Heritage - 2001 - Direct Electric Curing of Concrete and Mortars

DIRECT ELECTRIC CURING OF MORTAR AND
CONCRETE
laD Heritage B.EDg (HoDs), M.Se
A thesis submitted in partial fulfilment ofthe requirements of

Napier University for the degree of Doctor of Philosophy
School of the Built Environment

Napier University
Edinburgh, UK
2001
Abstract
Direct electric curing is the method by which the hydration reactions within concrete are
thermally accelerated by passing an alternating electric current through the sample. This
use of electricity as a means of supplying heat to young concrete and mortar has become
recognised as a more effective and energy efficient form of accelerating the curing
process than traditional steam curing.
The present investigation involves studying the affects of thermally accelerated curing
of cementitious materials in comparison to normally non-heated curing. Prior to
mixing, tests are performed on the constituents of the cementitious mixes. Results of
these tests are used in the creation of mix designs for mortar and concrete samples. A
range of concrete mixes are designed with and without additives and admixtures are
used to make cubes, slabs and beams.
From the testing results, the factors affecting the short and long term properties of
electrically cured cementitious materials are investigated and heating regimes are
presented to achieve specific properties of both strength and durability at specific ages.
A substantial section of reinforced concrete is required to gain representative results in
electrically curing reinforced concrete slabs. Compressive strength is difficult to
measure due to the electrical distortion affects of inserts and the damage caused by
coring so that at present, basic maturity concepts are used as a compressive strength
guide which limits the validity of results obtained. This research looks at refining these
concepts to include early age heating effects induced by direct electric curing.
The microstructural development of concrete when subjected to accelerated curing is

also considered. The affect of delay periods and maximum temperature is studied using
a scanning electron microscope and the results presented.
11
Declaration
This thesis is submitted to Napier University, Edinburgh for the Degree of Doctor of
Philosophy. The work described in this thesis was carried out under the supervision of
Dr. Fouad Khalaf, Dr. John Wilson and Prof. Alan Sibbald. The work was undertaken
in the School of the Built Environment, Napier University.
In accordance with Napier University regulations governing the Degree of Doctor of

Philosophy, the candidate submits this thesis as original unless otherwise referenced.
During this period of research the following papers have been presented and/or
published:
1. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Electrical curing of cement

based mortar'. Presented at the ih BCA Conference on Higher Education and the
Concrete Industry, Heriot Watt University, September 1997.
2. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Thermal acceleration of

hydration for mortars using direct electric curing'. Proc. 9th BCA Conference on
Higher Education and the Concrete Industry, Cardiff University, July 1999, pp. 103-
116.
3. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Thermal acceleration of

Portland cement concretes using direct electronic curing'. ACI Materials Journal,
Vol. 97, No.1, 2000, pp. 37-40.
4. HERITAGE, I. and KHALAF, F.M. 'Direct electric curing of concrete'. Proc.

lOth BeA Conference on Higher Education and the Concrete Industry, University of
Birmingham, June 2000, pp. 69-81.
Ian Heritage
111
Acknowledgements
The author acknowledges the assistance and encouragement of his supervisors. Dr.
Fouad Khalaf and Dr. John Wilson. Their helpful suggestions and technical advice
have greatly aided the author during his research. Additionally. their recommendations
and constructive discussions are much appreciated by the author.
The author wishes to thank Napier University for funding the project under an internal
research grant. Additional funding was provided by Napier University through payment
for the supervision and teaching of undergraduate students. The author thanks the
School of the Built Environment, Napier University for making facilities available for
the work. He is grateful to all School Support Staff members for their help with
laboratory testing, equipment repairs and administration assistance.
The author would also like to thank his friends and colleagues of the University,
especially Pauline Guo, who made his stay in Edinburgh such a unique experience.
The author dedicates this work to his parents, family and dog who have always given
him great support throughout his career. For Bailey.
iv
Contents
Abstract ii
Declaration iii
Acknowledgments iv
Contents v
List of Figures xii
List of Tables xxi
Chapter 1 - INTRODUCTION 1
1.1 History of Cementitious Materials 1
1.2 Accelerated Curing 3
1.3 Direct Electric Curing 4
1.4 Scope of the Investigation 5
1.5 Outline of Thesis 6
Chapter 2 - LITERATURE REVIEW 7

2.1 Introduction 7
2.2 Hydration of Cement 7
2.3 Accelerated Curing of Mortar and Concrete 9
2.4 Requirements for Accelerated Curing It
2.4.1 Economical considerations II
2.4.2 Parameters relating to materials 13
2.4.3 Sununary of requirements 14
2.5 Methods of Accelerated Curing 15

2.5.1 Steam curing 15
2.5.2 Autoclave 15
2.5.3 Microwave 16
2.5.4 Hot water 20
2.5.5 Hot air heating 20
2.5.6 Infra-red heating 20
2.5.7 Heated mix components 21
2.5.8 Additives and admixtures ..,.,
v
2.5.9 Special cements (RHPC) 22
2.5.10 Mass concrete 22
2.5.11 Climate heating 23
2.5.12 Electrical curing 25
2.5.13 Direct electric curing 26
2.6 Electrical Properties of Concrete 26

2.6.1 Introduction 26
2.6.2 Conduction paths 27
2.6.3 Electrochemical structure of cement paste 27
2.6.4 Electrolytic conductance 30
2.6.5 Ionic concentration and speed 30
2.6.6 Electrode/electrolyte interface 31
2.6.7 Polarisation 31
2.6.8 Temperature dependence 32
2.6.9 Resistivity 35
2.6.10 Impedance 36
2.6.11 Capacitance 36
2.6. I 2 Electrical properties as an indication of concrete properties 36
2.7 Direct Electric Curing 38

2.7. I Practical application of DEC 39
2.8 Heating Cycles 40

2.8.1 Heating period 41
2.8.2 Isothermal period 41
2.8.3 Cooling period 42

2.8.4 Curing cycle curve 42
2.8.5 Maturity 43
2.8.6 Thermal conductivity of concrete 48
2.8.7 Thermomechanical affects 49
2.9 Effects of Accelerated Curing on Mortar and Concrete 50

2.9.1 Fresh properties 50
2.9.2 Microstructure 50
2.9.3 C-S-H formation 50
2.9.4 Delayed ettringite formation 53
2.9.5 Microcracking 56
2.9.6 Pore structure 60
2.9.7 Compressive strength 63
vi
2.9.8 Durability 64
2.9.9 Setting times 64
2.9.10 Tensile. flexural strengths and modulus of rupture 65
2.9.11 Other properties 66
2.9.12 Effects of incorporating reinforcement 66
2.10 Additives and Admixtures 67

2.10.1 General 67
2.10.2 Accelerating admixtures 69
2.10.3 Retarding admixtures 69
2.10.4 Air entraining admixtures 70

2.10.5 Superplasticising admixtures 70
2.10.6 Pulverised fuel ash (PF A) 70
Chapter 3 - MATERIAL PROPERTIES AND TESTING PROCEDURES 74

3.1 Introduction 74
3.2 Constituent Materials Testing Procedures 74
3.2.1 Sieve analysis 74
3.2.2 Impact test 75
3.2.3 Density bottle 75
3.2.4 Speedy test 76
3.2.5 Vicat test 77

3.2.6 Loss on ignition (LOI) 77
3.2.7 Fineness 77
3.3 Constituent Material Properties 78

3.3.1 Coarse aggregate 78
3.3.2 Fine aggregate 78
3.3.3 Cement 79
3.3.4 Lime 80
3.3.5 Admixtures 80
3.3.6 Pulverised fuel ash 82
3.3.7 Water 82
3.4 Mix Design and Casting 82

3.4.1 Mortar 82
3.4.2 Concrete 83
3.4.3 Casting 83
vii
3.S Fresh Concrete Testing Procedures 84
3.5.1 Slump 84
3.5.2 Vebe consistometer 84
3.5.3 Air entrainment test 84
3.6 Hardened Concrete Testing Procedures 85

3.6.1 Compressive strength test 85
3.6.2 Splitting tensile test 85
3.6.3 Flexural tensile strength 87
3.6.4 Core samples 88
3.6.5 Initial surface absorption test (lSA T) 88
3.6.6 Porosity 91
3.6.7 Water absorption 92
3.6.8 Density 93
3.6.9 Ultrasonic pulse velocity 94
3.6.10 Schmidt hammer 95
3.6.11 Scanning electron microscope (SEM) 96
3.7 DEC Set-up 96

3.7.1 Equipment used 96
3.7.2 Cube casting and curing 97
3.7.3 Beam casting and curing 100
3.7.4 Slab casting and curing 101
3.7.5 Thermistor placement 102
3.7.6 Heating procedure 103
Chapter 4 - DIRECT ELECTRIC CURING OF MORTAR 105

4.1 Introduction 105
4.2 Water/Cement Ratio lOS
4.3 Heating Rates lOS
4.4 Resistance 109
4.5 Maturity 113
4.6 Maximum Temperature 114
4.7 Long-term Compressive Strength 115
4.S Durability 116
4.9 Isothermal Period 117
4.10 Delay Period 118
viii
4.11 Conclusions 119
Chapter 5 - DIRECT ELECTRIC CURING OF CONCRETE 122

s.i Introduction 122
S.2 Normally Cured Samples 122
S.3 Effect of Cover 123

S.4 Water/Cement Ratio 126
S.S Effect of Delay Period 133
S.6 Effect of Maximum Temperature 140
S.7 High Slump and High Strength Mixes 141
S.8 Resistance 144
S.9 Effect of Temperature on Resistance 146
S.10 Two Stage Heating 148
s.n Splitting Tensile Strength IS4
S.12 Flexural Tensile Strength IS7
S.13 Maturity lS8
S.14 Initial Surface Absorption Test 163
S.IS Porosity 166
S.16 Density 168
5.17 Water Absorption 168
S.18 Temperature Profiles 169
S.19 Scanning Electron Microscopy 177

5. I 9. I Microstructure I 77
5.19.2 Pore structure 178
5.19.3 Microcracking 179
5.20 Conclusions 186
Chapter 6 - DIRECT ELECTRIC CURING OF CONCRETE CONTAINING

ADMIXTURES AND ADDITIVES 191
6.2 Air Entraining Admixture 191
6.3 Superplasticising Admixture 200
6.4 Accelerating Admixture 202
6.S Retarding Admixture 203
ix
6.6 Pulverised Fuel Ash Addition 204
6.7 Conclusions 207
Chapter 7 - DIRECT ELECTRIC CURING OF CONCRETE CONTAINING

REINFORCEMENT 210
7.2 Normally Cured Slab 210
7.2.2 Core failure load 212
7.2.3 Temperature history 214
7.3 Electrically Cured Slab A 215

7.3.1 Visual inspection 215
7.3.4 Porosity 222
7.3.5 Density 213
7.3.6 Water Absorption 225
7.3.7 Temperature profile 225
7.4 Electrically Cured Slab B 226

7.4.1 Visual inspection 226
7.4.4 Temperature history 233
7.5 Conclusions 240
Chapter 8 - SUMMARY OF CONCLUSIONS AND FUTURE RESEARCH 243

8.2 Summary of Conclusions 243

8.2.1 DEC of mortar 243
8.2.2 DEC of concrete 244
8.2.3 DEC of concrete containing admixtures and additives 247

8.2.4 DEC of concrete containing reinforcement 248
8.3 Recommendations for Future Research 249
x
REFERENCES 252
Appendix A: Mix Design Sheets 267
Appendix B: Published Papers 273
xi
List of Figures
Page No.
Figure 2.1 Representation of the hydration products formed from

anhydrous clinker 8
Figure 2.2 Conduction paths through concrete 27
Figure 2.3 Schematic breakdown of concrete composition as relating to
conduction 29
Figure 2.4 Resistance versus temperature for metals and insulators 34
Figure 2.5 General temperature curve used for accelerated curing

methods 42
Figure 2.6 Dependence of rate of hydration of a cement paste on
temperature 47
Figure 2.7 Temperature functions of various formulae to predict
cement hydration rates 48
Figure 2.8 Localised damage caused by thermally induced stresses 58
Figure 2.9 Admixtures by type 68
Figure 3.1 Coarse and fine aggregate sieve results 79
Figure 3.2 Arrangement of splitting tensile test 86
Figure 3.3 Fracture planes of splitting tensile test 86

Figure 3.4 Arrangement of flexural tensile test 87
Figure 3.5 ISAT assembly 91
Figure 3.6 Set-up of ultrasonic pulse velocity meter 94
Figure 3.7 Scanning electron microscope set-up 96
Figure 3.8 Circuit diagram 97
Figure 3.9 Plastic cube moulds 98
Figure 3.10 Set-up for cube heating 99
Figure 3.11 Set-up for beam heating 100

Figure 3.12 Set-up for slab heating 101
Figure 3.13 Cube thermistor location 102
Figure 3.14 Beam thermistor location 102
Figure 3.15 Slab thermistor location (A) 103
Figure 3.16 Slab thermistor location (B) 103
Xl1
Figure 4.1 Compressive strength at 3 days versus W/C ratio for
electrically (2 hour delay, 30°CIhr rise, 80°C maximum) and
normally cured samples 105
Figure 4.2 Percentage and absolute gains of electrically cured (2hour
delay, 30°Clhr nse, 80°C maximum) specimens over
normally cured specimens for varying W IC ratios 106
Figure 4.3 Relationship between resistance and W IC ratio for
electrically (2 hour delay, 30°Clhr, SO°C maximum) cured
samples 107
Figure 4.4 Heating curve for mortar (W IC = 0.55, 2 hour delay,
40°Clhr rise, SO°C maximum) lOS
Figure 4.5 Percentage and absolute gains of differing heating rates (2
hour delay, SO°C maximum) on samples ofW/C = 0.57 109
Figure 4.6 Relationship between resistance immediately prior to
compressive strength testing and 3 days compressive
strength 110
Figure 4.7 Typical resistance graph (W/C = 0.55, 2 hour delay, 40°Clhr
rise, SO°C maximum) 111
Figure 4.S Corrected resistance versus time (W IC = 0.55, 2 hour delay,
40°C/hr rise, SO°C maximum) 112
Figure 4.9 Corrected resistance versus time (W/C = 0.55, 2 hour delay,
40°C/hr rise, 80°C maximum) 113
Figure 4.10 Values of maturity versus 3 days compressive strength for
samples ofW/C = 0.57 114
Figure 4.11 Effect of the maximum temperature on the 3 days
compressive strength (2 hours delay, 30°Clhr rise) 115
Figure 4.12 Effect of DEC (2 hour delay, 40°Clhr rise, SO°C maximum)
on long term compressive strength 116
Figure 4.13 Initial surface absorption of electrically and normally cured
samples at 28 days (W/C = 0.5, 2 hour delay, 40°C/hr rise,
80°C maximum) 117
Figure 4.14 Compressive strength after 3 days for isothermal and non
isothermal electrical (2hour delay, 40°Clhr nse, SO°C
maximum) curing lIS
Figure 4.15 Effect of delay period prior to heating on 3 day compressive
strength (W/C = 0.55, 40°Clhr rise, 80°C maximum) 119
Figure 5.1 Compressive strength versus time for normally water cured
samples 123
X11I
Figure 5.2 Compressive strength versus time for covered and
uncovered samples (1 hour delay, 40°CIhr rise, 80°C
maximum) 124
Figure 5.3 Resistance versus time for covered and uncovered samples
(1 hour delay, 40°CIhr rise, 80°C maximum) 125
Figure 5.4 Resistance versus time for covered and uncovered samples
(1 hour delay, 40°Clhr rise, 80°C maximum) 126
Figure 5.5 W/C versus compressive strength at 1 day (l hour delay,
40°CIhr rise, SO°C maximum) 127
Figure 5.6 Actual and percentage gain in compressive strength of
electrically (1 hour delay, 40°CIhr rise, SO°C maximum) and
normally cured samples at 1 day 127
Figure 5.7 Compressive strength versus W IC after 3 days air curing (1
hour delay, 40°CIhr rise, SO°C maximum) 128
Figure 5.S Compressive strength versus W/C after 3 days air curing (l
hour delay, 40°CIhr rise, SO°C maximum) 129
Figure 5.9 Actual and percentage gain of electrically cured over
normally cured samples for different W IC after 3 days air
curing (I hour delay, 40°Clhr rise, SO°C maximum) 130
Figure 5.10 Resistance versus time for different W IC (l hour delay,
Figure 5.11 Resistance versus time for different W IC (1 hour delay,
40°Clhr rise, SO°C maximum) 131
Figure 5.12 Corrected resistance versus time for different W/C (1 hour
delay, 40°Clhr rise, SO°C maximum) 132
Figure 5.13 Early age compressive strength (40°CIhr nse, 80°C
maximum) 133
Figure 5.14 Compressive strength versus delay period (40°Clhr rise,
80°C maximum, 1 day demoulding) 134
Figure 5.15 Percentage compressive strength achieved at different times
relative to 2S day compressive strength for various delay
periods (40°CIhr rise, SO°C maximum) 135
Figure 5.16 Compressive strength versus time for various delay periods
(40°CIhr rise, 80°C maximum) 136
Figure 5.17 Compressive strength versus delay period (40°Clhr nse,
80°C maximum, 3 day demoulding) 137
Figure 5.1S Compressive strength versus delay period (40°Clhr nse.
60°C maximum) 137
xiv
Figure 5.19 Percentage compressive strength achieved at different ages
relative to 28 day compressive strength for various delay
periods (40°CIhr rise, 60°C maximum) 138
Figure 5.20 Corrected resistance versus time for different delay periods
(40°Clhr rise, 80°C maximum) 139
Figure 5.21 Compressive strength versus time for 3 hour delay period
samples (40°CIhr rise, 60°C and 80°C maximum) 140
Figure 5.22 Compressive strength versus time for a high slump mix (60-
lS0mm) (l hour delay, 40°CIhr rise, SO°C maximum) 141
Figure 5.23 Compressive strength versus time for a high strength mix
(63N/mm2 target mean strength) (I hour delay, 40°Clhr rise,
SO°C maximum) 142
Figure 5.24 Resistance versus time for differing mix types (1 hour delay,
Figure 5.25 Resistance versus time for differing mix types (1 hour delay,
Figure 5.26 Compressive strength versus resistance at 1 day 145
Figure 5.27 Compressive strength versus resistance at demoulding for 3

day air cured samples 145
Figure 5.2S Temperature corrected and uncorrected resistance versus
time (1 hour delay, 40°CIhr rise, SO°C maximum) 146
Figure 5.29 Temperature corrected and uncorrected resistance versus
time (I hour delay, 40°Clhr rise, SO°C maximum) 147
Figure 5.30 Comparison of 1 and 2 stage heating regimes (3 hour delay) 14S
Figure 5.31 Comparison of I and 2 stage heating regimes (2 hour delay) 149
Figure 5.32 Compressive strength versus time for different delay periods
for two stage heating (15°Clhr rise for 1 hour + 40°Clhr rise,
80°C maximum) ISO
Figure 5.33 Compressive strength versus time for different delay periods
for 2 stage heating (40°CIhr rise for 1 hour + 20°C/hr rise,
SO°C maximum) 151
Figure 5.34 Percentage compressive strength gain versus time for 1 and
2 stage heating regimes (3 hour delay) 152
Figure 5.35 Corrected resistance versus time for 2 stage heating regimes
(3 hour delay) 153
Figure 5.36 Splitting tensile strength versus time electrically (3 hour
delay, 40°Clhr rise, SO°C maximum) and normally cured
samples 154
xv
Figure 5.37 Splitting tensile strength versus distance along beam at
differing ages 155
Figure 5.38 Percentage increase of splitting tensile strength at differing
ages 156
Figure 5.39 Comparison of percentage gain of compressive and splitting
tensile strength (3 hour delay, 40°CIhr rise, 80°C maximum) 157
Figure 5.40 Flexural tensile strength of electrically (3 hour delay,
40°CIhr rise, 80°C maximwn) and normally cured samples
at 28 days 158
Figure 5.41 Compressive strength versus maturity (normally cured) 159
Figure 5.42 Compressive strength versus maturity (electrically cured) 160
Figure 5.43 Compressive strength versus maturity (normally and
electrically cured) 161
Figure 5.44 Early age ($ I day) compressive strength versus maturity 161
Figure 5.45 Resistance versus maturity 162
Figure 5.46 ISA of electrically (3 hour delay, 40°Clhr rise, 80°C

maximwn) and normally cured samples at 28 days 163
Figure 5.47 Compressive strength versus ISA 10 164
Figure 5.48 Porosity versus ISA 10 165
Figure 5.49 Porosity versus time for electrically and normally cured
samples 166
Figure 5.50 Porosity versus compressive strength for electrically and
Figure 5.51 Density versus time for electrically and normally cured
samples 168
Figure 5.52 Water absorption versus time for normally and electrically
cured samples 169
Figure 5.53 Temperature (DC) colour key to Figures 5.54, 5.57, 7.16 and
7.27 171
Figure 5.54 Temperature cross section centrally perpendicular to
electrodes 172
Figure 5.55 Temperature distribution along centre of beam at a depth of
50mm (3 hour delay, 40°Clhr rise, 80°C maximwn) 174
Figure 5.56 Temperature distribution along centre of beam at a depth of
20mm (3 hour delay, 40°Clhr rise, 80°C maximum) 175
XVI
Figure 5.57 Cross section temperature map of the beam at the end of
heating (276 minutes) (3 hour delay, 40°CIhr rise, 80°C
maximum) 177
Figure 5.58 1380X magnification of normally cured concrete (1 day) 180
Figure 5.59 1190X magnification of electrically cured (40°Clhr nse,
60°C maximum, 3 hour delay) concrete (1 day) 180
Figure 5.61 3090X magnification of normally cured concrete (1 day) 181
Figure 5.62 3980X magnification of electrically cured (40°CIhr rise,
Figure 5.63 616X magnification of normally cured concrete (28 days) 182
60°C maximum, 3 hour delay) concrete (28 days) 183
Figure 5.65 lSI OX magnification of electrically cured (40°Clhr nse,
Figure 5.66 2440X magnification of electrically cured (40°CIhr nse,
Figure 5.68 267X magnification of normally cured concrete (l day) 185
Figure 6.1 Density versus ultrasonic pulse velocity at 28 days (0.5 hour
delay, 40°Clhr rise, 80DC maximum for 4 hours) 192
Figure 6.2 Ultrasonic pulse velocity for normally cured samples of
different percentages of air entrainment 192
Figure 6.3 Ultrasonic pulse velocity for electrically (0.5 hour delay,
40°CIhr rise, 80°C maximum for 4 hours)cured samples of
different percentages of air entrainment 193
Figure 6.4 Compressive strength versus ultrasonic pulse velocity for
electrically (0.5 hour delay, 40°Clhr rise, 80°C maximum
for 4 hours) and normally cured samples 194
Figure 6.5 Compressive strength versus time for electrically (0.5hr
delay, 40DCIhr rise, 80DC maximum for 4hrs) and normally
cured samples with 9.2% air entrainment 195
XVII
Figure 6.6 Compressive strength at 28 days versus percentage air
entrainment (0.5 hour delay, 40°CIhr rise, 80°C maximum
for 4 hours) 196
Figure 6.7 Compressive strength for normally and electrically (3hr
delay, 40°CIhr rise) cured samples with 5.0% air
entrainment heated to different maximum temperatures 197
Figure 6.8 Air entrained (5.0%) compressive strength versus time for
electrically (3hr delay, 40°Clhr rise, 80°C maximum) and
Figure 6.9 Resistance versus time for different percentages of air
entrainment (0.5 hour delay, 40°Clhr rise, 80°C maximum
for 4 hours) 199
Figure 6.10 Resistance versus time for air entrained and non-air
entrained samples (3hr delay, 40°Clhr rise, 60°C maximum) 200
Figure 6.11 Compressive strength versus time for electrically (3hr delay,
40°Clhr rise, 60°C maximum) and normally cured samples
with and without superplasticiser 20 I
Figure 6.12 Resistance versus time of electrically (3hr delay, 40°Clhr
rise, 60°C maximum) cured concrete samples with and
without superplasticiser 201
40°CIhr rise, 60°C maximum) and normally cured samples
with and without accelerator 202
Figure 6.14 Retarder compressive strength versus time for electrically
(3hr delay, 40°Clhr rise, 60°C maximum) and normally
cured samples 203
with and without PFA 204
Figure 6.16 Effect of delay period on compressive strength for 30% PF A
cement replacement (40°CIhr rise, 60°C maximum) 205
Figure 6.17 Resistance versus time for electrically (2hr delay, 40°CIhr
rise, 60°C maximum) cured samples with and without PFA 206
Figure 6.18 ISA for electrically (2hr delay, 40°CIhr nse, 60°C
maximum) and normally cured concretes containing PF A 207
Figure 7.1 Schmidt hammer number and 150mm cube compressive
strength versus time for normally cured slab 21 I
Figure 7.2 Slab Schmidt hammer number versus 150mm cube
compressive strength for normally cured samples 211
xviii
Figure 7.3 Core failure load and l50mm cube compressive strength
versus time for normally cured samples 212
Figure 7.4 Core compressive strength versus actual compressive
strength for normally cured slab 213
Figure 7.5 Temperature versus time for a normally cured slab 214
Figure 7.6 Average compressive strength versus time for electrically
and normally cured slab (derived from Schmidt hammer) 216
Figure 7.7 Average compressive strength versus distance from centre
reinforcement bar (derived from Schmidt hammer) 217
of slab parallel to reinforcement bars (derived from Schmidt
hammer) 218
Figure 7.9 Average core compressive strength versus time for
electrically and normally cured slab (derived from core
results) 219
Figure 7.10 Compressive strength at 56 days perpendicular to
reinforcement bars (derived from core results) 220
Figure 7.11 Compressive strength at 56 days (derived from core results) 221
Figure 7.12 Average compressive strength versus time (derived from

core and Schmidt hammer results) 222
Figure 7.13 Porosity over slab surface at 56 days 223
Figure 7.14 Density over slab surface at 56 days 224
Figure 7.15 Water absorption over slab surface at 56 days 224
Figure 7.16 Temperature profile along Slab A at 320 minutes after

gauging with water 225
and normally cured slab (derived from Schmidt hammer
results) 227
Figure 7.18 Average compressive strength versus perpendicular distance
from central reinforcement bar (derived from Schmidt
hammer results) 228
of Slab B parallel to reinforcement bars (derived from
Schmidt hammer results) 229
and normally cured slab (derived from core results) 230
Figure 7.21 Compressive strength at 28 days perpendicular to
reinforcement bars (derived from core results) 231
xix
Figure 7.22 Compressive strength at 28 days (derived from core results) 232
Figure 7.23 Average compressive strength versus time (derived from
core and Schmidt hammer results) 233
Figure 7.24 Temperature profile perpendicular to the reinforcement bars
at the centre of Slab B at a depth of 75mm 234
Figure 7.25 Temperature profile perpendicular to the reinforcement bars
at the centre of Slab B at a depth of20mm 235
Figure 7.26 Temperature history of differing depths at the centre of Slab
B and reinforcement bar 236
Figure 7.27 Temperature profile over time parallel to central
reinforcement bar 238
xx
List of Tables
Page No.
Table 2.1 Main compound composition of Portland cement 7

Table 2.2 Energy requirements of various accelerated curing methods 12
Table 2.3 Comparison of concrete resistance, permeability and ASTM
C1202 36
Table 2.4 Coefficients of linear thermal expansion 58
Table 2.5 Summary of the effects of PFA on the properties of
concrete 72
Table 3.1 Grading limits for coarse aggregates (BS 882) 74
Table 3.2 Grading limits for fine aggregates (BS 882) 75
Table 3.3 Sieve analysis of coarse aggregate 78
Table 3.4 Impact values of coarse aggregate 78
Table 3.5 Sieve analysis of fine aggregate 79
Table 3.6 Properties of cement used 80
Table 3.7 Properties of superplasticiser 80
Table 3.8 Properties of air entrainer 81
Table 3.9 Properties of retarder 81
Table 3.10 Properties of accelerator 81
Table 3.11 Classification of concrete quality on the basis of pulse
velocity 82
Table 3.12 Mix proportions of mortar 83
Table 3.13 Summary of mix proportions for 1m3 of concrete for
various mix designs 95
Table 5.1 Temperature of 150mm cube at various ages 170
Table 5.2 Temperature of beam at various ages 176
Table 7.1 Temperature of Slab A at various ages 226
Table 7.2 Temperature of Slab B at various ages 240
xxi
Chapter 1 - INTRODUCTION
1.1 History of Cementitious Materials

We may never know who made the first concrete since early attempts probably resulted
in a weak material which would long since have vanished. Concrete did not
immediately become the material as we know it today but rather developed over many
centuries (Stanley 1999).
The oldest concrete so far discovered dates from around 7000BC from Galilee, Israel. It
is a floor consisting of a lime concrete, made from burning limestone to produce
quicklime, which, when mixed with water and stone sets to form an early concrete.
Evidence of the use of concrete from 5000-2500BC is scarce but it is likely that it was
used in some form, but has either been destroyed, not discovered yet or degraded.
In around 1950BC the first known illustration of the various stages in the manufacture
and use of mortar and concrete was produced at Thebes in Egypt. At this stage concrete
was mostly used as an infill material for stone walls with only much later it being used
as a structural material.
From the Egyptians, the art of making concrete spread around the eastern Mediterranean
and by 500BC the Greeks used it as a lime based composition to cover walls of sun
dried bricks and as a binding material between brick and stone.
It is likely the Romans developed the idea of making concrete from the Greeks and
examples of early Roman concrete can be found dating to around 300BC. It is from the
Romans that we get the word 'concrete'. It derives from the word 'concretus', which
means grown together or compounded.
During the second century BC the Romans quarried a pink volcanic ash from near
Pozzuoli and upon mixing with lime found they had a much stronger concrete than
anything previously attained. During this time concrete had become readily accepted,
such that by 13BC, Vitrivius had compiled a pamphlet outlining the preparation of
concrete which was used throughout the Roman Empire.
Another advance by the Romans was the use of precast concrete. An example of this is
at the Italian port of Puteoli where blocks were left for two months before being sunk
into position to fonn a breakwater.
The use of concrete may have declined with the withdrawal of the Romans but examples
of Saxon concrete mixers remain from around 700AD. With the advent of the Norman
conquest concrete regained a measure of its former usage with it being used in such
buildings as the White Tower in the Tower of London, Dover Castle and Salisbury
Cathedral.
Until the 18th century there was very little use of concrete by the British with the
exception of a breakwater built in 1670 at Tangier in North Africa. This was built by
Henry Shere who experimented with over 30 different mixes until he selected the one
with the greatest strength underwater.
In 1756 John Smeaton was commissioned to build the third lighthouse on the Eddystone
Rocks, 14 miles south-west of Plymouth (Stanley 1999). To achieve the strengths
required Smeaton began the first modern scientific investigations into concrete and
published these fmdings in 'A narrative of the Eddystone Lighthouse' .
Portland cement was patented on 21 st October 1824 by Joseph Aspin. He named it

'Portland' because he thought it resembled Portland stone in colour and went on to
create several cement making factories.
2
The cement making process was improved by Isaac Johnson who raised the temperature
at which the cement was fired. Johnson is often regarded as the father of modem
Portland cement.
Portland cement quickly replaced 'Roman' cement in mortars and renders but was not
generally mixed with aggregates to make concrete until the mid 1800s.
Although superior to the cements of the time, modem day Portland cement bares little
resemblance to the original Portland cement except in name and basic ingredients and
since the mid 1800s concrete has generally increased in production and quality until the
present day were it is the major construction material in the World (Leung and
Pheeraphan 1995) with production in the range of 10,000 million tons per year (Idom
1990).
1.2 Accelerated Curing

Curing has been defined in many different ways. Two examples are (Cather 1994):
I. 'Curing is the creation of an environment in which hydration reactions can proceed

to help fulfil the aim of producing concrete of adequately low porosity'.
2. 'Curing is adequate when the resulting concrete achieves an expected service
requirement' .
Curing can be shown to have a marked effect on the rate of hydration of cement. The
particular performance of the concrete to resist deterioration needs to be considered

carefully in the light of the potential benefits of accelerated curing. The specification is
currently based upon vague evidence and is considerably influenced by a preconception
of requirements and what was previously considered adequate.
The term accelerated curing denotes a means of achieving rapid gain of compressive
strength at an early age. The techniques available include the use of heat, the
3
introduction of admixtures and, since the comparison of curing times is normally based
on OPC concrete, the use of high early strength cement (Kirkbride 1971).
In the manufacture of precast concrete, accelerated curing is used to increase

productivity (Kirkbride 1971). Benefits are derived not only from the more effective use
of plant and labour but also from shorter storage periods and therefore from reductions
in stock and stacking space. Less quantifiable benefits such as the facility to satisfy rush
orders at short notice are also relevant.
1.3 Direct Electric Curing

Direct electric curing of cementitious materials is the procedure by which an alternating
electric current is passed directly through a sample (acting as a resistor) to produce a
heating effect, thus increasing the initial rate of hydration.
The main advantages of the DEC process are (Bredankamp et al 1993):
1. DEC is one of the most energy efficient methods for the accelerated curing of
concrete.
2. The electrical power demand for this process may be during off peak hours when
cheaper electricity rates are available.
3. Due to high preliminary strengths the curing process is completed before the start of
the next shift.
4. After the initial capital outlay for equipment, the runnmg costs of DEC are
substantially lower than that of externally applied heat (steam, autoclave, etc.)
cunng.
5. DEC offers exact control of the curing cycle, which leads to minimal breakage's and
improved quality control.
6. As compared to normal curing methods, the reuse of shuttering can be increased by
2-3 times. This improvement in reuse increases during winter months.
4
1.4 Scope of the Investigation
Prior to mixing, tests were performed on the constituents of the cementitious mixes.
The results from these tests were used in the creation of mix designs for both mortar and
concrete samples.
Tests were performed on the fresh and hardened concrete and the performance of Direct
Electrically Cured mortar and concrete was compared to non accelerated cured samples
of the same mix designs. The effects of a number of curing and mix variables was
investigated.
The main objectives and scope of this study are as follows:
1. To review the current knowledge of accelerated curing and electrical aspects of
cementitious materials.
2. To study the physical properties of DEC and compare to non accelerated cured
cementitious materials.
3. To develop heating regimes to achieve specific properties at specific ages for DEC.
4. To study the microstructural development of electrically cured concrete.
5. To study the effect of admixtures, additives and reinforcement on DEC.
6. To achieve methods of predicting the mechanical properties of DEC samples.
7. To determine the feasibility of DEC in various circumstances and applications.
5
1.5 Outline of Thesis
The structure of the thesis can be summarised as follows:
Chapter 1 Introduction, scope and aim of the present investigation.
Chapter 2 Literature review of previous investigations into accelerated curing,
electrical curing and electrical properties of cementitious materials.
Chapter 3 Presents the results of an experimental investigation into the physical

and mechanical properties of the materials used and the testing
procedures adopted.
Chapter 4 Gives the results of an experimental and theoretical investigation of
electrically cured mortar samples.
electrically cured concrete samples.
electrically cured concrete samples containing admixtures and
additives.
electrically cured concrete samples containing reinforcement.
Chapter 8 Summary of conclusions and recommendations for future research.
6
Chapter 2 - LITERATURE REVIEW
2.1 Introduction
This chapter provides a review of previous work covering the accelerated curing of
mortar and concrete, the electrical characteristics of cement based materials, additives
and admixtures.
2.2 Hydration of Cement

The chemical combination of cement and water, known as hydration, produces a strong
binding medium for the aggregate particles in concrete and causes an exothermic
reaction producing heat (Czernin 1980, Ramachandran, Feldman and Beaudoin 1981,
Jackson and Dhir 1988, Bazant and Kaplan 1996).
The phenomena that occur during the early age of concrete curing are complex with
coupling between chemical-thermal-hydration and mechanical effects. Among the main
attributes that govern the non-linear behaviour of early age concrete are stiffness
evolution, the development of thermal strains, creep and cracking (Niu, Tu, Liang and
Zhang 1995).
The main compounds of Portland cement that react during hydration can be seen in
Table 2.1. The abbreviations stated in the table denote the notation commonly used
within cement hydration analysis. These compounds react at different rates and their
proportions seriously affect strength development.
Table 2.1- Main compound composition of Portland cement
Compound Formula Abbreviation

Tricalcium silicate 3Cao.Si02 C3S
Dicalcium silicate 2Cao.SiO, C,S
Tricalcium aluminate 3Cao.Al203 C3A
Tetracalcium alumino ferrite 4 CaO.AI2O).Fe,03 C4AF
7
The main hydration process commences at the boundaries of particles, and the gel which
is formed grows outwards from the surfaces of the particles towards the water filled
capillary space as well as inwards from the boundaries of the cement particles (Bazant
and Kaplan 1996). The continued formation of the hydration products result in a
stiffening of the cement paste which results in 'final set' after a period of 4-6 hours after
the addition of water (Neville 1988). The setting process is governed by the reaction of
C3S with sulphates which produces ettringite crystallisation creating an eventual
interlocking of crystals, thus causing the set (Skalny, Jawed and Taylor 1978).
Anhydrous clinker Hydration products
Colloidal
C-S-H gel products of
variable
composition
CzS slow
Ca(OH)~
Portlandite
CJA fast Mainly
but crystalline
retarded composition
C,.AF by Sulpho-ferri-
gypsum aluminate hydrates
Gypsum
Figure 2.1 - Representation of the hydration products formed from anhydrous

clinker
The Portland cement paste eventually becomes a network consisting of hydrates of the
various compounds present (C-S-H gel), crystals of Ca(OHh, minor compounds,
unhydrated cement and water filled voids pores. The pore fluid is strongly alkaline (pH
8
12-13) and originates mainly from the hydrates of calcium. sodium and potassium oxides
(Jawed and Skalny 1991). The products of hydration have a low solubility in water as
shown by the stability of the hardened cement paste in contact with water (Neville
1988).
A representation of the hydration products formed from the anhydrous clinker can be
seen in Figure 2.1. The areas of the boxes give the approximate volume proportions of
the phases (Double 1983).
2.3 Accelerated Curing of Mortar and Concrete

There is a widely held view that curing of concrete is not given the care and attention
that it requires (Cather 1994).
In concrete construction, economic considerations force builders to remove fonnwork as

soon as the concrete has gained sufficient strength to safely support its self-weight and
imposed loads. As such, it is often required that the usual release time is shortened by
the use of accelerated curing methods (Carino et al 1983).
The need to re-use equipment at frequent intervals, thus increasing daily production
rates, makes it necessary, particularly in prefabrication, to accelerate the chemical
reactions during concrete hydration.
A higher rate of strength development in concrete can greatly facilitate several important
processes in the construction industry. For example, in the development of precast

concrete products, it is desirable to attain sufficient strength within a short period of time
so the moulds can be reused and the final products can be handled and delivered to site
as soon as possible (Leung and Pheeraphan 1994).
Precasting of concrete has numerous advantages, relating to using accelerated curing

techniques, over in-situ construction. These include (Richardson 1973, Levitt 1982):
9
1. The opportunity to speed up the on site processes of construction by delivery of
already high strength members.
2. The reduction of skilled trade personnel and batching plant size on site.
3. Closer control of the quality of concrete used.
4. Greater specialisation of personnel.
5. Reduction of the need for scaffolding and other supporting services.
6. A greater degree of specialisation in products produced because of the added
flexibility of the greater range of construction techniques possible.

7. The maximum reuse of moulds and other equipment.
8. The reduction of delays due to adverse weather conditions.
9. The increase of production off site thus allowing greater planning control.
10. A reduction in site overcrowding with materials.
11. The availability of specialist techniques such as accelerated curmg IS greatly
increased, as is the possibility of humidity control.
Under normal moist curing conditions at room temperature, it can take as long as one
day or longer to achieve a strength that is sufficient for handling without damage and it
may take several days before significant loads may be carried.
A higher rate of strength development can greatly facilitate several important processes
in construction. In the case of repairing localised damage in pavements, the traffic must
be controlled until the repair material has gained sufficient strength. This is typically in
the region of 14 N/mm2 (Parker and Shoemaker 1987). To minimise the effect of repair
on traffic flow, the repair material should develop strength that is sufficient within five
hours so that any repair work can be carried out during off peak times of the day.
Several non-chloride admixture accelerators have been developed but their long term
performance is not generally well understood. Also many accelerators require precise
control of dosage which can make their application difficult or expensive to adjust.
10
The strength gain in Portland cement concrete can also be accelerated by the application
of thermal energy and during the past several decades, economic and technological
changes have greatly enhanced the attractiveness of accelerated curing, especially in the
precast concrete industry (Wadhwa, Srivastava, Gautum and Chandra 1987).
Steam curing is the most popular form of accelerated curing, but others, such as vacuum,
accelerating agents, hot water, microwave, infra red and electrical heating are available.
Steam curing involves hot water installations carrying low grade heat and is therefore
relatively uneconomical. The use of hot air and infra red emissions need special care to
avoid significant loss of moisture (Wadhwa et al 1987). Other forms of accelerated
curing have different advantages and disadvantages and are discussed in greater detail in
Section 2.5.
2.4 Requirements (or Accelerated Curing
2.4.1 Economical considerations
The selection of a suitable method of accelerated curing (Federation Intemationale De
La Precontrainte 1982) depends on the product to be cured, the performance expected
and on the manufacturing conditions (rate of production, floor space, staff skill and
equipment available), but the cost of the process is often the influencing factor. The
cost can be assessed by examining the following factors:
I. Cost of the energy brought into the mould

• The cost of the electricity, gas or other fuel delivered to site on the basis of a five
year period.
• The cost of the transformation of this energy into usable energy for the type of
thermal curing selected.
• The cost of equipment necessary for the transformation of the electrical energy
into usable energy (boilers, pipes, insulation, transformers etc.)
11
• The cost of maintenance of the transformation system and the cost of transfer of
the energy within the plant.
• The cost of transfer of this usable energy from the transformation place to the
mould
2. Minimising the losses of energy
• The optimisation of the energy distribution circuits (reduction of paths)
• The thermal insulation of circuits and moulds
• Temperature regulation
3. Concrete preparation for thermal curing
• Reducing the quantity of water
• Using the heat of hydration to aid the accelerated curing
• Replacing the precuring phase at ambient temperatures by a precuring phase with

a slightly increased temperature in the case of a lower than required ambient
temperature
A main concern in any accelerated curing process is the energy required to cure a given
mass of concrete. The approximate values of energy consumption can be seen in Table
2.2 (Wadhwa et al 1987).
Table 2.2 - Energy requirements of various accelerated curing methods
Type of accelerated Energy (kW.b) Initial cost of Maintenance cost of

curing consumed per mJ of equipment equipment
concrete
Steam 160-180 High High
Infra red 120-160 Moderate Moderate
Hot water 140-160 Moderate Moderate
Electrical curing 40-50 Low Negligible
(Semi-direct)
Electrical curing 30-60 Very low Negligible
(DEC)
The table clearly shows that the initial cost of equipment and maintenance for DEC is
very low compared to the other types of accelerated curing.
12
2.4.2 Parameters relating to materials
(a) Cement
For Portland cements without additives, the tricalcium aluminate (C3A) content appears
to be important for the suitability of a cement in accelerated curing. An optimum value
for C3A lies between 9-10% (Federation Internationale De La Precontrainte 1982). A
high tricalcium silicate (C3S) content is advantageous, especially when C3S:C2S is
greater than 3. Also, a relatively high C3A content must be accompanied by a high S03
content, but this must not exceed 3-4%.
(b) Aggregates
As aggregates do not take a significant part in the chemical reactions during hydration,
the suitability for accelerated curing must be examined only as due to their thermal
properties. Such properties that must be considered are thermal expansion, thermal
conductivity, specific heat and diffusion.
Also, care must be taken when using lightweight aggregates due the extra percentage of
voids which requires a reduction in heating rate and maximum temperature.
(c) Water
Care must be taken to avoid excessive evaporation by the use of suitable heating
regimes and enclosures to allow adequate humidity in the environment of the concrete.
(d) Admixtures
The use of admixtures must be considered very carefully when in association with
accelerated curing as some admixtures alter properties at higher temperatures and as
such preliminary mixes should be carried out.
13
2.4.3 Summary of requirements
Studies (Hanson 1963 and Higginson 1961) have previously shown that high early age
compressive strengths can be achieved when expenditure of the energy used was not a
concern.
The specific factors that could influence the development of an energy efficient, high
early and long-term strength, curing technique are:
1. The cement should have a C3S content of 55 to 65 percent.

2. The cement should have a C3A content of8 to 14 percent.
3. The fineness of the cement should be as high as practicable.
4. The loss on ignition should be as low as possible.
5. The water cement ratio should be between 0.30 and 0.40.
6. The water reducing admixture dose and the time when added require consideration.
7. The initial temperature of the fresh concrete should be between 22 and 38°C and
should be kept between these limits during the initial delay period.
8. Concrete slump should be as low as possible.

