Early age shrinkage and bond at interface between repair material and concrete substrate

Effect of moisture condition of substrate on the early age restraint shrinkage of composite specimen from 0.5 hours to 90 days after adding water to new concrete 201 6.3.5.. LIST OF TABL

BETWEEN REPAIR MATERIAL AND CONCRETE

SUBSTRATE

KYAW MYINT LAY

(B E.)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHIOSOPHY

DEPARTMENT OF CIVIL ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2007

I begin by expressing my gratitude to my supervisor, Associate Professor K C

Gary Ong, for his guidance, advice, and patience throughout the course of this study I

will always be grateful for lessons learned under his tutelage

I would like to thank the staffs of the Concrete Technology and Structural Engineering Laboratory of the Department of Civil Engineering at the National University of Singapore for their kind assistance throughout the study

Finally, I express my deepest gratitude to my family for all their love, support,

and encouragement throughout my life Without their guidance and encouragement this

work would not have been possible Last but not least I would like to dedicate this work

to my beloved wife, Mya Nandar, and our daughter, Cheryl Lee @ Mya Cherry, for their

understanding, love and motivation throughout this study

TABLE OF CONTENTS

Pages TITLE PAGE

SUMMARY x

1.4 ‘Time-zero’ for early age shrinkage measurement 9

1.5 Effect of very early age shrinkage in the concrete structure 10

1.6 Effect of early age shrinkage on composite concrete structure 11

1.6.1 Tensile stress in the composite concrete 12

1.6.2 Debonding at interface of the composite concrete 12

1.8 Organization of thesis 14

CHAPTER 2 IMAGE ANALYSIS TECHNIQUE

2.1 Introduction 16 2.2 Application of image analysis used in the monitoring of movements 16

2.2.3 In-plane deformation or displacement measurement 17

2.3 Image analysis technique and equipment used in current study 19

2.3.2 Procedures in the image analysis technique 21

2.3.5.2 Analysis parameters in tracking the targets 26

2.4.2.2.Calculation of uncertainty in the calibration of the image

analysis with dial gauge (using 4288 pixels) 35 2.4.2.3 Calculation of uncertainty in the calibration of the image

analysis with dial gauge (using 12864 pixels) 37

2.5 Effect of the software 39

2.6 Effect of image acquisition technique and targets 40

2.7.1 Effect of lens distortion on the image analysis technique 45

2.7.2 Effect of distortion with different focal length tested using fixed

2.8.1 Effect of instability of camera mounting 54

2.8.2 Effect of leveling of camera and specimen 55

3.3.1 Laser sensor method used in this study 67

3.3.2 Moisture loss and temperature rise measurement 68

3.4 Behaviors of the concrete mixes at a very early age 71

3.4.2 Very early age total shrinkage of concrete 73

3.4.3 Stiffening effect on very early age total shrinkage 75

3.5 Correlation between image analysis and laser sensor techniques 76

3.6 Correlation between image analysis and Demec measurement 81

3.7 Factors affecting very early age shrinkage monitored using the image

3.7.1 Settlement of the targets and its effect 82

2.7.3.1.Shrinkage in the longitudinal direction 89

2.7.3.2.Shrinkage in the transverse direction 90

2.7.3.3.Effect of specimen size on mortar specimen 92

3.8 Summary 96

CHAPTER 4 EFFECT OF WATER AND SUPERPLASTICIZER ON

VERY EARLY AGE TOTAL SHRINKAGE

4.1 Introduction 99 4.2 Early age shrinkage of concrete in literature 99

4.3.1 Aggregates 100

4.5.3 Effect of moisture loss on very early age total shrinkage 106

4.5.4 Effect of moisture loss on very early age shrinkage (starting from

4.5.5 Effect of W/C ratio on very early age total shrinkage 112

4.6.3 Effect of W/C very early age total shrinkage of concrete 117

4.7 Effect of setting time on very early age total shrinkage of concrete 119

4.8.1 Empirical model for very early age shrinkage (before initial setting

time) 123 4.8.2 Empirical models for very early age shrinkage (after initial setting

time) 127 4.9 Effect of amount of water added on the very early age total shrinkage of

4.9.1 Specimen preparation and testing procedure 133

4.9.3 Expansion before and after setting time 135

4.10 Summary 136

CHAPTER 5 EFFECT OF SILICA FUME AND SHRINKAGE

REDUCING ADMIXTURE ON VERY EARLY AGE TOTAL SHRINKAGE

5.1 Introduction 137 5.1.1 Effect of silica fume on very early age properties of concrete 137

5.1.2 Effect of shrinkage reducing admixture on very early age

5.3.6 Empirical model for very early age shrinkage of concrete with

silica fume replacement before initial setting time 156

5.3.7 Empirical models for very early age shrinkage of concrete with

silica fume replacement after initial setting time 160

5.3.8 Effect of SRA on early age total shrinkage of concrete 162

5.4 Summary 168

CHAPTER 6 EARLY AGE DIFFERENTIAL SHRINKAGE AND

BOND STRENGTH DEVELOPMENT IN COMPOSITE CONCRETE

6.1 Introduction 170 6.1.1 Early age differential shrinkage in composite concrete 170

6.1.2 Effect of early age shrinkage on bond strength 172

6.1.3 Effect of moisture condition of substrate on bond strength 172

6.2.3 Specimen preparation and shrinkage monitoring 177

6.2.4.1 Existing bond strength test methods 178

6.2.4.2 Shear bond test method used in this study 180

6.3 Results and discussions on differential shrinkage of composite specimen 181

6.3.2 Drying shrinkage of monolithic specimen 184

6.3.3 Absorption and expansion of substrate concrete 185

6.3.4 Effect of moisture condition of substrate on restraint shrinkage 188

6.3.4.1 Very early age restraint shrinkage of composite specimen 188

6.3.4.2 Effect of moisture condition of substrate on the very early age

restraint shrinkage of composite specimen (during the first 24 hour after adding water to the new concrete) 195

