UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development, application and early-age monitoring of fiber-reinforced ‘crack-free’ cement-based overlays Gupta, Rishi 2008

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2008_spring_gupta_rishi.pdf [ 6.94MB ]
Metadata
JSON: 24-1.0063070.json
JSON-LD: 24-1.0063070-ld.json
RDF/XML (Pretty): 24-1.0063070-rdf.xml
RDF/JSON: 24-1.0063070-rdf.json
Turtle: 24-1.0063070-turtle.txt
N-Triples: 24-1.0063070-rdf-ntriples.txt
Original Record: 24-1.0063070-source.json
Full Text
24-1.0063070-fulltext.txt
Citation
24-1.0063070.ris

Full Text

i DEVELOPMENT, APPLICATION AND EARLY-AGE MONITORING OF FIBER-REINFORCED ‘CRACK-FREE’ CEMENT-BASED OVERLAYS  by  RISHI GUPTA  B.Eng. (Civil), Pune University, 1999 M.A.Sc. (Civil), The University of British Columbia, 2002      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES (Civil Engineering)      THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2008  © Rishi Gupta, 2008 ii ABSTRACT   In most industrialized countries, significant future activity in the construction sector will be related to repair and rehabilitation of aging infrastructure.  This will require use of durable and high performance repair materials.  Among various mechanisms cited for lack of durability in repairs, early-age shrinkage cracking in overlay materials is of utmost importance.  Fiber-reinforcement can be used to alleviate some of the issues related to plastic shrinkage.  However, quantifying the performance of cement-based composites under restrained shrinkage conditions remains an issue.  Various test techniques are available to measure free and restrained shrinkage, but do not simulate the real constraint imposed by the substrate on the overlay.  In this dissertation, an innovative test method called the bonded overlay technique is described.  An overlay of fiber-reinforced material to be tested is cast directly on a substrate, and the entire assembly is subjected to controlled drying.  Cracking in the overlay is then monitored and characterized.  During the development of this test method, instrumentation was included to enable measurement of the crack propagation rate using image analysis, evaporation rate, heat development, and strain using embedded sensors.  Using the above technique, the effect of mix proportion including variables such as water-cement ratio (w/c), sand-cement ratio (s/c), and coarse aggregate content were studied.  An increase in w/c from 0.35 to 0.6 significantly increased the total cracking. Addition of coarse aggregates reduced shrinkage cracking, however, for the range of s/c investigated, no definite conclusions could be drawn.  Mixes with 0-20% fly ash and a 7 iii lit/m3 dosage of shrinkage reducing admixtures indicated no significant reduction in cracking.  The influence of fiber geometry on cracking in overlays was also investigated. Fiber types included different sizes of polypropylene and cellulose fibers and one type of glass fiber (volume fraction ranging between 0-0.4%).  Glass fibers at a small dosage of 0.1% were the most efficient fiber and completely eliminated cracking.  Of the two field projects considered: one was a plaza deck at the UBC Aquatic Center, where cellulose fibers were used, and the second at the UBC ChemBioE building, where polypropylene fibers were used in slabs-on-grade.  Both overlays were instrumented with strain sensors, data from which were monitored over the Internet. Results clearly indicated that fibers reduced the strain development in fiber-reinforced overlays when compared to un-reinforced overlays.  An energy-based fracture model was proposed to predict maximum crack widths and in a second study, an equation was proposed to correlate early-age shrinkage and flexural toughness of cellulose fibers.  In both models, a reasonable correlation with the test data was observed.  In addition, factorial design method was used and a mathematical model was proposed to correlate different variables such as w/c, s/c, and fiber dosage.  iv TABLE OF CONTENTS  ABSTRACT   …………………………………………………………………………….. ii TABLE OF CONTENTS   …………………………………………………..................... iv LIST OF TABLES   ……………………………………………………………………… x LIST OF FIGURES   …………………………………………………………….............. xiii ACKNOWLEDGEMENTS   …………………………………………............................. xxiii   Chapter 1- INTRODUCTION  1.1 Overview   ………...................................................................................................... 1  1.2 Problem Definition and Scope   ……………………….............................................. 2   1.2.1 Test Techniques   ……………...……………………………………………… 4   1.2.2 Effect of Mixture Proportion and Admixtures   ………………………………. 4    1.2.3 Field Applications   …………………………………………………………... 5   1.2.4 Concluding Remarks   ………………………………………………………… 6   Chapter 2- LITERATURE REVIEW  2.1 Introduction   ……………………………………….………………………………. 8  2.2 Shrinkage of Cement-Based Materials   ……………………………………..…….. 8  2.3 Test Techniques   …………………………………………………………………… 15   2.3.1 Un-Restrained Tests   …………………………………………………………. 15   2.3.2 Restrained Tests   ……………………………………………………………... 16    2.3.2.1 Ring Type Tests   …………………………………………………….. 16    2.3.2.2 Bonded Overlay Technique Developed at UBC   ……………………. 20    2.3.2.3 Slab/Overlay Tests   ………………………………………………….. 21    2.3.2.4 Other Tests   …………………………….……………………………. 26  2.4 Effect Of Fiber Reinforcement On Shrinkage   ………..…………………………… 30  2.5 Effect of Mix Proportion, Admixtures, and Additives   …………………………… 34 v  2.6 Proposed Models    ………………………………………………………………… 38   Chapter 3- MATERIAL PROPERTIES  3.1 Introduction   ……………..…………………………………………….…………... 44  3.2 Plain Mixes   ……………….……………………………………………………….. 44  3.3 Fiber Reinforced Mixes   ..………………………………………………………… 47  3.4 Supplementary Cementing Materials (SCMs)   ..……………………….……….. 54   3.4.1 Fly Ash   ………………………………………………………………………. 54   3.4.2 Silica Fume   ………………………………………………………................ 56  3.5 Admixtures   ………………………………………………………………………... 57   3.5.1 Superplasticizer   ………………………………………………….…………... 57   3.5.2 Shrinkage Reducing Admixtures   ……………………………………………. 58   Chapter 4-DEVELOPMENT OF EXPERIMENTAL TECHNIQUE  4.1 Test Summary   …………………………………………………………………… 60  4.2 Introduction   ……………………………………………………………………….. 60  4.3 Bonded Overlay Test Technique   ……..………………………………..………… 61   4.3.1 Substrate Bases   …………………………………………………...………… 61    4.3.1.1 Development of Substrate Bases   …………………………..……….. 61   4.3.2 Environmental Chamber   …………………………………………….…… 69    4.3.2.1 Development of Environmental Chamber   ………………………… 69    4.3.2.2 Modified Environmental Chamber   ………………………….……… 70   4.3.3 Placement of Overlay 73   4.3.4 Crack Measurement   ………………………………………………………… 74   4.3.5 Effect of Different Concrete Substrates on Crack Characteristics (Results)   ...   77   4.3.5.1 Preliminary Observations 77   4.3.5.2 Results 78  4.4 Further Developments in the Test Method   ……………………………………… 80   4.4.1 Sensors   ……………………………………………………………………… 80 vi   4.4.2 Strain Development in the Overlay   ………………………….……….……… 83   4.4.3 Non-Contact Crack Measurement Using Image Analysis   ………………..… 88  4.5 Concluding Remarks   ……………………………………………………………… 92   Chapter 5- EARLY-AGE BEHAVIOR OF CEMENTITIOUS COMPOSITES  5.1 Summary   …………………………………………………………………….……. 95  5.2 Introduction   ……………………………………………………………………….. 95  5.3 Test Set-up and Variables Investigated   ...………………………………..……….. 96  5.4 Crack Measurements   ……..………………………………………………...……... 98  5.5 Results   …………………………………………………………………………….. 99   5.5.1 Specimen Moisture Loss   …………………………….……………….……... 99   5.5.2 Heat Evolution   ………………………………………………………….…… 100   5.5.3 Crack Analysis   ………………………………………………………….…… 103  5.6 Concluding Remarks   ……………………………………………………………… 114   Chapter 6- DEVELOPING ‘CRACK-FREE’ OVERLAYS USING FIBERS  6.1 Fiber Reinforced Mixes   ……………………….…………………………...……... 115  6.2 Results and Discussion   …………………..……………………………………….. 115  6.3 Effect on Total Crack Area   ……………………………………………………….. 117  6.4 Effect on Crack Width   ……………………….………………………………........ 123  6.5 Efficiency Factors   ……………………………………………………………......... 129  6.6 Effect on Average Number of Cracks   …………………………………………….. 133  6.7 Time to First Crack   …………………………………………………………......... 133  6.8 Concluding Remarks   …………………………………………………………........ 135    Chapter 7- FIELD MONITORING OF FIBER REINFORCED OVERLAYS  7.1 Introduction   ………………………………..…….…………………………........... 142 vii  7.2 Aquatic Center Project   ………………………………………………….………..... 143   7.2.1 Site Preparation and Mix Proportions   ………………………………….......... 144   7.2.2 Gauge Location and Data Monitoring   ……………………….………………. 145   7.2.3 Results - Strain from Embedded Sensors   …………………………...……….. 151   7.2.4 Non Destructive Testing   …………………………………………................... 155    7.2.4.1 Ultrasonic Pulse Velocity   …………………………………………. 157    7.2.4.2 Schmidt Rebound Hammer   …………………………………………. 160    7.2.4.3 Resistivity Measurements   ……………………………………........... 160   7.2.5 Results – NDTs   ………………………………………………………............ 163   7.2.5.1 Ultrasonic Pulse Velocities   ………………………………………… 163    7.2.5.2 Rebound Hammer Values   ……………………………………........... 166    7.2.5.3 Resistivity Values   ……………………………………………........... 169    7.2.5.4 Confirmatory Laboratory Testing   ……………………………........... 171  7.3 Demonstration Project at ChemBioE Building, UBC   ………………...….….......... 177   7.3.1 Project Details   ……………………………………………………………...... 177   7.3.2 Embedded Sensors   …………………………………………………………... 178   7.3.3 Mix Proportions and Placement Details   …………………………….……...... 182   7.3.4 Strain Monitoring   ………………………………...…………….…….…….... 186   7.3.5 Results and Discussion    ……………………………………………..……….. 187    7.3.5.1 Strain Readings   …………………………………….………………. 187    7.3.5.2 Material Tests   ……………………………………….…….………... 193    7.3.5.3 Non Destructive Tests (NDTs)   ………………………….………...... 200     7.3.5.3.1 Schmidt Hammer Measurements   ….………………………. 200     7.3.5.3.2 Ultrasonic Pulse Velocity   ………….…................................. 202     7.3.5.3.3 Resistivity   ………………………….…................................. 203    7.3.5.4 Laboratory Strain tests   …………………………………….………... 204  7.4 Concluding Remarks 207  Chapter 8- ALTERNATE PLASTIC SHRINKAGE PREDICTION TECHNIQUES  8.1 Introduction   ………………………………………………………………….……. 209  8.2 Compressive Strength at Early Ages   …………………………………..………….. 209 viii  8.3 Uniaxial Tensile Properties    ………………………………………………...…....... 210   8.3.1 Introduction    ………………………………………………………...……...... 210   8.3.2 Mixes and Test Set-up   ……………………………………………...……....... 211   8.3.3 Test Results   ………………………………………………………….………. 214   8.3.4 Analysis and Discussion   …………………………………………….............. 220  8.4 Correlating Early-Age Shrinkage and Flexural Toughness    ……...………….……. 227   8.4.1 Introduction    …………………………………………………...……..…….... 227   8.4.2 Procedure and Results    ……………………………………………....……..... 228   8.4.3 Analysis and Discussion    …………………………………...………..……..... 230  8.5 Factorial Design    ………………………………………………………...…...…… 232   8.5.1 Introduction    …………………………………………………………………. 232   8.5.2 Definition of Factorial Design   ……………………………………………...... 233   8.5.3 Calculation of “Effect” and “Interaction”   …………………………..……….. 234    8.5.3.1 Calculation of Main Effect   …………………………….....……….... 234    8.5.3.2 Calculation of the Main Effect of Variable B   ….…………………… 234    8.5.3.3 Calculation of B × C Interaction   ……………….….……….……….. 236    8.5.3.4 Calculation of A × B × C Interaction   ……………...………..………. 236   8.5.4 Geometric Representation   …………………...………………………………. 236   8.5.5 Calculation of Standard Error   …………………………………........……….. 239   8.5.6 Application of Factorial Design to Representative Test Results   …...……....... 240   8.5.7 Mathematical Model   ……………………………………………...….……… 244  8.6 Energy-Based (Fracture) Model   …………………………………………..………. 252   8.6.1 Introduction   ………………………………………………………….…......... 252   8.6.2 Model Theory   ……………………………………………………..…………. 252   8.6.3 Proposed Model   ……………………………………………………...………. 254   8.6.4 Predicted Values   ……………………………………………………………... 255   8.6.5 Use of Model (Summary) and Limitations    ………………………..………... 263 Chapter 9 - CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH  9.1 Conclusions   ……………………………………………………………………..… 266  9.2 Recommendations for Future Research   …..……………………………………….. 272 ix   BIBLIOGRAPHY   …..…………………………………………………........................... 275   APPENDICES   …………………………………………………………………………... 286    x LIST OF TABLES  Table 2.1 Effect of Fly-Ash on Shrinkage   ……………………………….. 37 Table 3.1 (a) Physical and Chemical Data for Cement   ………………………. 45 Table 3.1 (b) Specific Gravity and Mechanical Properties of Fibers   ………… 47 Table 3.2 Fiber Properties   ………………………………………………... 48 Table 3.3 Chemical and Physical Analysis of Fly-Ash   ………………….. 56 Table 3.4 Chemical and Physical Properties of Silica Fume (Source: Basalite Concrete Products)   …………………………………… 57 Table 3.5 Physical and Chemical Properties of SRA   …………………….. 59 Table 4.1 Different Types of Substrate Bases   ……………………………. 62 Table 4.2 Mix Proportions for Evaluation of Substrate Bases   ………….... 68 Table 4.3 Properties of Deformed Steel Rebars   ………………………….. 68 Table 4.4 Effect of Substrate Types on Crack Characteristics   …………… 79 Table 4.5 Comparison of Results from Measurements Using a Microscope and Image Analysis   ……………………………………………. 91 Table 5.1 Mix Proportion of Substrate base   ……………………………… 97 Table 5.2 Overlays with Varying w/c and s/c Ratios (Series M)   ………… 98 Table 5.3 Overlays with Varying Fly-Ash and SRA Dosages (Series F)   ... 98 Table 5.4 Evaporation Rate Before and After Demolding   ……………….. 101 Table 5.5 Crack Analysis (Microscope Method)   ………………………… 105 Table 5.6 Effect of w/c on Rate of Evaporation, Heat Evolution and Cracking   ……………………………………………………….. 110 Table 6.1 Fiber Dosages Investigated   ……………………………………. 116 xi Table 6.2 Detailed Test Results for Different Mixes   ..…………………… 128 Table 6.3 Effect of Fiber on “Time to first Crack’   …………………..…... 134 Table 7.1 Mix Proportions of Repair Concrete Used at the Parking Garage and Aquatic Center   ………………………………………….… 146 Table 7.2 Corrections for Pulse Velocity Due to Temperature Changes   … 159 Table 7.3 UPV Results for Plain Concrete and FRC   ……………..……… 164 Table 7.4 Compressive Strength using Schmidt hammer   ………………... 167 Table 7.5 Resistivity for Plain Concrete and FRC Overlays   ……….……. 169 Table 7.6 Lab UPV Values for Plain Concrete and FRC Prisms   …….….. 176 Table 7.7 Lab Resistivity Measurements   ………………………..……….. 176 Table 7.8 Concrete Mix Design   ………………………………………….. 184 Table 7.9 Fiber Dosage and Concrete Placement Details   ………………... 185 Table 7.10 Post-Setting Strain Increment Data (Electrical Sensors)   ………. 191 Table 7.11 Compression Test Results   ……………………………………... 194 Table 7.12 Toughness Values Using ASTM C 1609   ……………………... 199 Table 7.13 UPV Results   …………………………………………………… 203 Table 7.14 Onsite Resistivity Results   ……………………………………… 204 Table 8.1 (a) Compressive Strength and Elastic Modulus at Early Ages   ……. 210 Table 8.1 (b) Mixes Investigated under Early-Age Uniaxial Tension   ……….. 212 Table 8.2 Uni-axial Test Data   ……………………………………………. 223 Table 8.3 Fiber Dosage for Shrinkage and Toughness Tests   …………….. 228 Table 8.4 Summary: Shrinkage Crack Area and Toughness Data (ARS)  … 229 Table 8.5 Data from a 23 Factorial Design   ……………………………….. 234 xii Table 8.6 Effect of Variable B or “B Main Effect”   …………………..….. 235 Table 8.7 Mixes for 23 Factorial Design   …………………………..……... 241 Table 8.8 Crack Analysis Results   ………………………………………… 242 Table 8.9 Coefficients from Regression Analysis   ………………..………. 245 Table 8.10 Comparison of Predicted Data to Test Results   ……………….... 246 Table 8.11 Comparison of Predicted Values and Test Results   ……………. 249 Table 8.12 Comparison of Crack Widths (in mm)- Model vs. Test Results  .. 259    xiii LIST OF FIGURES  Figure 1.1 (a) Plastic Shrinkage (a) Restraint at the Base   …………………….. 2 Figure 1.1 (b) Plastic Shrinkage (b) Unrestrained   …………………………….. 2 Figure 2.1 (a) ACI Nomograph for Evaporation Rate Estimation   ………..…… 9 Figure 2.1 (b) Phenomenological Summary of Early-Age Volume Change   ..… 10 Figure 2.2 Restrained Ring Test   ……………………….…………………… 16 Figure 2.3 (a) Nordtest Shrinkage Test a) Test Specimen   ……………….……. 17 Figure 2.3 (b) Nordtest Shrinkage Test b) Cracked Specimen   …………….….. 17 Figure 2.4 Test Setup, Ring Type Specimen   ……………………………..... 18 Figure 2.5 Schematic of Tensile Stress Development in the Shrinking Overlay   …………………………………………………………. 20 Figure 2.6 Test Setup Using a Plywood Mold   ……………………..………. 21 Figure 2.7 Crack Measurement (Bayasi and McIntyre, 2002)   ……………... 21 Figure 2.8 (a) Plan and Profile View of the Environmental Chamber (Naaman et al., 2005)   ………………………………………………………... 22 Figure 2.8 (b) Test Setup (Naaman et al., 2005)   ……………………………….. 23 Figure 2.8 (c) Test Setup with Stress Riser   ……………………………………. 24 Figure 2.9 Test Setup (Najm and Balaguru, 2002)   ………………………… 25 Figure 2.10 Test Setup (Sanjuan and Moragues, 1994)   ……………………... 26 Figure 2.11 Test Setup with AE sensors (Kim and Weiss, 2003)   …………… 27 Figure 2.12 (a) Figure 2.12 (a)- Test Set-Up used by Bloom and Bentur (1995)   . 28 Figure 2.12 (b) Figure 2.12 (b)- Specimen Size and Grips used by Kovler (1994)……………………………………………………..……… 29 xiv Figure 2.12 (c) Cracking Test Apparatus or Bench   ……………………………... 30 Figure 3.1 (a) Gradation Analysis (a) Concrete Sand    ..……………………….. 46 Figure 3.1 (b) Gradation Analysis (b) 14 mm Coarse Aggregate   ……………… 46 Figure 3.2 SEM Images for Micro Synthetic Fibers PF10   ………………… 50 Figure 3.3 SEM Images for Macro Synthetic Fibers PF2   …………...…….. 51 Figure 3.4 SEM Images for Micro Synthetic Fibers PF8   ………………….. 51 Figure 3.5 SEM Images for Glass Fibers GF1   ……………………………... 52 Figure 3.6 SEM Images for Cellulose fibers CF1   ………………………….. 52 Figure 3.7 SEM Images for Cellulose fibers CF4   ………………………….. 53 Figure 3.8 (a) Morphology (a) Fly ash   ………………………………………… 55 Figure 3.8 (b) Morphology (b) Silica Fume   …………………………………… 55 Figure 4.1 (a) Substrate Types: (a) Smooth Finish   ……………………………. 64 Figure 4.1 (b) Substrate Types: (b) Aggregate Finish   …………………………. 64 Figure 4.1 (c) Substrate Types: (c) Staggered Protrusions   …………………….. 64 Figure 4.1 (d) Substrate Types: (d) Symmetric Protrusions   …………………… 64 Figure 4.2 Steel Substrate Base   …………………………………………….. 64 Figure 4.3 Dimensions: Substrate Base with Staggered Protuberances (Type C3)   ……………………………………………………………… 66 Figure 4.4 (a) Overlay Wrapping Around the Substrate (after Demolding) …………......................................................................................... 66 Figure 4.4 (b)  Dimensions of the Overlay Wrapping around the Substrate (in mm)   …………………………………………………………….. 67 Figure 4.5 Environmental Chamber   …………………………………….….. 69 Figure 4.6 (a) Environmental Chamber (a) 3D View   ………………………….. 71 xv Figure 4.6 (b) Environmental Chamber (b) Dimensioned Sketch   ……………... 72 Figure 4.7 Typical plot- Temperature and RH vs. Time   …………………… 73 Figure 4.8 Substrate Ready to Receive Overlay   …………………………… 74 Figure 4.9 Specimen After Cracking   ………………………………………. 75 Figure 4.10 Crack Measurement Using a Microscope   ……………………… 75 Figure 4.11 Finalized Substrate Type S3-R   ………………………………… 80 Figure 4.12 Sensors and Data Acquisition System   …………………………. 81 Figure 4.13 (a) Arrangement for Measuring Moisture Loss (a) Location of the Specimen in the Environmental Chamber   ……………………… 82 Figure 4.13 (b) Arrangement for Measuring Moisture Loss (b) Placement of the Specimen Atop the Weighing Scale   ……………………………. 82 Figure 4.14 Non-Contact Laser Sensor for Measuring Settlement   ………….. 83 Figure 4.15 Strain Gauges Mounted on Substrate Surface   ………………….. 84 Figure 4.16 Chairs to Place Strain Gauges in the Overlays   ………………… 84 Figure 4.17 (a) Strain Measurements in the Overlay Using Embedded Strain Gauges   ………………………………………………………….. 86 Figure 4.17 (b) Strain Measurements in Restrained Specimens   ………………… 87 Figure 4.18 Image Analysis: Crack area Masking   …………………………... 89 Figure 4.19 Variation of Crack Area Based on Measurement Technique   …... 92 Figure 5.1 Effect of Mix Proportion on Moisture Loss   ……………………. 101 Figure 5.2 Heat Evolution in 0.35 and 0.5 w/c Mixes   …………………….. 102 Figure 5.3 Heat Evolution in 0.40 and 0.60 w/c Mixes   ……………………. 102 Figure 5.4 Crack Evolution using Image Analysis   ………………………… 104 Figure 5.5 (a) Results of Series-M Tests   ………………………….…………… 108 xvi Figure 5.5 (b) Effect of Coarse Aggregates (Mixes M8, M8-1 and M8-2)   ……. 109 Figure 5.5 (c) Results of Series-F Tests   ……………………………………….. 109 Figure 5.5 (d) Results of Series-M Tests (Effect of s/c)   ………………………. 110 Figure 5.6 (a) Typical Crack Pattern (a) Effect of w/c   ………………………… 111 Figure 5.6 (b) Typical Crack Pattern (b) Effect of Coarse Aggregates   ………... 111 Figure 5.7 (a) Time to First Crack for Series-M   ……………………………….. 113 Figure 5.7 (b) Time to First Crack for Series-F   ………………………………... 113 Figure 6.1 (a) Crack Area vs. Fiber Volume (a)   ………………………………. 118 Figure 6.1 (b) Crack Area vs. Fiber Volume (b)   ………………………………. 119 Figure 6.1 (c) Crack Area vs. Fiber Volume (c)   ………………………………. 119 Figure 6.1 (d) Crack Area vs. Fiber Volume (d)   ………………………………. 120 Figure 6.2 (a) Average Crack Width vs. Fiber Volume   ………………………. 124 Figure 6.2 (b) Average Crack Width vs. Fiber Volume   ………………………. 124 Figure 6.2 (c) Average Crack Width vs. Fiber Volume   ………………………. 125 Figure 6.3 (a) Maximum Crack Width vs. Fiber Volume   …………………….. 126 Figure 6.3 (b) Maximum Crack Width vs. Fiber Volume   ……………………... 126 Figure 6.3 (c) Maximum Crack Width vs. Fiber Volume   ……………………... 127 Figure 6.4 (a) Crack Area Control Efficiency (η area) vs. Fiber Volume   ………. 131 Figure 6.4 (b) Crack Area Control Efficiency (η area) vs. Fiber Volume   ………. 131 Figure 6.5 (a) Crack Width Control Efficiency (η width) vs. Fiber Volume   …….. 132 Figure 6.5 (b) Crack Width Control Efficiency (η width) vs. Fiber Volume   …….. 132 Figure 6.6 Crack Pattern for C1   ……………………………………………. 136 xvii Figure 6.7 (a) Crack Pattern for PF1   …………………………………………... 137 Figure 6.7 (b) Crack Pattern for PF1   …………………………………………... 137 Figure 6.7 (c) Crack Pattern for PF1   …………………………………………... 137 Figure 6.7 (d) Crack Pattern for PF1   …………………………………………... 137 Figure 6.8 (a) Crack Pattern for PF2   …………………………………………... 138 Figure 6.8 (b) Crack Pattern for PF2   …………………………………………... 138 Figure 6.8 (c) Crack Pattern for PF2   …………………………………………... 138 Figure 6.8 (d) Crack Pattern for PF2   …………………………………………... 138 Figure 6.9 (a) Crack Pattern for PF8   …………………………………………... 139 Figure 6.9 (b) Crack Pattern for PF8   ………………………………………….. 139 Figure 6.9 (c)  Crack Pattern for PF8   ………………………………………….. 139 Figure 6.10 (a) Crack Pattern for CF1   ………………………………………….. 140 Figure 6.10 (b) Crack Pattern for CF1   …………………………………………... 140 Figure 6.10 (c) Crack Pattern for CF1   …………………………………………... 140 Figure 6.10 (d) Crack Pattern for CF1   ………………………………………….. 140 Figure 6.11 (a) Crack Pattern for GF1   ………………………………………….. 141 Figure 6.11 (b) Crack Pattern for GF1   …………..……………………………… 141 Figure 6.11 (c) Crack Pattern for GF1   …………………………..……………… 141 Figure 7.1 Cellulose Fibers   ………………………………………….…….. 145 Figure 7.2 (a) Parking Garage (a) A Strain Gauge   …………………………..... 147 Figure 7.2 (b) Parking Garage (b) Placement of Concrete   …………………….. 147 Figure 7.3 (a) Aquatic Center Plaza Deck (a) Slab Schematic   ………………… 148 xviii Figure 7.3 (b) Aquatic Center Plaza Deck (b) Location of Strain Gauges   …….. 149 Figure 7.3 (c) Aquatic Center Plaza Deck (c) Placement and Finishing of Concrete   ………………………………………………………… 149 Figure 7.4 Embedded Strain Gauge   ………………………………………... 150 Figure 7.5 Specially Designed Chairs for Strain Gauge   …………………… 150 Figure 7.6 Web-based Data Acquisition System WebDaq/100   ……………. 151 Figure 7.7 Strain Measurement from the Aquatic Center Overlay   ………… 152 Figure 7.8 Measured Strain from Aquatic Center and the Parking Garage   ... 154 Figure 7.9 Schematic Strain Variation in FRC and Influence of Strain Gauge Location   …..…………………………………………………….. 155 Figure 7.10 (a) Schematic of Pulse Velocity Apparatus   ………………………... 157 Figure 7.10 (b) Schematic Showing Placement of Probes for Indirect Measurement   …………………………........................................ 158 Figure 7.11 A PUNDIT   ……………………………………………………… 159 Figure 7.12 Grid Points Smoothened using the Sander   ……………………... 159 Figure 7.13 Schmidt Rebound Hammer   …………………………………….. 160 Figure 7.14 Resistivity Meter and Four-Probe Set-Up   ……………………… 162 Figure 7.15 Circuit Diagram of the Resistivity Meter   ………………………. 162 Figure 7.16 (a) 2D and 3D Plots of UPV (in m/sec)   …………………………… 165 Figure 7.16 (b) 2D and 3D Plots of Strength (in MPa)   ………………………..... 168 Figure 7.17 2D and 3D Plots of Resistivity (in KΩ-cm)   ……………………. 170 Figure 7.18 Range of Electrical Resistivity for a Variety of Materials   ……... 171 Figure 7.19 Smoothening Using a Sander   ……………………………..…… 172 Figure 7.20 Beam Arrangement for UPV Measurements (Schematic)   ……… 173 xix Figure 7.21 Indirect UPV Measurements on Prisms   ………………………… 174 Figure 7.22 Lab Resistivity Measurements   …………………………………. 175 Figure 7.23 Schematic of Concrete Placements   …………………………….. 178 Figure 7.24 (a) A Typical Optical Sensor (a) Installation of Fiber Optic Cable on GFRP Rebar   ……………………………………………………. 180 Figure 7.24 (b) A Typical Optical Sensor (b) Fiber Optic Cable on Cebar with Armour Tape and Angle Polished Connector   ………………….. 180 Figure 7.25 (a) Strain Gauges (a) Typical Details (Specially Designed Chairs)   ... 181 Figure 7.25 (b) Strain Gauges (b) Chairs Before Concrete Placement   …………. 181 Figure 7.25 (c) Strain Gauges (c) Partially Embedded Gauges During Concrete Placement………………………………………………..……….. 181 Figure 7.26 Schematic of Sensor Locations (P1 and P4 are Control and Fly Ash Placements Respectively; and F2, F3, and F5 Contain Novomesh 950)   ............................................................................ 182 Figure 7.27 Ready-Mix Plant, Richmond, B.C.   ……………………………... 183 Figure 7.28 Polypropylene Fibers   …………………………………………… 183 Figure 7.29 (a) Concrete Pour on Site (a) Embedded Strain Gauges (Electrical and Optical) ……………………………………………………… 186 Figure 7.29 (b) Concrete Pour on Site (b) Concrete Pour and Finishing Operation ……………………………………………………………………. 186 Figure 7.30 On-Site Data Acquisition System   ……………………………… 187 Figure 7.31 Presetting Averaged Signals from the Electrical Sensors (Tensile +ve Compression –ve).   …………………………………………. 189 Figure 7.32 Average Post-Setting Signals from the Electrical Sensors.   …….. 189 Figure 7.33 (a) Average Signals from the Optical Sensors   ...…………………… 192 Figure 7.33 (b) Average Post-Setting Signals from the Optical Sensors   ……….. 193 xx Figure 7.34 Cylinder and Prisms for Material Tests   . ……………………….. 194 Figure 7.35 Average Load vs. Deflection Plots for FRC (F2/F5 and F3) as Per C1609   …………………………………………………………… 196 Figure 7.36 Load vs. Deflection Plots for Three FRC (F3) Specimens Tested as Per C1609   …………………………………………………… 197 Figure 7.37 Load vs. Deflection Plots for Three FRC (F5) Specimens Tested as Per C1609   ……………………………………………………. 197 Figure 7.38 Post Crack Strength (PCS) Plots   ……………………………….. 198 Figure 7.39 Calculation of PCSm Values   ……………………………………. 198 Figure 7.40 Slabs-On-Grade Before Performing NDTs   …………………….. 200 Figure 7.41 On-Site Schmidt Hammer Results   ……………………………… 201 Figure 7.42 UPV Measurements Using PUNDIT   …………………………… 202 Figure 7.43 Electrical and Optical Strain Gauges (Laboratory test)   .. ………. 205 Figure 7.44 Electrical Sensor Data (Laboratory)   ……………………………. 205 Figure 7.45 Optical Sensor Data (Laboratory)   ……………………………… 206 Figure 8.1 Molds for Casting Briquette Specimens   ………………………... 212 Figure 8.2 Briquette dimensions for Uniaxial Tensile Test   ………………... 213 Figure 8.3 (a) Test Set-up (a) Overall Arrangement for Uniaxial Tension Tests    214 Figure 8.3 (b) Test Set-up (b) Test Specimen During Testing   ………………… 214 Figure 8.4 Load vs. Deflection for Mix C1 (Control)   ……………………… 215 Figure 8.5 Load vs. Deflection for Mix C2 (20% Fly-Ash)   ……………….. 216 Figure 8.6 Load vs. Deflection for Mix F1 (Polypropylene Fibers, PF8)   ….. 217 Figure 8.7 Load vs. Deflection for Mix F2 (Polypropylene fibers, PF1)   ….. 217 Figure 8.8 (a) Load vs. Deflection for Mix F3 (Polypropylene Fibers, PF2)   ….. 218 xxi Figure 8.8 (b) Load vs. Deflection for Mix F4 (Glass Fibers, GF1)   …………... 218 Figure 8.9 Load vs. Deflection for Mix F5 (Cellulose Fibers, CF1)   ………. 219 Figure 8.10 Load vs. Deflection for Mix F6 (Cellulose Fibers, CF5)   ………. 219 Figure 8.11 A Tested Specimen Showing Polypropylene Fibers Across the Crack   …………………………………………………………… 220 Figure 8.12 (a) Stress and Elongation for Different Mixes   …………………….. 224 Figure 8.12 (b) Comparison of Tensile Stress at 2 mm and Shrinkage Cracking ……………………………………………………………………. 226 Figure 8.13 Shrinkage and ARS Correlation   ………………………………... 230 Figure 8.14 Geometric Representation of a 23 Factorial Design   ……………. 237 Figure 8.15 (a) B Main Effect   …………………………………………………... 238 Figure 8.15 (b) B x C Interaction   ……………………………………………….. 238 Figure 8.15 (c) A x B x C Interaction   …………………………………………… 238 Figure 8.16 Main and Interaction Effects on Cracking Area and Average Width   …………………………………………………………… 243 Figure 8.17 Comparison of Proposed Model to Test Data (Crack Area) ………………………………………………………………...….. 247 Figure 8.18 Comparison of Proposed Model to Test Data (Crack Width) …………......................................................................................... 247 Figure 8.19 Effect of w/c Ratio on Predicted and Actual Test Data   ………… 250 Figure 8.20 Effect of %PP on Predicted and Actual Test Data   ……………... 251 Figure 8.21 (a) Stress Transfer in the Cohesive Zone in Front of a Stress-Free Crack: (a) Stresses Near the Crack Tip   ………………………… 253 Figure 8.21 (b) Stress Transfer in the Cohesive Zone in Front of a Stress-Free Crack: (b) Microcrack Zones   …………………………………… 253 Figure 8.22 (a) Fracturing in Tensile Bar: Distributed Cracking Before Peak   ….. 254 xxii Figure 8.22 (b) Fracturing in Tensile Bar: Localized Crack in a Narrow Zone   … 254 Figure 8.23 (a) Schematic Showing the Effect of Reduction in Fiber Volume Fraction on the Softening Curves   ………………………………. 257 Figure 8.23 (b) Plot of Post-Peak Stress vs. Maximum Crack Width (Glass, Cellulose, and Polypropylene Fibers)   ………………………….. 258 Figure 8.23 (c) Plot of Post-Peak Stress vs. Maximum Crack Width (Cellulose and Polypropylene Fibers) With an Insert Showing Magnified View   ……………………………………………………………. 258 Figure 8.24 (a) Comparison of Predicted Crack Widths With Test Data   ……….. 261 Figure 8.24 (b) Comparison of Cracking in Specimens With That Assumed in the Model   …………………………………………………………… 261 Figure 8.25 Relationship Between Fiber Volume Fraction and Factor ‘n’   ….. 263 xxiii ACKNOWLEDGEMENTS  I dedicate my thesis to the memory of my dad, Dr. Avinash Chander Gupta, who introduced me to the world of Civil Engineering and inspired me to aim high. Dad, because of you, I am where I am.   This thesis would not have been possible without the support I received from a number of people.  First and foremost, I would like to thank Dr. Nemy Banthia, my mentor, guide, and guru who always made me strive for the best.  I am truly indebted to him for giving me the opportunity to work on this project.  His guidance, technical expertise, innovative ideas, and continued encouragement were instrumental in allowing me to successfully complete this project.  I would also like to thank my mom, Saroj Gupta whose continued prayers and blessings I have sought to complete this thesis.  A big thank you to Aditi’s parent, Jayshree and Srinivas Kolachala, for their prayers, good wishes and blessings.  Thank you to my sisters, Neerja, Archana, and Tulika and sister-in-law Monisha and their families for continuing to believe in me.  In particular, I would like to acknowledge all my extended family in Vancouver for their support and encouragement, especially my Grandfather, Mr. Bajrangi Dass Chadha who has taught me to be very patient in life.  I would also like to thank my defense committee (especially Dr. Sidney Mindess and Dr. Pat McGrath) for their intellectual contributions and technical expertise that has allowed me to successfully complete this dissertation.  I would like to acknowledge the contributions of my colleagues: Reza, Fariborz, Vivek, Yashar, Manote and many visiting students from France, including Antoine Clement, Willy Breda and Maxime Trocme.  The assistance of technicians at the rusty hut including Doug Smith, John Wong, Bill, and Harald Schrempp is greatly appreciated.  I also acknowledge the involvement, financial and in-kind contributions of NSERC, ISIS Canada Research Network, Propex Inc., Pozzolanic International Limited, Rempel Concrete (Mr. Irv Lenz), on-site staff of Stuart Olson, and Cement Association of Canada (Andy Vizer). xxiv  I would also like to thank Mr. Ian Simpson, President of Ground-IT.com who helped us conduct preliminary resistivity tests and also provided the resistivity meter for testing. Assistance provided by other members of UBC such as Dr. Jesus Calvino-Fraga, (Instructor, Department of Electrical and Computer Engineering), and Ron Loewen (Dean’s Office, Applied Science) is also appreciated.  Lastly, I would like to thank my wife Aditi, without whose enduring love and support, I would have never made it so far.  Thank you for being so patient and being there for me every step of the way. Without you, this journey would never have been easy. Your everlasting smile gave me the strength to complete this project.       1 CHAPTER 1- INTRODUCTION  1.1 Overview Cement-based materials shrink due to loss of moisture and due to self desiccation (internal consumption of free available water from the mix for continued hydration process).  Cement paste in its plastic state is thus subject to volumetric contractions, which can reach values as high as one percent of the absolute volume of dry cement. Rapid loss of moisture leads to reduction in volume even before concrete has gained strength.  This reduction in volume in the absence of sufficient bleeding is termed as ‘plastic shrinkage.’  This volumetric shrinkage and the rate at which it takes place is a function of temperature, humidity, and wind velocity of the drying environment. Although some of the water lost this way is replenished by bleeding, if the surface moisture loss exceeds 0.5 kg/m2/h (Mindess et al., 1981), negative capillary pressures develop in the concrete causing internal compressive strains.  If concrete is restrained, these compressive strains may result in tensile stresses far in excess of those needed to cause cracking in young concretes with poorly developed strengths.  This leads to what is termed as “restrained plastic shrinkage cracking” and is found to be the cause for plastic shrinkage cracks as shown in Figure 1.1 (a) as opposed to free uniform shrinkage (Figure 1.1 (b)) where volume reduces but does not create any stresses due to lack of a restraint.  Restraint can also develop due to differential shrinkage between the layers of an overlay.  When rising bleed water from inside the concrete cannot replace the amount of surface water that has evaporated, tensile stresses are induced in the surface layers, also resulting in cracks.  The tensile force is caused by the water saturation front receding 2 below the concrete surface (the water flow is outward), which forms menisci between the fine particles of cement and aggregates.  If the concrete surface has started to set and has developed tensile strength that is sufficient to resist the tensile forces, cracks will not form. If the surface dries very rapidly, tensile stresses will far exceed the tensile strength of the material, which is very low at early ages.    (a) (b) Figure 1.1 - Plastic Shrinkage (a) Restraint at the base and (b) Unrestrained   1.2 Problem Definition and Scope  In most industrialized countries, significant future activity in the construction sector will be related to repair and rehabilitation of the aging infrastructure.  Experience has shown that thin repairs or patching debond more easily than thick structural repairs, 3 and need exists to develop high performance repair materials that produce durable, aesthetically pleasing and cost effective repairs.  Among the various mechanisms cited for lack of durability in repairs, early-age shrinkage cracking in the repair materials is the most important.  There are several reasons for this concern: first, for a better bond, cement-based repair materials generally carry large amounts of cements which increases their overall shrinkage; second, the substrate offers a high level of restraint thereby increasing the possibility of developing tensile stresses sufficient to cause tensile cracking in the overlays; and finally, it is often difficult, if not impossible, to properly cure repair materials.  Cracking not only creates easy access routes for deleterious agents to enter the overlay-substrate interface, but also allows for an early saturation of the overlay material resulting in freeze-thaw damage, swelling, scaling, discoloration and eventual debonding.  Early-age shrinkage of cement mortar and concrete has also posed a major concern for decades especially in applications such as slabs-on-grade, bridge decks, shotcrete tunnel linings, thin repair patches, tilt-up panels, and industrial floors, where the surface area to volume ratio is large.  Due to the increased exposed surface area, moisture loss from concrete is rapid, causing excessive volumetric shrinkage.  Moreover, in these applications, the restraint imposed by the substrate on the overlay is severe causing severe surface cracking.  Reduced durability of these surfaces and need for repeated remedial work has becomes far too common in these applications.  Cracking of these overlays is also unsightly.   4 1.2.1 Test Technique  Various test techniques are available to study the early-age properties of cement composites.  However, most techniques do not simulate the real stress conditions (described above) in a shrinking material under restrained conditions.  Lack of a test technique that simulates real field restraint conditions has motivated this research.  A novel test technique where the overlay is restrained using a sub-base with protuberances is used and the overlay is subjected to extreme environmental conditions.  Details about this technique and the instrumentation are presented in Chapter 4.  1.2.2 Effect of Mixture Proportion and Admixtures  Effect of water/cement (w/c) ratio and presence of aggregates are known to significantly affect properties of cement-based materials.  Using the novel technique developed during this project, influence of w/c ratio and aggregates was studied on early- age moisture loss, heat of hydration and shrinkage crack characteristics of cementitious composites.  This is described in Chapter 5.  Many solutions exist that can alleviate cracking in concrete overlays.  The most effective technique of mitigating plastic shrinkage cracking is by preventing the loss of water from the concrete surface by extended curing.  In some instances, however, curing alone is not adequate, and additional measures need to be adopted.  These include temperature control, shielding from high winds, reduced use of admixtures that prevent bleeding and the use of shrinkage reducing admixtures (Weiss & Shah, 1997 and Shah et al. 1992).  However, in recent years, use of fibers in concrete to reduce shrinkage cracking has become very popular.  