9. The delay period should be set as close to the initial setting time as possible.
10. The rate of temperature rise should be between 22 and 44°Clhr.
II. The maximum temperature should be between 71 and 82°C.
12. Appropriate and efficient thermal insulating barriers should be used to minimise
heat losses.
13. The applied heat can be turned off after several hours so long as efficient insulators
are used.
It has been stated (Pfeifer and Landgren 1982) that a difference in the maximum
temperature attained does not greatly effect the level of strength development at both
early and long term ages. It has also been shown that increases in concrete strength was
not dependent upon the length of the heating period with 3 hours heating showing only
an 8 percent loss when compared to a heating period of 10 hours.
14
2.5 Methods of Accelerated Curing
2.5.1 Steam curing
The reduction of the curing period in the precast concrete process is an important factor
for the productivity, capital saving and the reduction of workshop area. The traditional
thermal curing technique has long been used in the concrete industry. Steam at 90-
100°C is introduced into a chamber to heat up the precast concrete pieces. It can take
more than 10 hours to complete a curing cycle before demoulding. Research shows that
the quality of concrete is generally lower as the compressive strength after 28 days is on
average 10-15% lower than for concretes under normal curing conditions (Xuenquan,
Jianbgo and Mingshu 1987).
Conventional heating techniques, such as steam curing, rely on the conduction of heat
from the exterior to the interior (Leung and Pheeraphan 1994). The heating is therefore
non uniform and a long heating period is required to attain the required temperature.
Steam cunng and autoclaving are widely used in the precast concrete industry.
However, these require the use of expensive equipment and long processing time. Also,
the long term strength is adversely affected. This loss is attributed to the non uniform
formation of hydration products when reaction rates are high such as when temperature
is increased.
2.5.2 Autoclave
With the use of autoclaving (high pressure steam curing), specimens treated at 75°C and
90°C exhibit, even after only 8 hours curing, strengths higher than those attained at 25°C
and by comparison, for the specimens treated at 60°C, 15 to 36 hours are required to
achieve analogous results (Marcialis, Massidda and Sanna 1983).
If autoclaving is used in a mix containing PFA with prolonged hydration a gradual

combination of the calcium hydroxide with fly ash takes place.
15
The increase in temperature by autoclaving from 60°C to 75°C results in a reduction in
porosity whilst that from 75°C to 90°C gives rise to an increase in total porosity and a
shift of the pore sizes towards the coarser class sizes (Marcialis et al 1983).
2.5.3 Microwave
Microwave energy can heat materials that have a high dielectric constant, such as
concrete, both rapidly and uniformly.
Microwave energy generates heat in cementitious materials due to the dielectric nature
of the water contained within the mix (Sohn and Johnson 1999).
Microwave energy can heat a specimen uniformly and volumetrically, being

independent of the thermal conductivity of the specimen. It can also more easily
enhance the evaporation rate and allows better control of the energy absorption, thus
optimising the heating process.
Excessive microwave energy can lead to a long term decrease in strength due to over
evaporation and overheating.
Conventional thermal energy accelerates curing and reduces the induction period of
OPC mortars. Using steam, the samples must be cured in water in order to avoid
reducing the ultimate strength of the concrete at later ages. With steaming, heating
occurs at the edge of the sample causing thermal gradients which can lead to cracking
(Hutchinson, Chang, Jennings and Brodwin 1991).
With microwave curing the amount of energy must be carefully calculated: too much
energy and the water boils while too little has no effect (Hutchinson et al 1991).
16
Compared with conventional thermal curing, the most distinct features of microwave
curing are that the curing period is only 15-30 minutes, much shorter than for steam
curing and the strength of concrete, especially in the case of early ages, is improved
(Xuenquan et al 1987).
The permeability of microwave treated mortar is much lower than for untreated ones,
indicating that the former is denser in microstructure. Part of the free water could be
removed prior to setting during the short term microwave treatment, resulting in plastic
shrinkage of the mortar slurry and reducing the porosity (Xuenquan et aI1987).
When the starting time of the microwave treatment is delayed, the effect of the
microwave curing is reduced. For example, if the starting time is delayed to 45 and 120
minutes after mixing, the compressive strength after three days decrease to 90% and
83% by comparison to that of the specimen treated just after mixing. It is believed that
the network structure of cement is formed in this period, so that the effect of plastic
shrinkage is next to impossible under microwave treatment and the porosity of concrete
can not be reduced (Xuenquan et al 1987).
There are no distinct differences of the morphology between a specimen treated with or
without microwave energy as found by scanning electron microscope observation.
Unlike steam cured specimens, the 28 day compressive strength of microwave cured
specimens is also increased. The increase in later age strength can be attributed to the
removal of water from the fresh concrete when energy is applied, causing the collapse of
the capillary pores and subsequent density increase of the concrete (Leung and
Pheeraphan 1994).
The strength of microwave cured specimens have been found to be slightly higher than
that of conventionally cured control specimens. Microwave application can significantly
increase the degree of hydration during the first day, after which the microwave cured
17
specimens and the control specimen attain similar degrees of hydration (Leung and
Pheeraphan 1994).
For concrete specimens treated at lower microwave energy levels, accelerated cured
specimens and conventionally cured specimens exhibit almost identical strength at 7
days. Also, for all microwave energy levels, the 4.5 hour strength of concrete specimens
are higher than that for corresponding mortar specimens (Leung and Pheeraphan 1994).
It has been found (Leung and Pheeraphan 1994) that the most important parameter to
control was specimen temperature during the curing process using a feedback
temperature controller.
Previous studies (Leung and Pheeraphan 1994) have been uncertain of the thermal
history of the specimen during the accelerated phase of the curing as power, rather than
temperature, was usually used as a control.
Sohn and Johnson (1999) used a 30 minute delay with a 7°C/min rate of temperature
rise with a 2 hour isothermal period to heat their specimens. They found that with the
procedures used, overheating did not occur and as such, an isothermal curing phase
could be achieved using microwave curing. They also found that at 40°C maximum
temperature the strength values achieved were similar to normally cured mortars.
Compressive strength at 28 days reduced to 80% of the normally cured samples for a
maximum temperature of 60°C and at a maximum of 80eC the 28 day strength was
found to be only 50% of the normally cured specimens.
A reduced isothermal period was attempted and it was found that the 28 day strength of
the accelerated cured specimens could be maintained if the isothermal period of the
60eC maximum specimen was reduced to 48 minutes. The 80°C specimen always had a
reduced 28 day strength when compared to normally cured samples and the 40eC
maximum had a comparable strength with 120 minutes isothermal period.
18
Microwave curing allows a more uniform distribution of heat within the sample and as
such much less process time (less than one hour) is sufficient to provide high early
strength. When the process is applied before the concrete has set, the removal of water
causes the capillary pores to collapse which consequently densifies the concrete.
The effect of microwaving concrete on microcracking and pore structure has also been
investigated (Leung and Pheeraphan 1995). This was achieved by painting sections
with black paint and applying white zinc oxide which reveals pores and microcracks.
From this it was revealed that microwave cured samples generates many more pores and
microcracks. This is likely to be due to localised boiling of water leading to bubble
formation. However, in concrete, this effect seems to reduced when compared with an
equivalent mortar. This may be due to the presence of large aggregate particles in
concrete which restrain the expansion of the mortar which reduces the amount of
microcracks and number of pores thus improving strength.
It is clear (Leung and Pheeraphan 1994) that microwave heating generates many more
pores and microcracks in mortar specimens. One possible explanation for the improved
performance of microwave cured concrete over mortar specimens can then be given as
follows. During microwave heating, differential expansion of the aggregate and cement
gel and air bubble expansion due to localised boiling causes thermal stresses and can
lead to the formation of microcracks and pores. In concrete, where the presence of
aggregates restrains the expansion of the mortar, the amount of microcracks and pores is
reduced and the strength is comparatively improved.
It has been concluded (Hutchinson et al 1991) that microwave energy acts as an

accelerator only during the first 24 hours and that the induction period of the treated
specimens is significantly shorter. Also, it has been concluded that microwave heating
does not significantly influence compressive strength at 7 or 28 days.
19
2.5.4 Hot water
Hot water curing involves the input of heated water directly into the dry mix or
immersion directly after mixing to increase hydration reactions. No further heating is
generally used. This immersion method is covered by BS 1881: Part 112 where it is
used as an accelerated method of determining ultimate compressive strength. This
procedure uses water at temperatures of 35, 55 and 82°C.
It has been shown (Jonasson 1985) that when the mixing water is heated, this increases
the relative rate of hardening within concrete.
2.5.5 Hot air heating
Heating of concrete contained within shelters in Canada since the 1950s typically uses
two methods (Turenne 1985). These are circulation of hot air around the specimen at
the same pressure as the outside air and the pressurisation of the shelter by forcing hot
air from outside. Both methods can easily cause large differences in heating
distribution.
2.5.6 Infra-red heating

Competition in the construction industry and the rise of energy costs have forced the
need to perfect concrete casting methods. This has lead to shorter casting times and
more effective methods of heating concrete.
Infra-red heating developed at the end of the 1950s and this method has the advantages
of speed, directability and favourable heating costs.
It has been shown that for heating concrete, radiators with a radiation temperature of
800-900°C should be used (Penttala 1985).
20
It has also been shown (Penttala 1985), that for infra-red heat curing, the maximum
temperature of the concrete should be between 40°C and 60°C. At 60°C the long term
strength loss was shown to be between 25 and 30%.
2.S.7 Heated mix components
The heating of concrete has traditionally been carried out after the addition of water but
it is possible to heat components of the mix, or the mix itself during mixing. This
procedure has been used in Finland since the early 1980s (Kukko 1985).
The temperatures used in this method range from 40°C to 60°C because temperatures in
excess of 60°C cause a rapid loss of strength and workability difficulties.
It has been shown that with the use of preheated constituents the effectiveness of a
plasticiser is severely reduced.
It has been shown that the shrinkage of preheated concrete is approximately the same as
for ordinarily cured concrete of the same proportions. Important with respect to
shrinkage is the need to increase the water requirement of the mix for a given
workability. The risk of cracking is mainly dependent on the temperature difference
between the concrete and ambient air at the demoulding age. In many cases, the
difference between concrete temperature and ambient air is smaller than when concrete
is heated on site.
When usmg this method, it has been recommended (Kukko 1985) that the most
important tasks are to:
1. Ascertain rapid concreting and rapid heat insulation

2. Plan additional heating of the edges of members
3. Plan a temperature monitoring program
21
4. Plan for changes in ambient conditions
Heat insulation of 3-5cm of mineral wool has been recommended (Kukko 1985) for
wooden forms and 5cm for steel forms. However, special attention must be made to the
tightness of any insulation used and extra insulation should be used to protect exposed
corners of the heated sample.
2.5.8 Additives and admixtures
Accelerating admixtures increase the rate of strength gain in young concrete. They are
generally either calcium chloride (where no reinforcement is used) or calcium formate
(Neville 1988). Accelerating admixtures are covered in greater depth in Section 2.10.
2.5.9 Special cements (RHPC)
Rapid hardening Portland cement causes accelerated strength gain in concrete made with
it in place of ordinary Portland cement. This is due to both the increased fineness of
RHPC and the high C3S content which rapidly reacts with water during early hydration
(Neville 1988).
2.5.10 Mass concrete
Temperature rises in mass concrete can be up to 80°C. This can lead to maturity
differentials and corresponding strength variances.
For Portland cement mortars (Owens 1985) at 20°C the amount of Ca(OHh will
increase with time, indicating further hydration and strength gain but with an increase in
internal heat, the amount of Ca(OH)2 increases, but the strength falls. At 11 days there
can be as much Ca(OH)2 present as there is at one year at 20°C.
22
This reduction in strength would also imply a possible increase in permeability and an
abandoning of the usual relationship between permeability and water/cement ratio.
2.5.11 Climate heating
The strength of concrete in hot climates has been proved to adversely affected (Jalali and
Abyaneh 1995). The fmal strength is known to decrease with increased initial curing
temperatures even when other factors such as desiccation are precluded.
Physical causes for this effect are believed to be due to the large differences m
volumetric expansion of the different concrete constituents. The large increases In
volume of air and water with high temperatures creates internal stresses in the concrete.
If the tensile strength of the concrete is not sufficient to withstand these forces then
porosity increases and microcracking may occur which results in a lower long term
strength. The physical effect seems to be related to relatively high temperatures with
microcracks being observed at 85°C but not at 60°C (Jalali and Abyaneh 1995).
Possible chemical causes for the loss of strength could be that the chemical composition
of the hydrates is affected by the curing temperature, the microstructure of the hydrates
is partially changed or the degree of hydration is altered (Jalali and Abyaneh 1995).
Problems associated with hot weather concreting can include increased water demand,
increased rate of slump loss and setting, difficulty in pumping concrete, lower long term
strength and increased shrinkage due to an increase in the water content of concrete
(Gaynor, Meininger and Khan 1985).
It has been shown (Gaynor et al 1985) that were curing after placement was favourable
and prolonged mixing was not involved, the ultimate strength disadvantages In
increasing the concrete temperature from 23 to 38°C was less than 10%.
23
It has also been concluded that higher mixing temperatures (35°C) have little effect on
the drying shrinkage and that for an increase in mixing temperature from 18 to 35°C an
increase of 4.7kg/m3 of cement is required to maintain strength levels. Also the water
required to achieve a target slump has to be increased by around 2-6% and that strength
loss at 28 days is around 10%.
Hydration at low temperatures provides sufficient time for the hydration products to
diffuse in the interstitial space among the cement grains in a relatively uniform manner.
When accelerated climate curing occurs the hydration products tend to be diffused in a
non uniform way. This non uniformity results in a reduction in strength due to weak
zones in the matrix. Further to this, the high concentration of hydrated products in the
zone surrounding the cement grains retards subsequent hydration. This non uniformity
might be expected to decline as time passes with gradual diffusion of the hydration
products but the effect of increased initial curing temperature on C3S has been shown to
not be dissipated after 5 years of curing at 21°C. The effect of temperature on final
strength is probably a combination of its effects on a variety of physical and chemical
parameters (Jalali and Abyaneh 1995).
Some possible adverse effects of hot weather concreting include (Berhane 1992):
1. A reduction in setting time which creates difficulty in handling and finishing fresh
concrete.
2. Thermal and plastic shrinkage which may initiate and accelerate other deterioration
processes.
3. Reduced ultimate strength
4. Enhanced permeability and reduced durability due to macro and microcracking.
5. Formation of cold joints.
The ACI Manual of Concrete Practice (1992) refers to a maximum ambient temperature
of 32°C for the production of good quality concrete.
24
When concreting is carried out in hot weather, the strength of the concrete is adversely
effected by the early age heat increasing the rate of hydration of the cement which
causes a thicker than usual coating around the cement grains and also by the non
uniform distribution of hydration products within the cement matrix. Reduction in
strength also occurs due to the reduction of water available for hydration due to rapid
evaporation (AI-Gahtani, Abbasi and Al-Amoudi 1998).
It has been shown (Idorn 1990) that in the Middle East the temperature of the interior of
a concrete mix can be as high as 87°C after twelve hours were as at the comer of the
same mix the temperature is only 38°C. This is with an initial fresh concrete
temperature of 37°C. This led to early cracking of the structure even though laboratory
tested cubes from the same batch matched required specifications.
2.5.12 Electrical curing
Several methods of electrical curing have been tried (Kirkbride 1996). Internal methods
using either the reinforcement or special coils of wire as the heating element do not
appear to have enjoyed commercial success in the United Kingdom. The most
successful form of electrical curing element is the heating panel incorporating concrete
bonded to a steel plate. Electric heating cables have also been incorporated in concrete
beds. In general, with electrical curing, enclosures are necessary to retain moisture.
Electrical curing can be of an internal method (Wadhwa et al 1987). In the semi direct
method, either comparatively cheap electrical elements are embedded in the mass of the
concrete, or the ends of the reinforcement is connected to an electrical power source and
reinforcement is used as a direct heating element. Limitations of these methods include:
1. Loss of electrical heating elements in the mass of the concrete.

2. Difficulties in achieving a uniform temperature gradient in thick products.
25
Electrical curing is now being put to use in larger scales in the precast concrete industry
but has little usage in situ.
2.5.13 Direct Electric Curing
Direct electric curing is the process through which electricity is passed through concrete
or mortar in order to raise the temperature. DEC is covered in more depth in Section
2.7.
2.6 Electrical Properties of Concrete
2.6.1 Introduction
The electrical conduction through concrete can be divided into three parts:
1. Macroscopic being made up of conduction through the partially saturated aggregate
particles.
2. Microscopic involving the conduction through the hydrating cement phase.
3. Submicroscopic involving the electrolytic aspects of the aqueous phase of the
concrete.
(a) Macroscopic
The resistance of normal aggregates when compared to the cement paste is
approximately 100 times greater (Monfore 1968). The majority of current flow is
carried by the aqueous phase of the cement paste.
(b) Microscopic
Ohmic resistance is caused by the restriction of current flow through the hydrated
microstructure of the hydrating cement paste. Previous (Monfore 1968) conductivity
tests have shown that the conductivity of wet samples is approximately 10,000 more
than the equivalent oven dried samples which means that the solid phase of concrete can
be considered to be an insulator when dry.
26
2.6.2 Conduction paths
The paths of conduction stems from the principle that ionisation of water will conduct
electricity. A concrete mix will provide such a medium because of its heterogeneous
nature. There are three possible electrical paths through a fresh mix. Firstly, the
electrical current may pass through the cement paste only. The second path is the
converse, where the current passes through the aggregate only, this occurs where there is
physical contact between the aggregate particles. The third possible path is through
both the aggregate and the cement paste. These paths (McCarter and Curran 1984) are
shown in Figure 2.2.
a) Path through b) Path through c) Path through

cement paste aggregate only all constituents
Figure 2.2 - Conduction paths through concrete
2.6.3 Electrochemical structure of cement paste
In general (Farrar 1978, Sriravindrarajah and Swamy 1982) it has assumed that purely
resistive effects are present in concrete. The cement paste has a controlling influence
upon the overall electrical properties of concrete. Other work (McCarter and Afshar
1985) has used a combination of capacitance and resistance effects to model the
electrical properties of concrete. Previous models have assumed that conduction in a
cement paste is purely ionic through evaporable water in the continuous water filled
capillaries within the paste.
27
In most fine grained materials (such as cement) there are free unbound charges in the
aqueous phase and electrostatically held charges adjacent to the surfaces of the
individual particles. The amount of charges and the strength by which they are held
depends on:
1. The net electrical charge in the particle itself

2. The particle surface texture
3. The number of unsatisfied surface bonding sites
When an alternating current is applied, the free unbound charges within the aqueous
phase give rise to ionic conduction and bound particles oscillate with varying
amplitudes. The amplitude at which these oscillate is dependent on:
1. Type of charge
2. Degree of association with the surface of the particle

3. Temperature
4. Strength of the applied electric field
5. Frequency of the applied electric field
The electrochemical properties of cement pastes are time dependent and can be divided
into unhydrated cement grains, hydration by-products and the aqueous phase. A
summary of this can be seen in Figure 2.3.
Unhydrated cement particles and the solid products of hydration can be said to be non
conductive and as such, the only means of conduction of electricity through the
hydrating cement paste is through the aqueous phase which acts as an electrolyte due to
the presence of ionic compounds. This is shown by the fact that when concrete is oven
dried it can be classed as an insulator with a resistance of around 2-3x 103nm (Farrar
1978).
28
I Concrete I
I
Coarse aggregate particles in
mortar mix
I
I Mortar I
I
Fine aggregate particles in
cement paste
I
I Cement paste
I
I I
1
I Capillary pores
I
Cement gel
I Dry I I
Ions in solution
I I CSHgel I r
I Gel pores 1
I I
I
I I Unhydrated cement
I Disconnected
I Continuous
I particles
Figure 2.3 - Schematic breakdown of concrete composition as relating to

conduction
Because the aqueous phase acts as an electrolyte, concrete exhibits a negative
temperature coefficient of resistivity which is particular to electrolytic materials and as
such the resistance must be corrected for by a temperature coefficient of resistivity
(Wilson and Gupta 1996).
Electrical conduction through a damp cement paste is achieved by the mobility of ions
in the evaporable water content. The aqueous phase is initially saturated with Hydroxyl
and Calcium ions which mainly contribute to the conduction process. Ions such as
Sodium, Potassium, Aluminium, Silicates and other oxides (though highly soluble)
occur, but exist only at low levels of concentration, thus not significantly affecting
conduction (Neville 1988). Also, during the initial 24 hours after gauging with water,
29
Sulphates are present, but these precipitate into Sulphoaluminate hydrate and as such
can be disregarded.
2.6.4 Electrolytic conductance
Electrolysis can be defined as the changes in the chemistry of a liquid medium caused
by the passage of an electric current. Conduction is only possible in electrolytes where
at least some of the liquid contains oppositely charged ions. The electrodes through
which current enters and leaves the electrolyte are known as the anode and cathode, with
positive ions (cations) flowing towards the negatively charged cathode and negative ions
(anions) flowing towards the positively charged anode.
At the surface of both of the electrodes the methods of conduction change. Ions
discharge causing the substance concerned to come out of solution and are either
deposited on the electrode or liberated as bubbles of gas. Also fresh ions are formed
from the material of the electrode and passed into the electrolyte. This reaction causes
interference in the measuring of conductance and as such necessitates the use of high
frequency alternating current so as to reverse the minimal electrolysis taking place every
half cycle.
2.6.5 Ionic concentration and speed
As an ionic compound dissolves it splits into ions with the more dilute the electrolyte,
the more complete the ionic dissociation. As the solution becomes more concentrated
less ions split from the compound. Charge may be transferred at a greater rate at higher
levels of ionic concentration.
The transfer of charge is greater the faster the ionic speed. The ionic speed is dependent
on:
1. Electric field density. An ion experiences a force attracting it to the appropriate

electrode in proportion to the electric field density.
30
2. Viscosity of the solvent. With increasing viscosity the terminal velocity of the ion is
decreased.
3. The effect of asymmetry. A charged ion will tend to attract the ions of an opposite
charge and repel ions of a like charge. This reduces the speed of the ion as its
corresponding cloud of oppositely charged ions move in the opposite direction but
still attract the original ion causing a net reduction in velocity.
4. Electrophoretic effect. Momentum is transferred from the moving Ions to the
solvent and is also exchanged between ions of opposite charges as they travel in
opposite directions. This causes a greater loss of momentum than the solvent by
itself would otherwise exert.
2.6.6 Electrode/electrolyte interface
The charge of the electrode attracts the oppositely charged ions within the liquid
resulting in a layer of ions predominantly composed of either positive or negative
charges. This layer adjacent to the electrode in tum attracts ions of a specific type.
Thus, on either side of the interface layer are charges of the opposite charge. Within the
bulk of the sample ions are randomly distributed.
2.6.7 Polarisation
(a) Electrode polarisation
The polarisation of the electrodes is a result of the process by which charged particles
are transferred from the solution to the electrodes. These charged particles delay the
conduction process causing an increase in resistance and thus a rise in the potential
difference occurs. Hence the cathode becomes more negative and the anode more
positive.
The greatest problem to the application of direct current to a fresh cement paste is the
activation polarisation caused by the formation and liberation of hydrogen gas at the
31
electrodes which obstructs the flow of charges to the electrode surface. This is in the
form of:
2H- +2e---- > H2
In an electrolyte, polarisation effects known as back e.m.f. occur at the electrode

interface, resulting in an increased resistance. This can be minimised through the use of
alternating current which causes a capacitance to occur in parallel with the resistance of
the specimen. As such, concrete can be considered to be a network of capacitors and
resistors in parallel through the liquid and solid phases.
(b) Bulk electrolyte polarisation

A secondary polarisation effect occurs away from the electrodes in the bulk of the
electrolyte. This involves the alignment of dipoles and other charged particles in the
direction of the current flow (McCarter and Afshar 1985). If direct current is applied
this realignment is stable.
The magnitude of the stored charge IS defined as the dielectric constant and IS
independent of sample size:
Eqn.2.1
where:
Dielectric constant
C = Capacitance of specimen
Co = Capacitance of equivalent sample size in a vacuum
32
2.6.8 Temperature dependence
Electrolytes are affected by temperature in two ways. These are:
1. A rise in the temperature of the solvent reduces viscosity causing the mobility of the
ions to increase thus reducing resistance.
2. Ionic concentration increases with higher temperatures. This is due to increased
dissociation of ions within the electrolyte.
From the above, it can be seen that electrolytes have a negative temperature coefficient
of resistivity with an increase in temperature reducing resistance (Rait 1992).
Because of the high temperature coefficient of resistivity for ionic solutions, resistance
values corrected to a particular temperature must be used. These values are calculated
using the formula (Wilson and Gupta 1996):
Eqn.2.2
where:
n, = Corrected resistance (Cl)
R = Measured resistance (Cl)
a = Temperature coefficient (per CC) = 0.022 per cc

T = Temperature (CC)
To = Reference temperature CC) = 20 C c
This shows a considerable increase in resistance at higher temperatures. After an initial

period of constant resistance corresponding to the setting time, it can be seen that the
relationship becomes linear. In view of the published data (Wilson and Gupta 1996)
relating log (maturity) to strength, this suggests that corrected resistance is also closely
related to strength.
33
The electrical resistance of a material depends on its temperature. It is known that the
resistance of a metal (Figure 2.4a) increases as its temperature increases and conversely
the resistance of an insulator decreases with temperature as shown in Figure 2.4b.
Temperature eC) Temperature eC)
a) Metal or alloy b) Insulator or dielectric
Figure 2.4 - Resistance versus temperature for metals and insulators
The temperature coefficient, a, is used to indicate how the resistance of a conductor will
change with temperature (Rait 1992) and is found using the expression:
1
a= x Eqn.2.3
where:
Values of resistance (n)
Temperatures at which RI and R2 are respectively measured (OC)
This is generally of more importance in the production of electrical components which

employs almost pure metals as opposed to mortar and concrete which are a combination
of numerous compounds.
34
2.6.9 Resistivity
The resistivity (Om) of a certain conductor is dependent on:
1. The temperature
2. The specific conductor material
The resistivity of a sample of concrete of a specific composition is dependent on the

amount of free water present and the density of the sample. Increases in density are
caused by the continuing hydration of the cement paste which in turn causes an increase
in resistivity. Resistivity also increases due to the dehydration of the saturated pores by
evaporation.
Resistivity is a fundamental property of any material which depends on the chemical
composition and the physical treatment which the material has received (Rait 1992). It
can be determined by using the following equation:
pL
R = Eqn.2.4
A
where:
R = Resistance (n)
p = Resistivity (nm)
L Length (m)
A Area (nr')
Also conductivity (Srn-I) is defined as the reciprocal of resistivity. Hence:
1
00=- Eqn.2.5
P
35
The conduction of electricity through concrete is dependent upon both the ohmic
resistivity and ionic conductivity with the ohmic conductivity being dependent on the
fractional volume of aggregate and the proportional volume of the non conducting
cement compounds.
2.6.10 Impedance
The alternating current impedance of concrete is almost identical to its direct current
resistance (Neville 1988). This is due to the fact that the capacitive reactance is
significantly larger than its resistance and as such only the resistance contributes
significantly to its impedance.
2.6.11 Capacitance
The capacitance of concrete is small when using an alternating electric current (0.020
microfarads at 50Hz at 7 days) and decreases with age (0.003 microfarads at 50Hz at
113 days) (Neville 1988).
2.6.12 Electrical properties as an indication of concrete properties
It has been found (Zhao, Zhou, Zhu and Feng 1998) that concrete permeability is one of
the intrinsic properties of concrete and is directly related to its durability.
Table 2.3 - Comparison of concrete resistance, permeability and ASTM C1202
ASTM cuoa Alternatin I! Current

Charge Permeability of chloride Conductance Resistance
(Coulombs) (lO-s Siemens) (Ohms)
>4000 High >287 <348
2000-4000 Moderate 154-287 649-348
1000-2000 Low 87-154 1149-649
100-1000 VeI)' Low 27-87 3704-1149
<100 Negligible <27 >3704
36
The resistance of the specimen measured by alternating current (1V at 1kHz) was
converted into conductance and then compared to the conductance from ASTM C 1202.
The correlation with this was found to be 0.9867 with conductance = 0.0666026xcharge
+20.446 (Zhao et al 1998) and can be seen in Table 2.3.
The conductivity of pastes of ordinary Portland cements containing a variety of

admixtures at a frequency of 3kHz has been measured (Tamas 1982). The results
showed that:
1. In the case of cements, the conductivity of the paste increased during the initial
period, reached a first maximum after 1-3 hours and then decreased again until
around 6-10 hours whereupon a second maximum occurred.
2. This second maximum did not occur when pastes were made of clinker or pure
clinker materials.
3. The effect of several accelerating admixtures CNH.4HC03and CaCh) was to slightly
delay the appearance of the second maximum but in the use of formate accelerators
the second maximum occurs after a shorter time and was less intensive.
4. The effect of retarders such as citric acid was to greatly delay the appearance of the
second maximum.
5. With an increase in temperature of the paste, the second maximum occurred after a
shorter period of time.
Electrical conductivity is a consequence of the movement of charge carriers and in the

case of cements this is achieved by ions. Electrophoretic charge transport can be
excluded in view of the relatively large size of cement particles (Tamas 1982). Thus an
increase in the conductance of paste can be attributed solely to an increase in the
mobility or number of ions. The mobility of ions can be affected by the formation of
electrically insulating layers during hydration. However, the previous results can be
explained by changes in ion numbers.
37
The first conductivity maximum, which takes place shortly after gauging with water is a
H
consequence of hydrolysis, which yields Ca , OH" and in the case of gypsum also S04-
ions. These are readily absorbed by the formation of CSH and ettringite which
decreases the conductance again. I3-C2S reacts slowly with water and showed a flat
maximum approximately 6-7 hours after mixing with an overall conductivity much
lower than with C3S. At approximately 5 hours the conductance of each sample crossed
being due to the higher reactivity of C3S.
The second maximum was considered to be due to ettringite being converted into
monosulfate creating 2 mols of Ca ++ and S04 - ions per mol of ettringite. Also, the effect
of admixtures was attributed to the formation of monosulphate phases from ettringite.
It has also been found (McCarter, Chrisp and Starrs 1999) that electrical testing can also
be used to fmd a number of distinct regions of alkali-activated slag which rate the
chemical reactivity and rigidity of the mixture.
2.7 Direct Electric Curing

DEC was first introduced in the railway sleeper industry, utilising single cell moulds.
These techniques have now been largely superseded by multicell moulds on highly
automated production lines (Bredankamp et al 1993). Casting operations are normally
performed within normal working hours, with the accelerated curing taking place after
working hours.
In colder climates it has been the standard practice to stop most concrete construction
work during the winter months. However, with limited time and economical
considerations, this has lead to the implementation of direct electric curing (DEC).
It has been found (Bredankamp et al 1993) that at constant times and differing voltages
for DEC as compared to normal curing:
38
1. At approximately 80 hours after the concrete had been cast, the strength of the
naturally cured samples exceeded that of the DEC samples.
2. The DEC cycle causes no subsequent loss of strength at 28 days. This ultimate
strength depends on the initial temperature rise.
It has been stated (Wadhwa et al 1987) that direct electric curing is based on the fact that
concrete is a resistor of approximately SOO-ISOOQ/cm and as such can be heated when

an electric current is passed through it. Concrete has also a negative temperature
coefficient and behaves like an electrolyte and as such only alternating current can be
used.
It has also been noted (Bredankamp et al 1993) that DEC should have the following
guidelines when used in situ:
l. Before heating is started, a period of 3-5 hours after adding water is required to
allow the concrete to gain sufficient strength to withstand the thermal forces applied.
2. It is essential that the concrete must be adequately vibrated upon placing and that the
minimum water/cement ratio is used.
3. All efforts must be made to stop the evaporation of water from the surface of the
sample by the use of a suitable covering.
2.7.1 Practical application of DEC
The design criteria for heating of electrically cured concrete in a practical environment
require the following data (Kafry 1993):
1. Quantity of concrete to be cured in an 8 hour shift.

2. Number of shifts per day.
3. Demoulding strength required.
4. Mix proportions.
39
5. Reinforcement data.
6. Geometry of specimens.
7. Ambient temperature and moisture conditions.
8. Availability of electric power supply.
The actual design procedure consists of (Kafry I 993):
1. Calculating the power.

2. Calculating the working voltages and currents.
3. Designing the curing scheme.
4. Designing the control scheme.

5. Specifying the equipment.
This design procedure can be calculated by comparing maturity values of normally and
accelerated cured concrete with an emphasis on the analysis of electric field density
when reinforcement is used. The equipment specification is also relatively simple as
electrical curing equipment has already been implemented in numerous places. This
equipment has been shown to have adequate safeguards for worker safety and quality
control (Kafry 1993).
2.8 Heating Cycles

Due to the heterogeneous nature of concrete, being composed of a number of different
materials including water and air, the values of each materials coefficient of thermal
expansion differ considerably. Therefore, once heating has begun, the differential
expansion of the materials causes voids to form within the mix resulting in a reduction
of final strength.
Heat can be applied to fresh concrete using DEC, to drive off trapped air and allow
residual water to be drawn to the surface during this fluid-plastic stage, without any
detrimental effect on the concrete. However, with dry mixes, this should be avoided
40
due to the fact that the initial setting time for a relatively dry mix is considerably shorter
than that of a wet mix.
2.8.1 Heating period
This period, as the name suggests, is where heat is added to the system to accelerate the
curing process. Unlike other methods, DEC has the advantage that the heat is spread
evenly throughout the mix, therefore internal temperature differences are minimal.
Consideration should be given to the geometry and construction of the mould in that
stiff, deep moulds with small casting openings can withstand higher heating rates more
than weak, shallow and open moulds.
2.8.2 Isothermal period
The maximum temperature is held constant during this period by either reducing the
amount of energy being supplied or shutting of the power completely when the
maximum temperature is reached but supplying short bursts of energy to the system
when required to maintain the maximum temperature (Kirkbride 1971).
The energy required to maintain the maximum temperature at this stage is largely
dependent on the surface area of the unit being cured. Units with very large surface
areas (Le. long thin units) will loose heat more readily than those with small surface
areas which can store large amounts of heat energy within them (Kafry 1993).
The temperature may continue to rise for a period of time due to the heat of hydration
and it is this which necessitates the use of a predetermined maximum temperature to
avoid boiling of surplus water.
In Denmark (Danish Ministry of Transport 1985), the maximum curing temperature in

concrete can be no higher than 60°C because it is supposed that temperatures higher
41
than this value cause damage due to the formation of the microstructural cracks which
cannot later be remedied.
2.8.3 Cooling period
The time taken for this stage to complete is again dependent on the geometry of the unit
in that large surface areas tend to cool more quickly than small areas. However care
should be taken to ensure that the rate of cooling is not too rapid thus avoiding thermal
over-stressing.
2.8.4 Curing cycle curve
From the descriptions given previously, the general curve used In most forms of
accelerated curing is that shown in Figure 2.5.
Heating period
Isothermal
period
Natural cooling period
Natural curing period Time (Hours)
Figure 2.S - General temperature curve used for accelerated curing methods
A number of guidelines concerning various heating cycles have been proposed by

several authors (Kirkbride 1971, Orchard and Barnett 1971, Wilson and Gupta 1996).
These include:
42
1. The temperature should not be raised during the setting period and that even after
this period the rise in temperature should be limited to 22-33 CIhr
D
rise with a
maximum of 66-S2DC.
2. The rate of temperature rise should not exceed 25 CIhr.
D
3. To get the highest long term strength from a given mix there should be a delay of 3-
5 hours from the mixing of the concrete to the commencement of the temperature
rise, the rate of temperature rise should not exceed 20DCIhr and the maximum curing
temperature should not exceed SODC.
4. After approximately 6 hours the temperature may be raised to 100DC without serious
loss of long term strength.
5. Cracks are also liable to occur if the difference between the temperature at the core
of the unit and that at the surface is greater than 15DC.
It has also been stated that the increase in strength per hour reaches a constant low value
at an age of 24 hours for a maximum curing temperature of 90DC and the same low
value at increasing ages up to 42 hours as the maximum curing temperature is reduced to
20DC. Hence, there is no advantage in accelerated curing after 42 hours and very little
advantage after 24 hours.
2.8.5 Maturity
Saul's maturity law (Saul 1951) (where compressive strength relates to increases in time
and temperature of curing) is generally accepted as a way to correlate a relationship

between time and temperature of curing and compressive strength (Kirkbride 1971).
The concept is that the strength of any concrete can be related to its maturity expressed
as a product of time and temperature of curing. In fact, the maturity is usually calculated
from a base of -1 0 or -11DC. Although there is a conflict of opinion as to its reliability,
where in some cases it underestimates the actual strength achieved by high temperature
curing. The law provides a useful design aid in the calculation of a high temperature
curing cycle.
43
It has been presented (Neville 1988, Wilson and Gupta 1996) that graphs of log
(maturity) against strength will show a linear relationship except at low values of
maturity. Maturity is defined as:
Eqn.2.6
where:
M = Maturity (OC.hr)
t = Time (hr)
T = Temperature (OC)
= Maturity reference temperature (OC) = -11°C
Experiments have been carried out to relate the electrical resistance to log (maturity),
and to demonstrate how measurement of electrical resistance could be used to measure

the reduction in the time required using accelerated curing.
The maturity concept has been recommended by organisations such as the ACI as an
alternative to testing field cured cubes or cylinders. From the maturity values, the
strength of hardening concrete can be estimated. However, maturity implies that with
increasing age, concrete will continue to gain strength for an infinite amount of time.
It has also been shown (Jonasson 1985) that the maturity concept works well for
permanently high temperature concretes up to around 14 days but then over estimates
strength at later ages.
Alternatives have been suggested (Carino et al 1983) such as using a semi-logarithmic

function as an appropriate strength maturity relation.
S = a + b log(M) Eqn.2.7
44
where:
S Strength (N/mm2)
M Maturity (QC.hr)
a,b Regression coefficients
However, this equation is only a good approximation at intermediate maturity values.
It has also been suggested (Carino et al 1983) that the strength maturity relation should
be in the form of a hyperbola with the following equation:
Eqn.2.8
where:
S = Strength (Nzmrrr')
M = Maturity (QC.hr)
Su = Limiting strength as maturity tends to infinity (N/mm2)
A = Initial slope of the strength-maturity curve (NjoC.hr.mm2)
Parameters A and Su have a physical significance and are determined by a computerised
least squares fit of the data.
The temperature at early ages has been shown to have a disproportionate effect on the
early strength development compared with time. In addition, early age temperature
affects strength at later ages, with high early temperatures reducing long term strength
development.
Above a certain maturity value, when the C2S is the principal contributor to strength
gain, a rise in the initial maximum temperature does cause a comparative strength
45
increase at a given maturity. The conclusion to be drawn is that for a given constant
maturity a higher strength can be achieved if that maturity is reached rapidly (Orchard
and Barnett 1971).
Despite the mild atmosphere present in the UK, there are around 30 days during a
typical winter that the temperature falls below freezing. This can double when in the
north of Wales or Scotland or if above 350m.
It has been recommended (Pink 1985) that concrete should attain at least SN/mm2
compressive strength before being exposed to the damaging effects of frost.
In the temperature range of O°C to 20°C and during the first 72 hours Weaver and
Sadgrove (1971) have proposed the concept of 'equivalent age' based on a factor F from
which an equivalent maturity at 20°C can be calculated.
(t+16)2
F= 36 Eqn.2.9
where:
t = Temperature of concrete during curing (OC)
F = Equivalent age
The Arrhenius equation can be defined as (Naik 1985):
f(Tl = k. exp( - R~k) Eqn.2.10
where:
= Proportionality constant
= Temperature of concrete in degrees Kelvin
46
E Activation energy in kilo joules per mole
R Universal gas constant
The Arrhenius equation takes into account the activation energy for the hydration
process. Since this reaction is exothermic, for each temperature there is a slightly
different activation energy (E). Also, different cements have different compositions and
as such have different activation energies.
It has been reported (Naik 1985) that E could be predicted by the following relationship:
at Tc~20°C, E(Tc) = 33.3 kJ/mol
and at Tc~20°C, E(Te) = 33.3+ 1.47(20- Te) kl/rnol
0.01 .__..___.__..__._--,-__.____._~~_.__~~..__~,---,__..__.__._--,-__._~~~__.__~..__~,-----,
-20 o 20 40 60 so 100
Temperature (QC)
Figure 2.6 - Dependence of rate of hydration of a cement paste on temperature
Figure 2.6 shows the temperature function, illustrating the dependence of the rate of
hydration of the cement upon the temperature. In the 1950s, formulae by Nurse, Saul-
47
Bergstrom, Verbeck and Rastrup served well for classical concrete curing as long as
extreme temperatures were not attained. It can be seen from Figure 2.7 that for high
hydration temperatures the Rastrup formula does not apply, while the Nurse and Saul-
Bergstrom formulae lead to an underestimation of the hydration rate.
10 r-------------------------------,------------------------,n
o ~--~--~~--~--~----~--~----~--~----~--~----~--~
20 30 40 50 60 70 80
Temperature (OC)
Figure 2.7 - Temperature functions of various formulae to predict cement

hydration rates
2.8.6 Thermal conductivity of concrete
A knowledge of the thermal conductivity of concrete during the early hydration period
is essential to obtain an accurate estimate of the thermal stresses that will develop in a
concrete structure as a result of the heat produced by the binder during hydration
(Gibbon and Ballam 1998).
Traditionally, the values for the thermal conductivity used for heat models of concrete
structures have been based on the properties of the aggregate or the hardened concrete,
with a single value being assumed for all stages of hydration. This does not represent
48
the actual conditions of the concrete during hydration (Gibbon and Ballarn 1998). It has
been shown that the thermal conductivity of concretes and mortars decrease in the first 2
days by 10-20%, depending on mix constituents, whereupon small decreases continue to
occur during the following 7 days. Typical values of thermal conductivity are in the
range ofO.6-2.6 W/m.oC for a time after mixing ofO-4 days.
2.8.7 Thennomechanical affects
Fresh concrete consists of a mixture of materials with different coefficients of thermal

expansion; hardened cement paste and aggregates have relatively similar coefficients;
water and air are about 20 and 300 times higher respectively (Kirkbride 1996). Cold
fresh concrete placed in a mould and subsequently heated after 1 hour, undergoes a
series of differential expansions within its composition as the cement, aggregates, water
and air expand at vastly different rates. In these circumstances it may be postulated that
internal stresses within the concrete could occur. However, if heating is delayed for 3 to
5 hours the concrete at this time has achieved a degree of homogeneity and strength
which, if the rate of temperature rise is kept within limits, is sufficient to sustain the
disruptive forces.
Heat contours have been obtained (Orchard and Barnett 1971). The shapes of
temperature contours were very similar irrespective of temperature or time once
thermodynamic equilibrium had been established. A temperature differential of as high
as 16°C was obtained between the centre and the outside of the specimen when the
central temperature was 88°C and the ambient was 20°C. These temperature contours
could be reduced by insulating the specimens or by placing the specimens in a preheated
area to follow the heating regime.
It was also noticed (Orchard and Barnett 1971) that the failure pattern of electrocured
samples during compressive failure was consistently a near perfect cone fracture. This
49
was considered to be due to the variances in maturity across the sample which set up
regular patterns of internal stresses and differing strengths.
2.9 Effects of Accelerated Curing on Mortar and Concrete
2.9.1 Fresh properties
It has long been known (Abbasi and AI-Tayyib 1985) that when concrete ingredients are
mixed in hot weather rapid evaporation of water occurs. This results in a lower
effective water content and hence lower effective water/cement ratio. From this,
workability is reduced and also lower slump values are obtained.
2.9.2 Microstructure
A comparison of hot air cured samples (Bajza and Rousekova 1983) shows that
increasing curing temperatures lead to changes in the structure of the matrix; with the
more intensive the heat treatment, the coarser the structure.
2.9.3 C-S-H formation
Within the high volume production of precast concrete the use of elevated temperature
curing techniques to accelerate early strength development has become widespread
(Patel, Bland and Poole 1995). However, research has shown that such curing leads to
the development of a modified microstructure leading to the formation of secondary