6.3.4.3 Effect of moisture condition of substrate on the early age

restraint shrinkage of composite specimen (from 1 day to 90 days after adding water to new concrete) 199

6.3.4.4 Effect of moisture condition of substrate on the early age

restraint shrinkage of composite specimen (from 0.5 hours to

90 days after adding water to new concrete) 201

6.3.5 Effect of age of substrate on the early age differential shrinkage of

6.3.6 Effect of the silica fume and SRA on the early age differential

6.4 Results and discussions on bond strength development in composite

specimen 215 6.4.1 Effect of moisture condition of substrate on shear bond strength of

SUMMARY

With the development of high strength concretes which are more sensitive to cracking immediately after setting, there is more interest in the early age shrinkage of cementitious material Due to difficulties associated with the fixing and placing of targets for use with a number of shrinkage monitoring devices on fresh concrete, that has not set, information about very early age shrinkage was not well documented in the literature Very early age shrinkage may also have significant effects on induced stresses and bond strength development in the case of a composite specimen comprising freshly cast cementitious material on a hardened concrete substrate

The research was firstly focused on the development of a new approach for monitoring very early age shrinkage of concrete occurring immediately after mixing the concrete and continuing for the first 24 hours A new approach was developed by using digital image analysis that can monitor the very early age shrinkage of fresh cementitious material 30 minutes after water was added to the mix The new approach could also be used in monitoring of the early age differential shrinkage of composite specimens The result obtained from this new technique was correlated with other existing techniques, a laser sensor technique during the first 24 hours after adding water to the mix and with Demec Gauge measurement from 24 hours to 7 days after casting Then the source of errors that can occur in the image analysis method and ways to overcome the error were investigated

Once the new method was established, the phenomena causing the early age total shrinkage of concrete during the first few hours after adding water to the mix were investigated From experimental results, the three factors that significantly contribute to

the early age shrinkage were identified as moisture loss, time of hardening and W/C Then parameters that may affect the early age total shrinkage were studied, such as the use of superplasticizer, shrinkage reducing admixture and silica fume Then, a generalized model was proposed for very early age total shrinkage of concrete

With the knowledge of the very early age of concrete, the research extended to investigate the differential shrinkage of composite specimens comprising concrete substrate and freshly cast repair material Finally, the effect of differential shrinkage of composite on the bond strength development at the interface was measured That test result will help to define a proper strategy to improve the bond strength, for a more

durable repair system

Key words: bond strength, differential shrinkage, early age shrinkage, evaporation,

image analysis, shrinkage, shrinkage reducing admixture and silica fume

LIST OF TABLES

Pages

Table 2.1 – Parameters used in the tests for lens distortion test 47

Table 3.1 – Comparison of methods used for measuring of shrinkage at an early

age 66

Table 3.3 – Physical properties of concrete 71

Table 4.1 – Mix proportions of concrete with different W/C 102

Table 4.2 – Physical properties of concrete with Daracem 100 104

Table 4.3 – Physical properties of concrete with Adva 105 115

Table 4.4 – Shrinkage coefficient for different W/C ratios 124

Table 4.5 – Shrinkage strains at 48 hours after adding water to the mix (measured

starting from the initial setting time of concrete 129

Table 4.6 – Amount of mix water in construction grout 133

Table 4.7 – Physical properties of construction grout as specified by manufacturer 133

Table 5.1 – Mix proportions of concrete with silica fume 141

Table 5.2 – Mix proportions of concrete with shrinkage reducing admixture 142

Table 5.3 – Physical properties of concrete with silica fume 143

Table 5.4 – Physical properties of concrete with SRA 144

Table 5.5 – Shrinkage coefficient for the mixes with silica fume (Adva 105) 157

Table 5.6 – Shrinkage coefficient for the mixes with silica fume (Daracem 100) 157

Table 5.7 – Shrinkage strains at 48 hours after adding water to the mix (starting

Table 6.1 – Mix proportions of concrete used in composite specimen 175

Table 6.2 – Configurations of the composite specimen 176

Table 6.3 – Physical properties of concrete used in the composite specimens 182

Table 6.4 – Compressive strength development of RC, SRA, and SF mixes 182

Table 6.5 – Shear strength development of RC, SRA, and SF mixes 182

Table 6.6 – Shear bond strength of composite specimens with different moisture

Fig 1.3 – Chemical shrinkage and volumetric autogenous shrinkage of cement

paste with water-to-binder ratio of 0.30 [Justnes et al., 2000] 5

Fig 1.4 – Horizontal shrinkage, vertical shrinkage, volumetric autogenous

shrinkage and volumetric chemical shrinkage of mortar with W/C of

0.30 comparing with chemical shrinkage [Holt, 2001] 6

Fig 2.1 – Steel pins targets for image analysis 21

Fig 2.2 – Location of targets used in a prism 21

Fig 2.3 – Experimental set up for image analysis technique and laser sensor

technique 23

Fig 2.4 – Very early age shrinkage of a typical concrete prism specimen with

exposed trowelled surface analyzed using images (3072 x 2048 pixels)

and (9216 x 6614 pixels) during the first 6 hours after adding water to

Fig 2.5 – The location of two typical targets ((i) is the left target and (ii) is right

target) with time (from 30 min to 48 hours after adding of mix water)

