Li, and H. Xie, “A way for improving interfacial transition zone between concrete substrate and repair materials,” Cem. Concr. Res., vol. 32, no. 12, pp. 1877–1881, 2002. [22] G. Li, “A new way to increase the long-term bond strength of new-to-old concrete by the use of fly ash,” Cem. Concr. Res., vol. 33, no. 6, pp. 799–806, 2003. [23] M. D. Cohen, J. Olek, and W. L. Dolch, “Mechanism of plastic shrinkage cracking in portland cement and portland cement-silica fume paste and mortar,” Cem. Concr. Res., vol. 20, no. 1, pp. 103–119, Jan. 1990. [24] J. Zhou, K. Van Breugel, G. Ye, TU Delft: Civil Engineering and Geosciences: Design and Construction, and TU Delft, Delft University of Technology, “Performance of Engineered Cementitious Composites for Concrete Repairs.” 11-Jan-2011. [25] A. M. Vayburd and J. E. McDonald, “An evaluation of equipment and procedures for tensile bond testing of concrete repairs,” DTIC Document, 1999. [26] A. Garbacz, M. Górka, and L. Courard, “Effect of concrete surface treatment on adhesion in repair systems,” Mag. Concr. Res., vol. 57, no. 1, pp. 49–60, 2005. [27] P. J. Robins and S. A. Austin, “A unified failure envelope from the evaluation of concrete repair bond tests,” Mag. Concr. Res., vol. 47, no. 170, 1995. [28] L. Courard, J.-F. Lenaers, F. Michel, and A. Garbacz, “Saturation level of the superficial zone of concrete and adhesion of repair systems,” Constr. Build. Mater., vol. 25, no. 5, pp. 2488–2494, 2011. [29] F. Saucier and M. Pigeon, “Durability of new-to-old concrete bondings,” ACI Spec. Publ., vol. 128, 1991. [30] K. Neshvadian Bakhsh, “Evaluation of Bond Strength between Overlay and Substrate in Concrete Repairs,” 2010. [31] A. Momayez, M. R. Ehsani, A. A. Ramezanianpour, and H. Rajaie, “Comparison of methods for evaluating bond strength between concrete substrate and repair materials,” Cem. Concr. Res., vol. 35, no. 4, pp. 748–757, 2005. [32] S. Austin, P. Robins, and Y. Pan, “Shear bond testing of concrete repairs,” Cem. Concr. Res., vol. 29, no. 7, pp. 1067–1076, 1999. [33] O. R. Mata and R. A. Atadero, “Evaluation of Pull-Off Tests as a FRP–Concrete Bond Testing Method in the Laboratory and Field,” Pract. Period. Struct. Des. Constr., 2013. [34] D. Beaupré, “Bond Strength of Shotcrete Repair.” 57   [35] S. Austin, P. Robins, and Y. Pan, “Tensile bond testing of concrete repairs,” Mater. Struct., vol. 28, no. 5, pp. 249–259, 1995. [36] L. Courard, B. Bissonnette, A. Garbacz, A. Vaysburd, K. von Fay, G. Moczulski, and M. Morency, “Effect of misalignment on pull-off test results: numerical and experimental assessments,” ACI Mater. J., vol. 111, no. 2, 2014. [37] P. H. Emmons and A. M. Vaysburd, “Factors affecting the durability of concrete repair: the contractor’s viewpoint,” Constr. Build. Mater., vol. 8, no. 1, pp. 5–16, 1994. [38] V. C. Li, “High performance fiber reinforced cementitious composites as durable material for concrete structure repair,” Int. J. Restor., vol. 10, no. 2, pp. 163–180, 2004. [39] Cristina Zanotti, Sudip Talukdar, and Nemkumar Banthia, “A State-of-the-Art on Concrete Repairs and Some Thoughts on Ways to Achieve Durability in Repairs,” in Infrastructure Corrosion and Durability-a sustainability study, 2nd ed., 731 Gull Ave, Foster City.CA 94404, USA: OMICS Group International, 2015, pp. 115–129. [40] M. Naderi, “Effects of cyclic loading, freeze-thaw and temperature changes on shear bond strengths of different concrete repair systems,” J. Adhes., vol. 84, no. 9, pp. 743–763, 2008. [41] H. Beushausen and M. G. Alexander, “Failure mechanisms and tensile relaxation of bonded concrete overlays subjected to differential shrinkage,” Cem. Concr. Res., vol. 36, no. 10, pp. 1908–1914, 2006. [42] M. M. Sprinkel, Thermal compatibility of thin polymer-concrete overlays. 