Sulfate Resistance and Properties of Portland- Limestone ... · Sulfate Resistance and Properties...
198
Sulfate Resistance and Properties of Portland- Limestone Cements By Amir Mohammad Ramezanianpour A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Civil Engineering University of Toronto © Copyright by Amir Mohammad Ramezanianpour (2012)
Transcript of Sulfate Resistance and Properties of Portland- Limestone ... · Sulfate Resistance and Properties...
Sulfate Resistance and Properties of
Portland- Limestone Cements
By
Amir Mohammad Ramezanianpour
A thesis submitted in conformity with the requirements for the degree of
Doctor of Philosophy
Department of Civil Engineering
University of Toronto
© Copyright by Amir Mohammad Ramezanianpour (2012)
ii
“Sulfate Resistance and Properties of
Portland-Limestone Cements”
By
Amir Mohammad Ramezanianpour
A thesis submitted in conformity with the requirements for the degree of
Doctor of Philosophy
Department of Civil Engineering
University of Toronto
March 2012
Abstract
Portland-limestone cements (PLC) have been used in practice for a considerable period of
time in several countries. In 2008, the CSA A3000 cements committee approved the
addition of a new class of cement with up to 15% interground limestone. The CSA A23.1
concrete committee also approved the use of PLC in concrete in 2009. However, to date,
due to uncertainty about the performance of Portland-limestone cements in sulfate
environments, their use has not been allowed in sulfate exposures.
iii
In this study, the sulfate resistance of five different Portland-limestone cements and their
combinations with various amounts of supplementary cementitious materials (SCMs)
were examined. Besides the standard tests performed at 23 °C, a modified version of the
ASTM C1012 test was developed in this study (adopted in 2010 as CSA A3004-B) and
used to investigate the possibility of thaumasite form of sulfate attack at 5 °C.
It was found for tests conducted at 23 °C that while 100% cement mixes deteriorated in
sulfate exposure due to conventional sulfate attack, partially replacing the Portland
cements and Portland-limestone cements with 30% or 50% slag was effective in making
the mixes highly sulfate-resistant. In sulfate exposure at 5 °C, all of the 100% cement
mortar bars failed the test and had completely disintegrated due to the formation of
thaumasite. Partially replacing cement with 30% slag was effective in controlling the
deterioration at 5 °C only for Portland cements and not Portland-limestone cements.
However, all the combinations of the cements with 50% slag were resistant to the
thaumasite form of sulfate attack.
In a parallel study, the hydration of Portland-limestone cements and the relationship
between strength and porosity of mortar samples were examined. The results of hydration
studies revealed that the limestone portion of Portland-limestone cements reacts with the
alumina phases and produces carboaluminates, which contributes to the strength. As the
limestone content of the cement increased, the shift in the optimum level of SCM
providing maximum strength and minimum porosity was attributed to the availability of
more alumina, which allowed more limestone to participate in the hydration reactions,
forming additional carboaluminate hydrates.
iv
Acknowledgements:
I would like to take this opportunity to thank all the people who have helped me
throughout my life. I thank Professor Douglas Hooton for accepting me as his graduate
student, and allowing me to pursue my studies at the wonderful environment of Concrete
Materials Laboratory at the department of Civil Engineering, University of Toronto. I
have definitely benefited a lot from his advice and assistance. I would also like to thank
Professor Panesar, Professor Pressnail, and Professor Peterson who kindly accepted to be
my committee members, and Professor Beaudoin as the external examiner.
I should also thank the people of Concrete Materials Laboratory, namely Dr.
Terry Ramlochan, Urszula Nytko and Olga Perabetova, for helping me to get familiar
with the equipments and experiments. I have enjoyed working in this great environment.
Indeed, the aid of my friends during the past five years deserves plenty of
appreciation. Despite its rises and falls, it has been a pleasure for me to accompany these
nice people.
I convey my sincere and special thanks to my dear parents who have served me
since the beginning of my life in this world. I owe a huge portion of my academic
knowledge to my father as he is also a professor in concrete materials. Moreover, if it was
not for the efforts of my mother, I would not have been in this state now. Also, I thank
my sister and all the rest of my relatives and friends who have supported me.