Fig 2.6 – Shrinkage strains obtained using two typical targets with time (from 30

Fig 2.7 – Ratio of length difference to length of image 1 for the three pairs of

targets 30

Fig 2.8 – Image of testing frame and the 50 targets on the bar captured using the

Fig 2.9 – Correlation of movements measured using dial gauge and image

analysis of images captured using a focal length of 70 mm (4288

Fig 2.10 – Image of testing frame and the 50 targets on the bar capture using the

Fig 2.11 – Correlation of movements measured using dial gauge and image

analysis of images captured using a focal length of 18 mm (4288

Fig 2.12 –Variation of number of pixels between two set of fixed targets tested

Fig 2.13 –Variation of number of pixels between two set of fixed targets tested

Fig 2.14 – Effect of software on shrinkage monitoring using image analysis 40

Fig 2.15 – Image of two targets (8 mm size circle) and (4 mm size square) targets

points in center to center distance of about 12 mm 41

Fig 2.16 – Variation in pixel coordinate readings for fixed target point using 12

Fig 2.17 – Variation in pixel coordinate readings for fixed target point using 400

Fig 2.18 – Geometry of curvilinear distortion (a) stop at lens (b) symmetrical

design (c) Barrel distortion: stop at in front of positive lens (d)

Pincushion distortion: stop behind positive lens (Ray, 2002) 44

Fig 2.19 – General-purpose lens for 24 x 36 mm format C, center; S1, S2 sides

and E, edge of format; P, Pincushion, B, barrel; M, change at close

Fig 2.20 – Barrel distortion effect in the images captured using a focal length of

Fig 2.23 – Effect of lens distortion using lens at different focal lengths 49

Fig 2.24 – Effect of lens distortion on the images captured using a focal length of

Fig 2.25 – Effect of lens distortion on the images captured using a focal length of

Fig 2.26 – Effect of lens distortion using lens at different focal lengths 51

Fig 2.27 – Comparison of results obtained using fixed targets and moving targets 52

Fig 2.28 – Variation in the pixels/mm values with respect to the location of the

targets and the mean pixel/mm value captured using a focal length of

Fig 2.29 – Variation in the pixels/mm values with respect to the location of the

targets and the mean pixel/mm value captured using a focal length of

Fig 2.30 – Demonstration of misalignment of camera and specimen 56

Fig 3.1 – Experimental set up for measurement of autogenous shrinkage of

concrete gauge plug embedded (JCI Report, 1998) 60

Fig 3.2 – Experimental set up for the measurement of horizontal shrinkage of

Fig 3.3 – Autogenous shrinkage measurement using dilatometer [Jensen and

Fig 3.4 – Experimental set up for the measurement of autogenous shrinkage

Fig 3.5 – Early age shrinkage measurement using one laser sensor [Kaufmann et

Fig 3.8 – Location of target pins for the image analysis technique and reflected

plates for the laser sensor technique on the specimen tested 68

Fig 3.9 – Rate of moisture loss and temperature measured with tap water at 65 ±

Fig 3.10 – Rate of moisture loss from the specimen during the first 48 hours after

Fig 3.11 – Temperature rise of the specimens during the first 48 hours after

Fig 3.12 – Very early age total shrinkage of concrete specimen during the first 48