1983. [43] A. M. Vaysburd and P. H. Emmons, “How to make today’s repairs durable for tomorrow—corrosion protection in concrete repair,” Constr. Build. Mater., vol. 14, no. 4, pp. 189–197, 2000. [44] P. Carter, S. Gurjar, and J. Wong, “Debonding of highway bridge deck overlays.,” Concr. Int., vol. 24, no. 7, pp. 51–58, 2002. [45] J.-L. Granju, “Debonding of thin cement-based overlays,” J. Mater. Civ. Eng., vol. 13, no. 2, pp. 114–120, 2001. [46] A. Turatsinze, J.-L. Granju, V. Sabathier, and H. Farhat, “Durability of bonded cement-based overlays: effect of metal fibre reinforcement,” Mater. Struct., vol. 38, no. 3, pp. 321–327, 2005. [47] P. Kabele, “Assessment of structural performance of engineered cementitious composites by computer simulation.,” 2000. . [48] V. C. Li and M. Lepech, “Crack resistant concrete material for transportation construction,” in Transportation Research Board 83rd Annual Meeting, Washington, DC, Compendium of Papers CD ROM, Paper, 2004, pp. 04–4680. [49] J. Zhang and V. C. Li, “Monotonic and fatigue performance in bending of fiber-reinforced engineered cementitious composite in overlay system,” Cem. Concr. Res., vol. 32, no. 3, pp. 415–423, 2002. [50] M. Mahoutian, “Characteristics of concrete containing fly ash with mercury adsorbent,” in Masters Abstracts International, 2012, vol. 51. [51] A. C192/C192M, Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory. ASTM International West Conshohocken, Pa., 2000. [52] “ASTM C1583 / C1583M-13, Standard Test Method for Tensile Strength of Concrete Surfaces and the Bond Strength or Tensile Strength of Concrete Repair 58   and Overlay Materials by Direct Tension (Pull-off Method), ASTM International, West Conshohocken, PA, 2013, www.astm.org.” . [53] M. Naderi, “Friction-transfer test for the assessment of in situ strength and adhesion of cementitious materials,” Constr. Build. Mater., vol. 19, no. 6, pp. 454–459, 2005. [54] “ASTM C39 / C39M-15a, Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, ASTM International, West Conshohocken, PA, 2015, www.astm.org.” [55] MTI Instruments Inc., “MICROTRAK II STAND-ALONE (DISCONTINUED).” [Online]. Available: http://www.mtiinstruments.com/products/microtrakIIsal.aspx. [Accessed: 21-Sep-2015]. [56] I. S. O. Standard and B. ISO, “4287/1997,” Geom. Prod. Specif. GPS–Surface Texture Profile Method–Terms Defin. Surf. Texture Parameters, 1997. [57] National Physical Laboratory, “Feature Parameters,” SoftGauges for the Evaluation of Profile Surface Texture Parameters. [Online]. Available: http://resource.npl.co.uk/softgauges/SP_Feature_Para.htm. [Accessed: 27-Jan-2016]. [58] Image Metrology A/S, “Roughness Analysis for Profiles,” Image Metrology. [Online]. Available: http://www.imagemet.com/WebHelp6/Default.htm#RoughnessAnalysis/ProfileRoughness.htm#Profile. [Accessed: 27-Jan-2016]. [59] K. Foroutan-Pour, P. Dutilleul, and D. L. Smith, “Advances in the implementation of the box-counting method of fractal dimension estimation,” Appl. Math. Comput., vol. 105, no. 2, pp. 195–210, 1999. [60] A. Standard, “B46. 1 (2002) Surface Texture, Surface Roughness, Waviness and Lay,” Am. Soc. Mech. Eng. Am. Natl. Stand., pp. 1–98, 2002. [61] B. EN, “1504-3 (2005) Products and systems for the protection and repair of concrete structures. Definitions, requirements, quality control and evaluation of conformity. Structural and non-structural repair,” Br. Stand. Inst. [62] M. Ali, S. Kurihara, and S. Matsui, “Bonding shear strength at the interface between old and new concrete,” Technol. Rep. Osaka Univ., vol. 48, no. 231937, pp. 237–246, 1998. [63] N. Banthia and S. Dubeau, “Carbon and steel microfiber-reinforced cement-based composites for thin repairs,” J. Mater. Civ. Eng., vol. 6, no. 1, pp. 88–99, 1994. [64] Wetblasting.com, “Why Wet Blasting?,” Wetblasting.com. [Online]. Available: http://www.wetblasting.com/why-wet-blasting/. [Accessed: 11-Nov-2015]. [65] M. A. Costa, “Fractal description of rough surfaces for haptic display,” Stanford University, 2000.    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0303138/manifest

Comment

Related Items