Fibers mitigate cracking in two ways: first they 5 reduce the overall shrinkage strains and lower the possibility of tensile stresses exceeding the tensile strength, and second, if the cracks do occur, fibers bridge them effectively and prevent them from growing.  Different types of fibers affect early-age material property differently and depending on the volume fraction used in a mix and the effectiveness of fibers, shrinkage cracking can be completely eliminated.  The test technique described above was used to develop ‘crack-free’ overlays by studying the effect of mix proportion on plastic shrinkage cracking and different fiber types that could effectively control development of these cracks.  1.2.3 Field Applications  Performance of materials developed in a laboratory study can be very different from that under field conditions.  Certain factors such as substrate restraint condition, environmental conditions, thickness of the overlay, and construction practices can affect the performance and behavior of overlays on site.  In this context, the scope of this research project was extended to study the performance of some ‘crack-free’ overlay material developed during laboratory testing by applying using it in real field investigations.  Two instrumented field projects with embedded strain sensors were carried out to demonstrate reduction in early age strain in the material that corresponds to a reduced cracking potential.  Results from these studies including data from some non destructive tests (NDTs) are presented in Chapter 7.    6  1.2.4 Concluding Remarks  The completion of this research project has led to significant improvement in our understanding of the early-age behavior of cement-based composites both in the laboratory and in the field.  A novel test technique was developed and several modifications were made to the test set-up to monitor the behavior of the overlay using sensors.  Effect of mixture proportion on cracking was studied for unreinforced mixes using this technique. Water-cement ratio (w/c) and aggregate content were identified as the key parameters that affected cracking in the overlays.  For the range of sand-cement ratio (s/c) investigated, no definite trend in the change of crack area could be established. Various volume fractions of a total of sixteen different fiber types including cellulose, synthetic, and glass were used to reinforce the overlay material.  Glass fiber showed the greatest improvement in reducing crack widths and crack areas.  A small volume fraction of 0.1% was sufficient to eliminate any shrinkage cracking.  The test technique was also used to further compare and characterize the performance of various fiber types. Cellulose and synthetic fibers were selected to further study the performance of thin overlays and slabs-on-grade in real field conditions.  The test slabs and the overlay were instrumented and the data monitored over the internet.  The strain data indicated that the strain in the fiber-reinforced overlays was lower when compared to that of the unreinforced overlays, which indicated a lower cracking potential.  NDTs such as “rebound hammer tests,” “ultrasonic pulse velocity,” and “resistivity tests,” were performed to compare the performance of a reinforced overlay to that of an un-reinforced overlay.  These results are presented in details in Chapter 7. 7  Since hybrid fiber combinations are now becoming popular in the industry, it is recommended that future work also include study of such fiber combinations.  Future work can also include a comparison of performance of SRAs with that of fibers.  It is recommended that future work include use of automated image analysis software to monitor crack growth during laboratory experiments and use of low modulus strain gauges for field investigations. Some other recommendations for future research are described in Chapter 9.   8 CHAPTER 2- LITERATURE REVIEW  2.1 Introduction  As described in Chapter 1, this research project focused on studying the early-age cracking characteristics of cementitious composites, especially when used as bonded overlays.  Developing a test technique that could be used to measure restrained plastic shrinkage cracking of cement-based composites and using the technique to develop ‘crack-free’ overlays formed the core of this project.  Although significant amount of research has been carried out to understand free shrinkage of cement-based materials, limited research has been carried out on bonded overlays that have high potential of cracking. In this chapter, a review of the research carried out in this area is presented.  2.2 Shrinkage of Cement-Based Materials  Cement-based materials shrink due to loss of moisture (evaporation) under the influence of external factors.  The key external factors and their influence on evaporation rate is well summarized by the ACI evaporation nomograph (Figure 2.1a) contained in the ACI Manual of Concrete Practice, Section 305R, “Hot Weather Concreting,” (ACI 305R-96, 1996).  According to this nomograph, the rate of evaporation is affected by air temperature, relative humidity, concrete temperature, and wind velocity.  This nomograph is widely used to predict the evaporation rate under the given environmental or exposure conditions.  Other than the external factors, early-age shrinkage is affected by the hydration reaction and the physical properties of the materials used (particle size and distribution, 9 and the water-binder ratio).  Various factors are known to contribute to early-age volume change and a phenomenological summary described by Lange David in chapter 3 of the report edited by Bentur A. (2003) is reproduced in Figure 2.1b.  As far as volume change during the very early ages is concerned, which is the focus of this thesis, autogenous, plastic and drying shrinkage are the key components and these are briefly described below.    Figure 2.1a - ACI Nomograph for Evaporation Rate Estimation (Kosmatka et al., 2002)  10  Thermal Deformation External Heat Release Autogenous External Drying Basic Creep Drying Creep Redistribution Early Hydration  Influences from Shrinkage Shrinkage (under load, (Total Creep of Bleed Water Hydration no drying) less basic creep) or Water from Chemical Aggregate Shrinkage Cement Hydration SwellingCreepShrinkage Early Age Volume Change   Figure 2.1b - Phenomenological Summary of Early-Age Volume Change (Redrawn, adopted from Bentur, A, 2003) 11  Autogenous/Chemical Shrinkage: Chemical shrinkage is generally the driving force behind autogenous shrinkage.  Autogenous shrinkage occurs during the cement hydration stage when the total volume reduces without any loss of moisture to the environment. This type of shrinkage causes external macroscopical reduction of the cementitious system.  Chemical shrinkage on the other hand causes internal-microscopic reduction in the volume since the volume of the hydration products is less than the total volume of the constituents (water + cement) undergoing the reaction.  Hence, chemical shrinkage at the molecular level leads to autogenous shrinkage at a larger scale.  Plastic Shrinkage: In chapter 2 of the report edited by Bentur (2003), Bentur A defines plastic shrinkage as, “the external-macroscopial (bulk) linear and volume changes occurring in the system at the stage where it can be treated as a fluid, namely, before a solid self-supportive skeleton develops, roughly at about the setting time.” Cement pastes in the plastic state, may undergo volumetric contractions as high as 1% of the absolute volume of dry cement (Banthia & Nandakumar 2001). This is due to both the evaporation of mixing water and the autogenous process of concrete hydration.  Plastic shrinkage cracking is a more concerning issue with high performance concrete since such concretes generally have more fines that lead to less bleeding.  Shrinkage in the plastic stage results in shrinkage strains but also leads to subsidence or settlement. According to Qi et al. (2003) cracks can develop in the material because of the combined effects of shrinkage and settlement. According to Japan Concrete Institute (JCI), subsidence is defined as, “vertical displacement of the top 12 surface of cementitious materials before initial setting, due to bleeding, chemical shrinkage and so on.  The displacement is measured by the top surface of substantial solid phase and not by the surface bleeding water.”  Concrete subsides both after finishing and during bleeding and this subsidence is expected to be higher when the available water in concrete is higher.  According to Babaei and Fouladgar (1997), the typical rate of bleeding in concrete is 0.98 kg/m2/hr i.e. 0.2 lb/ft2/hr, which if exceeded by the evaporation rate for freshly cast concrete, can cause serious plastic shrinkage cracking. Lower w/c ratio will cause lesser bleeding and this reduces the allowable rate of evaporation for freshly placed concrete.  Hence, this requires that additional precautionary measures be taken.  Drying Shrinkage: Much like plastic shrinkage, drying shrinkage is caused due to loss of moisture caused by evaporation from the specimens. However, drying shrinkage is known to continue for a long period of time even after the cementitious material has gained full strength.  Hansen W. (1987) defines drying shrinkage as, “the time-dependent deformation due to loss of water at constant temperature and relative humidity (RH).” Even under the influence of varying temperature and humidity, concrete is expected to experience drying shrinkage.  Hansen further describes that the four key factors responsible for drying shrinkage are surface energy, capillary tension, movement of inter- layer water and the disjoining pressure.  As far as the development of cracks is concerned, the type and extent of restraint developed during shrinkage is equally vital.  The restraint can either be internal or external.  In chapter 3.6 of the report edited by Bentur (2003), Bisschop points out that 13 the internal restraint can develop either due to non-uniform shrinkage caused by a moisture gradient, due to the presence of stiff aggregates, or due to presence of steel rebars in reinforced concrete.  Thermal Expansion: Cement-based materials might also expand during the first few hours (6-12 hours) due to an increase in temperature that causes expansion of the material; also termed as thermal volume change (Lepage et al., 1999).  It should however be noted that initial expansion before setting can also be due to re-absorption of bleed water which is accumulated on the surface of concrete, at the time of setting (Bjontegaard 1999, and Bjontegaard and Sellevold, 2001).  Effect of Shrinkage on Cementitious Materials:  Plain, unreinforced cementitious materials are characterized by low tensile strengths, and especially low tensile strain capacities. As described earlier, concrete during setting shrinks, and if restrained, may crack under tensile stresses. Much of the cracking takes place in the cement-based materials during the first few hours of adding water to the cement. Some of the factors affecting restrained shrinkage are environmental conditions that control the rate of drying severity of restraint, and the properties of cement paste and aggregates.  Under restrained condition, this contraction can cause strains far in excess of those needed to cause cracking in young pastes with poorly developed strength. In spite of every effort, plastic shrinkage cracking still remains a real concern, particularly in large surface area placements like slabs on grade, thin surface repairs, patching, shotcrete tunnel linings, etc. In these applications, the exposed surface area per unit volume of the overlay 14 material is high and the old concrete substrate or the rock surface offers a high degree of restraint.  2.3 Test Techniques  Cracking in young concrete placed as an overlay is caused due to high restraint from the sub-base combined with significant loss of moisture at early ages. There are two ways of evaluating the performance of a cementitious material as an overlay: first under restrained conditions and second in unrestrained condition when the overlay is allowed to shrink freely.  2.3.1 Un-Restrained Tests  Free shrinkage tests help determine the potential of material to shrink but do not necessarily help characterize material based on their cracking potential.  This is big limitation of these tests especially for fiber-reinforced cement-based composites, where, the effect of fibers is predominantly after cracking.  According to Hossain et al. (2003) free shrinkage tests are not always sufficient for detecting cracking potential of materials since cracking is influenced by a complex combination of factors such as strength and stiffness development, creep, shrinkage, degree of restraint, and toughness.  Kronlof et al., 1995 and Qi et al., 2005 have tested cement-based materials to predict their durability in structures by measuring free shrinkage strains and settlement.  Existing free shrinkage tests are typically simple and mainly involve measurement of reduction in a certain dimension of the specimens (Kronlof et al., 1995, Kovler, 1995).  These specimens are generally in the shape of a prism.  Others have reported measuring shrinkage using non 15 contact laser (Qi et al., 2005), measuring strains using fiber bragg gratings to measure strains (Slowik et al., 2004), and relating bleeding and evaporation to plastic shrinkage cracking (Topcu and Elgun, 2004).  Measurement of strain in systems that do not crack can be useful in predicting the cracking potential.  Strain monitoring technique was used in this project for two field investigations, which is described in Chapter 7.  2.3.2 Restrained Tests  As opposed to the free shrinkage tests, there are several existing techniques of studying shrinkage induced cracking in cement-based materials.  These include, for example, a ring type specimen (Grzybowski & Shah, 1990), a linear specimen with anchored ends (Banthia et al., 1993), a linear specimen held between a movable and a fixed grip such that a complete restraint and one-dimensional fixity are achieved by returning the movable grip to the original position after shrinkage (Bloom & Bentur, 1995), and a plate type specimen where the restraint is provided in two orthogonal directions (Khajuria & Balaguru, 1992).  Discussion on some of these important tests is presented in the following sections.  2.3.2.1 Ring Type Tests  Ring type tests have been widely used to predominantly study shrinkage cracking in cement-based materials.  These tests are based on the principle that the steel ring resists the shrinkage of the concrete ring, resulting in development of tensile stresses, which can cause cracking. There is admittedly non-uniformity of the stress field in such tests, but it has been suggested that for certain geometries the difference between the 16 tangential radial stresses at the inner and outer surface of concrete ring are small enough (10 to 20%) to consider the stress field as uniform (Grzybowski and Shah, 1990). Test set-up used by Shah et al. (1998) where a concrete annulus is cast around a steel ring is shown in Figure 2.2.  This arrangement is very similar to that proposed in the AASHTO standard (AASHTO, 1998).  This type of steel ring test is also an ASTM standard (ASTM C 1581).  In this method, the compressive strain, the age at cracking, and tensile stresses developed in the material are measured.   Figure 2.2 - Restrained Ring Test (Shah et al., 1998)   A similar test method, standardized in Scandanavian countries, and used to evaluate the cracking tendency of concrete at early ages is herein called the Nordtest Method (1995) and has a designation NT BUILD 433.  In this method, ring specimens are tested under controlled conditions.  Figure 2.3a shows the test specimen and the set- up that consists of inner and outer rings that are seated on an oiled base plate. This is a simple method that was developed to determine the cracking tendency of concrete at early ages.  No research that compares this test method to other similar test methods could be found, however, a typical cracked specimen is shown in Figure 2.3b. 17   (a) Test Specimen   (b) Cracked Specimen Figure 2.3 - Nordtest Shrinkage Test (Nordtest, 1995) 18 Kovler et al. (1993) modified the conventional steel ring test to improve the long duration (generally 5-6 weeks) required for the cracks to appear during standard curing conditions.  They increased the crack sensitivity by using a Perspex core having a high coefficient of thermal expansion.  Initial tests did not significantly improve the duration of cracking.  Later, the Perspex core was further modified to include a stress concentrator.  This modification led to significant improvement, with the time required for the first micro crack to appear reducing to 1-2 minutes after exposure to drying conditions.  Toledo Filho et al. (2005) used a ring type specimen with an external plastic cylindrical mould of 150 mm internal diameter and 50 mm height.  The mixes were placed between the outer ring and the central concrete cubic core of 70 mm dimensions as shown in Figure 2.4.   Figure 2.4 - Test Setup, Ring Type Specimen (source: Toledo Filho et al., 2005) 19  In the ring tests described above cracks develop in the test specimens due to a combination of plastic and drying shrinkage and hence the true effect of only plastic shrinkage cannot be not clearly quantified.  2.3.2.2 Bonded Overlay Technique Developed at UBC  While effective for comparative measurements, most of the techniques described in the previous section produce stress fields in the specimen that are different from those occurring in reality, especially in the case of thin overlays.  A technique producing realistic shrinkage conditions was developed (Banthia & Campbell 1998, Banthia et al. 1996, Banthia & Yan 2000).  In this method, a layer of fresh concrete (or shotcrete) is placed directly on a fully hardened substrate. This ‘old substrate’ has surface protrusions, which enhance its roughness and in turn imposes a restraint uniformly at the base of a shrinking overlay. The whole assembly is then subjected to a drying environment that creates resultant stresses due to a combination of stresses due to base restraint and differential stresses due to shrinkage (Figure 2.5).  These resultant stresses in turn lead to cracking in the overlay.  This technique is fully described in detail in Chapter 4. Some other test techniques, where a slab-type specimen is tested are described in the following section.   20  Figure 2.5 – Schematic of Tensile Stress development in the shrinking overlay   2.3.2.3 Slab/Overlay Tests  In some test techniques, restrained shrinkage cracking of cementitious materials is measured using a slab that is cast directly over a rough surface that provides the restraint. Soroushian et al. (1993) tested slabs of size 533 x 838 mm, 38 mm in thickness that were exposed to environmental conditions.  After 5hrs of exposure, crack widths and lengths were measured.  This technique was later utilized by Bayasi and McIntyre (2002) to study the effect of fibers on plastic shrinkage of concrete.  In this technique, a concrete slab is cast in a plywood mold and is placed under controlled environmental conditions as shown in Figure 2.6 and crack characteristics are measured using a microscope (Figure 2.7).  Due to the aspect ratio of the slabs, generally cracks can develop randomly in any direction, which better represents the cracking pattern expected in thin overlays. However, some of the slab type specimens as described above are very large in size and hence limit the use of such tests.  21  Figure 2.6 - Test Setup using a Plywood Mold (source: Bayasi and McIntyre, 2002)   Figure 2.7 - Crack Measurement (source: Bayasi and McIntyre, 2002)  Much like the technique developed by Banthia et al. (1996), Naaman et al. (2005) have developed a test technique where the mix being investigated is cast directly over a grooved and hard concrete substrate that provides the restraint to the shrinking mix. Details of the environmental chamber and the substrate are shown in Figures 2.8 (a) and (b) respectively. Qi et al. (2003) also report using slab specimens with a stress riser but have conducted tests at accelerated drying conditions.  Stress risers can result in consistent crack patterns, however the restraint imposed by the stress riser used in these 22 tests do not represent the real restraint conditions.  Also, the thickness of the overlay used by Qi et. al is less than 35 mm when compared to a 60 mm overlay.  This thickness can again limit the use of fibers more than a certain length in the mixes.    Figure 2.8 (a) - Plan and profile view of the Environmental Chamber (Naaman et al., 2005)  23  Figure 2.8 (b) - Test Setup (Naaman et al., 2005)    Plate specimens simulating slabs-on-grade were also tested by Trottier et al. (2002).  Rectangular plate specimens 900 x 600 x 50 mm were used for the plastic shrinkage cracking tests. The molds were made of 20 mm-thick plywood with metallic and wooden restraints systematically placed at the bottom and on the sides of the mold.  Another technique where a stress riser is used to create cracking in slabs using a steel insert plate is described by Lura et al. (2007).  This test set-up (Figure 2.8 (c)) was recently adopted as an ASTM standard (ASTM C 1579) and is entitled, “Standard Test Method for Evaluating Plastic Shrinkage Cracking of Restrained Fiber Reinforced Concrete (Using a Steel Form Insert).” This technique can be used to study the effect of fibers and or other additives on the performance of concrete under restrained shrinkage conditions.  However, as described before such tests may not be best suited to study the 24 performance of fiber-reinforced overlays for the following reasons:  The thickness of the overlay being tested is only 36.5 mm over the middle stress riser, which does not simulate the real conditions of bleeding and strain development through the depth of the overlay.  In this test, the restraint developed by the stress risers may not simulate the conditions experienced by an overlay in real conditions and moreover, due to the thickness of the overlay, the maximum length of the fiber that can be tested would also be limited.     Figure 2.8 (c) - Test Setup with Stress Riser (Lura et al., 2007)     25 2.3.2.4 Other Tests  Najm and Balaguru (2002) have used a restrained shrinkage test where instead of developing restraint using the base, a strip of 12.5 x 25 mm wire mesh was nailed to the base to provide restraint along the perimeter.  The arrangement consisted of a mold constructed using a plywood base, a tile board glued to the top of the plywood to obtain a smooth non-absorbing surface, and Plexiglas edges.  A thin polyethylene sheet was placed at the top of the tile board to eliminate adhesion between the mortar and the tile board (Figure 2.9).   Figure 2.9 - Test Setup (Najm and Balaguru, 2002)   Unlike the slab specimens described above, where restraint is developed by a substrate, tests were conducted by Sanjuan and Moragues (1994) where only plastic shrinkage strains were measured to evaluate performance of mortars and concrete with admixtures and polypropylene fibers. Extensometers located on steel plates (F in Figure 2.10) and connected to other steel plates by a steel rod (E in Figure) were used.  Slab type specimens, 20 mm thick were used for the tests.  Test equipment consisted of a chamber, in which air flow speed and temperature were kept constant.  The equipment consisted of 26 a fan (B in Figure), an electrical resistance (C in Figure) and a control system (A in Figure) to hold the air flow speed and temperature constant.  The set-up included a thermometer (D in Figure) and relative humidity was recorded using a hygrometer (G in Figure).   Figure 2.10 - Test Setup (Sanjuan and Moragues, 1994)  Kim & Weiss (2003) and Hossain et al. (2003) describe a technique where a steel testing frame is used to test fiber reinforced mortars under restrained conditions and further used acoustic emission (AE) to monitor early age cracking as seen in Figure 2.11. This could be an advantage when precise measurement of “time to first crack” is required.  Square specimens (25 mm in size) were used with plastic notches placed 50 mm from the center to cause cracking at a known location.  Some of the tests described above will be described in greater detail in this section.  27  Figure 2.11 - Test Setup with AE sensors (Kim and Weiss, 2003)   Bloom and Bentur (1995) used 40 x 40 x 1000 mm long bars for free and restrained shrinkage tests using the set-up shown in Figure 2.12.  The specimens were wider at the end to fit into a fixed and movable grip. Movement of the grip was measured using a gauge (#9 in Figure 2.12 (a)).  The other parts of the equipment included: screw assembly (#7 and 8) connected to a grip (#5) through a load cell (#6).  In this test, the specimen is allowed to shrink and when 2 µm strain is recorded, restraint is achieved by moving the grips to their original position and the corresponding load is recorded.  28  Figure 2.12 (a) - Test Set-up used by Bloom and Bentur (1995)   The prototype developed by Bloom and Bentur was further modified by Kovler (1994) to produce an automated and high precision system using a closed-loop machine. In this technique, specimen size of 40 mm x 40 mm and 1.0 m long are held in special grips (Figure 2.12 b) and subjected to uniaxial shrinkage.  As opposed to the other techniques, in this technique creep and shrinkage strains can be distinguished, since two specimens (one under free and the other under restrained condition) are simultaneously tested.  This technique can be used to determine creep coefficients and elastic moduli, however since the tensile stresses developed in the specimens are slightly less than the concrete strength, cracks do not develop.  29  Figure 2.12 (b) - Specimen Size and Grips used by Kovler (1994)   A similar testing bench called “self-cracking test” developed at Laboratoires des Ponts et Chaussées (LCPC) in Paris was used by Paillere (1989) to study the effect of fibers on concrete containing silica fume.  This test bench (Figure 2.12 (c)) is used to maintain the length of the specimens constant and includes an arrangement to measure load using a dynamometer and change in length.  For studying fresh concrete, the bench is placed horizontally.  As opposed to the technique described above, full depth cracks develop in the specimens within a few days, which can be used to study the performance of fibers.  30  Figure 2.12 (c) - Cracking Test Apparatus or Bench (source: Paillere (1989))   2.4 Effect of Fiber Reinforcement on Shrinkage  Plastic shrinkage cracking has been a major issue in concrete construction. This type of cracking is often caused by an excessive evaporation of water that cannot be replenished through bleeding. Due to the negative effects of shrinkage cracking, numerous experiments and studies are conducted to control the severity of cracking.  Many solutions proposed to reduce retrained shrinkage cracking, include the use of shrinkage reducing admixtures (SRA) (Folliard and Berke, 1997).  Weiss and Shah (1997) have also investigated the use of SRA, shrinkage compensating concrete, and fiber-reinforcement to limit shrinkage cracking.  Among the many solutions proposed, the most promising one is the use of randomly distributed fibers of steel, polypropylene, 31 carbon, etc., that provide bridging forces across cracks and thus prevent them from growing and propagating.  The presence of fiber is expected to reduce both the lengths and widths of shrinkage induced cracks (Banthia et al. 1993, Bloom & Bentur 1995, Grzybowski & Shah 1990, Khajuria & Balaguru 1992) and reduce damage at the interface in repairs enhancing the strength of the interfacial bond (Banthia & Dubeau 1994).  In addition, fiber reinforcement (depending on fiber type) is expected to improve the mechanical performance, deformability, toughness, impact resistance and fatigue endurance of the overlay—properties that are highly desirable from a repair point of view (Emmons et al. 1993, Pigeon & Bisonette 1999, Rossi 1994).  Wang et al. (2001) conducted a study where the effect of fibers on concrete pore structure and the resulting effect on plastic shrinkage cracking of concrete was investigated.  Specimens were dried, and water loss rates and initial crack times were recorded.  Crack widths and areas were measured by means of image analysis, and the pore structure was analyzed with a mercury-intrusion porosimeter.  It was found that the addition of 0.1% volume of fibers increased the number of large pores in cement paste, thereby changing bleeding behaviour and reducing the plastic shrinkage crack area by 30 to 40%.  Another study by Soroushian and Ravanbakhsh (1998) involved the use of cellulose fibers at 0.06% volume fraction (0.9 kg/m3).  The test results showed that cellulose fibers are effective in reducing the plastic shrinkage of both normal and high- strength concrete.  Kronlof et al. (1995) studied the effect of polypropylene fibers (1% by volume) and found that free plastic shrinkage reduced by about 30% when the mixes contained fibers.  The tests were conducted at 20°C and a RH of 40%. Qi et al. (2005) 32 also investigated the effect of polypropylene fibers (19 mm long, 0.2% volume) on settlement of concrete during the plastic stage.  Settlement was measured using non contact laser and tests indicated that fibers not only resulted in lower but also more uniform settlement.  Bayasi and McIntyre (2002) used a technique developed by Soroushian et al. (1993) where a concrete slab was cast in a plywood mold to study the effect of fibers on restrained plastic shrinkage cracking.  They found that fibrillated polypropylene fibers reduce the evaporation rate by 10-15% and a small dosage of 0.1% by volume of fibers reduces the crack area, maximum width, and average width by approximately 90, 80, and 40%. Similarly, Byounggeon and Weiss (2003) also found that fibers arrest plastic shrinkage cracks and restrain their growth in mortars tested under restrained conditions. Monofilament polypropylene fibers and steel fibers when tested using the ring test were found to reduce restrained shrinkage cracking (Pease et al., 2005). Shah et al. (1998) studied the effect of steel, polypropylene, and cellulose fibers using the ring test and found that randomly distributed fibers can reduce the age at appearance of first visible crack and can significantly reduce crack widths. Different fiber compositions can alter the degree to which this occurs. Kovler et al. (1993) used a modified ring test and found that 0.1 and 0.2% of polypropylene fibers reduced restrained plastic shrinkage cracking. Qi et al. (2003) tested monofilament and fibrillated polypropylene fibers using slab specimens cast over a stress riser.  Clear reduction in crack width was observed with increasing dosage of fibers.  Naaman et al. (2005) evaluated the effect of various types of synthetic and flexible metallic fibers under restrained shrinkage condition.  They used prismatic specimens cast over grooved substrates and found that 0.2% of most fine 33 diameter fibers provided a reasonable control of cracking; cracking reduced to about 10% of that recorded for control mixes.  As opposed to this, Najm and Balaguru (2002) tested large-diameter polymeric fibers with diameter ranging between 0.15 – 0.64 mm.  Even the large diameter fibers reduced plastic shrinkage cracking: a 1% addition by volume reduced the crack width by 60%.  Trottier et al. (2002) tested plate specimens to simulate slabs-on-grade and compared the plastic shrinkage performance of Welded-Wire-Fabric (WWF) to concrete reinforced with low dosage (0.03% - 0.1%) of low-denier fibrillated and monofilament synthetic fibers.  Crack width in the specimens with WWF was the highest amongst all specimens investigated and crack area was about 82% of that recorded for plain concrete. Most fiber-reinforced concrete mixtures had lower crack areas, widths, and number of cracks when compared to the specimens containing WWF. Similar research was conducted by Voigt et al. (2004) where restrained shrinkage of FRC was compared to concrete reinforced with WWF.  Voigt et al. used restrained shrinkage ring tests according to AASHTO PP34 (1998).  Different fiber types and fiber blends were found to be effective in controlling cracking. When 0.25% volume of steel fibers were used, maximum crack width was comparable to that observed for specimens with WWF. Toledo Filho et al. (2005) investigated the effect of natural fibers on free and restrained plastic shrinkage of cement mortar.  Addition of 0.2% by volume of sisal fibers (25 mm long) significantly reduced the free plastic shrinkage.  Similarly, addition of 0.2% of sisal and coconut fibers delayed restrained shrinkage cracking and controlled crack development. Altoubat and Lange (2001) conducted restrained shrinkage tests on uniaxial dog bone shape specimens and investigated the effect of steel fibers.  They found that 34 steel fibers significantly delayed fracture of normal strength concrete without much influencing the stress at failure.  2.5 Effect of Mix Proportion, Admixtures, and Additives  Other than studying the effect of fibers, shrinkage tests have also been used to evaluate the effect of mix proportion, various additives and admixtures on various forms of shrinkage.  Factors such as water-cement or water-binder ratio (Altoubat & Lange, 2001; Sanjuan & Moragues, 1994; Igarashi et al., 2000), cement-sand ratio (Sanjuan & Moragues, 1994), drying conditions (Altoubat & Lange, 2001), curing conditions (Altoubat & Lange, 2001; Toledo Filho et al., 2005), mix proportions (Toledo Filho et al., 2005), and admixtures such as silica fume (Bayasi & McIntyre, 2002; Toledo Filho et al., 2005, Igarashi et al., 2000, Subramaniam et al., 2005), blast furnace slag (Toledo Filho et al., 2005), shrinkage reducing admixture (Pease et al., 2005), and fly ash (Atis, 2003, Subramaniam et al., 2005) have been investigated.  Most research described above has focused on studying the influence of drying shrinkage on performance of concrete. Effect of mix proportion on various forms of shrinkage cracking has been studied by many researchers (Uno 1998, Almusallam et al. 1998, Shaeles and Hover 1988, Wang et al. 2001, Topcu and Elgun 2004, Atis 2003, Lee et al. 2003, Naik et al. 1995, Nelson et al. 1992, Day 1990, Gettu et al. 2002, Bentz et al. 2001). Almusallam et al (1998) studied the cumulative effect of cement content, w/c, and hot and humid environment on time and intensity of cracking of slabs 450 x 450 x 20 mm in size and found that both cement content and w/c significantly affect plastic shrinkage cracking of concrete.  Similarly, Uno (1998) reported that factors such as w/c, fines content, member size, admixtures, and 35 building practices influence plastic shrinkage cracking but found the rate of evaporation of water to be the most important factor of all.  Plastic shrinkage cracking as discussed previously is significantly affected by bleeding and evaporation rates.  Topcu and Elgun (2004) investigated the influence of temperature, humidity, and windy conditions on rate of evaporation and bleeding.  They found that increase in cement content reduced bleeding rate and mixtures with higher mix water resulted in increased evaporation rates. Typical effects of fly ash in concrete mixtures are fairly well documented.  Fly ash reduces the water requirement for the same slump, reduces the bleed water, and increases the setting time.  In terms of shrinkage, the majority of studies have concluded that fly ash affects different forms of concrete shrinkage.  It has been reported that more durable composites can be produced by replacing part of cement with fly-ash, which is known to reduce cracking resulting from drying shrinkage (Atis 2003, , Naik et al. 1995, Nelson et al. 1992, and Day 1990), and autogenous shrinkage (Lee et al. 2003).  Effect of higher fly-ash replacement on drying shrinkage was investigated by Atis C (2003). Prisms were cast, cured in a curing room and length changes were recorded.  Drying shrinkage in fly ash concrete was up to 30% lower when compared to ordinary concrete for 50-70% replacements.  There are also studies that found that fly ash decreases the shrinkage strains of concrete resulting in less cracking.  Kosmatka et al. (2002) concluded that the effect of fly ash on drying shrinkage is “generally small and of little practical significance.” Sundaram et al. (1989) found that concrete containing large volumes of Type F (low calcium) fly ash had drying shrinkage strains comparable to or lower than the control. 36 Subramaniam (2005) found that increasing levels of ultra-fine fly ash replacement did not change the drying shrinkage of concrete.  A study was conducted by Li et al. (1999) where restrained drying shrinkage cracking of concrete was measured using ring-type specimens.  The effects of the use of different pozzolans on the total width of cracking were investigated, and it was found that crack widths increased with increasing quantities of fly ash. As seen above, there are inconsistencies in conclusions drawn from various studies and the effect of fly ash on plastic shrinkage is not completely understood.  Some specific findings indicating that the effect of fly-ash on shrinkage is not completely understood is summarized in Table 2.1.              37  Table 2.1 - Effect of Fly-Ash on Shrinkage  Reference Fly-ash Type Cement Replacement (%) Type of Shrinkage/ Test Method Used Effect on Shrinkage Atiş (2003) Class F 50-70% Drying shrinkage, Recorded length change in 50 x 50 x 200 mm prisms 30% reduction, 50% for mixes with superplasticizer Atis, (2003) High- calcium nonstandard 10%, 20%, 30% and 40%.  Drying shrinkage, Recorded length change in 25.3 x 25.3 x 284.6 mm prisms 40% reduction Gesoglu et al. (2006) Class C Various Free shrinkage + restrained shrinkage cracking on ring type specimens Reduction in free shrinkage and 35% reduction in crack width in restrained tests Class C 20% and 50% Naik et al. (1995) Class F 40% Drying shrinkage, Recorded  length change in 75 x 100 x 280 mm prisms. Reduction in shrinkage Subramaniam (2005) Class F Approximately 8-11% Free shrinkage on 100 x 100 x 400 mm prism + Restrained shrinkage with ring tests Increase in volume of fly-ash decreased the autogenous shrinkage; No change noted in drying shrinkage Li et al. (1999) Not stated 25% Unrestrained and restrained shrinkage using ring type specimens) Increase in free shrinkage and crack width    Shrinkage reducing admixtures (SRAs) are also known to be effective in reducing shrinkage cracking (Gettu et al. 2002 and Bentz et al. 2001). Pease et al. (2005) studied the effect of SRA in reducing restrained shrinkage.  They used restrained ring 38 specimens for the tests and found that increasing dosage of SRA reduced overall shrinkage.  SRAs also reduced shrinkage stresses at early ages and reduced the cracking potential.  Bentz et al. (2001) also investigated the effect of SRAs on drying shrinkage. They found that in sealed systems, SRAs helped maintain a greater internal RH and reduced autogenous deformation.  They also found that SRAs were particularly more effective in reducing self desiccation and drying shrinkage in mixes where the w/c ratio was low.  Clearly, a significant amount of work remains to be done to fully understand the influence of mix proportion, additives, and admixtures on shrinkage induced cracking in cementitious materials, especially when used in repairs and overlays.   