Finally, I express my genuine appreciation towards the Almighty, and also the one
for whom I am impatiently waiting.
v
Contents
I. Abstract .......................................................................................................................... ii II. Acknowledgements: .................................................................................................... iv
III. List of Abbreviations................................................................................................ vii IV. List of Tables............................................................................................................ viii V. List of Figures.............................................................................................................. ix 1. Introduction............................................................................................................... 1
1.1. Initiatives.................................................................................................................. 1 1.2. Portland-Limestone Cements................................................................................... 2
1.2.1. History............................................................................................................... 3 1.2.2. Current Situation............................................................................................... 7
1.3. Significance of Sulfate Attack ................................................................................. 8 1.4. Research Objectives............................................................................................... 10
2. Background ............................................................................................................. 12 2.1 Sulfate Attack.......................................................................................................... 12
2.1.1. Conventional Sulfate Attack ........................................................................... 13 2.1.1.1. Ettringite Formation................................................................................. 13 2.1.1.2. Gypsum Formation .................................................................................. 15
2.1.2. Thaumasite Sulfate Attack.............................................................................. 17 2.1.2.1. Mechanisms of Thaumasite Formation.................................................... 17 2.1.2.2. Thaumasite Deterioration......................................................................... 19
2.1.3. Mitigating Sulfate Attack................................................................................ 21 2.1.4. Tests for Sulfate Resistance ............................................................................ 23
2.2. Sulfate Resistance of Portland-Limestone Cements .............................................. 25 2.2.1. Conventional Sulfate Attack ........................................................................... 26 2.2.2. Thaumasite Sulfate Attack (TSA)................................................................... 36 2.2.3. Summary and Gaps ......................................................................................... 44
2.3. Portland-Limestone Cements and SCMs............................................................... 46 2.3.1. Hydration of Portland-Limestone Cements .................................................... 46 2.3.2. Compressive Strength and Porosity ................................................................ 49 2.3.3. Effect of SCMs ............................................................................................... 51
3. Experiments............................................................................................................. 54 3.1. Materials ................................................................................................................ 54 3.2. Tests on Sulfate Resistance.................................................................................... 56
3.2.1. Conventional Sulfate Attack ........................................................................... 57 3.2.2. Thaumasite Sulfate Attack.............................................................................. 59
3.3. Studying Thaumasite Sulfate Attack ..................................................................... 61 3.4. Compressive Strength, Porosity, and Hydration.................................................... 62
3.4.2. Compressive Strength ..................................................................................... 63 3.4.3. Porosity ........................................................................................................... 63 3.4.4. Hydration ........................................................................................................ 64
4. Results and Discussion............................................................................................ 65 4.1. Sulfate Resistance Tests......................................................................................... 65
4.1.1. Sulfate Attack at 23°C .................................................................................... 67
vi
4.1.1.1. 100% Cement at 23°C.............................................................................. 67 4.1.1.2. 70% Cement and 30% Slag at 23°C ........................................................ 74 4.1.1.3. 50% Cement and 50% Slag at 23°C ........................................................ 80 4.1.1.4. Overall Comparison ................................................................................. 87
4.1.2. Thaumasite Sulfate Attack.............................................................................. 90 4.1.2.1. 100% Cement at 5°C................................................................................ 90 4.1.2.2. 70% Cement and 30% Slag at 5°C .......................................................... 96 4.1.2.3. 50% Cement and 50% Slag at 5°C ........................................................ 103