hours starting 30 minutes after adding water to the mix and at the

Fig 3.13 – Comparison of setting time, temperature rise and shrinkage rate

measured for Mix 1 during the first 24 hours after adding water to the

mix 76

Fig 3.14 – Comparison of shrinkage rate reading between image analysis and

Fig 3.15 – Comparison between image analysis and laser sensor techniques

starting 30 minutes after adding water to the mix 79

Fig 3.16 – Comparison of total moisture loss from the mixes during the first 48

Fig 3.17 – Comparison between image analysis and laser sensor techniques

starting at the initial setting time of concrete 81

Fig 3.18 – Shrinkage of Mix M obtained using image analysis and the Demec

gauge on the same specimen (average of three readings) 82

Fig 3.19 – Illustration of effect of settlement on image analysis technique 83

Fig 3.20 – Location of 100 mm targets and 10 mm targets for image analysis 84

Fig 3.21 – Effect of settlement monitored concrete specimen cast using Mix RC 85

Fig 3.22 – Effect of expansion and settlement monitored in Mix L construction

Fig 3.25(a) – Effect of settlement monitored in Mix M construction grout

specimen starting 24 hour after adding mix water (specimen 1) 88

Fig 3.25(b) – Effect of settlement monitored in Mix M construction grout

specimen starting 24 hour after adding mix water (specimen 2) 89

Fig 3.26 – Comparison of very early age total longitudinal shrinkage of concrete

specimens of width 100 mm and 50 mm (setting time 5:15 hours) 90

Fig 3.27 – Comparison of very early age total longitudinal shrinkage of concrete

specimens of width 100 mm and 50 mm (setting time 10:20 hours) 90

Fig 3.28 – Comparison of very early age total transverse shrinkage of concrete

specimen of width 100 mm & 50 mm (setting time 5:15 hours) 92

Fig 3.29 – Comparison of very early age total transverse shrinkage of concrete

specimen of width 100 mm & 50 mm (setting time 10:20 hours) 92

Fig 3.30 – Comparison of very early age total longitudinal shrinkage of mortar

Fig 3.31 – Comparison of very early age total transverse shrinkage of mortar

Fig 3.32 – Comparison of very early age total longitudinal and transverse

Fig 3.33 – Comparison of very early age total shrinkage of concrete with

specimen of width 100 mm in the longitudinal and diagonally across

Fig 3.34 – Comparison of very early age total shrinkage of concrete with

specimen of width 50 mm in the longitudinal and diagonally across

Fig 4.1 – Sieve analysis result of fine aggregates 101

Fig 4.2 – Sieve analysis result of coarse aggregates (19 mm - 4.75 mm) 101

Fig 4.3 – Rate of moisture loss from the mixes with Daracem 100 105

Fig 4.4 – Temperatures of the specimens during the first 48 hours after adding

Fig 4.5 – Very early age total shrinkage of concrete cast with D 30 mix with

Fig 4.6 – Very early age total shrinkage of concrete cast with D 35 mix with

Fig 4.7 – Very early age total shrinkage of concrete cast with D 40 mix with

Fig 4.8 – Very early age total shrinkage of concrete cast with D 45 mix with

Fig 4.9 – Very early age total shrinkage of concrete cast with D 30 mix with

different curing conditions (plotted starting from the initial setting time

Fig 4.10 – Very early age total shrinkage of concrete cast with D 35 mix with

different curing conditions (plotted starting from the initial setting time

Fig 4.11 – Very early age total shrinkage of concrete cast with D 40 mix with

different curing conditions (plotted starting from the initial setting time

Fig 4.12 – Very early age total shrinkage of concrete cast with D 45 mix with

different curing conditions (plotted starting from the initial setting time

Fig 4.13 – Very early age total shrinkage of concrete cast with different W/C

with Daracem 100 (plotted starting from the initial setting time of the

Fig 4.14 – Very early age total shrinkage of concrete cast with different W/C

Fig 4.15 – Very early age total shrinkage of D series concrete mixes cast with

different W/C ratios (plotted starting from the initial setting time of the

mix) 114

Fig 4.16– Total shrinkage of D series concrete specimens cast with different W/C

ratios (plotted starting from 1 day after the casting of the mix) 115

Fig 4.17 – Temperatures of the specimens during the first 48 hours after adding

Fig 4.18–Rate of moisture loss from the specimen from the mix with Adva 105 117

Fig 4.19 (a) – Very early age total shrinkage of concrete cast with different W/C

Fig 4.19(b) – Very early age total shrinkage of A series concrete mixes cast with

different W/C ratios (plotted starting from the initial setting time of the

mix) 118

Fig 4.20 – Total shrinkage of A series concrete cast with different W/C ratios

(plotted starting from 1 days after the casting of the mix) 119

Fig 4.21 – Comparison of very early age shrinkage of mixes with Adva 105 and

Fig 4.22 – Comparison of rate of moisture loss between the D 30 and A 30 mixes 120

Fig 4.23 – Comparison of very early age total shrinkage of mixes with Adva 105

and Daracem 100 (plotted starting from the initial setting time of the

Fig 4.24 – Comparison of very early age total shrinkage of mixes with Adva 105

and Daracem 100 (plotted starting from 1 day after casting of the mix) 122

Fig 4.25 – Comparison of shrinkage coefficient of A and D mixes 125

Fig 4.26 – A comparison of measured values and predicted values for D 30 mix 126

Fig 4.27 – A comparison of measured values and predicted values for D 40 mix 126

Fig 4.28 – Comparisons of measured values and predicted values for A 30 and

Fig 4.29 – Shrinkage strains at 48 hours after adding water to the mixes

(measured starting from the initial setting time of mix) 129

Fig 4.30 – A comparison of measured and predicted values for the D 30 and D 40

mixes 130

Fig 4.31 – Comparisons of measured and predicted value for the A 30 and A 40

mixes 130

Fig 4.32 – Comparisons of measured and predicted value for the D 30 and D 40

mixes (measured starting 30 minutes during 48 hours after adding mix

water) 131

Fig 4.33 – Comparisons of measured and predicted value for the A 30 and A 40

mixes (measured starting 30 minutes during 48 hours after adding mix

water) 131

Fig 4.34 – Sieve analysis of construction grout before and after washing (ASTM

C33-90a) 132

Fig 4.35 –Very early age shrinkage of grout (with different amounts of mix water)

starting 30 minutes after adding water to the mix 134

Fig 4.36 – Rate of water loss from the construction grout specimen with different

Fig 5.1 – Temperature development in the A-30 & D-30 series 145

Fig 5.2 – Temperature development in the A-35 & D-35 series 146

Fig 5.3 – Rate of moisture loss during the first 30 hours after adding water to the

Fig 5.4 – Rate of moisture loss during the first 30 hours after adding water to the

Fig 5.5 – Rate of moisture loss during the first 30 hours after adding water to the

Fig 5.6 – Rate of moisture loss during the first 30 hours after adding water to the