2.6 Proposed Models  Various types of shrinkage were described in the sections above, however, as far as prediction of shrinkage is concerned, numerous proposed models only focus on predicting the drying shrinkage of cementitious materials. Goel et al. (2007) summarize and compare various existing models: ACI-209R-82 model, the B3 model, the CEB-FIP model code 1990, and the GL2000 (Gardner and Lockman, 2001), and Muller model (Muller et al., 1999). Expressions for some of these important models to predict creep and shrinkage as described by Goel et al. (2007) are presented below.  ACI 209 Code Provisions ACI-209R-82 recommends the following expression for shrinkage:  39 shu cc c csh ttT tttt εε )( )(),( −+ −=        2.1  where, tc= 7 days for moist cured concrete, and 1-3 days for steam cured concrete, εshu= ultimate shrinkage strain, 780 for standard conditions, t is age of concrete in days, Tc = 35 days for moist cured concrete and 55 days for steam cured concrete,  CEB-FIP Model Code 1990  According to this model, total shrinkage/swelling is calculated based on the following equation:  )}() 100 2(350{ }{10)]1.09(10160[),( 2 6 c c c RHcmsccsh ttA ttftt −+ −×−+= − µ ββε  2.2 scβ  is constant, depends on type of cement, fcm is the concrete mean compressive strength at 28 days in MPa, RHβ  is a constant that depends on relative humidity, µ  perimeter of the member in contact with atmosphere (mm), cA  cross-sectional area (mm 2), and other variables as defined before.  B3 Model Mean shrinkage strain in the cross section is given by:  )(),( tSktt hshucsh εε −=         2.3 40 where, kh is the humidity dependence, S(t) is the time curve, and other variables as defined before.  The ultimate shrinkage strain is given by: )( )6007()270)(091.0( 28.01.221 shcc c cmshu tE Efw τααε + ++= −     2.4 where, w is the water content in kg/m3, 1α  and 2α  are constants related to the cement type and curing condition, cE  is the modulus of elasticity of concrete at the age of 28 days (MPa), shτ  is the shrinkage half-time (days), and other variables as defined before.  The time function is given by: sh ctttS τ −= tanh)(          2.5 where, t and tc are the age of concrete and the age drying commenced, end of moist curing in days, respectively, τsh is the shrinkage half-time as given in Equation 2.6, h is the relative humidity of the environment at ambient temperature (in decimal), and other variables as defined before.  225.008.0 })/{2()(085.0 SVkft scmcsh −−=τ       2.6 where, ks is the cross-section shape correction factor, and V/S is the volume-surface ratio in mm.   41 GL2000 Model This model proposes the following equation:  ] )/(15.0 [)18.11(),( 2 4 SVtt tthtt c c shucsh +− −−= εε      2.8 where, variables as described before.  610301000 −×= cm shu f Kε         2.9 where, K is a shrinkage constant that depends on the cement type and other variables as described before.   Above-mentioned models were used for predicting creep and shrinkage of various grades of concrete by Goel et al. (2007). They found that the predictions from the GL2000 model were the closest to the experimental results.  Shrinkage estimated using the fib 2000 model has been found to be about 75% of the measured shrinkage (Bentur A, 2003). ACI 209 (2005) reports predictions from the four models described above were compared to that of the results in the RILEM data bank. GL2000 and B3 models provided the best fit for shrinkage strain and the CEB model underestimated the strain.  Coefficient of variation was calculated to compare model predictions with test data.  The variation was 35% and 36% for GL2000 and B3 respectively. Variation for CEB was 37% and finally ACI model resulted in a variation of 45%.  It was also noted that the use of more input data (test results) in the models improved the predictions of all models accept for the ACI model. 42  Many other microstructural models have been proposed to predict behavior and shrinkage of cement-based materials.  These models are based on the cement hydration process and are well described in chapter 5.1 by Charron J. –P., Marchand J., Bissonnette B., and Pigeon M. in the report edited by Bentur A (2003).  The model “CEMHYD3D” was developed at the National Institute of Standards and Technology in the US.  This model has been modified over the past few years and the latest version of this model is described by Bentz D.P. (2006a, 2006b, 2006c). “HYMOSTRUC” (Hydration Morphology and Structural Development) model was developed at Delft University of Technology in the Netherlands.  The details and application of this model to study formation of microstructure of cement and concrete is reported by Ye G. et al. (2003) and Princigallo A. et al. (2003).  Some other models such as “DuCON” (Ishida et al., 1998) developed at the Tokyo University and “ENPC” (Hua, C et al., 1995, 1997) that was developed in France, have been used to predict shrinkage of cement based materials. These models are based on the assumption that capillary effects such as the formation of capillary depression due to tension in the liquid phase are responsible for autogenous shrinkage.  Hua et al. (1995) reports good agreement between predicted values with measured values for a w/c ratio of 0.42.  At the age of two days, 60 µm was measured when compared to a strain of 57.6µm from the model. There exist some other empirical models that were developed to predict behavior of concrete at early ages including the effect of basic creep and autogenous shrinkage of concrete.  Description of models such as “CESAR’s model,” “LeRoy’s model,” “Granger’s model,” “Bazant-Baweja’s model,” and “DeShutter and Taerwe’s model” and a comparison of their predicted results is reported by Charron, J.-P, et al. (2001a, 2001b). 43 Charron J. –P., Marchand J., Bissonnette B., Pigeon M, and Gerard B have commented on the effectiveness of the above-mentioned empirical models in chapter 5.3 of the report edited by Bentur A (2003).  According to him, for the mixes investigated during their study, except for the CESAR’s model, all other models did not result in very reliable predictions.  Most models described above have been developed to predict creep and shrinkage in concrete in terms of the strain development.  Limited work has been done to predict early-age behavior (plastic shrinkage) of concrete, especially in terms of evaluating the shrinkage cracking of fiber-reinforced cement-based composites.  44 CHAPTER 3 - MATERIAL PROPERTIES  3.1 Introduction  Different plain and fiber reinforced cement based composites were investigated to determine their performance during plastic shrinkage using the UBC bonded overlay technique (described in detail in Chapter 4). In this technique, the mix being studied is subjected to extreme temperature conditions and restraint is developed by creating bond between the overlay and the substrate.  Laboratory experiments focused on mortar mixes where effectiveness of different fiber types was studied.  Effect of fly-ash and other admixtures on plastic shrinkage cracking was also investigated.  3.2 Plain Mixes  Type 10 (CSA type GU) cement manufactured by Lafarge Canada Inc was used for the mixes.  Cement complied with ASTM C 150 and CSA A-3001 03 specifications. A typical mill test report for the cement is given in Table 3.1(a).  Locally available concrete sand and 14 mm concrete aggregate were used for the laboratory mixes (source of the aggregates was Earl Creek).  Typical sieve/gradation analysis of the fine and coarse aggregate was obtained from the vendor and is reproduced in Figures 3.1 (a) and (b) respectively.  The sand had a Fineness Modulus (FM) = 2.55 and both fine and coarse aggregates were washed by the vendor.  Potable water acceptable for making concrete was used for all the mixes.   45  Table 3.1 (a)- Physical and Chemical Data for Cement (Courtesy, Lafarge Canada Inc.) Physical Data Chemical Analysis Percent Fineness by Air Permeability (m2/kg; ASTM C204) 387 Silica Dioxide (%SiO2; ASTM C114) 19.7 Fineness by 45µm (No. 325) Sieve (% passing) 98.7 Aluminum Oxide (%Al2O3; ASTM C114) 4.6 Compressive Strength (ASTM C109/C109M) MPa psi Ferric Oxide (%Fe2O3; ASTM C114) 3.5 3-day 31.5 4570 Calcium Oxide (%CaO; ASTM C114) 64.8 7-day 40.5 5880 Magnessium Oxide (%MgO; ASTM C114) 0.7 28-day  0 Sulphur Trioxide (%SO3; ASTM C114) 2.7 Previous month 28 day strength 49.9 7240 Loss on Ignition (%L.O.I.; ASTM C114) 2.6 Time of Set, Vicat (Initial minutes; ASTM C191) 90 Insoluble Residue (%, ASTM C114) 0.32 Autoclave Expansion (%, ASTM C151) 0.00 Free lime (% of CaO) 1.2 Air Content (%, ASTM C185) 3 Tricalcium Silicate (%C3S; ASTM C150) 70 Colour (Lafarge Index) 30 Tricalcium Aluminate (%C3A; ASTM C150) 6  Total Alkali as Sodium Oxide (%NaEq; ASTM C150) 0.48  46 0 10 20 30 40 50 60 70 80 90 100 10 5 2. 5 1. 25 0. 63 0. 31 5 0. 16 0. 08 P an Sieve Size (mm) Pe rc en ta ge Individual % Retained % Passing  (a) Concrete Sand 0 10 20 30 40 50 60 70 80 90 100 20 14 10 5 2. 5 0. 08 P an Sieve Size (mm) Pe rc en ta ge Individual % Retained % Passing  (b) 14 mm Coarse Aggregate Figure 3.1 (a) and (b) - Gradation Analysis (redrawn; data from Lafarge Canada Inc.)  47 3.3 Fiber-Reinforced Mixes  Sixteen different fiber types consisting of synthetic, cellulose, and glass fibers were used in this investigation.  A complete list of these fibers along with their key properties is given in Table 3.2.  All the polypropylene fibers (PF) except type PF10 had a specific gravity of 0.91, tensile strength of 375MPa and modulus of elasticity of 3.5 GPa.  Specific gravity, tensile strength, and modulus of elasticity (in this order) for PF10, cellulose fibers (CF) and glass fibers (GF) in listed in Table 3.1 (b).   Table 3.1 (b) - Specific Gravity and Mechanical Properties of Fibers Fiber Type Specific Gravity Tensile Strength (MPa) Elastic of Modulus (GPa) PF10 0.92 620 9.5 CF1 1.1 620-895 8.3 CF2-CF5 1.5 700 (approximate) 10-40 GF1 2.68 1700 72    48   Table 3.2 - Fiber Properties Fiber Designation Fiber type Length (mm) Diameter (microns) or denier Pictures PF1 FM  150 (Propex Inc.) Multi dimension (~7 to 20) 20  PF2 FM  300 (Propex Inc.) Multi dimension (~15-20) 2600  PF3 FM  MD (Propex Inc.) 12.5 + 19 (50% each) (fibrillated) 6 PF4 FM  Stealth (Propex Inc.) 12.5 (monofilament) 3   PF5 FM  Stealth (Propex Inc.) 12.5 (monofilament) 6  PF6 FM Stealth (Propex Inc.) 6.35 (monofilament) 6  PF7 Fibrillated (Propex Inc.) 12.5 (Fibrillated) 1000  PF8 Microfilament (Grace micro fiber) 20 (~25)*   * Estimated value 49  Table 3.2 (Continued)- Fiber Properties Fiber Designation Fiber type Length (mm) Diameter (microns) or denier Pictures PF9** Blended fibers, Novomesh 950 (Propex Inc) 50 (macro) + multi design gradation (micro) 0.83 (macro) + multi design gradation (micro) PF10 Macro fiber, Strux 90/40 (Grace) 40 Aspect ratio = 90, (440)  CF1 UltraFiber 500 (Buckeye) 2.1-2.5 (Sized) 2.5  CF2 Bleached, long fiber (Weyerhaeuser 187) CF3 Bleached (Weyerhaeuser 144)  CF4 Weyerhaeuser 161 CF5 Long fiber (Weyerhaeuser 177) 2-3 0.9-2.7  GF1 AntiCrack fiber (Saint-Gobain) 18 Monofilament (14)   Legend: PF- Polypropylene Fiber, CF- Cellulose Fiber, and GF- Glass Fiber ** Blend of polypropylene/polyethylene macro-monofilament fibers with sinusoidal deformations & 100% virgin micro- synthetic fibers containing no reprocessed olefin materials 50   Magnified images of selected fibers taken using Scanning Electron Microscope (SEM) are also shown in Figures 3.2 through 3.7.  The SEM was a Hitachi, model S-3000 N with a magnification range of 20 to 200,000 and an accelerating voltage range of 0.1 kV to 30 kV. It had three electron detectors (secondary electrons, back-scattered electrons and environmental secondary electrons), an x-ray detector and was variable-pressure capable, with a vacuum range from high vacuum up to 270 Pa.  The fiber in this study had a sputtered film (approximately 10 nanometers thick) of 60% gold and 40% palladium.  The accelerating voltage of 20 kV and a high vacuum of 0.01 Pascal or less was used.  Two images were taken for each fiber; first a low magnification image (45 x) and the other higher magnification (between 150 x and 500 x), which are presented in Figures 3.2-3.7.  SEM images clearly depicted any fiber fibrillations (branching of fibers) on macro fibers and depicted the microstructure of cellulose fibers.   a) Magnification X 45  b) Magnification X 500 Figure 3.2 - SEM images for micro Synthetic fibers PF1  51     a) Magnification X 45  b) Magnification X 400 Figure 3.3 - SEM images for macro Synthetic fibers PF2     a) Magnification X 45  b) Magnification X 150 Figure 3.4 - SEM images for micro Synthetic fibers PF8   52    a) Magnification X 45  b) Magnification X 150 Figure 3.5 - SEM images for Glass fibers GF1     a) Magnification X 45  b) Magnification X 450 Figure 3.6 - SEM images for Cellulose fibers CF1   53    X 45  X 200 Figure 3.7 - SEM images for Cellulose fibers CF4    Synthetic fibers (in particular for fiber type PF1 and PF2) are made using polypropylene; an addition polymer (monomers are joined by addition reaction without the loss of any atom or molecule) made from the monomer propylene, it is unusually resistant to many chemical solvents, bases and acids. Polypropylene has a melt point of 324°F (162°C), low electrical conductivity, good acid and salt resistance, low thermal conductivity, and minimal water absorption. Much like synthetic fibers, glass fibers used in this study (GF1) also have high resistance to aggressive chemicals, low electrical conductivity, and a softening point of 860°C.  These fibers are made using alkali-resistant glass and each fiber consists of 100 filaments.  Cellulose fibers on the other hand, are the primary structural component of plants. Cellulose chains are stretched out and lined up next to each other so they are straight to 54  form fibers with low diameters.  Fiber surface area of most cellulose fibers is very high (25 000 cm2/g in case of fiber type CF1).  3.4 Supplementary Cementing Materials (SCMs) 3.4.1 Fly Ash  To study the effect of fly-ash, an ASTM Class C fly-ash was used in some mixes. The chemical composition of the fly ash used in this study is presented in the form of the mill certificate received from the manufacturer (Table 3.3). Class C fly ash is produced by plants that burn subbituminous or lignite coal and typically has a CaO content of more than 20%, whereas Class F is produced by plants that burn anthracite or bituminous coal and its CaO content is less than 10%.  The microstructure (Figure 3.8 (a)) and composition of a material ultimately affects its behaviour in concrete.  Fly ash particles are spherical in nature and their sizes range between 10 to 20 micrometers.  Spherical particles help in improving the workability of the mix and since the fly ash particles are fine, they also reduce bleeding in concrete and make the mix denser by particle packing.  55   Figure 3.8 - Morphology (a) Fly ash, and (b) Silica Fume (adopted from Mindess et al., 2003)         56   Table 3.3 - Chemical and Physical Analysis of fly-ash   3.4.2 Silica Fume  Densified silica fume produced by Norchem Concrete Products was used for producing high strength concrete for the substrate bases used during laboratory shrinkage tests. 57  Particles of silica fume are much finer than cement or fly ash particles (morphology shown in Figure 3.8 (b)); the resulting concrete is denser and surface area of binding material is larger making silica fume a much more reactive material. Properties of the silica fume (received from the manufacturer) are given in Table 3.4.   Table 3.4 - Chemical and Physical Properties of Silica Fume (Source: Basalite Concrete Products) Chemical analysis  Physical characteristics  SiO2 93.0 % min. Specific surface area (m2/kg) 18,000 – 22,000 Fe2O3 0.80 % max. Specific gravity 2.2 Al2O3 0.40 % max. Bulk density (kg/m3) 250 – 325 CaO 0.60 % max. Fineness (ave. diam.) (µm) 0.1 – 0.2 MgO 0.60 % max. Percent passing 45 µm (%) 95 – 100 Na2O 0.20 % max. Particle shape Spherical K2O 1.2 % max. Form Amorphous C (Free) 2.0 % max. Canadian standard CSA – A3001-03 SO3 0.40 % max. L.O.I. 3.5 % max.    3.5 Admixtures 3.5.1 Superplasticizer  For preparing the substrate bases for laboratory testing, silica fume was used along with a superplasticizer (brand name Glenium 3000, manufactured by BASF 58  Admixtures Inc formerly known as Master Builder Technologies).  This is a ready-to-use high-range water-reducing admixture based on polycarboxylate chemistry. Glenium 3000 meets ASTM C 494 requirements for Type A, water-reducing, and Type F, high-range water-reducing, admixtures.  3.5.2 Shrinkage Reducing Admixtures  To compare the performance of various fibers with SRAs, a SRA produced by Grace Construction Products (brand name Eclipse) was used at the dosage rate recommended by the manufacturer.  According to the manufacturer, a dosage rate of 7.5 L/m3 (1.5 gal/yd3) results in ultimate shrinkage reductions of up to 50%.  The SRA is a liquid admixture for concrete or any Portland cement based material and contains no expansive material, but instead acts chemically to reduce shrinkage.  This SRA works on the principle of reducing the surface tension of water and pulling in on the walls of the pores, thus reducing shrinkage.  The SRA’s chemical family name is aliphatic propylene glycol ethers and its physical and chemical properties received from the manufacturer are given in Table 3.5.        59   Table 3.5 - Physical and Chemical Properties of SRA Physical State: Liquid Appearance/Odor: Light orange translucent liquid. Ether-like odor Odor Threshold: (ppm) Not Available pH: Not Available Vapor Pressure: (mm Hg) <0.1 mm Hg @ 68°F Vapor Density: (Air = 1) Not Available Solubility In Water: ~15% @ 70°F Specific Gravity: (Water = 1) 0.93-0.94 @ 77°F Evaporation Rate: (Butyl Acetate = 1) ~0.02 Boiling Point: ~414°F @ 760 mm Hg Viscosity: Unknown Bulk Density: (Pounds/Cubic Foot)(Pcf) Not Applicable % Volatiles (gr/L): (70°F) (21°C) 100 %  60  CHAPTER 4 - DEVELOPMENT OF EXPERIMENTAL TECHNIQUE   4.1 Test Summary  A novel test method for characterizing restrained plastic shrinkage cracking in cementitious materials was developed.  This technique is very different from other techniques used so far and simulates real stress conditions that develop in a shrinking overlay or repair material.  The development of this technique is described in this chapter.  4.2 Introduction  Different test methods available to measure shrinkage induced cracking have been described in Chapter 2.  Most of these methods do not produce stresses in the specimen that simulate real stress conditions in the field experienced by repair materials.  Hence, a technique producing realistic shrinkage conditions was developed (Banthia & Campbell 1998; Banthia et al. 1996; Banthia & Yan 2000), where large specimens (roughly 1m long) were used for testing.  In this earlier version of the method, the substrate base providing restraint was prepared by randomly placing rounded aggregates over fresh concrete. A layer of fresh concrete (or shotcrete) is placed directly on a fully hardened substrate.  This ‘old substrate’ has surface protrusions, which enhance its roughness and, in turn, impose a uniform restraint on the shrinking overlay.  The whole assembly is then subjected to a drying environment to induce cracking in the overlay.  61   The purpose of this research was to further understand the test procedure and significantly improve it.  During the course of this thesis, the substrate bases were standardized to improve reproducibility, significant alterations were made to the existing test equipment and additions were made to the test capabilities. Once improvements in the test were satisfactory, the objective of further tests was to study the effectiveness of various types of fibers and admixtures on early age restrained shrinkage cracking.  The test method characterizes the cracking performance of a cement-based material when used as a bonded overlay.  The technique is therefore useful in developing crack free repair material for infrastructure applications.  While this method is best suited to study the effect of fibers, it is also sensitive to change in mix proportion, addition of chemical and mineral admixtures such as fly ash and shrinkage reducing admixtures (SRAs).  It should however be noted that this technique simulates extreme conditions of plastic shrinkage.  Due to the high temperature in the environmental chamber, the material being tested is subjected a much different hydration and setting rates and this needs further investigation.  4.3 Bonded Overlay Test Technique 4.3.1 Substrate Bases 4.3.1.1 Development of Substrate Bases  As described earlier, in this test an overlay is subjected to conditions of restrained plastic shrinkage.  Hence, the effectiveness of the technique relies on development of bond or restraint between the overlay and the substrate. 62  The initial efforts to create a substrate base resulted in the use of a large base (roughly 1.0 m in length) with randomly placed rounded stones on fresh concrete (Banthia and Yan 2000, Banthia et al., 1996). To address handling issues and to reduce the size of the specimen, substrate bases were redesigned with a new size of 325 mm length and a cross-section of 100 x 40 mm.  Various types of substrate bases developed during the course of this project are given in Table 4.1. Some of the issues and concerns with each substrate base are described below in a chronological order.  Table 4.1 - Different Types of Substrate Bases Substrate Type/Material Substrate Designation Surface Protrusion Pattern Embedded Rebars S1 Symmetric N/A Steel S2 Steel Protrusions Staggered N/A C1 Smooth Concrete N/A ; C2 Rounded Stone Finish Random ; C3 : C3-R* Staggered ; C4 : Concrete C4-R* Concrete Protrusions Symmetric ;  * with rebar     63  Aggregate Finish (Type C2):  During the use of these bases (shown in Figure 4.1 (b)), two issues were identified.  Firstly, casting of the bases and hand placement of the aggregates in fresh concrete was a very cumbersome process.  Secondly, the randomness associated in placing the aggregates resulted in higher variability in test results.  Steel Substrate Base (Type S1 and S2):  Later, to address the above issues arising from the use of the substrate bases with aggregate finish, two types of steel bases with staggered and symmetric surface protuberances were designed (Figure 4.2 (a, b)).  This type of substrate base was however, also phased out soon after preliminary testing due to concerns with thermal incompatibility with the overlay material and due to concerns with lack of bond development. Moreover, recovering the steel base after testing was also a cumbersome process.  64   a) b) c) d)  Figure 4.1 - Substrate Types: a) Smooth Finish (Type C1), b) Aggregate Finish (Type C2), c) Staggered Protrusions (Type C3), and d) Symmetric Protrusions (Type C4)  (a) Symmetric Protrusions (Type S1) (b) Staggered Protrusions (Type S2) Figure 4.2 - Steel Substrate Base    65  Concrete Substrate Bases:  Since the premise of this test was to simulate real stress conditions that develop in the field, a substrate base representing the characteristics of the substrate in a new or repair project was developed.  This resulted in the use of high strength concrete for the substrate base with either staggered protrusions (Figure 4.1 (c)) or with symmetric protrusions (Figure 4.1 (d)) as in the case of the steel bases described above. To ascertain that the protrusions were providing the restraint, performance of substrates type C3 and C4 were compared to that for a smooth concrete substrate type C1 (Figure 4.1 (a)).  This comparison is made in section 4.3.5.  The protrusions on the concrete bases were semicircular in shape with a diameter of 18.5 mm and clear dimensions of all substrate bases (not including the surface protrusions) were: 40 mm × 100 mm × 325 mm (Figure 4.3).  Symmetric bases had 27 surface protrusions and the staggered base had 60% more protrusions representing higher roughness and higher restraint.  The actual protruded area of the symmetric and staggered bases was 7,258 mm2 and 11,559 mm2 respectively.  These bases were deliberately made shorter than the length of the overlay in order to avoid curling-up at the ends (Figure 4.4). This way, the overlay would be able to ‘wrap’ over the base and thus restrain upward curling.  A high strength concrete mix that developed compressive strength at 28 day (f’c) of approximately 85 MPa when tested in accordance to ASTM C 39 was used for the bases.  Concrete mix proportions are given in Table 4.2.   66    Figure 4.3 - Dimensions: Substrate Base with Staggered Protuberances (Type C3)   Substrate Base Overlay  Figure 4.4 (a)– Overlay Wrapping around the Substrate (after demolding)  67   Figure 4.4 (b) – Dimensions of the Overlay Wrapping around the Substrate (in mm)     Except for substrate type C2, the two other types of concrete substrate were evaluated with two sub types: first with reinforcing bars (substrate type C3-R and C4-R) and second without reinforcing bars (substrate type C3 and C4) to simulate different substrate stiffness.  All type C2 substrates with the aggregate finish also contained steel reinforcing bars.  Steel rebars (2 nos. 10 mm ∅) provided along the length of the substrate were expected to reduce the chances of breakage during handling and enhance the linear stiffness.  Chemical and mechanical properties of the steel rebars are given in Table 4.3. The steel rebars met the CSA G30.18-M92 400W specifications.   375 325 40 60 68   Table 4.2 - Mix Proportions for evaluation of Substrate Bases Cement Silica Fume Water Sand Aggregate Superplasticizer kg/m3 Substrate Base 535.5 59.5 166.6 809.2 809.2 1.61 Overlay Mortar for evaluation of substrate bases 1200 - 576 601 - -    Table 4.3 - Properties of Deformed Steel Rebars (Reproduced from the Mill certificate, courtesy: Harris Rebar, Delta, B.C.) Chemical Composition (%) Mechanical Properties C S P Si Mn Cr Ni Cu V Nb Mo Ceq Yield Point (MPa) Ultimate Tensile Strength (MPa) Elongation (%) Cold Bend 0.20 0.029 0.016 0.41 1.06 0.06 0.04 0.14 0.017 0.032 0.005 0.39 490 620 19.0 OK 69    4.3.2. Environmental Chamber 4.3.2.1 Development of Environmental Chamber  Initially, an environmental chamber measuring 1740 mm × 350 mm × 380 mm (Figure 4.5) was used.  The chamber was made in clear plastic so that the sides and top surface of three replicate test specimens could be observed.  Temperature and humidity inside the chamber was controlled using a heater with a fan at one end of the environmental chamber and test specimens were placed along the length of the environmental chamber.  After some initial testing, it was observed that due to a temperature and humidity gradient along the length of the chamber, specimens placed closer to the heater/blower cracked more compared to the one near the end, thus resulting in a high in-batch test variability.   Figure 4.5 - Environmental Chamber   70  4.3.2.2 Modified Environmental Chamber  A new environmental chamber measuring 1290 mm x 1390 mm x 280 mm was constructed.  A detailed dimensioned sketch and a photo are given in Figure 4.6.  As before, the chamber was made in clear plastic so that the sides and top surface of all three test specimens could be observed.  Since the environmental chamber now had three enclosures allowing heated air to escape the chamber through three 240 mm x 175 mm openings, the test specimens could be placed at the same distance from the blowers with more uniform shrinkage conditions.  The chamber contained three blowers at one end capable of circulating air to the other end of the chamber at a rate of about 0.160 m3/s, which would produce a velocity of approximately 13.5 km/hr.  The blowers consisted of in-built heaters and fans with the following specifications: Heaters:  Model# KHSS 464, 4800 Watts, and 240 Volts. Fans: Model- Dayton- Direct Drive Fan Motor, HP 1/30, RPM 1550, and Volts 115.   This new environmental chamber was used for conducting tests described in this thesis. The chamber is equipped with digitally adjustable humidity and temperature sensors/controllers (Brand: Omega, model: RHCN1) capable of recording humidity to ±1% and maintaining the temperature to ±1oC.  These controllers regulate the power supply to the heaters (with fans) as necessary, in order to maintain a constant temperature and humidity in the chamber.  In a test, typically, a temperature of 50°C is chosen, which results in a relative humidity less than 5% and this produces an evaporation rate of approximately 1.0 kg/m2/h from the specimen surface.  Details about the actual measured evaporation rate from specimens and the effect of mixture proportion on evaporation is provided in Chapter 5. 71  Figure 4.7 demonstrates the temperature and relative humidity (RH) recorded inside the environmental chamber during a typical test.  Notice the drop in the temperature and increase in the relative humidity that takes place during the demolding process that takes place 2h after the material is placed in the environmental chamber.    Figure 4.6 - Environmental Chamber a) 3D View  Data Acquisition System Blowers 1 2 3 Temp/RH Probe 72   Figure 4.6 - Environmental Chamber b) Dimensioned Sketch  73  0 10 20 30 40 50 0 5 10 15 20 Time (hrs) Te m pe ra tu re  (˚ C ) a nd  R H (% ) Temperature RH D em ol di ng  af te r 2 hr s  Figure 4.7 - Typical plot- Temperature and RH vs. Time  4.3.3 Placement of Overlay  A 60 mm deep layer of the overlay mixture to be investigated is placed on the substrate, which is in moist surface dry condition.  The substrate is moist cured for a minimum of 28 days.  Figure 4.8 shows a fully cured substrate (type S3-R) ready to receive the overlay mix in a mold.  The entire operation of placement of the overlay layer and finishing is completed in less than 10 minutes.  The entire assembly is then transferred to the environmental chamber, which is started ahead of time to allow the chamber to reach a steady state of temperature and humidity.  After two hours, the chamber is opened and the sides of the mold are removed (Figure 4.4) to expose the specimen to a uniform state of drying.  The 74  chamber is then closed and temperature is maintained and relative humidity monitored for the next twenty-two hours.   Figure 4.8 - Substrate Ready to Receive Overlay  4.3.4 Crack Measurement  After twenty-two hours, cracks developed in the overlay (Figure 4.9) are characterized using a 60X microscope (Figure 4.10) or by using image analysis, which is described in section 4.5.3.  Crack width is measured at 10 or more locations for every 100 mm of crack length.  The microscope used in this study had an accuracy of ± 0.02 mm and was used to characterize cracking in unreinforced overlays (Chapter 5) and in fiber-reinforced overlays (described in Chapter 6)  75   Figure 4.9 - Specimen After Cracking   Figure 4.10 - Crack Measurement Using a Microscope  Crack Calculations:  As described earlier, this technique can be very effectively used to evaluate the performance of fibers.  Hence, for illustrative purposes, crack calculations are described in the context of fiber reinforced concrete. In order to evaluate a particular fiber type, three specimens of plain concrete and three specimens of fiber-reinforced concrete are tested.  For each observed surface crack, width and the length are measured and used to calculate crack 76  areas.  The average crack width is calculated by taking an average of the crack widths that have been measured at several locations.  The Total Crack Area (Atotal ) is obtained by summing over all cracks in  a specimen. Atotal  = ∑= = ni i iilw 1                                                                             Equation (4.1) where,  wi = average crack width of the ith crack  li = length of the ith crack  n = number of cracks observed in a test  For each specimen, the Maximum Crack Width (wmax) is also recorded. The Crack Width Control Efficiency (η width) and the Crack Area Control Efficiency (η area) can then be calculated as: η width  = 100 )( max, max,max, x w ww plain frcplain −                                    Equation (4.2) η area  = 100 )( , ,, x A AA plaintotal frctotalplaintotal −                                Equation (4.3) where, wmax,plain = average of maximum crack widths observed in three replicate plain  concrete specimens wmax,frc = average of maximum crack widths observed in three replicate fiber-  reinforced concrete specimens 77  Atotal,plain = average of total crack areas observed in three replicate plain concretes  specimens Atotal,frc = average of total crack areas observed in three replicate fiber-reinforced  concrete specimens  Time to First Crack:  Since a see through plexiglass environmental chamber is used for testing, time to first crack can be recorded and related to the performance of different admixtures/additives in the overlay mix.  Typically, the first crack would appear after demolding the specimens at two hrs from the time of casting.   4.3.5 Effect of Different Concrete Substrates on Crack Characteristics  4.3.5.1 Preliminary Observations Preliminary tests were conducted on the various substrate bases described in Table 4.1.  Tests clearly indicated that there was no effect of including a rebar on cracking induced in the overlay.  However, since rebars were deemed necessary to prevent any breakage of substrate bases during handling tests were conducted only on bases with rebars.  Test results from the preliminary investigation (not presented here) were analyzed. Total crack area, crack width, and number of cracks were considered to determine the substrate base that produced the most consistent and repeatable cracking pattern in the overlays.  The most suitable substrate bases were identified as type C2 (aggregate finish) and 78  C3-R (concrete with staggered protrusions).  To study the effect of substrate roughness, material, and stiffness on the restraint imposed on the shrinking overlay and eventually on crack characteristics, confirmatory tests were conducted on these two types of substrate bases. Tests were also conducted on a smooth substrate (type C1) to test the material under the influence of minimal restraint on the shrinking overlay.  Results of these tests are presented in the next section.   4.3.5.2 Results  Substrate type C2 and C3-R were identified for further testing and to determine if the new concrete bases would reduce the variability in test results.  Tests were conducted on these substrates along with a smooth substrate base (Type C1) to determine the most effective concrete substrate and its effect on crack characteristics.  The average test results for three specimens are presented in Table 4.4 with standard deviation indicating in-batch variation provided in parenthesis.  For the smooth substrates, a very small crack was observed in one specimen (total crack area 12 mm2).  This crack was treated as a test artifact and was possibly caused due to improper demolding.  From the test results it was also clear that the surface protrusions were clearly responsible for providing the restraint and causing shrinkage cracking in the overlays.  Substrate type C2 (with aggregate finish) due to the presence of aggregate resulted in maximum crack area indicating development of high restraint. However, as mentioned before, casting of these substrate bases was very time consuming and subject to inconsistent surface roughness, resulting in poor in- 79  batch repeatability.  As seen in Table 4.4, the standard deviation for this type of substrate base was high when compared to type C3-R; with a 42% in-batch test variability for crack area.     Table 4.4 - Effect of Substrate Types on Crack Characteristics Substrate types Substrate Designation Average Total Crack Area (mm2) Average Crack width (mm) Average Maximum Crack Width (mm) Average Number of Cracks Smooth Finish C1 12* 0.3* 0.56* 0.7* Aggregate Finish C2 381 (42%) 0.885 (23%) 1.8 (34%) 4.7 (12%) Staggered Finish (W Rebar) C3-R 339 (24%) 1.24 (7%) 2.27 (18%) 2.7 (22%) * Test artifact: single crack created during improper demolding in one specimen     80   Figure 4.11- Finalized Substrate Type S3-R   The substrate type C3-R resulted in reasonable amount of cracking for the control mix but more importantly resulted in the lowest standard deviation (for crack area and widths) when compared to the other types of substrates.  This type of substrate also resulted in an average of 2.7 cracks in the overlays as compared to 4.7 cracks in the substrate with aggregate finish.  Even though the type C3-R substrate resulted in fewer cracks, the crack were better developed when compared to the more random cracking in type C2 substrate.  Hence, this substrate with staggered protrusions (Figure 4.11) was selected as the substrate base for all future testing.   4.4 Further Developments in the Test Method 4.4.1 Sensors  The test set-up described in the previous section was further developed and sensors were added to the set-up to gather additional test data. 81  Internal Temperature Measurements: The set-up was connected to a thermocouple (J type) to record internal temperature changes in the overlays, and ambient temperature and relative humidity inside the environmental chamber.  One end of the thermocouple was embedded in the test specimen and the other end was connected to a custom-designed signal conditioning system (Figure 4.12).     Power Supply- Thermocouple Thermocouple Signal Conditioning System Connections- Strain Gauges and Weighing Scale  Figure 4.12 - Sensors and Data Acquisition System  Evaporation Rate: The true moisture loss measured from the specimens can be very different from that measured from a water-filled container placed inside the environmental chamber. To measure moisture loss from within the specimens, the set-up was modified such that one 82  of the overlay specimens could be mounted atop a weighing scale to constantly monitor the actual weight change in the specimen (Figure 4.13).  The weighing scale had a least count of 0.1 gms and was capable of being connected to the data acquisition system for continuous and precise measurements.  The data acquisition system recorded room temperature, room humidity, temperature and humidity inside the chamber, internal concrete temperature, and moisture loss at a frequency of 0.20 hz.  (a) Location of the Specimen in the Environmental Chamber (b) Placement of the Specimen Atop a Weighing Scale Figure 4.13 - Arrangement for Measuring Moisture Loss   Other Sensors: In addition to the above-mentioned modifications, several other devices and sensors were added as described below to the environmental chamber to enhance the capabilities of the test equipment.  A new data acquisition with more data channels was designed to record data from these sensors/instruments (Figure 4.12).  The following 83  instruments/sensors were added to the environmental chamber, and selectively used during testing: • An additional probe to record room temperature and humidity, • An anemometer to verify air velocity in each enclosure, • A non-contact laser sensor to measure settlement of overlays (Figure 4.14), and • A plexiglass camera frame for acquiring non-contact digital images of cracking on the specimens. The procedure adopted to analyze digital images is given in section 4.5.3.   Figure 4.14 - Non-Contact Laser Sensor for Measuring Settlement  4.4.2 Strain Development in the Overlay  Measuring strain across the cross section of freshly cast material helps understand early-age characteristics of mortar.  Measuring strain in fresh concrete/mortar necessitates use 84  of non-contact or embedded strain sensors.  However, developing adequate bond between strain gauges and fresh concrete and gauge calibration at changing temperature conditions still remains a challenge.    