4.2. Evolution of Thaumasite Sulfate Attack.............................................................. 108 4.2.1. Type I Mortar Bars at 5°C ............................................................................ 108 4.2.2. GUL22 Mortar Bars at 5°C........................................................................... 115 4.2.3. Thaumasite Sulfate Attack and Expansion ................................................... 121 4.2.4. Progression of Thaumasite Sulfate Attack.................................................... 123
4.3. Compressive Strength, Porosity, and Hydration.................................................. 126 4.3.1. Compressive Strength ................................................................................... 126
4.3.1.1. 100% Cement......................................................................................... 126 4.3.1.2. 85% Cement and 15% Slag.................................................................... 127 4.3.1.3. 70% Cement and 30% Slag.................................................................... 129 4.3.1.4. 50% Cement and 50% Slag.................................................................... 130 4.3.1.5. 90% Cement and 10% Metakaolin ........................................................ 132 4.3.1.6. Summary................................................................................................ 133
4.3.2. Porosity ......................................................................................................... 136 4.3.2.1. 100% Cement......................................................................................... 136 4.3.2.2. 85% Cement and 15% Slag.................................................................... 138 4.3.2.3. 70% Cement and 30% Slag.................................................................... 139 4.3.2.4. 50% Cement and 50% Slag.................................................................... 141 4.3.2.5. 90% Cement and 10% Metakaolin ........................................................ 142
4.3.3. Combination of Compressive Strength and Porosity.................................... 144 4.3.3.1. 100% Cement......................................................................................... 144 4.3.3.2. 85% Cement and 15% Slag.................................................................... 146 4.3.3.3. 70% Cement and 30% Slag.................................................................... 148 4.3.3.4. 50% Cement and 50% Slag.................................................................... 150 4.3.3.5. 90% Cement and 10% Metakaolin ........................................................ 152
4.3.4. Hydration ...................................................................................................... 154 4.3.4.1. 100% Cement......................................................................................... 154 4.3.4.2. 85% Cement and 15% Slag.................................................................... 155 4.3.4.3. 70% Cement and 30% Slag.................................................................... 157 4.3.4.4. 50% Cement and 50% Slag.................................................................... 159 4.3.4.5. 90% Cement and 10% Metakaolin ........................................................ 160 4.3.4.6. Comparison ............................................................................................ 162
5. Conclusions and Recommendations........................................................................ 166 5.1. Conclusions.......................................................................................................... 166 5.2. Recommendations................................................................................................ 169
6. References .................................................................................................................. 171
vii
List of Abbreviations
ASTM American Society for Testing and Materials
CH Calcium Hydroxide
CSA Canadian Standard Association
C-S-H Calcium Silicate Hydrate
FTR Fourier Transform Infrared Spectroscopy
GU General Use
GUL General Use with Limestone
MK Metakaolin
MIP Mercury Intrusion Porosimetry
OPC Ordinary Portland Cement
PLC Portland-Limestone Cement
SCM Supplementary Cementitious Materials
SEM Scanning Electron Microscopy
TSA Thaumasite Sulfate Attack
XRD X-Ray Diffraction
viii
List of Tables
Table 1.1: Types of Cement in EN 197-1 (2000) ............................................................... 5 Table 2.1: Sulfate Resistance of Cement with Limestone Replacements......................... 26 Table 2.2: Effect of 30% filler on time to failure of mortar bars in 5% Na2SO4 .............. 27 Table 2.3: ASTM C 452 Expansion of Mortars................................................................ 28 Table 2.4: Time to Expansion of ASTM C 1012 Mortar Bars ......................................... 28 Table 2.5: Sulfate Resistance of Limestone Mixtures ...................................................... 31 Table 3.1: Chemical compositions and physical characteristics of the used cements ...... 55 Table 3.2: Chemical compositions of the slag and metakaolin ........................................ 56 Table 3.3: The cements used and their correspondent limestone content......................... 62 Table 4.1: Current CSA A3004-08 requirements for sulfate resistance of PLC .............. 66 Table 4.2: ACI 318-08 requirements for sulfate resistance .............................................. 67 Table 4.3: Length change of CSA A3004-A mortar bars made with 100% cement ........ 68 Table 4.4: Length change of CSA A3004-A mortar bars made with 70% cement and 30% slag .................................................................................................................................... 75 Table 4.5: Length change of CSA A3004-A mortar bars made with 50% cement and 50% slag .................................................................................................................................... 81 Table 4.6: Length change of CSA A3004-B mortar bars made with 100% cement......... 90 Table 4.7: Length change of CSA A3004-B mortar bars made with 70% cement and 30% slag .................................................................................................................................... 