ettringite. Mature pastes cured at temperatures of between 50 and 70°C have been
reported to exhibit a coarser C-S-H morphology and pore structure than that of pastes
cured under ambient conditions. The width of the C-S-H shell around the hydrating
grains has been observed to increase with increasing cure temperature, while the inner
C-S-H in a paste cured at 80°C has been reported to be denser than the outer hydrate. At
elevated temperatures a greater proportion of portlandite has been found to form dense
clusters as opposed to the more lamellar type morphology observed under ambient
conditions (Patel et al 1995).
50
Samples manufactured at curing temperatures up to 46°C show a similar microstructure
to that seen at 20°C with only a limited coarsening of the cement matrix at 42 and 46°C.
Concretes cured at higher temperatures exhibited a coarser microstructure than that of
concrete cured at 20°C, particularly with respect to portlandite (Patel et al 1995).
It is clear that concrete is a poor conductor of heat (Xuenquan et al 1987). It takes time
to raise the temperature of a specimen and as such differential temperature gradients will
be formed. Meanwhile, the early hydration of cement, especially for C3S, can be too fast
under such a high temperature, resulting in the formation of a large amount of very fine
C-S-H gel, which may surround the unhydrates, causing hindrance of diffusion and
further development of strength.
Concrete outside the laboratory cures at temperatures other than 20°C and often under
less than ideal moisture conditions. High curing temperatures may result from hot
weather, accumulated heat of hydration or applied heat. Heat curing is generally used in
the production of precast concrete products in order to increase the early age strength,
primarily to allow rapid production. The long-term properties are often observed to be
negatively influenced by elevated curing temperatures. Strength and other mechanical
properties are often reduced and permeability increased. It has been shown (Kjellsen
1995) that curing at elevated temperatures influences microstructure considerably.
It has also been shown (Kjellsen 1995) that curing at elevated temperatures leads to
reduced surface areas of the hydrates formed and a coarser pore structure. Also that a
coarser pore structure of the outer product phase prevails at mature ages as a result of
isothermal curing at elevated temperatures. It has been shown (Kjellsen 1995) that the
distribution of reaction products is much more uniform in pastes cured at relatively low
temperatures. In pastes cured at high temperatures, relatively dense shells of reaction
products form around the hydrating cement grains, while the outer product phase
remains relatively porous as relatively little reaction product is formed in this phase.
51
Curing at temperatures above 70°C at early ages may influence the formation of
ettringite decisively as this phase may not form at higher temperatures. At even higher
temperatures monosulphate also becomes unstable and hydrogamet may appear.
Ettringite may also reform at later ages, resulting in destructive cracking. Another effect
of heat curing is caused by a too rapid temperature rise at early ages, which may lead to
microcracking and increased porosity due to differences in the thermal properties of the
concrete constituents.
Although reaction products with less chemically bound water have been reported to
form at elevated temperatures (Kjellsen 1995), no indications have been found that the
reaction products formed at 50°C contained less overall non-evaporable water than the
reaction products formed at 20 or SoC. Assuming a constant water chemically bound in
the C-S-H gel, the greater total oxides in the inner product of a sample cured at 50°C
suggest that this phase has a somewhat lower fine porosity than the outer product or
hydrated phases of a specimen cured at 5°C.
Oxide totals have indicated (Kjellsen 1995) that the local finer porosity of the outer C-S-
H gel is largely independent of curing temperature. At higher curing temperatures the C-
S-H in the inner product apparently becomes denser primarily due to lower fine porosity
and the outer product contains relatively coarse pores.
Microanalytical and microscopic examination have supported the idea that the early
curing temperature has relatively little influence on the final distribution of C-S-H. The
inner product of heat cured concrete does not appear denser than that of concrete cured
at 20°C, and the outer product does not appear to have a higher porosity. With a 0.50
W/C ratio paste non-isothermally cured, the specimen reached considerably less maturity
at the end of its heat curing period and showed, in contrast to a heat cured concrete,
some influence of heat curing as regards C-S-H distribution (Kjellsen 1995).
52
For many precast applications concrete undergoes a period of curing at elevated
temperatures to accelerate strength development. It has been found (Scrivener 1992)
that in virtually every case, concrete continues to hydrate after the heat treatment has
ceased.
As an example of microstructural development for a concrete sample cured at room

temperature, placed in a bath at 200e for 4 hours, heated to 800e over 4 hours and
maintained at this temperature for 16 hours, then cured normally until tests at 35 and 135
days, it has been reported (Scrivener 1992) that immediately after heating the
microstructure of the paste reveals rims of product 1-2Jlm thick around the cement
grains. After 35 days a region of darker inner product e-S-H develops inside the bright
rim and at 135 days this dark region had grown thicker as the cement grains continue to
react. The outer part was said to have formed at 800e while the inner part formed at
200e in a much greater period of time. It was also reported that there were no
significant differences in composition at either 35 or 135 days.
2.9.4 Delayed ettringite formation
The observed expansion of some laboratory and field Portland cement mortars and
concretes exposed to elevated temperatures has been attributed to delayed ettringite
formation (DEF) (Yang, Lawrence, Linsdale and Sharp 1999).
Others have stated (Shayan and Ivanusec 1996, Shayan and Quick 1992) that DEF alone
cannot cause deterioration by itself and other factors such as Alkali Silica Reaction
(ASR), freeze-thaw cracks or microstructural damage are the major factors.
Expansions have been reported (Odler and Chen 1995) in cement pastes cured to a
maximum of 80-90oe.
53
According to investigations by optical microscopy and scanning electron microscopy
(SEM), many ettringite bands can be observed in concretes and mortars particularly in
the transition zone between paste and aggregate.
There are two principle hypotheses concerning ettringite formation. These are the
ettringite crystal growth pressure theory and the uniform paste expansion theory (Yang
et al 1996).
The former hypothesis suggests that pressure from ettringite crystals growing m
preformed microcracks in the transition zone between the paste and the aggregate
widens these cracks and causes the observed expansion. It has been suggested that the
degree of supersaturation of ettringite in the pore solution of the cementitious material
was insufficient to cause sufficient crystal growth pressure to achieve this expansion.
However, it has also been suggested (Diamond 1996) that on the basis of basic concepts
of fracture mechanics that a relatively small degree of ettringite supersaturation is
enough for ettringite growth to expand the preformed cracks as the stress concentration
at the tip of a long crack in concrete is large due to the brittle nature of concrete in
tension.
The uniform paste expansion theory suggests that expansion of the cement paste matrix
occurs, leaving gaps around the cement grains. Subsequently, newly crystallising
ettringite rapidly fills the cracks produced by this expansion to result in ettringite band
formation. Ettringite formation does not contribute directly to the observed expansion.
It has been found (Yang et al 1996) that the widths of the ettringite bands around the
aggregate were proportional to the width of the aggregate. This has been attributed to
the effect of uniform expansion of the cement paste. Microscopic investigations
(Diamond 1996) showed that many ettringite and rim cracks form as portions of a
continuous crack network distributed throughout the concrete, which develops partially
around the aggregate grains. The results of X-ray microanalysis (Yang et al 1996)
suggests that portions of sulphate and aluminate species may be mixed with C-S-H gel
54
immediately after the heat cure due to dissolution and decomposition of ettringite. The
sulphate content of the C-S-H gel was found to decrease, while ettringite recrystallised
during the period of expansion. At the end of the expansion, the gel was seen to have
only mixed with monosulfate. It has also been suggested that recrystallisation of
ettringite initially occurs in C-S-H gel and then in the cracks and voids which may then
cause the uniform expansion of the paste.
Most of the ettringite in the mortars normally cured formed within seven days of
hydration commencing. Ettringite could not be detected in the heat cured samples
immediately after curing (Yang et al 1999). However, the overall amount of ettringite
in the heat cured samples was seen to increase after seven days of water storage. It was
generally shown to have a lower ettringite content than the normally cured specimens at
all ages.
For mortars cured at 100°C hydrogamet was detected but monosulfate could not be
detected. In contrast, monosulfate was found in the normally cured specimens but
hydrogamet was not found to be present even after hydration for one year.
The microstructure at one year for normally cured specimens showed only empty cracks
which may have been caused by the drying process necessary for SEM. Several regions
of high monosulfate concentrations appeared in the mortar but only a few concentrations
of ettringite could be observed.
For heat cured (lOO°C) specimens no concentrations of ettringite could be found

immediately after heating and there were found to be no significant differences in the
frequency or width of cracks between the cement matrix and aggregate when compared
to normally cured specimens (Yang et al 1999).
Indications of an initial ettringite band formation and areas of ettringite concentration

could be identified in the heat cured mortar after 90 days. These bands were seen to
55
develop both in the transition zone between the cement matrix and the aggregate and in
the cement paste matrix itself. Most of the bands seemed to form directly from the gel
hydration products rather than having been deposited in preformed empty cracks as an
inert filling. Also, the cracks linking the ettringite bands were observed to be of a lesser
diameter than the ettringite bands.
In heat cured mortars after 155 days of water storage at room temperature, ettringite
bands were shown to have developed (Yang et al 1999) both in the aggregate-paste
transition zone and in the cement-paste matrix, in the region near and parallel to the
surface of the prisms. More ettringite bands were observed in the transition zone than in
the paste matrix. These bands passed through the transition zones along one or two
sides of the aggregate rather than completely around the aggregate.
At later stages of the prism expansion, ettringite bands were seen to develop from the
surface region to the central area of the prism. Many large grains were seen (Yang et al
1999) to be covered with ettringite bands from one, two, or occasionally three sides.
Some ettringite bands were seen to have developed in the paste matrix and formed
bridges between the aggregate grains.
From the above results it was concluded that there was little evidence to support the
paste expansion hypothesis. It has also been found (Lawrence 1993) that DEF is
dependent on the cement composition, with certain cements showing expansions, while
others show none.
2.9.5 Microcracking
The phenomena that occur during the early age of concrete curing are complex with
coupling between chemical-thermal-hydration and mechanical effects. Among the main
attributes that govern the non-linear behaviour of early age concrete are stiflhess
56
evolution, the development of thermal strains, creep and cracking (De Borst, Van Den
Boogard, Van Den Bogert and Sluys 1993).
Recent work indicates a growing concern that concrete cured at elevated temperatures
may show abnormal expansion and associated microcracking (Patel et al 1995). In most
of these studies delayed ettringite formation was identified as the probable cause of the
expansion.
The chemical processes that occur during cement hydration in the first days after casting
are accompanied by an increase in temperature and volume change. However, because

the stiffness of the concrete at this stage is relatively low, a change in volume results in
only moderate, mainly compressive, stresses. As the hydration of the cement continues,
the stiffness of the concrete increases significantly. During the hardening phase, if
concrete temperature drops, significant tensile stresses may develop which lead to the
formation of cracks (Niu et al 1995).
Microcracking has been observed in many concretes, forming a prominent network in
concretes cured at 85°C, and also in eighteen year old precast concrete.
The fissures within the most extensively microcracked concretes are characterised by
hydrate infill, notably ettringite with some associated portlandite. Where microcracks
occur at the paste/aggregate interface, the ettringite occurs as a crust like layer on the
surface of coarse aggregate particles.
The presence of microcracks increases moisture mobility within the concrete which may
produce density gradients within the matrix leading to further microcracking. The
domination of the matrix by a network of microcracks is conductive to the formation of
secondary ettringite (Patel et al 1995).
57
Table 2.4 - Coefficient of linear thermal expansion
Component Coefficient of linear thermal expansion

(per°C)
Gravel 4-5xlO-6
Limestone 2-3xlO-6
Sand 2xlO-6
Cement 11-20x1O-O
Owing to the considerable differences in the coefficient of thermal expansion of

aggregate and cement matrix that can be seen in Table 2.4 (El Hussein and Abd El
Halim 1993, Neville 1988), the aggregate will attempt to expand at a different rate than
the cement matrix during temperature changes. This can be seen for the cooling case in
Figure 2.8.
Cement Aggregate Particle

Matrix
Crack
Figure 2.8 - Localised damage caused by thermally induced stresses
The cement matrix/aggregate interaction with temperature changes and the development
of thermally induced stresses at high temperatures has been described (El Hussein and
Abd El Halim 1993). Since expansion of the aggregate during thermal heating is
limited by the cement matrix, a contact pressure will develop at the aggregate/cement
matrix interface. Therefore, the contact pressure produced by differential expansion
will result in radial and tangential stresses in the cement matrix surrounding the
58
individual aggregate particle. As a result, the cement matrix around the aggregate will
be under tension.
A number of assumptions have been made concerning Figure 2.8. These include:
I. No stress relaxation through creep occurs.

2. Uniform temperature throughout is assumed.
3. Coefficient of thermal expansion and Poisson's ratio are assumed constant
throughout the temperature range.
Once these assumptions have been made, Figure 2.8 can be modelled as:
Eqn.2.11
where:
P = Contact pressure
L\T = Change in temperature (OC)
ai, a2 = Coefficients of thermal expansion for aggregate and cement (OC I)
VI,V2 = Poisson's ratio for aggregate and cement
EhE2 = Young's modulus for aggregate and cement
In Equation 2.11, K can be calculated by:
Eqn.2.12
where:
a = Radius of aggregate (mm)
b Radius of cement matrix coating (mm)
59
From this the tangential stress (crt) and radial stress (crR) can be calculated by:
Eqn.2.13
and
Eqn.2.14
Localised damage in the form of microcracks resulting from high tensile stresses may
result in reducing the mix resistance to externally applied loads.
2.9.6 Pore structure
In general cement porosity is composed of a well interconnected porosity, composed of

tubular pores, and a disconnected porosity of rounded pores. Also cement total porosity
decreases due to the hydration process but at the same time chemical shrinkage
generates new pores which increase total porosity at the time of cement setting. Pores
generated by shrinkage are only free pores well interconnected between each other.
Curing under pressure can delay the appearance of these pores due to shrinkage
(Parcevaux 1984).
Up to cement setting, trapped porosity is solely determined by the original dispersion of

cement grains in water and increases sharply during cement hardening due to the
formation of hydrated species in the intergranular porosity or because of the plugging of
some pores through microporous hydration products.
Curing temperature was found to have little influence on cement pore size distribution
except when it exceeds 80°C. Above 80°C, cement hydration proceeds at a higher rate
and gives rise to larger pores through production of steam.
60
The total pore size distribution of hardened cement pastes has been described
(Parcevaux 1984) as being composed of a threshold diameter and several families of
pores, macro or capillary porosity, micro or C-S-H gel porosity and mesoporosity
(porosity between particulate hydration products). Macroporosity has been found to be
composed of pores wider than a few thousand Angstroms (A) and gel porosity narrower
than a few hundred Angstroms. The effect of curing conditions on the pore size
distribution of hardened cement can be categorised as the width of the pores increasing
with a reduction in curing temperature. It has been found (Parcevaux 1984) that as
curing temperature increases from 25 to 65°C, the macroporosity increases while the
micro and mesoporosity decreases without any change in overall porosity. The
relationship between pore size and mechanical properties has been found to be
controversial.
Samples cured at higher temperatures have been shown (Goto and Roy 1981) to have
lower porosities. This is reasonable because the percentage hydration for samples cured
at higher temperatures is larger than for those cured at lower temperatures, for early age
samples and the porosity decreases with increased percentage hydration. However, there
is evidence that the degree of hydration for samples cured at higher temperatures will
eventually be smaller to those cured at lower temperatures for longer times.
With samples cured at 27°C, a peak occurs in the pore size distribution at 75-140A,
revealing no pores larger than 4300A in radius, while samples cured at 60°C had two
peaks, at 75-140A and at 1400-230oA. It was also noted that the total pore volume
measured by mercury intrusion porosimetry was 19% and 13.5%, instead of 33.9% and
37.9% for samples cured at 60°C and at 27°C, respectively. The differences reflected
the volumes of the larger pores, greater than 7.51lm in radius.
With additional water curing the values of total pore volume were seen (Bajza and
Rousekova 1983) to decrease. Analysis of pore size distribution data shows that the
61
amount of small pores increases with time of heat treatment as well as by additional
water curing.
Specifications dealing with curing of concrete have been limited to specifying the
desirable properties of the constituents without reference to the quality of the end
product.
It is widely accepted that proper curing of concrete is essential if the concrete is to

achieve its full potential. It is also true that the curing phase of concrete construction is
neglected. This is because at present it is difficult to have any verifiable control over
how well the concrete is being cured.
The capillary pores represent space not filled by gel or other solids such as calcium
hydroxide and the gel pores are characteristic features of the structure of the gel formed
by the hydration products.
Poor curing affects only the exposed surface for a depth of approximately 3cm (Senbetta
and Scholer 1984).
It has been found (Bajza and Rousekova 1983) that after 1 day, specimens cured at 80°C
have a lower pore volume of pores of radius 3.75-7.S0nm than specimens normally
cured at 20°C.
The proportion of small pores in the pore size distribution was shown to increase with
the time of heat treatment as well by additional water curing. It has also been found
(Bajza and Rousekova 1983) that with an increase in the temperature of curing, coarser
structure of the cement paste was apparent.
62
2.9.7 Compressive strength
The primary benefit of heat curing is early compressive strength (Kirkbride 1971).
Approximately 60% of the 28 day strength can be achieved in 24 hours. In many types
of precast manufacture a much lower percentage strength is required for the purposes of
demoulding and to facilitate mechanical handling, and two production cycles a day can
be achieved.
It has been shown (Hutchinson et al 1991) that the adverse effect on long-term strength
of a high initial temperature increases as the water/cement ratio increases. It has also
been stated (Hutchinson et al 1991) that the adverse effect on strength at later ages after
initially high temperature curing could be reduced considerably by the addition of
gypsum to the mix. A likely cause of this is that the gypsum interferes with the phases
of hydration that readily form in the presence of heat (C3A and C4AF). As such these
phases are not disturbed by the later hydration of slower acting phases (C2S and C3S)
which reduces microstructural damage and increases long-term strength.
It has also previously been stated (Hutchinson et al 1991) that little difference is
produced in the 24 hours compressive strength if the delay before commencement of
accelerated curing was 1 hour or 4 hours. Similarly there is little difference in the rates
of temperature rise of8.75 and 35°CIhr. This is contradictory to the expected changes in
maturity. However, there is some evidence that a delay of 4 hours before accelerated
curing produces a higher long term strength than a 1 hour delay period. It has also been
stated that the varying types of cement gained strength at approximately the same rates.
The 28 days strength of microwave cured concrete is seemingly unaffected by

temperatures up to 80°C whilst a small loss was noticed at 90°C. This is a substantial
improvement over steam curing were it has been reported that large ultimate strength
losses for temperatures of 95°C occur unless a delaying period of up to 7 hours is
achieved (Hutchinson et al 1991).
63
2.9.8 Durability
Durability of accelerated cured concrete has been a major concern. The following states
the primary durability characteristics of accelerated concrete (Kirkbride 1971):
1. Permeability: Permeability is one of the most important properties of concrete

relating to its durability (Goto and Roy 1981). The permeability is influenced not
only by the total pore volume, but also by the pore size distribution. The
permeability values decrease exponentially with decreasing W IC ratio, at a constant
curing temperature. Samples, which are cured at higher temperatures, have higher
permeabilities and therefore have a coarser, more connected, pore structure. Steam
cured specimens kept moist after heating show similar permeabilities as water cured
samples.
2. Freeze/thaw resistance: Approximately the same for both accelerated and normally
cured after 7 days.
3. Sulphate attack: Steam cured concrete appear to have greater resistance to sulphate
attack.
2.9.9 Setting times
The setting of ordinary Portland cement paste occurs primarily by the hydration of C3S
with the initial set being identified by the rapid rise in heat of hydration and the final set
being identified by the approximate peak temperature. (Neville 1988).
The setting times can be estimated by the use of Vicat apparatus (BS 4550: Part 3).
When a proportion of the cement is replaced by PFA or GGBS, the setting times are
generally increased when compared to ordinary Portland cement only. This is due to the
hydration of the cement is retarded until calcium hydroxide is liberated from the
hydrating cement. It has been reported (Gebler and Klieger 1986) that increases in the
initial and final sets can be as high as 25 and 45 minutes respectively for a replacement
level of 25% PFA. Setting times can also be further increased by the alteration of PFA
64
class, grading or composition as well as increasing the setting times when a greater
degree of replacement is used.
For a ordinary Portland cement with or without PF A inclusion (Mailvaganam, Bhagrath

and Shaw 1983) reported that the initial set can double when curing temperature
changes from 20 to 5°C. It has also been found (Kanazawa, Yamada and Sogo 1992)
that reductions in setting times of concrete occur up to a temperature of 80°C where the
cement was 50% replace by GGBS.
Results have shown that (Eren, Brooks and Celik 1995):
1. For the same temperature, increasing the level of replacement of PFA or GOBS
increases the setting times.
2. For the same temperature and replacement level, the setting times of OOBS cements
are less than for PFA concrete.
3. For the same temperature between 6 and 20°C, increasing the level of replacement
of ordinary Portland cement by GOBS increases the setting times and for higher
temperatures the setting times are reduced with no consistent trend of replacement
level.
4. The general effect of an increase in temperature is to decrease the setting times, but
the reductions are different for the initial set, final set and type of concrete.
2.9.10 Tensile, flexural strengths and modulus of rupture

Tensile and flexural strengths are accelerated in the same manner as compressive
strength. However, it has been noted that microcracking normally associated with
accelerated curing affects tensile and flexural strength to a slightly greater degree than
compressive strength due to the nature of the applied load.
65
It has been found (Abbasi and Al-Tayyib 1985) that the modulus of rupture and the
tensile splitting strength of concrete prepared and cured in hot weather conditions are
reduced with the increase in the maximum temperature of concrete at preparation. This
is the case even when all necessary precautions have been taken to achieve the required
compressive strength and shows that tensile strength is more affected by increased
temperatures.
2.9.11 Other properties
A number of other properties are important when evaluating the desirability of

accelerating curing. These include:
1. Shrinkage. A significant reduction of drying shrinkage has been found for steam
cured specimens when compared to water cured specimens (Kirkbride 1971).
2. Creep. Reductions of up to 50% have been reported for steam cured specimens
(possibly due to the rapid drying out of samples) (Kirkbride 1971).
3. Loss on ignition. At 90eC, the LOI values are found (Bajza and Rousekova 1983) to
be lower than those for samples hydrated at 80eC; this being possibly caused by
premature evaporation of water at the low relative humidity of the hot curing air, and
the subsequent lack of water for hydration. Additional water curing, at room
temperature, increases the water content but its relative increase diminishes with an
increase in temperature of hot air curing.
2.9.12 Effects of incorporating reinforcement
Incorporating reinforcement to a mix that is to be subject to most forms of accelerated