Figure 4.15 - Strain Gauges Mounted on Substrate Surface    Figure 4.16 - Chairs to Place Strain Gauges in the Overlays  85    Strains developed within the substrate were measured by installing strain gauges on the substrate surface in the longitudinal and transverse directions (Figure 4.15).  Chairs were designed to locate another strain gauge in embedded condition within the overlay (Figure 4.16).  The sensors were located at the center of each overlay 10 mm below the top surface. Data from two strain sensors embedded in two specimens are plotted in Figure 4.17 (a). Notice that there is an increase in the strain initially followed by a decrease.  Increase in the tensile strain during the first few hours can be attributed to the thermal expansion of the material (material is initially at room temperature) due to heating in the chamber as well as due to the initial heat of hydration.  This agrees with findings of Yang et al (2004) shown in Figure 4.17 (b), where an increase in strains was observed due to volume change due to heat of hydration followed by a rapid drop in strains due to sudden cooling and rapid drying.  From the strain gauges installed on the substrate base, little or no change in strain was observed in the substrate itself during a plastic shrinkage test.  This further implied that the substrate was stable with respect to its environment.  86          -1500 -1000 -500 0 500 1000 1500 0 2 4 6 8 10 12 14 16 18 20 Time (hrs) M ic ro  s tr ai n (1 0 m m  b el ow  s ur fa ce ) Average Specimen 2 Specimen 1 Demoulding  Figure 4.17 (a) - Strain Measurements in the Overlay Using Embedded Strain Gauges  87   Figure 4.17 (b) - Strain Measurements in Restrained Specimens (from Yang, et al (2004))    Demolding, which occurred two hours after casting was followed by a rapid drop in the measured strains.  This may be attributed to two factors: first, as the environmental chamber is opened during demolding, cold air from outside enters the chamber and lowers the chamber temperature; and second, demolding causes an increase in the exposed surface area and a rapid increase in the drying rate leading to the development of compressive strains that counteract the hitherto created tensile strains due to expansion.  After these initial fluctuations, there is gradual build-up of compressive strains in the overlay.    88  4.4.3 Non-Contact Crack Measurement Using Image Analysis  A non-contact method to determine the growth of cracks during the first 24 hours (when the environmental chamber cannot be disturbed) was developed.  High resolution images of specimens were taken at suitable intervals using a camera fixed at a constant distance from the specimen.  Images are processed using commercially available image analysis software.  This technique was only used to determine crack area and not for calculating crack widths.  The general principle to measure the crack area involves counting the number of pixels of the entire surface of the beam and relating it to the known area of the specimen and hence determining a “calibration factor.”  Later the crack area is “masked” or isolated from the non-crack area by comparing the difference in color of the pixels within a crack (darker) and that on the rest of the specimen (lighter).  The number of pixels within the masked area is related to the crack area by using the predetermined calibration factor.  The following procedure was adopted for image analysis: • Open the digital image using the software (Corel Photo Paint) and select a mask around each crack, • Adjust Brightness/Contrast/Intensity for clarity, • Sometimes a colored image can be transformed into gray scale for better clarity, • Adjust Brightness/Contrast/Intensity again if required, • Use the mask function and set the tolerance that defines the color range.  Several iterations might be required before a fixed tolerance could be used for masking. Instead of using a fixed tolerance, sample color (also called “seed color”) of the pixels 89  within a crack could also be picked (function typically called ‘pick color’) to mask the crack, • After the crack is satisfactorily masked, the number of pixels can be calculated using the histogram function (see Figure 4.18), • The number of pixels can then be converted to crack area using the predetermined calibration factor, and • Repeat this procedure for images successively acquired to determine the rate of crack area increase.   Figure 4.18- Image Analysis: Crack area Masking  90    To compare the results obtained from manual measurements using a microscope and that obtained from image analysis, a study was conducted on eight different types of mixes. The w/c and s/c in the mixes ranged from 0.35 to 0.5 and 0.4 to 0.6 respectively.  Detailed test results for these mixes are presented in the next chapter (Table 5.2).  Crack areas calculated from microscope measurements after the test had ended (22 hrs from demolding) and from analyzing images taken at the same time are compared in Table 4.5 and plotted in Figure 4.19. With the exception of Mix 1 where there was minimal cracking, it is evident that crack areas obtained from image analysis were higher (a maximum % increase of 44 for mix M5) for some mixes and lower (a maximum % reduction of 42 for Mix M6) for others when compared to the crack areas calculated form manual measurements.  However, from Figure 4.19, it is evident that results from both techniques followed the same trend and were in good agreement. It was evident that mixes M7, M8, and M10, which had higher w/c ratio had more cracking when compared to the other mixes that had comparatively lower w/c ratio.       91   Table 4.5- Comparison of Results from Measurements using a Microscope and Image Analysis Mix Average area using Microscope (mm2) Average area using Image Analysis (mm2) % Difference (Image Analysis to Microscope) M1 22.1 35.8 62 M2 135.5 157.8 16 M4 285.4 252.6 -11 M5 128.5 184.5 44 M6 259.9 150.7 -42 M7 627.9 399.1 -36 M8 554.7 469.2 -15 M10 498.6 390.9 -22    92     Figure 4.19- Variation of Crack Area Based on Measurement Technique   4.5 Concluding Remarks The development of a novel technique that can be used to evaluate the performance of cement-based overlays under restrained shrinkage conditions was described in this chapter and the following conclusions were made: 93  • During the preliminary investigation various types of substrate bases were studied and their influence on repeatable cracking in overlays was determined.  The effect of substrate stiffness, roughness, and type of material was studied to select an appropriate substrate. • Substrate base with staggered protrusions was selected based on the restraint developed by this type of substrate on the overlay and because of the repeatable cracking observed with these overlays.  During the preliminary investigation, an in- batch standard deviation of 24% and 7% was observed in total crack area and average crack width.  These were deemed acceptable for a test procedure, which was developed to study the performance of unreinforced and fiber-reinforced overlay during early age. • A fully instrumented environmental chamber was constructed where the temperature was controlled and the relative humidity was monitored.  The environmental chamber was equipped with a data acquisition to record the temperature and relative humidity in the room and inside the chamber.  The environmental chamber was equipped to precisely record the moisture loss from the specimens, record strain from up to six embedded strain gauges, measure the internal temperature from the specimens, measure wind velocity using an anemometer, and a non-contact laser sensor to measure settlement. • Factors for measuring the efficiency of fibers in terms of total crack area and maximum crack width were defined.  To measure the crack area and width, a non- contact image analysis technique was also developed in addition to measuring crack 94  characteristics using the visual method (microscope).  A 40% difference in results was noted between the image analysis technique and the visual method.  However, it was concluded that the image analysis technique could be instrumental in monitoring the crack propagation during early ages as opposed to only measuring cracking after the test had ended at 24 hrs. 95  CHAPTER 5 - EARLY-AGE BEHAVIOR OF CEMENTITIOUS COMPOSITES  5.1 Summary  This chapter describes the influence of mixture proportions on plastic shrinkage cracking in cementitious repairs and overlays.  The following variables were studied: water- cement ratio (w/c), sand-cement ratio (s/c) aggregate-cement ratio (a/c), fly-ash content, and the presence of a shrinkage reducing admixture (SRA).  The bonded overlay technique described in Chapter 4 was used to characterize the performance of twelve different mixes with water-cement ratio (w/c) ranging between 0.35 and 0.6 and sand-cement ratio (s/c) ranging between 0.4 and 0.6.  In addition to measuring crack width, crack area, and time to first crack for different mixes, moisture loss from the specimens, heat evolution, and the rate of crack growth were also determined.  5.2 Introduction  The effect of loss of moisture from cement-based materials and its effect on the in- service performance have already been discussed in the previous chapters.  The loss of moisture from freshly placed mortars and concrete depends on several factors including external environmental conditions that are generally difficult to control and factors such as mixture proportions that are comparatively easier to control.  In this chapter, effect of mix proportion on evaporation rate, heat evolution, and crack characteristics such as crack time, crack widths, and crack areas has been investigated using the bonded overlay technique described in the previous chapter.  This chapter describes the effect of increasing a/c (coarse aggregate-cement ratio), addition of Class C fly ash, and a 96  shrinkage reducing admixture. Depending on availability, fly-ash is currently being utilized to replace parts of cement to make concrete more economical, reduce green house gas emissions created during the production of cement, and to improve fresh properties and durability of concrete.  Hence, the aim of this study was also to assess the effect of fly-ash on restrained plastic shrinkage cracking of cement mortars.  It should be noted that the mixes used in this study were selected based on their high potential for shrinkage.  These mixes had a high paste content, which is expected to shrink the most and a low coarse aggregate content, which provides the internal restraint and reduce shrinkage.  The intention was to use these mixes to create a high potential for cracking.  These mixes could then be used as control mixes and their results could be compared to that of fiber- reinforced mixes to determine efficiency of fibers in reducing the cracking potential under identical conditions.  The comparison between unreinforced and fiber-reinforced mixes based on their crack characteristics is made in Chapter 6.  Typical mixture proportions used in practice can be very different from that described in this chapter, since they are much leaner and have significant restraint due to presence of coarse aggregates.  Hence, to further understand the performance of unreinforced mixes with more realistic mixture proportions and environmental conditions, field studies were conducted and are described in Chapter 7.  5.3 Test Set-Up and Variables Investigated  The test technique described in Chapter 4 was used to study the influence of mixture proportion on shrinkage induced cracking.  Type 10 Portland cement, river sand and water at room temperature were used for casting the overlays. Mix proportions for substrate bases are 97  shown in Table 5.1.  Two series of tests were performed: Series-M where the water-cement ratio (w/c), sand-cement ratio (s/c), and coarse aggregate-cement ratio (a/c) were varied, and Series-F where fly ash content and presence of shrinkage reducing admixture (SRA) were investigated.  Three specimens were tested for each variable.  Table 5.1- Mix Proportion of Substrate Base Ingredient Cement Silica Fume Water Sand Coarse Aggregate* Superplasticizer Mass by volume (kg/m3) 535.5 59.5 166.6 809.2 809.2 1.61 * 10 mm maximum size   Series-M (Table 5.2) consisted of 12 overlay mixes with varying combinations of w/c and s/c.  The cement content for the control mix was 1200 kg/m3, with a w/c and s/c equal to 0.5, resulted in a sand and water content of 600 kg/m3.  The water and sand content was corrected for the moisture content in sand.  Mix M8 was further modified to study the influence of coarse aggregates on crack characteristics.  To achieve this, two additional mixes M8-1 and M8-2 were tested where the total a/c was 0.6 and 0.7 respectively.  Both mixes M8- 1 and M8-2 had the same s/c ratio of 0.5.  Series-F consisted of mixtures with various Class C fly-ash dosages and one mix with SRA, as indicated in Table 5.3.  Fly-ash met the ASTM requirements for Class C fly-ash. Since fly-ash was expected to improve the workability, mix M5 was used as the control mix for Series F.  Mix M5 had a lower w/c = 0.4 and s/c = 0.5.  Mixes F2 to F5 had a water- cementitious (w/cm) ratio of 0.4 and s/c = 0.5.  Mix F6 had the same mixture proportion as M5 with the addition of the SRA.  98  Table 5.2- Overlays with varying w/c and s/c ratios (Series M) w/c Mix Designation s/c a/c M1 0.4 M2 0.5 0.35 M3 0.6 M4 0.4 M5 0.5 0.40 M6 0.6 M7 0.4 M8 0.5 - M8-1 0.5 0.6 M8-2 0.5 0.7 0.50 M9 0.6 M10 0.4 M11 0.5 0.60 M12 0.6 -   Table 5.3 - Overlays with varying fly-ash and SRA dosages (Series F) Mix Designation Admixture Dosage F1 (Control) s/c = 0.5 w/c = 0.4 - 0% F2 5% F3 10% F4 15% F5 Fly-Ash* 20% F6 SRA 7 liters/m3 * (% cement replacement)   5.4 Crack Measurements  Time required for first crack to develop was recorded in all cases. Crack areas and widths after 24 hours were measured using a microscope as described in Chapter 4. In some 99  cases, crack evolution was also monitored using image analysis (technique described in Chapter 4) to determine growth of cracks during the first 24 h. Toward this end, images were taken at 3, 4, 5, 6, and 24 hrs after casting so that the progression of cracks could be observed and quantified.   5.5 Results  During this study, the average laboratory temperature and RH recorded over a period of one month was 23°C and 49% respectively. As stated earlier, temperature and humidity inside the chamber were 50 ± 2°C and 5 ± 1% respectively.  5.5.1 Specimen Moisture Loss  Moisture loss recorded for series M is presented in Figure 5.1.  Note: Data for Mix M5 could not be recorded both prior to and after demolding, and for Mix M12 data only after demolding could be recorded.  Average calculated evaporation rates before and after demolding are presented in Table 5.4 where the post-demolding evaporation rates were averaged over a period of 16 h after demolding.  Before demolding, evaporation rate for mixes with w/c of 0.35 and 0.40 ranged between 0.89 and 0.99 kg/m2/h and for mixes with w/c of 0.50 and 0.60 ranged between 1.10 and 1.37 kg/m2/h. Mix M11 was an outlier with the highest total moisture loss of 155 g before demolding. Notice a steep increase in the rate of moisture loss at demolding and large evaporation rates for up to 6 h after casting. Mixes with low w/c attained their stable mass earlier than those with a higher w/c.  During this time, evaporation rates ranged between 0.30 and 0.48 kg/m2/h. Wang et al (2001) measured the 100  evaporation rates for mixes during early ages and reported an increase in the percentage mass loss in the first 90 minutes from 14% to 20% when the w/c was increased from 0.35 to 0.55. In this study, on the other hand, for an increase in w/c from 0.35 to 0.6, the percentage mass loss only increased from 4% to 8%. This may be related to the different specimen surface to volume ratios used and the different environmental conditions imposed.  For the range of s/c used in this study, no noticeable trend in the influence of s/c on the moisture loss.  5.5.2 Heat Evolution  Heat evolution was recorded in terms of temperature rise for series M. For clarity sake, results for mixes with w/c of 0.35 and 0.5 are compared in Figure 5.2 and those for mixes with w/c of 0.4 and 0.6 are compared in Figure 5.3.  Notice an initial drop in temperature during casting, followed by a steep rise in temperature for the next 3-4 hours due to a combination of high temperature inside the chamber and the heat of hydration.  Most mixes attained peak temperatures between 4-5 hours from the time of casting.  Considering Mixes M5 and M10 as outliers, it was evident that mixes with a lower w/c reached higher overall temperatures.  Once again, the effect of s/c on heat evolution was not clear. Approximately 10 hours after casting, specimen temperature dropped and attained the temperature in the chamber (50°C).   101  -900 -800 -700 -600 -500 -400 -300 -200 -100 0 0 2 4 6 8 10 12 14 16 18 20 Time (hrs) M o i s t u r e  L o s s  ( g ) M3 M2 M1 M6 M8 M4 M7 M9M11 M12M10  Figure 5.1 - Effect of Mix Proportion on Moisture Loss  Table 5.4 - Evaporation Rate Before and After Demolding Mix Designation M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Before Demolding 0.99 0.97 0.97 0.89 - 0.92 1.10 1.21 1.37 1.25 2.07 - Moisture Loss, kg/m2/h  After Demolding 0.31 0.34 0.30 0.39 - 0.36 0.45 0.40 0.43 0.48 0.48 0.45  102  15 25 35 45 55 65 0 5 10 15 20 Time (hrs) Te m p (° C ) M2 M3 M1 M9 M8 M7  Figure 5.2 - Heat Evolution in 0.35 and 0.5 w/c Mixes   15 25 35 45 55 65 0 5 10 15 20 Time (hrs) Te m p (° C ) M5 M4 M6 M10 M11 M12  Figure 5.3 - Heat Evolution in 0.40 and 0.60 w/c Mixes  103  5.5.3 Crack Analysis  Rate of crack growth during the first 24 hours was evaluated using image analysis. Randomness associated with masking the cracks resulted in the inability to determine crack areas precisely.  However, it was evident that for most mixes, rate of crack growth was significant in the first 4 hours (rate was higher for mixes with higher w/c) and minimal thereafter.  Some typical plots are shown in Figure 5.4. Overall summary of crack analysis including crack area, average width, maximum crack width, number of cracks and time to first crack for all mixes are presented in Table 5.5.  These data are further plotted in Figures 5.5 (a), (b), (c), and (d). In Figure 5.5 (a) data from Mixes M1 to M12 are given. In Figure 5.5 (b) results for Mixes M8, M8-1 and M8-2 are given to illustrate the influence of coarse aggregate.  In Figure 5.5 (c), data from Series-F are plotted and finally to study the effect of increase in s/c for a constant w/c (0.5 in this case), a representative graph is plotted in Figure 5.5 (d).  Typical crack patterns are shown in Figure 5.6 (a).   104  0 100 200 300 400 500 600 0 5 10 15 20 25 Time (Hrs) C ra ck  A re a (m m 2 ) M2 (w/c=0.35) M5 (w/c=0.4) M8 (w/c=0.5)  Figure 5.4 - Crack Evolution using Image Analysis            105   Table 5.5 - Crack Analysis (Microscope Method) Mix Designation w/c Admixture Admixture Dosage Average Time to Crack (min) Average Crack Area (mm2) Average Crack Width (mm) Average Maximum Crack Width (mm) Average number of cracks M1 128 22 0.43 0.70 0.5 M2 118 136 0.98 1.60 1.5 M3 0.35 120 14 0.36 0.90 1.3 M4 120 285 1.22 2.50 2.5 M5 134 129 0.87 1.70 1.5 M6 0.4 131 260 1.25 2.60 2.3 M7 142 628 1.62 4.00 4.3 M8 148 534 1.82 4.10 3.3 M9 0.5 138 555 1.62 3.80 3.5 M10 139 499 1.07 2.30 4.8 M11 136 795 1.79 4.80 4.5 M12 0.6    _    _ 139 732 1.75 5.20 5 F1 (Control for F series) _ 0% 170 184 0.91 1.61 1.5 F2 5% 173 207 1.70 2.05 1.3 F3 10% 156 210 1.68 2.05 1.3 F4 15% 163 259 1.86 2.22 1.5 F5  Fly-ash* 20% 185 169 1.22 1.77 1.5 F6 0.4 SRA 7 lit/m3 143 175 0.49 1.87 1.8   * % cement replacement   106   Notice in Figure 5.5 (a) that w/c significantly influenced the crack areas and crack widths, and this is in agreement with the findings of Wang et al (2001).  For example, for s/c = 0.5, crack area increased from 135 mm2 to 794 mm2 with an increase of w/c from 0.35 to 0.6 and crack width increased from 0.98 to 1.79 mm. With the exception of Mix M8, crack widths for all other mixes were directly proportional to the observed crack area.  It should however be noted that increase in cracking was observed for the w/c ratio range studied, since the following relationship was met: 1≥ teBleedingRa nRateEvaporatio   This trend could reverse for different environmental conditions and for a different w/c range.  Influence of s/c on crack area and widths was not consistent and no correlation could be established. This is evident in Figure 5.5 (d), where initially the crack area and the crack widths drop when s/c increases to 0.5 but increase when the s/c increases to 0.6.  To further assess the influence of w/c, results for mixes with the same w/c were averaged and are presented in Table 5.6.  Temperature vs. time curves (Figures 5.2 and 5.3) were integrated for a period of 10 hrs from the time of casting.  This aggregate value of heat generated, which was a function of high temperature inside the environmental chamber and the heat of hydration was termed “thermal output.”  This value was computed in oC-h units for various w/c mixes (Table 5.6).  Notice a decrease in the thermal output with an increase in the w/c. Both crack area and average crack width increased with an increase in the evaporation rate. This is contrary to the findings of Shaeles and Hover (1988) who did not find any correlation between evaporation rate and 107  severity of cracking for mixes with w/c ratio ranging between 0.5 and 0.7 and s/c ratio ranging between 0.3 and 0.45.  The effect of adding coarse aggregates on crack area, average and maximum crack widths is shown in Figure 5.5 (b).  The crack area reduced to 30 mm2 and the average crack width reduced to 0.29 mm for Mix M8-2 where the a/c ratio was 0.7.  The crack patterns are shown in Figure 5.6 (b).  This clearly indicated that addition of a very small volume of coarse aggregates can provide significant internal restraint reducing the shrinkage of the cement phase of concrete.  Similar observations were made by Hobbs D. W. (1974), who also found that a change in aggregate volume fraction significantly affected shrinkage strains.  Data from Series-F in which the influence of fly-ash and SRA was investigated are plotted in Figure 5.5 (c). Average number of cracks recorded in the F series was between 1 and 2.  Keep in mind the expected in-batch standard deviation from the test technique, the following conclusions were made.  A slight increased crack area and width was observed with an increase in the percentage of fly-ash replacement; however, the trend reversed for a fly-ash content of 20%. Mix F5 (20% replacement) had lower crack area than both the control (F1) and SRA mix (F6).  There thus appears to be a threshold value of fly-ash content beyond which fly-ash is effective in controlling shrinkage induced cracking. Below this threshold, one can expect an increase in the amount of shrinkage induced early-age cracking. This needs further investigation. Increase in cracking for mixes containing fly-ash may be attributed to lower initial matrix strength and greater amount of water available for evaporation. 108   Mix F6 containing SRA, resulted in lower average crack width when compared to the control mix, however, very similar total crack area and maximum crack widths were recorded.    0 1 2 3 4 5 6 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Mix Designation C ra ck  W id th  (m m ) 0 100 200 300 400 500 600 700 800 C ra ck  A re a (m m 2 ) Average Maximum Width Average Width Crack Area Trendline (Crack area) w/c→ 0.35 0.500.40 0.60  Figure 5.5 (a) - Results of Series-M Tests  109  0 1 2 3 4 5 6 M8 M8-1 M8-2 Mix Designation C ra ck  W id th  (m m ) 0 100 200 300 400 500 600 700 800 C ra ck  A re a (m m 2 ) Average Maximum Width Average Width Crack Area Figure 5.5 (b) - Effect of Coarse Aggregates (Mixes M8, M8-1 and M8-2)  0.0 1.0 2.0 3.0 4.0 5.0 6.0 F1 (0%) F2 (5%) F3 (10%) F4 (15%) F5 (20%) F6 (SRA) Mix Designation C ra ck  W id th  (m m ) 0 100 200 300 400 500 600 700 800 C ra ck  A re a (m m 2 ) Average Maximum Width Average Width Crack Area Figure 5.5 (c) - Results of Series-F Tests  110   0 1 1 2 2 3 3 M4 M5 M6 Mix Designation (w/c = 0.5) C ra ck  W id th  (m m ) 0 100 200 300 400 500 600 700 800 C ra ck  A re a (m m 2 ) Average Maximum Width Average Width Crack Area  Figure 5.5 (d) - Results of Series-M Tests (Effect of s/c)    Table 5.6 - Effect of w/c on Rate of Evaporation, Heat Evolution and Cracking w/c Property 0.35 0.4 0.5 0.6 Before Demolding 0.98 0.91 1.23 1.66 Avg. Evaporation Rate (kg/m2/h) After Demolding 0.32 0.38 0.43 0.47 Avg. Thermal Output (°C- h) 506 481 475 477 Avg. Time to First Crack (min) 122 128 143 138 Avg. Crack Area (mm2) 57.3 224.6 572 674.9 Avg. Crack Width (mm) 0.59 1.11 1.62 1.54 Avg. Maximum Crack Width (mm) 1.07 2.27 3.97 4.10    111   w/c = 0.6 w/c = 0.5 w/c = 0.4 w/c = 0.35 Figure 5.6 - Typical Crack Pattern (a) Effect of w/c   M8 M8-1 M8-2 Figure 5.6 - Typical Crack Pattern  (b) Effect of coarse aggregates 112  Time to First Crack  Time to first crack was recorded from the time the specimens were placed in the environmental chamber to the first instance when a crack was sighted on the specimens. ‘Time to first crack’ is plotted for Series-M in Figure 5.7 (a).  Notice that for most cases, the first crack appeared between 2 - 2.5 h after casting.  Time to first crack was lower in mixes with lower w/c which agrees with the findings of Wang et al (2001). Average time to first crack increased by 16 minutes when w/c increased from 0.35 to 0.6.  The influence of s/c on time to crack could not be established. Average evaporation rate during the first 2-2.5 h ranged between 0.4 and 0.8 kg/m2/h which according to Almusallam et al (1998) is within the critical range of 0.2-0.7 kg/m2/h at which cracking can be expected and lower than the value of 1 kg/m2/h suggested by ACI 305 (1996). Results for series F are presented in Figure 5.7 (b).  With the exception of F3 and F4, time to first crack increased with an increase in the fly-ash content. Significantly lower time to first crack (143 minutes) was recorded when SRA was used (F6).  Hence, in general, lower time to first crack was recorded for mixes with lower w/c or with lower fly-ash content.  This finding does not completely fit with the fact that a lower w/c and lower fly-ash content lead to a much faster rate of strength gain and formation of a structure at early age that provides the tensile strength to the material.  113  100 120 140 160 180 200 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Mix Designation Ti m e to  fi rs t c ra ck  (m in )  Figure 5.7 (a) - Time to First Crack for Series-M  100 120 140 160 180 200 F1 F2 F3 F4 F5 F6 (SRA) Mix Designation Ti m e to  fi rs t c ra ck  (m in )  Figure 5.7 (b) - Time to First Crack for Series-F     114  5.6 Concluding Remarks Based on the tests described in this chapter the following conclusions were drawn: • Increase in w/c ratio significantly increased the amount and rate of moisture loss (evaporation rate) from the overlays.  Area under the time-temperature curve was calculated in °C-h and termed as “heat evolution.”  Mixes with lower w/c ratio resulted in lower overall heat evolution.  For the range of s/c ratio investigated, no particular trend could be established. • The rate of crack growth was significantly higher during the first four hours. Mixes with higher w/c ratio had a higher crack growth rate when compared to the mixes with lower w/c ratio.  Increase in w/c ratio (within the selected range) significantly increased the observed crack area and crack width.  For the small range of s/c investigated, no definite trend with the crack characteristics could be established. • Time to first crack was lower for mixes with lower w/c ratio and lower fly-ash content when compared to their respective control mixes. • Addition of a low volume of coarse aggregates significantly reduced the crack area and width developed during the first 24 hrs. • Considering the in-batch standard deviation associated with the test method, the effect of fly ash dosage on crack area and crack width could not be clearly established.  Use of a 7 lit/m3 dosage of SRA significantly reduced the average crack width, however, this resulted in no significant reduction in total crack area and the average maximum crack width.  This implies that the type and dosage of SRA was not very effective. 115  CHAPTER 6 - DEVELOPING ‘CRACK-FREE’ OVERLAYS USING FIBERS  6.1 Fiber Reinforced Mixes  In this section, the effect of different fiber types on cracking was investigated using the bonded overlay technique described in Chapter 4.  For this study, the control mix described in Chapter 4 with w/c and s/c equal to 0.5 were used.  Different fiber types investigated are described in Chapter 3 and the fiber dosages investigated are described in Table 6.1.  Typical fiber dosage was studied, which varied between 0 to 0.4% by volume. Eight types of polypropylene fibers with different fiber geometry were studied.  Glass fibers and five types of cellulose fibers were also investigated.  Results of all fibers were compared to an unreinforced mix (a total of 44 fiber reinforced mixes were investigated).  6.2 Results and Discussion  Crack widths were measured using the procedure described in Chapter 4 and total crack area was determined for every specimen using Equation 4.1.  A minimum of three samples were tested for every mix.  The effect of different fibers on controlling cracking was evaluated by comparing averaged “crack area,” “average crack width,” and “maximum crack width” to values determined for the control or unreinforced mix. Results have been arranged according to fiber material and are presented graphically in Figures 6.1 to 6.3.  Glass fiber mix GF1 is included with polypropylene fibers.  For clarity, not more than four polypropylene fibers are plotted in one figure.   116   Table 6.1 - Fiber Dosages Investigated Fiber/Mix Designation Fiber type Fiber Dosage (%) C1 - 0 PF1 FM  150 0.033, 0.066, 0.1, 0.3 PF2 FM  300 0.033, 0.066, 0.1, 0.3 PF3 FM  MD 0.1, 0.2, 0.3 PF4 FM  Stealth 0.1, 0.2 PF5 FM  Stealth 0.1, 0.2, 0.3 PF6 FM Stealth 0.1, 0.2, 0.3 PF7 Fibrillated 0.1, 0.2, 0.3 PF8 Grace MicroFiber 0.033, 0.066, 0.1 CF1 Buckeye (UltraFiber 500) 0.033, 0.066, 0.1, 0.3 CF2 Weyerhaeuser 187 0.1, 0.2, 0.3, 0.4 CF3 Weyerhaeuser 144 0.1, 0.2 CF4 Weyerhaeuser 161 0.2, 0.3 CF5 Weyerhaeuser 177 0.2, 0.4 GF1 Saint-Gobain (AntiCrack fiber) 0.033, 0.066, 0.1   117  These data along with average ‘number of cracks’ observed for different mixes and crack control efficiencies (calculated using equations 4.2 and 4.3) for various fibers are presented in Table 6.2.  Crack control efficiencies for various mixes are plotted against fiber volume fraction in Figures 6.4 and 6.5.  Typical crack patterns observed for some of the fibers are presented in Figures 6.6 to 6.11.  6.3 Effect on Total Crack Area  Since all the fiber-reinforced mixes were being compared with the control mix C1, values for six specimens were averaged and the average total crack area was calculated as 305 mm2.  In general, crack area reduced with an increase in fiber dosage (with the exception of CF1), this trend is clearly seen in Figure 6.5 (a-c) and is also illustrated by the crack patterns in Figures 6.6 to 6.11.  The technique developed by Naaman et al. (2005) was described in Chapter 2.  This technique is used to measure restrained plastic shrinkage.  Naaman conducted a study and measured the total crack area for mixes reinforced with different types of carbon, synthetic, and steel fibers.  These results are presented in Figure 6.1 (d).  Naaman et al. found that polyvinyl alcohol fibers were the most effective followed by carbon fibers.  Steel fibers (FMF in Figure) were the least effective in controlling cracking at volume fractions lower than 0.25%.  However, as was pointed out in Chapter 2, this test technique utilizes an overlay that is less than 35 mm deep and the environmental conditions used are much different (temperature between 35 and 40° C) from that used in the bonded overlay technique.  This combined with the presence of coarse aggregates might be responsible for the lower total crack area observed in this study (approximately 80 mm2).  Several other studies described in 118  Chapter 2 have investigated the use of fiber in reducing restrained plastic shrinkage craking, however, most of them have focused on drying shrinkage as opposed to early- age plastic shrinkage.   0 100 200 300 400 500 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  A re a (m m 2 ) PF1 PF2 PF3 PF4  Figure 6.1 (a)  119  0 100 200 300 400 500 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  A re a (m m 2 ) PF5 PF6 PF7 PF8 GF1  Figure 6.1 (b)  0 100 200 300 400 500 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  A re a (m m 2 ) CF1 CF2 CF3 CF4 CF5  Figure 6.1- (c)  120   (d) Figure 6.1 - Crack Area vs. Fiber Volume (Naaman et al., 2005)  Cellulose fiber:  Data for the cellulose fiber CF1 were inconsistent and when compared to the crack area for the control mix (C1), crack area increased at low fiber dosage of 0.033%, then reduced for 0.066% and 0.1% and then reverted back to that recorded for the control mix for a further increase in fiber volume.  Since these fibers were sized and treated, which prevented easy dispersion and mixing of the fibers and hence led to the high variability in test results.  Moreover, if the graph (Figure 6.1 (c)) is divided into two regions, first with fiber dosage between 0% and 0.066 % and the second between 0.066% and 0.3%.  Within the first region the data is not consistent as the fiber dosage is very small and the test method 121  is not very sensitive to such small fiber dosages.  Some of the preliminary testing with these fibers indicated that at such low fiber concentrations these fibers were ineffective in controlling cracks.  Dosage rates between 0.066% and 0.1% constitute a practical range in which all fibers begin to significantly affect the shrinkage cracking and produce consistent trends.  In this fiber dosage range, all fibers demonstrate a steady decrease in average crack area.  In the case of PF1, the average crack area reduced to 11 mm2 at 0.3% fiber volume fraction.  PF2 demonstrates the same effect but is overall less effective in decreasing the average crack area.  Polypropylene fiber: Results further indicate that while polypropylene fibers in general are effective in controlling plastic shrinkage cracking in concrete, a finer fiber is generally more effective than a coarser one (PF4 and PF8 were more effective than other comparable coarser fibers), and a longer fiber is often more effective than a shorter one. Amongst the polypropylene fibers, PF8, which was 20 mm long was more effective than other types (PF5, PF6, and PF7) that were 12.5 mm or shorter.  Fibrillated polypropylene fiber PF2 at a very low fiber volume of 0.033% was also an outlier as the effect of this fiber on reducing crack area was lower than the test accuracy (in-batch test variability). Amongst synthetic fibers (PF3, PF5, and PF6) having the same denier (6 denier), fiber PF3 was the most effective, followed by PF5 and PF6.  This clearly indicated that a longer fiber of the same denier is more effective. This further implies that polypropylene fibers develop a poor bond with the cementitious matrix at early age and a longer fiber length is necessary for an efficient transfer of stress across a crack.  Also a comparison between fiber PF4 with fiber PF5 indicates that a finer denier fiber is more effective than 122  a coarser denier fiber.  This is expected as a finer fiber would have a larger surface area over which it would bond with the cementitious matrix and thus result in a greater transfer of tensile stress to the fiber.  Also, at a given fiber volume fraction, a finer fiber will have more number of fibers crossing a crack, hence a higher crack growth resistance.  Fiber fibrillations provide an effective mechanical anchorage sufficient to overcome the otherwise poor adhesion between fiber and the matrix. It is also likely that the fibrillated fibers disperse better than their monofilaments counterparts.  To study the effect of fiber fibrillation, monofilament fibers were compared to fibrillated fibers.  Crack areas for fibers PF7 (fibrillated) when compared to monofilament fibers PF4 and PF5 of the same length indicate that fiber PF4 was the most effective followed by PF7 and then PF5.  Also fibrillated fibers PF3 when compared to PF1, indicate that PF1 and PF3 were very similar in reducing the crack area (even though PF3 had much lower denier), but the reduction in crack area with PF2 was marginal.  Unfortunately, the length and denier of these fibers was not identical, hence the effect of fibrillation on crack area could not be clearly established.  Glass fibers and general discussion:  For most fibers, significant reduction in crack area was observed at 0.3%.  For some fibers (CF2 and CF5) at 0.4%, cracks were more or less eliminated producing a ‘crack free’ overlay material.  Glass (GF1) and polypropylene fiber (PF8) were the most effective in reducing crack area and more or less eliminated cracking at 0.1%; crack area reduced to 31 and 0 mm2 respectively.  It is believed that fiber dispersion for these two fibers was also much superior as compared to the other fibers.  Other fibers that significantly reduced cracking were PF4, PF7, PF3, and PF1 in 123  order of decreasing effectiveness.  Fibers of the same length (12.5 mm) were also compared.  Fiber type PF3, which has 50% by mass of the 12.5 mm fiber, performed significantly better than the 6d fiber (PF5).  Hence, in terms of the fiber material, for the fibers investigated, they were ranked in the order of decreasing effectiveness as follows: Glass → Polypropylene → Cellulose  For the cellulose fibers, conclusions were drawn based on the visible trends and extrapolated values to other fiber volume fractions.  Other than fiber CF1, the trend for all cellulose fibers was similar and no definite conclusions could be drawn to distinguish performance of different fiber types based on fiber length and surface coatings (bleached and unbleached).  Except for the mix CF1, increase in the volume fraction of other cellulose fibers beyond 0.2%, resulted in significant reduction in crack areas (steep slope of the curve in Figure 6.1).  Results also demonstrate that appropriate addition of these fibers (CF2 and CF5 at 0.4%) can eliminate all plastic shrinkage cracking.  6.4 Effect on Crack Width Average crack width and maximum crack width are plotted against fiber volume fraction for different fibers in Figures 6.2 (a-c) and 6.3 (a-c) respectively.  Average crack width is indicative of the overall crack width on the surface of the specimens, however, the maximum crack width is useful when evaluating a structure for serviceability and durability.  The crack width and length are the only measured parameters which are then used to calculate other parameters such as the crack area.  As such the trends observed 124  with the crack widths are to a large extent consistent with those observed in the case of crack areas in the previous section. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  W id th  (m m ) PF1 PF2 PF3 PF4  Figure 6.2 (a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  W id th  (m m ) PF5 PF6 PF7 PF8 GF1  Figure 6.2 (b) 125  0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) A ve ra ge  C ra ck  W id th  (m m ) CF1 CF2 CF3 CF4 CF5  (c) Figure 6.2 - Average Crack Width vs. Fiber Volume  Average and maximum crack width for control mix (C1) was 2.17 and 2.77 mm respectively.  