Fig 5.7 – Effect of silica fume on very early age total shrinkage of A-35 mixes

starting 30 minutes after adding water to the mix and after initial

Fig 5.8 – Effect of silica fume on very early age total shrinkage of the A-30

series starting 30 minutes after adding water to the mix and after the

Fig 5.9 – Effect of silica fume on drying shrinkage of the A-30 and A-35 series

starting 24 hours after adding water to the mixes 152

Fig 5.10 – Effect of silica fume on very early age total shrinkage of the D 35

series starting 30 minutes after adding water to the mix and after initial

Fig 5.11 – Effect of silica fume on very early age total shrinkage of the D 30

series starting 30 minutes after adding water to the mix and after initial

Fig 5.12 – Effect of silica fume on drying shrinkage of concrete mixes A-30 and

A-35 series starting from 24 hours after adding water to the mixes 156

Fig 5.13 – Comparison shrinkage coefficient of mixes with different silica fume

content 157

Fig 5.14(a) – Very early age shrinkage of the D-30 mixes before the initial

Fig 5.14(b) – Very early age shrinkage of the D-35 mixes before the initial

Fig 5.15 – Shrinkage strains at 48 hours after adding water to the mixes with

different silica fume content (starting from initial setting time of

concrete) 161

Fig 5.16 – A comparison of measured and predicted shrinkage strains for the

Fig 5.17 – A comparison of measured and predicted shrinkage strains for D-35

mixes 162

Fig 5.18 – Rate of moisture loss from the specimen with W/C 0.30 and different

Fig 5.19 – Temperature development in the SRA 10, SRA 15 & SRA 20 with

Fig 5.20 – Very early age total shrinkage of mixes with a W/C ratio of 0.30

Fig 5.21 – Very early age total shrinkage of mixes with a W/C ratio of 0.30 and

different SRA contents (plotted starting from the initial setting time of

Fig 5.22 – Very early age total shrinkage of mixes with a W/C ratio of 0.30 and

different SRA contents (plotted starting from 1 day after casting) 167

Fig 5.23 – A comparison of effect of sealing and effect of SRA in reducing the

Fig 6.1 – Sieve analysis of coarse aggregate used (10 mm maximum size

Fig 6.2 – Composition and location of targets in the composite specimen 176

Fig 6.3 – Some common methods used in measuring bond strength of composite

Fig 6.4 – Test set up and specimen geometry used in direct shear test (Ray et al.,

2005) 179

Fig 6.5 – Experimental set up for shear bond strength (front view, side views

Fig 6.6 – Sawing of composite specimen into 5 cubes for shear bond tests 181

Fig 6.7 – Sawing operation to obtain 100 mm cubes for testing 181

Fig 6.8 – Compressive strength development of the RC, SRA, and SF mixes

Fig 6.9 – Shear strength development of the RC, SRA, and SF mixes during the

Fig 6.10 – Ratio of shear strength to compressive of the RC, SRA, and SF mixes

Fig 6.11 – Drying Shrinkage of substrate and repair concrete from 1 day to 90

Fig 6.12 – Moisture loss and shrinkage of OD substrate during drying at 105˚ C 187

Fig 6.13 – Comparison of moisture absorption and expansion of the AD and OD

specimen (50 mm x 100 mm x 500 mm) upon being submerged in

Fig 6.16 – Comparison of very early age free shrinkage of monolithic specimen

(50 mm thick) and restrained shrinkage of the OD-90 composite

specimen, plotted from 0.5 hours to 24 hours after adding water to the

Fig 6.17 – Comparison of very early age free shrinkage of monolithic specimen

(50 mm thick) and restrained shrinkage of the AD-90 composite

specimen, plotted from 0.5 hours to 24 hours after adding water to the

Fig 6.18 – Comparison of very early age free shrinkage of monolithic specimen

(50 mm thick) and restrained shrinkage of the SD-90 composite

specimen, plotted from 0.5 hours to 24 hours after adding water to the

Fig 6.19 – Specimen and experimental set up used for the early age shrinkage

Fig 6.20 – Early age free shrinkage of concrete with and without metallic fiber

tested in monolithic specimen; unsealed specimen (a) (b) and sealed

Fig 6.21 – Early age shrinkage of concrete with and without metallic fiber,

monolithic specimen (a) (b) and composite specimens (e) (f) (Granju

Fig 6.22 – Very early age differential shrinkage of the OD-90 composite

specimen, plotted starting from 0.5 hours to 24 hours after adding

Fig 6.23 – Very early age differential shrinkage of the AD-90 composite

specimen, plotted starting from 0.5 hours to 24 hours after adding

Fig 6.24 – Very early age differential shrinkage of the SD-90 composite

specimen, plotted starting from 0.5 hours to 24 hours after adding

Fig 6.25 – Determination of “break point” in the AD-90 composite specimen 198

Fig 6.26 – Differential shrinkage of the OD-90 composite specimen, plotted

starting from 1 day to 90 days after adding water to the RC concrete

mix 200

Fig 6.27 – Differential shrinkage of the AD-90 composite specimen, plotted

starting from 1 day to 90 days after adding water to the RC concrete

mix 201

Fig 6.28 – Differential shrinkage of the SD-90 composite specimen, plotted

starting from 1 day to 90 days after adding water to the RC concrete

mix 201

Fig 6.29 – Differential shrinkage of the OD-90 composite specimen, plotted

starting from 0.5 hours to 90 days after adding water to the RC

Fig 6.30 – Differential shrinkage of the AD-90 composite specimen, plotted

starting from 0.5 hours to 90 days after adding water to the RC

Fig 6.31 – Differential shrinkage of the SD-90 composite specimen, plotted

starting from 0.5 hours to 90 days after adding water to the RC

Fig 6.32 – Comparison of early age restrained shrinkage of the RC concrete layer

of the composite specimens with different moisture conditions plotted

Fig 6.33 – Comparison of early age restrained expansion of the substrate layer of

the composite specimen with different moisture conditions plotted

Fig 6.34 – Comparison of expansions of the monolithic and OD-90 substrate

specimen, plotted starting from 30 minutes to 90 days 206

Fig 6.35 – Comparison of expansions of the monolithic and AD-90 substrate

specimens, plotted starting from 30 minutes to 90 days 207

Fig 6.36 – Very early age differentiate shrinkage of the AD-3 composite

specimen, plotted starting from 0.5 hours to 90 days after adding water

Fig 6.37 – Very early age differentiate shrinkage of the AD-28 composite

specimen, plotted starting from 0.5 hours to 90 days after adding water

Fig 6.38 – Differential shrinkage of the AD-3 composite specimen, plotted

starting from 1 day to 90 days after adding water to the RC concrete

mix 210

Fig 6.39 – Differential shrinkage of the AD-28 composite specimen, plotted

starting from 1 day to 90 days after adding water to the RC concrete

mix 211

Fig 6.40 – Very early age differential shrinkage of the SF-90 composite specimen,