97 Table 4.8: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag at 5°C....................................................................................................................... 103 Table 4.9: Compressive strength of mortar cubes made with 100% cement.................. 126 Table 4.10: Compressive strength of mortar cubes made with 85% cement and 15% slag......................................................................................................................................... 128 Table 4.11: Compressive strength of mortar cubes made with 70% cement and 30% slag......................................................................................................................................... 129 Table 4.12: Compressive strength of mortar cubes made with 50% cement and 50% slag......................................................................................................................................... 131 Table 4.13: Compressive strength of mortar cubes made with 90% cement and 10% MK......................................................................................................................................... 132 Table 4.14: Total Porosity of mortar cubes made with 100% cement............................ 137 Table 4.15: Total Porosity of mortar cubes made with 85% cement and 15% slag ....... 138 Table 4.16: Total Porosity of mortar cubes made with 70% cement and 30% slag ....... 140 Table 4.17: Total Porosity of mortar cubes made with 50% cement and 50% slag ....... 141 Table 4.18: Total Porosity of mortar cubes made with 90% cement and 10% MK ....... 143
ix
List of Figures
Figure 1.1: CEN Data on types of cement produced in Europe.......................................... 6 Figure 1.2 : the concentration of sulfate ions in groundwater in Southern Alberta.......... 10 Figure 2.1: Sulfate concentration that is required for a reaction of portlandite to gypsum or syngenite, calculations (line) and experimental measurements (dots). ........................ 16 Figure 2.2: Soft, white mush consisting of mainly thaumasite and gypsum. ................... 20 Figure 2.3: Sulfate expansion isobars (microstrain) for different w/c & limestone levels30 Figure 2.4: ASTM C 1012 Expansions of Cements with and without 20% Limestone ... 33 Figure 2.5: ASTM C 1012 Expansions of Cements with C3A Contents less than 5% ..... 34 Figure 2.6: Photos of concrete cubes: High C3A PLC (12% limestone) at 0.40 W/C...... 38 Figure 2.7: Photos of concrete cubes: High C3A PC at 0.50 W/C (Holton 2003) ............ 38 Figure 2.8: The section view of prisms stored in 1.8% MgSO4 at 5°C for 5 years .......... 40 Figure 2.9: XRD patterns of pastes exposed to Na2SO4 solution (left) and MgSO4 solution (right) for 360 days ........................................................................................................... 43 Figure 2.10: Expansion of prisms in Na2SO4 solutions .................................................... 44 Figure 2.11: Phase equilibrium of limestone .................................................................... 47 Figure 2.12: Relative compressive strength and porosity of PLC samples ...................... 51 Figure 2.13: Hypothetical effect of availability of more alumina on strength of PLC..... 52 Figure 2.14: 28 day EN 196 mortar strengths for Portland blended cement predicted from the thermodynamic model................................................................................................. 53 Figure 3.1: Comparator used to measure the length of mortar bars.................................. 58 Figure 3.2: Mortar samples stored in 5°C refrigerator...................................................... 60 Figure 4.1: Expansion of CSA A3004-A mortar bars made with 100% cement............. 68 Figure 4.2: Time to failure of CSA A3004-A mortar bars made with 100% cement...... 69 Figure 4.3: CSA A3004-A mortar bars made with 100% cement after 18 months of exposure at 23°C............................................................................................................... 70 Figure 4.4: XRD analysis of 100% Type I mortar after 18 months of exposure to sulfate solution at 23°C ................................................................................................................ 71 Figure 4.5: XRD analysis of 100% GU mortar after 18 months of exposure to sulfate solution at 23°C ................................................................................................................ 72 Figure 4.6: XRD analysis of 100% GUL11 mortar after 18 months of exposure to sulfate solution at 23°C ................................................................................................................ 72 Figure 4.7: XRD analysis of 100% GUL13 mortar after 18 months of exposure to sulfate solution at 23°C ................................................................................................................ 73 Figure 4.8: XRD analysis of 100% GUL22 mortar after 18 months of exposure to sulfate solution at 23°C ................................................................................................................ 73 Figure 4.9: Expansion of CSA A3004-A mortar bars made with 70% cement and 30% slag .................................................................................................................................... 75 Figure 4.10: Expansion of CSA A3004-A mortar bars made with 70% cement and 30% slag .................................................................................................................................... 76 Figure 4.11: CSA A3004-A mortar bars made with 70% cement and 30% slag after 24 months of exposure at 23°C.............................................................................................. 77 Figure 4.12: XRD analysis of 70% Type I and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 78
x