As in the case of crack area, crack width also reduced with an increasing fiber volume fraction.  Average and maximum crack width data followed a similar trend. Results for PF1, PF2, and CF1 for fiber dosage rates below 0.1% were inconsistent. Inconsistency also occurred further when average crack width for PF2 and CF2 increased to 0.83 and 1.42 mm respectively at fiber dosage of 0.3% as compared to 0.79 and 0.56 mm at 0.1% fiber dosage.  Glass fiber GF1 was the most effective in reducing the maximum crack width from 2.77 mm for control mix to 1.02 and 0.11 mm for fiber dosage of 0.033% and 0.066% respectively.  For mixes with cellulose fibers (CF2 and 126  CF5), a dosage of 0.4% was required to reduce the maximum crack width to 0.6 and 0.31 mm, respectively. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) M ax  C ra ck  W id th  (m m ) PF1 PF2 PF3 PF4  Figure 6.3 (a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) M ax  C ra ck  W id th  (m m ) PF5 PF6 PF7 PF8 GF1  Figure 6.3 (b) 127  0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Fiber Volume (%) M ax  C ra ck  W id th  (m m ) CF1 CF2 CF3 CF4 CF5  (c) Figure 6.3 - Maximum Crack Width vs. Fiber Volume 128  Table 6.2 - Detailed Test Results for Different Mixes Mix/ Fiber Desig- nation Fiber Type Fiber Volume Fraction (%) Average Area (mm2) Average Width (mm) Max Width (mm) No. of Cracks Crack Area Control Efficiency (ηarea) % Crack Width Control Efficiency (ηwidth) % C1 Control 0.0 305 2.17 2.77 2.3 0 0 0.033 292 0.93 2.19 3 4.3 20.8 0.066 317 1.87 3 2 -3.9 -8.4 0.1 211 1.03 2.2 2 30.7 20.5 PF1 FM 150 0.3 11 0.12 0.5 0.3 96.3 81.9 0.033 220 0.90 1.88 4 28.0 32.0 0.066 400 1.65 3.1 2 -31.3 -12.0 0.1 217 0.79 2.2 3 29.0 20.5 PF2 FM 300 0.3 195 0.83 2.28 2 35.9 17.6 0.1 212 1.48 1.92 3.0 30.5 30.6 0.2 59 0.52 0.54 2.0 80.5 80.5 PF3 FM  MD 0.3 13 0.18 0.42 2.0 95.9 84.8 0.1 121 1.02 1.32 2.0 60.3 52.3 PF4 FM  Stealth 0.2 4 0.06 0.18 0.3 98.8 93.5 0.1 216 1.16 1.32 3.7 29.2 52.3 0.2 120 0.85 1.04 2.7 60.8 62.4 PF5 FM  Stealth 0.3 102 0.68 0.89 7.0 66.6 67.8 0.1 258 1.38 2.00 3.3 15.5 27.7 0.2 243 1.32 1.42 3.0 20.4 48.7 PF6 FM Stealth 0.3 154 1.08 1.40 3.0 49.4 49.4 0.1 173 0.93 1.02 2.7 43.3 63.1 0.2 43 0.47 0.54 2.0 85.9 80.5 PF7 Fibrillated 0.3 31 0.34 0.38 1.3 89.8 86.3 0.033 158 1.11 1.94 1.3 48.1 29.9 0.066 212 0.78 1.45 2.7 30.4 47.6 PF8 Grace MicroF iber 0.1 31 0.31 0.43 0.3 89.7 84.3 0.033 394 1.64 2.8 2 -29.3 -1.2 0.066 288 1.85 4 2 5.5 -44.6 0.1 198 0.56 1.86 3 35.0 32.8 CF1 Buckey e (UltraFi ber 500) 0.3 295 1.42 2.96 2 3.2 -7.0 0.1 284 1.17 1.31 2.5 6.8 52.6 0.2 240 1.72 2.74 2.0 21.5 0.8 0.3 65 0.41 0.79 2.3 78.7 71.5 CF2 Weyerh aeuser 187 0.4 27 0.27 0.60 0.7 91.1 78.3   129   Table 6.2 - (Continued)- Detailed Test Results of different mixes Mix/Fi ber Design ation Fiber Type Fiber Volume Fraction (%) Average Area (mm2) Average Width (mm) Max Width (mm) No. of Cracks Crack Area Control Efficiency (η area) % Crack Width Control Efficiency (η width) % 0.1 242 2.33 3.00 2.7 20.5 -8.4 CF3 Weyerh aeuser 144 0.2 230 1.83 2.50 1.7 24.5 9.6 0.2 280 2.00 2.50 2.5 8.4 9.6 CF4 Weyerh aeuser 161 0.3 145 1.53 2.50 3.3 52.5 9.6 0.2 215 1.92 3.00 2.0 29.4 -8.4 CF5 Weyerh aeuser 177 0.4 10 0.10 0.31 0.3 96.6 88.8 0.033 56 0.54 1.02 1.0 81.7 63.3 0.066 9 0.06 0.11 0.7 97.1 96.1 GF1 Saint- Gobain (AntiCr ack fiber) 0.1 0 0 0 0 100.0 100.0   6.5 Efficiency Factors  To further quantify the efficiency of every fiber in controlling shrinkage cracking, ‘Crack Width Control Efficiency (ηwidth)’ and ‘Crack Area Control Efficiency (ηarea)’ were calculated according to Equations 4.2 and 4.3.  These values are included in Table 6.2 and presented in Figures 6.4 (a), (b) and 6.5 (a), (b).  It is evident from these figures that glass fiber GF1 was very efficient in controlling the crack area and had 100% crack and width control efficiency at 0.1% fiber volume.  Fibers PF4 at 0.1% had crack area and crack width control efficiencies of 60.3% and 52% respectively.  Similarly, PF8 at 0.1% was more efficient with area and width control efficiency of 89% and 84% respectively.  At 0.1% fiber dosage, all other fibers 130  had less than 45% crack area control efficiency.  However, at the same volume fraction, except for CF2, CF3, and CF4 (where ηwidth reduced or was negligible) all other fibers had a crack width control efficiency of more than 20%.  Crack area and width control efficiencies were both greater than 49% at 0.3% fiber dosage for all fibers except PF2, CF1, and CF4.  Note that this conclusion does not apply to fiber type CF3 and CF5, since they were not tested at 0.3% fiber dosage.     131  -40 -20 0 20 40 60 80 100 0.0 0.1 0.2 0.3 0.4 Fiber Volume (%)C ra ck  A re a C on tr ol  E ffi ci en cy  (% ) PF1 PF2 PF3 PF4 PF5 PF6 PF7 PF8  Figure 6.4 (a) -40 -20 0 20 40 60 80 100 0.0 0.1 0.2 0.3 0.4 Fiber Volume (%)C ra ck  A re a C on tr ol  E ffi ci en cy  (% ) CF1 CF2 CF3 CF4 CF5 GF1  (b) Figure 6.4 - Crack Area Control Efficiency (ηarea) vs. Fiber Volume 132  -40 -20 0 20 40 60 80 100 0.0 0.1 0.2 0.3 0.4Fiber Volume (%)C ra ck  W id th  C on tr ol  E ffi ci en cy  (% ) PF1 PF2 PF3 PF4 PF5 PF6 PF7 PF8  Figure 6.5 (a) -50 -30 -10 10 30 50 70 90 0.0 0.1 0.2 0.3 0.4 Fiber Volume (%) C ra ck  W id th  C on tr ol  E ffi ci en cy  (% ) CF1 CF2 CF3 CF4 CF5 GF1  (b) Figure 6.5 - Crack Width Control Efficiency (ηwidth) vs. Fiber Volume 133   6.6 Effect on Average Number of Cracks  The average ‘number of cracks’ observed on the specimens are given in Table 6.2. On an average 2-3 cracks were observed on most specimens including the control (C1). For mix C1, instances where fewer cracks were observed, resulted in larger crack widths. In general, the ‘number of cracks’ were not affected at small fiber dosages.  However, as the fiber dosage increases, and the fiber efficiency in reducing cracking improved, the number of cracks reduced.  PF5 was an exception where at 0.3% fiber dosage, the average number of cracks increased although the crack area and crack widths were lower when compared to the values at 0.1% fiber volume.  6.7 Time to First Crack  To study the effect of fiber on the initiation of the crack, “Time to first crack” as defined in Chapter 5 (section 5.5.3) was recorded for some mixes in this study.  This is reported in Table 6.3.  Cracking of specimens due to plastic shrinkage is a direct function of the rate of bleeding and evaporation (Topcu and Elgun, 2003).  Depending on the type of fiber (hydrophobic or hydrophilic), water can be retained in the mix and depending on the dispersion of fibers in the mix, the voids and bleeding channels can be significantly modified affecting bleeding.  The average time to first crack for the control mix was 148 minutes, which was 28 minutes after demolding the specimens.  For the synthetic fibers, PF1 and PF2 (treating mix PF1 with 0.066% volume of fibers as an anomaly), the time to first crack increased to more than 163 minutes at 0.3% fiber volume.  This meant that fibers beyond a certain 134  fiber volume delayed the onset of shrinkage cracking.  As far as the cellulose fiber CF1 is concerned, once again treating 0.1% fiber volume as an anomaly, a similar trend was observed.  At 0.3% fiber volume the time to first crack was 162 minutes.  However, the delayed onset of cracking always did not result in lower crack areas, as in the case of CF1.  Table 6.3 - Effect of Fiber on “Time To First Crack’ Mix/Fiber Designation Fiber Material Fiber Volume Fraction (%) Average Time (min) C1 - 0 148 0.033 133 0.066 165 0.1 147 PF1 Polypropylene 0.3 166 0.033 139 0.066 148 0.1 140 PF2 Polypropylene 0.3 163 0.033 150 0.066 150 0.1 131 CF1 Cellulose 0.3 162    135  6.8 Concluding Remarks Chapter 6 describes the performance of several different fiber types under restrained plastic shrinkage conditions that was measured using the bonded overlay technique. Based on the test results described in this chapter, the following conclusions can be drawn: • Overall, in terms of the fiber material, glass fibers were the most effective in reducing cracking in the overlays, followed by polypropylene, and then cellulose fibers.  Glass fiber completely eliminated cracking at a small fiber volume fraction of 0.1%.  The high efficiency of glass fibers was possibly due to the combined effect of longer length, small diameter, and bond strength of the fibers. • Amongst all the polypropylene fibers, fiber type PF8 was the most effective in controlling cracking in the overlays.  These fibers reduced the maximum crack width by 47%.  Influence of fiber geometry on shrinkage cracking was also observed from the test results.  Finer fibers were more effective than coarser fibers, and longer fibers were more effective than the shorter ones. • Except for cellulose fiber type CF1, for all other fibers an increase in the fiber dosage resulted in a reduction in the total measured crack area.  A fiber volume fraction of cellulose fibers higher than 0.1% was required to cause notable reduction in crack areas.  For some of the cellulose fibers, poor dispersion was attributed to their poor performance, however a volume fraction of 0.4% for some cellulose fibers completely eliminated cracking. • Average crack widths and maximum crack widths followed a similar trend as in the case of crack areas.  Glass fibers, at only a fiber volume fraction of 0.066% 136  reduced the maximum crack width by 96%.  At the same volume fraction, synthetic fiber type PF8 reduced the average crack width and maximum crack width to 0.78 and 1.45 mm respectively. • In general, at high fiber volume fractions, the number of cracks reduced and the time to first crack increased for most fiber types. • Crack width control efficiency and crack area control efficiency for all fiber types were useful in quantifying the efficiency of fibers in reducing cracking.  Glass fibers at a fiber volume fraction of 0.1% had an efficiency of 100% in controlling crack areas and widths.  In terms of crack control efficiencies, at a fiber volume fraction of 0.1%, synthetic fiber PF8 ranked second.  At this volume fraction, all other fibers had an efficiency lower than 45%.    Figure 6.6 - Crack Pattern for C1  137    a) PF1  0.033% b) PF1  0.066%   c) PF1  0.1% d) PF1  0.3% Figure 6.7 - Crack Pattern for PF1   138   a) PF2  0.033% b) PF2  0.066%   c) PF2  0.1% d) PF2  0.3% Figure 6.8 - Crack Pattern for PF2      139    a) PF8  0.033% b) PF8  0.066%  c) PF8  0.1 Figure 6.9 - Crack Pattern for PF8        140     a) CF1  0.033% b) CF1  0.066%   c) CF1  0.1% d) CF1  0.3% Figure 6.10 - Crack Pattern for CF1    141       a) GF1  0.033% b) GF1  0.066%   c) GF1  0.1% Figure 6.11 - Crack Pattern for GF1  142  CHAPTER 7 - FIELD APPLICATIONS AND HEALTH MONITORING OF FIBER REINFORCED OVERLAYS  7.1 Introduction The science of monitoring structural health in a continuous or periodic manner is commonly referred to as ‘Structural Health Monitoring’ (SHM).  Merging Civil Engineering with electronics is now being called ‘Civionics,’ and according to ISIS Canada, “Civionics is to Civil Engineering what Avionics is to Aerospace” (ISIS Canada 2004).  SHM can be successfully performed in many different ways including monitoring strain variation using embedded sensors and by other non-destructive tests (NDTs). Many researchers have described the applications of fiber optic sensors for monitoring bridge structures (Tennyson et al. 2001, Shehata & Rizkalla 1999, Fuhr & Huston 1998). Several studies have been carried out to investigate the effects of fiber reinforcement on development of strain in polymeric (synthetic) matrices using embedded fiber-optic sensors and bragg-grating sensors (Bao et al. 2002; Botsis et al. 2004; Jinno et al. 2003a; Jinno et al. 2003b; Kalamkarov et al. 1999).  In the case of cement-based matrices, on the other hand, only limited number of studies is available where embedded sensors were used to measure early age strains (Al-Obaid 1989; Gao et al. 2004; Habel et al. 1997; Slowvik et al. 2004).   Although there are many field applications where Fiber-Reinforced Concrete (FRC) has been used, however, no study could be found in the literature that dealt with early age monitoring of FRC using embedded strain gauges.  The effect of mixture proportion and that of fibers in reducing restrained plastic shrinkage cracking was described in Chapters 5 and 6 respectively.  The bonded overlay 143  technique previously described in Chapter 4 was used for this purpose.  Using the findings from Chapter 5 and 6, types and volume fractions of fibers that prevent cracking under severe environmental conditions were identified and “crack-free” overlay materials were developed.  To predict the behavior of such materials, when used for real applications, under real environmental conditions and using real mixes, performance of unreinforced and FRC mixes were compared.  Since, the FRC slabs were not expected to crack under less severe conditions (real conditions) for shrinkage, sensors were embedded in concrete to monitor early-age strain development.  NDTs were conducted to determine any internal cracking or cracking not visible to the naked eye.  In this chapter, two field studies, a repair project and a new construction project are reported.  A brief summary of a previously conducted repair project in a parking garage involving the use of carbon fiber reinforced concrete is also described.  In all the three field studies, sensors were embedded in the overlays to monitor in-situ strains as a function of time.   7.2 Aquatic Center Project  In this section, experience gained from a previously conducted field study where carbon fiber reinforced concrete was used in a parking garage in Downtown Vancouver is briefly described.  Afterwards, a study where the plaza deck slab at UBC Aquatic Center was instrumented is described in details.  Performance of these slabs was monitored by measuring strain remotely using a web-based data acquisition system and by performing 144  a combination of NDTs (rebound hammer, ultrasonic pulse velocity, and resisitivity measurements).  7.2.1 Site Preparation and Mix Proportions  At the Aquatic Center project, a 32 MPa concrete (see mixture proportions in Table 7.1) was used which was reinforced with cellulose fibers (Figure 7.1).  Properties of the cellulose fibers used at the Aquatic Center project were similar to CF4 and CF5 (unbleached cellulose micro-fibers with no surface treatment) described in Table 3.2 in Chapter 3.  A fiber dosage of 0.2% by volume was added in the concrete used at the Aquatic center.  Industrial standard 12.7 cm x 12.7 cm (4” x 4”) wire mesh was used at half the overlay depth.  An additional 2.25 l/m3 of superplasticizer was added to FRC to compensate for the loss of workability due to the fibers and achieve the same workability as plain concrete.  The substrate at the Aquatic Center was cleaned manually before placing the 200 mm overlay.  Aquatic centre plaza deck was 2.1 x 8.5 m in size.  At the parking garage site, a 40 MPa repair concrete (see mixture proportions in Table 7.1) was placed as an overlay with an average thickness of 75 mm.  Typical properties of carbon fibers used at the parking garage were as follows: tensile strength = 2111 MPa, tensile modulus = 232 GPa, elongation to break = 1.0%, fiber diameter = 9-11 microns, and fiber density = 1900 kg/m3.  Prior to placement of the overlay, the surface was thoroughly roughened using an industrial sand blaster and an industrial shot blaster. All loose concrete particles were removed, and steel reinforcement was cleaned of any attached rust.  The substrate at this site was very different from that at the Aquatic center. 145  The substrate consisted of a structural slab with top surface containing rusted rebars as opposed to a membrane on top of a structural slab at the Aquatic center.   Figure 7.1 - Cellulose fibers   7.2.2 Gauge Location and Data Monitoring  Figures 7.2 and 7.3 show the placement of strain gauges in the parking garage and Aquatic Center projects, respectively.  Electrical strain gauges (Figure 7.4) were located 75 mm and 25 mm below the finished surface of the overlay, respectively, in the parking garage and Aquatic Center.      146   Table 7.1 - Mix Proportions of Repair Concrete Used at the Parking Garage and Aquatic center Ingredient Parking garage Aquatic center Cement, CSA, Type 10 (kg/m3) 320 310 Fly ash (kg/m3) 70 77.5 Silica fume (kg/m3) 23 - 14mm aggregates (kg/m3) 1035 950 Sand (kg/m3) 730 950 Water (kg/m3) 150 174 Water reducing admixture (ml/100kg of cementitious) 330  580 Air entraining admixture (ml/100kg of cementitious) 130  - Strength (MPa) 40 32 Carbon Fibers (kg/m3) 1.9 kg/m3 (0.1%) Cellulose Fibers (kg/m3)  3.0 kg/m3 (0.2%)   147  Strain gauge on rebar  Figure 7.2 (a)   (b) Figure 7.2 - Parking garage (a) a strain gauge and (b) placement of concrete   At the parking garage, gauges were embedded at two different locations (1 and 2) in plain and fiber concrete.  Specially prepared chairs were used for mounting the gauges. 148  Detailed description of strain gauge layout, calibration, and placement of concrete may be found elsewhere (Banthia & Nandakumar 2001; Banthia et al. 2004).  For both sites, two placements were prepared side-by-side: one with fiber reinforcement and the other without fiber reinforcement (hereafter termed ‘plain’).    Four gauges were placed along the center line in each overlay, sets of two at an angle of ±45° to nullify any preferential strain from the transverse or the longitudinal direction.   Figure 7.3- (a)  149  Control Overlay (4.25 m) FRC Overlay (4.25 m) Storm water drain Location of finishing joint 2.1 m wide  Figure 7.3 (b)   (c)  Figure 7.3 - Aquatic Center Plaza Deck (a) Slab Schematic (b) Location of Strain Gauges, and (c) Placement and Finishing of Concrete.  150   Figure 7.4- Embedded Strain Gauge   Figure 7.5 - Specially Designed Chairs for Strain Gauge   The gauges were placed 25 mm below the finished surface using special chairs (Figure 7.5) and were connected to a web-based data acquisition system called WebDaq/100 (Figure 7.6). This system was capable of recording the data from the eight sensors and transferring it remotely via the web/internet.  Details about this data acquisition system can be found in Banthia et al. (2004).  The strains were monitored intermittently for a period of eight months starting from the time of placement of 151  concrete.  Sample data acquired continuously every 10 minutes during the first couple of weeks is presented in the results.   Figure 7.6 - Web-based Data Acquisition System WebDaq/100  7.2.3 Results- Strain from Embedded Sensors  Averaged strain readings from the Aquatic Center are plotted in Figure 7.7. Sensors in both the plain overlay and the FRC overlay recorded compressive strain.  The strain in FRC overlay at any given time was lower than the strain of the plain overlay. This is in agreement with the findings of Al-Obaid (1989) who also found that shrinkage strains decrease when fibers are used.  Due to the inability to accurately calibrate and ‘zero’ the strain sensors in-situ, only a comparative assessment between sensors at a given site is possible.  Since the strain sensors in-situ did not start measurements at 152  ‘zero’, the recorded strain values at the Aquatic Center were higher than expected.  For plotting purposes, the strain values recorded from the two projects are compared on a log scale in Figure 7.8.  Compressive strains were given a negative sign to identify them as such.  0 2000 4000 6000 8000 10000 0 5 10 15 Time ( Days) C om p.  s tr ai n ( X 10 -6 ) FRC Plain  Figure 7.7- Strain Measurement from the Aquatic Center Overlay    The average maximum strains recorded during the first 2 weeks after casting at the Parking Garage are compared with those at the Aquatic Center in Figure 7.8.  Notice that unlike the Aquatic Center, the FRC overlay in the Parking Garage recorded a tensile strain.  These observations, while appear contradictory, actually are not, and indicate a 153  common mechanism in both overlays.  When the overlay shrinks (Figure 7.9), it develops a compressive strain near the surface, and this compressive strain results in tensile stresses due to the geometric restraint.  In Figure 7.9, a linear distribution in strains through the depth is assumed which is somewhat different from the observations of Gao et al. (2004) but adopted nonetheless for simplicity and ease of illustration.  If the induced tensile stress exceeds the tensile strength, cracking would occur.  Reduced strains in the FRC overlay (Figure 7.7) thus indicate a reduced possibility of cracking in such an overlay.  Some distance away from the top surface, the compressive strains would revert to tensile strains thus creating a neutral axis and preserving the equilibrium of the overlay.  In the case of the Aquatic Center, where the sensors were placed closer to the surface, sensors in both the plain and the FRC overlay remained above the neutral axis and recorded compressive strains.  In the case of the FRC overlay, assuming that the severe drying occurs only above the neutral axis and the material below the neutral axis develops strains similar to the plain placement, a reduced level of strain at the level of the sensors implied that the neutral axis had moved upwards.  This hypothesis is further supported by the common observations previously reported in section 2.4 that cracks in FRC placements remain shallower and narrower than plain placements.  154  -4.5 -3 -1.5 0 1.5 3 0 10 20 30 40 50 60 70 80 Depth of gauge from surface M ic ro  s tr ai n (lo g va lu es ) Aquatic-FRC Aquatic-Plain Garage-Plain-1 Garage-Plain-2 Garage-FRC-1 Garage-FRC-2  Figure 7.8 - Measured Strain from Aquatic Center and the Parking Garage    In the case of the Parking Garage, on the other hand, sensors were placed at a greater depth from the surface and an upward movement of the neutral axis in the case of FRC resulted in the sensor in this case falling below the neutral axis.  Sensors in the FRC overlay, therefore recorded tensile strains.  Reduced compressive strains at the surface would, once again, reduce the possibility of cracking in the FRC overlay and indeed, much reduced cracking was recorded in the FRC overlay in the Parking Garage (Banthia & Nandakumar 2001).  155   -ε (Comp.) +ε (Tension) Plain C t FRC Substrate Interface Gauge location-Garage Gauge location-Aquatic Neutral Axis FRC- Stresses in Overlay & Substrate Neutral Axis Stress σ ft-Strength   Figure 7.9- Schematic Strain Variation in FRC and influence of Strain Gauge Location  7.2.4 Non Destructive Testing The device used for measuring ultrasonic pulse velocity (UPV) through concrete is often termed as PUNDIT, an acronym for Portable Ultrasonic Non-destructive Indicating Tester.  UPV measurements (ASTM C 597, 1998) can be used to evaluate cracking in concrete, since ultrasonic pulses take much longer to travel through cracks, cavities and other defects especially when they are not saturated with water and are air- filled.  UPV values can be used to study the uniformity and density of concrete placements. In general, a higher velocity would suggest a superior placement since the pulse would encounter fewer defects, voids, and cracks during its travel. A denser specimen thus has a higher velocity while a sample with numerous pores would have a lower velocity.  This can thus help detect the changes in concrete properties, severity of deterioration or cracking. Malhotra, et al. (1991) describe the aforementioned method to 156  be very effective in studying the homogeneity of concrete and evaluating relative quality of concrete. The other applications of this method are to check the density of concrete for evaluating effectiveness of consolidation and for locating areas of honeycombed concrete. It is generally advisable to use NDTs in combination; for example UPV and Schmidt rebound value measurements are recommended to be combined for better reliability and understanding (Shah 2002, Pascale et al. 2003, Qasrawi 2000, and Ward & Langan 1994).  Hence, in this investigation along with UPV, Schmidt rebound numbers and electrical resistivity values were measured for plain concrete and FRC overlays at the UBC Aquatic Center.  Historically, resistivity has been measured for soil, but the technique has also been used for measuring resistivity of concrete on site (Broomfield 1997).  Electrical resistivity measurements of the overlay were used to identify zones of non-conformities.  Non-conformities in concrete can arise due to chloride concentration, moisture content and regions of cracking, which would decrease and increase the resistivity respectively.  These measurements were made using a resistivity meter based on the four-probe Wenner principle.  In this technique, current is applied between the outer probes and potential drop measured between the inner probes to eliminate any effects due to the surface contact resistances (Broomfield 1997).  NDTs were conducted on the overlays at the Aquatic Center eight months after casting to assess the performance of the overlays.  The concrete deck slab was divided into a 60 cm x 60 cm grid for conducting NDTs.   157  7.2.4.1 Ultrasonic Pulse Velocity  A 60 cm probe distance was chosen based on preliminary UPV tests that indicated this distance to be the optimum for UPV measurements.  Figure 7.10 (a) adopted from ASTM C 597 (1998) shows the schematic of typical pulse velocity apparatus.  The most accurate method of measurements is by pacing the transmitting and receiving probes directly across a sample.  However, for the slabs in this investigation were accessible only from the top and hence, the measurements had to be made on the surface using the indirect method.  Figure 7.10 (b) shows the placement of the probes on the slabs. According to ASTM C 597, such measurements are indicative of only surface layers. This was acceptable in this study, as the aim of measuring UPV was to study only the overlay and not the supporting slab.   To Transmitting Transducer From Receiving Transducer   Figure 7.10 (a) - Schematic of Pulse Velocity Apparatus (adopted from ASTM C 597- 97) 158          Figure 7.10 (b) - Schematic Showing Placement of Probes for Indirect Measurement    Figure 7.11 shows a Portable Ultrasonic Non-Destructive Digital Indicating Tester (PUNDIT) that was used to conduct testing on the same day to maintain similar environmental conditions; temperature during the testing period was about 20°C.  Naik and Malhotra (1991) suggest correction values for UPV due to change in temperature and moisture.  At 20°C, the correction for air-dried concrete and water-saturated concrete is 0% (Table 7.2). Hence, even if the cellulose fibers were to absorb and retain any moisture, the error in UPV would be negligible. The instrument was calibrated several times in the field to ensure precise measurements. Surface of the slab at each grid point was smoothened using a sander (Figure 7.12) before taking the readings. A water-based gel “Aquasonic100” (used for ultrasounds) was utilized to ensure proper contact between the UPV probes and concrete surface. This gel facilitated easy cleaning without leaving a greasy surface. Transmitting Transducer Receiving Transducer 159   Table 7.2 - Corrections for Pulse Velocity Due to Temperature Changes (Malhotra and Carino 1991) Correction (%) Concrete Temperature (°C) Air Dried Concrete Water Saturated Concrete 60 +5 +4 40 +2 +1.7 20 0 0 0 -0.5 -1 Under -4 -1.5 -7.5    Figure 7.11- A PUNDIT  Figure 7.12- Grid Points Smoothened using the Sander  160  7.2.4.2 Schmidt Rebound Hammer A square grid of size 60 cm was marked on the overlay.  Schmidt rebound hammer (Figure 7.13) was used to evaluate the compressive strength at the grid points. Testing was conducted according to ASTM C 805-97 (1998) and readings were taken perpendicular to the deck slab surface.  Three rebound values (R) were recorded at each grid point and average compressive strengths were determined using a calibration chart.  Figure 7.13 - Schmidt Rebound Hammer   7.2.4.3 Resistivity Measurements Resistivity was measured using a four-probe set-up (Figure 7.14), also called a Wenner probe. A resistivity meter with the following specifications was connected to the probe to determine resistivity values (Davis Inotek Instruments 2004). The details of this instrument are as follows: (a) Type: Megger Rechargeable Ground/Earth Resistance Tester, (b) Test frequency 128 Hz, open circuit voltage 50 V RMS, 161  (c) Short circuit current 10 mA maximum, accuracy ±2% of reading ±3 digits, and maximum service error ±5% of reading ±3 digits, (d) Fully automatic operation, autoranging from 10 mOhms to 20 kOhms, and (e) Conformation to BS7671, BS7430, BS6651 and VDE 0413   Resistivity measurements can be made through three or four contact points (measurement terminals).  The tester was capable of making three or four terminal measurements, and, in this investigation, four terminal measurements were made. The instrument generated a ±5 Amp square wave current, which prevented polarization at the two probes. Polarization refers to formation of charges at the probes that can induce voltage errors in the readings.  The circuit diagram of the resistivity meter is shown in Figure 7.15. The four-probe set-up was specially designed with copper probes spaced 50 mm apart; each probe loaded with a spring to ensure proper contact with uneven concrete surfaces. To facilitate appropriate connection with the resistivity meter, banana plug sockets were used. A salt free electrode gel “Spectra 360,” generally used in the medical field was applied between the copper probes and concrete surface to ensure proper surface conduction. Square wave current (±AC) was applied through the two outer probes and the drop in potential was measured between the two inner probes.  162  Four-probe set-up Copper probes; spaced 50 mm apart  Figure 7.14 - Resistivity Meter and Four-Probe Set-up  V 120 V from plug V 1000 Ω Voltmeter to measure ‘I’ across the ‘R’ Transformer  Figure 7.15 - Circuit Diagram of the Resistivity Meter  The instrument displays a warning when sensing abnormally high current and voltage by lighting indicators ‘Rc’ and ‘Rp’ respectively. When tests were conducted on 163  plain concrete and FRC overlays in dry state, both lights (Rc and Rp) flashed, indicating very high resistance values (higher than 20 kΩ on the tester). However, after a couple of days of exposure to rain, the overlays were in moist surface dry condition and readings could be taken. Resistivity was measured for both overlays and values were expected to provide useful comparison between plain concrete and FRC overlays. Readings were taken across the grid; longitudinally and transversely. For some readings the ‘Rp’ indicator flashed indicating higher than normal potential drop, however, these readings corroborated well with adjacent readings. The resistivity ‘ρ’ of the semi-infinite, homogenous concrete overlays was calculated according to equation (7.1) as described by Broomfield (1997).  IaV /2πρ =  Equation (7.1)  where, ‘a’ is the electrode spacing (5 cm in this investigation), ‘I’ is the applied current, and ‘V’ is the potential measured across the inner probes. ‘V/I’ is the resistance ‘R’ measured by the resistivity meter and displayed on its screen.  7.2.5 Results- NDTS 7.2.5.1 Ultrasonic Pulse Velocities “Tecplot 8” was used to prepare 2-d and 3-d contour plots for better visualization (Figure 7.16 (a)).  It is evident from Figure 7.16 (a) that UPV was much lower in the FRC overlay and in the region close to the storm water drain (location shown in Figure 7.3 (b)).  The average UPV through both overlays was 3040 m/sec (Table 7.3); comparable to that through typical concrete specimens, which is generally between 3500 and 4500 164  m/sec (Malhotra and Carino 1991).  The average values indicated an overall value for easy comparison between placements.  Average UPV in plain concrete and FRC was 3329 m/sec and 2745 m/sec respectively; 21% higher in plain concrete.  Maximum, minimum, and standard deviation for both overlays combined were 4225 m/sec, 2135 m/sec and 518 m/sec respectively.  Refer to Appendix A for detailed UPV test results.  Table 7.3 - UPV Results for Plain Concrete and FRC UPV (m/sec) Overlay Average Standard Deviation Maximum Minimum Both overlays 3040 518 4225 2135 Un- reinforced 3329 468 4225 2135 FRC 2745 385 3974 2158  Lower average UPV values in FRC were attributed to the presence of hydrophilic cellulose fibers that can lead to higher water retention and porous interfaces.  To confirm this hypothesis, further confirmatory laboratory testing was conducted, and is described in section 7.2.5.4.  165              Figure 7.16 (a) - 2D and 3D Plots of UPV (in m/sec) FRC FRC  Drain Location 166  7.2.5.2 Rebound Hammer Values Figure 7.16 (b) shows the 2-D and 3-D contour plots of average compressive strength measured across the overlay slabs.  Average, maximum and minimum strengths for both overlays combined were 38 MPa, 50 MPa, and 23 MPa respectively (Table 7.4).  This was calculated to get a better sense of the overall strength of the overlay.  The average compressive strength of FRC was 39 MPa; similar to that for plain concrete.  Refer to Appendix B for rebound numbers and strengths measured at all nodes across the grid.  According to Malhotra (1976), strength estimation of concrete using Schmidt hammer is generally only with an accuracy of ±15-20% for specimens cast, cured and tested in the lab.  In this investigation, overall standard deviation for strength measured using Schmidt hammer was only 4.8 MPa (coefficient of variation of 12.5%).  Schmidt hammer values were hence very uniform and consistent.  In the vicinity of the storm water drain low strength values were recorded. This is consistent with low UPV measurements in the same region. Since Schmidt rebound hammer measures the surface hardness of concrete, similar rebound numbers for FRC and plain concrete indicated that the fibers do not negatively affect the hardness or compressive strength of concrete, even though they might affect the UPV values as described in the previous section.      167    Table 7.4 - Compressive Strength Using Schmidt Hammer Compressive Strength (MPa) Overlay Average Standard Deviation Maximum Minimum Both overlays 38 4.8 50 23 Un- reinforced 38 5.2 47 23 FRC 39 4.5 50 29         168                Figure 7.16 (b) - 2D and 3D Plots of Strength (in MPa)  FRC FRC 169  7.2.5.3 Resistivity Values Overall, resistivity of both slabs was within the range of resistivity values for concrete (dry-indoors) and concrete (saturated) as described by Whiting and Nagi (2003) in Figure 7.17. Resistivity across the overlays is shown in Figure 7.17.  It is clear from the plots that the electrical resistivity of the plain concrete overlay was much higher than the section reinforced with cellulose fibers. The vicinity of storm water drain had very low resistivity values. Average resistivity values of plain concrete and FRC were 117 and 71 KΩ-cm respectively. The overall standard deviation was 37 KΩ-cm (COV ~ 41%). Average, minimum, maximum, and standard deviation for un-reinforced concrete and FRC are given in Table 7.5.  It is evident that the average resistivity of FRC was much lower when compared to un-reinforced concrete. Detailed resistivity readings are included in Appendix C.  Table 7.5 - Resistivity for Plain Concrete and FRC Overlays Electrical resistivity (KΩ-cm) Overlay Average Standard Deviation Maximum Minimum Both overlays 90 37.4 176 39 Un- reinforced 117 39.9 176 43 FRC 71 19.4 111 39  170            167.6 43 124.643 116.043 73 .0 42 64.4419 10 7. 44 2 55.8418 47.2417 0 200 400 600 800 0 100 200   0 200 400 600 80 X (cm)0 167.6 43 124.643 9 8.84 24 73.042 47.2417 55.8418 64.4419 10 7. 44 2 167.643 133 .243  Figure 7.17 - 2D and 3D plots of Resistivity (in KΩ-cm) FRC FRC 171    Figure 7.18 - Range of Electrical Resistivity for a Variety of Materials (Whiting and Nagi, 2003)  7.2.5.4 Confirmatory Laboratory Testing  Laboratory tests were performed to further investigate the influence of fibers on UPV and electrical resistivity.  The purpose of the UPV investigation was to study pulse attenuation due to fibers and to determine the extent of variability in its measurement. 172   Prisms 100 x 100 x 350 mm in size were cast in the laboratory using plain concrete and FRC (cellulose 0.2%) to replicate the conditions at the Aquatic Center.  The target strength of the matrix was 40 MPa and an additional dosage of 1.25 l/m3 of superplasticizer was added to the FRC mix to simulate field conditions. Four sides of the prism were smoothened using a sander (Figure 7.19) and were placed on steel roller supports (Figure 7.20) to avoid the influence of the underlying base on UPV. Direct, semi-direct, and indirect UPV measurements were conducted on lab specimens (Figure 7.21).  Average, standard deviation, maximum, and minimum UPV values are given in Table 7.6.   Figure 7.19 - Smoothening Using a Sander  173  Transmitting probe Receiving probe Supports  Figure 7.20 - Beam Arrangement for UPV Measurements (schematic)  The top surface of the specimens was uneven and thus resulted in variability in readings.  It is clear that cellulose fibers attenuate ultrasonic pulses through concrete, which is in line with field observations.  The exact reason for this attenuation could not be identified, however, it is hypothesized that this might have been due to a combination of the fibers themselves, change in moisture content, porosity, and matrix quality. Standard deviation for indirect measurements was similar in field and lab tests, 14% and 12%, respectively. The average of indirect UPV measurements through FRC was 7% lower than plain concrete in the laboratory when compared to a drop of 17% observed in the field.  Formation of micro-cracks is known to increase material attenuation and alter the ultrasonic wave speed (Young-Chul and Kundu 2001). Laboratory specimens were not subjected to any loading before the tests, hence no formation of micro-cracks was expected.  Therefore, reduction in UPV in FRC was due to fibers alone and not due to formation of any micro-cracks due to loading.  Hence using the hypothesis that fibers 174  attenuated UPV on-site as well, no conclusions with regards to cracking of overlays could be drawn.   Figure 7.21 - Indirect UPV Measurements on Prisms   Lab Resistivity Results  Specimens described above, were tested at an age of 28 days in moist surface-dry condition.  Resistivity values were recorded on all four longitudinal sides (Figure 7.22) and are presented in Table 7.7.  175   Figure 7.22 - Lab Resistivity Measurements  Resistivity of plain concrete and FRC was 23 and 22.2 kΩ-cm, respectively. These values are not comparable to that recorded in the field due to different moisture conditions and very different age of concrete.  Since the lab specimens had much higher moisture content, resistivity was significantly lower than the field values. The effect of fiber on resistivity in the lab could not be clearly established.  