plotted starting from 0.5 hours to 90 days after adding water to the SF

Fig 6.41 – Very early age differential shrinkage of the SRA-90 composite

specimen, plotted starting from 0.5 hours to 90 days after adding water

Fig 6.42 – Comparison of very early age shrinkage of the RC, SF and SRA

Fig 6.43 – Comparison of expansion of the substrate of the AD-90, SF-90 and

Fig 6.44 – Differential shrinkage of the SF-90 composite specimen, plotted

starting from 1 day to 90 days after adding water to the SF concrete

mix 215

Fig 6.45 – Differential shrinkage of the SRA-90 composite specimen, plotted

starting from 1 day to 90 days after adding water to the SRA concrete

mix 215

Fig 6.46 – Interface of the SD-90 composite specimens before shear bond testing 219

Fig 6.47 – Interface of the AD-90 composite specimens before shear bond testing 220

Fig 6.48 – Interface of the OD-90 composite specimens before shear bond testing 220

Fig 6.49 – Sawn cube specimens after testing with interface between the RC

concrete layer and the substrate concrete layer exposed 221

Fig 6.50 – Shear bond strength of development of the composite specimens with

Fig 6.51 – Coefficient of variation (COV) of the bond test specimens with

Fig 6.52 – Variation of shear bond strength of the AD-90 composite specimen

Fig 6.53 – Variation of shear bond strength of the OD-90 composite specimen

Fig 6.54 – Shear bond strength of development of composite specimens with

Fig 6.55 – Coefficient of variation (COV) of the test results of composite

Fig 6.56 – Sawn cube specimens after testing with interface between the RC, SF

and SRA concrete layer and the substrate concrete layer exposed 229

Fig 6.57 – Shear bond strength of development of composite specimens with RC,

Fig 6.58 – Coefficient of variation (COV) of the composite specimens with RC,

Chapter 1 Introduction 1.1 Background

Shrinkage or volume change of concrete play an important role in the design of concrete structures that are sensitive to cracking such as water retaining structures, massive concrete pours and relatively thin repaired structures According ACI 224R-98, cracking due to shrinkage cannot be eliminated in most structures Other common problems that are attributed to the shrinkage of concrete are losses of pre-stress in pre-stressed concrete members and debonding of concrete in repair structure

In conventional design of such structures, shrinkage or volume change is assumed

to begin at the time of loading or drying In reality, volume change commences immediately after the cement and water come in contact during concrete mixing, an unavoidable phenomenon Even when the concrete curing condition is ideal, significant shrinkage could occur during the first day after casting Shrinkage which occurs within the first day could give rise to cracking as the tensile strength of the concrete is very low

As mentioned previously, early age shrinkage is commonly defined as the shrinkage during the first day, while the concrete is setting and starting to harden (Holt, 2000) Long term shrinkage refers to the shrinkage of the concrete at an age of after 24 hours Shrinkage occurring after 24 hours is easy to monitor as the concrete specimen is demolded and standardized shrinkage measurements method can be used In this study

“very early age shrinkage” is used to denoted shrinkage to be monitored starting before the setting time of concrete At such very early ages the concrete is still soft and there are difficulties in the monitoring of shrinkage of a semi-solid material These difficulties

have hindered comprehensive the physical testing and understanding of the factors influencing very early age shrinkage (Holt, 2004) The amount of very early age shrinkage depends on many factors, including the properties of the material, temperature, relative humidity of the environment, and the size of the specimen Detail mechanism of very early age shrinkage will be discussed in the following sections

1.2 Mechanisms of early age shrinkage

Studies on very early age shrinkage of concrete have identified some mechanisms which induce or influence shrinkage at a very early age as,

gm of cement (about 10 % by volume of paste) (Neville, 1995) A recent study by Holt E

E (2001) showed that chemical volumetric shrinkage for a W/C 0.30 cement paste have

about 2 ml per 100 gm of cement after 24 hours of hydration as shown in Fig 1.1

Fig 1.1 – Volumetric total chemical shrinkage of cement paste with different W/C (Holt,

2001)

1.2.2 Autogenous shrinkage

In the literature, it is noted that different definitions have been used to define more or less the same phenomenon of autogenous shrinkage Jensen and Hansen (1996) defined autogenous deformation as bulk deformation of a closed isothermal cement past system JCI (1998) defined “autogenous shrinkage as the macroscopic volume reduction

of cementitious materials when concrete hydrates after initial setting” Autogenous shrinkage does not include volume change due to loss or ingress of substances,

temperature variation, application of an external force and restraint In another study,

Holt, 2004 defined autogenous shrinkage as an external volume change occurring with no moisture being transferred to the surrounding environment Jensen and Hansen (1996) mentioned that the unrestrained bulk deformation of a sealed cement paste at a constant temperature has been called as chemical shrinkage, bulk chemical shrinkage, chemical volume change, self desiccation shrinkage, autogenous deformation, autogenous shrinkage, autogenous volume change, endogenous shrinkage, and indigenous shrinkage