Figure 4.13: XRD analysis of 70% GU and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C..................................................................................................... 78 Figure 4.14: XRD analysis of 70% GUL11 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 79 Figure 4.15: XRD analysis of 70% GUL13 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 79 Figure 4.16: XRD analysis of 70% GUL22 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 80 Figure 4.17: Expansion of CSA A3004-A mortar bars made with 50% cement and 50% slag .................................................................................................................................... 82 Figure 4.18: Expansion of CSA A3004-A mortar bars made with 50% cement and 50% slag at 23°C....................................................................................................................... 82 Figure 4.19: CSA A3004-A mortar bars made with 50% cement and 50% slag after 24 months of exposure ........................................................................................................... 83 Figure 4.20: Comparison of expansion of mortar bars made with 30% and 50% slag at 23°C after 2 years ............................................................................................................. 84 Figure 4.21: XRD analysis of 50% Type I and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 85 Figure 4.22: XRD analysis of 50% GU and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C..................................................................................................... 85 Figure 4.23: XRD analysis of 50% GUL11 and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C................................................................................................. 86 Figure 4.24: XRD analysis of 50% GUL13 and 50% slag mortar after 2 years of exposure to sulfate solution 23°C..................................................................................................... 86 Figure 4.25: XRD analysis of 50% GUL22 and 50% slag mortar after 2 years of exposure to sulfate solution 23°C..................................................................................................... 87 Figure 4.26: 6 month expansion of CSA A3004-A mortar bars with different slag replacement levels............................................................................................................. 88 Figure 4.27: 12 month expansion of CSA A3004-A mortar bars with different slag replacement levels............................................................................................................. 88 Figure 4.28: 6 month expansion of CSA A3004-A mortar bars with different limestone contents ............................................................................................................................. 89 Figure 4.29: 12 month expansion of CSA A3004-A mortar bars with different limestone contents ............................................................................................................................. 89 Figure 4.30: Length change of CSA A3004-B mortar bars made with 100% cement ..... 91 Figure 4.31: Time to failure/disintegration of CSA A3004-B mortar bars made with 100% cement..................................................................................................................... 91 Figure 4.32: CSA A3004-B mortar bars made with 100% cement after disintegration at 5°C .................................................................................................................................... 92 Figure 4.33: XRD analysis of 100% Type I mortar after 365 days of exposure to sulfate solution at 5°C .................................................................................................................. 93 Figure 4.34: XRD analysis of 100% GU mortar after 330 days of exposure to sulfate solution at 5°C .................................................................................................................. 94 Figure 4.35: XRD analysis of 100% GUL11 mortar after 210 days of exposure to sulfate solution at 5°C .................................................................................................................. 94
xi
Figure 4.36: XRD analysis of 100% GUL13 mortar after 180 days of exposure to sulfate solution at 5°C .................................................................................................................. 95 Figure 4.37: XRD analysis of 100% GUL22 mortar after 90 days of exposure to sulfate solution at 5°C .................................................................................................................. 95 Figure 4.38: Length change of CSA A3004-B mortar bars made with 70% cement and 30% slag............................................................................................................................ 97 Figure 4.39: Expansion of CSA A3004-B mortar bars made with 70% cement and 30% slag at 5°C......................................................................................................................... 98 Figure 4.40: CSA A3004-B mortar bars made with 70% cement and 30% slag after 24 months of exposure at 5°C................................................................................................ 99 Figure 4.41: XRD analysis of 70% Type I and 30% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 100 Figure 4.42: XRD analysis of 70% GU and 30% slag mortar after 2 years of exposure at 5°C .................................................................................................................................. 101 Figure 4.43: XRD analysis of 70% GUL11 and 30% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 101 Figure 4.44: XRD analysis of 70% GUL13 and 30% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 102 Figure 4.45: XRD analysis of 70% GUL22 and 30% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 102 Figure 4.46: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag.......................................................................................................................... 104 Figure 4.47: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag.......................................................................................................................... 104 Figure 4.48: CSA A3004-B mortar bars made with 50% cement and 50% slag at 5°C after 24 months of exposure............................................................................................ 