curing can be said to be similar to varying the aggregate. This is because any accelerated
curing technique acts solely on the cement hydration characteristics of the mix and
leaves both aggregate and reinforcement unaffected. However, due to differentials in the
coefficient of heat expansion, thermal stresses will occur at the boundary between
66
reinforcement and cement matrix. However, it has been shown (Wilson and Gupta
1996) that even with incorporated reinforcement, structural gain is still readily
achievable and for a range of heating regimes up to maximum of 80°C with no adverse
effect due to using reinforcement being noted.
In the case where heating is intended to be due to current passing through the
reinforcement, one end was connected to the live voltage and the other end to neutral.
Assuming that the neutral and the mould are at earth potential, there will be leakage
currents along the length of the reinforcement into the concrete which will generate heat
by direct electric conduction.
In the case where direct electric conduction is intended, current must be passed through
the reinforcing bars, which will heat up due to their inherent resistance (Wilson and
Gupta 1996). The heating of the concrete is thought to take place due to both the ohmic
heating of the concrete itself and the direct conductive heating caused by the ohmic
heating of the reinforcement.
2.10 Additives and Admixtures
2.10.1 General
Additives and admixtures are materials that are added to a mix to alter either the short
term or long term properties of concrete. Admixtures are generally added during the
mix and additives are normally used as a partial replacement for cement. A
classification of admixtures can be seen in Figure 2.9 (Rixom 1977).
67
til 'I)
::J
o .5 Q'"
Q) 'I)
~ .~8
~ 8 ::J Q)
til 0 ~ (,)
Cd .... -e0l)
;>, .~ ;>, ("d
If .s
::J
-
.........
s::
........ Q)
e
v Q)
.5 o
~ cx:l
0 V'J.
s:::
....
til Q)
~
<Ii'
-;
til
c.c::-e
....CdQ) -eU
Q)
.s
c:: ....
III
Q) -
.::!
;>,
....
- -
::J Cd_ til
e o ~
0 U .... '" ....
V (,)
....'" Cd
-c
V .=.... N ("d Cd ::J C. ~-;
'"
.... (,)
S:::J::~eeo~ El b
-
Cd 0
o ::J u._ bI) U 0 ~.:: El ::J .... ~
<Il .... -e cx:l
.... ("d
v .......
Q) ...... .r;; s::: 0 s::: (,) _e. c..= .- (,) ;>
'Q~ ....
El El El -; ~ to;;,.s ~ ~ t:l..
....v
Cd (,) _U .= ::S::J..c:-eQ.:s-ec:: U
..c: .- .- ;>, - .r;;:S 01) (,) ;>. U
Q) (,) -e -e - U Q.._.r;; :s ("d c:: -
Cd .';:: o 0 0·- 0 .....- - ~ ._ 0 ;>
~~~ ...:l tIltllt:l..OU~O~ Nt:l.. t:l..
01)
c::
(,) 01)
::J .5
-e (,)
Q)
.... ::J 01)
.... -e c:: '"....

...
Q)
Cd
v
....
....
<Il
-;
.;:
(,)
:s
-e
....o
~ Q)
o P-
...... ~ ....Cd
I\)
.... ..c:
01)
.s... ~
...... e c. .s '"....U
c:: c:: '"
v
Cd
....
01)
c:: 01)
.--
o
.9 .-·60
-e
E
- -~....
Q)
v
(,)
....
Cd
.s....
:s
o.... ~
("d
'"o....
....
(,)
c:: .::
cE
'"
(,) Q) o :s ("d
<-' ~ ~ <-'
1 ~ I I
U
I I I
......
.5
01)
....
~
<Il
....
I\)
.~
(,)
r
'"
:s
o
.g
(,) 01)
c:: o .... U
c::
-
III
QJ ';;';
o
.... 01) ("d ("d
.....
.... c.
.... .5 C.
(,)
11) ';: U c. ....
I\)
.... III
("d ("d «i 8 C.
::J ::J
~Q. ~ V'J.
""'
..
'"
~
...::s
.~
e
~
<
Figure 2.9 - Admixtures by type
68
2.10.2 Accelerating admixtures
These materials shorten the setting time of cement based mixes and/or increase the rate
of strength gain. The most popular accelerator was calcium chloride (Neville 1988) but
due to problems with reinforcement corrosion, other compounds were used. All
accelerators increase the rate of hydration of the different chemical constituents in the
cement powder.
2.10.3 Retarding admixtures
Retarding admixtures slow down the rate of setting of cement based materials. They are
usually used to offset heat build up in large mass concrete structures or for retaining
workability for extended periods of time. They are generally in the form of sugars or
hydroxycarboxylic acids (Rixom 1977). It is thought that retarding admixtures form a
film on the C3A compound and as such interfere with hydration through the barrier to
water. This film then breaks down which allows hydration to continue. Also, retarders
interact with the C3S compound, which allows retardation of up to several days.
Compared to normally mixed concrete, hot-mixed concrete requires more water for a
given consistency. Any elevation of the mix temperature accelerates the changes in
consistency and as such there is a more limited time from mixing to compaction.
It has been concluded that the use of retarding admixtures delay the chemical reactions
in hot-mixed concrete, and stabilise the mobility of the fresh mixes for at least 20
minutes. It has also been found that the use of a retarder accelerates strength
development, and the 28 day compressive strength of hot-mixed concrete is similar to
that of normally cured concrete of the same mix (Berge 1976).
69
2.10.4 Air entraining admixtures
Air entraining admixtures do exactly as their name implies and creates numerous air
bubbles (O.OS-I.2Smm diameter) within the concrete mix. This improves resistance to
freezing conditions due to the presence of large chambers for the freezing water to
expand into and can also improve workability. However, such a large amount of voids
inevitably causes both short and long term compressive strength loss. Examples of
materials commonly used as air entrainers are; natural wood resins, certain fats and oils
and lignosulphates (Rixom 1977).
2.10.5 Superplasticising admixtures
The addition of a plasticiser to a mix allows a number of properties of both plastic and
hardened concrete to be altered. These include (Rixom 1977): greater workability to be
achieved for any given water/cement ratio, consistent workability for a reduction in
water content and a slight increase in air content.
2.10.6 Pulverised fuel ash (PFA)
Pulverised fuel ash is created in electricity generating power stations when pulverised coal
(passing 75J.U11
sieve) is combusted. Generally, more than 85% of the ash produced in this
manner is composed of amorphous glass and compounds formed from silicon, aluminium,
iron, calcium and magnesium.
Physically, PFA consists of hollow, spherical particles generally between 1 and ISOJ.ll11
although, especially in the largest particles, both shape and surface texture can change
significantly.
It has long been known (Davis, Carlson, Kelly and Davis 1937) that the addition of PFA to
a concrete mix will increase its workability for a given water content. This is generally
considered to be due to the spherical form ofPFA which gives rise to a ball-bearing effect.
70
Portland cement grains tend to coagulate in the fresh concrete producing a non uniform
hydrate structure. The reason that PFA inclusion results in improvement of the workability
is that the fme sized spherical particles physically disperse the cement coagulations which
frees more paste to lubricate the aggregates. This in turn produces a more homogenous
matrix as hydration proceeds (Dhir 1986). It follows that fmer ashes increase workability
more than coarse ashes (Ivanov and Zacharieva 1982).
For a concrete mix incorporating PFA designed for equivalent properties of workability,
28 day strength and 20°C water curing as an OPC mix, it has been shown (Berry and
Malhotra 1980) that there is very little difference in strength values of the concretes up to
28 days, but thereafter, the PFA concrete goes on to develop significantly higher strength
values. Initially the concrete containing PFA will not be as strong as the equivalent OPC
mix but at later ages and depending on PFA quality the mix containing PFA is
significantly stronger.
The addition of PFA to OPC concrete significantly changes the permeability characteristics
of the concrete. It has been suggested (Bakker 1983) that PFA inclusion decreases
permeability due to the size and number of pores. This occurs because of increased
precipitation of gel products in the paste as a result of the pozzolanic reaction of PFA.
This has the effect of reducing leaching of Ca(OH)2 in later life which would further
increase the concrete permeability.
The rate at which carbonation occurs is dependent on a number of factors. These include
the permeability of the concrete, its degree of saturation and the mass ofCa(OHh available
for reaction.
As has been stated previously the inclusion of PFA decreases the long term permeability
and reduces the amount of Ca(OH)2 available for reaction. Due to this restriction of the
amount of Ca(OH)2 available for reaction, carbonation in concretes containing PFA has
been seen to be reduced slightly (Ho and Lewis 1983).
71
PFA has been widely accepted as a method of combating alkali aggregate reaction (AAR).
However, it has also been stated (Mehta 1983) that PFAs vary in effectiveness and that
some PFAs (mostly high calcium PFAs containing large amounts of soluble alkali
sulphates) increase AAR.
The reduction in AAR has been said to be (Hobbs 1982) because PFA acts like a cement
with an alkali content ofO.2 % by weight thus diluting the total alkali content.
Table 2.5 summarises the effects of PFA on the properties of fresh and hardened concrete
(Bamforth 1992).
Table 2.5 - Summary of the effects of PFA on the properties of concrete
Property Effect ofPFA

Water demand Reduced with increasing proportion ofPFA
Workability At constant WIC ratio, slump of PF A concrete is increased
Bleed Reduced bleed for PF A concretes
Setting time Increased byPF A use
Standard cured strength Early strength gain is reduced, but 10nKterm strength increased
Heat cycled strength For equivalent grade mixes, PF A concrete will have higher heat
cycled strength at 28 days
Air cured strength Air curing affects OPC and PF A concretes equally
Dryin2 shrinka2e Reduced with PF A inclusion
Sulphate resistance Increased compared with OPC concrete
Acid resistance Increased due to lower permeability
Water and gas permeability Comparable with OPC at 28 days, but reduced to a greater extent
with age
Carbonation Small increase
Chloride ingress Chloride diffusion is decreased as is deterioration under cyclic salt
exposure
If 30% of the cement is replaced by a class F fly ash (LOI = 3%, amount passing 451lm
sieve = 5.8%) then it has been shown that:
1. At 20°C and at 11 days, the amount of Ca(OH)2 present is proportional to the

reduced amount of cement present, indicating no reaction, but at 1 year about 30% of
the Ca(OH)2 has been removed, indicating considerable reaction.
72
2. At 11 days and as the internal heat increased, the strength increases as the Ca(OH)2
content is reduced.
3. At 1 year, the strengths are significantly greater than those of the 100% OPC
mixtures, and further hydration has been shown to have taken place over the full
range of heated mixes. This indicates that at temperatures over 50°C, the
pozzolannic reaction is retarded by the increase in Ca(OH)2.
Permeability tests carried out show that with OPC mixes, as internal heat increases, so
does permeability, but with OPCIPFA mixes the opposite is true.
73
Chapter 3 - MATERIAL PROPERTIES AND TESTING
PROCEDURES
3.1 Introduction
This chapter provides results of the properties of the constituent materials used, test set-
up and the experimental procedure used for the preparation and testing of samples.
All electrically and normally cured mortar samples consisted of 150x 150x 150mm cubes
unless otherwise stated.
3.2 Constituent Materials Testing Procedures
3.2.1 Sieve analysis
The sieve analysis (BS 812: Part 103) was used to determine the particle size
distribution for both fine and coarse aggregate.
Table 3.1 - Grading limits for coarse aggregates (BS 882)
Percentage by mass passio2 BS sieve

Graded a22re ate Single-sized a22re2ate
Sieve size 40-Smm 20-Smm 14-Smm 40mm 20mm 14mm lOmm
(mm)
50.00 100 - - 100 - - -
37.50 90-100 100 - 85-100 100 - -
20.00 35-70 90-100 100 0-25 85-100 100 -
14.00 - - 90-100 - - 85-100 100
10.00 10-40 30-60 50-85 0-5 0-25 0-50 85-100
5.00 0-5 0-10 0-10 - 0-5 0-10 0-25
2.36 - - - - - - 0-5
It entails the sieving of a random dry sample of material through successively smaller
sieves and recording the amount retained on each. This is converted into a percentage
passing and is then compared to tables specifying the grading limits (BS 882) for
various classes of aggregates (Tables 3.1-3.2).
74
Table 3.2 - Grading limits for fine aggregates (BS 882)
Percentage by mass passing BS sieve

Sieve size Additional limits for grading
(mm) Overall limits C M F
10.00 100 100 100 100
5.00 89-100 89-100 89-100 89-100
2.36 60-100 60-100 65-100 80-100
1.18 30-100 30-90 45-100 70-100
0.60 15-100 15-54 25-80 55-100
0.30 5-70 5-40 5-48 5-70
0.15 0-15 0-15 0-15 0-15
3.2.2 Impact test
The impact test apparatus (BS 812: Part 112) gives a relative measure of the resistance
of an aggregate to sudden shock or impact. The impact value was determined by
dropping a standard hammer mass onto an aggregate and measuring the weight of fmes
caused by the impact. The sample for the test was obtained by sieving the aggregate
through 14 and 10mm sieves and retaining the mass retained on the 10mm sieve. The
fines were calculated by sieving the impacted sample through a 2.36mm sieve and a
percentage passing value was then calculated.
The maximum allowable impact values for concrete aggregates are stated in BS 882 as:
• 25% when the aggregate is to used for heavy duty concrete flooring.
• 30% when the aggregate is to be used for pavement wearing surfaces.
• 45% when to be used for other concretes.
3.2.3 Density bottle
The density bottle test in accordance with BS 1377: Part 2 was used to determine the
density of the fine aggregate used in all of the mixes.
This was achieved by weighing a sample of fine aggregate within a SOml density bottle,
then adding distilled water until the bottle is full. The bottle was then placed in a
75
vacuum air stirred to remove trapped air. This was then repeated until all air was
removed. The bottle was then transferred to a temperature bath at 25°C and
subsequently weighed in air when containing the water and fine aggregate and the water
alone. From this the density (p.) was calculated by:
Eqn.3.1
where:
mt = Mass of bottle (g)
m2 = Mass of bottle + sample (g)
m3 = Mass of bottle + sample + water (g)
I114 = Mass of bottle full of water only (g)
m3-m2 = Mass of water used (g)
m2-mt = Mass of sample used (g)

= Volume of soil (ml)
3.2.4 Speedy test
The 'Speedy' test was carried out to determine the moisture content of the fine
aggregate used. The test involved the mixing of a standard amount of fine aggregate
with calcium carbide in a sealed pressure vessel. The calcium carbide reacted with
water present in the fme aggregate to produce a gas which pressurised the vessel. This
excess pressure was then measured and a direct reading of the percentage of water
present in the surface of the fine aggregate was then found.
This test was carried out for every mix to allow the free water content to be accurately
determined and altered were necessary.
76
3.2.5 Vicat test
The Vicat test determines the initial and final setting times of the cement in accordance
with BS 4550: Part 3. The test involved measuring the penetration of a needle into a
standard cement paste mix. When this paste stiffened sufficiently for the needle to
penetrate to a point 5 ± 1mm from the bottom of the mould, initial set was said to have
taken place. BS 12 specifies a minimum time of 45 minutes for ordinary and rapid
hardening Portland cements.
The fmal set was determined using a similar needle with a 5mm circular cutting edge set
5mm behind the leading edge of the needle. Final set was said to have taken place when
this needle made an impression on the surface of the paste, but the circular cutting edge
failed to do so. BS 12 states that the final set should be no more than 10 hours for
ordinary Portland cements.
3.2.6 Loss on ignition (LOI)
The loss on ignition (LO!) test was carried out on the cement and PFA to ascertain the
level of carbon present. This involved the heating of pre-weighed samples in a kiln to a
temperature of 975°C ± 25°C for 1 hour to burn off any carbon. Samples were then
weighed again to establish the percentage carbon present (BS EN 196: Part 2).
3.2.7 Fineness
The most commonly method of determining fineness of PFA is by calculating the

percentage retained on a 45J.1msieve (BS 451: Part 2). This is because it is quick to
perform, easily repeatable and uses cheap equipment (Malhotra 1987). However, this only
considers the particle size at one cut off, thus enabling two differently size distributed
PFAs to be considered equal fineness. Even so, it has been shown that the residue retained
on a 45J.1m sieve is roughly proportional to the pozzolanic activity index and water
requirement (Dhir 1986).
77
3.3 Constituent Material Properties
3.3.1 Coarse aggregate
A coarse aggregate of20mm single sized crushed granite was used subject to BS 812:
Part 103. The aggregate impact value was found to be approximately 8.5% in
accordance with BS 812: Part 112. The relative density was found to be 2.7 according
to BS 812: Part 2.
Results for the coarse aggregate properties can be seen in Tables 3.3-3.4 and Figure 3.1.
Table 3.3 - Sieve analysis of coarse aggregate
BS sieve size Mass Mass Percentage Percentage Cumulative %

(mm) retained passing retained passing retained
(2) (2)
20 240 2121 10.2 89.8 10.2
14 1629 492 69.0 20.8 79.2
10 464 28 19.7 1.2 98.8
5 24 4 1.0 0.1 99.9
2.36 0 4 0.0 0.1 99.9
Tray 4 0
Table 3.4 - Impact values of coarse aggregate
Sample Test 1 Test 2

Mass or aggregate pre-test (g) 539 465
Fractions passin2 2.36mm sieve(g) 38 47
Impact value 7% 10%
Average impact value 8.5%
3.3.2 Fine aggregate

The fine aggregate used was found to be class M grading according to BS 812: Part 103
with a relative density of 2.7 in accordance with BS 1377.
Results for the fine aggregate sieve analysis can be seen in Table 3.5 and Figure 3.1.
78
Table 3.5 - Sieve analysis of fine aggregate
BS sieve size Mass Mass Percentage Percentage Cumulative %

(mm) retained passing retained passing retained
(2) (2)
10 0.4 117.9 0.3 99.7 0.3
5 0.6 117.3 0.5 99.2 0.8
2.36 19.0 98.3 16.1 83.1 16.9
1.18 19.7 78.6 16.7 66.4 33.6
0.60 25.4 53.2 21.4 45.0 55.0
0.30 23.9 29.3 20.2 24.8 75.2
0.15 13.6 15.7 11.5 13.3 86.7
Tray 15.7 0
Fine aggregate
80
Coarse aggregate
0.1 10 100
Sieve size (mm)
Figure 3.1 - Coarse and fine aggregate sieve results
3.3.3 Cement
The cement used was ordinary Portland cement in accordance with BS 12 and of class
42.5N. Referral was made to the manufacturers literature and a summary of results can
be seen in Table 3.6.
79
Table 3.6- Properties of cement used
Cement property Value

Initial settin2 time 270min
Final setting time 360 min
Fineness (4SJ'm residue) 345m-lkg
WI 0.85%
3.3.4 Lime
The lime used being derived from calcium carbonate limestones and chalks being
subject to the tests and requirements of BS 890.
3.3.5 Admixtures
(a) Superplasticiser
The superplasticiser used was the FOSROC Conplast M I. Reference was made to the
manufacturers literature and a summary of properties can be seen in Table 3.7.
Table 3.7 - Properties of superplasticiser
Chemical basis Sulphonated melamine polymers

Appearance Clear liquid
Specific gravity 1.10 at 20°C
Chloride content Nil
Air entrainment <1%
Alkali content 40.0g Na20 equivalent/litre of admixture
Typical dosage 2l/100kg of cement
(b) Air entrainer

The air entrainer used was the FOSROC Conplast AE 316. Reference was made to the
80
Table 3.8 - Properties of air entrainer
Chemical basis Blend of synthetic and natural occurring

surfactants
Appearance Brown liquid
Alkali content <15.0g Na20 equivalentllitre of admixture
Typical dosage 0.031/1 OOkgof cement for a 5% air entrainment
(c) Retarder
The retarder used was the FOSROC Conplast R. Reference was made to the
Table 3.9 - Properties of retarder
Chemical basis Hydroxycarboxylic materials

Appearance Brown liquid
Air entrainment Slightly less than normal equivalent mix
Alkali content <55.0g Na20 equivalentllitre of admixture
Typical dosage OAIIIOOkg of cement
(d) Accelerator
The accelerator used was the FOSROC Conplast A650. Reference was made to the
Table 3.10 - Properties of accelerator
Chemical basis Selected inorganic materials

Appearance Clear liquid
Specific gravity 1049 at 20°C
Air entrainment <1%
Alkali content 5.0gNa20 equivalent/litre of admixture
Typical dosage OAIII00kg of cement
81
3.3.6 Pulverised fuel ash
The pulverised fuel ash was found to have a fineness of 18% passing a 45J.lm sieve and
a loss on ignition of 3.8%.
3.3.7 Water
Ordinary tap water was used in accordance with BS 3148.
3.4 Mix Design and Casting
3.4.1 Mortar
Having found the values of specific gravity for each material, mix proportions with
respect to mass was calculated. Table 3.12 shows the mix proportions by volume
(cement:lime:sand of 1:1;4:3) used to produce the mortar samples. The mix is
proportioned to the required mass of cement and the required masses for 1m3 of mortar
can also be seen in Table 3.12. These values were calculated from Eqn. 3.2:
3.15) (0.25 x 2.3) (3 x 2.68)

Proportions compared with cement = ( 3.15: 3.15 : 3.15 Eqn.3.2
Therefore (by mass): cement:lime:sand = 1:0.183:2.55
Table 3.12 - Mix proportions of mortar
Proportion by volume Mass (kWm1

Cement 1 741
Lime y.. 136
Sand 3 1890
The water content was calculated directly from the cement content as required by the
W/C ratio.
82
3.4.2 Concrete
All concrete mixes were designed with reference to Department of the Environment
guidelines (I992). A summary of the mix proportions for 1m3 can be seen in Table
3.13. The full mix design sheets can be seen in Appendix A.
Table 3.13 - Summary of mix proportions for 1m3 of concrete for various mix
designs
Cement Fine Coarse

Mix type content Water aggregate aggregate
(kg/ml) (kg/m') (kg/mJ) (kg/mJ)
C30 (Slump 10-30) 345 190 743 1162
C50 (Slump 10-30) 475 190 568 1207
C30 (Slump 60-180) 409 225 751 995
C30 (Slump 10-30) 442 190 684 1070
2% air entrained
C30 (Slump 10-30) 442 190 652 1021
5% air entrained
PFA addition was based solely on a percentage of the cement values with no alteration
to the overall mix design.
The addition of a superplasticiser, retarder and accelerator replaced an equivalent mass
of water within the mix.
3.4.3 Casting
All electrically and normally cured samples (cubes, beams and slabs) were cast In
accordance with BS 1881: Parts 108 and 125. Samples were demou1ded after a period
of 1 or 3 days and placed in a water tank at a temperature of 20DC ± 1DC in accordance
with BS 1881: Part 111.
83
3.5 Fresh Concrete Testing Procedures
3.5.1 Slump
The slump test indicates the workability of fresh concrete and is extensively used
throughout the world. This was achieved by filling and compacting (3 layers each
tamped 25 times) a standard sized cone with concrete and then the cone was removed
cleanly and vertically. The concrete then 'slumped' to a reduced height and this
reduction was measured and compared to qualitative tables relating to workability. The
slump test was used in accordance with BS 1881: Part 102.
A slump test was carried out for every mix to ensure compliance with the original mix
design.
3.5.2 Vebe consistometer
The Vebe consistometer test was also used to indicate the workability of fresh concrete.
This was achieved by placing a sample of concrete formed in a similar manner to the
slump test in a bowl of standard dimensions. A standard plate (free to move vertically)
was then placed onto the sample and the apparatus was vibrated. The time taken for the
sample to completely cover the lower surface of the plate as the concrete collapses and
fills the bowl was then recorded and compared to workability charts and the mix design.
The Vebe test was used in accordance with BS 1881: Part 104.
A Vebe test was carried out for every mix to ensure compliance with the original mix
design.
3.5.3 Air entrainment test
The amount of air entrained in a concrete sample can be assessed by the use of the air
entrained test in accordance with BS 1881: Part 106.
84
This test involved placing a standard compacted sample of concrete within an airtight
vessel. The vessel was then filled with water to a datum and a standard calibration air
pressure of 12SkN/mm2 was pumped into the vessel. This resulted in a reduction of
water level as the air contained within the concrete compressed and from this a direct
percentage value of air entrained was found.
The air entrained test was carried out for every mix containing air entraining admixture.
3.6 Hardened Concrete Testing Procedures
3.6.1 Compressive strength test
The crushing machine used for the compressive strength tests was a 2MN Avery-Denison
machine, which complied with BS 1881: Part 115. All compressive strength tests were
carried out on 15Oxl50xl50mm cubes in accordance with BS 1881: Part 116.
3.6.2 Splitting tensile test
The splitting tensile test in accordance with BS 1881: Part 117 IS a method for
determining the indirect splitting tensile strength of a cubic sample.
The test involves the placing of the sample in the arrangement shown in Figure 3.2
whereupon an increase in load of 0.02-0.04 (N/mm2)/s is imposed on the sample. The
tensile splitting strength «(Jet) can then be calculated by:
2F
(J et = 1tld Eqn.3.3
where:
F Failure load (N)
= Length of specimen (mm)
d Cross sectional dimension of the specimen (mm)
85
Tensile tests were carried out solely on concrete beams at varying ages and were carried
out at least a distance d from the end surface of the beams and other tensile testing sites
in such a way that the fracture plane crossed the trowelled surface. A diagrammatic
representation of this can be seen in Figure 3.3.
Steel loading plate
Sample
d=lOO
Fracture plane
Hardwood packing
1=100
Figure 3.2 - Arrangement of splitting tensile test
Concrete beam
Trowelled
face
)=100
Figure 3.3 - Fracture planes of splitting tensile test
86
3.6.3 Flexural tensile strength
The flexural tensile strength of concrete is determined by means of a constant moment

in the central part of a beam using a two-point loading system (Figure 3.4). A load is
applied without shocking at a rate of 0.06 ± 0.04 (N/mm2)/s until failure occurs. The
flexural tensile strength (fef) is then calculated by:
FI
fcf = dd"2 Eqn.3.4
I 2
where:
F = Failure load (N)
d) = Depth of beam (mm)
d2 = Width of beam (mm) = 100mm
= Maximum distance between rollers (mm)
Beams were tested after 28 days in accordance with BS 1881: Part 118 on both normally
and electrically cured specimens. The test set-up was in a similar manner to the splitting
tensile test with the crack face running perpendicular to the unloaded trowelled surface.
Loading rollers
d d
Supporting roller
L=3d
L=4d
Figure 3.4 - Arrangement of flexural tensile test
87
3.6.4 Core samples
The examination and compression testing of cores cut from hardened concrete is a well
established method, enabling visual inspection of the interior regions of a member as
well as an estimation of strength by testing (Bungey 1982).
Cores were taken vertically from the cast slabs at vanous locations between the
reinforcement on the underneath side. The cores were cut by means of a rotary cutting
tool with a diamond bit and were of 50mm diameter. This diameter was chosen to allow
a sufficient number of cores to be taken for the analysis of compressive strength
distribution throughout the slab. Core length was also 50mm and each core was cut in
such a manner that the section 50mm below the surface was removed and discarded.
The cores were then ground at each end to allow a parallel surface normal to the vertical
axis of the core at either end of the sample.
Initially a visual inspection of the cores was carried out and following this a
compressive strength test was performed on each sample at a rate of 12-24(N/mm2)/min.
The visual inspection allowed a qualitative analysis of crack location and size.
The compressive strength of the cores was correlated by initially coring a normally
cured slab which was based upon a mix design of known compressive strength at
various ages. The electrically cured cores were then correlated to this curve to estimate
the equivalent 150mm cube compressive strength.
Density, porosity and water absorption determination was carried out at various ages on
the core samples taken from the slabs.
3.6.5 Initial surface absorption test (ISAT)
The ISA test measures the water absorption into the concrete specimen. This is dependent
on capillary suction which in tum is dependent on the pore radius, temperature,
88
connections within the pore network, water concentration gradient and degree of
saturation. The water absorption in tum affects the durability characteristics of concrete.
The ISA T (Levitt 1969) was initially used as a quality control test for the manufacturing of
precast concrete curbs. The test was later adopted for general concrete testing as part of
BS 1881: Part S. The test apparatus can be seen in Figure 3.S. The cap void covers a
surface area of the sample ofSOOOmm2•
The ISA T theory is based upon viscous flow through a fine capillary which is modelled by
Poiseulles empirical equation for the rate of flow for straight, parallel cylindrical
capillaries (Levitt 1969) as shown below:
dv Hr4
-=-- Eqn.3.5
dt 8LJ-l
where:
dv/dt = Rate of flow (m3/s)
H = Head pressure (m)
r = Radius of capillary (m)
L = Length of water filled capillary (m)
J-l = Viscosity of water (m2/s)
This is different to the actual case for the convoluted pore network that exists in concrete
but does serve as a useful approximation.
In Eqn. 3.2 the average pore radius (r), pressure head (H = 200mm) and viscosity (u) are
assumed to be constant. The capillary length (L) is assumed to be proportional to the
volume of pores filled by water (v), hence:
dv a
--- Eqn.3.6
dt v
89
where:
a Constant
This can then be integrated to give:
v2 = 2at + C Eqn.3.7
where:
C = Constant
This can then be substituted into Eqn. 3.6 to give:
dv 0
-=af5+C Eqn.3.8
dt
For a more applicable approach to concrete testing, it has been shown that the equation is
in the form:
dv
-=at -n
+C Eqn.3.9
dt
where:
n = Material constant
The value of n is constant for a given material but has been shown to vary (Dhir and Byars
1991) from approximately 0.3 to 0.7 depending on the quality of the concrete tested. The
value of n can be taken from the graph of Log (t) against Log (dv/dt). This produces a
straight line from which the value of n is equal to the gradient of the line.
The value of n is proportional to the rate of decrease in absorption. Hence a high value of
n is equivalent to a relatively rapid decrease in absorption rates which has been shown
(Dhir and Byars 1991) to be caused by silting of the pores, decreasing capillary pressure
90
caused by moisture within the concrete or depth related changes of microstructure. A
lower value of n is equivalent to a reduced rate of decrease in the rate of absorption which
has been similarly shown to be caused by lateral spread of the test area, micro cracking or
increases in the effective pore radius due to capillary flushing.
All ISA testing was carried on two 150x150x150mm specimens in accordance to BS
1881: Part 5.
Plan view of cap
Figure 3.5 - ISAT assembly
3.6.6 Porosity
Porosity measurements were taken of various samples of normally and electrically cured
specimens at various ages (DeVenny and Khalaf 1998).
The test was achieved by drying the samples (min 100g) in an oven at 105°C ± 5°C for a
period of 24 hours. The samples were then weighed and placed in a vacuum at 600mm
91
below atmospheric pressure for a period of 0.5 hours. The vacuum chamber was then
filled with distilled water and allowed to soak for a period of 24 hours. The samples
were then weighed in air and water and from this the porosity was calculated by:
(m - m )
P wa da xl00 Eqn.3.10
(mwa - mw)
where:
P == Porosity (%)
Illda Mass of dry sample in air (g)
mwa == Mass of saturated sample in air (g)
mw Mass of saturated sample suspended in water (g)
3.6.7 Water absorption
Water absorption was calculated for various samples of normally and electrically cured
specimens at various ages (DeVenny and Khalaf 1998).
Water absorption was achieved by drying the samples (min lOOg)in an oven at 10Soe ±
5°C for a period of 24 hours. The samples were then weighed and placed in a vacuum
at 600mm below atmospheric pressure for a period of 0.5 hours. The vacuum chamber
was then filled with distilled water and allowed to soak for a period of 24 hours. The
samples were then weighed in air and water and from this the water absorption was
calculated by:
m -m
W= wa da xl00 Eqn.3.l1
md•
where:
W Water absorption (%)
92
lllda Mass of dry aggregate in air (g)
mW8 Mass of saturated aggregate in air (g)
3.6.8 Density
Density measurements in accordance with BS 1881: Part 114 were taken of various
samples of normally and electrically cured specimens at various ages.
The volume of the samples (cube or core) was calculated by:
V= rna -mw Eqn.3.12

1000
where:
v Volume of sample (rrr')
Mass of sample in air (kg)
mw Mass of sample suspended in water (kg)
The samples (minimum of 1OOg)were then dried in an oven at 1Osoe ±soe for a period
of 24 hours. The samples were then weighed in air and from this the density was
calculated by:
rna
p=- Eqn.3.13
V
where:
p
The specific gravity could then be found by dividing the density by the density of water
(1000kglm\
93
3.6.9 Ultrasonic pulse velocity
This test involves a pulse of sound at various frequencies (depending on transducer

selection) being passed through a sample of known length. This pulse will travel faster
in a dense medium and as such can give a relative indication of compressive strength.
However, again this indication can only be assumed if the sample being tested is of type
that mirrors the specimen in-situ. Many factors greatly influence the readings achieved
from this piece of apparatus. These include temperature, stress history, path length,
moisture content, aggregate type and the presence of reinforcement.
From these variations it can be assumed that for a comparison with compressive
strength, variations of around 20% can be expected. However, concrete can be
qualitatively classified according to Table 3.11. The ultrasonic pulse velocity test
conforms to BS 1881: Part 203. A diagrammatical layout of the test can be seen in
Figure 3.6.
Trigger
Pulse generator
Time measuring
circuit
Transmitter Receiver Amplifier
Sample
Figure 3.6 - Set-up of ultrasonic pulse velocity meter
94
Ultrasonic tests were carried out on 150x150x150mm cubes at various ages.
Table 3.11 - Classification of concrete quality on the basis of pulse velocity
Longitudinal pulse velocity Quality of concrete

(km/s)
>4.5 Excellent
3.5-4.5 Good
3.0-3.5 Doubtful
2.0-3.0 Poor
<2.0 Very poor
3.6.10 Schmidt hammer
This test is based on the principal that the rebound of an elastic mass will depend on the
surface hardness of the test subject in question. From this test, information can be
gathered on surface characteristics up to 30mm deep in concrete. The results give only a
measure of the relative hardness of the cover zone and as such, a direct comparison with
other properties, such as compressive strength, is difficult unless a good degree of
calibration can be achieved. This calibration must be based on the particular mix under
consideration, and the concrete surface, curing history and age of specimens should
correspond as closely as possible to the in-situ concrete. The results from a Schmidt
hammer are significantly influenced by: coarse aggregate type, moisture condition,
curing history, cement type, cement content, member mass, degree of compaction,
surface type, surface carbonation, stress state and temperature (Nasser and Al-Manaseer
1987).
The range of correspondence between Schmidt hammer and compressive strength

results are not better than 15% and in most cases a value of 25% is more likely.
Schmidt hammer tests, in accordance with BS 1881: Part 202, were carried out on both
the electrically cured and normally cured concrete slabs over the lower surface at regular
spacing intervals at various ages.
95
3.6.11 Scanning electron microscope (SEM)
The scanning electron microscope was used to determine the microscopic structure of
both electrically and normally cured concrete samples at various ages. Specimens were
first coated in platinum to present a suitably reflective surface to the electrons.
A representation of the SEM can be seen in Figure 3.7. The electron gun produces a
stream of electrons, which travel through condenser lenses and condenser apertures to
form a tight, coherent beam which then interact with the sample and are detected by
various instruments to produce a visual image.
1.... ------- Electron gun

Electron beam
••••• +-Objective aperture
__ ---- Scan coils
Figure 3.7 - Scanning electron microscope set-up
3.7 DEC Set-up
3.7.1 Equipment used
An alternating power supply which consisted of a Regavolt Variac (step-down)

transformer was used to supply current through the cubes.
96
A Hydra 2625A data logger was configured to take temperature readings from a series
of thermistor wires placed within the electrically cured samples.
Two Thurlby 1504 r.m.s. multimeters and a Feedback EW604 electronic wattmeter were
used in the electrical circuit.
A diagrammatic representation of the connection of the power supply and the specimens
to be electrically cured can be seen in Figure 3.8. The circuit consists of the electrically
cured specimens connected to the alternating current power supply with a voltmeter
across the supply and an ammeter in series.
Ammeter Wattmeter Voltmeter
Supply
240V 50Hz
\
Cementitious samples
/
Transformer Variac
Figure 3.8 - Circuit diagram
3.7.2 Cube casting and curing

150x l50x 150mm plastic moulds were used for casting the electrically cured specimens.
These moulds were externally reinforced to reduce deformation at the higher
temperatures used as shown in Figure 3.9.
97
Two stainless steel plates of dimensions I50x250x2.5mm were used as electrodes
during the electrical curing. These were placed at opposite faces of the cube prior to the
placement of the material in the cube as shown in Figure 3.10 and were not coated in
demoulding agent due to the possibility that this procedure may seriously affect the
contact properties between the plate and the mix.
150
2.5
~~
I+- Steel plates ..
150 \ External
~ Steel plates .... 150
re
Base
/
Side elevation Plan elevation
Figure 3.9 - Plastic cube moulds
A thermocouple temperature sensor was embedded in the centre of each sample

immediately after vibration at a depth of approximately 20mm unless otherwise stated.
A set of three 150mm steel cubes were used as moulds for the normally cured cubes.
A polystyrene box was constructed and used to insulate the cubes being cured
electrically. The cube moulds were placed on a polystyrene slab 50mm thick which
formed the base of the insulating container and the remaining cover (also 50mm thick)
placed over the top.
98
Power, voltage and current level measurements were taken every 15 minutes during
heating and after this had been completed the levels of resistance were checked at
regular intervals by applying a power of 20W for a maximum of 20 seconds so as not to
significantly begin heating again.
The electrical resistance was calculated by:
Eqn.3.14
where:
R = Total resistance of samples (Q)
P = Power(W)
I = Current (A)
Electrical connectors
To power
supply
2.Smm stainless steel

sheet electrode
Figure 3.10 - Set-up for cube heating
99
3.7.3 Beam casting and curing
Four 1OOxlOOx400mm wooden moulds were used for casting the electrically cured
beams. These moulds were externally reinforced to reduce deformation at higher
temperatures.
Two stainless steel plates of dimensions 1OOx150x2.5mm were used as electrodes

during the electrical curing. These were placed at opposite faces of the beam prior to
the placement of the material in the mould as shown in Figure 3.11.
A thermocouple temperature sensor was embedded in the centre of each sample

immediately after vibration at a depth of approximately 20mm unless otherwise stated.
Similar moulds without electrode plates were used for the normally cured beams.
A polystyrene box was constructed and used to insulate the beams being cured
electrically. The beam moulds were placed on a polystyrene slab 50mm thick which
formed the base of the insulating container and the remaining cover (also 50mm thick)
placed over the top.
2.5mm stainless steel
To power
supply
Electrical connectors
Figure 3.11 - Set-up for beam heating
100
3.7.4 Slab casting and curing
A 400x600x 150mm wooden mould was used for casting the electrically cured slab.
This mould was externally reinforced to reduce deformation at the higher temperatures
used.
A stainless steel base plate of dimensions 400x600x2.5mm was used as an electrode

during the electrical curing. The other electrode consisted of the internal reinforcement
bars (l6mm diameter) as shown in Figure 3.12.
Thermocouple temperature sensors were embedded immediately after vibration in

various places within the sample at various depths.
The normally cured slabs were cast in a similar mould not containing an electrode plate.
A wooden lid was constructed to insulate the slab being cured electrically. This was
solely for heat insulation and no special care was taken to make the lid airtight.
300
600
..............................
J...40
·f
To power supply 2.5mm stainless steel
160m diameter Electrical connectors

reinforcing bars
Figure 3.12 - Set-up for slab heating
101
3.7.5 Thermistor placement
As stated previously, in the majority of experiments carried out, a thermistor

temperature sensor was manually placed in the centre of each sample approximately
20mm below the surface. However, a number oftests were carried out on cubes, beams
and slabs in which more thermistors were placed to monitor temperature gradients. The
thermistors were attached to 2mm diameter wooden dowel rods prior to pouring of the
concrete and were mounted on an external framework for stability throughout concrete
compaction. Figures 3.13-3.16 show the placement of the thermistors in the various test
set-ups.
37.5 37.5 37.5 37.5

IE ~IE ~
Steel plate 20
1 0<) 11
30
2 0 10
25
3
25
30
20
Plan view Cross-section (X-X)
Figure 3.13 - Cube thermistor location
100
1-5 6-10 11-15 16-20 20

15
tee Iplate 2
15
3
15
4 15
e
Concrete 5 20
Figure 3.14 - Beam thermistor location
102
Rebars
;~
x
300
_______ x
150
150
. Concrete ~--------~~._--~
Steel plate
!
130
1< >10101
100 50 50
Figure 3.15 - Slab thermistor location (A)
300
Rebars
20
30
300
X X 25
L_ ______
j.
25
300 30
20
Steel plate
5050 100
Figure 3.16 - Slab thermistor location (B)
3.7.6 Heating procedure
The experiments carried out involved the heating of samples in the arrangement shown
in Figure 3.8 with the use of a Variac power controller.
103
The samples were heated after the addition of water at the end of a variable natural
curing period (0.5-4 hours). This natural curing time was chosen to allow a degree of
structural strength to develop before heating stresses were incurred. Then a rise in
temperature was applied by manually varying the power input until the samples reached
a maximum value. At this stage samples were either allowed to cool in situ or an
isothermal period of varying length was achieved. Samples were then demoulded and
placed in a water tank at a temperature of 20°C ± 1°C (BS 1881: Part 111).
104
Chapter 4 - DIRECT ELECTRIC CURING OF MORTAR
4.1 Introduction
This chapter provides the results and discussion of the experimental and theoretical
investigation carried out on electrically cured mortar specimens. All testing was carried
out on ISOxISOxISOmm cubes unless otherwise specified. These results are
summarised in Figures 4.1-4.15.
4.2 Water/Cement Ratio

Figure 4.1 shows the relationship between the compressive strength and water/cement
ratio for both electrically and normally cured specimens. The figure shows that for all
W/C ratios, the electrically cured specimens achieved a higher 3 days compressive
strength than the normally cured specimens. Both types of specimen showed a general
trend of reduction in strength as the W/e ratio increased (Neville 1988).
30
2S •
N
--
Ii
Ii
.._
--
Z
.=
'Sil
20
1:1
...e IS
'"~
.~
'"
:l ..
Cl.
10
Ii • Electrically cured
Cl
U
• Normally cured
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
045 05 0.55 0.6 0.65 0.7 0.75 08
W/C
Figure 4.1 - Compressive strength at 3 days versus W/C ratio for electrically (2
hour delay, 300Clhr rise, 80°C maximum) and normally cured samples
105
With the application of electrical cunng, advantages concerning workability
requirements were substantial. An equivalent electrical 3 days compressive strength
was achievable with a 0.15 difference in W Ie ratio from the normally cured specimens.
Figure 4.2 shows the relationship between the percentage strength gain of electrically
cured over normally cured specimens for a range of WIe values. The figure shows that
the percentage increase in strength over normally cured samples was proportionately
higher at greater WIe ratios. As such the reduction of strength to be expected for
increasing WIe ratios was less for the electrically heated specimens. The actual
compressive strength gain at differing wle ratios decreased slightly for the 0.50, 0.57
and 0.62 W Ie ratio, showing that the actual cement content affected the strength gain.
45 10
40 9
35
• Percentage gain
---
~
0
'-'
30 • Actual gain
...=>
rl
0;=
6
')5 !..
~- IIQ
~
s1:1"
~ 5 S·
20 ---2:......
..
..." 4
e
eN
=-" 15
3'-'
10
• • 2
04 045 0.5 055 0.6 065 0.7 0.75 0.8
w«:
Figure 4.2 - Percentage and absolute gains of electrically cured (2hour delay,
300Clhr rise, 80°C maximum) specimens over normally cured specimens for
varying W IC ratios
From this it can be stated that the amount of cement present per rrr' reduced the apparent
increased rate of hydration but it seems evident that the reduced amount of cement
106
present was still achieving accelerated hydration reactions. The 0.75 w/e ratio points
may be due to an apparent decrease in WIe ratio due to water evaporation, thus causing
an equivalent strength gain. The other points, though still accelerated, may have been
offset by this reduction in water content to give a partially dehydrated mix, thus
reducing the apparent strength gain.
Figure 4.3 shows resistance versus time for various W/C ratios. This indicated that with
a higher W Ie ratio, resistance decreased. This was due to the increased conductivity of
the samples for the increased amount of water present. The relationship between
resistance and WIC ratio is apparent throughout heating. It can also be seen that
resistance increased over time in all cases and this was due to free water being bound
within hydration products, thus reducing conductivity.
100
90
80
-+-W/C =0.5
70
.-. ___ W/C = 0.57
'"El .......W/C =0.75
~ 60
~
Qj
Col 50
iii
!!
.!!l 40
rl
C(
30
.....
20
10
.__. ~
0
0 50 100 150 200 250 300 350 400
Time (minutes)
Figure 4.3 - Relationship between resistance and W/C ratio for electrically (2 hour
delay, 30oClhr, 800C maximum) cured samples
107
4.3 Heating Rates
Figure 4.4 shows a typical temperature versus time curve for a mortar sample. This
shows that the temperature during the linear rise can be accurately controlled and that
the maximum rate of cooling was approximately 3°Clhr. This is within the allowable
cooling rate given in several standards (Highways Agency 1998, Price 1969) but it may
be that for other geometrys or mixes, this may not be the case. Specifically, at edges
and comers, temperature losses must be minimised by the addition of a limited amount
of current to reduce temperature loss or by the use of insulation (Price 1969, Kirkbride
1971). The figure also shows that for the sample used and the thermistor position,
temperature time lags were not significant (Kafry 1993). This time lag is where the
temperature at the thermistor position continues to increase after power has stopped
being applied.
90
80
70
0'60
._
0
-=e
~ so
~40
..
a
~ 30
20
10
0
0 500 1000 1500 2000 2500 3000 3500
Time (minutes)
Figure 4.4 - Heating curve for mortar (W/C = 0.55, 2 hour delay, 40oC/hr rise,
80°C maximum)
Figure 4.5 shows the effects of differing heating rates (to a maximum of 80°C) on
samples of W/C = 0.57. It can be seen from this relationship that the maximum 3 day
108
compressive strength gain was achieved when a heating rate of approximately 40°CIhr
was used. This suggested that above this heating rate any increase in the hydration of
cement would be offset by the structural damage that would be occurring due to the
higher thermal stresses imposed on the samples. Higher rates of temperature rise could
not be achieved with the equipment used due to the likelihood of separation of the steel
electrodes from the mortar. The actual compressive strength gain can be explained by
the fact that the lower rates of temperature gain caused the samples to be at relatively
high temperatures (60-80°C) for significantly longer periods, thus allowing a
significantly higher rate of moisture loss. This may have caused the samples to become
increasingly desiccated and as such have relatively smaller strength gains.
25 r----------------------------------------------------------------------------------------------------, 10
20
15
.-. 10
~
•._, ...
~
=!.
=
.; IJQ
CIII
GI
CIII
0 ~---------------------- __ ~----------------------------------------------~o =s·
III
~
..
-=GuI
GI
Q.,
-5
~
e
El
-10
• Percentage gain
-5
-15
• Actual gain
-20
-25
0 10 15 20 25 30 35 40 45
Rate of beating (OC/br)
Figure 4.5 - Percentage and absolute gains of differing heating rates (2 hour delay,
80°C maximum) on samples ofW/C = 0.57
4.4 Resistance
Figure 4.6 shows the relationship between the resistance of electrically cured mortar
cubes measured immediately prior to compressive strength testing and the compressive
strength for W/C = 0.57. The samples shown used different rates of heating rise and/or
109
maximum temperatures. This indicated that there was no relationship between electrical
resistance and compressive strength for mortars over the range tested. This was
different to the behaviour of concrete and may have been due to intrinsic variability in
the electrically cured test specimens or to the short length of time before testing. At 3
days, the mortar had not had time to significantly use up the excess water (the main path
for conduction) in hydration product formation. As such, it would be reasonable to
expect similar values of resistance for different strengths at early ages.
30
....-
a
25
•
a •
~
-=;,
r:::I
Q>
20
•• ••
•
•
:: 15
.,'"
'"
Q>
rl
... 10
Q,
El
e
U
0
105 110 115
Resistance (ohms)
Figure 4.6 - Relationship between resistance immediately prior to compressive

strength testing and 3 days compressive strength
Figure 4.7 shows a typical resistance versus time curve for W/C = 0.55 with a 2 hour
delay and a 40°CIhr rate of rise to a maximum temperature of 80°C. It can be seen from
this figure that the resistance initially reduced during the first 60-80 minutes of heating
(Tamas 1982). This was partially due to temperature effects reducing resistance as the
mortar behaves as a dielectric but was also due to water absorption by the aggregates
present within the mix reducing the effective path length of the current, thus reducing
110
resistance. This is derived from McCarter and Curran (1984) where it has been stated
that current can flow in any or all of three paths:
• cement paste only

• aggregate only
• cement and aggregate combination.
It thus follows that if the aggregate starts in an initially dry state (insulator), then current
will not be able to flow through it and as such will have to flow along a comparatively
higher resistance path. With the absorption of water into the aggregates the resistance is
reduced (conductor) within and the current can follow a path of least resistance at least
partially through the aggregate, thus reducing the overall resistance. Resistance may
also be reduced by the increased levels of soluble ions (Ca ++, OH" and S04 -) in the pore
water during heating.
130
120
110
100
-'"
B
..a
90
80
--
Q
~
....
70
-Cl
eI
.l!l
rl
g:
60
50
40
30
20
10
0
500 1000 1500 2000 2500 3000 3500
Time (minutes)
Figure 4.7 - Typical resistance graph (W/C = 0.55, 2 hour delay, 40°C/hr rise, 80°C
maximum)
111
The resistance after approximately 200 minutes after gauging with water began to
increase (Tamas 1982). This was due to the hydration of the cement reducing the
amount of free water available for conduction of electricity. It was also likely to be in
some extent initially to be due to direct water loss from the samples by evaporation due
to the higher temperatures used.
140 r-----------------------------------------------------------, 90
120 • 80
•• 70
.-
a'"
100
• 60 -l
e
-=e
'-'
~
...
80
• ---+- Uncorrected
--
resistance
Corrected resistance
50 "g
Dl
a-
...
I:l
...
eo!
60 . .•.. Temperature 40 ;:
'"
.~ 'Cl
......•... n
cz::
...•......•.....•
30 '-'
40
20
20
10
0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0
0 500 1000 1500 2000 2500 3000 3500
Time (minutes)
Figure 4.8 - Corrected resistance versus time (W IC = 0.55, 2 hour delay, 40°C/hr
rise, 80°C maximum)
Figure 4.8 shows resistance, corrected resistance and temperature versus time. The
corrected resistance was calculated using Equation 2.2 (Wilson and Gupta 1996) and
using the value of exfor concrete as an estimate for mortar. While likely to be relatively
inaccurate, this estimation shows that the apparent resistance was likely to be seriously
different to the actual resistance. This would be significant at early ages during the high
temperature phase if resistance was being used to monitor the hydration reactions of the
sample. The corrected values of resistance can be seen to be significantly higher than
the uncorrected values of resistance. This takes into account the fact that the mortar acts
as a dielectric and as such, the apparent resistance at higher temperatures reduces.
112
Figure 4.9 shows the initial period of heating shown in Figure 4.8. From this it can be
seen that the corrected resistance did not decrease in value as did the uncorrected
resistance. However, this is not a contradiction of the explanation given for Figure 4.7.
This is because the corrected resistance showed a relatively slow rate of resistance rise
initially as the aggregates within the sample absorbed the free water and then as the
aggregates became saturated, the rate of rise of resistance was solely due to the moisture
loss used in the hydration reaction. The effect of the reduced rate of resistance rise was
relatively small when compared to the corrected resistance values for concrete, as within
mortar, the only aggregate used is sand, with a relatively low and quick absorption of
water. Also, within concrete, the large aggregate particles when acting as an insulator
cause a greater effective path length than the small aggregate particles within mortar.
80 90
70 ............. 80
_._ Uncorrected resistance
__ Corrected resistance
........
70
60
. . ...... Temperature
--
til
B 50
60 0-3
"El
-=~ 50 ~
~
y
1:1
!!
40
40 ~
-
"I
~
I:
.~ 30
til
~
--
30 '-'
0
("')
20
20
10 10
0
0 50 100 150 200 250 300 350 400
Time (minutes)
Figure 4.9 - Corrected resistance versus time (W/C = 0.55, 2 hour delay, 400C/hr
4.5 Maturity
Figure 4.10 shows values of maturity versus 3 days compressive strength for samples of
W/C = 0.57. A linear best-fit line can be achieved for the tests containing no isothermal
113
period (coefficient of determination equal to 0.72) but if the test involving a I hour
isothermal period was taken into account then the coefficient of determination decreased
to 0.64.
From this it can be stated that maturity was not a good indicator for final strength values
where variable heating strategies have occurred, but was more consistent for a range of
maturity values for a specific heating methodology. The range of maturity values was
extremely small and indicated that the proportionally large maturity values incurred
during cooling virtually overwhelmed any maturity differences incurred during the
initial rapid heating.
3600
3580
3560
3540
.-
10
.t:I
U 3520
,_.
0
C 3500
'1:
~
= 3480
~
3460
• No isothermal period
3440 • Isothermal period (I hour)
3420
3400
10 15 20 25
Compressive strengtb (N/mml)
Figure 4.10 - Values of maturity versus 3 days compressive strength for samples of
W/C=O.S7
4.6 Maximum Temperature

Figure 4.11 shows the relationship between maximum temperature of electrically cured
specimens and compressive strength. The figure indicates that for the small range tested
(approximately 15% variance), a relationship exists between maximum temperature and
114
compressive strength. This differed to concrete based work where a maximum
temperature of 80°C is not recommended (Price 1969, Kirkbride 1971). The higher
temperature tolerance of mortars can be attributed to absence of coarse aggregate
particles in the mix. Coarse aggregate particles have a much higher coefficient of
thermal expansion (4-5xlO-6/0C) than fine aggregate (2xlO-6/0C) and as such an increase
in maximum temperature creates greater thermally induced stresses between the matrix
and the aggregate (El Hussein and Abd El Halim 1993).
o L_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
60 65 w ~ w 85 90
Maximum temperature (Oq
Figure 4.11 - Effect of the maximum temperature on the 3 days compressive

strength (2 hours delay, 30°Clhr rise)
4.7 Long-term Compressive Strength

Figure 4.12 represents the relationship between compressive strength and age for
electrical (W/C = 0.5, 2 hour delay, 40°CIhr, 80°C maximum) and normal curing. This
shows that compressive strength was increased at early ages using DEC but the rate of
gain of strength was decreased when compared to normal curing (Kafry 1993). This
was likely to be due to microcracking caused by increased thermomechanical stresses
when heating. The heating increases the rate of hydration at early ages but damages the
115
microstructure because at the time of maximum temperature, the structure of the mortar
is not strong enough to resist the differential expansion of the different constituents and
as such is damaged. This thermomechanical damage was also caused by the rate of
increase in temperature chosen. The fact that samples were cured for three days in air
was also apparent as the normally cured samples did not achieve the long-term strength
development that would be expected if the samples had been water cured after 1 day
from casting (Neville 1988).
35 r---------------------------------------------------------,
30
«:
a
~ 25
.Cl
t., 20
-.::e
1:1
'"
~ 15
'" • Electrically cured

~
=. 10
• Normally cured
a
~
u
o 10 15 20 25 30
Time (days)
Figure 4.12 - Effect of DEC (2 hour delay, 40°C/hr rise, 80°C maximum) on long
term compressive strength
4.8 Durability
Figure 4.13 shows the initial surface absorption of electrically and normally cured
samples at 28 days (W/C = 0.5, 40°CIhr heating rate to a maximum of 80°C). This
shows that durability was not adversely affected by DEC for the particular mix and
heating regime chosen. This was likely to be due to the modification of capillary pores
caused by the increased temperature used which more than offset the damage caused to
the microstructure by thermomechanical stresses from the differential expansion of the
116
constituents. The compressive strength of both the normally and electrically cured
samples were approximately equivalent (4% difference), but the values of surface
absorption at 10 minutes (ISA 10) were significantly different (9% difference).
02 ,------------- -,
0.15
• Electrically cured
.-..
--El'"
.....-.. • Normally cured
:::;
01
El
'-'
'-'
«
-
tr.J
0.05
o 10 20 30 40 50 60 70
Time (min)
Figure 4.13 - Initial surface absorption of electrically and normally cured samples
at 28 days (W/C = 0.5, 2 hour delay, 40°Clhr rise, 80°C maximum)
4.9 Isothermal Period

Figure 4.14 shows the compressive strength after 3 days (W/C = 0.55) for a non-
isothermal period and an isothermal period of 1 hour at 80°C for a linear rise of 40°CIhr
and a delay period of 2 hours. This shows that the non-isothermal period samples had
an increased rate of compressive strength gain compared to the isothermal period
samples. This small reduction in strength may have been due to excessive water loss
during the isothermal period, causing desiccation of the samples. Both heating regimes
achieved a compressive strength gain over the normally cured samples in the order of 8
and 15% (1.6 and 3.4N/mm2) for isothermally and non-isothermally electrically cured
samples respectively.
117
Normally cured
Figure 4.14 - Compressive strength after 3 days for isothermal and non isothermal
electrical (2hour delay, 400Clhr rise, 80°C maximum) curing
4.10 Delay Period

Figure 4.15 shows the compressive strength of electrically cured samples of differing
delay periods at 3 days. The samples were of W/C = 0.55 with a 40°CIhr rate of rise to
a maximum temperature of 80°C. This shows that the delay period determined to a
large extent the expected compressive strength with the 2 hour delay attaining
approximately 11% greater compressive strength than the 1 hour delay sample. This
was due to the damage caused to the structurally less well-formed 1 hour delayed
sample from the thermally induced stresses present during heating. Both delay periods
showed a greater compressive strength at 3 days than the normally cured samples.
118
Normally cured
Figure 4.15 - Effect of delay period prior to heating on 3 day compressive strength
(W/C = 0.55, 400Clhr rise, 80°C maximum)
4.11 Conclusions
1. Over a range of W/C ratios electrically cured mortar samples achieved a higher
initial compressive strength than normally cured samples and the proportional
effectiveness of electrically cured mortar specimens was found to be increased at
higher WIC ratios.
2. Resistance can be used as a guide to the amount of water in mortar with higher W/C
ratios giving lower resistances at all ages.
3. For a 2 hour delay period, a rate of temperature rise of 40°CIhr gave the highest
initial (3 day) compressive strength for the 0.57 W/C ratio mortar samples with a 2
hour delay.
4. There appeared to be little correlation between 3 days mortar compressive strength

and resistance.
119
5. The initial drop in resistance during heating was due to water absorption by
aggregates which reduced the effective path length of the electrical current.
6. Mortar behaved as a dielectric having a reduced apparent resistance at higher

temperatures. The effect of temperature during heating seriously affected the
resistance properties of the mortar samples. This effect became negligible after
approximately 3 days.
7. Maturity was found to be a good indication of compressive strength when DEC was
used.
8. The value of maturity during cooling was significantly higher than the maturity
measured during the initial heating of samples. This meant that accurate monitoring
of temperature after the initial heating concluded was required in order to estimate
compressive strength.
9. For mortar, a maximum temperature of approximately 80°C provided the greatest

initial (3 day) compressive strength for a 30°Clhr rise with a delay period of2 hours.
10. Long-term (28 day) compressive strength was not adversely affected by DEC in the
samples tested although the rate of compressive strength development at later ages
(>28 days) was reduced when compared to normally cured samples. The reduced
rate of compressive strength development at later ages in mortar was likely to be due
to differential expansion of the aggregates and the cement matrix incurred by
heating.
11. The durability of mortar samples (as measured by the ISAT) did not appear to be
adversely effected by DEC. This was likely to be due to capillary pore modification
during heating.
120
12. The inclusion of an isothermal period at maximum temperature appeared to have an
adverse effect on initial (3 days) compressive strength. This was likely to be due to
increased microstructure damage incurred at the maximum temperature and
increased temperature gradients within the sample.
13. The delay period chosen had a significant effect on initial (3 days) compressive
strength. Results show that a 2 hours delay produced a higher compressive strength
than a 1 hour delay.
121
Chapter 5 - DIRECT ELECTRIC CURING OF CONCRETE
5.1 Introduction
This chapter provides the results and discussion of an experimental and theoretical
investigation carried out on electrically and normally cured concrete samples. These
results can be seen in Figures 5.1-5.69 and Tables 5.1-5.2.
All testing was carried out on ISOxlS0xlSOmm cubes of a C30 (10-30mm slump) mix
unless otherwise specified and the mixing, casting, curing and monitoring was
implemented in accordance with the experimental procedures stated in Chapter 3.
5.2 Normally Cured Samples

Figure 5.1 shows compressive strength versus time relationship for normally cured
samples of mix design target mean strength = 43N/mm2 and a slump = 10-30mm. This
shows the variability of mix compressive strengths used over the investigation. A
logarithmic best fit curve (compressive strength = 12.03 x Ln (time) + 9.64) was used
with a coefficient of determination = 0.95. This curve was used in any further
comparisons of compressive strength.
A regression analysis was carried out on the data contained in Figure 5.1 as another
method to determine the acceptability of the best-fit curve. To determine whether x
provides information in predicting y the hypotheses for this test are:
Ho:B=O
Ho:BoO
Testing the hypothesis involved a t-test of the regression line slope with degrees of
freedom = 38. From this it was found that the t-stat from the regression = 26.3. The t-
stat @ 99% level = 2.42 and since 2.42<26.3 it can be said with 99% confidence that
122
B<>0 and the null hypothesis can be rejected. This gives an added indication that the
best-fit curve used is valid.
70
60 •
.-.a •
El
.....
Z
50
•
'-'
.=
bIl 40
•
=~...
...
'"~
• y = 12.03Ln(x)
R2
+ 9.6444
= 0.9529
.~ 30
'"
~
...'"
El -'0
Q.
=>
U
10
0
•
0 10 15 20 25 30 35 40 45 50
Time (days)
Figure 5.1 - Compressive strength versus time for normally water cured samples
5.3 Effect of Cover

Figure 5.2 shows the effect of covering of electrically cured samples on 3 day
compressive strength. This shows that, for the heating regime used (1 hour delay,
40°Clhr rise to 80°C maximum), removing the 50mrn thick expanded polystyrene cover
reduced compressive strength by 8.3N/mrn2. The uncovered compressive strength can
be seen to be lower than the equivalent age normally cured sample. This was due to the
excess moisture loss incurred by the uncovered sample (Sohn and Johnson 1999). With
the covered samples, moisture was retained within the covering, creating a high
humidity environment that allows a degree of moist curing. With the uncovered
samples, evaporation could readily escape into the atmosphere, thus allowing
desiccation, particularly of the surface region of the concrete. The removal of the cover
may also have effected compressive strength by the loss of thermal insulation. This
123
created higher temperature differentials within the concrete, thus creating higher
temperature induced stresses and a further reduction in compressive strength (Kirkbride
1996).
Normally cured
Figure 5.2 - Compressive strength versus time for covered and uncovered samples
(1 hour delay, 40°C/hr rise, 80°C maximum)
Figure 5.3 shows uncorrected resistance versus time for covered and uncovered
electrically cured samples of the same configuration as shown in Figure 5.2. This shows
that the uncovered samples attained a significantly larger resistance value as heating
occurred until approximately 1200 minutes whereupon resistance values almost ceased
to rise. This was due to an increased moisture loss from the uncovered samples creating
a desiccated environment which initially caused the increase in resistance. The low
resistance rise after 1200 minutes was due to the lack of free water retarding hydration.
This confirmed the relatively lower 3 day compressive strength of the uncovered
samples shown in Figure 5.2. The covered samples showed an initially lower resistance
value as the amount of free water was proportionally higher, but then continued to
increase resistance as the free water was bound in hydration reactions. The [mal
124
resistance values for the covered samples seemed to indicate a more fully hydrated
sample and a corresponding compressive strength (Orchard 1973). These results show
the importance of using identical sample environments if resistance values are to be
used in assessing the degree of hydration achieved or WIC ratio of concrete.
700 r---------------------------------------------------------~
600
500
-+-Covered
200 ___ Uncovered
100
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 500 1000 1500 2000 2500 3000 3500 4000 4500
Time (minutes)
Figure 5.3 - Resistance versus time for covered and uncovered samples (1 hour
delay, 40°C/hr rise, 80°C maximum)
Figure 5.4 shows the resistance for the initial period of heating for the electrical heating
regimes used in Figure 5.2. This shows an initial resistance difference between the
mixes which indicated that the mixes had slightly different W IC ratios. However, this
difference was not believed to be significant. The figure also shows that the uncovered
samples had an identical degree of change in resistance as the covered samples until the
maximum temperature was approached at approximately 120 minutes and onwards (60-
80°C). At this time a larger rate of rise of resistance could be observed. This was due
to the fact that significant levels of evaporation can only be achieved at temperatures in
the region above 60°C. This may be important in further investigations of DEC where
insulation is not available and a heating regime reducing moisture loss is required.
125
200
180
160
-a'"
.1:1
140
120
,e
"
eo. 100
-1:1
CIII
'"
·il
IX
80
60
-+-Covered
---- Uncovered
40
20
0
0 50 100 150 200 250 300 350 400
Time (minutes)
Figure 5.4 - Resistance versus time for covered and uncovered samples (1 hour
delay, 40°Clhr rise, 80°C maximum)
5.4 Water/Cement Ratio

Figure 5.5 shows the relationship between the compressive strength and W/C ratio for
both electrically and normally cured specimens with a heating delay of 1 hour to a
maximum temperature of 80°C. The figure shows that for all W/C ratios, the
electrically cured specimens achieved a higher 1 day compressive strength compared to
the normally cured specimens. Both types of specimen showed the normal trend of
reduction in strength as the W/C ratio increased (Goto and Roy 1981, Neville 1988).
Figure 5.6 shows that the percentage compressive strength gains of electrically cured
over normally cured specimens for the above range of WIC values vary from 60% for
W/e=O.55 to 219% for W/e=O.7. It can also be seen from this figure that the actual
compressive strength gain was higher at higher W Ie ratios. This was due to the
increased water content of the mixes, which allowed a degree of moisture loss, without
incurring desiccation effects. As such the reduction of strength to be expected for
increasing WIe ratios at 1 day was less for the electrically heated specimens.
126
25 ,---------------------------------------------------------------,
• Normally cured
05 0.55 0.6 0.65 0.7 0.75
w«:
Figure 5.5 - W/C versus compressive strength at 1 day (1 hour delay, 40°C/hr rise,
80°C maximum)
250.-------------------------------------------------------------,20
•
200
15
.-
~
t..,
1:1 150
.;
Ij.I)
"
Ij.I)
!!
1:1
100
"!::
"
Q.,
• Percentage gain
• Actual gain
50
o L_~~~ __ ~~~~ __ ~~~~ __ ~~~~ __ ~~~~ __ ~~~~ __ ~ 0

0.5 0.55 0.6 0.65 0.7 0.75
w«:
Figure 5.6 - Actual and percentage gain in compressive strength of electrically (1

hour delay, 40°C/hr rise, 80°C maximum) and normally cured samples at 1 day
127
Figure 5.7 represents compressive strength versus W/C ratio after 3 days of air curing.
This shows the effect of low W IC ratio adversely affecting compressive strength. The
assumed compressive strength for normally cured samples would be in the region of
2
75N/mm but the actual compressive strength was 28.2N/mm2 for normally cured and
2
37.2N/mm for electrically cured samples. This shows the likely presence of an
excessively dry mix where honeycombing and poor particle bonding is present (Neville
1988). However, the electrically cured samples still showed a comparative strength gain
over the normally cured samples and as such DEC may be used to improve otherwise
dry concrete.
80 ,----------------------------------------------------------,
70
• Normally cured
A Electrically dry
x Normally dry
10
0.3 0.35 04 045 0.5 0.55 0.6 0.65 07 0.75

W/C
Figure 5.7 - Compressive strength versus W/C after 3 days air curing (1 hour
delay, 40°Clhr rise, 80°C maximum)
Figure 5.8 shows the compressive strength versus W/C ratio for a range of mixes of
adequate W IC ratio after 3 days of air curing. It can been seen from this figure that the
electrically cured samples attained a significantly higher compressive strength after 3
days than the equivalent normally cured samples. This compressive strength gain was
128
approximately equal for all WIC ratios tested and as such a compressive strength gain
could be assumed for all WIC ratios used in the investigation.
40 ,----------- -,
35
--e
....e 30 • Electrically cured
• Normally cured
Z
:; 25
b'Il
1:1
~ 20
III
";.
.~ 15
e
C.
ae 10
o L-~~~~ __ ~~~~~ __ ~~~~~~~~~~~~ __ ~~~~~
0.5 055 0.6 0.65 07 07<
W/C
Figure S.S - Compressive strength versus WIC after 3 days air curing (l hour
delay, 40°Clhr rise, SO°C maximum)
Figure 5.9 shows the percentage and actual compressive strength gain after 3 days for
various WIe ratios. This shows that the actual gain in compressive strength remained at
approximately a constant level while the percentage gain rose proportionally with the
increase in WIC ratio. This shows that DEC was proportionally more effective at higher
WIe ratios and as such significantly higher performance could be gained with increasing
w«: ratios.
129
90 r--------------------------------, 10
80
•
70
;0-
"
6 ~
IJQ
)- 2!.=
Z
4 a
51""
• Percentage gain 3--
20 • Actual gain
10
o ~~~~~~-~~~~-L~~~~~~~~~-~~~~~~~O
05 0.55 0.6 0.65 0.7 075
W/C
Figure 5.9 - Actual and percentage gain of electrically cured over normally cured
samples for different W/C after 3 days air curing (1 hour delay, 40 C/hr rise, 80 e D D
maximum)
Figure 5.10 shows the relationship between resistance and w/e ratio. The figure clearly
shows that the resistance values were significantly affected by the amount of water
present within the mix (Neville 1988). The initial reduction in resistance was due to the
absorption of water by the aggregates reducing path length for the current and increased
levels of soluble ions in the pore water. The temperature at demoulding was
approximately equal to the reference temperature used in Equation 2.2 and as such the
final resistance may be considered to be at least a guide to the water content in concrete
of identical mix design, temperature history and geometry (Khalaf and Wilson 1999).
130
&hl r-------------------------------------------------------------~
500
__ 41X)
a
-=.._.co
Ii
CJ 300
a
~
.~
'"
0:: 200 -+- W/C = 0.55
___ W/C = 0.6
....... W/C = 0.63
lOO ""* W/C = 0.7
o 500 I()()() 1500 2()()() 2500 3()()() 3500 4000
Time (minutes)
Figure 5.10 - Resistance versus time for different W/C (1 hour delay, 40°C/hr rise,
80°C maximum)
140
120
100
.-.
'"El
.1:1
80
~
Ii
CJ
a
~ 60
'"
.~
0::
40 -+- W/C = 0.55
___ W/C =0.6
....... W/C = 0.63
20 ""*W/C =0.7
0
0 50 100 150 200 250 300
Time (minutes)
Figure 5.11 - Resistance versus time for different W/C (1 hour delay, 40°C/hr rise,
80°C maximum)
131
Figure 5.11 represents the resistance for varying W/C ratios over time for the first 250
minutes after gauging with water. The relationship shows that initially the W/C ratio
had only a minor effect on resistance, but at later ages as hydration continued, the
resistance value differences became more pronounced. At 250 minutes, compressive
strength can be in the region of 1 day normally cured compressive strength and as such
differences in WIC ratio and strength are more pronounced than at 60 minutes. This
may be useful in accelerating the accurate assessment of WIC ratios using resistance
measurements.
Figure 5.12 shows the corrected resistance using Equation 2.2 against time for differing
W/C ratios.
250 r---------------------------------------------------------~
200
.-.
III
e
-=~
~ 150
=
!
.~
E
~ 100
~
e -+- W/C = 0.55
u ---W/C =0.6
50
...... W/C=0.63
"""*" W/C = 0.7
o 50 100 150 lOO 250 3()O
Time (minutes)
Figure 5.12 - Corrected resistance versus time for different W/C (1 hour delay,
40°C/hr rise, 80°C maximum)
This again shows a similar relationship to Figure 5.11, but without temperature
dependent influences on resistance. These temperature independent resistances showed
a reduced rate of gain initially caused by a combination of resistance increase due to
132
hydration and a resistance drop caused by the absorption of free water into the
aggregates (Khalaf and Wilson 1999).
5.5 Effect of Delay Period

Figure 5.13 shows the very early age strength properties for 1 stage heating with
different delay periods (43N/mm2 target mean strength mix). The delay period of 1 hour
matched the normally cured compressive strength at 24 hours in only 4 hours after
gauging with water. The 1 day strength of electrically cured specimens was also
significantly larger than the normally cured specimens, with increases of up to 18N/mm2
being readily achievable. The 1 hour delayed samples were of a higher 4 hour
compressive strength than the other delays due to the increased maturity values (Saul
1951) at the time of testing.
30 ,--------------------------------------------------------,
25
• Age 4 hours
• Age 24 hours
Normal 24 hours
o L_ ~ ~ ~ ~ ~ ~ ~ ~
o
Delay (hours)
Figure 5.13 - Early age compressive strength (40°Clhr rise, 80°C maximum)
Figure 5.14 shows the relationship between compressive strength over differing ages for
a 43N/mm2 target mean strength mix with various delay periods prior to electrical
133
curing of specimens to a maximum of 80°C. The figure shows that at I day the I hour
delay time had a greater compressive strength than the 4 hours delay time, but at 7 and
28 days the situation was reversed (Pfeiffer and Landgren 1981). This was probably due
to microcracks caused by thermomechanical forces exerted upon the structurally less
well formed 1 hour delayed specimen that reduced the ultimate strength with age. It
also shows that the 3 hours delayed specimen was the only specimen to achieve its
target mean compressive strength at 28 days.
50 ~__~__~__~__~__~__~__~__-__- - __
- __-__-__-__-__-__-__-__- - __-__-_-_- __-__-__- - __-_- __-__-__-__~ ~__
~__~__~__~__~__~__~__~__~ ~__~__~__~__~__~__-__-__~ ~__~__~_~__~_~
Nonnal28
.-, 40
Ne
e
Z.._
•
.
~ 30 Nonnal7
1)1)
•
-
&:I
~
II)
~
.~ 20
e
Co
.Age
.Age
1 day
7 days
ee A Age 28 days
U 10 __________________ ______________________ __ _ - - __- __- - - - __- __- __
Nonnall
o L_ ~ ~ ~ ~ ~ __ ~~ __ ~ ~~ __ ~~ __ ~
o 4
Delay (Hours)
Figure 5.14 - Compressive strength versus delay period (40°C/hr rise, 80°C
maximum, 1 day demoulding)
Figure 5.15 shows the percentage of compressive strength achieved at different ages
relative to 28 day compressive strength for various delay periods to a maximum of
80°C. This shows that the 1 hour delay achieved the majority of its 28 day strength
(71 %) after only 1 day, whereas the normally cured specimens only achieved
approximately 20% at the same age. The other electrically cured specimens also
achieved high proportions of strength (55-65%) after 1 day. However, the 1 hour delay
period had a final lower compressive strength (32N/mm2) than any of the other regimes
134
tested (40, 46, 41 and 49N/mm2 for the 2, 3, 4 hours delay and normally cured
specimens respectively). This was likely to be due to increased microcracking in the I
hour delayed specimen, as the mix had not acquired enough strength to resist the
thermomechanical stresses caused by heating.
• Delay 1 hour
• Delay 2 hours
A Delay 3 hours
X Delay 4 hours
X Normally cured
o L-~~~~~~~~~~~~~~~~~~-L~~~~~~~~~~
o 10 15 20 25 30
Time (days)
Figure 5.15 - percentage of compressive strength achieved at different ages relative

to 28 day compressive strength for various delay periods (40°C/hr rise, 80°C
maximum)
Figure 5.16 represents the compressive strength over differing ages for a mix with
various delay periods prior to heating. The mix is a 43N/mm2 target mean strength with
slump of IO-30mm and was demoulded after 1 day. This shows that 3 hours delay prior
to heating was the most beneficial to strength gain. It also shows that normal curing had
a lower compressive strength than any heating regime at 1 day but had greater
compressive strength than any heating regime at 28 days.
135
Extrapolated to ages of 100 days shows that all mixes should continue to gain
compressive strength at varying rates and that all of the mixes, except the delay for 1
hour heating regime, would attain the target mean strength of 43N/mm2.
~ r---------------~--r---~~--------_,----r_~--._--._~~
so I
I I
I I
r--~;-I hour dcl;~ .L.
• 2 hour delay
~ 3 hour delay
10 X 4 hour delay
X Normally cured
10 100
Time (days)
Figure 5.16 - Compressive strength versus time for various delay periods (40°C/hr
Figure 5.17 shows the effect of delay period on the 3 day compressive strength. This
shows that for the heating regime used, a 2 hour delay was more effective (17.1 Nzmrrr',
62 % gain) than a 1 hour delay period before DEC (Xuenquan et al 1986). However, the
1 hour delay period still showed a 7.5 N/mm2 (27%) gain in compressive strength. This
may be adequate where time constraints are more important to the end user than a higher
compressive strength.
136
2 Hour delay
Percentage
gain
maximum, 3 day demoulding)
~ c
__
~__
~
__
~__
~
__
~__
~
__
~__
~
__
~__
~
__
~__
~
__
~__
~
__
~__
~
__
~__ =_~_=_=_=_=_=_=_=_=__
=__
=__
=__
=__
~__
~__
~__
~__
~__
~__
~__
~__
~__ =_=_~_~_~_=_=_=_=_=_=_=_=_=_=_=_=_=_=_~__
~__
Nonnal28
50
Ne
e
~ 40
-
-=
OIl
=
CI.l
.:: 30
Normal7
'"...
CI.l
';;; I
-
• Age 1 day
.Age 7 days J
E
Q,
20 '---~ Age 28 days _
e
o
u
10 Normal I
o ~----~----~----~------~----~----~----~------~----~----~
o 2
Delay (Hours)
maximum)
137
Figure 5.18 shows the relationship between compressive strength over differing ages for
a 43N/mm2 mix with various delay periods prior to electrical curing of specimens to a
maximum of 60°C. The figure shows a similar relationship to the differing delay
periods with a maximum temperature of 80°C (Figure 5.14), but with comparatively
higher electrically cured compressive strengths at 60°C for all delays at 28 days and no
significant losses of compressive strength at 1 and 7 days. This was likely to be due to
the reduced microstructural damage incurred by the relatively low (when compared to
the 80°C maximum) temperature gradients present at 60°C.
100
~
• 9()
'-'
...~
.Cl
80
1:1
...t
'"II;. 70
.;;
..'"
II
e,
60
la 50
Q
~
~
III
." 40
• Delay 2 hours
...
QC
M
..
Q 30 • Delay 3 hours
...~
III
1:1
20
4 Delay 4 hours
..
II
~
10
x Normally cured
=-
0
0 10 15 20 25 30
Time (Days)
Figure 5.19 - Percentage compressive strength achieved at different ages relative to

28 day compressive strength for various delay periods (40°C/hr rise, 60°C
maximum)
Figure 5.19 shows the percentage of compressive strength achieved at different ages
relative to 28 day compressive strength for various delay periods to a maximum of
60°C. The figure shows a closer degree of percentage gain for the electrically cured
samples of both short and long term ages than Figure 5.15. The electrically cured
samples achieved approximately 50% of their respective 28 day strengths (44.1, 48.4
138
and 40.2N/mm2 for the 2, 3 and 4 hour delays respectively) at 1 day. As with Figure
5.15, the 2 hours delay period achieved a larger percentage of its 28 day strength at 1
day than any other delay period, but the percentage increase after this time was not as
great as the other delay periods.
180
160
-+- I hour delay
--
III
E
140
--- 2 hour delay
...._ 3 hour delay
.c
--"
e
IJ
I~u
-+- 4 hour delay
.."=
.~
III
lOO
.." 80
'B
~.....
Q
6IJ
U
40
Normally cured resistance
20
0
0 50 100 150 200 250 300 350
Time (minutes)
Figure 5.20 - Corrected resistance versus time for different delay periods (40°C/hr
Figure 5.20 shows the corrected resistance for a number of different delay periods
(40°C/hr rise, 80°C maximum). From this it can be seen that the initial resistance of
each sample corresponded to the resistance at that time of a normally cured sample.
This shows that the resistance of a normally cured sample reduced slightly until
approximately 150 minutes after gauging with water, whereupon it rose. This was
partially due to moisture absorption of water by the aggregates, thus changing them into
a more conductive material, hence reducing resistance. It may also be partially due to
increased levels of soluble ions (Ca ++, OH- and S04-) in the pore water during heating.
The rise in resistance corresponded to free water being used in hydration reactions
(Khalaf and Wilson 1999). The rise in resistance of the normally cured samples was not
139
as pronounced as the rise in all of the electrically cured samples, showing that hydration
was proceeding at an accelerated pace. The resistance at the end of heating can be seen
to be dependent on the initial delay period. This could be used as a check to estimate
the initial delay period and thus be an aid in predicting the short and long-term
properties of the sample in question.
5.6 Effect of Maximum Temperature

Figure 5.21 shows the effect of maximum temperature on compressive strength for the 3
hour delayed samples. From this it can be seen that the 80°C maximum temperature
achieved a higher initial 1 day compressive strength (28.3N/mm2) than the 60°C
maximum temperature (21.6N/mm2). This situation was reversed at 28 days with the
60°C maximum attaining a compressive strength of 48.4N/mm2 compared to 46N/mm2
for the 80°C maximum.
00 ,----------------------------------------------------------,
50
• Maximum SOC
• Maximum 60C
• Normally cured
o 10 15 20 25 30
Time (days)
Figure 5.21 - Compressive strength versus time for 3 hour delay period samples
(40°Clhr rise, 60°C and 80°C maximum)
140
The initial increase in strength was due to the initially increased rate of hydration
achieved by using higher temperatures and the relatively higher maturity values at 1 day
(Saul 1951). At later ages, damage caused by the increased thermomechanical stresses
induced, reduced long-term compressive strength values. Both electrically cured
samples attained their target mean strength (TMS) of 43N/mm2 at 28 days but did not
attain the same compressive strength as the normally cured samples at 28 days.
5.7 High Slump and High Strength Mixes

Figure 5.22 shows the compressive strength over time for a high slump (60-180mm)
mix with TMS = 43N/mm2• This shows a similar relationship to mixes of slump 10-
30mm with the compressive strength of the electrically cured samples being initially
higher (4N/mm2) with the situation reversing after several days. From this it can be
inferred that DEC could be applied to various mix designs as long as preliminary testing
has been carried out on the specific mix.
70
60
N-E!
--E!
SO
z;
'-'
-= ..
DIl 40
III
!;
...
III
II>
30
;;;
.. III
II>
Q,
20
E! • Normally cured
e
U
10
0
0 10 IS 20 25 30
Time (days)
Figure 5.22 - Compressive strength versus time for a high slump mix (60-1S0mm)
(1 hour delay, 40°C/hr rise, SO°C maximum)
141
Figure 5.23 shows the compressive strength over time for a high strength mix with TMS
2
= 63N/mm • This shows a similar relationship to mixes of TMS = 43N/mm2 with the
compressive strength of the electrically cured samples being initially higher with the
situation reversing after several days. The initial compressive strength gain of the
electrically cured samples was found to be 9N/mm2 which is equal to a 24% increase
relative to normally cured samples. This larger percentage increase was thought to be
due to the proportionally higher amount of cement within the mix. It gives evidence
that the effectiveness of DEC is improved with the amount of cement/m' within the
mix. This was due to the increased hydration per kg of cement associated with DEC.
100 ,---------------------------------------------------------,
• Normally cured
10
o 10 15 20 25 30
Time (days)
Figure 5.23 - Compressive strength versus time for a high strength mix (63N/mm2
target mean strength) (1 hour delay, 40°Clhr rise, 80°C maximum)
Figure 5.24 shows resistance versus time for various mix types. From this it can be seen
that the different mix types have a large effect on resistance. The C30 mix with 10-
30mm slump had a greater resistance at later ages than the C30 mix with 60-J80mm
slump. This was due to the proportionally higher percentage of water to the rest of the
mix constituents thus reducing the resistance of the high slump mix. The amount of
142
water as a percentage of total constituents was 10.4% and 8.4% for the high and low
slump mixes respectively. These percentages were used, as the W/C ratio in the two
mixes were identical (0.55). This was not true for the C50 mix which had a water:
(cement + aggregates) ratio of 8.4%, but had a greater mass of cement per m', thus
reducing the amount of free water present in the mix after a degree of hydration had
occurred, thus increasing resistance (Neville 1988).
~ ,---------------------------------------------------------,
R(KI
700
-. 600
III
8
-=,e.soo
41
Col
c
!!.wo
III
°5
IX 300
-+- C 30 (l0-30mm slump)

200
___ C 50 (10-30mm slump)
lOO -+-C 30 (60-180mm slump)
o SOO 1000 ISOO 2000 2S00 3000 3500 4000 4500
Time (minutes)
Figure 5.24 - Resistance versus time for differing mix types (1 hour delay, 40°C/hr
Figure 5.25 shows the first 300 minutes of resistance measurements in Figure 5.24.
This shows the time during and immediately after accelerated curing. From this it can
be seen that the C50 mix had an initially similar resistance value to the other mixes, but
increased at a more rapid rate, due to free water being used at an increased rate for
hydration due to the proportionally higher cement in the mix. The lower resistance
values of the high slump mix over time were due to the higher free water content that
saturated the mix and was only partially absorbed by the aggregates at early ages, thus
reducing the effective path length for the current to flow through.
143
140
I ~(1
IO(_)
--'"E
.c:I
,e 80
~
...
..
1:1
60
'"
.;;
~
cz::
40
-+- C 30 (10-30mm slwnp)
...... C 50 (10-30mm slump)

~O
.,._ C 30 (60-1S0nun slump)
0 50 100 150 200 250 300

Time (minutes)
Figure S.2S - Resistance versus time for differing mix types (1 hour delay, 40°C/hr
S.8 Resistance
Figure 5.26 shows the relationship between electrical resistance of specimens at 1 day
and the compressive strength for a range of curing regimes, mixes and ages. From this
it can be seen that resistance was not a good indicator of compressive strength at early
ages (1 day). This was because the hydration had not occurred enough to allow
differences in resistance to be measured. Within the mix at this time were still large
quantities of free water which allowed effectively small path lengths to be achieved by
the electricity. This can be seen by the small differences in resistance measured (270-
370 ohms).
144
SO
~5
M- 40
E
E
35
~
••
;! 30
• •
.....
DII
1:1
....
<II
25
• • • • •
..
'iii 20
<II •
e. 15
e
Q
U 10
200 220 240 260 280 300 320 340 360 380 400
Resistance (ohms)
Figure 5.26 - Compressive strength versus resistance at 1 day
60 ~------------------------------------------------------------~
50
y=32.714Ln(x)-17I.75
• R2 = 0.8721
10
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
lOO 200 300 400 500 600 700 8UO 900
Resistance (ohms)
Figure 5.27 - Compressive strength versus resistance at demoulding for 3 day air
cured samples
145
Figure 5.27 shows the relationship between electrical resistance of specimens
immediately prior to testing (after 3 days of air curing) and the compressive strength for
a range of curing regimes, mixes and ages. The figure shows that, for concrete,
electrical resistance was a good indicator of compressive strength (compressive strength
= 32.71 x Ln (resistance) + 171.75) and could be used over a range of varying mixes,
heating regimes and ages. This means that the compressive strength of electrically
cured specimens can be estimated indirectly from its final resistance without the need to
test cubes, cores or other types of sample. The coefficient of determination (R2) for the
linear fit was found to be 0.87 with a maximum deviation from the best-fit line of
6N/mm2. If resistance is to be used as a non-destructive test (NDT) for geometrically
different electrically cured specimens, calibration curves relating compressive strength
and resistance, made of the same materials, are needed.
5.9 Effect of Temperature on Resistance

Figure 5.28 shows the effect of temperature on the resistance of concrete samples.
700 ,--------------------------------------------------------, 90
600
•• A
80
70
500
.-.
...
Ei
-=-e. • 60 ..,
til
a
"'"
1:1
400
.. -+-- Uncorrected resistance
_ Corrected resistance
501
:
...
..
1:1
!! .. ,. .. Temperature 40 ;:
•i!l 300
...
"
IX
...... ··6· 30 ,_,
.-.
o°
200
···6·· .,. .. 6 .
..... ···6 20
100
10
0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~o
() 500 1000 1500 2000 2500 3000 3500 4000 4500
Time (minutes)
Figure 5.28 - Temperature corrected and uncorrected resistance versus time (1

hour delay, 40°C/hr rise, 80°C maximum)
146
The figure shows that the effect of heating on resistance was substantial with the
corrected resistance being 150 ohms higher than the apparent resistance. This is
important if resistance is to used in predicting concrete hydration or WIe ratio. The
corrected and uncorrected resistances only converged after a period of approximately 66
hours when the temperature of the samples reduced to 20°e. It can also be seen from
this figure that the maximum rate of cooling at the centre of the surface at a depth of
20mm was approximately 4°Clhr. This is within the specifications laid down by the
Highways Agency (1988) but may not be true if other mix types and geometrys are
chosen.
300 100
90
250
80
1··········....
~200
e
Ii. ...... 70
~
ID
..1:1
Ii.
60 a
e 'Cl
ID
'-'
" .....,
Al
-
Col
1:1
CIl
.:1
150
• 50
1:1
ID
-..
III
"
Cl:: 100 • 40
30
D°
-+- Uncorrected resistance 20
50 ___ Corrected resistance
10
...... Temperature
0 0
0 50 100 150 200 250 300 350 400
Time (minutes)
Figure 5.29 - Temperature corrected and uncorrected resistance versus time (1

hour delay, 40°Clhr rise, 80°C maximum)
Figure 5.29 shows the first 400 minutes of resistance measurement in Figure 5.28.
From this it can be seen that the apparent effect of the initial lowering of resistance was
reduced to a slower rise in resistance than would be expected if hydration alone was
responsible for the increase. However, the effect was still present, showing that water
was being absorbed by the aggregates, thus lowering the rate of rise of resistance. This
147
also shows the need to correct for temperature when resistance is a governing factor in
estimating the degree of hydration or any other derived property, such as mix type or
Ws: ratio (Wilson and Gupta 1996).
5.10 Two Stage Heating

The two stage heating regimes entailed a variable rate of rise in temperature. These
were:
I. 15°Clhr for 1 hour with 40°Clhr to 80°C maximum (referred to as 2 stage low/high).
2. 40°Clhr for I hour with 20°CIhr to 80°C maximum (referred to as 2 stage high/low).
00 ,----------------------------------------------------------,
50
--
"'e
a
~40
-=
-=~
~ )0
..
'"
4i
'Wi
Ec. 20 • I Stage
Ei • 2 Stage (lowlhigh)
Q
U .2 Stage (high/low)
10
X Normally cured
0
0 10 15 20 25 )0
Time (days)
Figure 5.30 - Comparison of 1 and 2 stage heating regimes (3 hour delay)
Figure 5.30 shows a comparison of the 1 and 2 stage heating regimes used for a 3 hour
delay period. This shows increased damage caused by the high/low regime at later ages
and also shows the increased I day strength of the initially high rate of heating. The
low/high regime exhibited a greater 28 day strength than the other electrically cured
148
specimens and was almost equivalent to the normally cured mix. This illustrates the
importance of the very early stages of curing during heating and the need to be able to
monitor temperature throughout the specimen were relatively minor differences have
large, long term ramifications to compressive strength.
Figure 5.31 shows a comparison of the 1 and 2 stage heating regimes used for a 2 hour
delay period. This shows increased damage to the structure of concrete at later ages in
all electrically cured samples due to the lack of strength present after 2 hours delay to
resist the thermomechanical stresses imposed during heating. The 2 stage heating
regimes showed a slightly increased compressive strength after 1 day when compared to
the 3 hour delay period in Figure 5.30, but the I stage linear heating regime did not.
This may be because with the 2 stage heating regime, the concrete did not achieve its
maximum temperature until later than the 1 stage regime. This would increase the
maturity value of the concrete at 1 day, but the affect on maturity at later ages is
negligible when compared to the damage caused to the structure of the concrete by the
earlier delay period.
~ r----------------------------------------------------------,
50
--
...
Ei
Ei
.._
;to: 40
._,
-..
01:1
till
-
1:1
GO
30
'"
GO
~
.;;]
E 20
.1 Stage
=-Ei • 2 Stage (low/high)
e
U • 2 Stage (high/low)
10
X Normally cured
0
0 10 15 20 25 30
Time (days)
Figure 5.31 - Comparison of I and 2 stage heating regimes (2 hour delay)
149
Figure 5.32 shows the importance of the delay period for 2 stage curing (lowlhigh). The
difference of only I hour delay period significantly altered both the short and long term
strength development of concrete. With a relatively early delay period (2 hours), I day
strength was increased when compared to a 3 hour delay period. This situation was then
reversed as the initial damage caused to the structure of the early delay period specimen
significantly reduced its long-term strength (Xuenquan et al 1986).
ffi r----------------------------------------------------------,
50
• 2 Hour delay (low/high)
• 3 Hour delay (lowlhigh)
x Normally cured
10
o 10 15 20 25 30
Time (days)
Figure 5.32 - Compressive strength versus time for different delay periods for two
stage heating (15°C/hr rise for 1 hour + 40°C/hr rise, 80°C maximum)
Figure 5.33 shows the importance of the delay period for 2 stage curing (high/low).
From this it can be seen that the high/low regime for a delay period of 3 hours
significantly reduced long-term compressive strength when compared to the lowlhigh
regime shown in Figure 5.32. This was likely to be due to the higher temperatures being
reached at an earlier stage with the high/low regime, thus causing greater damage at later
ages. The 1 day compressive strengths for equivalent delay periods were similar due to
the similar maturity values achieved (Saul 1951).
150
~l r-------------------------------------------------------~
so
• 2 Hour delay (high/low)

• 3 Hour delay (high/low)
X Normally cured
10
o 10 15 20 25 30
Time (days)
Figure 5.33 - Compressive strength versus time for different delay periods for 2
stage heating (40°Clhr rise for 1 hour + 20°Clhr rise, 80°C maximum)
Figure 5.34 shows the percentage compressive strength gain of 1 and 2 stage electrically
cured samples (3 hour delay) over normally cured samples. From this it can be seen that
1 stage and 2 stage (high/low) achieved approximately 200% of the compressive
strength of the normally cured samples at I day. The 2 stage (lowlhigh) achieved
approximately 150% of the normally cured compressive strength at the same age. This
was likely to be due to the time at which the samples reach maximum temperature.
With both the I stage and the 2 stage (high/low) temperatures in excess of 60°C were
reached within I hour of heating (4 hours after gauging with water), but this temperature
was not reached until 1.75 hours after heating for the 2 stage (lowlhigh) heating rate.
This is critical to the effect on the acceleration of hydration where later higher
temperatures have a reduced effect on the initial strength (Orchard and Barnett 1971).
The 2 stage (high/low) heating rate attained a slightly higher percentage gain at I day
than the I stage heating rise due to the 2 stage samples being at higher temperatures for
longer, thus having increased maturity. The normally cured samples attained a greater
degree of compressive strength than the electrically cured samples between 7 and 10
151
days. The 28 day compressive strength of the electrically cured samples were all lower
than the normally cured samples. This was due to thermomechanical damage caused to
the structure of the concrete. This was confirmed by the greater reduction of the 2 stage
(high/low) heating rise. This was because this 2 stage heating rate maintained high
temperatures (60-80°C) for a comparatively longer period than the other heating rates,
thus causing more damage. This was also confirmed by the 2 stage (lowlhigh) heating
rate that allowed a greater degree of structural formation to occur before attaining high
temperatures (60-80°C), thus allowing less damage to occur and allowing a
comparatively high 28 day compressive strength.
250r---------------------------------------------------------
.1 Stage
.2 Stage (lowlhigh)
• 2 Stage (high/low)
----------------------------------------------------------_-_-------------_-----
o 10 15 20 25 30
Time (days)
Figure 5.34 - Percentage compressive strength gain versus time for 1 and 2 stage
heating regimes (3 hour delay)
Figure 5.35 shows the corrected resistance for 2 stage heating regimes, with a delay
period of 3 hours. From this it can be seen that the lowlhigh rise had reduced resistance
values when compared to the high/low rise. This reflected the degree of accelerated
hydration present within the samples (Wilson and Gupta 1996). The high/low rise
attained a higher temperature earlier and as such the degree of hydration was increased,
152
thus reducing the free water content and increasing resistance. The lowlhigh rise
increased resistance substantially after 60°C had been reached and the rate of resistance
rise was greater than the equivalent temperature rise in the high/low sample. This was
also true of the low rise section of the high/low heating regime. After approximately