In this thesis, the term total chemical shrinkage was used to describe the reduction

in the absolute volume of the hydrated cement paste due to hydration of cement paste The term autogenous shrinkage was used to describe the bulk deformation of a sealed cement paste at a constant temperature For a cement paste system, the autogenous shrinkage would be smaller than the total chemical shrinkage With hydration of cement a skeleton is formed in the cement paste and the rigidity of the skeleton increases As a result, part of the shrinkage due to hydration is compensated by the formation of voids in the cement paste system Thus the hydrates could not shrink as much as the total chemical shrinkage.The autogenous shrinkage or bulk reduction in overall volume would

be lower than the total chemical shrinkage

In a study, Setter and Roy (1978) reported that about 0.7 % volumetric autogenous shrinkage was found for Type I cement paste with a W/C of 0.30 at the age of

20 hours (Fig 1.2) They also reported that the volumetric autogenous shrinkage was

high during the first hours of hydration, since the paste was not yet strong enough to withstand the stresses that are developed during the total chemical shrinkage

Fig 1.2 – Volumetric autogenous shrinkage of cement paste [Setter and Roy, 1978]

In another study, Justnes et al [2000] illustrated a comparison between chemical shrinkage and volumetric autogenous shrinkage immediately after concrete was mixed, and showed that those two measurements are reasonably similar during the first few

hours as shown in Fig 1.3 However after some time these curves began to diverge from

one another and this divergence occur at ages ranging from 1 hour to 3 hours after initial setting time depending on the cement type and W/C In a recent study, Mounanga et al (2006) illustrated a comparison between chemical shrinkage and volumetric autogenous shrinkage for a cement paste with W/C 0.25 They reported that the curves diverged less than 2 hours after mixing when the specimen was cured at a temperate of 50˚ C

The above mentioned volumetric autogenous shrinkage was achieved by measuring the volumetric change of the paste sample sealed in a rubber bag When tests are carried out using prism or slab specimens, a portion of the volumetric shrinkage will induce as horizontal shrinkage and other portion will register as vertical shrinkage or settlement

Fig 1.3 – Chemical shrinkage and volumetric autogenous shrinkage of cement paste with

water-to-binder ratio of 0.30 [Justnes et al., 2000]

1.2.3 Settlement

Settlement is generally defined as the downward movement of the constituents of the mix after compaction In parallel with the settlement due to downward movement of the constituents, vertical shrinkage due to hydration of the paste would also induce settlement of the paste during the very early age In literature, the volume shrinkage before the setting of the paste was attributed to the vertical shrinkage and no horizontal shrinkage was assumed to have occurred In a study, Holt E E (2001) demonstrated a comparison between the vertical shrinkage (settlement) and horizontal shrinkage during

the first 12 hours after casting as shown in Figure 1.4 The volumetric autogenous

shrinkage was calculated using simple equations, then correlated the volumetric chemical shrinkage and autogenous shrinkage The results showed that a large amount of horizontal shrinkage was registered in some specimens even before the setting time However, she could not demonstrate the very early age shrinkage quantitatively during that period due to the limitations of the measurement method used

Fig 1.4 – Horizontal shrinkage, vertical shrinkage, volumetric autogenous shrinkage and

volumetric chemical shrinkage of mortar with W/C of 0.30 comparing with chemical shrinkage [Holt, 2001]

1.2.4 Shrinkage due to moisture loss

When concrete is exposed to a dry environment, the drying shrinkage, deformation arising from loss of water from the concrete to the environment, occurs simultaneously with autogenous shrinkage The mechanism of drying shrinkage was explained by the capillary tension theory One of the main causes of drying shrinkage is the surface tension developed in the small pores of the cement paste of concrete When these pore water is loss, a meniscus forms at the air/water interface Surface tension in this meniscus pulls the pore wall inwards and the concrete responds to these internal force by shrinkage Thus, cracking of exposed surfaces usually attributed to plastic shrinkage, caused by insufficient protection is not uncommon in practice Aïtcin et al (2004) reported that the presence of bleed water prevents the development of plastic shrinkage However, high performance concrete with low W/C ratio has very little bleed water which could be quickly removed by the surrounding dry environment As a result, high plastic shrinkage may develop within the first few minutes after its placing As discussed previously, due to the measurement difficulty during plastic stage, quantitative measuring of the effect of drying on very early age shrinkage during plastic period is very limited in the literature

1.2.5 Thermal effect

Another factor affecting the early age shrinkage strain is thermal expansion due

to the heat of hydration of cement.Thermal expansion refers to the volume changes that occur when concrete undergoes temperature fluctuations During the early ages, concrete temperature is a function of the heat of hydration and climatic conditions The heat generated due to hydration results in a temperature rise in the concrete as a function of

the thermal conductivity and specific heat of the paste and aggregate The typical heat of hydration of OPC for the first 3 days after mixing would be about 265 J/g (Mindess et al., 2004) This thermal effect may be significant during the first 24 hours after water is first added to the concrete mix

The rise in the temperature in the mix is accompanied by an increase in the total volume, a process which occurs concurrently with the shrinkage of the hydrated cement paste Strains associated with a change in temperature depend on the magnitude of temperature drop or rise and the coefficient of thermal expansion of material The coefficient of thermal expansion of concrete also varies with time especially when concrete is in a fresh state accompanied by the effects of an increasing Young’s modulus with time (Turcry et al., 2002)