105 Figure 4.49: XRD analysis of 50% Type I and 50% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 106 Figure 4.50: XRD analysis of 50% GU and 50% slag mortar after 2 years of exposure at 5°C .................................................................................................................................. 106 Figure 4.51: XRD analysis of 50% GUL11 and 50% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 107 Figure 4.52: XRD analysis of 50% GUL13 and 50% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 107 Figure 4.53: XRD analysis of 50% GUL22 and 50% slag mortar after 2 years of exposure at 5°C .............................................................................................................................. 108 Figure 4.54: XRD analysis of 100% Type I mortar after demolding ............................. 109 Figure 4.55: XRD analysis of 100% Type I mortar after 28 days of exposure to Na2SO4 at 5°C .................................................................................................................................. 110 Figure 4.56: XRD analysis of 100% Type I mortar after 56 days of exposure to Na2SO4 at 5°C .................................................................................................................................. 110 Figure 4.57: XRD analysis of 100% Type I mortar after 91 days of exposure to Na2SO4 at 5°C .................................................................................................................................. 111 Figure 4.58: XRD analysis of 100% Type I mortar after 120 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 111
xii
Figure 4.59: XRD analysis of 100% Type I mortar after 150 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 112 Figure 4.60: XRD analysis of 100% Type I mortar after 180 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 112 Figure 4.61: XRD analysis of 100% Type I mortar after 210 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 113 Figure 4.62: XRD analysis of 100% Type I mortar after 240 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 114 Figure 4.63: XRD analysis of 100% Type I mortar after 270 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 114 Figure 4.64: XRD analysis of 100% Type I mortar after 300 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 115 Figure 4.65: XRD analysis of 100% GUL22 mortar after demolding............................ 116 Figure 4.66: XRD analysis of 100% GUL22 mortar after 28 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 117 Figure 4.67: XRD analysis of 100% GUL22 mortar after 56 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 117 Figure 4.68: XRD analysis of 100% GUL22 mortar after 91 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 118 Figure 4.69: XRD analysis of 100% GUL22 mortar after 120 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 119 Figure 4.70: XRD analysis of 100% GUL22 mortar after 150 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 119 Figure 4.71: XRD analysis of 100% GUL22 mortar after 180 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 120 Figure 4.72: XRD analysis of 100% GUL22 mortar after 210 days of exposure to Na2SO4 at 5°C .............................................................................................................................. 120 Figure 4.73: Expansion of type I and GUL22 mortar bars exposed to Na2SO4 at 5°C .. 122 Figure 4.74 : Mush accumulated at the bottom of container due to thaumasite sulfate attack ............................................................................................................................... 124 Figure 4.75 : Mortar pieces after extensive cracking due to thaumasite sulfate attack .. 124 Figure 4.76: Compressive strength of mortar cubes made with 100% cement .............. 127 Figure 4.77: Compressive strength of mortar cubes made with 85% cement and 15% slag......................................................................................................................................... 128 Figure 4.78: Compressive strength of mortar cubes made with 70% cement and 30% slag......................................................................................................................................... 130 Figure 4.79: Compressive strength of mortar cubes made with 50% cement and 50% slag......................................................................................................................................... 131 Figure 4.80: Compressive strength of mortar cubes made with 90% cement and 10% MK......................................................................................................................................... 133 Figure 4.81: 28 day compressive strength of mortar cubes made with different SCMs. 134 Figure 4.82: 56 day compressive strength of mortar cubes made with different SCMs. 135 Figure 4.83: 28 day compressive strength of mortar cubes with various slag replacement levels ............................................................................................................................... 135 Figure 4.84: 56 day compressive strength of mortar cubes with various slag replacement levels ............................................................................................................................... 136
xiii