176  Table 7.6 - Lab UPV values for Plain Concrete and FRC Prisms UPV (m/sec) Average Standard Deviation Concrete type Overall Direct Semi- direct Indirect* Overall Direct Semi- direct Indirect* Max Min Plain 4564 4812 4375 4487 438 278 11.6 514 5073 3382 FRC 4274 4574 4111 4164 416 250 7.8 465.8 4834 3175 * values for casting face not included  Table 7.7 - Lab Resistivity Measurements Resistivity values (kΩ-cm) Specimen 1 Specimen 2 Specimen 3 Material Side 1 Side 2 Side 3 Side 4 Side 1 Side 2 Side 3 Side 4 Side 1 Side 2 Side 3 Side 4 Average Standard Deviation Plain concrete 19.9 23.4 22.9 22 22.9 23.3 22 22.6 20.4 25.1 28.3 22.5 23 2.16 FRC 22.6 23.3 23.6 21.7 20 22.8 19.5 19.8 25.1 21.3 25.5 21.5 22.2 1.96  177 7.3 Demonstration Project at ChemBioE Building, UBC 7.3.1 Project Details A 6.6 m x 6 m concrete slab was cast for the loading dock of the ChemBioE building located at 2360 East Mall at the University of British Columbia, Vancouver. Asphalt finish was upgraded to a 150 mm thick concrete slab-on-grade (with no steel reinforcement) for the loading bay of the building, which is expected to experience significant loads during its service life.  As opposed to the Aquatic center project, since there were no reinforcing bars in the slabs, the effect of shrinkage on concrete alone could be studied.  The slab was subdivided into five sections; four of size 3.3 m x 2 m and one 6.6 m x 2 m in size (Figure 7.23).  In two of the five sections (P1 and P4), control (plain concrete) and high volume fly-ash concrete were placed. The other three were reinforced with the synthetic fibers to study the effect of the following: low fiber dosage, high fiber dosage and high fiber dosage in a joint free slab.  Slab F5 was the only slab without a joint over a span of 6.0 m.  178  Type F2, High Dosage FRC (Vf = 0.5%, 4.5 kg/m3) Type P1, Control (32 MPa) Type F3, Low Dosage FRC (Vf = 0.33%, 3.0 kg/m3) Type P4, High Volume Fly-Ash (Type CI 40%) Type F5, Joint-free Floor Technology (Vf = 0.5%, 4.5 kg/m3) Figure 7.23 - Schematic of Concrete Placements  7.3.2 Embedded Sensors Electrical and optical embedded strain sensors were used in this investigation to monitor the development of early age shrinkage strains in concrete. Commercially available vinyl coated electrical strain gauges similar to the ones used at the Aquatic Center (Figure. 7.4) were used.  The optical sensors consisted of a fiber bragg grating (FBG) installed at mid-length of a Glass Fiber Reinforced Polymer (GFRP) rebar (Figure. 7.24) and finally armoured with a rubber tape.  The modulus of elasticity of the GFRP 179 rebar was 42 GPa and that of the electrical sensor as reported by the manufacturer was 2.75 GPa.  A GFRP rebar was chosen to mount the FBG sensor due to its lower elastic modulus and higher resistance to corrosion when compared to steel rebars.  A single mode fiber optic patch cable with a FC/APC (angle polished connector) was used to connect the fiber optic sensors to the data acquisition system.  Due to the very sensitive nature of the fiber optic cables, the embedded portion of optical cables and electrical wires were protected using vinyl tubing. A fiber optic sensor is used for precise measurements and is based on a Fabry- Pérot interferometer.  The part that senses the strain change consists of two parallel mirrors (semi-reflective) placed perpendicular to the longitudinal direction of the optical fiber.  This part is also known as the Fabry-Pérot cavity.  The light is originated by a LED source and it travels through the fiber cable until it reaches the cable end.  An interference created by the reflections is monitored and a signal is sent back to the read- out unit.  The device correlates the change in reflections over a certain length, called cavity length.  Since the Fabry-Pérot cavity is bonded to the materials subjected to strain, its length varies with the applied strain field.  In the readout unit each value of the Fabry- Pérot cavity length is associated with a certain strain value.   180 a) Installation of Fiber Optic Cable on GFRP Rebar b) Fiber Optic Cable on Rebar with Armour Tape and Angle Polished Connector Figure 7.24 - A Typical Optical Sensor   Strain gauges were mounted on specially designed chairs (Figure 7.25), 25 mm (1”) below the top finished surface and equidistant from the slab edges. Detailed location of the strain sensors is shown in the schematic in Figure 7.26. Every placement contained two electrical (long and short direction) and one optical strain sensor along the short direction.  Two temperature sensors were also installed alongside sensors ‘O1’ and ‘E4’ in placements P1 and P4 respectively, to monitor temperature changes. This allowed correcting strain data for any thermal strains.  181  (a) (b)  c) Figure 7.25 - Strain Gauges (a) Typical Details (specially designed chairs), (b) Chairs before Concrete Placement, (c) Partially Embedded Gauges during Concrete Placement    182  F5 F2 P4 F3 P1  Figure 7.26 - Schematic of Sensor Locations (P1 and P4 are control and fly ash placements respectively; and F2, F3, and F5 contain Novomesh 950) [Mix design details in Table 7.8]  7.3.3 Mix Proportions and Placement Details  Mix design for various concrete mixes batched at a ready-mix plant (Figure 7.27) is given in Table 7.8. As indicated in this table, 40% cement was replaced by fly ash in the mix used for test slab P4.  Mix proportion for all other placements was similar and polypropylene fibers were added in the concrete for slabs F2, F3, and F5.  Fibers used in the study were produced by “Propex Concrete Systems” and consisted of a blend of 183 macro and micro polypropylene fibers (Figure 7.28).  Properties of this fiber (PF9) have been described in Table 3.2 in Chapter 3.  The fibers were pre-mixed at the ready-mix plant as recommended by the manufacturer. Details of the fiber dosage are given in Table 7.9 below.   Figure 7.27 - Ready-Mix Plant, Richmond, B.C. Figure 7.28 - Blended Polypropylene Fibers           184  Table 7.8 - Concrete Mix Design Concrete Placements P1, F2, F3, F5 P4 Strength, MPa 32  @ 28 days 32  @ 56 days Class (CSA 23.1, concrete exposure class) C2 C2 Cement Type GU, kg 340 230 Fly Ash, Type CI, kg 0 150 20mm Agg, kg 705 705 14mm Agg, kg 425 425 Sand (SSD), kg 685 615 Water Reducer, Grace WRDA 64 300* ml/100 300 ml/100 Water, litres 155 145 Target Slump +/- 20 mm 70 70 Air, % 5-8 5-8 Maximum W/CM Ratio 0.45 0.45 * Dosage was adjusted by the ready-mix supplier for FRC mixes to achieve the target slump     185    Table 7.9 - Fiber Dosage and Concrete Placement Details Placement # Notation Admixture Dosage Concrete type 1 P1 None 32 MPa Plain Concrete (Control), no fly-ash 2 F2 Novomesh 950 Fiber (4.5 kg/m3) 32 MPa, no fly-ash 3 F3 Novomesh 950 Fiber (3.0 kg/ m3) 32 MPa, no fly-ash 4 P4 Fly-Ash (40% cement replacement) 32 MPa with fly-ash 5 F5 Novomesh 950 Fiber (4.5 kg/m3) to study the effect of no joints 32 MPa, no fly-ash   Four different types of concrete mixes described in Table 7.9 above, were poured over a period of 3 days.  Since there were no reinforcing bars in the slabs, no external vibration was required to place the concrete.  Concrete was finished (Figure 7.29) and protected by a plastic sheet soon after finishing was completed for at least three days. 186 Even though this study involved measuring strain development due to plastic shrinkage, the slabs were protected to avoid excessive loss of moisture simulate real site conditions.  a) Embedded Strain Gauges (electrical and optical) b) Concrete Pour and Finishing Operation Figure 7.29 - Concrete Pour on Site  7.3.4 Strain Monitoring  The data acquisition (DAQ) system was secured on-site, which included a system (manufactured by “Iders Inc.”) for gathering data from optical sensors and another custom designed at UBC to log data from the electrical sensors (Figure 7.30). The Iders optical DAQ system is designed to monitor strain from FBG sensors embedded within a structure.  The advantage of the system is that it monitors all FBG channels on a common time base, which simplifies processing of data. The unit used in this study had a total of eight channels. A custom software was used to acquire data onto a PC. As opposed to this, the system for electrical sensors was custom designed at UBC to connect up to 13 channels (details provided in section 7.2.2). The equipment was configured to enable data monitoring over the Internet using “WebDaq” (Web-based data acquisition). 187   Computer for data logging    System for Optical sensors    System for electrical strain gauges including WebDaq Figure 7.30 - On-site Data Acquisition System  7.3.5 Results and Discussion 7.3.5.1 Strain Readings  Data was collected from optical and electrical sensors at 100 Hz and 0.01 Hz respectively. Enormous amount of data acquired from the optical data acquisition system was analyzed and condensed by averaging data for every 50 seconds. Since, the gauges reposition and attain a new initial strain immediately after concrete placement, the strain values recorded from all sensors were “zeroed” after final set to have a fair comparison between different strain sensors. 188 All electrical sensors recorded a tensile strain during the first few hours during casting and finishing operations when the gauges readjusted and realigned in concrete. After this initial tensile loading, compressive strains developed during the first 24 hrs. Compressive strains in the FRC placements at the end of 24 hrs were much lower than those in the other placements.  Placement F3 with low volume of fibers developed about 100 µ strains after 24 hours, which was similar to that in placements P1 and P4. Data were acquired in two sets.  One prior to setting (this included the initial temperature rise) and one after setting (this included acquisition after proper bonding). At the start of both these stages, all sensors were “zeroed” to have a fair comparison between different strain sensors. Further, in the case of the electrical sensors, the signals from the gauges in longitudinal and transverse direction in each placement were very similar, indicating that strain in all directions of the placement was similar. For a simplified comparison, strains recorded from the two electrical sensors in each placement were averaged (except E5-T). 189  -300 -200 -100 0 100 200 300 5 7 9 11 13 15 17 19 21 23 25 Time (hrs) M ic ro  S tr ai n F3 F2 P1 P4 F5  Figure 7.32 - Average Post-Setting Signals from the Electrical Sensors.  -300 -200 -100 0 100 200 300 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Time (hrs) M ic ro  S tr ai n P4 P1 F3  Figure 7.31 - Presetting Averaged Signals from the Electrical Sensors (Tensile +ve Compression –ve).  190  Figure 7.31 shows the presetting strains recorded with the electrical sensors. Notice that these sensors recorded a tensile strain during the first few hours after casting. This is expected as the temperature of concrete will increase due to the generation of the heat of hydration. Figure 7.32 shows the average post-setting signals from the electrical sensors. Notice that the sensors in F2 and F5 (FRC with a higher fiber dosage rate) recorded low values of strains implying a reduced potential for cracking. P1, P4 and F3, on the other hand, recorded greater strains than those in F2 and F5 placements, and these strains were compressive in nature.  As will be noted later, unfortunately, the compressive strength of concrete in placement F3 was abnormally low which may have led to greater bleed rates, lower elastic modulus and hence the higher strains noted in this placement.  The strain results of placement F3 may also indicate that a fiber dosage of 3.0 kg/m3 (Vf = 0.33%) is not sufficient to bring about a notable reduction in the measured strains.  This volume fraction of fiber was in excess of what was required to produce “crack-free” overlays described in Chapter 6.  As described earlier, even though the FRC slabs were not expected to crack, the strains were monitored because the reduction in strain signifies a lower probability or potential for cracking.  The similar magnitude of strains in P1 and P4 is somewhat surprising as one would expect the placement with high fly-ash content (P4) to record somewhat higher strains due to reduced initial bleeding and greater potential for subsequent water loss.  The reduced nature of strains in F2 and F5 is encouraging. Similar strains recorded in F2 and F5 is also encouraging as it implies that a joint-free placement (F5) does not necessarily increase the risk of cracking. 191   Table 7.10 - Post-Setting Strain Increase (Electrical Sensors) Placement Short direction Long direction Average of Strain in long and short direction P1 -134 -146 -140 F2 -61 -24 -42.5 F3 -12 -37 -24.5 P4 -146 -122 -134 F5 -12 - -12    Strains recorded during the first 24 hrs including the presetting strains are presented in Figure 7.33 (a).  Post-setting strain increase for the first 24 hours was calculated and is given in Table 7.10 for electrical sensors. The post-setting data from the optical sensors are plotted in Figure 7.33 (b) after zeroing the strains.  As seen in this figure, the strains recorded in all optical gauges were low when compared to that measured using the electrical strain gauges.  To protect the FBG mounted on the GFRP rebar, a rubber tape was used which unfortunately appears to have prevented the development of bond between the sensors and concrete.  192 -500 -400 -300 -200 -100 0 100 200 300 400 500 0 5 10 15 20 Time (Hrs) M ic ro  S tr ai n F3 P1 P4 F5 F2  Figure 7.33 (a)- Average Signals from the Optical Sensors  193   7.3.5.2 Material Tests  As indicated before, cylinders and prisms were cast (Figure 7.34) to determine the compressive strength (ASTM C 39) and flexural toughness (ASTM C 1609) values. The cylinders were protected using plastic sheets and left on-site for the first day to expose them to the same exposure conditions as the slabs.  They were later moved to the laboratory.  Results from the compression tests are presented in Table 7.11.  Notice that the control (P1) and high volume fly ash mixes (P4) had similar compressive strengths (22 and 24 MPa respectively) at 56 days.  The mixes with a high fiber dosage had a compressive strength of about 32 MPa which was the only mix that came close to the target strength of 32 MPa (see Table 7.11). The mix with low fiber dosage (F3)  Figure 7.33 (b)- Average Post-Setting Signals from the Optical Sensors  -300 -200 -100 0 100 200 300 5 10 15 20 25 Time (Hrs) M ic ro  S tr ai n P1 P4 F5 F2 F3 194 developed a very low strength only of 17 MPa. This led to the high strain readings as was described previously.   Figure 7.34 - Cylinder and Prisms for Material Tests  In Table 7.11, the results from the Schmidt Hammer Tests are also given. These test are described later.  Table 7.11 – Compression Test Results Placement Compressive Strengths at 56 days (ASTM C39) Ratios of Compressive Strengths from Schmidt Hammer Tests (described later) P1 (Control) 22.3 1.00 P4 (High Volume Fly ash) 24.2 0.94 F2 (High fiber volume) 32.4 2.15 F3 (Low fiber volume) 17.0 0.61 F5 (High fiber volume) 32.4 2.07  195  The average flexural toughness (load-displacement) curves for low fiber dosage (F3) and high fiber dosage (F2 and F5) fiber concretes are given in Figure 7.35.  The load vs. deflection curves for the three specimens tested for F3 and F5 are presented in Figure 7.36 and 7.37 respectively.  Notice that the curves for F5 are very consistent.  On the contrary, for the mix F3, due to the lower fiber dosage, higher instability was noticed after the peak load.  Due to a test error, data for F3-2 was lost beyond 1 mm of specimen deflection.  These curves were analyzed as per the post crack strength (PCS) method (Banthia and Trottier 1995) and according to ASTM C 1609 (2006).  Results for the PCS analysis are given in Figure 7.38.  Toughness values calculated according to ASTM C 1609 are presented in Table 7.12. For a beam tested on a span L, with a width and depth, respectively, of b and h, the post-crack strength PCSm at a deflection of L/m is given by (Figure 7.39):                                                                                                                                       (7.2)    The terms used in the above equation are described in Figure 7.39.  PCSm has units of stress and at a deflection equal to δpeak , the PCSm value would coincide with the MOR of the beam. Specimens cast using concrete mix type F2/F5 with a high fiber dosage had significantly higher toughness than the mix with a lower fiber dosage (F3). The un-reinforced concrete specimens (P1 and P4), as expected, did not have any post 2 , )( )( hb m L LE PCS peak mpost m ××− ×= δ 196 crack toughness and hence only peak load values could be recorded.  The peak loads for P1 and P4 were 9.1 kN and 11.8 kN respectively with an expected linear loading curve (as observed for the other mixes in Figures 7.35 and 7.36).    Figure 7.35 - Average Load vs. Deflection Plots for FRC (F2/F5 and F3) as per C1609   0 2 4 6 8 10 12 14 0.0 0.5 1.0 1.5 2.0 Displacement (mm) Lo ad  (k N ) F5 F3 197 0 2 4 6 8 10 12 14 16 0 0.5 1 1.5 2 Displacement (mm) Lo ad  (k N ) F3-3 F3-2 F3-1  Figure 7.36 -  Load vs. Deflection Plots for Three FRC (F3) Specimens Tested as Per C1609  0 2 4 6 8 10 12 14 16 0 0.5 1 1.5 2 Displacement (mm) L oa d  ( kN ) F5-2 F5-1 F5-3  Figure 7.37 -  Load vs. Deflection Plots for Three FRC (F5) Specimens Tested as Per C1609 198  0 1 2 3 4 5 6 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 L/m P C S (M P a) Average F5 Average F3  Figure 7.38 - Post Crack Strength (PCS) Plots         Figure 7.39 - Calculation of PCSm Values [See Equation 7.2] L o a d Deflection Pre-peak Energy(Epre) Post-peak Energy (Epost,m) Epost,m δpeak l/m 2 , )( )( hb m L LE PCS peak mpost m ××− ×= δ 199   Table 7.12 - Toughness Values Using ASTM C 1609  Mix F3 (Low Fiber Dosage) Mix F5 (High Fiber Dosage) Specimen 1 2 3 Average 1 2 3 Average P1 (kN) 11.53 11.04 8.96 10.51 13.86 11.16 13.00 12.68 f1 (MPa) 3.46 3.31 2.69 3.15 4.16 3.35 3.90 3.80 δ1 (mm) 0.017 0.055 0.044 0.039 0.039 0.022 0.056 0.04 P100,0.5 (kN) 2.45 1.84 1.96 2.09 4.91 4.91 5.52 5.11 f100,0.5 (MPa) 0.736 0.552 0.589 0.626 1.472 1.472 1.656 1.53 P100,2.0 (kN) 1.96 - 3.80 2.88* 4.54 5.64 4.91 5.03 f100,2.0 (MPa) 0.589 - 1.141 0.865* 1.362 1.693 1.472 1.51 T100,2.0 (Joules) 5.40 - 6.16 5.78* 10.67 11.30 10.69 10.88 * Average of two values    In summary, increase in fiber volume from 3.0 kg/m3 to 4.5 kg/m3 significantly improved the toughness of concrete. Peak load and post crack strength (at 2 mm deflection) for F2/F5 (high volume fiber concrete) when compared to F3 (low fiber volume concrete) was 150% and 115% higher respectively.  It should also be noted that the compressive strength of F3 as placed at site was 48% lower when compared to F5.   200  7.3.5.3 Non Destructive Tests (NDTs) As in the case of the Aquatic Center project, to further study the performance of various slabs (Figure 7.40), Ultrasonic pulse velocity (UPV), electrical resistivity, and schmidt rebound measurements were conducted.  The slab-on-grade was divided into a grid of size 0.3 x 0.25 m.            Figure 7.40 - Slabs-On-Grade before Performing NDTs  7.3.5.3.1 Schmidt Hammer Measurements  The procedure described earlier in section 7.2.4.2 was used to measure Schmidt rebound numbers for the slab-on-grade at 56 days.  Results are given in Figure 7.41. Note that in the case of the Schmidt hammer, due to a number of intervening  F2 201 uncertainties, the measured values represent not so much the absolute values of compressive strength but probable values and only provide a comparative assessment from one placement to another.  Since the rebound numbers were being compared for different concrete slabs, to keep the effect of moisture and temperature the same, readings were taken on the same day within a span of a couple of hours.  On a comparative basis, the relative values of the compressive strengths measured using the Schmidt hammer correlated well with the results obtained from ACTM C39 compression tests as indicated in Table 7.11.  00. 511. 51. 9 2. 252. 753. 253. 754. 14. 555. 56 0 1.2 2.4 3.6 4 0 5 10 15 20 25 30 35 MPa X - Direction (m) F2 F5  P1 P4 F3   Figure 7.41 - On-Site Schmidt Hammer Results 202  7.3.5.3.2 Ultrasonic Pulse Velocity A PUNDIT was used to measure pulse velocities of the concrete placements indirectly by placing the probes on the surface of concrete at a known distance (Figure 7.42). The procedure previously described in section 7.2.4.1 was used.   Figure 7.42 - UPV Measurements Using PUNDIT   Velocities between different points on the grid were computed; average, minimum, and maximum values are presented in Table 7.13. Results clearly indicated that velocities for FRC slabs F2 and F3 (except for joint-free FRC slab F5) were higher than un-reinforced concrete placements. Even though F2 and F5 were cast using the same concrete, significantly lower velocities were recorded for slab F5; suggesting some effect of not having a joint on UPV values.  Further investigation is required to make any conclusive comments.  203   Table 7.13 - Averaged UPV Results Velocity (m/s) P1 F2 F3 P4 F5 Average 2400 2899 2754 2291 2122 Max 3080 3709 3341 3158 2604 Min 1656 1948 1953 1880 1634   7.3.5.3.3 Resistivity  The method described in section 7.2.4.3 earlier was used to measure the resistivity of the various slabs. Average, maximum, and minimum resistivity values recorded between the grid points are presented in Table 7.14. It is believed that the moisture content of all the slabs was high due to significant rainfall a few days prior to testing. Average resistivity values recorded for all slabs were lower than 1.34KΩ·cm and values were very similar for different concrete types. In the context of the joint-free floor, this meant that there was no adverse effect of increasing the joint spacing.       204   Table 7.14 – Onsite Resistivity Results Resistivity Values (kΩ·cm) Slab Designation Average Resistance (Ω) Maximum Minimum Average F5 41.50 1.46 1.17 1.30 P4 41.68 1.43 1.23 1.31 F3 41.00 1.49 1.17 1.29 F2 40.23 1.48 1.19 1.26 P1 42.62 1.66 1.21 1.34   7.3.5.4  Laboratory Strain Tests To develop an understanding of the general behavior and consistency of strain readings from electrical and optical strain gauges, further laboratory tests were conducted.  Electrical and optical strain gauges similar to the ones used on site were mounted side by side (Figure 7.43) in three different specimens to maintain similar conditions.  Note these strain gauges were different from the ones described in section 4.4.2 of Chapter 4.  The shrinkage measurement technique described in Chapter 4 using an unreinforced mortar overlay was used. The strain gauges were 10 mm from the top of an overlay and strains were monitored for the first three hours in the environmental chamber at an elevated temperature of 50°C. The aim of the exercise was to study the consistency of results and hence the test was terminated at about three hours after casting 205 to recover the strain gauges.  The initial setting time of the cement used was 90 minutes, however at elevated temperatures of 50°C, the initial setting time was expected to reduce significantly.  Rivera-Villarreal R. (1986) tested different types of cement mortars at temperatures ranging from 15 to 38°C and reported a drop of 60% or more in the initial setting time.   Figure 7.43 - Electrical and Optical Strain Gauges (Laboratory Test)  -400 -200 0 200 400 600 800 0 0.5 1 1.5 2 2.5 3 3.5 Time (hrs) M ic ro  S tr ai n Specimen 1 Specimen 2 Specimen 3  Figure 7.44 - Electrical Sensor Data (Laboratory) 206  -400 -200 0 200 400 600 800 0 0.5 1 1.5 2 2.5 3 3.5 Time (hrs) M ic ro  S tr ai n Specimen 2 Specimen 1 Specimen 3  Figure 7.45 - Optical Sensor Data (Laboratory)  Initially, the gauges along with the substrate base were placed inside the environmental chamber to receive the overlay.  Substrate base and the gauges gained temperature inside the environmental chamber, while the overlay mix was only at room temperature.  Strain readings recorded for electrical and optical strain gauges are presented in Figures 7.44 and 7.45 respectively. In general, both types of gauges demonstrated a similar trend; initial compression due to a drop in temperature while pouring the mix and then tensile strain due to a combination of high temperature in the environmental chamber and the heat of hydration. The data clearly supported previously made assumptions that strain in optical sensors did not fully develop due to lack of bond. Electrical strain gauges due to their better bond showed larger strain change during the first three hours and the trend from three different specimens was very consistent. On the 207 contrary, even though the optical gauges were initially consistent, they later recorded a strain change of only 200 microns and moreover the results were not consistent. This corroborates well with the on-site strain data presented earlier.   7.4 Concluding Remarks Based on the field investigations described and results presented in this chapter, the following conclusions were drawn: • Embedded strain gauges, both electrical and optical can be used to monitor the early-age behavior of unreinforced and FRC slabs and overlays.  The monitoring can be done over the internet for sites at remote locations.  Further research is needed to address the issues associated with bonding of FBGs when used as an embedded strain gauge in concrete. • Strain development during early-ages can help predict the probability and potential for plastic shrinkage cracking.  This is especially useful when no visible cracking is expected due to restrained plastic shrinkage.  Use of cellulose and polypropylene fibers indicated a reduction in measured strain during early-ages, hence indicating a much reduced cracking potential. • High volume fiber dosage used for the joint-free slab-on-grade indicated very similar strains when compared to a smaller slab with the same FRC mix.  This indicated that the risk of cracking in the joint free slab was not higher than the smaller slab. 208 • The strain values recorded in the slab cast using a high volume fly-ash (40% cement replacement) concrete mix were similar to that for the control mix (no fly- ash), indicating that the cracking potential at early-ages was no different for the slab with high volume fly-ash. • Strain monitoring can be complemented with NDTs for complete health monitoring of structures.  In addition to using a rebound hammer, indirect UPV measurements were useful in slabs to qualitatively study the performance of the slabs.  Laboratory confirmatory tests indicated that cellulose fibers attenuate pulse velocities.  This should be considered during field investigation. 209 CHAPTER 8 - ALTERNATE PREDICTION TECHNIQUES  8.1 Introduction  In order to effectively predict the behavior of cementitious materials at early ages using alternate techniques and relate the findings to the previously described test method in Chapter 4, it was essential to develop an understanding of the basic material properties at early ages.  These findings are briefly described in sections 8.2 and 8.3.  These data were further used to develop and study the effectiveness of alternate methods of predicting early age properties of cement-based composites.  8.2 Compressive Strength at Early Ages  Compressive strength of cementitious materials is the fundamental mechanical property that dictates the performance of the material under early age loading conditions. To evaluate early compressive strength, the control mix M8 (w/c = 0.5, s/c = 0.5, and cement = 1200 kg/m3) described in Chapter 5 was cast into 100 mm cube molds. Compressive strength of other fiber-reinforced mixes described in Chapter 6 was expected to be similar, since all mixes had identical cement and water contents and nominal dosage of fibers.  Test samples were cast and placed in the environmental chamber under the same temperature and relative humidity regime as described in Chapter 4.  As previously described in Chapters 4 and 5, time to first crack for most mixes during restrained plastic shrinkage conditions was between 2-3 hrs.  Hence, compressive tests were performed at approximately 3 hrs from the time of casting.  These tests were performed using a closed-loop highly sensitive Instron Universal Testing 210 Machine (model 8802 with a capacity of 250 kN).  Load values along with cross-head displacements were recorded.  Approximate elastic modulus (E) values along with the peak load and strength values are given in Table 8.1a.  Compressive strength of the material was about 0.3 MPa and the elastic modulus was 2.14 MPa after 3 hrs of exposure in the environment chamber at 50°C. As expected, these values steadily increased with time indicating increased hydration and formation of a stronger microstructural skeleton. The peak stress increased by 48% in 25 minutes.  These values, even though approximate, were critical in understanding the strength gain at early ages at elevated temperatures.  Table 8.1a - Compressive Strength and Elastic Modulus at Early Ages Time since casting (min) Peak Load (kN) Peak Stress (MPa) Elastic Modulus “E” (MPa) 180 2.94 0.29 2.14 190 3.43 0.34 2.14 200 4.05 0.4 5.83 205 4.29 0.43 5.88   8.3 Uniaxial Tensile Properties 8.3.1 Introduction 211  As previously described, under restrained shrinkage conditions, cement based composites crack when the tensile stresses exceed the tensile strength of the material. The test described in this section was used to determine the enhancement in tensile strength of the material (if any) due to the addition of fibers at early ages and also to determine the effectiveness of the fibers to carry load after the first crack.  This would indicate the ability of fibers to inhibit crack growth under restrained shrinkage conditions.  8.3.2 Mixes and Test Set-up  Different mixes investigated under uniaxial tension are described in Table 8.1b. A fiber dosage (Vf) of 0.1% by volume was selected for all fiber reinforced mixes since 0.1% was the fiber dosage for some of the most effective fiber types that resulted in no shrinkage cracks.  Briquette specimens were cast using specially prepared molds (Figure 8.1) and were subjected to the environmental conditions described in Chapter 4 to simulate conditions similar to those experienced by the shrinking overlay in a typical shrinkage test.  Specimen size and geometry are shown in Figure 8.2.  Briquette specimens were 25.4 x 25.4 mm in size at the critical cross section and the failure zone was 6 mm long. These briquettes have been previously used by Banthia et al. (1994 and 1995) to perform tensile impact tests and they have also described the test set-up in detail.  It should be noted that these tests require careful consideration of eccentricities that may be introduced during testing and the instability that is observed after the peak load.   212    Table 8.1b - Mixes Investigated under Early-Age Uniaxial Tension Fiber type Mix Designation Admixtures Vf (%) Control C1 - - Admixture C2 Fly ash (20% cement replacement) - Grace MicroFiber (PF8) FT1 Fibermesh 150 (PF1) FT2 Fibermesh 300 (PF2) FT3  Polypropylene Saint-Gobain (AntiCrack) (GF1) FT4 Glass Buckeye (UltraFiber 500) (CF1) FT5 Weyerhaeuser (CF5) FT6  Cellulose   0.1%     Figure 8.1 - Molds for Casting Briquette Specimens    213  Figure 8.2 - Briquette Dimensions for Uniaxial Tensile Test [Adopted from Banthia et al. (1994)]   Tensile tests were conducted between 120 and 155 minutes from the time of casting corresponding to the time when restrained shrinkage cracks were expected to appear.  Test equipment is shown in Figure 8.3.  The equipment consisted of a fixed clamp and a moving clamp, which was connected to an LVDT and a load cell.  The load cell had a capacity of 267 N and was sensitive enough to capture small load changes during testing.  The deflection rate was controlled by changing the speed of the motor that powered the moving clamp.  The initial speed was set at 0.3 mm/min till the first crack appeared (corresponding to the peak load) and was increased to 2.3 mm/min after the peak. The data was collected using a data acquisition system.  214 LVDT Moving Clamp  Figure 8.3 (a)   Fixed Clamp Crack Location  b) Figure 8.3 - Test Set-up a) Overall Arrangement for Uniaxial Tension Tests, b) Test Specimen During Testing  8.3.3 Test Results  The load vs. deflection graphs for different mixes are plotted in Figures 8.4 to 8.10 and the test results have been summarized in Table 8.2.  A typical failed specimen is 215 shown in Figure 8.11, where pulled-out and fractured fibers across the crack can be noticed.  The unsupported length at the start of the test was 6 mm and hence calculation of percent elongation was based on this length.  The load vs. deflection graphs for the five control samples tested are plotted in Figure 8.4.  Some deviation in the peak load was observed depending on the location of the crack and the actual area of the cracked section.  Peak tensile stress was calculated based on measured areas and was found to be between 0.04 and 0.08 MPa.  As expected no post peak ductility was observed due to lack of fibers.  Test results for mix containing fly-ash are plotted in Figure 8.5, as expected the average tensile strength was 0.037 MPa, lower than that for control.  In most specimens, some minimal initial non-linear strain was measured probably due to initial displacement between the specimen and the grips.  0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) C1-1 C1-2 C1-3 C1-4 C1-5  Figure 8.4 - Load vs. Deflection for Mix C1 (Control) 216 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) C2-1 C2-2 C2-3  Figure 8.5 - Load vs. Deflection for Mix C2 (20% Fly-Ash)   Load vs. deflection plots for the fiber reinforced mixes are plotted in Figures 8.6 to 8.10.  As described before, all the fiber reinforced mixes contained 0.1% volume of fibers and the fiber types used in various mixes are given in Table 8.1.  It is evident that fibers added post peak load carrying capacity even at early ages suggesting their effectiveness in controlling propagation of cracks after their initiation.  Tests were continued to a deflection, until fibers lost their load transfer capacity.  Glass fibers had exceptional load carrying capacity with a plateau near the peak load indicating multiple cracking, strain hardening, and high effectiveness of fibers in controlling propagation of cracks.  217 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT1-1 FT1-2 FT1-3  Figure 8.6 - Load vs. Deflection for Mix FT1 (Polypropylene Fibers, PF8)  0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT2-1 FT2-2 FT2-3  Figure 8.7 - Load vs. Deflection for Mix FT2 (Polypropylene Fibers, PF1) 218 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT3-1 FT3-2 FT3-3  Figure 8.8a - Load vs. Deflection for Mix F3 (Polypropylene Fibers, PF2) 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT4-1 FT4-2 FT4-3  Figure 8.8b - Load vs. Deflection for Mix F4 (Glass Fibers, GF1) 219 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT5-1 FT5-2 FT5-3  Figure 8.9 - Load vs. Deflection for Mix F5 (Cellulose Fibers, CF1) 0 2 4 6 8 10 12 0 2 4 6 8 10 Deflection (mm) Lo ad  (l bs ) FT6-1 FT6-2 FT6-3  Figure 8.10 - Load vs. Deflection for Mix F6 (Cellulose Fibers, CF5)  220  Figure 8.11 - A Tested Specimen Showing Polypropylene Fibers Across the Crack   8.3.4 Analysis and Discussion The load vs. deflection data was further analyzed and the following aspects were calculated: • Peak tensile stress based on the maximum recorded load and the measured area at the failed cross-section.  This stress represents the limit, which when exceeded will cause the first crack to appear in a shrinking overlay, • Stress at 2 mm:  Representing residual strength of the material at a typically observed crack width during a shrinkage test. • Elongation at Peak:  Representing the strain capacity of the material at the peak load even before a crack has appeared in the material.  This value would indicate the effect of fibers and admixtures on the material’s intrinsic strain capacity before cracking.  221 The above-mentioned properties of the various mixes are presented in Table 8.2 and graphically shown in Figure 8.12 (a) where tensile stress at peak, at 2 mm deflection, and elongation at peak load in percentage are plotted.  Peak Stress: Average peak tensile stress for mix C2 was 0.037 MPa, 62% lower when compared to the control mix C1 due to the lower matrix strength.  This was expected as the mix C2 contained 20% less cement, which was replaced with fly-ash that reduced the early-age matrix strength.  Peak stress values for all fiber reinforced mixes were comparable to the control mix C1 (between 0.061 and 0.071 MPa) with the exception of mix FT1.  Stress Carried at 2 mm: Un-reinforced mixes had no residual strength at 2 mm deflection.  Mixes (FT5 and FT6) containing cellulose fibers had marginal residual strength.  Polypropylene fibers retained between 15 and 25% and the mixes ranked (in decreasing order):  FT1 (fiber type PF8), FT2 (fiber type PF1), and FT3 (fiber type PF2).  In terms of residual strength at 2 mm, the most effective fibers were glass fibers (GF1) that retained 85% of its peak tensile strength even at a deflection of 2 mm.  This trend completely corroborated with the effectiveness of fibers in controlling crack area (at Vf = 0.1%) described in Chapter 6, hence indicating that residual strength under uniaxial tension is a good indicator of FRC performance under restrained plastic shrinkage conditions. For the same volume fraction, polypropylene fibers, PF1, which is a finer fiber as compared to PF2, indicated a much higher residual strength at 2 mm deflection and a higher percentage elongation at the peak. 222  Elongation at Peak: Percentage elongation at peak was calculated based on an unsupported specimen length of 6 mm.  As seen in Figure 8.12 (a), the elongation at peak for all unreinforced specimens was similar at about 13.5%.  All fibers increased the strain capacity before reaching the peak load when compared to the unreinforced mixes. Marginal increase was observed with mixes FT1, FT3, FT5, and FT6.  Polypropylene fiber mix FT2 and glass fiber mix FT4, on the other hand, increased the elongation at peak by more than 76% when compared to the control. 223 Table 8.2 - Uni-Axial Test Data Material Mix Designation Testing Age (min) Cracked area (mm2) Average Peak Load (N) Average Tensile Stress (MPa) Stress at 2 mm (MPa) Elongation at Peak (mm) % Elongation at Peak Control C1 130 614 37.3 0.060 0 0.8 13.5 Fly ash (20% cement replacement) C2 132 672 25.0 0.037 0 0.8 13.6 Polypropylene (Grace) FT1 137 648 31.2 0.048 0.015 1.0 16.4 Polypropylene (FM 150) FT2 140 648 42.9 0.067 0.010 1.5 25.0 Polypropylene (FM 300) FT3 135 624 40.1 0.064 0.004 0.9 15.8 Glass (Saint-Gobain) FT4 137 624 38.4 0.061 0.052 1.4 23.8 Cellulose (Buckeye) FT5 140 624 44.4 0.071 0.002 1.0 16.0 Cellulose (Weyerhaeuser) FT6 130 624 39.2 0.063 0.003 0.9 15.3   224 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 C1 C2 FT1 (Poly, PF8) FT2 (Poly, PF1) FT3 (Poly, PF2) FT4 (Glass, GF1) FT5 (Cellulose, CF1) FT6 (Cellulose, CF5) Mix Type S t r e s s  ( M P a ) 0 5 10 15 20 25 30 E l o n g a t i o n  a t  P e a k  ( % ) Tensile Stress at Peak Tensile Stress Carried at 2mm Elongation at Peak  Figure 8.12 (a) - Stress and Elongation for Different Mixes 225  Discussion: This study clearly indicated that certain fibers are effective in marginally increasing the tensile strength of the material itself and some increase both the residual strength and elongation capacity at early ages.  Fly-ash reduces the early-age tensile strength of the material and would thus be expected to increase restrained plastic shrinkage cracking as described in Chapter 6.  In general, glass fibers were found to be the most effective followed by polypropylene fibers and the cellulose fibers. These findings are in complete agreement with the effect of these fibers on shrinkage cracking as described in Chapter 6.  In Figure 8.12 (b) a plot showing comparison of tensile stresses carried at 2 mm (recorded from briquette specimens) and shrinkage crack area recorded from testing for mixes containing identical fiber dosage is made.  