60°C the rise in resistance of the high/low regime was significantly greater than the
resistance increase in the initial low section of the lowlhigh regime. This was likely to
be due to a combination of the following two factors:
1. At temperatures above 60°C, water loss through evaporation may significantly affect
resistance values.
2. The increase in the rate of hydration may be temperature specific, with greater
increases over certain temperature regions.
450
400
350
-.
III
e
-=e
--
II
101
I;
300
...~ ~5U
.~
III
.. 200 ___ 2 stage (low/high)

...
'=
II
t.. 15U
.........2 stage (high/low)
Q
U
lOO
50
0
0 200 400 600 gOO 1000 1200 1400 1600
Time (minutes)
Figure 5.35 - Corrected resistance versus time for 2 stage heating regimes (3 hour
delay)
The resistance values after heating reflected the relative compressive strengths of the
samples after I day with the lowlhigh regime having a lesser resistance, thus indicating
153
a lower degree of hydration and a lower compressive strength. At demoulding, the rate
of resistance increase in the lowlhigh regime was greater than the high/low which would
indicate a greater degree of hydration at later ages (Wilson and Gupta 1996). This was
in agreement with the later 28 day compressive strength values which showed the
lowlhigh regime having greater compressive strength.
5.11 Splitting Tensile Strength

Figure 5.36 shows the splitting tensile strength of beams over time. From this it can be
seen that tensile strength of electrically cured samples followed a similar relationship to
that of compressive strength as shown in Figure 5.21 (Neville 1988) with an initial
increase over normally cured samples at 1 day and a reduction of tensile strength at 28
days.
6 ,----------------------------------------------------------.
• Electrically cured (inner)
x • Electrically cured (outer)
X Normally cured
o 10 15 20 25 )0
Time (days)
Figure 5.36 - Splitting tensile strength versus time electrically (3 hour delay,
It can also be seen that the inner (centre) section attained a higher tensile strength when
compared to the outer (1/4) section. This was due to the higher temperatures achieved in
154
the centre section, thus increasing the initial rate of hydration and maturity. The outer
sections eventually attained a higher tensile strength due to the increased
thermomechanical damage caused to the inner section by the relatively higher
temperatures used.
Figure 5.37 shows the splitting tensile strength against position on an electrically cured
beam for various ages. From this it can be clearly seen that the position that tensile
strength is tested is crucial to both short and long-term properties. This shows that the
inner section attained the greatest initial tensile strength, but the lowest long-term tensile
strength. The situation was reversed as the distance from the centre increased and the
figure shows an extrapolated maximum tensile strength of approximately 6.4N/mrn2 at
28 days for the extreme edges of the beam. This shows the importance of the testing
position of accelerated cured beams where temperature histories are certainly different
for each section (Figure 5.57).
Centre of beam
.Age 1 day
.Age 7 days
A Age 28 days
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-200 -150 -100 -50 o 50 100 150 200
Distance along beam (mm)
Figure 5.37 - Splitting tensile strength versus distance along beam at differing ages
155
Figure 5.38 shows the percentage gain in splitting tensile strength of electrically cured
beams (3 hour delay, 40°CIhr rise, 80°C maximum) for inner and outer sections. It can
again be seen that the inner sections attained a significantly greater gain in tensile
strength over normally cured concrete at earlier ages than the outer sections. Both outer
and inner electrically cured samples attained significantly higher tensile strengths at I
day than the normally cured samples (60-80%), thus indicating that the rate of hydration
had been increased. The 28 day tensile strength can be seen to be 14.5% less than the
normally cured samples. This was at the worst case of the centre of the beam were
temperature induced stresses were highest, thus reducing tensile strength to the greatest
degree.
100 r---------------------------------------------------------~
80
'""'
~ 60
--.;
•
1:1 • Electrically cured (inner)
~ • Electrically cured (outer)
..~"
Cl!
1:1
40
"..
1.1
"
Q., 20
u 10 15 20 25 30
Time (days)
Figure 5.38 - Percentage increase of splitting tensile strength at differing ages
Figure 5.39 shows the percentage gain oftensile strength of electrically cured beams and
compressive strength of equivalently cured cubes (3 hour delay, 40°Clhr rise, 80°C
maximum). The percentage gains for tensile strength at I, 7 and 28 days were
substantially lower than the compressive strength gains for an equivalent mix and
156
heating regime (195%, 12% and -7.3% gain). This was due to the amount of
microcracks present within the mix. During compressive testing, these cracks were
forced to close, thus not affecting the compressive strength greatly, but during tensile
testing, these microcracks acted as stress concentrators and allowed larger crack
propagation to occur, thus decreasing the tensile strength to a greater degree. This
decrease in tensile strength from what would normally be expected means that extra
consideration needs to be given to the design of concrete in tension when using
electrically accelerated curing. Substantial pretesting to derive the tensile properties of
the concrete along the length of the beam would certainly need to be carried out prior to
any assumptions being made about tensile strength (Neville 1988).
250 r---------------------------------------------------------~
200
~ 150
•,_. • Tensile gain (inner)
1:1
'iI
II!) • Compressive gain
~IOO
!
1:1
~
~ 50
-50 L._...__...__...__~....___.'--''--''--'---'~~~~__._~~~~__'_~~~~__._~~~~__'
o 10 15 20 25 30
Time (days)
Figure 5.39 - Comparison of percentage gain of compressive and splitting tensile

strength (3 hour delay, 40°Clhr rise, 80°C maximum)
5.12 Flexural Tensile Strength

Figure 5.40 shows the effect of electrical curing on 28 day flexural tensile strength.
This shows an equivalent reduction in flexural tensile strength (15%) as with splitting
157
tensile strength when compared to normally cured samples. This was again due to
microcracks acting as stress concentrators as tensile forces occur during loading.
Figure 5.40 - Flexural tensile strength of electrically (3 hour delay, 40°C/hr rise,
80°C maximum) and normally cured samples at 28 days
5.13 Maturity
Figure 5.41 shows compressive strength of normally cured samples versus maturity.
From this it can be seen that compressive strength for the C30 mix followed a
logarithmic curve of equation:
compressive strength = 11.244 x Ln (maturity) - 63.038 Eqn.S.1
The coefficient of determination (R2 = 0.9241) was only slightly less than the coefficient
of determination for the logarithmic curve used to describe Figure 5.1 (R2=0.9529) and
may be ascribed to minor temperature fluctuations within the curing tank (20°C± 1°C)
that is allowed in BS 1881. This shows that for the mix used, log maturity was a good
indicator of compressive strength (ASTM CI074). This was to be expected, as
158
compressive strength at constant temperatures and moisture conditions generally follows
a logarithmic curve (Neville 1988).
70
60 II •
....""" I
a I Y = 11.244Ln(x) - 63.038 I I
Ii
a 50 R2 = 0.9241
~ I I
.,I;
C'1:1JJ
...
41
I.
40
I! •
I
'"41
;..
.;;;
[I
30
I
I
I
I Ii
I
I. I I I
=-e 20
I:>
o
10
0
100 1000 10000 100000
Maturity (OChr)
Figure 5.41 - Compressive strength versus maturity (normally cured)
Figure 5.42 shows compressive strength of electrically cured samples of various heating
regimes versus maturity. From this it can be seen that compressive strength for the C30
mix followed a logarithmic curve of equation:
compressive strength = 6.244 x Ln (maturity) - 20.35 Eqn.5.2
The coefficient of determination (R2) was equal to 0.8459. The gradient of the best-fit
line can be seen to be significantly less than the equivalent normally cured samples
indicating that in general, the rate of compressive strength gain in electrically cured
samples was lower than normally cured samples. Also, the intercept for the electrically
cured samples is shown to be at a greater level indicating that initial strength was higher.
The coefficient of determination was reduced when compared to normally cured
159
maturity because classical maturity calculations do not take account of extreme
temperatures or the time at which they occur (Saul 1951). However, it can be seen that
classical maturity values can at least give an indication of compressive strength
providing previous data has been acquired concerning the particular mix.
70
60
--aa
....
<, so
~ y = 6.2447Ln(x) - 20.35
...
.Cl R2 = 0.8459
~40
...~
10
'"~ 30 I
.;;:;> , I
'"~10 I I
Q, 20
a
Cl
•
u
10
0
100 1000 10000 100000
Maturity (OChr)
Figure 5.42 - Compressive strength versus maturity (electrically cured)
Figure 5.43 shows compressive strength of both electrically cured samples of various
heating regimes and normally cured samples versus maturity. From this it can be seen
that standard maturity values for normally cured concrete underestimated electrically
cured compressive strength at early ages and overestimated electrically cured
compressive strength at later ages. This was because classical maturity values do not
account for the greater effect of early age temperature rises on hydration (Saul 1951),
nor do they take into account damage caused to the structure of the concrete at these
early ages by thermally induced stresses.
160
60 I I
Ma
~ 50 I!I I
.Cl
tn40
I
=e
... I
'"~ 30
'r;; I I
e
Cl. 20
II ~L _ _l__J L,
a
Q
I • Normally cured
u
10
O~ __ ~ __ ~-L~LLLLL_ __ ~ __ ~~~~LLL- __ -L__ ~~~~LW
100 1000 10000 100000
Maturity (DChr)
Figure 5.43 - Compressive strength versus maturity (normally and electrically

cured)
35
I I I I
30
y = 10.162Ln(x) - 48.627
:: I
I
1
i
I
I
i,
1 I
R2 = 0.8779 I
/:1
I
I II I
i I
...
.Cl
• •
I
~ 20 V
~ V
'"~ V
~)t
I
'r;;
15
I I I
E I. I I I I I
Cl. 10
El I I
I
U
o
/v. 1
I
I I
I
II
I
I
,
I I
,
I
100 1000 10000
Maturity (OChr)
Figure 5.44 - Early age (~ 1 day) compressive strength versus log maturity
161
Figure 5.44 shows the early age compressive strength of electrically cured samples of
various heating regimes versus maturity. From this it can be seen that compressive
strength for the C30 mix followed a logarithmic curve of equation:
compressive strength = 10.162 x Ln (maturity) - 48.627 Eqn.5.3
The coefficient of determination (R2) was equal to 0.8779. From this it can be seen that
although compressive strength of electrically cured samples did rise with maturity, the
variability caused by the effects of temperature at certain times caused a large degree of
variability. As such, classical maturity values can only give an indication of the early
age compressive strength of electrically cured samples (ASTM CI 074).
400
350
• •
• ••
.-til
e
.a 250
300
• ..•..
--
0
GO
Col
1:1
200
S
.ra
til ISO
GO
i::.:::
100
SO
o 200 400 600 800 1000 1200 1400 1600 1800 2000
Maturity (OChr)
Figure 5.45 - Resistance versus maturity
Figure 5.45 shows the relation between electrical resistance and maturity values for
electrically cured samples. From this it can be seen that there appeared to be no
correlation between resistance and maturity. This would seem to be a contradiction to
162
previous results, which showed that both resistance and maturity can be used to indicate
compressive strength. The reason for the apparent lack of correlation was that all
resistance measurements shown were taken at very early ages (1 day) and as such the
effectiveness of changes in resistance predicting compressive strength due to hydration
progression was reduced.
5.14 Initial Surface Absorption Test

Figure 5.46 shows the Initial Surface Absorption Test (ISAT) (Levitt 1969) results for a
1 stage heating period with a 1 hour delay period after 28 days. This shows that the
surface absorption of electrically cured concrete was significantly lower than the
equivalent normally cured specimen. This would seem to be in contradiction with the
comparatively lower compressive strength results for the electrically cured specimens
and may have been due to the refinement of the capillary pores during heating, thus
causing a reduction in moisture flow.
0.25 ,-------------------------------,
02
--+- Normally cured

-.. ---- Electrically cured
'"
::::::_0.15
....Ii!
;::;;
Ii!
~
-
~ 01
005
o 10 20 30 40 50 60 70 80 90 100 110 120 130
Time (minutes)
Figure 5.46 - ISA of electrically (3 hour delay, 40°C/hr rise, 80°C maximum) and
normally cured samples at 28 days
163
Other aspects of durability, especially Delayed Ettringite Formation (Lawrence 1993),
have not been considered at the present time. This is an important area of concern with
the relatively high temperatures used in the presented test set-up and needs to be fully
evaluated before larger scale practical applications can be developed.
Figure 5.47 shows the relationship between compressive strength at various ages and the
10 minute surface absorption result (lSA 10) for both electrically and normally cured
samples for a C30 (10-30mm slump) mix design. From this it can be seen that the
normally cured samples followed the expected relationship of a lowering of compressive
strength as ISA 10 increased (Dhir and Byars 1991). This was due to compressive
strength being related to the degree of hydration experienced. As hydration progressed,
the capillary pore structure of normally cured samples gradually filled with hydration
products, thus reducing the absorption of water.
60
--aa
...
50
.....
~ 4(J
-=i.! •
..
=e 30
• • •
III
GO
.~
..
'"'"
GO
C.
20 • Normally cured
a
0 • Electrically cured
u
10
0
0 0.05 0.1 0.15 U.2 0.25 U.3
ISA 10 «mllm2)/s)
Figure 5.47 - Compressive strength versus ISA 10
164
The electrically cured samples, however, follow the opposite trend (dotted line). This is
because the ISA readings were all been taken at relatively high ages (>28 days).
Electrically cured samples that achieved high long-term compressive strengths generally
had relatively little damage to the structure of the cement matrix due to the thermal
stresses induced. The electrically cured samples that attained relatively low long-term
compressive strengths were generally the ones with a shorter initial delay period (1-2
hours) or high maximum temperature, thus incurring a larger amount of cement matrix
damage by the thermally induced stresses. The damage occurred because the cement
matrix was still fluid when heating occurred and as such had little inherent strength to
counteract the induced stresses. For these mixes, the stresses induced a refinement of
the capillary pore structure, thus reducing the amount of water flow through the
hardened concrete.
20
19
18
17
• • •
~ ...............•....• Ii···· • •
-'""
~
0
16
•
.f' 15
'"...
Q
Q
•
Q., 14
• Normally cured
12
II
10
0 0.05 0.1 0.15 0.2 0.25 03
ISA 10 «mllm2)/s)
Figure 5.48 - Porosity versus ISA 10
Figure 5.48 shows the relationship between porosity and ISA 10 for electrically and
normally cured samples. From this it can be seen that relatively small changes in
165
porosity occurred. However, even with this small range there appeared to be opposite
trends in the electrically (dotted line) and normally (solid line) cured relationships. The
normally cured samples appeared to have the expected increase in ISA 10 value as
porosity increased (Neville 1988). This was to be expected, as with increased voids,
surface absorption of water should increase. The electrically cured samples followed
the opposite trend with ISA 10 increased as porosity decreased. This may have been
because, with short delay periods (before the initial set), air bubbles trapped within the
mix were free to expand from the increase in temperature. This caused an increase in
porosity but also corresponded with a refinement of capillary pores, thus reducing ISA
10. If heating occurred at later delay periods (after the initial set), then the air bubbles
were constrained from expanding, thus causing expansion stresses, which in tum created
microcracks which can readily increase the network of connected pores and capillaries.
5.15 Porosity
Figure 5.49 shows the change in porosity over time for C30 mixes of both electrically
and normally cured samples.
20,--------------------------------------------------------
•
~....•...............•............................. •• .
._
.......
•
15 •
•
.f' 10
• NonnaJly cured
'"f
:. • Electrical1y cured
o 100 200 300 400 500 600 700 800 900
Time (days)
Figure 5.49 - Porosity versus time for electrically and normally cured samples
166
The electrically cured samples (dotted line) were of various heating regimes. Both
showed little difference in porosity up to periods of 800 days and remained around 16-
18% porosity. This shows that porosity, as measured by the experimental set-up used
(DeVenny and Khalaf 1998), was not adversely affected by the use of electrical curing.
Figure 5.50 shows the relationship between porosity and compressive strength for a C30
mix for electrically and normally cured mixes. The electrically cured samples were of
various heating regimes. From this, it can be seen that porosity did not appear to change
significantly with changes in compressive strength. Both electrically (dotted line) and
normally (solid line) cured mixes had a porosity of 14-17% over the range of
compressive strengths tested. This was likely to be due to the age of the samples (>28
days) by which time porosity changes little.
20,--------------------------------------------------------,
- ..
.......................................................... \ .
15
•
• Normally cured
o 10 20 30 40 50 60 70
Compressive strength (N/mml)
Figure 5.50 - Porosity versus compressive strength for electrically and normally
cured samples
167
5.16 Density
Figure 5.51 shows the changes in density over time for electrically and normally cured
concrete samples. The electrically cured samples were of various heating regimes.
From this it can be seen that both normally (solid line) and electrically (dotted line)
cured samples had virtually identical densities (2300-2400kg/m3) over periods up to 800
days. This was clearly due to the fact that identical mixes (C30) were used in both
circumstances and although electrical curing does inevitably lead to initial water losses,
in the long-term this did not affect density.
3~ ,---------------------------------------------------------,
2500
. :'":
.a
............•..........•....................................•. ~ _..,
• .
• Normally cured
I~
500
lOO 200 300 400 500 600 700 800 900
Time (days)
Figure 5.51 - Density versus time for electrically and normally cured samples
5.17 Water Absorption

Figure 5.52 shows the water absorption of electrically (dotted line) and normally (solid
line) cured samples at various ages. The electrically cured samples were of various
heating regimes and all mixes are designed as C30 (slump lO-30mm). From this it can
be seen that water absorption remained fairly constant over the long-term, though a
168
slight reduction was apparent. This was to be expected, as with the time scale used (up
to 800 days), relatively little reduction in pore volume occured.
10 ,----------------------------------------------------------,
~ ~ .111 . •• ••
• ••
Cl
Cl 6
• •
..
;:
C.
Cl
III
.c
"..
Gi
4
• Normally cured
~
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 100 200 300 400 500 600 700 800 900
Time (days)
Figure 5.52 - Water absorption versus time for normally and electrically cured
samples
5.18 Temperature Profiles

A full list of temperatures for a 150x 150x150mm cube heated to 80CC at a rate of
40cCIhr with a delay period of 3 hours over various ages can be seen in Table 5.1.
Figure 5.53 shows the colour temperature scale used for defining the temperatures in
Figures 5.57, 5.54, 7.16 and 7.27. Each colour represents a range of 1"C.
Figure 5.54 shows a graphical representation of Table 5.1. From this it can be seen that
prior to heating, the temperature distribution was approximately uniform (± 1"C) across
the sample at each time period selected. At 60 minutes, it can be seen that the exterior
regions were at a higher temperature than the internal regions. This was due to the
ambient temperature (20cC) being higher than the temperature of the concrete. This
169
allowed heat conduction through the electrode plates into the concrete but did not occur
at the base due to the cubes sitting on 50mm thick expanded polystyrene which acted as
a thermal barrier to the lower ambient temperatures. The overall temperature did rise
from approximately 12°C immediately after mixing to 17°C prior to heating. This was
due to exothermic hydration reactions occurring as the concrete cures normally (Neville
1988).
Table 5.1 - Temperature of lSOmm cube at various ages
Time after Sensor number from Chapter 3

gauging with 1 2 3 4 5 6 7 8 9 10
water (min)
0 12.4 12.4 12.4 12.4 12.5 12.5 12.5 12.5 12.5 12.4
60 14.1 13.8 13.7 13.7 13.7 14.3 14.0 14.0 13.9 13.9
160 17.1 17.1 17.0 17.0 17.0 17.2 17.2 17.3 17.2 17.2
180 17.5 17.5 17.5 17.5 17.4 17.6 17.7 17.7 17.7 17.6
195 27.5 27.4 27.1 26.1 24.7 27.2 27.6 27.1 26.3 24.7
210 37.5 37.6 37.1 35.4 33.1 36.5 37.3 36.6 35.2 32.6
225 47.5 48.6 47.9 45.7 42.6 46.0 47.7 46.9 45.1 41.6
240 57.5 60.1 59.6 56.9 53.2 55.5 58.7 58.0 56.0 51.7
255 67.5 71.9 71.6 68.8 64.5 64.9 69.6 69.1 67.2 62.5
270 77.5 84.3 84.4 81.8 77.2 75.0 82.0 81.8 80.1 75.0
274 80.0 88.0 88.0 85.5 81.0 77.3 84.6 84.9 83.3 78.2
285 75.6 84.4 85.7 84.7 81.6 73.2 81.1 82.5 82.1 78.7
300 72.0 80.7 82.7 83.3 81.9 70.6 78.0 80.4 81.1 79.6
510 52.0 54.1 54.1 53.4 52.2 51.4 53.1 54.2 52.6 51.6
1550 26.0 26.2 26.2 26.3 26.3 26.0 26.1 26.2 26.3 26.3
1710 25.7 25.8 25.8 25.9 25.9 25.4 25.8 25.8 25.9 25.9
Sensor number
II 12 13 14 15 16 17 18 19 20
0 12.4 12.5 12.5 12.5 12.6 12.6 12.7 12.7 12.8 12.7
60 15.0 14.6 14.4 14.4 14.4 15.3 15.1 15.1 15.0 15.0
160 17.6 17.5 17.5 17.5 17.4 17.8 17.9 17.9 17.9 17.8
180 18.1 17.9 17.9 17.9 17.9 18.1 18.2 18.3 18.3 18.1
195 27.2 27.2 26.8 25.6 24.3 26.6 26.7 27.1 27.4 25.1
210 34.9 36.0 35.3 33.6 31.4 33.7 34.4 34.9 35.0 31.8
225 43.6 45.6 44.9 42.9 39.9 41.5 43.4 43.7 43.5 39.8
240 52.7 56.0 55.3 52.8 49.3 50.0 53.1 53.8 53.6 48.6
255 61.8 66.4 66.0 63.2 59.3 58.9 63.3 64.4 64.2 5S.1
270 71.2 77.6 77.6 75.0 70.8 67.9 74.0 75.2 75.8 69.0
274 73.S SO.7 80.9 7S.1 73.8 70.6 77.1 79.1 79.3 72.3
285 70.3 77.7 79.3 78.2 75.3 67.7 75.5 77.4 77.2 73.0
300 6S.4 75.5 77.9 77.9 76.4 66.3 74.1 76.5 76.9 74.1
510 51.4 52.2 52.2 51.6 50.7 50.1 51.5 51.9 51.9 51.2
1550 25.9 26 26.1 26.1 26.1 25.8 26.0 26.1 26.1 26.1
1710 25.5 25.7 25.7 25.7 25.7 25.5 25.6 25.7 25.6 25.7
170
During heating, it can be seen that the maximum temperature was always in the central
region and 'sank' as the temperature increased. This may have been due to the
increased insulation on the bottom surface. As the temperature increased, a greater
amount of heat was lost through the exposed upper surface, thus requiring higher
currents to attain the same rate of heating which increased the temperature gradients and
thus increased the temperature of the lower, better insulated regions.
After heating, the cube gradually cooled, with a greater rate of cooling and higher
temperature gradients at the surface which was due to the thermal insulating properties
of the plastic mould. After 1 day the temperature had again reached an approximate
equilibrium condition across the section at a temperature of 25°C.
.0-1 .1-2 02-3 03-4 .4-5 .5-6 .6-7 .7-8 .8-9 .9-10
010-11 .11-12 .12-13 .13-14 .14-15 .15-16 .16-17 017-18 D 18-19 019-20
.20-21 .21-22 .22-23 23-24 .24-25 .25-26 .26-27 027-28 .28-29 .29-30
.30-31 .31-32 .32-33 .33-34 .34-35 .35-36 .36-37 .37-38 .38-39 .39-40
.40-41 041-42 .42-43 .43-44 .44-45 045-46 .46-47 .47-48 .48-49 .49-50
.50-51 .51-52 .52-53 .53-54 54-55 .55-56 056-57 .57-58 058-59 059-60
.60-61 061-62 62-63 063-64 .64-65 65-66 066-67 067-68 .68-69 .69-70
070-71 .71-72 072-73 073-74 074-75 075-76 076-77 077-78 078-79 079-80
80-81 081-82 082-83 083-84 084-85 085-86 086-87 Cl87-88 .88-89 089-90
.90-91 .91-92 .92-93 .93-94 94-95 .95-96 .96-97 .97-98 098-99 .99-100
Figure 5.53 -Temperature (OC) colour key to Figures 5.54, 5.57, 7.16 and 7.27
171
Omin 60min
20
u-c n-e
~50
'-'
;: 75
c.
~
Q
100
14°e
130
160 min 180 min

l) ~
~
/'
....._ .._
195 min 210 min
225 min 240 min
Width (mm)
Figure 5.54 - Temperature cross section centrally perpendicular to electrodes
172
255 min 270 min
84°C
274 min
300 min 510 min
1550 min 1710 min
Figure 5.54 - Temperature cross section centrally perpendicular to electrodes

(continued)
173
Figure 5.55 shows the temperature profile along the beam (400xIOOxIOOm) at various
ages during heating (3 hour delay, 40°CIhr rise to a maximum of 80°C) at a depth of
50mm. From this it can be seen that the central region, at the end of heating, attained a
higher temperature (89.6°C) than the outside regions (81.3°C) and the point where the
temperature datum was calculated from (80°C at a depth of20mm). This was due to the
relatively small temperature gradients that were present laterally (Figure 5.57) and the
relatively high thermal insulating properties of concrete which resulted in the highest
temperature being within the internal central region (Bazant and Kaplan 1996).
Figure 5.55 - Temperature distribution along centre of beam at a depth of 50mm

(3 hour delay, 40°Clhr rise, 80°C maximum)
Figure 5.56 shows the temperature profile along the beam (400x 1OOxIOOmm)at various
ages during heating at a depth of 20mm. From this it can be seen that the maximum
temperature at the surface was always lower than the temperature within the concrete
(Figure 5.55). This was due to the vertical temperature gradients inherent within
electrically heated concrete that show the transfer of heat from the 'hot' core to the
relatively cool ambient atmosphere. It can also be seen that immediately prior to heating
174
(180 minutes). the outside regions were at a slightly higher temperature than the inner
regions. This was again due to the influence of the ambient temperature. which at this
time was higher than the concrete temperature and as such heated up the concrete from
the outside inwards.
lOO
-+-180mm
90 ....... 195min
...... 210mm
80 *""225mm
....... 240mm
70 -+-255mln
G
0 -+-270min
'-' 60
.=:.
e -276mm
50
41
=.
e
41
Eo-
40
.. • ~ • • ..
30
• • • .. • • •
20
• • • t • • •
10
0
·300 ·200 ·100 0 100 200 300
Distance along beam (mm)
Figure 5.56 - Temperature distribution along centre of beam at a depth of 20mm

(3 hour delay, 40°C/hr rise, 80°C maximum)
A full list of temperatures for a beam (400x 1OOx1OOmm)with a 3 hour delay period
heated to 80°C at a 40°Clhr rise over various ages can be seen in Table 5.2. From this it
can be seen that the temperature gradients present increased during heating, especially
near the upper surface, but decreased after heating ended until after 1 day where a
uniform temperature of 25°C (± 1°C) was attained.
175
Table 5.2 - Temperature of beam (400xlOOxlOOmm) at various ages
Time after Sensor number ~from Chapter 3)

gauging (min) 1 2 3 4 5 6 7 8 9 10
180 16.3 16.5 16.6 16.8 16.9 16.3 16.4 16.6 16.8 17.1
195 26.3 26.25 25.8 25.5 25.1 25.6 25.7 25.4 25.0 24.4
210 36.3 36.0 35.0 34.1 33.2 34.8 34.9 34.1 33.1 31.7
225 46.3 46.5 45.8 45.0 44.0 44.2 45.0 44.7 43.6 42.1
240 56.3 57.5 57.7 57.0 56.1 53.8 55.9 56.2 55.3 53.8
255 66.3 69.1 70.5 70.4 69.5 62.9 66.7 68.0 67.8 66.6
270 76.3 79.8 83.4 84.0 83.9 71.9 78.0 80.8 81.8 81.2
276 80.0 84.6 89.6 91.2 90.1 75.l 82.4 86.1 88.0 88.1
285 76.9 82.8 88.1 90.2 90.4 73.0 80.5 84.9 87.7 88.6
300 72.6 77.4 82.8 85.6 86.6 69.3 75.7 80.3 83.8 85.8
345 66.9 69.4 72.4 74.1 74.5 65.1 68.3 70.8 72.9 74.3
1375 25.5 25.6 25.6 25.6 25.6 25.5 25.6 25.6 25.7 25.7
1515 25.2 25.3 25.3 25.3 25.3 25.2 25.3 25.3 25.3 25.3
Sensor number
11 12 13 14 15 16 17 18 19 20
180 16.6 16.6 16.8 16.9 17.1 16.9 17.1 17.2 17.3 17.4
195 25.8 25.8 25.6 25.0 24.7 26.1 26.1 25.8 25.4 24.8
210 35.l 35.0 34.3 33.1 32.3 35.3 35.0 34.4 33.5 32.2
225 44.2 44.9 44.5 43.5 42.5 44.l 44.4 44.0 43.1 41.8
240 53.4 55.4 55.8 55.0 54.0 52.9 54.2 54.3 53.6 52.0
255 62.3 65.8 67.2 67.2 66.5 61.7 64.2 65.3 64.9 63.3
270 70.8 76.3 79.3 80.6 80.4 69.7 73.8 76.1 76.6 75.1
276 74.0 80.6 84.5 86.7 86.9 72.8 78.0 81.3 82.0 81.2
285 71.9 78.6 83.0 86.0 86.6 69.7 75.0 78.5 80.l 79.5
300 68.0 73.6 7S.1 S1.9 83.4 65.7 70.2 73.8 75.9 76.0
345 63.2 65.9 6S.3 70.7 71.6 60.1 62.1 64.1 65.2 65.1
1375 25.4 25.4 25.5 25.5 25.5 25.3 25.3 25.4 25.4 25.4
1515 25.1 25.2 25.2 25.2 25.2 25.1 25.1 25.1 25.1 25.1
Figure 5.57 shows a cross section temperature map of the beam (400xIOOxIOOmm) at
the end of heating (276 minutes). From this it can be seen that the central point of
maximum temperature has 'sunk' in a similar manner to Figure 5.54 to a point 20mm
above the bottom surface. It can also be seen that the maximum temperature gradients
occurred vertically near to the upper surface with comparatively small temperature
gradients appearing laterally. This was due to the current flowing uniformly laterally
throughout the beam and the relatively low ambient temperature present.
176
20
35 ~
'S.
91°C c:r
65 a,!
-133 -66 o 66 133 200
Length (mm)
Figure 5.57 - Cross section temperature map of the beam at the end of heating (276
minutes) (3 hour delay, 40°Clhr rise, 80°C maximum)
5.19 Scanning Electron Microscopy

Three curing regimes were chosen for analysis by the use of a Scanning Electron
Microscope (SEM). These were:
1. Normally cured (sample A)

2. DEC: 3 hour delay, 40°C/hr rise, maximum 60°C (sample B)
3. DEC: 2 hour delay, 40°C/hr rise, maximum 80°C (sample C)
After 1 day all samples were transferred to a water tank in accordance with the
procedures stated in Chapter 3.
5.19.1 Microstructure
Figures 5.58, 5.59 and 5.60 show scanning electron microscope images of concrete
samples cured according to curing regime A, Band C respectively, 1 day after mixing.
All samples showed a dense microstructure associated with the development of C-S-H
gel. At 1 day normally cured concrete showed a more uniform microstructure as seen in
Figure 5.58 than the concrete heated to 80°C maximum (Figure 5.60) which exhibited a
much coarser microstructure. Sample C also showed a much higher level of hydration
than the other two samples with ettringite crystals clearly visible in Figure 5.60 as
filamentous structures growing out from the cement grains into the water filled pores.
177
This was the result of the high maximum temperatures applied, which accelerated the
initial hydration process.
Sample A showed a higher degree of unhydrated cement particles at 1 day than samples
B and C which implied that elevated temperature speeds up the hydration process of
OPC cement.
At higher magnifications, shown in Figures 5.61 and 5.62 respectively, it can be seen
that the ettringite crystals have become larger in sample C when compared to sample A.
The larger needle-like crystals (ettringite) form at this stage of hydration as a result of
the reaction of C3A and gypsum in solution. It can also be seen that the microstructure
in sample C is coarser and more obviously crystalline in form.
At 28 days there is little difference in the microstructure of all of the concrete samples
scanned as shown in Figures 5.63, 5.64 and 5.65, but the heat cured samples appeared
less uniform. The C-S-H gel had grown during the continued hydration and had became
more compact in all samples. There was also a relative scarcity of residual unhydrated
cement grains in all samples showing that a high proportion of the possible hydration
had taken place by 28 days in both electrically and normally cured samples.
5.19.2 Pore structure

It can be observed from Figures 5.58, 5.60 and 5.66 respectively that all samples had a
visible pore structure. The larger pores of the concrete samples can be seen to be of the
hollow shell type, but the samples also contained capillary pores and spaces.
It appears from Figures 5.58, 5.60 and 5.66 that sample C had a coarser pore structure
than sample B, which in turn had a coarser pore structure than sample A. Heat curing
has generally been reported to result in a coarsening of the pore structure (Kjellsen
1995).
178
At 28 days, the pores in the normally cured concrete appeared to be less numerous
(Figure 5.63) but the pores in the heated concrete (Figures 5.64 and 5.67) appeared to be
more apparent but not as extensive as at 1 day. The high decrease in porosity of sample
A was because during hydration calcium ions may have diffused into the previously
formed hollow shells and precipitated as calcium hydroxide (portlandite). This was
likely not to have happened in the heat cured concrete due to the greater maturity at the
time of exposure to water curing which may have caused less production of portlandite
to fill the hollow shells.
5.19.3 Microcracking
Figures 5.63, 5.67, 5.68 and 5.69 respectively show images of the concrete samples in
the aggregate/concrete matrix region. All of the concrete samples exhibited
microcracking in the bulk cement paste and especially at the aggregate/cement matrix
zone. However, the extent of the cracks at the aggregate interface was found to be more
prevalent in the mixes that had been electrically cured. No apparent differences in crack
number and width seemed apparent between the electrically cured mixes. This may
have been due to the small sample sizes available. This kind of microcracking is likely
to be caused as a result of the differential expansion of the aggregate during heat curing
and the subsequent cooling period.
179
Figure 5.58 - 1380X magnification of normally cured concrete (1 day)
Figure 5.59 -1190X magnification of electrically cured (40°Clhr rise, 60°C

maximum, 3 hour delay) concrete (1 day)
180
Figure 5.60 -1060X magnification of electrically cured (40°C/hr rise, 80°C
181
Figure 5.62 - 3980X magnification of electrically cured (40°Clhr rise, 80°C
Figure 5.63 _ 616X magnification of normally cured concrete (28 days)
182
maximum,3 hour delay) concrete (28 days)

183
Figure 5.66 - 2440X magnification of electrically cured (40°C/hr rise, 60°C

maximum,2 hour delay) concrete (28 dayS)
184
Figure 5.69 -1030X magnification of electrically cured (40°Clhr rise, 60°C

185
5.20 Conclusions
1. The effect of removing the 50mm thick polystyrene cover was found to cause a
reduction in initial strength of DEC concrete when compared to normally cured
concrete. This was likely to be due to excessive moisture loss from the concrete and
higher internal temperature gradients necessary to maintain the same heating regime
as for DEC concrete cured under insulation cover.
2. The resistance of uncovered samples was found to attain a higher value than covered
samples up to 1800 minutes after mixing, but after that time covered samples were
found to have a higher resistance. This was likely to be due to the loss of moisture
from the uncovered samples which initially increased resistance but reduced further
hydration due to desiccation.
3. At 1 day, all samples of varying W/C ratios cured using DEC attained a higher
compressive strength than normally cured samples. The increase became slightly
higher with increases in WIC ratios but this effect was not apparent for compressive
strength at 3 days.
4. The effect of W/C ratio on resistance values was seen to be apparent, with higher
WIC ratios producing lower resistance values.
5. The compressive strength of normally cured concrete at 24 hours was attained using
DEC in 4 hours and compressive strengths of up to 26N/mm2 were readily
achievable using DEC after 1 day.
6. Short delay periods (1 hour) tended to give higher initial compressive strengths, but
reduced long-term compressive strengths when compared to longer delay periods (4
hours). The reduction in long-term compressive strengths was likely to be due to
increased microcracking in the short delay period samples which reduce long-term
strength. The greatest long-term compressive strength (~28 days) was found to
186
occur when heating commenced at 3 hours after mixing. This was due to high early
age compressive strengths combining with limited microstructural damage to
produce relatively high long-term compressive strengths.
7. DEC concrete attained a large proportion (up to 71%) of its 28 day compressive
strength within 1 day.
8. The temperature corrected resistance (Equation 2.2) at the end of heating was found
to be dependent on the delay period chosen prior to heating.
9. For equivalent mixes and heating regimes, a higher maximum temperature produced
higher I day compressive strengths but lower long-term (~28 day) compressive
strengths. This was likely to be due to increased thermomechanical damage and
moisture loss at the higher temperatures.
10. DEC was found to be applicable to different mix designs with initial (~l day)
increases but long-term (~28 days) reductions in compressive strength, when

compared to equivalent normally cured concrete. This was found to be the case for
both high slump and high strength mixes.
II. The resistance at 1 day was not found to be proportional to compressive strength but
at 3 days a logarithmic relationship was found to exist.
12. Concrete was found to behave as a dielectric with a lower apparent resistance at
higher temperatures. This implies that concrete temperature should be closely
monitored if resistance is to be used in predicting the progress of hydration.
13. Significantly different short and long-term compressive strengths were found to be
achievable using a 2 stage heating rise with a high/low heating regime producing
higher short-term (1 day) but lower long-term (28 days) compressive strengths.
187
14. It was found that the effect of the delay period prior to heating of the 2 stage samples
effected the short and long-term compressive strength in a similar manner to 1 stage
heating regimes.
15. The splitting tensile strength of concrete cured using DEC was found to follow a
similar trend as compressive strength of DEC cubes with higher initial temperatures
causing increased initial tensile strength but lower long-term tensile strength.
16. It was found that, when compared to normally cured samples, the DEC splitting
tensile strength had a lower percentage tensile strength increase at 1 day and a higher
percentage tensile strength loss at 28 days than the equivalently cured DEC
compressive strength tests. This was likely to be due to microcracks acting as stress
concentrators during tensile loading.
17. Flexural tensile strength was found to act in a similar manner to splitting tensile
strength.
18. Saul's maturity values were found to be more accurate in determining the
compressive strength of normally cured samples than electrically cured samples.
Maturity values were also found to underestimate the compressive strength of
electrically cured samples at early ages and overestimate compressive strength at
later ages.
19. There was found to be no direct relationship between resistance and maturity for the
samples tested.
20. Durability of concrete (as measured by ISAT) did not appear to be adversely effected
when using DEC. This was likely to be due to the refinement of capillary pores
during heating.
188
21. It was found that ISA 10 values were related to compressive strength. For normally
cured samples, ISA 10 values increased as compressive strength decreased, but for
electrically cured samples, the opposite was true. This was due to low compressive
strengths in electrically cured concrete being related to microstructural damage
which also caused capillary pores to alter, thus reducing water absorption.
22. Porosity, density and water absorption was found to be similar for both electrically
and normally cured samples.
23. Natural temperature rises within the concrete were apparent prior to heating and
were due to exothermic hydration reactions.
24. Temperature gradients increased during heating and also the maximum temperature
within the electrically heated (150x 150x150mm) cubes and (400x IOOxIOOmm)
beams were in the central region with minimum temperatures immediately adjacent
to the exposed concrete surface.
25. Within the (400xIOOxlOOmm) beams, larger temperature gradients were seen to
occur vertically (parallel to the electrodes) than laterally (perpendicular to the
electrodes). This was thought to be due to the thermal insulating properties of the
concrete and mould and the relatively low ambient temperatures.
26. Using a scanning electron microscope (SEM), both normally and electrically cured
samples were found to exhibit a dense microstructure associated with C-S-H
formation at all ages, but electrically cured samples tended to have a less uniform
structure with coarser pores being more prevalent.
27. Results show that higher concentrations and lengths of ettringite were found after 1
day in electrically cured samples.
189
28. The microstructure at 28 days appeared to be similar for normally and electrically
cured samples with little unhydrated cement present.
29. All samples have a degree of aggregate/cement matrix fracturing, but the electrically
cured samples were found to have wider, more numerous aggregate/cement matrix
cracking. No differences in the degree of this cracking could be found between
different heating regimes used to electrically cure samples. The increased cracking
was thought to be caused by differential thermal expansion during initial heat curing.
190
CONTAINING ADMIXTURES AND ADDITIVES
6.1 Introduction
investigation of electrically cured concrete containing admixtures and additives. All
testing was carried out on 150x150x 150mm, C30 (l 0-30mm slump) cubes unless
otherwise specified. These results can be seen in Figures 6.1-6.18.
The mixing, casting, curing and monitoring was implemented in accordance with the
experimental procedures stated in Chapter 3.
6.2 Air Entraining Admixture
Figure 6.1 shows the relationship between density at 28 days and ultrasonic pulse
velocity (UPV) for both electrically and normally cured concretes containing different
amounts of air entrainment. The percentage of air entrainment for each sample can be
seen to be adjacent to the data points on Figure 6.1. From this, it can be clearly seen
that, as expected, the UPV tended to increase with density. The relationship (density =
253.3 x (UPV) + 1381) between the two properties was not strong (coefficient of
determination (R2) = 0.64) and so only gave an indication of the density for a particular
concrete. It also appeared that, for the small range tested, the electrically cured and
normally cured samples could be said to have separate linear changes in UPV as density
increased. This may have been due to differences in pore structure in the electrically
cured samples which locally increased density but did not affect UPV because they did
not form a continuous structure within the cement matrix (Patel et al 1994, Kjellsen
1995).
191
3000 r-----------------------------------------------------------,
3.9% •• 2.7% 2.7%
~SlX)
9.20%I.,_-------:-:~~-___;~
..
...~
92%. 3.9%
~ooo
y=253.3x+1381
R2=O.6441
1000 • Normally cured

500
35 4.5
UPV (kmls)
Figure 6.1 - Density versus ultrasonic pulse velocity at 28 days (0.5 hour delay,
40°Clhr rise, 80°C maximum for 4 hours)
35
.2.7%
.3.9%
&9.2%
15
05
o 10 15 20 25 Jil
Time (days)
Figure 6.2 - Ultrasonic pulse velocity for normally cured samples of different
percentages of air entrainment
192
Figure 6.2 shows the relationship between UPV and age for normally cured samples of
different air entrainment. From this it can be seen that UPV increased slightly over time
for the samples tested. This was due to minor increases in density as the concrete
continued hydration. It can also be seen that the 9.2% air entrainment mix had a
reduced UPV value at later ages. This was due to the relatively large amount of air
entrainment reducing density. The difference between the 2.7% and 3.9% air
entrainment could not be identified within the accuracy of the instrumentation used
(Bungey 1982).
45
35
:;====
~
:
A : :
i-- 25
--
..:.I:
;>
2
.2.7%
=-
;;J
.3.9%
I5 69.2%
05
o L_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 10 15 20 25 30
Time (days)
Figure 6.3 - Ultrasonic pulse velocity for electrically (0.5 hour delay, 40°C/hr rise,
80°C maximum for 4 hours) cured samples of different percentages of air
entrainment
Figure 6.3 shows the relationship between UPV and age for electrically cured (0.5 hour
delay, 40°Clhr rise, 80°C maximum, isothermal period for 4hrs) samples of different air
entrainment. This shows a similar trend to Figure 6.2 by which UPV increased over
time. The differences between samples appeared far greater than Figure 6.2 with the
mix containing no air entraining agent (2.7%) having a lesser UPV than the 3.9% air
193
entrainment. This may have been due to variability of the samples or the difference in
microstructure of adding bubbles affecting UPV. The mix containing 9.2% arr
entrainment had a significantly lower UPV at all ages than any of the other samples. At
28 days it appeared that UPV gave an indication of the proportion of air entrained within
the electrically cured mixes.
Figure 6.4 shows the relationship between compressive strength and UPV for
electrically and normally cured samples of differing air entrainment and ages. From this
it can be seen that a relationship exists (Bungey 1982) between the two variables
(compressive strength = 0.007 x (UPVi9782) for a combination of electrically and
normally cured mixes (coefficient of determination (R2) = 0.8478) with differing air
entrainment and ages.
60
--
...
El
50
• Normally cured
••
..
e
Z •
-=--
...
DJ)
iii
...t 30
'"
"~
';J
..=-
fi
20 •
u
e
Q
y=O.007X59782
R2=O.8478
•
10 •
0
25 3S 4S
UPV (km/s)
Figure 6.4 - Compressive strength versus ultrasonic pulse velocity for electrically
(0.5 hour delay, 40°Clhr rise, 80°C maximum for 4 hours) and normally cured
samples
194
Figure 6.S shows the relationship between compressive strength and time for normally
and electrically (O.Shr delay, 40°C/hr rise, 80°C maximum, isothermal period for 4hrs)
cured samples of 9.2% air entrainment. From this it can be seen that the expected
increase in initial (1 day) compressive strength did not occur. The electrically cured
samples did not attain the normally cured compressive strength at any age and the rate of
compressive strength gain over time was proportionally reduced. This was due to the
extreme heating regime used together with the large proportion of air entrained. With
only 0.5 hours delay prior to heating combined with a 4 hour isothermal period at 80°C,
the large percentage of air entrained caused a large amount of damage to the structurally
unformed cement matrix. This damage reduced the initial strength and also seriously
retarded the subsequent compressive strength development of the electrically cured
sample (El Hussein and Abd El Halim 1993).
35
",'-"
a 30
.§
-
~
-=
Cl)
==
~
25
20
'"~
>
.~ 15
~
Q.,
a
Q 10
U • Normally cured
- r - r~
10 100
Time (days)
Figure 6.S - Compressive strength versus time for electrically (O.Shr delay, 400C/hr
rise, 80°C maximum for 4hrs) and normally cured samples with 9.2% air
entrainment
195
Figure 6.6 represents the effect of air entrainment on 28 day compressive strength for
normally and electrically (O.5hr delay, 40°CIhr increase, 80°C maximum for 4hrs) cured
concrete. From this it can be clearly seen that the 28 day strength of the 3.9% and 9.2%
air entrained concrete was lower than the 2.7% concrete. This was because the 2.7% air
entrained concrete did not contain an air entraining admixture and the 3.9% and 9.2%
mixes were originally designed (Department of the Environment 1992) for an equivalent
TMS of 43N/mm2 with an estimated air entrained content of 2% and 5% respectively. It
can be seen that as the percentage of air entrainment increased the difference between
the normally and electrically cured concrete became greater. This could be due to
increased microcracking in the electrically cured samples as the amount of air
entrainment increased. During heating, the air within the entrained bubbles attempts to
expand causing tensile stresses in the surrounding cement shell and thus microstructural
damage (El Hussein and Abd El Halim 1993).
~ ,----------------------------------------------------------,
50
--aa
...
.....
z~ 40
-
01:1
011
1:1
-~
...
'"
GI
;;
30
..
rl
Cl.
20
a
Q
• NonnaIly cured
u
10
0
0 6 7 8 9 10
Air entrainment (%)
Figure 6.6 - Compressive strength at 28 days versus percentage air entrainment

(0.5 hour delay, 40°Clhr rise, 80°C maximum for 4 hours)
196
Figure 6.7 represents the effect of maximum temperature on electrically cured (3br
delay, 40°CIhr rise) samples with 5.0% air entrainment. From this it can be seen that
the mix electrically cured to a maximum of 60°C attained a slightly increased
compressive strength at all ages. This was likely to be due to the increased stresses
incurred at SO°C. This caused microfractures upon the entrained air cement shells
which caused structural damage. This was not apparent in mixes not containing air
entrainer (Figure 5.21) and was because the extra damage incurred by the
micro fracturing of the entrained air bubble shells outweighed the usual gains in
hydration (maturity) at early ages.
50 ,----------------------------------------------------------,
45
.-.. 40
NE
E 35
~
'-'
~ 30
~
1:1
~
.: 25
'"~
._> 20
~
~ 15
• Normally cured
=
U 10 • 60C maximwn
... 80C maximwn
o 10 15 20 25 30
Time (days)
Figure 6.7- Compressive strength for normally and electrically (3hr delay,
400C/hr rise) cured samples with 5.0% air entrainment heated to different
maximum temperatures
Figure 6.S shows the effect over time of 5.0% air entrainment for equivalent strength
mixes (Department of the Environment 1992) for normally and electrically (3hr delay,
40°Clhr rise, SO°Cmaximum) cured samples. From this it can be seen that the 5.0% air
entrained normally cured samples attained a similar compressive strength at all ages,
197
thus indicating that air entrainment was approximately 5.0%. It can also be seen that the
usual degree of compressive strength gain of electrically cured samples did not occur in
the air entrained samples. This was due to the effect of the large number of bubbles
entrained within the cement matrix. During heating, the air contained within the
bubbles attempts to expand, thus causing a far greater degree of structural damage to the
surrounding cement matrix. This caused an initial reduction in the expected
compressive strength gain and also significantly affected the long-term compressive
strength. This shows that extra care and test samples need to be considered when using
admixtures combined with DEC to ensure that properties are not assumed on the basis
of previous non-admixture mixes (Highways Agency 1998).
~ ,----------------------------------------------------------,
50
.5% air entrained (normally cured)

.5% air entrained (electrically cured)
• Normally cured
10
X Electrically cured
o L_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 10 15 20 25 30
Time (days)
Figure 6.S - Air entrained (5.0%) compressive strength versus time for electrically
(3hr delay, 400Clhr rise, SooC maximum) and normally cured samples