Figure 4.85: Total porosity of mortar cubes made with 100% cement........................... 137 Figure 4.86: Total porosity of mortar cubes made with 85% cement and 15% slag ...... 139 Figure 4.87: Total porosity of mortar cubes made with 70% cement and 30% slag ...... 140 Figure 4.88: Total porosity of mortar cubes made with 50% cement and 50% slag ...... 142 Figure 4.89: Total porosity of mortar cubes made with 90% cement and 10% MK ...... 143 Figure 4.90: Relative strength and porosity of mortar cubes made with 100% cement at 28 days ............................................................................................................................ 145 Figure 4.91: Relative strength and porosity of mortar cubes made with 100% cement at 56 days ............................................................................................................................ 145 Figure 4.92: Relative strength and porosity of mortar cubes made with 85% cement and 15% slag at 28 days......................................................................................................... 147 Figure 4.93: Relative strength and porosity of mortar cubes made with 85% cement and 15% slag at 56 days......................................................................................................... 147 Figure 4.94: Relative strength and porosity of mortar cubes made with 70% cement and 30% slag at 28 days......................................................................................................... 149 Figure 4.95: Relative strength and porosity of mortar cubes made with 70% cement and 30% slag at 56 days......................................................................................................... 149 Figure 4.96: Relative strength and porosity of mortar cubes made with 50% cement and 50% slag at 28 days......................................................................................................... 151 Figure 4.97: Relative strength and porosity of mortar cubes made with 50% cement and 50% slag at 56 days......................................................................................................... 151 Figure 4.98: Relative strength and porosity of mortar cubes made with 90% cement and 10% MK at 28 days......................................................................................................... 153 Figure 4.99: Relative strength and porosity of mortar cubes made with 90% cement and 10% MK at 56 days......................................................................................................... 153 Figure 4.100: XRD analysis of 100% Type I pastes at various ages.............................. 154 Figure 4.101: XRD analysis of 100% GUL13 pastes at various ages ............................ 155 Figure 4.102: XRD analysis of 85%Type I and 15% slag pastes at various ages .......... 156 Figure 4.103: XRD analysis of 85%GUL13 and 15% slag pastes at various ages......... 157 Figure 4.104: XRD analysis of 70%Type I and 30% slag pastes at various ages .......... 158 Figure 4.105: XRD analysis of 70%GUL13 and 30% slag pastes at various ages......... 158 Figure 4.106: XRD analysis of 50%Type I and 50% slag pastes at various ages .......... 159 Figure 4.107: XRD analysis of 50%GUL13 and 50% slag pastes at various ages......... 160 Figure 4 4.108: XRD analysis of 90%Type I and 10% metakaolin pastes at various ages......................................................................................................................................... 161 Figure 4.109: XRD analysis of 90%GUL13 and 10% metakaolin pastes at various ages......................................................................................................................................... 162 Figure 4.110 : XRD analysis of 100%Type I and GUL13 pastes at 28 days ................. 163 Figure 4.111: XRD analysis of 85%Type I and GUL13 with 15% slag pastes at 28 days Figure 4.112 : XRD analysis of 70%Type I and GUL13 pastes with 30% slag at 28 days......................................................................................................................................... 164 Figure 4.113: XRD analysis of 50%Type I and GUL13 pastes with 50% slag at 28 days Figure 4.114 : XRD analysis of 90%Type I and GUL13 pastes with 10% metakaolin at 28 days ................................................................................................................................. 165
1
1. Introduction
1.1. Initiatives
Concrete is the second most consumed material in the world after water. It is
estimated that approximately 10 km3 of concrete is required annually, which corresponds
to about 1.5 m3 per person (Sharp et al, 2010). Ordinary concrete typically contains about
12% cement. Currently, Portland cement is the most common type of cement used to
bond concrete in many parts of the world. Recent statistics indicate that about 3 billion
tons of cement is produced every year (Cembureau, 2010).
The main component of Portland cement is ground clinker, which mainly consists
of calcium silicates, with some aluminum- and iron-containing phases. Clinker is
manufactured by mixing limestone (calcium source), clay (silica and alumina sources)
and iron ore in a rotary kiln and heating the mix to over 1450 °C. This high temperature,
which is usually reached by burning fossil fuels, leads to chemical reactions that
transform raw materials into clinker. This results in large emission of greenhouse gases,
especially carbon dioxide, both from the burning of fossil fuels and from decalcification
of limestone.
It is often stated that the production of 1 ton of cement results in emission of 0.8
ton of CO2 (Lounis and Diagle, 2010). In fact, estimations show that the manufacture of
Portland cement is responsible for between 5 to 8% of global CO2 emissions (Sharp et al,
2010). In fact, production of cement is expected to further increase because of the
exponential growth rates in developing countries, such as China and India, which are the
major cement producers and consumers in the world. Therefore, it is imperative that steps
be taken in order to reduce the CO2 released into the atmosphere. This involves either
incorporating new environmentally friendly manufacturing technologies or finding
substitute materials to replace a major part of Portland cement for use in concrete
industry (Malhotra, 2000). Due to the limitations involved in reducing CO2 emissions
from alternative raw materials and fuels or by improving kiln efficiency, the first option
seems to be not practical. Hence, probably the most effective means of achieving
2
significant reductions in CO2 emissions lies in the replacement of Portland cement by
other suitable materials, or alternatively by reducing the clinker component of Portland
cement.