It can be clearly seen that for most fibers, an increase in the tensile strength at 2 mm reduced the shrinkage cracking potential.  Amongst the various properties investigated, residual strength was the most effective parameter in indicating effectiveness of fibers to mitigate restrained plastic shrinkage cracking.  The uniaxial tests described above are one of the first attempts in quantifying the performance of early age properties of fiber-reinforced cement-based materials especially at early ages, after they have been exposed to severe conditions of drying that is conducive to shrinkage cracking.  Limited studies report work similar to the one described above (Komlos 1979, Kovler 1994, Kovler & Bentur 1997).  226 0.000 0.010 0.020 0.030 0.040 0.050 0.060 F4 (Glass, GF1) F1 (Poly, PF8) F2 (Poly, PF1) F3 (Poly, PF2) F5 (Cellulose, CF1) F6 (Cellulose, CF5) C1 Mix Type Te ns ile  S tr es s (M Pa ) 0 50 100 150 200 250 300 350 To ta l C ra ck  A re a (m m 2 ) Tensile Stress Carried at 2mm Shrinkage Results (Total Crack Area)  Figure 8.12 (b) - Comparison of Tensile Stress at 2 mm and Shrinkage Cracking  227 8.4 Correlating Early-Age Shrinkage and Flexural Toughness 8.4.1 Introduction  Toughness of FRC is a key factor in characterizing the material’s mechanical properties and long term durability.  Similarly, plastic shrinkage tests are indicative of the performance of FRC overlays at early ages and determine the ultimate durability of the material.  Fibers in both cases enhance the performance of the material by bridging across cracks and providing crack growth resistance. One of the aims of this study was to determine the correlation between shrinkage and flexural toughness performance of fiber-reinforced cementitious composites.  A pilot study was undertaken to establish such a correlation, if it existed. Independent shrinkage and toughness tests on fiber-reinforced materials were conducted, data were analyzed and a mathematical correlation was determined.  Cellulose fiber (CF5) was selected for this pilot study with fiber volume fraction ranging between 0.2 and 0.4%.  Furthermore, it was decided that polypropylene macro fiber PF10 (1.0 kg/m3 or 0.12% by volume) be also tested to determine if the correlation obtained for the micro cellulose fiber was applicable to a macro fiber of a different material.  Properties of these fibers have been previously described in Chapter 3 (Table 3.2).  Shrinkage tests were conducted using the technique described in Chapter 4 and flexural toughness tests were conducted according to ASTM C 1399-02 (1999).  Shrinkage was measured in terms of crack area, whereas the toughness was measured as the Average Residual Strength (ARS) which is derived from load carrying capacity in the post-cracking regime at pre selected displacements.   228 8.4.2 Procedure and Results  Dosages of cellulose and polypropylene fibers used for these tests are given in Table 8.3.  For both tests four samples were tested at each of the fiber dosages.  In both tests, results were compared to an unreinforced control mix.   Table 8.3 - Fiber Dosage for Shrinkage and Toughness Tests Fiber type Fiber Content (%) Un-reinforced (Control) - Cellulose (CF5) 0.2, 0.3, 0.4 Polypropylene (PF10) 0.12   A typical 30 MPa concrete mix was used for the toughness tests and testing was conducted using a closed-loop Instron UTM machine.  Test set-up was similar to that described in Chapter 7 (section 7.3.5.2).  If the 30 MPa concrete mix were to be used for the shrinkage tests, then probably minimal or no cracking would be observed in the control and certainly no cracking would be expected in the FRC mix.  Hence, to study the effect of the fibers on reducing shrinkage induced cracking, the control mix described in Chapter 6 was used.  Shrinkage crack area and ARS values (calculated according to ASTM C 1399) are presented in Table 8.4.  ARS increased from 0.08 - 0.38 MPa when dosage of cellulose fiber increased to 0 - 0.4%.  It should be noted that since a very 229 sensitive closed-loop set-up was used for testing, a non-zero ARS was recorded for the control specimens due to aggregate interlocking.   Table 8.4 - Summary: Shrinkage Crack area and Toughness Data (ARS) Fiber Type Fiber Content (%) Average Residual Strength (MPa) Crack Area (mm2) Control mix 0 0.08 225 0.2 0.2 113 0.3 0.24 24 Cellulose (CF 5)  0.4 0.38 0 Polypropylene (PF 10) 0.12 0.91 183   ARS for the mix containing polypropylene fiber (PF10) was the highest (0.91 MPa).  Results from the shrinkage tests indicated that crack area reduced from 225 mm2 for control to no cracking at 0.4% fiber volume of cellulose fiber. ARS for micro cellulose fibers at 0.4% was 0.38 MPa, 58% lower than that for macro polypropylene fiber mix containing only 0.12% fibers by volume.  Furthermore, assuming a linear reduction in crack area for cellulose fibers with fiber volume fraction, for a 1.0 kg/m3 (Vf = 0.12%) fiber dosage, one can expect a similar crack area as the polypropylene fiber. Thus indicating that macro polypropylene fibers when compared to cellulose fibers are equally effective in reducing crack area. 230  8.4.3 Analysis and Discussion ARS and shrinkage data are plotted in Figure 8.13.  The plot suggests that as the crack area increases the ARS decreases. There does not appear to be a direct correlation between ARS and crack area in a general sense, but for a given fiber type (cellulose for example) there appears to be a correlation.  Exponential statistical regression was used to determine the mathematical correlation between toughness and crack area for cellulose fiber.  This equation is given below:  )0065.0(3444.0 CAT −=    Equation 8.1  where, T  is average residual strength (ARS), measured in MPa, and AC is crack area in (mm2)   231 ARS = 0.3444e-0.0065Ac (R 2 = 0.9334) 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 250 CRACK AREA (mm2) A R S (M Pa ) 0.4% CF5 0.2%  CF5 0.3%  CF5 Control 0.12% PF10  Figure 8.13 - Shrinkage and ARS Correlation  Notice that although the relationship between the parameters of shrinkage and toughness is only an estimate, the model fitted well to the test data with an R2 value of 0.933.  In other words, crack areas determined using the shrinkage technique described in Chapter 4 can also be predicted by measuring ARS values for FRCs in question. However, it should be noted that this regression model is valid only for cellulose fibers and the function or correlation may not be valid for other fibers.     232 8.5 Factorial Design 8.5.1 Introduction In addition to various models described in Chapter 2 that predict early-age strain, several models exist that predict early age characteristics of cement composites subjected to shrinkage (Mabrouk et al. 2004, Van Zijl et al. 2001, Sanjuan & Moragues 1994, Toledo Filho & Sanjuan 1999, and Radocea 1994).  Scientific experiments in the past have been modeled by keeping all parameters constant and varying one variable at a time. This traditional method is very time consuming and does not incorporate the correlation between variables.  In this section, a statistical prediction model based on factorial design is described.  This is a structured method that provides an efficient way of studying properties of a material that depends on several factors (Sanjuan & Moragues 1994 and Toledo Filho & Sanjuan 1999) as in the case of cement composites. Properties such as w/c, s/c, fiber content, etc. are considered as variables and their effect on restrained shrinkage cracking is investigated. Factorial design enables establishing correlation between different variables.  In this investigation, multiple linear regression analysis was used to develop a mathematical model that predicts restrained shrinkage cracking of a representative mix.   This study was aimed at understanding the effect of different factors/parameters on cracking of cement composites under restrained shrinking conditions. The factors considered in the investigation are: • w/c ratio • s/c ratio 233 • Volumetric percentage of fiber (Vf)  A factorial design analysis was conducted to determine the relative importance of different parameters and their interaction on the cracking area and the average crack width.  8.5.2 Definition of Factorial Design  Factorial design is a form of mathematical analysis which enables one to study the influence of several parameters with only a few tests. Contrary to “one factor at a time” approach, in which factors are varied one at a time, in the factorial design approach, much fewer tests need to be carried out to determine the effect of parameters studied, and to determine the interaction between them.  The interaction between parameters cannot be evaluated using “one factor at a time” approach because we assume that the factors act additively, which is not always the case. In this study, three factors mentioned earlier are investigated with an upper “+” and lower level “-” for each factor.  This method is described in detail by Box et al. (2005) and can be further explained using the following example. Example  In order to describe this method, let’s consider the effect of three factors: A, B and C on given variable “y”.  Let’s also assume that each factor has two levels : - and +. The levels signify a selected lower and an upper bound value. If using factorial design, we need 2 × 2 × 2 = 23 = 8 tests. On the contrary, based on “one factor at a time” approach, 8 tests for each factor would be required, resulting in a total of 234 24 tests.  According to factorial design, the 8 tests are the different arrangements of the 3 factors, which are shown in Table 8.5.  Table 8.5 - Data from a 23 Factorial Design Factors Variable Test A B C y 1 - - - y1 2 + - - y2 3 - + - y3 4 + + - y4 5 - - + y5 6 + - + y6 7 - + + y7 8 + + + y8  8.5.3 Calculation of “Effect” and “Interaction” 8.5.3.1 Calculation of Main Effect  “Effect of a factor” in factorial design refers to the change in “y” as that factor changes from “-” to “+.”  8.5.3.2 Calculation of the Main Effect of Variable B  The main effect of variable B is also known as “B main effect.”  In this example there are two tests for each arrangement of A and C: one with the lower level of B and 235 one with the upper. Since the two tests differ only in the B factor, the B effect will be the difference between the + and the − level. Then, the B total effect is the average of the four pairs of tests.  Table 8.6 - Effect of Variable B or “B Main Effect” Difference between + and − levels of B A C y3-y1 - - y4-y2 + - y7-y5 - + y8-y6 + +   Hence, B’s main effect can be calculated as:  4 6521 4 8743 yyyyyyyyBEffect +++−+++=     Equation 8.2   As seen from Equation 8.2, the main effect of B is basically the difference between two averages: the average response for the + level and the average response for the − level.  −+ −= yyBEffect        Equation 8.3 236   8.5.3.3 Calculation of B × C Interaction  This interaction is calculated by the difference between the average B effect at C’s upper level and the average B effect at C’s lower level (and vice versa).  Half of the difference is termed as B x C interaction. From Table 8.5,  4 6543 4 8721 yyyyyyyyCB +++−+++=×    Equation 8.4   8.5.3.4 Calculation of A × B × C Interaction  Two values of the B × C interaction can be calculated for the experiment, one for each level of A. Half the difference between these two values is defined as the three factor interaction (A × B × C interaction) and given by Equation 8.5.  4 7641 4 8532 yyyyyyyyCBA +++−+++=××  Equation 8.5   8.5.4 Geometric Representation  The 23 factorial design can be represented as a cube in which each corner is a test. The three factors A, B and C form the axes of the space (Figure 8.14). 237  A B C y1 Y4 Y2 Y5 Y3 Y7 Y8 Y6  Figure 8.14 - Geometric Representation of a 23 Factorial Design    Previously described effects and interactions are graphically shown in Figure 8.15.  Main effects may be viewed as a contrast between observations on parallel faces of the cube; the interaction is a contrast between results on two diagonal planes and the three factor interaction is a contrast between the two tetrahedrals.  238   a) b) c)  Figure 8.15 - a) B Main Effect, b) B x C Interaction, c) A x B x C Interaction 239  8.5.5 Calculation of Standard Error  Standard error is calculated as described by Box et al. (2005).  First of all, the variance is calculated for all tests using the following equation:  ∑ = = n i in 1 22 1 σσ         Equation 8.6  Where, σi2 is the variance for one test and is defined by Equation 8.7  ∑ −−= m i ii yym 22 )( 1 1σ       Equation 8.7  where, m is the number of runs for one test  Standard error is then calculated using Equation 8.8  σ n e 2=          Equation 8.8 Where n is the total number of runs  According to the technique described in Chapter 4, three runs (m=3) are made for each test, hence in this case for the 8 tests,  n = 24.    240 8.5.6 Application of Factorial Design to Representative Test Results  The factorial design technique described above was applied to some test results described in Chapters 5 and 6 to develop equations that describe the effect of mix proportion and fiber inclusion on restrained shrinkage cracking.  It was evident from the test results in Chapters 5 and 6 that w/c, s/c, and addition of fibers have an important effect on cracking.  Factorial design was used to quantify this effect and determine the interaction that existed between the factors affecting shrinkage cracking.  The control mix described in Chapter 5 with w/c ratio and s/c ratio both equal to 0.5 was used and values surrounding this mix were studied.  Polypropylene fiber (PF8) with a fiber length of 20 mm was selected for this study and commonly used range of 0 to 0.066% by volume of fibers was considered.  Fiber properties have been described in Chapter 3.  The parameters, their levels and the resulting mixes/tests forming the corners of a 23 factorial cube are given in Table 8.7.  Test results for the first four mixes have already been described in Chapter 5, remaining fiber reinforced mixes were tested to complete the test matrix for the factorial design.         241 Table 8.7 - Mixes for 23 Factorial Design Parameters Mix Designation w/c s/c Vf (%) Factorial Levels (w/c, s/c, Vf) 4-4 0.4 0.4 0 (-, -, -) 4-6 0.4 0.6 0 (-, +, -) 6-4 0.6 0.4 0 (+, -, -) 6-6 0.6 0.6 0 (+, +, -) 4-4-6 0.4 0.4 0.066 (-, -, +) 4-6-6 0.4 0.6 0.066 (-, +, +) 6-4-6 0.6 0.4 0.066 (+, -, +) 6-6-6 0.6 0.6 0.066 (+, +, +)   Crack analysis was conducted as described in Chapter 4 and crack area and width along with the standard deviation according to equation 8.7 were calculated for all mixes. These results are shown in Table 8.8.  242  Table 8.8 - Crack Analysis Results Standard Deviation (%) Mix Designation w/c s/c Vf (%) Crack Area (mm²) Average Crack Width (mm) Maximum Crack Width (mm) Crack Area Average Crack Width Maximum Crack Width 4-4 0.4 0.4 0 285 1.21 1.93 28.9 34.8 21.8 4-6 0.4 0.6 0 260 1.20 2.15 31.1 5.3 14.0 6-4 0.6 0.4 0 499 1.11 2.20 6.5 18.7 3.7 6-6 0.6 0.6 0 732 1.79 4.03 25.4 30.2 23.0 4-4-6 0.4 0.4 0.066 111 0.64 1.07 10.2 28.6 23.1 4-6-6 0.4 0.6 0.066 9 0.09 0.12 173.2 173.2 173.2 6-4-6 0.6 0.4 0.066 619 1.92 3.62 18.3 19.0 20.8 6-6-6 0.6 0.6 0.066 519 1.48 2.75 27.6 36.7 3.4   243  Main and interaction effects were calculated for the parameters and are presented graphically in Figure 8.16.  The figure shows the effect and interaction of parameters on crack area and crack width.  In this case, the “effect” of a parameter is defined as the difference in crack area or width for lower and upper values of the given parameter.   Main Effect -200 -100 0 100 200 300 400 500 average s/c w/c %PP s/c-w/c s/c-%PP w/c-%PP s/c-w/c- %PP standard error C ra ck in g A re a (m m ²) -0.55 -0.35 -0.15 0.05 0.25 0.45 0.65 0.85 1.05 1.25 W id th  (m m ) Cracking area Average width  Figure 8.16 - Main and Interaction Effects on Cracking Area and Average Width    In Figure 8.16, the first two bars represent average values of crack width and crack area of the data represented in the 23 factorial cube.  The last two bars represent the standard error in the test data for crack width and crack area, which can be used to study 244 the significance of the test results.  The other bars in the figure indicate the effect of parameters on crack area and crack width, which are similar except in the case of s/c. This means that crack area is generally proportional to the crack width.  Keeping the average and standard error in mind, w/c followed by fiber dosage (%PP) are the most important factors that affect cracking; the range of s/c studied did not affect the results considerably.  A small fiber dosage of 0.066% reduces crack area by 129 mm2.  As far as the interactions are concerned, s/c − %PP and w/c − %PP were the only two significant interactions when compared to the standard error in the data; fibers interacted with both w/c and s/c to reduce crack area and width.   8.5.7 Mathematical Model  Factorial design method is further used to predict plastic shrinkage cracking in cementitious composites using a multi-exponential regression model (as shown in Equation 8.9)  PPcwcs mmmby %3 / 2 / 1 ×××=      Equation 8.9  where, y= “Crack Area” or “Crack Width,”  b, m1, m2, and m3 are all coefficients and their values are presented in Table 8.10,  s/c is the sand to cement ratio,  w/c is the water to cement ratio, and 245  %PP is the percentage of the volume fraction of polypropylene fiber  The coefficients in Equation 8.9 were evaluated by least-square regression analysis based on the data presented in Table 8.8 and are presented in Table 8.9.  Note that values have been rounded off to three decimal places.   Table 8.9 - Coefficients from Regression Analysis Coefficient b m1 m2 m3 Crack Area 18.079 0.053 9550.356 5.84E-08 Crack Width 0.2809 0.117 182.905 2.266E-05    The first eight mixes were studied for factorial design and are described in Table 8.9.  To assess the effectiveness of the model in predicting crack area and crack width for mixes not used in formulating the equation, test results of additional mixes reported in Chapters 5 and 6 with other combinations of w/c, s/c, and Vf were used.  Some of the mixes reinforced with polypropylene fibers were used, as this equation was developed for synthetic fiber only.  These mixes are highlighted in grey in Table 8.10.  Test data for these mixes along with results predicted by the model are presented in Figures 8.17 and 8.18 and are summarized in Table 8.10.  Standard error was calculated according to Equation 8.8.  246  Table 8.10 - Comparison of Predicted Data to Test Results Test Results Prediction Model Mix Designation w/c s/c Vf (%) Crack Area (Standard Error) in mm² Average Crack Width (Standard Error) in mm Crack Area (mm2) Crack Width (mm) 4-4 0.4 0.4 0 285 (83) 1.21 (0.42) 218 0.96 4-6 0.4 0.6 0 260 (81) 1.20 (0.06) 121 0.62 6-4 0.6 0.4 0 499 (33) 1.11 (0.21) 1361 2.71 6-6 0.6 0.6 0 732 (186) 1.79 (0.54) 756 1.77 4-4-6 0.4 0.4 0.066 111 (11) 0.64 (0.18) 73 0.47 4-6-6 0.4 0.6 0.066 9 (16) 0.09 (0.16) 40 0.31 6-4-6 0.6 0.4 0.066 619 (113) 1.92 (0.37) 453 1.34 6-6-6 0.6 0.6 0.066 519 (143) 1.48 (0.54) 252 0.87 5-4 0.5 0.4 0 628 (96) 1.64 (0.33) 544 1.61 5-6 0.5 0.6 0 450 (148) 1.62 (0.16) 302 1.05 5-5-6 0.5 0.5 0.066 212 (35) 0.82 (0.06) 135 0.64 5-5-3 0.5 0.5 0.033 157 (74) 1.21 (0.34) 234 0.91 5-5-10 0.5 0.5 0.1 10 (18) 0.14 (0.25) 77 0.45 5-5 0.5 0.5 0 264 (33) 2.18 (0.85) 406 1.30    247 Cracking area -100 100 300 500 700 900 1100 1300 4-4 4-6 6-4 6-6 4-4 -6 4-6 -6 6-4 -6 6-6 -6 5-4 5-6 5-5 -6 5-5 -3 5-5 -10 5-5 Mix Designation (w/c-s/c-%PP) C ra ck in g ar ea  (m m ²) Results prediction  Figure 8.17 - Comparison of Proposed Model to Test Data (Crack Area)  Average width -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4-4 4-6 6-4 6-6 4-4 -6 4-6 -6 6-4 -6 6-6 -6 5-4 5-6 5-5 -6 5-5 -3 5-5 -10 5-5 Mix Designation (w/c-s/c-%PP) A ve ra ge  w id th  (m m ) Results Prediction  Figure 8.18 - Comparison of Proposed Model to Test Data (Crack Width) 248   Figure 8.17 and 8.18 compare the laboratory test results with the predictions from the regression model for crack area and width respectively.  Standard error for each test is also included in the plots.  The predicted crack area and width values corroborated well with test data for a wide range of mixes both fiber reinforced and unreinforced mixes except for mix 6-4, where the predicted crack area and width exceeded the test results. The calculated percentage error in the predicted values compared to test result error is presented in Table 8.11.  The average percentage error in crack area and crack width was 108% and 62% respectively, which indicates good corroboration with test results considering that only a few mixes resulted in high error because the crack area and width values were very low.  Fiber reinforced mixes 4-6-6 and 5-5-10 developed very small crack area and width, hence the percentage error was higher than 200% for both area and width; average percentage error neglecting these values dropped to 46% and 36% respectively for crack area and crack width.  The average percentage error further dropped to 34% and 26% for crack area and width respectively by considering data for mix 6-4 as an outlier.        249 Table 8.11 - Comparison of Predicted Values and Test Results Error (%) Mix Designation Crack Area Crack Width 4-4 24 21 4-6 53 48 6-4 173 145 6-6 3 1 5-4 13 2 5-5 54 40 5-6 33 35 4-4-6 34 26 4-6-6 330 229 6-4-6 27 30 6-6-6 52 41 5-5-3 49 24 5-5-6 36 22 5-5-10 633 209   Discussion: As was discussed earlier, the main effect of w/c ratio and fibers was dominant on the crack characteristics and hence the effect of these factors on crack area and width evaluated using the proposed model was compared to actual test data.  The effect of w/c ratio alone was evaluated by averaging the results of mixes with the same w/c but different s/c (except for w/c = 0.6, where predicted crack area and width for mix 6-4 were very large and were ignored).  Similarly, effect of fiber was studied for mixes by fixing the w/c and s/c ratios.  These results are presented in Figures 8.19 and 8.20.  It is clear that the model is quite effective in predicting the crack area and width for cement- 250 based composites and can be useful in predicting the effect of w/c and fibers on crack characteristics.  It should be also noted that some variability in the actual test results is introduced due to the fact that some of the data were adopted from previously completed test programs.  0 100 200 300 400 500 600 700 800 0.400 0.500 0.600 w/c ratio C ra ck  a re a (m m 2 ) 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 C ra ck  w id th  (m m ) Test crack area Predicted area Predicted width Test crack width  Figure 8.19 - Effect of w/c Ratio on Predicted and Actual Test Data  251 0 50 100 150 200 250 300 350 400 450 0.000 0.033 0.066 0.100 %PP C ra ck  a re a (m m 2 ) 0.00 0.50 1.00 1.50 2.00 2.50 C ra ck  w id th  (m m ) Test crack area Predicted area Predicted width Test crack width  Figure 8.20 - Effect of %PP on Predicted and Actual Test Data 252 8.6 Energy-Based (Fracture) Model 8.6.1 Introduction Sections 8.2 and 8.3 described testing the properties of mortar mixes at early-age and the recording of compressive strength and tensile strength data.  In sections 8.4 and 8.5, techniques such as correlating early-age shrinkage and flexural toughness, and factorial design that predict the behavior of mixes at early-age under restrained shrinkage conditions were described.  In this section, an energy-based fracture model is described.  Fracture energy curves were first developed based on the uniaxial tensile tests performed at early age. Maximum crack widths predicted by this model were further compared to that recorded during shrinkage tests.  8.6.2 Model Theory  Hillerborg (Hillerborg et al. 1976) proposed the Fictitious Crack Model (FCM) based on the Dugdale-Barenblatt plastic crack-tip zone in the 1970s.  The FCM is still widely used for analyzing crack growth in cementitious composites. According to the FCM, as soon as the tensile strength of the material is exceeded, the crack does not become unstable or stress free immediately, but instead the stress carried decreases with increase in the crack width.  Stresses near the crack tip and their transfer in the cohesive zone or process zone is shown in Figure 8.21.  253  (a) (b) Figure 8.21 - Stress Transfer in the Cohesive Zone in Front of a Stress-Free Crack: (a) Stresses Near the Crack Tip, (b) Microcrack Zones (adopted from Van Mier J.G. M., 1996)   According to Hillerborg et al. 1976, in an ideal situation, the parameters for the model should be derived from a stable uniaxial tensile test.  However, according to FCM, the stress-deformation diagram for the material can also be split into pre-peak and post- peak regions (as shown in Figure 8.22).   254 (a) (b) Figure 8.22 - Fracturing in Tensile Bar, (a) Distributed Cracking Before Peak, (b) Localized Crack in a Narrow Zone (reproduced from Van Mier J.G. M., 1996)   8.6.3 Proposed Model  The uniaxial tensile data recorded for mixes at early ages was used to develop a fracture-based energy model.  As observed from the tensile tests, a more or less expected linear stress-strain behavior was observed before cracking.  After cracking, since the use of strain is not valid anymore, the data was expressed as stress vs. crack width (this data was presented in the previous sections).  The area under the post-peak portion of the plot is equal to the fracture energy (Gf) and was calculated using Equation 8.10.  dwwG cw f )( 0 ∫= σ         Equation 8.10 255 where, ‘σ(w)’ is the stress, which is a function of the crack width ‘w’, ‘wc’ is the critical or maximum crack width where ‘σ(w)’ = 0 (traction free). These variables are shown in Figure 8.22 (b).  ‘σ (w)’ described in Equation 8.10, is usually determined from uniaxial tests or from a third-point bend test and functions are used to fit the data.  In the proposed model, the data from the uniaxial tests was used and a power law adopted from one proposed by Foote et.al (1986) that best described these data was used. This law is described in Equation 8.11.  n ct w w w f ]1[ −=σ       Equation 8.11  where, ‘ft’ is the peak tensile strength, which can be taken to be 0.075 MPa from tests and ‘n’ is an index number that produces different softening curves (n = 1 indicates linear softening) for different fibers when used at different volume fractions.  8.6.4 Predicted Values  The model was used to determine the maximum crack widths for three very different fiber types- glass, polypropylene (manufactured by Grace), and Cellulose (manufactured by Weyerhaeuser).  Initially, a fiber volume equal to 0.1% was selected since test data were available for this fiber dosage.  The value of ‘wc’ recorded during the tensile tests was used and for different values of ‘σ’, the corresponding crack width was 256 determined using Equation 8.11.  The values of ‘n’ and ‘wc’ used for the model are given in Table 8.12.  The value of ‘wc’ was obtained from the tests and ranged from 2 mm to 10 mm.  Since ‘wc’ represents the width at which the crack becomes traction free, it was assumed that the value of ‘wc’ was only a function of the fiber material and geometry and not a function of the fiber volume fraction.  Hence, the material properties such as ‘ft’ and ‘wc’ were further used to predict the fracture energies for other volume fractions.  Figure 8.23 (a) shows a schematic indicating softening of the fracture curves with reduction in fiber volume fraction or the increase in the value of factor ‘n’.  The value of ‘n’ between 1 and 36.5 gave the best suited softening curves. These softening curves are shown in Figure 8.23 (b) for glass fibers and 0.1% and 0.4% of polypropylene and cellulose fibers. Figure 8.23 (c) shows the softening curves for other volume fractions of cellulose and polypropylene fibers.  The insert in Figure 8.23 (c) is a magnified view of the initial portion of the curves clearly indicating that the fracture softening curves of the fibers at the selected dosages are very similar.    257        Crack-opening w ft wc S tr es s Linear Softening <Vf or >n  Figure 8.23 (a) - Schematic Showing the Effect of Reduction in Fiber Volume Fraction on the Softening Curves    258 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 1 2 3 4 5 6 7 8 9 10  Maximum Crack Width , w (mm) Po st  P ea k  St re ss  (M Pa ) Glass (0.033%) Glass (0.066‐0.1%) Cellulose (0.4%) Poly (0.1%)  Figure 8.23 (b) - Plot of Post-Peak Stress vs. Maximum Crack Width (Glass, Cellulose, and Polypropylene Fibers)  0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0  Maximum Crack Width , w (mm) Po st  P ea k  St re ss  (M Pa ) Poly (0.033%) Poly (0.066%) Cellulose (0.1%) Cellulose (0.2%) Cellulose (0.3%) 0.04 0.04 0.04 0.05 0.05 0.05 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20  Figure 8.23 (c) - Plot of Post-Peak Stress vs. Maximum Crack Width (Cellulose and Polypropylene Fibers) with an Insert Showing Magnified View  259  Table 8.12 - Comparison of Maximum Crack Widths (in mm)- Model vs. Test Results  0% Vf = 0.033% Vf = 0.066% Vf = 0.1% Vf = 0.2% Vf = 0.3% Vf = 0.4% Model  wc  n w n w n w n w n w n w Glass 5.5 19 1.02 1 0.27 1 0.27 - - - - - - Poly (Grace) 10 36.5 1.82 36.30 1.41 23 0.43 - - - - - - Cell (Weyer) 2 2.77   - - - - 6.505 1.31 6.5 1.23 6.3 0.8 5.8 0.61 Testing Glass 1.02 0.11 0 - - - Poly (Grace) 1.94 1.45 0.43 - - - Cell (Weyer) n/a   2.77   n/a   - n/a   - n/a   1.31 n/a   2.74  0.79 n/a   0.60    From the uniaxial tensile tests, the average tensile strength of the mixes was 0.075 MPa.  Average pre-peak energy for the mixes was determined and was equal to 0.02 N/mm.  Since this accumulated energy is released at the peak (when the matrix cracks), the energy is dissipated across the produced crack and results in stress softening and an increase in the crack width.  Using this understanding of cementitious composites, fracture energy (from equation 8.10) was equated to the unstable energy (0.02 N/mm) and the corresponding crack width required to dissipate this energy was calculated. These predicted widths “w” are shown in Table 8.12.  For computational purposes, the fracture energy curve was subdivided into several elements and a spreadsheet was developed to perform the analysis. Since this model well represents what the top layer of the shrinking overlay experiences, the predicted values were compared to the maximum crack widths recorded from restrained shrinkage tests.  These values are also reported in 260 Table 8.12 and a comparison shown graphically in Figure 8.24 (a).  Even though the number of cracks in a specimen under shrinkage conditions (reported in Chapter 4) could be more than one, formation of a single crack was assumed in the model.  This meant that the strained system went from a loaded system to an unloaded system by releasing the strain at a single crack location as opposed to multiple locations as shown in Figure 8.24 (b).  Overall, a very good agreement with the test results was observed.  For polypropylene fibers, using the proposed values of factor ‘n’, the predicted values of crack widths were very close to the test values at 0.033%, 0.066%, and 0.1%.  The test values for cellulose fibers at small dosages were not very conclusive as described in the earlier chapters.  However, the crack widths predicted by the model indicated a gradual reduction in crack width with an increase in fiber volume fraction, which is expected.  At fiber dosages of 0.3% and 0.4% of cellulose fibers, predicted values fitted quite well with the test data.  Finally for glass fibers, the fit with the test data was reasonable for fiber volume fraction less than 0.033%, but the model was limited to predicting 0.27 mm as the lowest crack width for higher volume fractions.  This is attributed to the fact that this model is based on the premise that a softening would be observed after the peak as opposed to some hardening that was observed for glass fibers during testing. Also, the glass fibers were exceptional in reducing the crack widths, and in fact reduced the crack width to almost zero at a fiber dosage of 0.066%; these values did not corroborate very well with the values predicted by the model.  261 0 0.5 1 1.5 2 2.5 3 3.5 0.00% 0.05% 0.10% 0.15% 0.20% 0.25% 0.30% 0.35% 0.40% Fiber Volume Fraction (%) M ax im um  C ra ck  W id th  (m m ) Poly (Tests) Poly (Model) Cellulose (Tests) Cellulose (Model) Glass (Tests) Glass (Model)  Figure 8.24 (a) - Comparison of Predicted Crack Widths with Test Data  Multiple Cracking in Specimens (Average Crack Width) Assumed Single Crack in model (Maximum Crack Width)  Figure 8.24 (b) - Comparison of Cracking in Specimens with that Assumed in the Model  262  The model was extended to predict crack widths at other fiber dosages using the previously defined values of ‘wc’, ‘ft’, and the unstable ‘pre-peak’ energy.  The values of ‘n’ used for the prediction are reported in Table 8.12.  Glass fibers were the most effective fibers in reducing shrinkage cracking.  This was not completely captured by the model.  For the given parameters, the model predicted a minimum crack width of 0.27 mm, which was much larger than that observed at 0.066% and 0.1%.  A good fit was however, observed for a fiber dosage of 0.033%. Reasonable agreement with the test data was observed for the cellulose fibers, especially at 0.3 and 0.4% fiber dosage.  This data is given in Table 8.12 and also plotted in Figure 8.24.  Finally, to predict the results from the model for other fiber volume fractions, the relationship between ‘n’ and ‘Vf’ was plotted (Figure 8.25).    263  Figure 8.25 - Relationship Between Fiber Volume Fraction and Factor ‘n’   8.6.5 Use of Model (Summary) and Limitations The model described above is proposed for polypropylene fiber (Grace) and Glass fibers within the fiber dosage range of 0-0.1% and for cellulose fibers for the range 0.1 to 0.4%. Since glass fibers were very effective in reducing plastic shrinkage cracking even at very small dosages, the model only predicts a crack width of 0.27 mm for fiber volume fraction larger than 0.033%. This implies that for glass fibers there is initially some strain hardening followed by strain softening.  In order to determine the crack width, the first step is to use the following equations to determine the power softening factor ‘n’ for the given volume fraction. 264  Polypropylene Fibers 24.2 + )(V56299 +)(V106- =n f 2 f 7 ×××      Equation 8.12  Glass Fibers 37+)(V-54545=n f×        Equation 8.13 Note: Equation 8.13 is limited to a Vf range of 0 to 0.033%  Cellulose Fibers 6.24 + )(V387 +)(V1.23x10- =n f 2 f 5 ××      Equation 8.14   Next, the values of ‘wc’ and ‘ft’ that are dependent on the matrix and fiber characteristics (obtained from tensile tests) are used as input parameters and substituted in Equation 8.11.  For different values of ‘σ(w)’ ranging between ‘ft’ and zero, the values of crack width ‘w’ can be obtained, hence predicting a fracture energy curve.  The area under this stress-crack opening curve is determined in small increments and cumulative energy is also determined.  The final input in this model is the initial energy that can be calculated as the area under the pre-peak portion of a stress-deformation curve in tension (0.02 N/mm, in this study).  The final step is to determine when the cumulative energy after the peak stress equals this unstable initial energy before peak stress.  The corresponding crack width predicts the expected average crack width determined from the restrained plastic shrinkage test described in Chapter 4.  265  Limitations of the Model: The model assumes a stable uniaxial test, which means that immediately after the peak load, the release of the energy is gradual.  This is not exactly what was observed during the uniaxial tests since an-open-loop test set-up was used.  The model is limited to the fiber types and volume fractions described above but can be extended to other fiber types as well. Using the parameters described above, the model is limited to predicting crack widths larger than 0.27 mm only, for glass fibers.  This is because the model is based on a fracture softening behavior of FRC, which did not completely apply to glass fibers that resulted in some strain hardening after the peak load.  266 CHAPTER 9 - CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH  9.1 Conclusions  1.  Test Technique The UBC bonded overlay technique was further developed to study the effectiveness of fibers, influence of mixture proportion (w/c and s/c ratio), additive such as fly-ash and admixture such as shrinkage reducing admixtures in controlling plastic shrinkage cracking. Significant instrumentation was also added to the environmental chamber as listed below: • Probes to continuously monitor temperature and relative humidity inside and outside the chamber and to control temperature inside the chamber using a data logging system. • A precise weighing balance to measure loss of moisture from a shrinking overlay. • Non-contact laser sensor (not used in this study) to measure settlement and thermocouple to measure temperature inside a representative sample. • Strain gauges, both electrical and optical (FBG mounted on a GFRP rebar) to measure strain from six specimens. • A camera frame and other accessories for image analysis and to monitor crack growth during the first 24 hrs.   267  2.  Mixture Proportion • The effect of w/c and s/c on evaporation rate, area under the time-temperature curve (termed “heat evolution”), time to crack, total crack area, average crack width, and maximum crack width was studied in this research. For the range of s/c investigated in this project, a definite trend could not be established. • An increase in w/c from 0.35 to 0.5, decreased the total measured internal heat of hydration by 6%.  Evaporation rate before and after demolding increased by 69% and 47% respectively when w/c increased from 0.35 to 0.6.  This suggests that the existing ACI nomographs for predicting evaporation rate do not account for the mixture proportion, or more importantly the w/c ratio, and hence are not sufficiently accurate.  It is recommended that a fourth set of curves for different mixes be added to the existing ACI nomographs or correction factors be used to modify the evaporation predicted by the nomographs. • The average time to first and second crack to appear increased with increasing w/c for the mix proportions and test conditions evaluated.  For w/c=0.35, average crack area and average crack width were 57 mm2 and 0.59 mm respectively. These values were significantly lower than the average values recorded for mixes with w/c=0.6.  The w/c was found to be a prime factor contributing to shrinkage cracking. • To study the effect of coarse aggregates, the s/c ratio was increased from 0.5 to 0.7 with 28% well graded coarse aggregate.  Coarse aggregate at very low addition rates provided restraint and significantly reduced shrinkage cracking. 268    3.  Effect of Fibers • Fibers were highly effective in controlling plastic shrinkage cracking in concrete and reduced the total crack area and crack width. Based on fiber material, for the same volume of fibers, glass fibers followed by synthetic fibers were the most effective in inhibiting crack growth.  