Figure 6.9 represents the effect of air entrainment on resistance for electrically (0.5hr
delay, 40°CIhr increase, 80°C maximum, isothermal period for 4hrs) cured concrete
during heating. From this it can be seen that an initial reduction in resistance was
apparent in all samples. This was likely to be due to water absorption by the coarse and
198
fine aggregates after mixing (Khalaf and Wilson 1999) and increases in the level of
soluble ions (Ca++, OH- and S04-) in the pore water during heating .. There appeared to
be no direct relationship between the amount of air entrainment and resistance but this
apparent effect may have been due to the altered (when compared to water alone)
electrical properties of the air entraining agent.
80
70
60
-.
'"e 50
.1:1
,_,
Q
"~
(j
1:1
40
...
'"
.;;
30
~"
20
10
0
0 50 100 150 200 250 300 350 400
Time (minutes)
Figure 6.9 - Resistance versus time for different percentages of air entrainment
(0.5 hour delay, 40°Clhr rise, 80°C maximum for 4 hours)
Figure 6.10 shows the effect of 5.0% air entrainment on resistance during (3hr delay,
40°CIhr rise, 60°C maximum) and after heating. It can be seen that the mix containing
air entrainment had a higher resistance value at all ages. This was likely to be due to the
larger proportion of voids within the mix that dry out during hydration, thus causing
localised pockets of high resistance and thus increasing overall resistance. Both air
entrained and non-air entrained samples showed the expected initial reduction in
resistance associated with the absorption of water into the aggregates within the mix
(Khalaf and Wilson 1999) and increases in the level of soluble ions in the pore water.
199
450
-lOO
350
__ 300
'"El
.1:1
~ 250
~
-"
CJ
1:1
200
.l!l
'"~
Cl:: 150
-+- 5% air entrainment

100
____0% air entrainment
50
0
0 200 400 600 800 1000 1200 1400 1600
Time (minutes)
Figure 6.10 - Resistance versus time for air entrained and non-air entrained
samples (3hr delay, 40°C/hr rise, 60°C maximum)
6.3 Superplasticising Admixture

Figure 6.11 shows the effect of replacing water with 2 litres of superplasticiser per
100kg of cement. From this it can be seen that, for normally cured concrete, the mix
containing superplasticiser attained a higher compressive strength over all ages.
For electrically cured concrete (3hr delay, 40°CIhr rise, 60°C maximum) it can be seen
that the mix containing no superplasticiser attained a slightly greater compressive
strength at 28 days than the equivalent electrically cured sample containing
superplasticiser. The opposite was true for normally cured mixes. This was due to the
setting time being increased when superplasticiser was added and thus the mix
containing the superplasticiser acted at 3 hours as a mix not containing superplasticiser
at I or 2 hours. This means that greater microstructural damage occurred during heating
resulting in long-term (28 days) compressive strength reductions.
200
60
50
N
~
El
El
;Z 4()
'-'
.1:1
...~
c:I
..."
loo
til
30
• Normally cured (superplasticiser)

~
0;;;"
til • Electrically cured (superplasticiser)
~
e,
20
& Normally cured (no superplasticiser)
El
Q
U X Electrically cured (no superplasticiser)
10
o L_~~~~~~~~~~~~~~~~~~~~~~~~ __ ~~~~~~~
o 10 15 20 25 30
Time (days)
Figure 6.11 - Compressive strength versus time for electrically (3hr delay, 400C/hr
rise, 60°C maximum) and normally cured samples with and without
superplasticiser
350
300
250
~
'"
El
.1:1
Q 200
'-'
.....
"u1:1
til 150
o~
~
100 -+-- Electrically cured (superplasticiser)
____Electrically cured (no superplasticiser)

50
0
0 200 400 600 800 1000 1200 1400 1600 1800
Time (minutes)
Figure 6.12 - Resistance versus time of electrically (3hr delay, 400C/hr rise, 60°C
maximum) cured concrete samples with and without superplasticiser
201
Figure 6.12 shows the effect of replacing water with 21 of superplasticiser per 100kg of
cement upon resistance during (3hr delay, 40°Clhr rise, 60°C maximum) and after
heating. From this it can be seen that the mix containing the superplasticiser initially
had a slightly higher resistance but at later ages the mix not containing superplasticiser
had a slightly higher value. This initially higher resistance may have been due to the
inherent superplasticiser resistance properties which then acted as a water reducer, thus
lowering the resistance of the more workable concrete. The lower final resistance of the
superplasticiser mix may indicate the relative level of hydration between the two
samples.
6.4 Accelerating Admixture

Figure 6.13 shows the effect of replacing water with 0.4 litres of accelerator per lOOkg
of cement. From this it can be seen that the normally cured mix containing the
accelerator achieved a higher compressive strength at all ages than the mix containing
no accelerator.
~ r-----------------------------------------------------------,
50
N
--ee
Z 40
-e
'-'
-=till
-....
1:1
til
30
..a
•r;;
til
20
• Normally cured (accelerator)
ee
u • Electrically cured (accelerator)
10 • Normally cured (no accelerator)
x Electrically cured (no accelerator)
0
0 10 15 20 25 30
Time (days)
Figure 6.13 - Compressive strength versus time for electrically (3hr delay, 40°C/hr
rise, 60°C maximum) and normally cured samples with and without accelerator
202
The electrically cured (3hr delay, 40°Clhr rise, 60°C maximum) mix containing the
accelerator achieved a lower compressive strength increase (when compared to the
equivalent normally cured mix) at early ages (1 day) than the mix not containing
accelerator. At later ages (28 days) both mix types attained a similar compressive
strength loss when compared to the equivalent normally cured samples. This was
because the setting time of the mix containing accelerator was reduced, thus the mix
containing the accelerator acted at 3 hours after mixing in a similar manner to the mix
not containing the accelerator at 4 or 5 hours. This means that the effect of heating upon
hydration was reduced at early ages, but microstructural damage was also reduced due to
the relatively increased structural strength at the beginning of heating.
6.S Retarding Admixture

Figure 6.14 shows the effect of replacing water with 0.4 litres of retarder per 100kg of
cement.
~ r----------------------------------------------------------,
50
....-..
e
....Ze 40
..
'-'
.1:1
~
..'"
Cl
f::
41
30
.e
'"'"
41
~ 20 • Normally cured (retarder)
a
Q • Electrically cured (retarder)
u
10 ~ Normally cured (no retarder)
X Electrically cured (no retarder)
0
0 10 15 20 25 )0
Time (days)
Figure 6.14 - Retarder compressive strength versus time for electrically (3hr delay,
400C/hr rise, 60°C maximum) and normally cured samples
203
It can be seen that the effect of adding a retarder was similar to that of adding a
superplasticiser. It can be seen that with electrical curing (3hr delay, 40°CIhr rise, 60°C
maximum) the mix containing the retarder attained a relatively higher initial
compressive strength increase (compared to an equivalent normally cured mix) than an
electrically cured mix not containing retarder. This situation was reversed at 28 days
with the electrically cured mix containing retarder incurring the lowest compressive
strength of the samples tested. This was due to the same reasons as Figure 6.11.
6.6 Pulverised Fuel Ash Addition

Figure 6.15 shows the effect of 30% PFA cement replacement on compressive strength
over time for normally and electrically (2hr delay, 40°Clhr rise, 60°C maximum) cured
concrete.
~ ,-----------------------------------------------------------,
50
• Normally cured (PFA)

• Electrically cured (PFA)
10 • Normally cured (no PFA)

X Electrically cured (no PFA)
o 10 15 20 25 30
Time (days)
Figure 6.15 - Compressive strength versus time for electrically (2hr delay, 400Clhr
rise, 60°C maximum) and normally cured samples with and without PFA
From this it can be seen that the normally cured mix containing PFA attained a
significantly lower compressive strength at later ages (28 days) than the sample not
204
containing PFA (Berry and Malhotra 1980). This was likely not to be the case at much
later ages where PFA has been shown to have increased compressive strength over
concrete not containing PFA (Neville 1988). The initial compressive strength gain of
the electrically cured PFA samples was reduced when compared to the electrically cured
mix not containing PFA. This may have been due to damage incurred during heating,
where the PFA mix had an increased setting time and thus did not attain the same degree
of strength to be able to resist the thermomechanical stresses imposed during heating as
the mix not containing PFA. It may also have been due to the addition of another solid
material with a different coefficient of thermal expansion to the other mix components.
45 r-----------------------------------------------------------,
40 NorDlll28
",'-" 35
Ea
8
Z
'-'
30
.1:1 Nortml7
C"Il 15
..e'"
c:I -
..
Q,j 20
·i
a. IS
ee Nonnall
U 10
• 1 day • 7 days • 28 days

o L_ ~ ~ ~ ~ ~ ~ ~ ~ ~ __ ~
o 4
Delay (hours)
Figure 6.16 - Effect of delay period on compressive strength for 30% PFA cement
replacement (400C/hr rise, 60°C maximum)
Figure 6.16 shows the effect of delay period on the compressive strength of electrically
(40°Clhr rise, 60°C maximum) cured mixes containing 30% PFA cement replacement.
From this it can be seen that the 3 hour delayed samples performed best in both the short
and long-term. The 4 hour delayed samples did not attain a significant degree of
compressive strength improvement at early ages. This was a similar result to Figure
205
5.18 and was due to similar reasons. The differences between the two figures were
likely to be due to the different setting times between mixes with and without PFA
(Bamforth 1992).
Figure 6.17 shows the effect of 30% PFA cement replacement upon resistance during
and after heating (2hr delay, 40°CIhr rise, 60°C maximum). From this it can be seen
that the initial resistance of the mix containing PFA was higher than the mix not
containing PFA. This was likely to be due to the physical form and resistance properties
of PFA (Dhir 1986) which may have increased the resistance of the cementIPF A matrix.
The final resistance of the PFA mix was lower than the equivalent mix not containing
PFA. This reflected the relatively reduced level of hydration present and the reduced
final compressive strength of the mix containing PFA.
400
350
300
.-.
'"Ii! 250
-=
Cl
'-'
.."'"
y 200
1:1
et
.;;
150
=:"
100 -+- Electrically cured (PF A)
___ Electrically cured (no PFA)

50
0
U 200 400 600 SUU 1000 1200 1400 1600 1800
Time (minutes)
Figure 6.17 - Resistance versus time for electrically (2hr delay, 400C/hr rise, 60°C
maximum) cured samples with and without PFA
Figure 6.18 shows the initial surface absorption characteristics of normally and
electrically (2hr delay, 40°CIhr rise, 60°C maximum) cured samples at 28 days. From
206
this it can be seen that the electrically cured samples attained a lower ISA value
throughout the test than the normally cured samples. This was a similar result to Figure
5.46 and was again likely to be due to collapsed capillary pores incurred during heating.
02 r---------------------------------------------------------,
0.15
• Normally cured
----e
-.. • Electrically cured
...'"
~
e
'-'
0.1
'-'
<
rn
....
005
o 10 20 30 40 so 60 70
Time (minutes)
Figure 6.18 - ISA for electrically (2hr delay, 400Clhr rise, 60°C maximum) and
normally cured concretes containing PF A
6.7 Conclusions
1. There exists a relationship between ultrasonic pulse velocity (UPV) and density for
both normally and electrically cured concrete samples.
2. The UPV values for both electrically and normally cured samples were found to give
an approximate estimation of percentage of air entrainment in concrete at 28 days.
3. It was found that UPV could be used to give an indication of compressive strength
for all concrete samples tested.
207
4. It was found that a compressive strength loss of electrically cured samples
containing high levels of air entrainment could be achieved in extreme
circumstances if very short delay periods were chosen and a 4 hour isothermal
period was applied at 80°C. This was likely to be due to increased microstructural
damage incurred around the entrained air bubble cement paste shells during heating
and cooling.
5. The reduction of compressive strength at 28 days for electrically cured air entrained
concrete when compared to normally cured concrete was found to be dependent on
the percentage of air entrainment, with higher levels of air entrained producing
greater reductions in compressive strength.
6. The maximum temperature of electrically cured air entrained concrete was found to
influence compressive strength at all ages, with higher temperatures producing lower
compressive strength. This was likely to be due to increased microcracking around
the entrained air bubbles at the higher temperatures.
7. The effectiveness of electrical curing in improving compressive strength on mixes

containing air entrainment was found to be reduced when compared to similarly
designed mixes not containing air entrainment. This was also due to increased
microcracking around the entrained air bubbles incurred during heating.
8. Resistance during heating was found not to be a good indicator of percentage of air
entrainment.
9. Results show that normally cured samples containing a superplasticiser attained a

higher compressive strength at all ages than a similar mix without superplasticiser.
The electrically cured concrete samples containing superplasticiser cured after a
delay of 3 hours acted in an identical manner to a heated mix without
superplasticiser with a delay period of 1 or 2 hours. This was due to the
208
superplasticiser mix having a greater period before the initial set which effectively
reduced the delay period.
10. Resistance was not found to be a good indicator of the presence of superplasticiser
added to the mix.
11. It was found that normally cured samples containing an accelerator attained a higher
compressive strength at all ages than a similar mix without accelerator. The
electrically cured mix with accelerator cured after a delay of 3 hours acted in a
similar manner to a heated mix without accelerator with a delay period of 4 hours.
This was due to the mix with accelerator having a reduced period before initial set
which effectively reduced the delay period.
12. It was found that both normally and electrically cured concrete containing a retarder
behaved in a similar manner to concrete containing a superplasticiser.
13. It was found that with cement replacement by PFA the effectiveness of electrical
curing in improving compressive strength was reduced at 1 day. This was likely to
be due to the concrete containing PFA having an extended period before initial set,
thus being less able to resist the thermomechanical stresses incurred during heating.
14. Resistance was not found to be a good indicator of the presence ofPFA but may be
used as an indication of the extent of hydration.
209
CONT AINING REINFORCEMENT
7.1 Introduction
investigation of normally and electrically cured concrete containing reinforcement.
These results can be seen in Figures 7.1-7.27 and Tables 7.1-7.2
The mixing, casting, curing and monitoring was implemented in accordance with the
experimental procedures stated in Chapter 3.
Electrically cured slabs A and B and the normally cured slabs all consisted of a C30 mix
with a TMS = 43N/mm2 and a slump of 10-30mm. The electrically cured slabs had a
delay period of 4 hours, with a linear temperature rise of 40°Clhr to a maximum of
60°C. The temperature datum was taken from position 1 for Slab A and position 3 for
Slab B (see Figures 3.15 and 3.16).
7.2 Normally Cured Slab
Figure 7.1 shows Schmidt hammer number for normally cured slab and
150x 150x 150mm cube compressive strength versus time. From this it can be seen that
both Schmidt hammer and compressive strength followed a similar trend of increased
strength over time (Xuenquan et a1 1986). There was a variability inherent in the
Schmidt hammer results due to only the immediate surface of the concrete being tested
(Bungey 1982). As such, aggregate particles or air voids immediately below the surface
could significantly higher or lower the Schmidt hammer number respectively.
210
40 r--------------------------------, eo
35
so I"'l
..
Jl 30
c
0-
n>
..,
e:s 4U
e
3
iii
... 25
.,
"0
~ ~
e '"::;:r
e
·s-
n>
'"...
30
] 20
.,
n>
'0
:I
.c
CJ
15
• Schmidt hammer number zo
...=-
!IQ
~
..;: 10 • Cube compressive strength ~
V3 9
10 ~
o 10 15 20 25 30
Time (days)
Figure 7.1- Schmidt hammer number and 150mm cube compressive strength
versus time for normally cured slab
40 ~--------------------------------,
35
t
.c 30
e:s
1:1
t 25
e
.=e y = 0.4076x + 14.264
-
:9
e
.c
20
15
~
..;: 10
V3
o 10 20 30 40 50 60
Cube compressive strength (N/mml)
Figure 7.2 - Slab Schmidt hammer number versus 150mm cube compressive
strength for normally cured samples
211
Figure 7.2 shows the relationship between average Schmidt hammer number and
compressive strength of 150mm cubes for normally cured samples over time. From this
it can be seen that the relationship can be assumed to be approximately linear which is
given by:
Schmidt hammer number = 0.407 x (cube compressive strength) + 14.264 Eqn.7.1
From this, approximate actual compressive strengths for the mix used (C30) could be
derived from the Schmidt hammer results gained from the electrically cured slabs.
7.2.2 Core failure load
Figure 7.3 shows the average core failure load for a normally cured slab and cube
compressive strength versus time.
2S~ .- --,60
50 Q
C"
~
r:>
Q
40 51
'Cl
~
'"'":;;.
30 ~
'"...
~
• Core failure load =
IJQ
20 ;.
• Cube compressive strength
Z
a-
10 EL
'-'
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0
o 10 IS 20 2S 30
Time (days)
Figure 7.3 - Core failure load and 150mm cube compressive strength versus time
for normally cured samples
212
From this it can be seen that both core failure load and compressive strength followed a
similar trend of increased strength over time. This increase was due to the continued
hydration incurred by both samples which increased the strength of the cement matrix
and thus increased both the cube and core crushing values (Neville 1988).
Figure 7.4 shows the relationship between average core failure load and cube
compressive strength for normally cured samples over time. From this it can be seen
that the relationship can be assumed to be approximately linear and can be given by:
core failure load = 414 x (cube compressive strength) +1279.9 Eqn.7.2
From this, approximate actual compressive strengths for the mix used (C30) could be
derived from the core failure load gained from the electrically cured slabs.
25000 ,...-----------------------------,
•
Z 20000
--
'C
~
oS
~ 15000
..:
:i y = 411.4x + 1279.9
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 10 w ~ ~ 50 60
Cube compressive strength (N/mm2)
Figure 7.4 - Core compressive strength versus actual compressive strength for
normally cured slab
213
7.2.3 Temperature history
Figure 7.5 shows the temperature changes in a normally cured slab. The thermistor
temperature sensors were placed in a similar arrangement to the electrically cured Slab
A. Individual sensor placements have not been identified due to the similarity of all
results. From this it can be seen that the temperature rose for approximately the first
300 minutes after gauging with water, whereupon a slight lowering oftemperature could
be seen to occur by 1300 minutes. This was due to Alite (impure C3S) hydration
reactions producing an exothermic chemical reaction (Neville 1988, Gibbon and Ballam
1998). The temperature rise was slight due to the placement of the thermistors on the
surface of the normally cured slab. This reduced the apparent heating effect due to the
effect of surface cooling by the atmosphere and the thermal insulation of the concrete.
Within the centre of the slab, temperature rises would be expected to be slightly higher.
All thermistor placements showed approximately the same temperature at all ages
(±2°C) and the average temperature over the entire time tested was approximately 23°C.
o 200 400 600 800 1000 1200 1400
Time (minutes)
Figure 7.S - Temperature versus time for a normally cured slab
214
7.3 Electrically Cured Slab A
7.3.1 Visual inspection
Visual observations were carried out on the slab during and after heating. It was
observed that during heating, visible surface cracks along the reinforcement bars
(approximately Imm wide) occurred after the first IS minutes and subsequently steam
was seen to be rapidly emitted from the underside of the slab. Concurrent to this
audible 'splintering' sounds were heard to be coming from the sample and air bubbles
were seen to be emitted from specific points across the surface. The bubble emission
sites remained in a fixed position throughout the test and as such it can be assumed that
pathways of gas were present in the body of the concrete that allowed the heated air and
steam to escape. This would damage the structure of the concrete in specific regions,
whilst leaving relatively undamaged sections between the emission points. This was
likely to be due to the placement of the temperature sensors near the surface of the slab
which resulted in the internal overheating of the sample in order to achieve the 60°C
maximum temperature. This required excessive voltages to be applied initially and as
such caused localised boiling which induced the observed cracks, bubbles and sounds.
Figure 7.6 shows the derived average (70 results per point) Schmidt compressive
strength for electrically cured Slab A and the normally cured compressive strength over
time. From this it can be seen that the electrically cured slab attained (on average) a
higher initial compressive strength than the normally cured samples. This situation was
reversed at later ages (>7 days) were the normally cured samples continued to gain
compressive strength at a greater rate than the electrically cured slab (Hutchinson et al
1991). This showed the same trend as the electrically cured cube samples not
containing reinforcement. The percentage gain, when compared to the normally cured
slab, for the electrically cured slab was 167.5% at I day and -28.5% at 28 days. The
percentage gain for the electrically cured cube, when compared to the normally cured
cube, was 98.8% at I day and -18.9% at 28 days. The greater increase in percentage
215
gain initially and the greater decrease in percentage gain at 28 days of the electrically
cured slab was due to the increased temperature internally required to achieve the
superficially similar heating regime to the cube. This extra temperature within the slab
was due to the electric field density around the reinforcement bars being substantially
higher than the surrounding areas. This field density then heated up the surrounding
concrete at a much higher rate than the surface areas and then heat was transferred solely
by conduction to the thermistors placed at 20mm below the surface. The increased
temperature internally was slowly distributed (Gibbon and Ballam 1998) throughout the
sample after heating had ended, thus increasing the early age maturity, but at later ages,
the higher initial internal temperatures caused greater structural damage than the cube
heating regime.
70
--
Ne 60
~
..
-=
~
50
..
1:1
~
'"
~
40
.~ •
..
'"'"~
Co
30
El • Electrically cured slab A

~
f.j
20
~
..~
~ • Normally cured slab
...
0( 10
0
0 10 20 30 40 50 60
Time (days)
Figure 7.6 - Average compressive strength versus time for electrically and
normally cured slab (derived from Schmidt hammer)
Figure 7.7 represents the average (10 results per data point) compressive strength
(derived from Schmidt hammer) perpendicular to the reinforcement bars (along the
length of the slab) of Slab A at various ages. The centre of the slab is marked by the
216
vertical dotted line. From this it can be seen that the 1 day compressive strength varied
significantly across the width of the slab. Where reinforcement bars are present (-100, 0
and 100mm), a greater degree of 1 day compressive strength was achieved than in the
regions between the reinforcement bars. This was due to the extra heating to be
expected around the reinforcement bars as electric field density is increased. This trend
was reversed at later ages (28 and 56 days) where the regions between the reinforcement
bars showed a greater compressive strength. This was likely to be due to the damaging
effects of temperature gradients causing thermomechanical damage in the regions
adjacent to the reinforcement bars, thus reducing long term strength. This trend was
similar to cube samples electrically cured containing no reinforcement, where greater
temperatures at an early age result in long-term structural damage.
=_=_=_~__
50 ~__~__~__~__~__~__~__~__~__~ ~__~__~__~__~__~__~__~__~__~ ~__~__~__~__~__~__~__~ ~__~_.~__~__~__~__~__~__~ ~__~__~__~__~__~__~__~__~ ~__~__~__~__~__~__~__ ~__
~__
45 Normal28
--
.. ~ 40
~
-= 35
biJ
=
...
~ 30 Nonnal7
'"
GI
.~ 25
'"rl
a 20
-+-1 day
El
Q ___ 28 days
CJ 15
t, ....... 56 days
~
~ 10 --------------------------------------------------------------r-------------------------------------------------------------
~ Nonnall
-200 -150 -100 -50 0 50 100 150 lOO
Distance perpendicular from central rebar (mm)
Figure 7.7 - Average compressive strength versus distance from centre

reinforcement bar (derived from Schmidt hammer)
It can also be seen from the figure that all regions of the slab (at the surface) achieved
the characteristic design strength (C30) but only two sections achieved the TMS
(43N/mm2) at 28 days (-50 and 50mm) (Wilson and Gupta 1996). There was a degree
217
of overlap present where it appeared the 1 day compressive strength was greater than the
28 day compressive strength. This was likely to be due to the innate variability of the
Schmidt hammer apparatus. However, confidence can be given to the general results
due to the apparent symmetry on either side of the centre reinforcement bar.
(derived from Schmidt hammer) parallel to the reinforcement bars of Slab A at various
ages. From this it can be seen that at all ages there was a greater similarity in results
when compared to the compressive strength perpendicular to the reinforcement bars.
This was due to a more uniform temperature distribution in this direction. The
temperature parallel to the reinforcement bars would be expected to be similar apart
from excess heating towards the centre of the sample (Gibbon and Ballam 1998). This
was confirmed by the slightly higher compressive strengths in the central region at 1 day
and the reduced compressive strengths at 28 days. The outer regions showed a
comparatively higher compressive strength at 28 days, due to reduced microstructural
damage from heating effects.
50 • __ . __.. _
45 Nonnal28
-
1-
oC:I
CIf)
1:1
40
35
-"
~ 30
'"
.i:: 25
Nonnal7
-+-1 day
....... 28 days
'":I
6. 20
ae
CJ 15
"~
CIf)
t 10 • .. -· ...... •• .. -- .. • .. ··- .... --- .. •• .. - .. -· .. • .... --- .. --·· .. -f .. --- .. -··· ... -- ... ---------- •• - ... --- .•• -- .. ------ .• ---- ••. - .
.:: Nonnal I
o L-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
·300 ·200 ·100 o 100 200 300
Distance from centre ofslab parallel to rebars (mm)
Figure 7.8 - Average compressive strength versus distance from centre of sla b
parallel to reinforcement bars (derived from Schmidt hammer)
218
Figure 7.9 shows the average (14 results per data point) core compressive strength for
electrically cured Slab A and the normally cured slab over time. From this it can be
seen that the electrically cured slab attained (on average) a higher initial compressive
strength than the normally cured samples (Wilson and Gupta 1996). This situation was
reversed at later ages (>3 days) were the normally cured samples continued to gain
compressive strength at a greater rate than the electrically cured slab. This showed the
same general trend as the electrically cured samples not containing reinforcement. The
percentage gain, when compared to the normally cured slab, for the electrically cured
slab was 129.2% at 1 day and -38.5% at 28 days whereas the percentage gain for the
electrically cured cube was 98.8% at 1 day and -18.9% at 28 days.
70 r-----------------------------------------------------------,
• Electrically cured slab A
• Normally cured slab
o 10 20 30 40 50 60
Time (days)
Figure 7.9 - Average core compressive strength versus time for electrically and
normally cured slab (derived from core results)
Figure 7.10 shows the compressive strength (derived from core results) at 56 days
perpendicular to the reinforcement bars at various locations along the length of Slab A.
219
From this it can be seen that the central regron attained a significantly lower
compressive strength than the outer regions (Owens 1985). This was due to the long-
term damage caused to the central region by thermomechanical stress caused by the
initial heating.
70 ,----------------------------------------------------------,
60
;:
I)J) 40
1:1
t
'"
~ 30 1-+-50~
.;;;
....... 130mm'
~
Cl. 21Omml
ee 20
1 ........
""*"290mm!
U
!_'_370mm
10
l;4S0mml
-+- 53 Omm I
o ~~~~~~~~~~~~~~~~~~_L~~~~~~~~~~_J
·200 ·150 -100 -50 o 50 100 150 200
Distance from central rebar (mm)
Figure 7.10 - Compressive strength at 56 days perpendicular to reinforcement bars

(derived from core results)
Figure 7.11 shows the differing compressive strengths (N/mm2) (derived from core
results) over the surface of Slab A at 56 days. From this it can be clearly seen that
compressive strength was approximately uniform parallel to the reinforcement bars
along the entire length of the slab. The central region clearly showed a reduced
compressive strength as in Figure 7.10. From this it can be stated that, for the particular
slab tested, compressive strength was mainly dependent on the distance perpendicular to
the reinforcement bars and once this has been estimated, then the compressive strength
can be assumed to be approximately equal throughout the length of the slab. This would
not be valid for larger sections where heat insulation effects could significantly increase
the temperatures in the central region, thus affecting structural damage. It would also
220
not be valid for different geometrys (Leung and Pheeraphan 1994) or reinforcement bar
layouts.
.60-65
.55-60
050-55
~ .45-50
~ .40-45
I 35-40
.30-35
.25-30
.20-25
L115-20
Length (mm)
Figure 7.11- Compressive strength at 56 days (derived from core results)
Figure 7.12 shows a comparison of the average derived compressive strengths for Slab
A for both core and Schmidt hammer results over time. From this it can be seen that the
compressive strength derived from the cores was lower than the compressive strength
derived from the Schmidt hammer results over the entire time tested (56 days). This
was due to the cores being taken from a greater depth than the Schmidt hammer which
is only a surface measurement of strength. Thus the cores contained a different
temperature history throughout their length and as such were going to fail under load at
the weakest point. This was at the point closest to the surface at early ages (1 day) and
at the point closest to the reinforcement bars at later ages (>28 days). This implied that
extra care should be taken when deciding upon a method for testing compressive
strength in electricalIy cured samples as normally cured concrete (excluding mass
concrete) tends to have a similar temperature history within its structure and as such
tends to have a more uniform compressive strength throughout.
221
4< ,----------- --,
•
•
.Core
• Schmidt hammer
o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 10 20 30 40 50 60
Time (days)
Figure 7.12 - Average compressive strength versus time (derived from core and
Schmidt hammer results)
7.3.4 Porosity
Figure 7.13 shows the porosity (%) at different points over Slab A at 56 days. From this
it can be seen that the porosity only varied between 13.8% and 17.4% over the entire
surface of the slab. This was similar to the porosity of the normally cured concrete at
the same age (approximately 16%). It can also be seen that porosity appeared to be
roughly similar parallel to the reinforcement bars, with areas of greatest porosity
corresponding roughly to the area along the reinforcement bars. This was not for
certain, however, due to the small differences that were apparent. The apparent increase
in porosity may have been due to localised boiling of the free water within the mix,
causing an increase in pore size and microcracks through the expansion of water
(through conversion to steam) and air. However, this may have been in conjunction
with the collapse of capillary pores during heating which would lower the porosity
around the reinforcement bars (Parcevaux 1983).
222
.17.5-18
.17-17.5
16.5-17
.16-16.5
~
go
mI5.5-16
I .15-15.5
014.5-15
014-14.5
.13.5-14
.13-13.5
Length(mm)
Figure 7.13 - Porosity over slab surface at 56 days
7.3.5 Density
Figure 7.14 represents the distribution of density (kg/rrr') over the surface of Slab A at
56 days. From this it can be seen that the density was approximately equal over the
entire slab (2225-2350kg/m3). This confirmed the results from Figure 5.51, which
showed that the heating regime used did not significantly affect density. This was
clearly due to the fact that the concrete used was of a uniform initial density and
although electrical curing did inevitably lead to initial small water losses, in the long-
term this did not affect the overall density.
223
112475-2500
02450-2475
.2425-2450
.2400-2425
:::;; 112375-2400
s:
~ .2350-2375
I 112325·2350
.2300-2325
02275-2300
02250-2275
.2225-2250
.2200-2225
·75 o 75 150 225
Lengtb(nun)
Figure 7.14 - Density over slab surface at 56 days
8.5-9
.8-8.5
:1 .7.5-8
~ .7-7.5
I 06.5-7
06-6.5
.5.5-6
.5-5.5
·ISO ·75 o 75 150
Length(mm)
Figure 7.15 - Water absorption over slab surface at 56 days
224
7.3.6 Water Absorption
Figure 7.15 shows the distribution of water absorption (%) characteristics over the
surface of Slab A at 56 days. From this it can be seen that the water absorption showed
a similar distribution to the porosity given in Figure 7.13. This also showed relatively
small differences in water absorption over the entire surface and was due to porosity and
water absorption being closely related (Neville 1988).
7.3.7 Temperature profile
The temperature profile along the surface of Slab A was found to be inaccurate. This
was due to the manual placement of the thermistors at a depth of approximately 20mm.
Thus a small deviation from this depth incurred a relatively large distortion of the
temperature profile due to the temperature gradients inherent during electrically
accelerated curing.
Width (mm)
Figure 7.16 - Temperature profile along Slab A at 320 minutes after gauging with
water
225
An example of the surface profile at 320 minutes after the gauging with water can be
seen in Figure 7.16 (same colour key as Figure 5.53) and a full list of temperatures over
various ages can be seen in Table 7.1. These showed that temperature generally tended
to be relatively decreased at the outside edges when compared to the central region.
Table 7.1 - Temperature of Slab A at various ages
Time after Sensor number (from Chapter 3

gauging witb 1 2 3 4 5 6 7 8 9 10
water (min)
240 15.8 15.8 15.7 15.7 15.7 15.6 15.7 15.6 15.5 20.6
250 23.1 24.6 22.2 36.5 23.1 23.0 22.2 24.0 26.5 22.2
260 33.1 35.4 28.1 43.2 35.7 29.3 30.8 34.5 32.4 22.5
270 38.5 39.3 31.8 46.1 41.9 33.2 35.6 38.8 34.7 23.1
280 45.4 45.0 37.1 57.5 48.6 39.2 41.6 44.7 40.8 23.8
290 51.0 50.5 41.5 62.7 55.4 43.9 46.6 50.0 44.4 24.8
300 55.3 56.1 47.2 70.6 62.0 50.3 51.6 56.2 51.0 24.9
305 56.9 57.9 48.8 72.1 64.7 51.9 53.7 58.4 52.5 23.8
310 58.1 59.0 51.1 72.9 66.3 54.2 55.2 59.9 54.8 24.2
315 59.1 59.4 52.3 73.1 67.0 55.5 56.3 60.6 56.0 24.2
320 59.8 59.5 53.2 73.0 67.1 56.4 57.0 60.7 56.8 24.3
1165 36.6 35.7 31.0 36.6 35.9 34.2 33.5 33.0 34.6 20.9~-
1370 29.7 29.2 25.2 29.9 29.0 28.1 28.2 27.8 28.4 22.0
7.4 Electrically Cured Slab B
7.4.1 Visual inspection
Visual observations were carried out on the slab during and after heating. It was
observed that at no time did visible cracks appear on the surface of the slab and no
'excessive' localised steam was seen to be emitted from the slab.

(derived from Schmidt hammer results) for Slab B and the compressive strength of
normally cured cubes. From this it can be seen that the electrically cured slab attained a
higher initial compressive strength but at later ages (>4 days) the normally cured
samples attained a higher compressive strength. The initial compressive strength gain of
the electrically cured slab was due to the increased hydration caused by the higher
226
temperatures and the later reduction in strength was due to damage (Kirkbride 1996)
caused to the structure of the cement matrix incurred during the initial heating phase.
This followed the same trend as that of an equivalently cured cube containing no
reinforcement. The percentage gain, when compared to equivalent normally cured
samples, for the electrically cured slab was 153.6% at 1 day and -38.7% at 28 days
whereas the percentage gain for the electrically cured cube not containing reinforcement
was 98.8% at I day and -18.9% at 28 days. This was again likely to be due to the
greater heating surrounding the reinforcement bars which increased the early age
maturity of the slab but induced long-term structural damage.
~ r-----------------------------------------------------------,
.-..
Ne 50
e
Z'-'
oCl
DiI 40
.t..
'"~
.. 30
.;;
t
Co
•
eo
(.,I
'0
- • Electrically cured Slab B
~
~ • Normally cured slab
E
~ 10
«
o 10 15 20 25 ]0
Time (days)
normally cured slab (derived from Schmidt hammer results)
(derived from Schmidt hammer) perpendicular to the reinforcement bars at various
distances from the central reinforcement bar at various ages. From this it can be seen
that at I day the compressive strength above the reinforcement bars (-100, 0, lOOmm)
tended to be higher than the compressive strength between the reinforcement bars. This
227
effect was reduced as the Schmidt hammer is a surface measurement of compressive
strength (Bungey 1982) and temperature effects were minimised due to the thermal
insulation of the concrete (Gibbon and Ballam 1998). This reduced temperature
differences at the surface, which in tum reduced the effective differences in maturity.
The situation was reversed at 10 days where the compressive strength was lower in the
areas adjacent to the reinforcement bars when compared to the areas between
reinforcement bars. At 28 days the compressive strength appeared to be contradictory to
the 10 days result. This may have been due to a combination of the innate variability of
the Schmidt hammer test and the relatively minor differences in surface temperature.
However, the highest reduction in compressive strength at 28 days was apparent above
the reinforcement bars at -100 and 100mm distance from the central reinforcement bar.
50 ~ __~__~__~__~__-__-__-__-__- - __-__-__-__-__-__-__-__- - __
- __-__-__-__-__-__-__- ~_.-__-__-__-__-__-__~__~ ~__~__~__~__~__~__~__~__~ - __
~__~__-__-__-__-__-__-__- - __, _
Nonnal28
45
Ne
El 40
Z
-
:;
~
1:1
~ 30
35
Nonnal7
'"
41
.. 25
.w;
fl
a 20
-+-1 day
El
e ___ 10 days
<.I 15
41
~ ...... 28days
t"
..
<
10 ------------------------------------------------------------~--------------------------------------------
Nonnall
--------- --- - -- --
o
-200 -150 -100 -50 o 50 100 ISO 200
Distance perpendicular from central rebar (mm)
Figure 7.18 - Average compressive strength versus perpendicular distance from

central reinforcement bar (derived from Schmidt hammer results)
(derived from Schmidt hammer results) parallel to the reinforcement bars at various
ages. The compressive strength at 1 day showed a similar compressive strength along
228
the reinforcement bars (20N/mm\ This may have again been due to the small
temperature differences that occurred on the surface (Figure 7.27). At 10 and 28 days
the compressive strength appeared to be slightly reduced in the central region. This was
due to the slightly higher temperatures present at the central region which caused a
greater degree of structural damage than the outer regions. The effect was not as
pronounced as for the compressive strength of cubes containing no reinforcement
because compressive strength derived from crushing values was a more accurate method
than compressive strength derived from Schmidt hammer (Bungey 1982).
45
Ne
a 40
Z
...~
:; 35
1:1
...
~ 30
'"
"~
'fil
25
~
:;. 20 ....... 1 day
ae ---10 days
... 15
........28 days
"~1'1
~ 10 --_----- ----_ -_ ------_-- -------------------------------------~------------------------------_------------ - --_-------- - -_---
>
-< Nonnall
o L_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
·300 ·200 .100 0 lOO 20() 300
Distance parallel to rebars (mm)
Figure 7.19 - Average compressive strength versus distance from centre of Siab B
parallel to reinforcement bars (derived from Schmidt hammer results)
Figure 7.20 shows the average (14 results per data point) electrically cured Slab B
(derived from core results) and normally cured slab compressive strengths over time.
From this it can be seen that the electrically cured compressive strength attained a
significantly higher value at initial ages (I day) than the normally cured slab. Though
229
the compressive strength of the electrically cured slab did not achieve the same rate of
strength increase after 1 day, the 28 day compressive strength of both samples were
approximately the same (50 N/mm2). This long-term strength equivalence was due to
the averaging effect of the core failure loads, as within the outer region, compressive
strengths were significantly higher than the inner regions. Also, the alteration of the
depth at which the maximum temperature was measured for this slab was different to
that of the cube set-up.
~ .---------------------------------------------------------~
• Electrically cured Slab B
• Normally cured
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 10 15 20 25 30
Time (days)
normally cured slab (derived from core results)
Within the cubes the maximum temperature differed by a maximum of 8.2°C when the
assumed temperature was 60°C at a depth of 20mm, but within the slab a difference of
up to 22.5°C occurred when a depth of 75mm was chosen as a datum and as such a
greater degree of temperature range can be assumed to occur within a slab. This shows
that the position of the temperature sensors used to determine the heating regime is
critical to the short and long-term properties of the concrete, especially when complex
geometrys and/or reinforcement bars are used.
230
Figure 7.21 shows the distribution of compressive strength perpendicular to the
reinforcement bars from the central reinforcement bar along the length of Slab B at 28
days. From this it can be seen that over the entire length of the slab, compressive
strength at the centre was significantly less than at the outer edges. This was due to the
long-term damage to the central structure of the concrete that attained significantly
higher temperatures, and thus higher thermomechanical stresses. than the outer regions
(Orchard and Barnett 1971).
~ r---------------------------------------------------------~
80
«: 70
El
El
Z,_, 60
of
: 50
b...
GI 40
.::... -+-100mm
~ 30 -e-200mm
Co.
El .........300mm
e
U 20
""*""400mm
10 ____500mm
o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~_J
-zoo -150 -lOO -50 o 50 100 ISIl 200
Width (mm)
Figure 7.21- Compressive strength at 28 days perpendicular to reinforcement bars

(derived from core results)
The compressive strength profile parallel to the reinforcement bars may have been more
complicated than the figure shows. This was due to the fact that no cores could be taken
immediately above the reinforcement bars. However, it could be expected from
previous results (Figure 7.18), that the 28 day compressive strength was lower in the
regions directly above the reinforcement bars. This was again due to the increased
temperatures in the region causing increased thermomechanical damage to the long-term
structure of the concrete. The effect on internal compressive strength was likely to be
231
greater than the effect on surface compressive strength (derived from Schmidt hammer
results) due to the higher temperature gradients internally (Figures 7.24 and 7.25).
Figure 7.22 shows the compressive strength (N/mm2) distribution (derived from core
results) over the surface of Slab B at 28 days. This again shows that the compressive
strength was higher in the outer regions than the inner regions reaching a maximum of
2
75N/mm and a minimum of 30N/mm2. This minimum value was equivalent to the
characteristic strength but did not reach the target mean strength (TMS = 43N/mm2) .
I \" • 75-80
.70-75
, '" I
e-
to
1:1
ag.
I
.65-70
.60-65
.55-60
050-55
.45-50
~. ~" .40-45
35-40
"I I
Width(mm)
I
~
I
II
.30-35
Figure 7.22 - Compressive strength at 28 days (derived from core results)
Figure 7.23 shows the comparison of the average compressive strength derived from the
Schmidt hammer and core failure load results for Slab B over time. From this it can be
seen that the compressive strength results were significantly different for the two
methods of estimating compressive strength. The compressive strength derived from
the core failure load results attained a higher value than the values derived from the
Schmidt hammer results. This was the opposite result to Figure 7.12 and was due to the
position of placement of the temperature sensors. In Slab B, the temperature sensor
232
used to determine the heating regime was placed at a depth of 75mm at the centre of the
surface of the slab. This means that the temperature in the centre of the slab did not
reach temperatures at which excessive structural damage may occur (approximately
>60°C) but did have a higher maturity value than the corresponding surface points. In
Slab A the temperature was measured at a depth of 20mm which was likely to cause an
excessive degree of structural damage in the centre of the slab where the expected
higher temperatures (Orchard and Barnett 1971) were not monitored.
~ ,-----------------------------------------------------------,
•
•
• Core
• Schmidt hammer
o 10 15 20 3U
Time (days)
Figure 7.23 - Average compressive strength versus time (derived from core and
Schmidt hammer results)
7.4.4 Temperature history
Figure 7.24 shows the temperature profile at vanous distances from the central
reinforcement bar perpendicular to the reinforcement bars at a depth of 75mm at the
centre of the Slab B during heating. From this it can be seen that in the vicinity of the
reinforcement bars (-100, 0, 100mm) a significantly higher temperature was attained.
This was due to the concentration of the field density around the reinforcement bars
233
which thus increased the temperature. It can also be seen that the outer edges attained a
lower temperature at all ages than the sensors placed between the reinforcement bars.
This was due to cooling incurred at the external surface (Orchard and Barnett 1971) and
shows the importance of adequate monitoring of the temperature gradients, thermal
stresses and maturity. A possibility of reducing these gradients may be to reduce the
spacing between reinforcement bars.
80
-+-240min
___ 250min
70 ---a- 260min
""*- 270min
_'_280min
60
-+-290min
-+-299min
G
0 50
'-
-""
I..
::I
~
I..
40
Co
e 30
"
Eo-
20
10
0
-~OO -150 -100 ·50 o 50 lOO 150 200
Width (mm)
Figure 7.24 - Temperature profile perpendicular to the reinforcement bars at the

centre of Slab 8 at a depth of 75mm
Figure 7.25 shows the temperature profile at various distances from the central
reinforcement bar perpendicular to the reinforcement bars at a depth of 20mm at the
centre of the slab during heating. From this it can be seen that the overall temperatures
and temperature gradients present near the surface were greatly reduced when compared
to the equivalent temperature gradients at greater depths. This was due to the insulating
thermal properties of concrete (Gibbon and Ballam 1998) and also by the comparatively
similar field densities that were likely to be present away from the immediate vicinity of
234
the reinforcement bars. This again shows the importance of knowing the temperature
history of the entire sample in order to estimate its short and long-term properties.
RCI r-------------------------------,
........24Umin
-e-250min
70 ....... 260min
--*- 270min
--lIE- 280min
60
....... 290min
-+-299min
10
o ~~~~~~~_L~~~~~~~~~~~~~~~~~~~~~~
·200 ·150 ·100 ·50 o 50 100 150 200
Width (mm)
Figure 7.25 - Temperature profile perpendicular to the reinforcement bars at the

centre of Slab B at a depth of 20mm
Figure 7.26 shows the temperature profile of certain depths at the centre of the slab and
of a thermistor directly in contact with the reinforcement bar. From this it can be seen
that the reinforcement bar did not attain as high a temperature as the concrete
immediately above it. This shows that the reinforcement bar did not act significantly in
heating directly the concrete. This was due to its low electrical resistance which did not
allow significant heating (Wilson and Gupta 1996). The reinforcement bar may have
acted as a cooling agent, allowing the heat to escape into the environment by direct heat
conduction to the external portion of the reinforcement bar surface which was exposed
to the ambient temperatures present (approximately 20°C).
235
w,------------------------------- ,
70
--+- Reinforcement bar
--- 20mm depth
60
L...._ 5~~ depth
G
~
50
f
~
..
Q,l
Cl.
40
~ 30
Eo<
20
10
o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o 200 400 600 800 1000 1200 1400 1600 1800
Time (min)
Figure 7.26 - Temperature history of differing depths at the centre of Slab Band
reinforcement bar
Figure 7.27 shows the cross sectional temperature along the length of the central
reinforcement bar for various ages. From this it can be seen initially (prior to heating at
240 min), temperatures were evenly distributed with a minimum temperature of 19°C in
the upper region «50mm) and a maximum temperature of 21°C in the lower region
(> 1OOmm depth). This initial variation was caused by the exothermic hydration
reactions as the concrete cured normally and the increase at lower depths was due to the
added insulation of the mould shuttering.
During heating, the maximum temperature was always between the reinforcement bar
(40mm) and the bottom electrode (150mm). This was due to the current being passed
only between the two electrodes, thus increasing the temperature rise of the concrete in
this area. The maximum temperature throughout heating was also always in the central
region of the slab. This was due to the thermal insulating properties of concrete (0.6-2.6
WIm°C) (Gibbon and Ballam 1998) which did not allow the temperature to be quickly
distributed.
236
The maximum temperature gradient could be seen to occur vertically in the central
region and decreases with depth. This attained a maximum of approximately
O.25°Clmm at 299 minutes and was due to the comparatively low ambient temperature
(approximately 21°C at the surface). The vertical temperature gradients were not as
large at the lower surface due to the thermal insulating properties of the mould