1.3 Very early age total shrinkage

In practice, the above mentioned mechanisms occur simultaneously and interact with each other in varying degrees depending on the mix proportions, environment and specimen configuration These mechanisms and their interaction with each other especially within the first 24 hours are difficult to study and model independently Relations relating chemical shrinkage, settlement, autogenous volumetric and horizontal shrinkage during the initial phase (before and during setting) of concrete are available but rather case specific Many different forms of shrinkage could develop simultaneously at a

very early age In this thesis, the term ‘very early age total shrinkage’ was used to

describe the sum of all possible types of length change including thermal expansion (when present) that may developed in the specimen measured starting just a few minutes

after casting

1.4 ‘Time-zero’ for early age shrinkage measurement

The starting time of shrinkage measurements is also not well established in the literature There is a lack of a standard that describe when shrinkage measuring should be started This gives rise to difficulties in comparing between different materials and in identifying the appropriate shrinkage strains (Weiss, 2003).The starting time of measurements would depend on the purpose of the study In general, distinctions could

be made based on whether a structural or a material study is carried out From the structural point of view, the term ‘Time-zero’ is generally defined as the time at which the concrete develops sufficient strength From that time on, tensile strains could be correlated to tensile stresses in the concrete if Young’s Modulus is known

The Japan Concrete Institute’s Technical Committee on autogenous shrinkage of concrete (1998) recommended time-zero for autogenous shrinkage measurements as the time of initial set excluding the strains generated within the concrete when in a fresh stage They stated that autogenous shrinkage is generally used for the prediction of stress related cracking in hardened concrete However, some researchers [Bentur A (2003)), Weiss (2003)], pointed out that the setting time measurement is rather arbitrary and strength gain should begin at or slightly after the time of initial setting So the initial setting time cannot be assumed to correspond precisely to the time-zero to be used for shrinkage measurements

Bentur A (2003) also stated that stresses will be generated when autogenous shrinkage starts to diverge from chemical shrinkage He defined time-zero as the timewhen self supporting skeleton develops and the material might be considered to become solid-like This, he pointed out, may be roughly around the setting time but

acknowledged that it is difficult to pinpoint this exactly Alternatively, Weiss (2003) stated that the early age hydration process and strength development are dependent on the curing temperature of the concrete Both the temperature and hydration mechanism could affect stress development in the concrete, and he recommended the use of maturity based concepts to predict time zero for prediction of stress in concrete

Measurements using embedded gauges (Zhang et al., 2003) may also provide an indirect indication of time-zero as no shrinkage would be recorded before the concrete gains sufficient strength However, there is no clear understanding of the strength level at which concrete stress would be related to strain In addition, strains measured using the embedded gages appear to be slightly lower than those measured using other methods especially during the early stage This has been attributed to the stiffness of the strain gauge As discussed there is no conclusive recommendation of a common starting point for monitoring autogenous shrinkage of concrete from the structural point of view in the literature However, from the material point of view, it is generally accepted that shrinkage measurement should be started just after mixing of the concrete Due to the limitations of the methods available, most of the measurement methods could not measure the shrinkage starting from that time (the details of measurement techniques will

be discussed in Chapter 3)

1.5 Effect of very early age shrinkage in the concrete structure

The effect of early age shrinkage will be more significant in concrete with low W/C than those with high W/C Since the gain in strength and Young’s modulus, Ec, of the former is faster It appears that such types of concrete are particularly sensitive to cracks before or at the onset of setting (plastic shrinkage cracking) or few days after the

hardening caused especially by early age shrinkage of concrete Holt (2004) reported that

if W/C ratio is lowered, the risk of cracking due to total early age shrinkage becomes greater Aïtcin et al (2004) also reported that for concrete with low W/C, autogenous shrinkage starts to develop very rapidly as soon as the hydrated cement paste becomes structured, particularly intensely during the first 24 hours

Another factor that could exaggerate the early age shrinkage of concrete is loss of moisture from the concrete.For a specimen moisture loss will be from the surfaces of the specimen, and that would cause higher shrinkage at the surface and lower shrinkage at the core of the concrete inducing differential shrinkage resulting in tensile stresses being generated If the tensile stress is higher than the tensile strength of the concrete, the surface concrete will crack The additional shrinkage in the long term will exacerbate

surface cracks

1.6 Effect of early age shrinkage on composite concrete structure

Similarly, in the case of composite constructions such as repair of deteriorated bridge decks; strengthening or renovating pavements or slabs-on-grade, moisture evaporates mainly from the top surface resulting in higher shrinkage at the top Unless the cementitious material used for new layer is shrinkage free, differential shrinkage strains may cause high stress concentration at the interface between the old substrate and the new material The new material layeris partially restrained from shrinking because it

is attached to substrate below which does not shrink as much as the new material, resulting in tensile stresses at and near the top surface Such a mechanism would cause premature failure of the overlay structure Long-term observations (Shah et al., 1992) of

manyoverlays have shown that cracking due to differential shrinkage is a very common problem

1.6.1 Tensile stress in the composite concrete

In such composite concrete systems, due to the differential shrinkage of the new material, the stresses developed will take the form of compression in the substrate and tension in the new material If the structural member is loaded in flexure, compression and tension zones will be formed and additional stresses will be presented in the member (Xu, 1999) These stresses reduce uniformly with the distance from the interface, and for

a number of combinations of material thickness and modulus of elasticity there will be stress reversals towards the outer layers of each material (Xu, 1999) Tensile strength can resist stresses caused by the restrained shrinkage of the repair material If the stresses exceed the tensile capacity of the concrete, cracking can occur at the top surface of the repair material or at the interface between the substrate and repair material A portion of the tensile stresses, however, can relieved by tensile creep and this may delay cracking in

the repair material

1.6.2 Debonding at interface of the composite concrete

Another form of premature failure that may occur in composite concrete is debonding especially near or at boundaries, normally at an early age Research showed that the durability of thin concrete repair is generally related to the durability of the bond between the old and the new concrete (Robin and Austin, 1995) The very early age shrinkage would affect the bond strength and hence the durability of the repaired system

A major cause of bond failure has been attributed to very early age differential shrinkage of the two materials Differential shrinkage results in internal stresses