Glass fibers outperformed all fibers, a reduction of 97% in crack area was observed at 0.066% and cracks were completely eliminated at 0.1%. Glass and polypropylene fibers (PF8) at a dosage of 0.1% completely eliminated shrinkage cracking resulting in ‘crack-free’ overlay materials. • Among the various synthetic fibers investigated, fiber type PF8 (properties in Table 3.2) was the most effective. In general, longer and lower denier fibers were more effective in reducing crack areas and crack widths. Finally, fibrillated fibers were more effective in controlling shrinkage cracking than their comparable monofilament counterparts. • Two types of coated and uncoated cellulose (Weyerhaeuser) fibers were not very effective in controlling plastic shrinkage cracking when used at dosages less than 0.2%. However, these fibers showed reduction in crack areas and widths when fiber volume was greater than 0.3%. For these fibers, different length and coatings resulted in no evident change in crack characteristics.  Cellulose fiber (CF1) resulted in increase in cracking when fiber dosage increased from 0.1 to 0.3%.  It 269 is hypothesized that the fiber dispersion is an issue for this fiber in mortar mixes due to the absence of coarse aggregates, unless the “chips” are wet dispersed in a mix and special instructions are followed.  4. Effect of Fly-Ash and SRAs • Considering the in-batch standard deviation associated with the bonded overlay test method, the effect of Class C fly ash dosage (0 to 20%) on crack area and crack width could not be clearly established. • SRA reduced the time to first crack when compared to the control mix but resulted in similar total crack area.  However, the average crack width reduced by 46% when compared to the control.  5. Field Investigation at the Aquatic Center • Average strains recorded in plain concrete were higher than those in FRC, indicating development of lower tensile stresses and higher resistance to cracking in FRC. • Field measurements also indicated significant ultrasonic pulse attenuation in the FRC slab due to the presence of cellulose fibers in concrete; this was confirmed in the laboratory tests where a 7% reduction in UPV was recorded for FRC.  UPV measurements were also useful in identifying areas with unconformities in the field.   270  6.  Field Investigation at the Chemical and Biological Building • Internal strain developed in slabs-on-grade at early ages monitored with the aid of traditional and fiber optic sensors indicated that the strains developed in FRC placements were lower than those developed in the unreinforced placements.  For the joint-free flooring (span of 6 m) there was no increased cracking risk as long as appropriate amount of fiber reinforcement was present. • Change in strain from electrical sensors was calculated during the first 72 hours. Compressive strain in FRC placements was less than one-third the strain recorded in control placement.  This implied that in-service cracking potential in FRC placements is lower and one can expect a better long-term durability. • The strain values in the slab cast using a high volume fly-ash (40% cement replacement) concrete mix were similar to that for the control (no fly-ash), indicating that the cracking potential at early-ages for the slab with fly ash was no different than the control slab.  Thus high volume fly ash could be used without significantly affecting early-age strain. • Electrical strain gauges installed on-site were more sensitive to strain change and recorded higher strain values when compared to the traditional strain gauges. This was due to the lower elastic modulus of electrical strain gauges when compared to the ‘E’ for GFRP rebar on the which the optical sensors were mounted.  Laboratory tests confirmed that electrical strain gauges were more responsive to strain changes during early-age setting of mortar and resulted in consistent trends. 271 • It is concluded that for proper structural health monitoring, NDTs can be successfully used to corroborate and supplement findings from strain measurements.  Simple Schmidt hammer test confirmed that the strength of on- site placements was lower than the target strength for most placements.  7. Alternate Prediction Techniques • A simple relationship between the parameters of shrinkage and flexural toughness (ARS values) was developed for cellulose fiber reinforced composite. The model fitted well with the test data with an R2 value of 0.933.  This indicated that crack area determined using the shrinkage technique can also be predicted by calculating ARS values with good agreement. • Factorial design analysis confirmed that w/c ratio is the most significant factor that affects crack areas and widths.  Increase in w/c ratio increased both crack area and crack width.  A small range of s/c ratio studied, did not affect the crack characteristics significantly.  The effect of polypropylene fibers was studied that indicated that addition of fibers clearly reduced cracking.  The proposed model predicts crack area and widths and the results corroborate well with the actual test data. • Uniaxial tensile tests on young fiber reinforced mortar specimens clearly indicated that certain fibers are effective in marginally increasing the tensile strength of the material itself and some increase both the post-peak residual strength and elongation capacity.  Fly-ash reduces the early-age tensile strength of 272 the material and would thus be expected to increase restrained plastic shrinkage cracking. • From the uniaxial tests, glass fibers were found to be the most effective followed by polypropylene fibers and cellulose fibers. These findings are in agreement with the effect of these fibers on shrinkage cracking.  Amongst the various properties investigated using the uniaxial tests, residual strength was the most effective in relating effectiveness of fibers in improving uniaxial tensile properties to behavior of material under restrained plastic shrinkage conditions. • An energy-based fracture model was proposed for polypropylene fiber (PF8) and glass fibers within the fiber dosage range of 0 to 0.1% and for cellulose fibers for the volume fraction range of 0.1 to 0.4%.  A power softening fracture response was assumed and a relationship between initial energy before the peak load in uniaxial tension and after the first crack was developed.  The model predicted maximum crack widths, which except for very low volume fraction of glass fiber (where some strain hardening was observed), agreed well with test results. Equations were also proposed to determine power softening factors for any fiber dosage within the ranges specified above.   9.2 Recommendations for Further Research  In this study, basic image analysis was utilized to measure crack area during the first 24 hrs, future research can focus on use of an image analysis software that can fully automate the process of measuring crack widths. 273  Different fibers are very effective in reducing crack area and crack width, however, the influence of a combination of two or more fibers (hybrid fiber combination) on inhibiting plastic shrinkage cracks is not completely understood.  Even though a blended fiber combination was used for the ChemBioE field project, future studies should not only focus on investigating the use of combination of fibers of different sizes (both macro and micro), but also on fibers made using different materials in the laboratory.  The bonded overlay technique is designed to study the effect of adding additives and admixtures in mixes that have high content of paste and are exposed to severe environmental conditions.  Hence, more field investigations should be conducted to study the behavior of real mixes under real field conditions.  Since glass fibers were found to be the most effective in controlling shrinkage cracking (described in Chapter 6), future field applications could be considered using this fiber.  Future study can also be undertaken to study the effect of mixture proportion and fiber reinforcement under less severe conditions in the environmental chamber.  Use of optical strain gauges at the Chemical and Biological Engineering project resulted in inconsistent results due to lack of bonding between the gauges and concrete. Modified optical strain gauges where FBG is installed on a longer rebar made using lower modulus material should be used on future projects.  Different admixtures, including fibers have been identified to produce a ‘crack- free’ overlay and their field performance has also been monitored at early ages. However, since durability of concrete is an important factor, further studies should also focus on evaluating the long term permeability and durability of these materials.  Several other forms of NDTs such as penetration resistance methods, acoustic emission methods 274 and radar-based tests can be used for further investigation.  Radar-based tests are known for their capabilities to scan, inspect and analyze ground and bridge decks and according to Geophysical Survey Systems, Inc., (2004), ground penetrating radar products can be used to define areas of deterioration within concrete deck resulting from chloride-induced corrosion.  The fracture based energy model described in this project was developed for the critical time period when the shrinkage cracks develop.  Further research is needed to study the constitutive properties of the overlay material at different time intervals. 275 BIBLIOGRAPHY   AASHTO PP34. (1998). Standard Practice for Estimating the Crack Tendency of Concrete. American Association of State Highway and Transportation Officials, Washington, D.C., pp. 132. ACI 209.1R-05. (2005). Factors Affecting Shrinkage and Creep of Hardened Concrete. Technical Report by ACI Committee 209, American Concrete Institute. ACI 305R-96. (1996). Hot Weather Concreting. Manual of Concrete Practice, Part 2. Farmington Hills: American Concrete Institute. Almusallam, A. A., Maslehuddin, M.; Abdul-Waris, M.; and Khan, M. M. (1998). Effect of mix proportions on plastic shrinkage cracking of concrete in hot environments. Construction and Building Materials, 12 (6-7), 353-358. Al-Obaid, Y.F. (1989). Drying shrinkage of glass fiber reinforced concrete. Cement, Concrete and Aggregates, 11(2), 119-120. Altoubat, S.A. and Lange, D.A. (2001). Creep, Shrinkage, and Cracking of Restrained Concrete at Early Age. ACI Materials Journal, 98 (4), 323-331. ASTM C 39/C39M-05e1. (1998). Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens. American Society of Testing and Materials, 04.02, Philadelphia. ASTM C 597-97. (1998). Standard Test Method for Pulse Velocity Through Concrete. American Society of Testing and Materials, 04.02, Philadelphia, 291-293. ASTM C 805-02 (1998). Standard Test Method for Rebound Number of Hardened Concrete. American Society of Testing and Materials, 04.02, Philadelphia. ASTM C 1399. (1999). Test Method for Obtaining Average Residual-Strength of Fiber-Reinforced Concrete. American Society of Testing and Materials, Philadelphia.04.02, 677-681. ASTM C 1579 (2006). Standard Test Method for Evaluating Plastic Shrinkage Cracking of Restrained Fiber Reinforced Concrete (Using a Steel Form Insert). American Society of Testing and Materials, Philadelphia.04.02. ASTM C 1581 (2004). Standard Test Method for Determining Age at Cracking and Induced Tensile Stress Characteristics of Mortar and Concrete under Restrained Shrinkage. American Society of Testing and Materials, Philadelphia.04.02. ASTM 1609/C. (2006). Standard test method for Fiber-Reinforced Concrete (Using beam with Third-Point Loading). American Society of Testing and Materials, 04.02, Philadelphia. 276 Atis, C.D. (2003). High-volume fly ash concrete with high strength and low drying shrinkage. ASCE Journal of Materials in Civil Engineering, 15 (2), 153-156. Babaei, K., and Fouladgar, A. M. (1997). Solutions to Concrete Bridge Deck Cracking. Concrete International, 24-37. Banthia, N., Azzabi, M. and Pigeon, M. (1993). Restrained Shrinkage Cracking in Fiber Reinforced Cementitious Composites. Materials and Structures. RILEM (Paris), 26 (161). 405-413. Banthia, N., Chokri, K., Ohama, Y., and Mindess, S. (1994). Fiber-Reinforced Cement Based Composites Under Tensile Impact. Advanced Cement Based Materials, 1, 131-141. Banthia, N., Moncef, A., Chokri, K., and Sheng, J. (1995). Uniaxial tensile response of microfiber reinforced cement composites. Materials and Structures, 28, 507- 517. Banthia, N. and Campbell, K. (1998). Restrained Shrinkage Cracking in Bonded Fiber Reinforced Shotcrete, RILEM- Proc. 35, The Interfacial Transition Zone in Cementitious Composites, Eds. Katz, Bentur, Alexander and Arligui, E and F N. Spon, 216-223. Banthia, N., and Dubeau, S. (1994). Carbon and Steel Micro-Fiber Reinforced Cement Based Composites for Thin Repairs. ASCE Journal of Materials in Civil Engineering, 6 (1), 88-99. Banthia, N., Gupta, R., and Mindess, S. (2004). Developing crack resistant FRC overlay materials for repair applications. Sixth RILEM symposium on Fibre Reinforced Concretes (BEFIB 2004), Varenna-Lecco, Italy, 99-106. Banthia, N., and Nandakumar, N. (2001). Carbon Fiber Reinforced Concrete in Parking Garage Repair. In N. Banthia, K. Sakai, & O.E. Gjory (eds.). Proceedings of the third International Conference Concrete under Severe Conditions environment and loading, CONSEC'01, Vancouver, BC, Canada. The University of British Columbia, Dept. of Civil Engineering: Vancouver. 2, 1748-1760. Banthia, N. and Trottier, J.-F. (1995). Test Methods of Flexural Toughness Characterization: Some Concerns and a Proposition”, Concrete International.: Design & Construction, American Concrete Institute, 92 (1), 48-57. Banthia, N., and Yan, C. (2000). Shrinkage Cracking in Polyolefin Fiber Reinforced Concrete. ACI Materials Journal. 97 (4), 432-437. Banthia, N., YAN, C. and Mindess, S. (1996). Restrained Shrinkage Cracking in Fiber Reinforced Concrete: A Novel Test Technique. Cement and Concrete Research, 26 (1), 9-14. 277 Bao, X., Huang, C., Xeng, X., Arcand, A., and Sullivan, P. (2002). Simultaneous strain and temperature monitoring of the composite cure with a Brillouin-scattering-based distributed sensor. Optical Engineering, 41 (7), 1496-1501. Bayasi, Z., and McIntyre, M. (2002). Application of Fibrillated Polypropylene Fibers for Restraint of Plastic Shrinkage Cracking in Silica Fume Concrete. ACI Materials Journal,  99 (4), 337-344. Bentur Arnon (Editor). (2003). Early Age Cracking in Cementitious Systems. Report (Report 025) of RILEM Technical Committee 181-EAS - Early age shrinkage induced stresses and cracking in cementitious systems. Bentz, D.P., Geiker, M. R., Hansen, K. K. (2001). Shrinkage-reducing admixtures and early-age desiccation in cement pastes and mortars. Cement and Concrete Research, 31 (7), 1075-1085. Bentz, D.P. (2006a). Modeling the influence of limestone filler on cement hydration using CEMHYD3D. Cement and Concrete Composites, 28 (2), 124-129. Bentz, D.P. (2006b). Capillary porosity depercolation/repercolation in hydrating cement pastes via low-temperature calorimetry measurements and CEMHYD3D modeling. Journal of the American Ceramic Society, 89 (8), 2606-2611. Bentz, D.P. (2006c). Quantitative comparison of real and CEMHYD3D model microstructures using correlation functions. Cement and Concrete Research, 36 (2), 259- 263. Bjøntegaard Ø, Sellevold E.J. (2001). Thermal dilation and autogenous deformation. Proc. of the RILEM Int. Conf. on Early Age Cracking in Cementitious Systems, Ed. by K Kovler, A Bentur, Haifa, Israel, 63–70. Bjøntegaard Ø. (1999). Thermal Dilation and Self Dessication as Driving Forces to Self-Induced Stresses in High Performance. Ph.D. thesis, Division of Structural Engineering, The Norwegian University of Science and Technology, Norway. Bloom, R. and Bentur, A. (1995). Free and Restrained Shrinkage of Normal and High Strength Concrete. ACI Materials Journal, 92 (2), 211-217. Botsis, J., Colpo, F., and Humbert, L. (2004). Residual strain characterization using an embedded FBG sensor: Measurements and simulations. ANTEC 2004 - Annual Technical Conference Proceedings, Chicago, IL. US, 3, 3982-3986. Box G., Hunter S. and William G. (2005). Statistics for Experimenters: Design, Innovation, and Discovery. 2nd edition. Broomfield, J. P. (1997). Corrosion of steel in concrete: understanding, investigating, and repair, London; New York: E & FN Spon. 278 Byonggeon, K., Weiss, J. (2003). Using acoustic emission to quantify damage in restrained fiber-reinforced cement mortars. Cement and Concrete Research, 33 (2), 207- 214. Charron, J.-P., Marchand, J., Bissonnette, B., Gerald, B. (2001). A comparative study of phenomenological models used to describe the behaviour of concrete at any early aged. Part 1. Canadian Journal of Civil Engineering, 28 (2), 314-322. Charron, J.-P., Marchand, J., Bissonnette, B., Gerald, B. (2001). A comparative study of phenomenological models used to describe the behaviour of concrete at any early aged. Part 2. Canadian Journal of Civil Engineering, 28 (2), 323-331. Davis Inotek Instruments < http://www.davis.com/showpage.asp?L3ID=153> accessed August 9, 2004. Day Robert L. (1990). Strength, durability and creep of fly-ash concrete. Serviceability and Durability of Construction Materials - Proceedings of the First Materials Engineering Congress, Part 2 (of 2). ASCE: Boston, 864-873. Emmons, P. H., Vaysburd, A. M., & McDonald, J. E. (1993). A Rational Approach to Concrete Repairs. Concrete International: Design and Construction, 15 (9): 40-45. Folliard, K. J., and Berke, N. S. (1997). Properties of high-performance concrete containing shrinkage-reducing admixture. Cement and Concrete Research, 27 (9), 1357- 1364. Foote, R.M.L., Mai, Y.-W, and Cotterel, B. (1986). Crack growth resistance curves in strain softening materials, Journal of the Mechanics and Physics of Solids, 34 (6), 593. Fuhr, P. L., Huston, D. R. (1998). Corrosion detection in reinforced concrete roadways and bridges via embedded fiber optic sensors. Smart Materials and Structures. 7 (2), 217-28. Gao, X., He, Z., Yang, Y., Zhou, Z., and Ba, H. (2004). Distribution of shrinkage strain and induced cracks of a round restrained concrete plate at early age. Kuei Suan Jen Hsueh Pao/ Journal of the Chinese Ceramic Society, 32 (3), 334-339. Gardner, N. J., and Lockman, M. J. (2001). Design provisions for drying shrinkage and creep and normal-strength concrete. ACI Materials Journal, 98 (2), 159- 167. Geophysical Survey Systems, Inc. <http://www.geophysical.com/BridgeScanbyGSSI.pdf>, accessed July 16, 2004. Gettu, R., Roncero, J., Martin, M. (2002. Long-term behavior of concrete incorporating a shrinkage-reducing admixture.  Indian Concrete Journal, 76 (9), 586-592. 279 Gesoglu, M., Ozturan, T, and Guneyisi, E. (2006). Effects of cold-bonded fly ash aggregate properties on the shrinkage cracking of lightweight concretes. Cement and Concrete Composites, 28 (7), 598-605. Goel, R., Kumar, R., and Paul, D.K. (2007). Comparative study of various creep and shrinkage prediction models for concrete. Journal of Materials in Civil Engineering, 19 (3), 249-260. Grzybowski, M. and Shah, S.P. (1990). Shrinkage Cracking of Fiber Reinforced Concrete. ACI Materials Journal, 87 (2), 138-148. Habel, W.R., Hofmann, D., and Hillemeier, B. (1997). Deformation measurements of mortars at early ages and of large concrete components on site by means of embedded fiber-optic microstrain sensors. Cement & Concrete Composites, 19 (1), 81-102. Hansen, Will. (1987). Drying Shrinkage Mechanisms in Portland Cement Paste. Journal of the American Ceramic Society, 70 (5), 323-328. Hillerborg, A., Modeer, M. and Petersson, P.-E. (1976). Analysis of crack formation and crack growth in concrete by means of fracture mechanics and finite elements, Cement and Concrete Research, 6, 773. Hobbs, D. W. (1974). Influence of Aggregate Restraint on the Shrinkage of Concrete.  American Concrete Institute Journal Proceedings, 71 (9), 445-450. Hua, C., Acker, P., and Ehrlacher, A. (1995). Analyses and models of the autogenous shrinkage of hardening cement paste. I. Modelling at macroscopic scale. Cement and Concrete Research, 25 (7), 1457-1468. Hua, C, Ehrlacher, A., and Acker, P. (1997). Analyses and models of the autogenous shrinkage of hardening cement paste II. Modelling at scale of hydrating grains. Cement and Concrete Research, 27 (2), 245-258. Hossain, A., Pease, B., and Weiss, W. J. (2003). Quantifying Early-Age Stress Development and Cracking in Low w/c Concrete Using the Restrained Ring Test with Acoustic Emission. Transportation Research Record, Concrete Materials and Construction, 1834, 24-33. Igarashi, S., Bentur, A. and Kovler, K. (1999). Stresses and creep relaxation induced in restrained autogenous shrinkage of high strength pastes and concretes. Advances in Cement Research, 11 (4), 169-177. Igarashi, S., Bentur, A. and Kovler, K. (2000). Autogenous shrinkage and induced restraining stresses in high strength concretes. Cement & Concrete Research, 30 (11), 1701-1707. 280 Ishida, T., Chaube, R., Kishi, T., Maekawa, K. (1998). Micro-physical approach to coupled autogenous and drying shrinkage of concrete. Proceedings of International Workshop on Autogenous Shrinkage of Concrete, JCI, Edited by E. Tazawa, Hiroshima, Japan, 271-280. ISIS Canada Update: Meshing Research and Civionics to Revolutionize the Design of Civil Engineering Infrastructure accessed July 16, 2004 http://www.isiscanada.com/latest/main.html Jinno, M., Sakai, S., Osaka, K., Fukuda, T. (2003a). Smart autoclave processing of thermoset resin matrix composites based on temperature and internal strain monitoring. Advanced Composite Materials: The Official Journal of the Japan Society of Composite Materials, 12 (1), 57-72. Jinno, M., Sakai, S., Osaka, K., and Fakuda, T. (2003b). Internal strain monitoring by optical fiber sensor and cure shrinkage during curing of CFRP. Journal of the Society of Materials Science, Japan, 52 (6), 688-694. Kalamkarov, A. L., Fitzgerald, S. B., and MacDonald, D. O. (1999). Use of Fabry Perot fiber optic sensors to monitor residual strains during pultrusion of FRP composites. Composites Part B: Engineering, 30 (2), 167-175. Khajuria, A. and Balaguru, P. (1992). Plastic Shrinkage Characteristics of Fiber Reinforced Cement Composites. Fiber Reinforced Cement and Concrete (Ed. R.N. Swamy), E&FN Spon, London, 82-90. Kim, B., Weiss, W. J. (2003). Using acoustic emission to quantify damage in restrained fiber-reinforced cement mortars. Cement and Concrete Research, 33 (2), 207- 214. Komlos, K. (1979). Uniaxial tensile strength of early age fibre concretes. Mater Constr Mater Struct, 12 (69), 201-206. Kosmatka, S.H., Kerkhoff, B., Panarese, W.C., MacLeod, N.F., and McGrath, R.J. (2002). Design and Control of Concrete Mixtures, EB101, 7th edition, Cement Association of Canada. Kovler, K. (1994). Testing system for determining the mechanical behaviour of early age concrete under restrained and free uniaxial shrinkage. Materials and Structures, 27, 324-330 Kovler, K. (1995). Interdependence of creep and shrinkage for concrete under tension. ASCE Journal of Materials in Civil Engineering, 7 (2), 96-101. Kovler, K., Bentur, A. (1997). Shrinkage of early age steel fiber reinforced concrete.  Archives of Civil Engineering, 43 (4), 431-439. 281 Kovler, K., Sikuler, J. and Bentur, A. (1993). Restrained shrinkage tests of fiber reinforced concrete ring specimens: effect of core thermal expansion. Materials & Structures, 26, 231-237. Kovler, K., Igarashi, S., and Bentur, A. (1999). Tensile Creep Behavior of High Strength Concretes at Early Ages. Materials and Structures, 32, 383-387. Kronlof, A., Markku; L., Pekka, S. (1995). Experimental study on the basic phenomena of shrinkage and cracking of fresh mortar. Cement and Concrete Research, 25 (8), 1747-1754. Lee H.K., Lee K.M., Kim, B.G. (2003). Autogenous shrinkage of high- performance concrete containing fly-ash. Magazine of Concrete Research, 55 (6), 507- 515. Lepage, S., Baalbaki, M., Dallaire E. and Aitcin, P.-C. (1999). Early Shrinkage Development in a High Performance Concrete. Cement, Concrete and Aggregates, 21 (2), 31-35. Li, Zongjin, Qi, M., Li, Zailiang, and Ma, B. (1999). Crack width of high- performance concrete due to restrained shrinkage. ASCE Journal of Materials in Civil Engineering, 11 (3), 214-223. Lura, P., Pease, B., Mazzotta, G. B., Rajabipour, F., Weiss, J. (2007). Influence of Shrinkage-Reducing Admixtures on Development of Plastic Shrinkage Cracks. ACI Materials Journal, 104 (2), 187-194. Mabrouk, R., Ishida, T., Maekawa, K. (2004). A unified solidification model of hardening concrete composite for predicting the young age behavior of concrete. Cement and Concrete Composites: Early Age Concrete - Properties and Performance. 26 (5): 453-461. Malhotra, V.M. (1976). Testing hardened concrete: nondestructive methods, Iowa State University Press. Malhotra, V. M. and Carino, N. J. (1991). CRC handbook on Nondestructive testing of concrete. Boca Raton : CRC Press. Mindess, S., Young, J.F. (1981). Concrete. Prentice Hall Inc., Englewood Cliffs, New Jersey, 481–485. Mindess, S., Young, J. Francis, and Darwin, D. (2003). Concrete, 2nd Edition, Prentice Hall, NJ. Muller, H. S.; Bazant, Z. P.; and Kuttner, C. H. (1999). Data Base on Creep and Shrinkage Tests. Rilem Subcommittee 5 Report RILEM TC 107-CSP, RILEM, Paris, pp. 81. 282 Naaman, A. E., Wongtanakitcharoen, T., and Hauser, G. (2005). Influence of Different Fibers on Plastic Shrinkage Cracking of Concrete. ACI Materials Journal, 102 (1), 49-58. Najm, H., and Balaguru, P. (2002). Effect of Large-Diameter Polymeric Fibers on Shrinkage Cracking of Cement Composites. ACI Materials Journal, 90 (4), 345-351. Naik, T. R., Ramme, B. W., Tews, J. H. (1995). Pavement construction with high- volume class C and class F fly-ash concrete. ACI Materials Journal, 92 (2), 200-210. Nelson P., Sirivivatnanon V., Khatri R. (1992). Development of high volume fly- ash concrete for pavements. Proceedings - Conference of the Australian Road Research Board, Pavement Performance and Materials, 16 (2), 37-47. Nordtest Method NT Build 433. (1995). Concrete: Cracking Tendency – Exposure to Drying during the first 24 hours. Nordtest, Finland. Paillere, A. M., Buil, M. and Serrano, J. J. (1989). Effect of fiber addition on the autogeneous shrinkage of silica fume concrete. ACI Materials Journal, 86 (2), 139-144. Pascale, G., Di Leo, A., and Bonora, V. (2003). Nondestructive assessment of the actual compressive strength of high-strength concrete. Journal of Materials in Civil Engineering, 15 (5), 452-459. Pease, B. J., Shah, H. R., Hossain, A. B., and Weiss, W. J. (2005). Restrained Shrinkage Behavior of Mixtures Containing Shrinkage Reducing Admixtures and Fibers. International Conference on Advances in Concrete Composites and Structures (ICACS), Chennai, India. Pigeon, M., and Bissonette, B. (1999). Tensile Creep and Cracking Potential. Concrete International. 21 (11), 31-35. Princigallo, A., Lura, P., Van Breugel, K., and Levita, G. (2003). Early development of properties in a cement paste: A numerical and experimental study. Cement and Concrete Research, 33 (7), 1013-1020. Qasrawi, Hisham Y. (2000). Concrete strength by combined nondestructive methods simply and reliably predicted. Cement and Concrete Research, 30 (5), 739-746. Qi, C., Weiss, J., and Olek, J. (2003). Characterization of plastic shrinkage cracking in fiber reinforced concrete using image analysis and a modified Weibull function. Materials and Structures, 36 (260), 386-395. Qi, C., Weiss, W. J., and Olek, J. (2005). Assessing the settlement of fresh concrete using a non-contact laser profiling approach. International Conference on Construction Materials: ConMat'05, Vancouver, Canada. 283 Radocea, A. (1994). Model of plastic shrinkage. Magazine of Concrete Research, 46 (167), 125-132. Rivera-Villarreal R. (1986). Effect of Temperature on the Properties of Mortars and Superplasticized Concrete Containing Low-Calcium Fly Ash. American Concrete Institute, Special Publication, SP91-09, 219-230. Rossi, Pierre. (1994). Steel Fiber Reinforced Concrete (SFRC): An Example of French Research. ACI Material Journal, 273-279. Sanjuan M.A. and Moragues A. (1994). Model for predicting plastic shrinkage of polypropylene reinforced mortars. Journal of Materials Science, 29 (11), 495-504. Shaeles, C. A. and Hover, K. C. (1988). Influence of mix proportions and construction operations on plastic shrinkage cracking in thin slabs. ACI Materials Journal, 85 (6), 495-504. Shah, Chanrakant B. (2002). NDT of earthquake-affected structures in Gujarat: Case study.  Indian Concrete Journal, 76 (6), 367-370. Shah, S. P., Karaguler, M. E., Sarigaphuti, M. (1992). Effects of shrinkage- reducing admixtures on restrained shrinkage cracking of concrete. ACI Materials Journal, 89 (3), 291-295. Shah, S. P., Weiss, W.J., and Yang, W. (1998). Shrinkage Cracking-Can It Be Prevented? Concrete International, 20 (4), 51-55. Shehata, E., Rizkalla, S. (1999). Intelligent sensing for innovative bridges. Journal of Intelligent Material Systems and Structures, 10 (4), 304-313. Slowik, V., Schlattner, E., and Klink, T. (2004). Experimental investigation into early age shrinkage of cement paste by using fibre Bragg gratings. Cement and Concrete Composites: Early Age Concrete - Properties and Performance, 26(5), 473-479. Soroushian, P. and Ravanbakhsh, S. (1998). Control of Plastic Shrinkage Cracking with Specialty Cellulose Fibers. ACI Materials Journal, 95 (4), 429-435. Soroushian, P., Mirza, F., Alhozajiny, A. (1993). Plastic Shrinkage Cracking of Polypropylene Fiber Reinforced Concrete. ACI Materials Journal, 92 (5), 553-560. Subramaniam, K.V., Gromotka, R., Shah, S.P., Obla, K., and Hill, R. (2005). Influence of ultra-fine fly ash on the early age response and the shrinking crackage potential of concrete. ASCE Journal of Materials in Civil Engineering, 17 (1), 45-53. Sundaram, V.S., Carette, G.G., and Malhotia, V.N. (1989). Properties of concrete incorporating low quantity of cement and high volumes of low-calcium fly ash. Proceedings of the Third International Conference on Fly Ash, Silica Fume, Slag, and Natural Pozzolans in Concrete, Vol. 1. 284 Tennyson, R. C., Mufti, A. A., Rizkalla, S., Tadros, G. & Benmokrane, B. (2001). Structural health monitoring of innovative bridges in Canada with fiber optic sensors. Smart Materials and Structures, 10 (3), 560-573. Toledo Filho, R.D., Sanjuan, M.A. (1999). Effect of low modulus sisal and polypropylene fibre on the free and restrained shrinkage of mortars at early age. Cement and Concrete Research, 29 (10),1597-1604. Toledo Filho, R. D., Ghavami, K.; Sanjuan, M. A.; England, G. L. (2005). Free, restrained and drying shrinkage of cement mortar composites reinforced with vegetable fibres. Cement and Concrete Composites, 27 (5), 537-546. Topcu, I. B., and Elgun, V. B. (2004). Influence of concrete properties on bleeding and evaporation. Cement and Concrete Research, 34 (2), 275-281. Trottier, J-F, Mahoney, M, and Forgeron, D. (2002). Can Synthetic Fibers Replace Welded-Wire Mesh in Slabs-on-Ground? Concrete International, 24 (11), 59-68. Uno, P.J. (1998). Plastic shrinkage cracking and evaporation formulas. ACI Materials Journal, 95 (4), 365-375. Van Mier Jan G. M. (1996). Fracture Processes of Concrete-Assessment of Material Parameters for Fracture Models, CRC Press. van Zijl, G.P.A.G., de Borst, R., Rots., J.G. (2001). A numerical model for the time-dependent cracking of cementitious materials. Journal for Numerical Methods in Engineering, 52 (7), 637-654. Voigt,T., Bui,V. K., and Shah, S.P. (2004). Drying Shrinkage of Concrete Reinforced with Fibers and Welded-Wire Fabric. ACI Materials Journal, 101 (3), 233- 241. Wang, K., Shah, S. P. and Phuaksuk, P. (2001). Plastic Shrinkage Cracking in Concrete Materials – Influence of Fly Ash and Fibers. ACI Materials Journal, 98 (6), 458-464. Ward, Michael A., Langan, Brian W. (1994). Strength evaluation of in-situ concrete by rebound hammer and core testing. Cement, Concrete and Aggregates, 16 (2), 181-185. Weiss, W. J, Shah, S. P. (1997). Recent trends to reduce shrinkage cracking in concrete pavements. Proceedings of the Airfield Pavement Conference, Aircraft/Pavement Technology: In the Midst of Change, 217-228. Whiting, David A., and Nagi, Mohamad A. (2003). Electrical resistivity of concrete- A literature review. R&D Serial No. 2457, Portland Cement Association, Skokie, Illinois, USA, 57 pages. 285 Yang, E. I., Yi, S. T., & Lee, H. J. (2004). Mechanical characteristics of axially restrained concrete specimens at early ages. Journal of Materials in Civil Engineering. 16 (1), 35-44.  Ye, G., Van Breugel, K., and Fraaij, A.L.A. (2003). Experimental study and numerical simulation on the formation of microstructure in cementitious materials at early age. Cement and Concrete Research, 33 (2), 233-239.  286 APPENDICES- A, B, AND C  Note: (‘X’ and ‘Y’ are the coordinates in the longitudinal and transverse direction of the overlay respectively).  Plain concrete starts with co-ordinates x,y (0,0) 287 APPENDIX A- UPV Measurements Dist. X (cm) Dist. Y (cm) Time (µ Sec) Z “Pulse Velocities” (m/sec) Comments 0 30 182 3296.7 0 90 193 3108.8 0 150 190 3157.9 30 0 193 3108.8 30 60 148 4054.1 30 120 153 3921.6 30 180 189 3174.6 60 30 152 3947.4 60 90 149 4026.8 60 150 142 4225.4 90 0 187 3208.6 90 60 147 4081.6 90 120 148 4054.1 90 180 167 3592.8 120 30 188 3191.5 120 90 174 3448.3 120 150 189 3174.6 150 0 188 3191.5 150 60 153 3921.6 150 120 152 3947.4 150 180 171 3508.8 180 30 193 3108.8 180 90 173 3468.2 180 150 191 3141.4 210 0 187 3208.6 210 60 194 3092.8 210 120 168 3571.4 210 180 169 3550.3 240 30 188 3191.5 240 90 176 3409.1 240 150 171 3508.8 270 0 190 3157.9 270 60 173 3468.2 270 120 173 3468.2 270 180 176 3409.1 300 30 172 3488.4 300 90 157 3821.7 300 150 175 3428.6 330 30 281 2135.2 Finishing Joint 330 90 231 2597.4 330 150 199 3015.1 Dist. X (cm) Dist. Y (cm) Time (µ Sec) Z “Pulse Velocities” (m/sec) Comments 360 0 257 2334.6 360 60 198 3030.3 360 120 196 3061.2 360 180 245 2449.0 390 30 193 3108.8 390 90 194 3092.8 390 150 255 2352.9 420 0 192 3125.0 Plain-FRC Transition 420 60 193 3108.8 420 120 197 3045.7 450 180 196 3061.2 450 30 172 3488.4 450 90 151 3973.5 450 150 172 3488.4 480 0 197 3045.7 480 60 203 2955.7 480 120 184 3260.9 480 180 182 3296.7 510 30 195 3076.9 510 90 200 3000.0 510 150 198 3030.3 540 0 200 3000.0 540 60 201 2985.1 540 120 214 2803.7 540 180 195 3076.9 570 30 223 2690.6 570 90 219 2739.7 570 150 213 2816.9 600 0 197 3045.7 600 60 225 2666.7 600 120 215 2790.7 600 180 196 3061.2 630 30 249 2409.6 630 90 250 2400.0 630 150 244 2459.0 660 0 260 2307.7 660 60 248 2419.4 660 120 241 2489.6 660 180 240 2500.0   288  APPENDIX A (Contd.)  Dist. X (cm) Dist. Y (cm) Time (µ Sec) Z “Pulse Velocities” (m/sec) Comments 690 30 255 2352.9 690 90 278 2158.3 690 150 256 2343.8 720 0 247 2429.1 720 60 244 2459.0 720 120 255 2352.9 720 180 243 2469.1 750 30 245 2449.0 750 90 245 2449.0 750 150 250 2400.0 765 0 115 2608.7 765 60 123 2439.0 765 120 125 2400.0 765 180 119 2521.0 780 30 236 2542.4 780 90 248 2419.4 780 150 245 2449.0  289  APPENDIX B- Schmidt Rebound Values  Schmidt Rebound (R) Readings Dist. X (cm) Dist. Y (cm) 1 2 3 Avg. Strength (alpha=- 90) psi Strength (MPa) 0 0 34 30 33 32 4897 34 0 60 36 34 - 35 5550 38 0 120 32 34 35 34 5223 36 0 180 40 39 42 40 6857 47 60 0 32 37 32 34 5223 36 60 60 35 33 - 34 5305 37 60 120 34 30 37 34 5223 36 60 180 38 35 - 37 5918 41 120 0 31 33 - 32 4815 33 120 60 40 38 - 39 6530 45 120 120 32 36 - 34 5305 37 120 180 36 35 - 36 5673 39 180 0 35 41 38 38 6285 43 180 60 40 38 - 39 6530 45 180 120 42 36 - 39 6530 45 180 180 31 36 33 33 5142 36 240 0 39 37 - 38 6285 43 240 60 33 36 - 35 5428 38 240 120 40 36 - 38 6285 43 240 180 34 32 - 33 5060 35 300 0 30 22 - 26 3345 23 300 60 35 36 - 36 5673 39 300 120 34 34 - 34 5305 37 300 180 37 35 - 36 5795 40 360 0 29 27 33 30 4243 29 360 60 32 34 31 32 4897 34 360 120 34 37 34 35 5550 38 360 180 33 33 34 33 5142 36 Schmidt Rebound (R) Readings Dist. X (cm) Dist. Y (cm) 1 2 3 Avg. Strength (alpha=-90) psi Strength (MPa) 420 0 30 30 28 29 4162 29 420 60 32 32 32 32 4815 33 420 120 35 34 34 34 5387 37 420 180 32 34 32 33 4978 34 480 0 34 32 34 33 5142 36 480 60 36 34 38 36 5795 40 480 120 36 36 36 36 5795 40 480 180 34 36 32 34 5305 37 540 0 32 35 32 33 5060 35 540 60 34 34 32 33 5142 36 540 120 34 35 33 34 5305 37 540 180 32 34 32 33 4978 34 600 0 34 34 32 33 5142 36 600 60 36 35 35 35 5632 39 600 120 35 37 35 36 5713 39 600 180 35 34 34 34 5387 37 660 0 32 33 33 33 4978 34 660 60 35 34 33 34 5305 37 660 120 40 37 40 39 6530 45 660 180 35 37 33 35 5550 38 720 0 34 35 34 34 5387 37 720 60 36 40 36 37 6122 42 720 120 39 35 37 37 6040 42 720 180 34 38 34 35 5632 39 780 0 34 37 34 35 5550 38 780 60 42 37 40 40 6693 46 780 120 38 39 34 37 6040 42 780 180 40 36 39 38 6367 44 810 0 33 34 33 33 5142 36 810 60 41 40 44 42 7183 50 810 120 38 42 38 39 6612 46 810 180 40 40 36 39 6448 45   290 APPENDIX C-Electrical Resistivity Measurements Dist. X (cm) Dist. Y (cm) Meter reading (KΩ) Resistivity (KΩ-cm) 0 30 4.39 137.92 0 90 4.92 154.57 0 150 4.8 150.80 30 0 4.12 129.43 30 60 4.95 155.51 30 120 5.58 175.30 30 180 5 157.08 60 30 3.73 117.18 60 90 4.25 133.52 60 150 5.15 161.79 120 30 4.22 132.58 120 90 5 157.08 120 150 4.6 144.51 150 0 3.55 111.53 150 60 4.39 137.92 150 120 3.88 121.89 150 180 4.47 140.43 180 30 4 125.66 180 90 4.45 139.80 180 150 5.61 176.24 240 30 2.13 66.92 240 90 3.51 110.27 240 150 2.41 75.71 270 0 1.812 56.93 270 60 2.06 64.72 270 120 2.35 73.83 270 180 1.9 59.69 300 30 2.19 68.80 300 90 2.24 70.37 300 150 2.17 68.17 360 30 1.377 43.26 360 90 2.16 67.86 360 150 2.12 66.60 390 0 1.23 38.64 390 60 1.672 52.53 390 120 2.12 66.60 390 180 1.97 61.89 Dist. X (cm) Dist. Y (cm) Meter reading (KΩ) Resistivity (KΩ-cm) 420 30 1.554 48.82 420 90 2.07 65.03 420 150 2.55 80.11 480 30 1.96 61.58 480 90 2.1 65.97 480 150 2.18 68.49 510 0 1.422 44.67 510 60 1.7 53.41 510 120 1.936 60.82 510 180 1.7 53.41 540 30 1.715 53.88 540 90 1.744 54.79 540 150 1.79 56.23 600 30 1.52 47.75 600 90 1.66 52.15 600 150 2.35 73.83 630 0 1.6 50.27 630 60 1.566 49.20 630 120 2.54 79.80 630 180 2.37 74.46 660 30 1.98 62.20 660 90 2 62.83 660 150 2.72 85.45 720 30 2.29 71.94 720 90 2.18 68.49 720 150 2.78 87.34 750 0 2.8 87.96 750 60 2.7 84.82 750 120 3.04 95.50 750 180 3.54 111.21 780 30 3.23 101.47 780 90 3.26 102.42 780 150 3.44 108.07 810 30 2.9 91.11 810 90 3.13 98.33 810 150 3.43 107.76  Note-Values in bold indicate that neither ‘Rp’ nor ‘Rc’ lights were flashing 291  APPENDIX D-Laboratory UPV Measurements  Mix Type Plain concrete Age >28 days Probe Type of Time UPV Comment Distance Measurement (Micro-Sec) (m/sec) Spec. # (mm) 100 Direct 22.1 4524.9 88.7 3382.2 Casting Face 63.4 4731.9 Side 2 63.1 4754.4 Side 3 300 Indirect 63.1 4754.4 Side 4 329 Semi direct 75 4386.7 1 350 Direct 69.2 5057.8 100 Direct 22 4545.5 86.9 3452.2 Casting Face 63.8 4702.2 Side 2 63.2 4746.8 Side 3 300 Indirect 62.8 4777.1 Side 4 329 Semi direct 75.4 4363.4 2 350 Direct 69 5072.5 100 Direct 21.7 4608.3 69.2 4335.3 Casting Face 63.5 4724.4 Side 2 62.7 4784.7 Side 3 300 Indirect 63.8 4702.2 Side 4 329 Semi direct 75.2 4375.0 3 350 Direct 69.1 5065.1  292 APPENDIX D-Laboratory UPV measurements (Contd.)  Mix Type FRC (Cellulose 0.2%) Age >28 days Probe Type of Time UPV Comment Distance Measurement (Micro-Sec) (m/sec) Spec. # (mm) 100 Direct 23 4347.8 90.6 3311.3 Casting Face 69 4347.8 Side 2 68.6 4373.2 Side 3 300 Indirect 69.7 4304.2 Side 4 329 Semi direct 80 4112.5 1 350 Direct 72.4 4834.3 100 Direct 23.1 4329.0 94.5 3174.6 Casting Face 70.5 4255.3 Side 2 67.2 4464.3 Side 3 300 Indirect 68.9 4354.1 Side 4 329 Semi direct 79.9 4117.6 2 350 Direct 73.9 4736.1 100 Direct 22.9 4366.8 77.5 3871.0 Casting Face 65.3 4594.2 Side 2 66.9 4484.3 Side 3 300 Indirect 67.6 4437.9 Side 4 329 Semi direct 80.2 4102.2 3 350 Direct 72.5 4827.6    

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0063070/manifest

Comment

Related Items