shuttering.
The temperature gradient along the length of the reinforcement bar was relatively minor
in comparison (maximum ofO.OSoC/mm at 380 minutes). This was due to the similarity
of the layout along the length of the slab, with a reinforcement bar always in the same
relative position. Thus, the only thermal gradients that could occur laterally were those
influenced by the bulk thermal insulation properties of the concrete and as such the
concrete tended to have a higher central temperature (Orchard and Barnett 1971).
Relatively high temperature gradients could be seen I.S hours after heating had ended.
The temperature contours were not centred upon the central reinforcement bar position,
but instead spread evenly throughout the length of the sample. The equivalent
temperature regions could be seen to be enlarging as heat is dissipated to the ambient
atmosphere at the edges of the sample (Orchard and Barnett 1971). After I day, the
temperature recorded at the central region was 32°C, which was still significantly higher
than the ambient temperature.
237
240 min 250 min
19°C 21°C
e
<'I>
~
<'I>
"C "C
er er
[
'-'
[
'-'
260 min 270 min
24°C 27°C
~
<'I>
e
<'I>
"C "0
er er
--.
[
'-' i
280 min 290 min
31°C 34°C
I::'
<'I>
"C
~
"C
er
.-.
er
§
'-'
~
'-'
Length (mm) Length (mm)
Figure 7.27 - Cross-sectional temperature profile over time along the

length of the central reinforcement bar
238
299 min 320 min
37°C 40°C
0 0
~ ~
"C "C
!:T !:T
9'
a
'-"
~
'-"
380 min 1395 min
29°C
e
.g'=' ~
't:I
!:T !:T
~
'-"
~
'-"
1465 min 1590 min
-300 o -300 o
Length (mm) Length (mm)
Figure 7.27 - Cross-sectional temperature profile over time along the

length of the central reinforcement bar (continued)
239
A full list of temperatures over various ages can be seen in Table 7.2.
Table 7.2 - Temperature of Slab B at various ages
Time after Sensor number (from Chapter 3

II gauging (min) 1 2 3 4 5 6 7 8 9 10
240 19.6 20.1 20.5 21.0 21.4 20.1 20.7 20.9 21.1 21.3
250 21.5 24.9 27.1 27.0 27.2 21.9 27.0 26.9 26.9 26.6
260 24.9 30.6 33.8 34.1 34.0 24.7 31.5 32.2 32.8 32.3
270 29.5 36.6 40.6 41.5 41.1 28.1 35.4 37.0 38.7 37.9
280 34.3 42.5 47.1 48.3 48.0 31.5 39.7 31.8 44.0 43.2
290 39.1 48.5 53.8 55.1 54.8 34.7 44.6 47.0 49.4 48.6
299 43.3 53.9 60.0 61.2 60.9 37.5 48.9 51.6 54.3 53.5
320 49.0 55.2 58.9 60.8 61.1 40.0 45.3 48.7 52.2 52.6
380 52.2 57.0 59.3 60.4 60.8 40.9 45.3 48.3 51.4 52.4
1395 31.8 32.2 32.4 32.4 32.4 29.8 29.9 30.2 30.5 30.5
1465 30.7 31.3 31.5 31.6 31.6 29.2 29.6 29.8 30.0 30.1
1590 30.1 30.4 30.6 30.6 30.6 28.9 29.1 29.2 29.4 29.4
Sensor number
11 12 13 14 15 16 17 18 19 20
240 19.7 20.2 20.7 2J.1 21.4 19.7 20.7 19.7 20.6 19.9
250 22.1 25.8 27.2 27.4 27.3 2J.1 26.2 21.7 24.9 24.9
260 26.0 31.8 34.2 34.5 33.9 24.1 32.8 24.6 30.0 29.9
270 30.8 37.7 41.1 41.8 40.6 28.5 39.8 28.2 35.2 34.9
280 35.5 43.3 47.2 48.3 47.0 33.3 46.2 31.8 39.9 40.3
290 40.0 49.1 53.5 54.8 53.5 37.8 52.6 35.3 44.7 46.4
299 44.1 54.3 59.3 60.7 59.2 41.8 58.4 38.5 49.0 51.2
320 47.9 53.2 56.9 59.0 58.6 48.2 58.4 42.0 49.1 52.2
380 49.9 54.1 56.6 58.2 58.3 50.9 58.4 45.0 50.9 54.6
1395 31.6 31.9 32.1 32.2 32.1 31.7 32.3 31.0 31.5 31.7
1465 30.8 31.1 31.4 31.4 31.4 30.8 31.4 30.4 30.8 30.7
1590 29.9 30.2 30.3 30.4 30.3 30.0 30.5 29.5 29.9 30.0
7.5 Conclusions
1. Both the Schmidt hammer and core failure load was found to be acceptable in
estimating compressive strength for electrically cured slabs.
2. Cracks were found to be present on the surface of the concrete slab which used a
depth of 20mm as a temperature datum during heating but no cracks were found on
the slab that used a depth of 75mm as a temperature datum.
240
3. Similar compressive strength trends were found for both the electrically cured slabs
and cubes with higher initial compressive strength and reduced compressive strength
at later ages (when compared to normally cured samples).
4. It was found from the Schmidt hammer results that the regions directly above the
reinforcement bars attained a higher initial compressive strength but a lower long-
term compressive strength when compared to the regions between the reinforcement
bars. This was due to higher temperatures in the immediate vicinity of the
reinforcement bars as the electric field density increased around the electrodes.
5. The results show that compressive strength was similar along the entire length of the
concrete slab and that a slight increase was apparent in the central region at early
ages. This was due to the increased temperature of the central region because of the
thermal insulating properties of concrete.
6. Core derived compressive strength showed higher strengths parallel to the

reinforcement bars in the outer regions at 56 days than the inner regions. This was
due to the reduced thermomechanical damage in the outer regions from relatively
low temperatures incurred during heating.
7. Differences were found between the strength predicted from the Schmidt hammer
and core failure load. This was found to be due to the Schmidt hammer being a
surface measurement and the cores being taken from within the concrete where
different temperature histories occurred during heating which resulted in different
values of compressive strength.
8. Porosity, density and water absorption were found to be approximately similar over
the entire surface of the slab.
241
9. The temperature immediately above the reinforcement bars was found to be greater
than the temperature in the region between the reinforcement bars. This was due to
electric field density increases around the reinforcement bars which increased the
heating rate in this area.
10. The temperature of the reinforcement bar was approximately the same as the
surrounding concrete at the depth at which it was placed. This implied that the
reinforcement bars did not heat up directly, but instead acted as pure electrodes.
This was likely to be due to the relatively low resistance of reinforcement bars which
precluded ohmic heating.
11. The maximum temperatures were seen to be present between the reinforcement bars
and the plate electrode. This was found to be due to current passing only through
this region between the electrodes. The area above this region was only heated by
thermal conduction and as such attained relatively high thermal gradients between
the reinforcement bar and the ambient exposed surface.
242
Chapter 8 - SUMMARY OF CONCLUSIONS AND FUTURE
RESEARCH
8.1 Introduction
This chapter provides a summary of conclusions from the results and discussions of
Chapters 4-7 and also suggestions for future research in the field of Direct Electric
Curing (DEC).
8.2 Summary of Conclusions

The experimental and theoretical results presented in this thesis prove that mortar and
concrete can be successfully produced using DEC with a wide range of strength and
durability properties using alternative heating regimes, mixes and reinforcement. In
general, it was found that the initial strength properties of both mortar and concrete were
substantially improved using DEC. The samples show higher initial strengths but
slightly lower long-term strengths.
8.2.1 DEC of mortar
It was found that both the short-term (3 days) and long-term (28 days) compressive
strengths could be favourably altered using DEC by the use of variable delay periods,
rates of temperature rise, maximum temperatures and isothermal periods.
Over a range of W/C ratios electrically cured mortar samples achieved substantially
higher initial (3 days) compressive strengths than normally cured samples and the
proportional effectiveness of electrically cured mortar specimens was found to be
increased at higher WIC ratios.
Long-term (28 day) compressive strength was not adversely affected by DEC in the
samples tested although the rate of compressive strength development at later ages (>28
days) was reduced when compared to normally cured samples. The reduced rate of
243
compressive strength development at later ages in mortar was likely to be due to
differential expansion of the aggregates and the cement matrix incurred by heating.
The initial drop in resistance during heating was due to water absorption by aggregates
which reduced the effective path length of the electrical current. Resistance can be used
as a guide to the amount of water in mortar with higher W/C ratios giving lower
resistances at all ages but there appeared to be little correlation between 3 days mortar
compressive strength and resistance. Mortar behaved as a dielectric having a reduced
apparent resistance at higher temperatures. This effect became negligible after
approximately 3 days when the sample temperature approached ambient levels.
Maturity was found to be a good indication of compressive strength when DEC was
used. The value of maturity during cooling was significantly higher than the maturity
increase measured during the initial heating of samples. This meant that accurate
monitoring of temperature after the initial heating was concluded was required in order
to estimate compressive strength.
The durability of mortar samples (as measured by the ISAT) did not appear to be
adversely effected by DEC. This was likely to be due to capillary pore refinement
during heating.
8.2.2 DEC of concrete
It was found that, for a range of different concrete mixes, short-term compressive
strengths could be substantially improved by using DEC and varying delay periods, rates
of temperature rise and maximum temperatures.
The compressive strength of normally cured concrete at 24 hours was attained using
DEC in 4 hours and compressive strengths of up to 26N/mm2 were readily achievable
244
using DEC after 1 day. At 1 day, all samples of varying W/C ratios cured using DEC
attained a higher compressive strength than normally cured samples.
Short delay periods (1 hour) tended to give higher initial compressive strengths, but
reduced long-term compressive strengths when compared to longer delay periods (4
hours). The reduction in long-term compressive strengths was likely to be due to
increased microcracking in the short delay period samples which reduce long-term
strength. The greatest long-term compressive strength (~28 days) was found to occur
when heating commenced at 3 hours after mixing. This was due to high early age
compressive strengths combining with limited microstructural damage to produce
relatively high long-term compressive strengths.
For equivalent mixes and heating regimes, a higher maximum temperature produced
higher I day compressive strengths but lower long-term (~28 day) compressive
strengths. This was likely to be due to increased thermomechanical damage and
moisture loss at the higher temperatures.
The splitting tensile strength and flexural tensile strength of concrete cured using DEC
was found to follow a similar trend as compressive strength of DEC cubes with higher
initial temperatures causing increased initial tensile strength but lower long-term tensile
strength.
Concrete was found to behave as a dielectric with a lower apparent resistance at higher
temperatures. This implied that concrete temperature should be closely monitored if
resistance is to be used in predicting the progress of hydration. The resistance at I day
was not found to be proportional to compressive strength but at 3 days a logarithmic
relationship was found to exist.
It was found that, when compared to normally cured samples, the DEC splitting tensile
strength had a lower percentage tensile strength increase at 1 day and a higher
245
percentage tensile strength loss at 28 days than the equivalently cured DEC compressive
strength tests. This was likely to be due to microcracks acting as stress concentrators
during tensile loading.
Saul's maturity values were found to be more accurate in determining the compressive
strength of normally cured samples than electrically cured samples. Maturity values
were also found to underestimate the compressive strength of electrically cured samples
at early ages and overestimate compressive strength at later ages.
Durability of concrete (as measured by ISAT) did not appear to be adversely effected
when using DEC. This was likely to be due to the refinement of capillary pores during
heating. It was also found that ISA 10 values were related to compressive strength.
Porosity, density and water absorption was found to be similar for both electrically and
normally cured samples.
Temperature gradients increased during heating and also the maximum temperature
within the electrically heated (150x150x150mm) cubes and (400xlOOxlOOmm) beams
were in the central region with minimum temperatures immediately adjacent to the
exposed concrete surface.
Using a scanning electron microscope (SEM), both normally and electrically cured
samples were found to exhibit a dense microstructure associated with C-S-H formation
at all ages, but electrically cured samples tended to have a less uniform structure with
coarser pores being more prevalent. It was also found that higher concentrations and
lengths of ettringite were found initially in electrically cured samples.
246
8.2.3 DEC of concrete containing admixtures and additives
Using combinations of DEC and admixtures or additives for a similar mix design could
achieve a large variability of the short-term (1 day) and long-term (28 days) compressive
strengths.
The maximum temperature of electrically cured air entrained concrete was found to
influence compressive strength at all ages, with higher temperatures producing lower
compressive strength. This was likely to be due to increased microcracking around the
entrained air bubbles at the higher temperatures. The effectiveness of electrical curing
in improving compressive strength on mixes containing air entrainment was found to be
reduced when compared to similarly designed mixes not containing air entrainment.
The reduction of compressive strength at 28 days for electrically cured air entrained
concrete when compared to normally cured concrete was found to be dependent on the
percentage of air entrainment, with higher levels of air entrained producing greater
reductions in compressive strength.
Normally cured samples containing a superplasticiser attained a higher compressive

strength at all ages than a similar mix without superplasticiser. The electrically cured
concrete samples containing superplasticiser cured after a delay of 3 hours acted in an
identical manner to a heated mix without superplasticiser with a delay period of I or 2
hours. This was due to the superplasticiser mix having a greater period before the initial
set which effectively reduced the delay period. It was also found that both normally and
electrically cured concrete containing a retarder behaved in a similar manner to concrete
containing a superplasticiser.
It was found that normally cured samples containing an accelerator attained a higher
compressive strength at all ages than a similar mix without accelerator. The electrically
cured mix with accelerator cured after a delay of 3 hours acted in a similar manner to a
heated mix without accelerator with a delay period of 4 hours. This was due to the mix
247
with accelerator having a reduced period before initial set which effectively reduced the
delay period.
It was found that with cement replacement by PFA the effectiveness of electrical curing
in improving compressive strength was reduced at 1 day. This was likely to be due to
the concrete containing PFA having an extended period before initial set, thus being less
able to resist the thermo mechanical stresses incurred during heating.
There exists a relationship between ultrasonic pulse velocity (UPV), density, percentage
of air entrainment and compressive strength for both normally and electrically cured
concrete samples.
8.2.4 DEC of concrete containing reinforcement
The compressive strength of the slabs tested after accelerated curing by DEC was found
to be dependent on the method of testing (Schmidt hammer or cores) and the placement
of the temperature datum.
Similar compressive strength trends were found for both the electrically cured slabs and
cubes with higher initial compressive strength and reduced compressive strength at later
ages (when compared to normally cured samples).
Both the Schmidt hammer and core failure load was found to be acceptable in
estimating compressive strength for electrically cured slabs although differences were
found between the strength predicted from the Schmidt hammer and core failure load.
This was found to be due to the Schmidt hammer being a surface measurement and the
cores being taken from within the concrete where different temperature histories
occurred during heating which resulted in different values of compressive strength.
248
It was found from the Schmidt hammer results that the regions directly above the
reinforcement bars attained a higher initial compressive strength but a lower long-term
compressive strength when compared to the regions between the reinforcement bars.
This was due to higher temperatures in the immediate vicinity of the reinforcement bars
as the electric field density increased around the electrodes.
The results showed that compressive strength was similar along the entire length of the
concrete slab and that a slight increase was apparent in the central region at early ages.
This was due to the increased temperature of the central region because of the thermal
insulating properties of concrete.
Porosity, density and water absorption were found to be approximately similar over the
entire surface of the slab.
The temperature immediately above the reinforcement bars was found to be greater than
the temperature in the region between the reinforcement bars. This was due to electric
field density increases around the reinforcement bars which increased the heating rate in
this area. The maximum temperatures were seen to be present between the
reinforcement bars and the plate electrode. This was found to be due to current passing
only through this region between the electrodes.
8.3 Recommendations for Future Research

Further research will be required in order to produce workable standards for the use of
DEC in producing concrete. Until this is achieved and a recognition of the benefits of
DEC over more traditional forms of accelerated curing become more widespread, the
DEC of concrete will remain a highly specialised and relatively unused procedure.
The durability of concrete cured using DEC could be further investigated by testing of
specimens for resistance to frost damage. This would involve freezing and thawing
specimens for a specified number of cycles and then comparing results with other forms
249
of curing. DEC undoubtedly has major benefits compared to normal curing In
resistance to frost attack due to the heating of concrete in the process. However,
investigations need to completed to study the effect of edges and comers present in most
cementitious products.
The resistance of DEC cured specimens should also be tested regarding other aspects of
durability including sulphate attack, chloride ingress and carbonation. This can be
achieved by utilising a number of standard testing procedures to again compare DEC
results with other forms of curing.
A special area for concern regarding heat accelerated hydration is delayed ettringite
formation. Long-term testing is required to ensure DEC is not especially susceptible to
this form of durability problem.
The concerns considering shrinkage of specimens cured using DEC need to be assessed
and compared to other forms of accelerated and normal curing. This is due to the
relatively large amount of moisture lost with DEC. The effects of shrinkage may be
reduced by the increased early age strength but the long term effects of DEC need to be
investigated.
The long term creep of specimens cured using DEC also needs to be considered. This
will be greatly influenced by the relatively low gain in strength after the initial curing as
well as the usual factors affecting the creep of concrete.
The further development of DEC into practical applications such as precast products
requires research. This should concentrate on the innate variability of specimens due to
reinforcement and geometry affecting the electrical field density and heat generation.
The investigation of DEC as relating to in-situ situations needs to be considered. Health

and safety issues and practicality of use are of importance due to the inherently
250
dangerous nature of passmg large currents through specimens in the relatively
uncontrolled environment of a construction site. This could be achieved by small scale
field trials and the analysis of previous in-situ procedures used in different countries.
A detailed investigation of the economical aspects of DEC in real situations is required

to understand the feasibility of using DEC as compared to other accelerating curing
methods. This will require an analysis of capital expenditure, maintenance, training,
cost of use, resale value and productivity levels.
A computer modelling of DEC as applied to differing geometrical forms and mixes by

the use of finite element analysis (FEA) is necessary. This will require the modelling
specifically of the changing electrical field density and resistance. The resistance is
affected by time, properties of the mix constituents, varying temperature and moisture
content. The electrical field density is affected by the placement of reinforcement, the
geometry of the specimen, the resistance and the required temperature rise.
251
REFERENCES
ABBASI, A.F. and AL-TAYYIB, A.J. 'Effect of hot weather on modulus of rupture
and splitting tensile strength of concrete'. Cement and Concrete Research, Vol. 15,
1985. pp. 233-244.
ACI COMMITTEE 308. 'Recommended practice for hot weather concreting'.

American Concrete Institute, Detroit, ACI 305R-77, 1992.
AL-GAHTANI, H.J., ABBASI, A.G.F. and AL-AMOUDI, O.S.B. 'Concrete

mixture design for hot weather: experimental and statistical analyses'. Magazine of
Concrete Research, Vol. 50, No.2, 1998, pp. 95-105.
AMERICAN SOCIETY FOR THE TESING OF MATERIALS. ASTM CI074-

1998. 'Standard practice for estimating concrete strength by the maturity method'.
AMERICAN SOCIETY FOR THE TESING OF MATERIALS. ASTM C1202-

1997. 'Standard test method for electrical indication of concretes ability to resist
chloride ion penetration'.
BAJZA, A. and ROUSEKOV A, I. 'Effect of heat treatment conditions on the pore

structure of cement mortar'. Cement and Concrete Research, Vol. 13, 1983, pp. 747-
750.
BAKKER, R.F .M. 'Permeability of blended cement structures'. Proceedings of the 1st
International Conference on the use of Fly Ash, Silica Fume, Slag and Other Mineral
By-products in Concrete, ACI Special Publication SP-79, 1983, pp. 589-605.
252
BAMFORTH, P.B. 'Performance of concrete structures'. Proceedings from the
National Seminar on the Use ofPFA in Construction, February 1992.
BAZANT, Z.P. and KAPLAN, M.F. 'Concrete at high temperatures'. Longman,

1996.
BERGE, O. 'Improving the properties of hot-mixed concrete using retarding

admixtures'. ACI Journal, Vol. 73, No. 29,1976, pp. 394-398.
BERHANE, Z. 'The behaviour of concrete in hot climates'. Materials and Structures,

Vol. 25,1992, pp. 157-162.
BERRY, E.E., and MALHOTRA, V.M. 'Fly ash for use in concrete - a critical
review', American Concrete Institute Journal, Vol. 77, No.8, 1980, pp. 59-73.
BREDENKAMP, S., KRUGER, D. and BREDENKAMP, G.L. 'Direct electric

curing of concrete'. Magazine of Concrete Research, Vol. 45, No. 162, March 1993, pp.
71-74.
BRITISH STANDARDS INSTITUTION. BS 12: 1996. 'Portland cement'.
BRITISH STANDARDS INSTITUTION. BS EN 196: Part 2: 1995. 'Chemical

analysis of cement'.
BRITISH STANDARDS INSTITUTION. BS 451: Part 2: 1994. 'Method of testing

fly ash: determination of fineness by wet sieving'.
BRITISH STANDARDS INSTITUTION. BS 812: Part 2: 1995. 'Method for

determination of physical properties'.
253
determination of particle size distribution'.

determination of aggregate impact value (AN)'.
BRITISH STANDARDS INSTITUTION. BS 882: 1983. 'Aggregates from natural

sources for concrete'.
BRITISH STANDARDS INSTITUTION. BS 890: 1995. 'Specification for building

limes'.
BRITISH STANDARDS INSTITUTION. BS 1377: Part 2: 1990. 'Methods of

testing for soils for civil engineering purposes. Classification tests'.

testing hardened concrete for other than strength'.

determination of slump'.

determination ofVebe time'.

determination of air content offresh concrete'.

making test cubes from fresh concrete' .
254
BRITISH STANDARDS INSTITUTION. BS 1881: Part 111: 1983. 'Method of
normal curing of test specimens'.

accelerated curing of test cubes'.
BRITISH STANDARDS INSTITUTION. BS 1881: Part 114: 1983. 'Methods for

determination of density of hardened concrete' .
BRITISH STANDARDS INSTITUTION. BS 1881: Part 115: 1986. 'Specifications

for compression testing machines for concrete' .

determination of compressive strength of concrete cubes' .

determination of tensile splitting strength' .

determination of flexural strength'.
BRITISH STANDARDS INSTITUTION. BS 1881: Part 125: 1986. 'Methods for

mixing and sampling of fresh concrete' .
BRITISH STANDARDS INSTITUTION. BS 1881: Part 202: 1986.

'Recommendations for surface hardness testing by rebound hammer'.
BRITISH STANDARDS INSTITUTION. BS 1881: Part 203: 1986.

'Recommendations for measurement of velocity of ultrasonic pulses in concrete'.
255
BRITISH STANDARDS INSTITUTION. BS 3148: 1980. 'Water for making
concrete'.

testing cement'.
BUNGEY, J. H. 'The testing of concrete in structures'. Surrey University Press, 1982.
CARINO, N.J., LEW H.S. and VOLZ, C.K. 'Early age temperature effects on
strength prediction by the maturity method'. ACI Journal, Vol. 80, No. 10, 1983, pp.
93-101.
CATHER, B. 'Curing: the true story?'. Magazine of Concrete Research, Vol. 46, No.
168, March 1994, pp. 157-161.
CZERNIN, W. 'Cement chemistry and physics for civil engineers'. Georg Godwin
Ltd, 1980.
DANISH MINISTRY OF TRANSPORT. 'The road directorate: the concrete of the

Faroe bridges'. 1985, pp. 156.
DAVIS, R.E., CARLSON, R.W., KELLY, J.W. and DAVIS, H.E. 'Properties of
cements and concrete containing fly ash'. Proceedings ACI 33,1937, pp. 577-612.
DE BORST, R., VAN DEN BOOGARD, A.H., VAN DEN BOGERT, P.A.J. and
SLUYS, L.J. 'Computational issues in time-dependent deformation and fracture of
concrete'. Proceedings s" RILEM International Symposium on Creep and Shrinkage of
Concrete, 1993.
256
DEPARTMENT OF THE ENVIRONMENT. 'Design of normal strength concrete
mixes'. Building Research Establishment, 1992.
DEVENNY, A.S and KHALAF, F.M. 'New tests for brick porosity and water
absorption'. 5th International Masonry Conference, British Masonry Society, No.8,
1998, pp. 13-15.
DHIR, R.K. 'Pulverized-fuel ash'. Cement Replacement Materials, Surrey University

Press, 1986, pp. 197-255.
DHIR, R.K. and BYARS, E.A. 'PFA concrete: near surface absorption properties'.
Magazine of Concrete Research, Vol. 43, No. 157, December 1991, pp. 219-232.
DIAMOND, S. 'Processes and problems'. Cement and Concrete Composites, Vol. 18,
No.3, 1996, pp. 205-215.
DOUBLE, D.D. 'New developments in understanding the chemistry of cement

hydration'. Proceedings of a Royal Society discussion meeting, 16-17 February 1983,
pp.53-66.
EL HUSSEIN, H.M. and ABD EL HALIM, A.O. 'Differential thermal expansion

and contraction: a mechanistic approach to adhesion in asphalt concrete'. Canadian
Journal of Civil Engineering, Vol. 20, 1993, pp. 366-373.
EREN, 0., BROOKS, J.J. and CELIK, T. 'Setting times of fly ash and slag-cement
concretes as affected by curing temperature'. Cement, Concrete and Aggregates,
CCAGDP, Vol. 17,No.1, 1995,pp.II-17.
FARRAR, J.R. 'Electrically conductive concrete'. GEC Journal of Science and

Technology, Vol. 45, No. I, 1978, pp. 45-48.
257
FEDERA TION INTERNATIONALE DE LA PRECONTRAINTE. 'Acceleration
of concrete hardening by thermal curing: guide to good practice'. 1982, pp. 1-16.
GA YNOR, R.D., MEININGER, R.C. and KHAN, T. S. 'Effect of temperature and

delivery time on concrete proportions'. Temperature Effects on Concrete, ASTM STP
858, 1985, pp. 68-87.
GEBLER, S.H. and KLIEGER, P. 'Effect of fly ash on physical properties of

concrete'. Proceedings 2nd International Conference on Fly Ash, Silica Fume, Slag and
Natural Pozzolans in Concrete, American Concrete Institute, ACI SP 91, Vol. 1, 1986,
pp. I-50.
GIBBON, G.J. and BALLIM, Y. 'Determination of the thermal conductivity of

concrete during early stages of hydration'. Magazine of Concrete Research, Vol. 50,
No.3, 1998, pp. 229-235.
GOTO, S. and ROY, D.M. 'The effect of W/C ratio and curing temperature on the
permeability of hardened cement paste'. Cement and Concrete Research, Vol. 11, 1987,
pp. 575-579.
HANSON, J.A. 'Optimum steam-curing procedures in precasting plants'. ACI

Journal, Vol. 60, No.1, 1963, pp. 75-100.
HIGGINSON, E.C. 'Effect of steam curing on the important properties of concrete'.

ACI Journal, Vol. 58, No.3, 1961, pp. 281-198.
HIGHW AYS AGENCY. 'Manual of contract documents for highway works, Vol. I,
Specification for Highway Works'. Series 1700, 1998, pp. 14-16.
258
HO, D.W.S. and LEWIS, R.K. 'Carbonation of concrete incorporating fly ash or a
chemical admixture', Proceedings of the 151International Conference on the Use of Fly
Ash, Silica Fume, Slag and Other Mineral By-Products in Concrete, Montebello,
Canada, ACI Special Publication, SP-79, 1983, pp. 333-346.
HOBBS, D.W. 'Influence of PFA and GGBS upon expansion caused by the alkali-
silica reaction'. Magazine of Concrete Research, Vol. 34,1982, pp. 83-93.
HUTCHINSON, R.G., CHANG, J.T., JENNINGS, H.M. and BRODWIN, M.E.

'Thermal acceleration of Portland cement mortars with microwave energy'. Cement and
Concrete Research, Vol. 21, 1991, pp. 795-799.
IDORN, G.M. 'Concrete curing technology'. Concrete and Structures, United States
Strategic Highway Research Program, Thomas Telford, 1990, pp. 87-102.
IVANOV, Y. and ZACHARIEVA, S. 'Influence of fly ash on rheology of fresh

concrete'. Proceedings of the International Symposium on the Use of PFA in Concrete,
Leeds, April 1982, pp. 133-141.
JACKSON, N. and DHIR. R.K. 'Civil engineering materials'. Macmillan Education

Ltd, 1988.
JALALI, S. and ABYANEH, M. Y. 'Prediction of final concrete strength in hot

climates'. Magazine of Concrete Research, Vol. 47, No. 173, 1995, pp. 291-297.
JA WED, I. and SKALNY, J. 'Hardened mortar and concrete with fly ash'. Fly Ash in
Concrete: Properties and Performance, RILEM Report 7, E & FN Spon, 1991, pp. 61-62.
JONASSON, J.E. 'Early strength growth in concrete - preliminary test results

concerning hardening at elevated temperatures'. Proceedings 3rd International Rilem
259
Symposium on Winter Concreting, Espoo, Finland, November 25-27, 1985, pp. 249-
254.
KAFRY, 1.0. 'Direct electric curing of concrete: basic design'. Whittle Publishing
Services, 1993.
KANAZA W A, K., YAMADA, K. and SOGO, S. 'Properties of low-heat generating

concrete containing large volumes of blast-furnace slag and fly ash'. Proceedings 4th
International Conference on Fly Ash, Silica Fume, Slag and Natural Pozzolans III
Concrete, American Concrete Institute, ACI SP 132, Vol. 1, 1992, pp. 97-118.
KHALAF, F.M. and WILSON, J.G. 'Electrical properties of freshly mixed concrete'.
Journal of Materials in Civil Engineering, Vol. 11, No.3, 1999, pp. 242-248.
KIRKBRIDE, T. 'Review of accelerated cunng procedures'. Precast Concrete,

February, 1971, pp. 93-106.
KJELLSEN, K.O. 'Heat curing and post-heat curing regimes of high-performance

concrete: influence on microstructure and C-S-H composition'. Cement and Concrete
Research, Vol. 26, No.2, 1996, pp. 295-307.
KUKKO, H. 'Use of heated fresh concrete'. Proceedings 3rd International Rilem

Symposium on Winter Concreting, Espoo, Finland, November 25-27, 1985, pp. 147-
159.
LA WRENCE, CD. 'Laboratory studies of concrete expansion arising from delayed

ettringite formation'. British Cement Association, Publication C/16, 1993.
260
LEUNG, C.K.Y. and PHEERAPHON, T. 'Microwave curing of Portland cement
concrete: experimental results and feasibility for practical applications'. Construction
and Building Materials, Vol. 9, No.2, 1995, pp. 67-73.
LEUNG, C.K.Y. and PHEERAPHON, T. 'Very high early strength of microwaved

cured concrete'. Cement and Concrete Research, Vol. 25, No.1, 1994, pp. 136-146.
LEVITT, M. 'Non destructive testing of concrete by initial surface absorption'.

Proceedings of the Symposium on Non-Destructive Testing of Concrete and Timber,
London, June 1969.
LEVITT, M. 'Precast concrete materials, manufacture, properties and usage'. Applied

Science Publishers, 1982.
MAILVAGANAM, N.P., BHAGRATH, R.S. and SHAW, K.L. 'Effect of

admixtures on Portland cement incorporating blast-furnace slag and fly ash'.
Proceedings l " International Conference on Fly Ash, Silica Fume, Slag and Natural
Pozzolans in Concrete, American Concrete Institute, ACI SP 79, Vol. 1, 1983, pp. 519-
535.
MALHOTRA, V.M. 'Supplementary cementing materials for concrete'. CANMET,

1987, pp. 139-142.
MARCIALIS, A., MASSIDDA, L. and SANNA, U. 'Low temperature steam curing

hydration of lime-fly-ash compacts'. Cement and Concrete Research, Vol. 13, 1983, pp.
887-894.
MCCARTER, W.J. and. AFSHAR, A.B. 'Diagnostic monitoring of the physio-

chemical processes in hydrating cement paste'. Cement, Concrete and Aggregates, Vol.
7, No.2, 1985, pp. 57-68.
261
MCCARTER, W.J., CHRISP, T.M. and STARRS, G. 'The early hydration of alkali-
activated slag: developments in monitoring techniques'. Cement and Concrete
Composites, No. 21,1999, pp. 277-283.
MCCARTER, W.J. and. CURRAN, P.N. 'The electrical response characteristics of

setting cement paste'. Magazine of Concrete Research, Vol. 36, No. 126, March 1984,
pp.42-49.
MEHTA, P.K. 'Pozzolanic and cementitious byproducts as mineral admixtures for

concrete - a critical review'. Proceedings of the Is1 International Conference on the use
of Fly Ash, Silica Fume, Slag and Other Mineral By-products in Concrete, ACI Special
Publication SP-79, 1983, pp. 1-46.
MONFORE, G.E. 'Electrical resistivity of concrete'. Journal PCA, Research and

Development Laboratories, May 1968, pp. 35-48.
NAIK, T .R. 'Maturity functions for concrete cured during winter conditions'.
Temperature Effects of Concrete, ASTM STP 858,1985, pp. 107-117.
NASSER, K.W. and AL-MANASEER, A. 'Comparison of nondestructive testers of

hardened concrete'. ACI Materials Journal, Vol. 84, No. 38, pp. 374-380.
NEVILLE, A.M. 'Properties of concrete'. Longman, 1988.
NIU, Y.Z., TU, CiL, and ZHANG, S.W. 'Modelling oftherrnomechanical damage of
early-age concrete'. Journal of Structural Engineering, Vol. 121, No.4, 1995, pp. 717-
726.
262
ODLER, I. and CHEN, Y. 'Effect of cement composition on the expansion of heat
cured cement pastes'. Cement and Concrete Research, Vol. 25, No.4, 1995, pp. 853-
862.
ORCHARD, D.F. 'Concrete technology'. Applied Science Publishers, 1973.
ORCHARD, D.F. and BARNETT, A.M. 'The electrocuring of concrete test

specimens'. JMSLA, Vol. 6, No.3, 1971, pp. 617-642.
OWENS, P.L. 'Effect of temperature rise and fall on the strength and permeability of
concrete made with and without fly ash'. Temperature Effects on Concrete, ASTM STP
858,1985, pp. 134-149.
PARCEVAUAX, P. 'Pore size distribution of Portland cement slurries at very early

stages of hydration (influence of curing temperature and pressure)'. Cement and
Concrete Research, Vol. 14, 1984, pp. 419-430.
PARKER JR., F. and SHOEMAKER, W. 'Rapid-setting pavement patching

materials'. New Pavement Materials, ASCE, 1987, pp. 81-95
PATEL, H.H., BLAND, CH. and POOLE, A.B. 'The microstructure of concretes
cured at elevated temperatures'. Cement and Concrete Research, Vol. 25, No.3, 1995,
pp. 485-490.
PENTTALA, V. 'Heating of concrete by infra-red radiation'. Proceedings 3rd

International Rilem Symposium on Winter Concreting, Espoo, Finland, November 25-
27, 1985, pp. 65-82.
263
PFEIFER, D.W. and LANDGREN, R. 'Energy-efficient accelerated curing of
concrete for plant-produced prestressed concrete'. PCI Journal, March 1982, pp. 94-
107.
PINK, A. 'Comments on the Rilem recommendations on concreteing in cold weather

with particular reference to current practice in the united kingdom'. Proceedings 3rd
International Rilem Symposium on Winter Concreting, Espoo, Finland, November 25-
27,1985,pp.29-39.
PRICE, R.E. 'Recommended practice for atmospheric pressure steam curing of

concrete'. ACI Committee 517, ACI Journal, August 1969, pp. 629-646.
RAITT, G. 'Physics in engineering 16 - 19'. Cambridge University Press, 1992.
RAMACHANDRAN, V.S., FELDMAN, R.F. and BEAUDOIN, J.J. 'Concrete

science'. Heydon & Son Ltd, 1981.
RICHARDSON, J.G. 'Precast concrete production'. Cement and Concrete

Association, 1973.
RIXOM, R.M. 'Concrete admixtures: use and applications'. Construction Press, 1977.
SAUL, A.G.A. 'Principles underlying the steam curing of concrete at atmospheric

pressure'. Magazine of concrete research, Vol. 8, No.6, 1951, pp. 127-140.
SCRIVENER, K.L. 'The effect of heat treatment on inner product C-S-H'. Cement
and Concrete Research, Vol. 22, 1992, pp. 1224-1226.
SENBETTA, E. and SCHOLER, C.F. 'A new approach for testing concrete curing
efficiency'. ACI Journal, Vol. 81, No. 12, 1984, pp. 82-86.
264
SHAY AN, A. and IVANUSEC, I. 'An experimental clarification of the association of
delayed ettringite formation with alkali-aggregate reaction'. Cement and Concrete
Composites, Vol. 18, Issue 3, 1996, pp. 161-170.
SHAYAN, A. and QUICK, G.W. 'Microscopic features of cracked and uncracked

railway sleepers'. ACI Materials Journal, Vol. 89, 1992, pp. 348-361.
SKALNY, J., JAWED, I. and TAYLOR, H.F.W. 'Studies on hydration of cement -

recent developments'. World Cement Technology, September 1978, pp. 183-195.
SOHN, D. and JOHNSON, D.L. 'Microwave curing effects on the 28-day strength of
cementitious materials'. Cement and Concrete Research, Vol. 29, 1999, pp. 241-247.
SRIRA VINDRARAJAH, R. and SWAMY, R.N. 'Development of a conductivity

probe to monitor setting time and moisture movement in concrete'. Cement, Concrete
and Aggregate, Vol. 4, No.2, 1982, pp. 73-80.
STANLEY, C. 'Concrete through the ages'. British Cement Association, 1999, pp. 2-
15.
TAMAS, D.F. 'Electrical conductivity of cement pastes'. Cement and Concrete

Research, Vol. 12, 1982, pp. 115-120.
TURENNE, R.G. 'Curing of concrete in shelters'. Proceedings 3rd International Rilem

Symposium on Winter Concreting, Espoo, Finland, November 25-27, 1985, pp. 53-63.
WADHWA, S.S., SRIVASTAVA, L.K., GAUTAM, D.K. and CHANDRA, D.

'Direct electric curing of in situ concrete'. Building Research and Practice, Journal of
the CIB, March / April, 1987.
265
WEA VER, J. and SADGROVE, B.M. 'Striking times of formwork'. Construction
Industry Research and Information Association, CIRIA Report 36, October 1971.
WILSON, J.G. and GUPTA, N.K. 'Assessment of concrete by measurement of its

electrical resistance'. Building Research and Information, Vol. 24, No.4, 1996.
XUENQUAN, W., JIANBGO, D. and MINGSHU, T. 'Microwave curing technique

in concrete manufacture'. Cement and Concrete Research, Vol. 17, 1987, pp. 205-210.
YANG, R., LAWRENCE, C.D., LYNSDALE, C.J. and SHARP, J.H. 'Delayed
ettringite formation in heat-cured Portland cement mortars'. Cement and Concrete
Research, Vol. 29,1999, pp. 17-25.
ZHAO, T.J., ZHOU, Z.H., ZHU, J.Q. AND FENG N.Q. 'An alternating test method
for concrete permeability'. Cement and Concrete Research, Vol. 28, 1998, pp. 7-12.
266
APPENDIX A
Mix Design Sheets
267
("oncrtlir .il dnilDror-. Job tide Nonnal rrax
Refereatt
S... e I.... or caleubltion Valuel
II Charactensuc strength 30 N/mm J
at -'2"-"__ davs
I, StMldard deVIlbon Flg.3 ____________________________

-------------~-------------.,
NlmmJ or no data __ -'-_ N/mm~
13 Margin Cl __ 1.6_4_ x 8 _ _......;1",3 __ :-.I/mm'
or
Specified
14 Target me_ Slrength Cl __ 3_0_ • 13 _ ___ ..;.43,-_ N/nv'"

IS Cemmlt)"})e Specfied OPC 'SRp(' 'RHPC'
16 Aagregale type coarse Crushed '-""hod
Aggregate type fine Crushed '-""hod
11 Free-watft/certJenl rene Table 2. Fig 4 o.SS
I" Maxt",,,m jrre- SfWcifi~d 06 } Use the lower value
055
water/cement ratIo
21 SIUfllIor vebe nrre Specified SI"",, mm or Vebe nme
22 MaXlmwn IlUP'egale mu Specified 20 nUll
23 Free-wetet c.ontml TableJ 190 190 kg/m'
31 Cement content CJ 190 055 345 kg/m

,
3! .\la.tm"." cement
contf!"t ___________ kg/m'
33 ."""",,"" cement
content __ .:;300::::....
__ kg/m'
uselllf£J2
useJ3If>3.1 345 kg/m'
34 Modified free-waterJcemerlt nrbo
41 RellD ve d.lUlty of 27 known/ .........

1l1li"(1&10ISSD)
42 C()I1(rete dlllSlly FI8 S 2440 kg/m'
43 TotailggTtpte cauent (4 2440 345 190 1905 kg/m'
51 GnKbng of fine Peredage pusmg 06 mm sieve 4S "

aggregate
52 Proporbon of fine Flg.6 39
aggrepte
S3 Fane aggregate content 19O' 0)9 143 kg/m
I
Coarse a&IVe.. te content 1905 143 1162 kg/m I
Quoatid .. Fine aares.te Coane -aRret_ie (ka)

(1<&) lOmm 10 mm 40 mm
per III\ (10 nearest 5 Lg) ~ I90 _ 143

per Inal n." of 003 m' 10 35 51 22,29
hen. In italiCSare optional IImbog values that may be speofied (see Secbon 7)
I N/tm12 - I Ms/m2 - I MPa(see footnoc.eto Sectton 3)

OPe ... ordmary Portland cemren. SkP( - swphlte-resistm8 Porthmd Cet18lt; R.HPC - rapd-hardenmg PonllWld CetTlCl1t
Relative denlny .. speCIfic waVIly (see fOOCllOleto Para!i 4) SSD - based 00 a saturated smace-dry basis
268
Coac:relr .i,dnila fOnl Job tide HIgh strength
S!aJje .... Refncace

or caicul.dGII Valucs
-i
II Charectensnc stlWlglh 5peafied ___ -'2:;;8__
50 N/nm~ at days
12 Statdard deVl300n FIg. 3

Proportim dofOC1>v.
____________________________
------------~--------------%
N/1TI11': or no data ___ .::..__N/mm"
13 Marlll" Cl (k= ~I~.6~4 __ __ 1_64_ x 8 _ __ ---""'-3 __ N/mm~
or
Specified ______ N/T1un'
I. Target ~ strenJlh Cl __ 5_0_ __ __13 _

15 Cetll!I1l type Specified OPC SR PC RHPC I
16 Aggregate type coarse Crushed _
Aggregate type fine Crushed_

17 Free-water/cenan rlbo Table 2. Ftg 4 04
18 Marimllm .fr~~- Specifitd 06 } Use the lower value 04
walfT/c~m'n' mho
2 I SllIII1' or vebe • .., Speafied 51_ 10 - 30 mm or Vebe time
22 MaXlITUn aagregace SlU 5peafied 20 mm

23 I
Free-water content Table 3 190 190 kg/m
31 Cement content Cl 100 04 475 kg/Oil

32 Akrflmllm cr"'~nl
content Specified ___________ kg/m
33 Mil"",,,,,, c,m~nt
conwnt ____ 3"'OO;.;;_ kglm'
usellif£3.2
use)3 if», 1
34 MoQ.fied free-wllericement rMlO
41 Relltl ve dawty of
agreglle (SSO)
2.7 kno"""_
42 Ccecreee denllty FI85 2440 kg/m
43 Talal aureJa1e ceaen C4 2440 475 190 1775 kg/m'
SI Grado8offine Percentage passms 0,6 mm SIeve 45 -"0
aggreJMl:
52 Propornoo of fin. 32 ·/t
agro ....
I
53 Fme a",epe content 1775 032 568 k8lm
, 4 Coane awepte cord." 1775 568 1207 kg/m'
Quaoucin Fine aureaate Co_ne ltIarea-Ce (q,)

0<&) to mm 20 mm 40 mm
per m (to nearest 5 kl)

I 568 1207
per mallnx of 003 rn' 1425 57 1704 36,21
hems In uahes are optionat hmnng values tNl may be specified (see Secbon 7)
1 NiiTWlL!'" I Mlhnl" I MPa (see footnote to Section 3)

OPr - on:knary P\)nland etnwll. SRPC .. sulphatf'olftllbng Portl .. d eemmt; RHPC' ., r.pid-h.rdening Portland eemenl
Relative denstty - speafle .... ~ty (1ft footnote to Para 54) SSD" based 00 a salurated SurfiCe-dry basis
269
Con(~1r mil dHicn'~
Job tide High slump
Stat:~ I-
Values
II ("haraC1~sne strength Specified JO N/mmJ at 2:0:8'--_ days

--{ Proportion defective
--------"-------_.'.
"
13
Standard
Margan
de\1atl.OIl _______________ N/mmJ or no data
__ -,,--_N/IllIll~
Cl (k~ __ ~IM~ _ __ 1_64_ x __ 8__ _---C';:..l __ Nzmrn'
Specified ____ N/nun'

I. TarBet me .... $trl!njth C2 __ 1_0_ ' __ 11__ __ .;:4;:..1
_Nillun.'
II Cement type Specified OPe 'SRI'(; 'RHI'(;
16 Aggregate type coarse Crushed _
A8iJeIPllel}'J>e fine Crushed ~

17 Free-wlter/Cert1l!flt ratio Table 2. Fig 4 0.55
18 Maxlm"mfru- __________ 0"'.6"- } Use the lower value 055
SfHqfitd
water cemem rauo
21 SIWJ1) or Yebe time Specified SI_ _____ ----'60=-..;1"'80"- mill or vebe time
" Ma\unum aggregate StU Specified 20

21 Free-water content Table 3 225 225 kg/me
31 Ceman content Cl 225 055

12 Ala:n"".,. crmem
cantt'" Spu/fied kg/m'
11 ,\ftn/m'M' cemeeu
cmft", Specified lOO kg/m'
use311f£12
use33if>31 409 kglm1
14 Modified Iree-weter/cerrent reno
41 Rei," Ye dslSlty of 2.7 known! assumed

agregate (SSD)
" 2 ( cecreee density Fig 5 2380 kg/m'
n Total aggregate conten C4 2380 409 lIS 1746 kg/m
51 Gndlngoffine Percentage passmg 0 6 mm sieve 41

agregale
52 Propomm of fine 4J
aggregate
Fme aggre-gate content 1746 043 751 kg/Ill
I
Coarse aggregate content 1746 751 995 kg/m

,
Cemeot W.lrr Coarle aurel_1t (k&)

(kc) (kc or L) IOmm lOmm 40mm
per m' (to neareat 5 kg) 751
per mal m\ of OUl rn' 1227 675 22.53
Items In lIahes are optIonal hmhng \'alurs thatllll)' be specified (see Secnoe 1)
I N/nn2 '" I MIJ.im2 - I MPI (see footnote to 5ectJon ))

OPe '" orchnary Ponl.,d cement. SRPC - sulphate--reststing Portl.,d cement. RHPC tepid-hardeang Portland cement $-
Relatlve denSity .~speafic gravity (see footno(e to Para 3 4) SSD - based on I saturated surface-df)' basis
270
Job tidr All entramcd (2~o)
R.rrre.te
II... .rtaltal.ci_
11 CharactenstJc strength Speafied JO
----{ P""",""on d.r.ct.ve

12 Standard deVlauon
IJ (k= ~l.M~ _
Speafied
14 Tatgd mean utength Cl

141 .4.uconIenI: ___ 2 0/,
1"' Z MocbfiC'd target mean strength 43 (1-0 05h -=- _
15 Cement type Specified OPC 'SRP(' '!YIP('
I. Agrepte type coarse Crushed _
Agrepte type fine Crushed _

17 Free-watcr/cemcn 1lItlo Tabie 2. Fig 4 0.43
18 Maxi",,.,,, jru. Spmfi<d 0.6 } Use the lower value
04J
'ttal,,,"c"'Nnl mho
21 Slwnp Of Yebe lime Specified Slump 10 - 3D mm or Vebe time
.- \b"unum aggregate size Specfied zu

lJ Free-water contC!fl( Table J 100 100 kg/m
I
31 Cement content Cl 190 043 442

,
kg/m
J2 .~"'fll,"C'''''nt
conl,nI Spmfi<d kg/m·'
)3 "fmJ",,,m c,,,,,nt
conl,nl Spuljwd 300 kg/m'
use),1 if£3.2
use 1.3 If:> 3,1 442 kg/l1l~
) 4 Mo(ilfied free-water/cement TlUO
41 Relative densityof' 2.7 known/ assumed

__ ,,(550)
42 C oncme denSI(," FigS 2440 ( 10 x x 27) 2386 kg/m
<3 Total aureaate contenl C4 2386 442 190 1754 kg/m
'1 Gradu~ of fine Peroentagc paulng 0 6 mm sieve 45 %
-,.
agrepte
>2 Proportion of fine JQ %
I
5J Fmc auregate cceteu 1754 0.39 684 kS:!m
I
C O.fU agregate COllltn( 1754 684 1070 kg/m
Cemenl W.ter Coarsr a&&rrlatr (kit:)

Ike) Ike" L) IOmm 1:0mm 40mna
per m 1 (to nearest
per tn.l nu" of

S kg)
OOJ m
. ~
J1 26
190
__
57
684
2052
Items 10 ltahcs .re optional hmlung values that may be specified (see Secnon 7)
I N/mm2 .. I MB/m2 - 1 MPII (SH footnote to Secacn 3)

OPe ,. on:hnary Portland cement: SRPC '" sulphate-reilltlng Portland cement, RHPC' ,. rapid-hardening Portland cement
Relative denslf)' .. specsfic vavwt)' (see fOOblote [0 PIIf1I 5 4) SSD - based on a saturated surfa~ry basIS
271
Job tid~ Au entrained (5%"
Rtf~rt.u
eruleala ... V.I~J
II Cbarsceensec strenglh Specified )0 N/mml at 28 days
---{ Propomon d.rect.ve ',.
12 Stand.ud devrancn Fig J N/nulll cr uo data N/mm-
13 Margan Cl Ik· ~I~M~ _ __ 1_04_ x 8 13 Nnnm"
Speafied N/mll\;
14 Tarxrt mean strt1l8lh __ 13

__
30 43 N/mm:
1'1 AIr CCln:en( __ 5__ %
I. 2 Modified ta'iet mean strength __ '_)_ 11-0055, -=-__ 50 N/mm l
15 Cemmttypc OPC~RHP('
1.6 Ag~te type coarse [rushed , __
AgregJte type fine Crushed ,............

17 F ree--wlter/cemenr. ranc Table 2, Fig 4 04)
18 ,\lnx"'''um fru- Sl"c'fi,d 06 } Use the lower value ou
...a"rlC"~m mho
II Slump or veee time Speafied Sh."" 10- )0 mm or Vebe lime
II Ma'Umum agg~te SIll: 10

I) Free-water content: Table 3 190 IQIl kg/m
,
31 Cemen( content IQIl 04) 442 kg/m

,
31 M,nt",,,,,, ('fWnI
con"nl kg/m
33 ,\Itn"""", r,,,,,,,,
cOt",m 300 kg/m
use 3 Ilf £ 3.2

use 3 3 if > ] I 4.2 kg/m
,
14 Modified free-wlter/cement rauo
41 RelluVf denSity of 21 known! ItHttIfl6d

_'.1550)
41 Coecreee dmJlt~' FI85 2440 ( 10 x x 21 ) 2305 kg/m
4J Taul agreg.ale content C. 2)05 442 IQIl 167) kg/m
51 Gradl,.. of fine Pe:n:entage passmg 0 6 mm sieve 41 0,'0
Igreple
52 Proportion of fine )0 %
5) Fine agregale content 1673 03Q 652 kg/m

,
5. Co.rse~ecOf"(mI 167) 651 1011 kg/m
,
Cement W.ter Co.rse IlIrr&ate (ke)

0<&) 0.1.' L) lOmm ZOmm 40mm
pet m 1 (10 oeerest ~ kg) 652 __ 1_02_1_

per tnal nux of 003 m' 1)26 57 JQ5b J063
Items In ltahcs art' optlONI hmltmg "aJues that may be specified (see Section 7)
I N/mmZ - I M8ImZ = I MPa (see foomote to SectIon J)
OPe' - ordinary Portland cement. SRPC" - sulphate-reSISting Portland cement, RHPC '"' rapld"hardenlng Portland cement
Relative densl1\ ... speafic 811\11\'(~footnote to Para 54) SSO - based on a saturated surface-dry basis
272
Appendix B page 273 onwards has not been
digitised at the request of the University.

Heritage - 2001 - Direct Electric Curing of Concrete and Mortars

Uploaded by

Copyright:

Available Formats

Heritage - 2001 - Direct Electric Curing of Concrete and Mortars

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Heritage - 2001 - Direct Electric Curing of Concrete and Mortars

Uploaded by

Copyright:

Available Formats

DIRECT ELECTRIC CURING OF MORTAR AND

laD Heritage B.EDg (HoDs), M.Se

A thesis submitted in partial fulfilment ofthe requirements of

School of the Built Environment

used to make cubes, slabs and beams.

A substantial section of reinforced concrete is required to gain representative results in

electrically curing reinforced concrete slabs. Compressive strength is difficult to

The microstructural development of concrete when subjected to accelerated curing is

In accordance with Napier University regulations governing the Degree of Doctor of

1. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Electrical curing of cement

2. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Thermal acceleration of

3. HERITAGE, I., KHALAF, F.M. and WILSON, J. 'Thermal acceleration of

4. HERITAGE, I. and KHALAF, F.M. 'Direct electric curing of concrete'. Proc.

and constructive discussions are much appreciated by the author.

him great support throughout his career. For Bailey.

Chapter 2 - LITERATURE REVIEW 7

2.5 Methods of Accelerated Curing 15

2.6 Electrical Properties of Concrete 26

2.7 Direct Electric Curing 38

2.8 Heating Cycles 40

2.8.2 Isothermal period 41

2.8.3 Cooling period 42

2.8.7 Thermomechanical affects 49

2.9 Effects of Accelerated Curing on Mortar and Concrete 50

2.10 Additives and Admixtures 67

2.10.2 Accelerating admixtures 69

2.10.3 Retarding admixtures 69

2.10.4 Air entraining admixtures 70

2.10.6 Pulverised fuel ash (PF A) 70

Chapter 3 - MATERIAL PROPERTIES AND TESTING PROCEDURES 74

3.2.2 Impact test 75

3.2.3 Density bottle 75

3.2.4 Speedy test 76

3.2.5 Vicat test 77

3.3 Constituent Material Properties 78

3.3.2 Fine aggregate 78

3.4 Mix Design and Casting 82

3.6 Hardened Concrete Testing Procedures 85

3.7 DEC Set-up 96

Chapter 4 - DIRECT ELECTRIC CURING OF MORTAR 105

4.10 Delay Period 118

Chapter 5 - DIRECT ELECTRIC CURING OF CONCRETE 122

S.3 Effect of Cover 123

S.18 Temperature Profiles 169

S.19 Scanning Electron Microscopy 177

5.19.2 Pore structure 178

5.19.3 Microcracking 179

5.20 Conclusions 186

Chapter 6 - DIRECT ELECTRIC CURING OF CONCRETE CONTAINING

Chapter 7 - DIRECT ELECTRIC CURING OF CONCRETE CONTAINING

7.2.2 Core failure load 212

7.2.3 Temperature history 214

7.3 Electrically Cured Slab A 215

7.3.2 Schmidt hammer 215

7.3.3 Core failure load 219

7.3.4 Porosity 222

7.3.5 Density 213

7.3.6 Water Absorption 225

7.3.7 Temperature profile 225