Replacement materials that react with calcium hydroxide are commonly known as
“Supplementary Cementitious Materials (SCMs)”. They include fly ash, ground
granulated blast furnace slag (GGBFS), pozzolans, silica fume, metakaolin, etc. These
replacement materials can be added separately to the concrete, allowing a reduction in the
cement content of concrete, or used to replace the clinker in blended/composite cements.
Blended cements, i.e. cements comprising of supplementary cementitious
materials such as ground granulated blast furnace slag, fly ash, pozzolans, and silica
fume, are being produced by many cement manufacturers, and are used extensively in the
concrete sector. According to Mehta (2007), the use of supplementary cementitious
materials has increased from 10% in 1990 to about 15% in 2005, and it is viable to
increase this number to about 50% in 2020. SCMs are beneficial not only in the sense
that they contribute to sustainability and reduction in CO2 emissions, but also due to the
potential ability of these materials to enhance the properties and performance of concrete.
In addition to that, some of the SCMs such as slag and fly ash are byproducts of other
industries, and their use could help preserve non-renewable resources.
1.2. Portland-Limestone Cements
One of the materials that has been introduced to Portland and blended cements as
a constituent is limestone or calcium carbonate (CaCO3). This has led to the production
of Portland-limestone cement, i.e. cements that have been interground with limestone.
Most Portland cement specifications allow the use of up to 5% limestone. Beyond that,
Portland-limestone cements are categorized based on the percentage of limestone added
to the cement. Portland-limestone cements consisting of limestone from 5% up to about
40% are being produced and used in various countries around the world, with the most
commonly used cement type in Europe being CEM II/A composite cement with 5-20%
limestone. Also, different standards have stated specifications with regards to the amount
3
of limestone used in Portland-limestone cements. In 2008, CSA A3001 adopted a new
class of Portland-limestone cements with up to 15% interground limestone.
Perhaps the main advantage of producing Portland-limestone cement is its
contribution to sustainable development. By introducing limestone into cement, the total
volume of cement would increase, or in other words, the amount of clinker required to
produce a certain amount of cement would decrease. This would result in a substantial
amount of energy saving in the production of cement as the consumption of natural raw
materials and the fuel needed for production of clinker would be reduced. Moreover, it
would contribute to sustainable development due to the reduction in greenhouse gas
emissions, mostly CO2 and NOx, involved in the pyroprocessing of clinker. On this basis,
the future world production of Portland-limestone cement is expected to increase.
Nevertheless, it should be noted that all the aforementioned benefits can only be achieved
provided that Portland-limestone cement has similar performance characteristics to
Portland cement, and has no adverse effects on the properties of concrete.
The properties of Portland-limestone cements have been the subject of numerous
studies. Researchers have studied the effect of using Portland-limestone cement with
various limestone contents on fresh properties, mechanical properties, and durability of
concrete. Overall, the data found in the literature seems to be inconsistent in some
specific areas, especially in cases where the limestone content of Portland-limestone
cement is greater than 5%. Data reported in the literature is apparently affected by the
quality and particle size distribution of the limestone used. Also whether the limestone
was interground, blended, or added at the mixer seems to have an influence on the results.
Hence, it is important that these factors be taken into consideration when interpreting the
data.
1.2.1. History
The use of Portland-limestone cements has been in practice for a considerable
period of time in several countries. It seems that European countries have been the
leaders in producing and using Portland-limestone cement. According to Schmidt (1992),
4
Heidelberg Cement, the main cement producer in Germany, used to produce an energy-
saving cement which contained 20% interground limestone and was used for special
applications as early as 1965. In 1979, a new standard introduced in France permitted the
incorporation of up to 35% of slag, fly ash, natural or artificial pozzolans, and limestone
in a new type of cement called CPJ (Moir 2003). In this new standard, four different
strength classes were introduced with upper and lower strength limits, which encouraged
the use of limestone to control the strength development of the lower strength class
cements while maintaining an appropriate level of cement fineness (Kelham, 1998).
Later on, a specific cement designated as PKZ which consisted of 85+/-5%
clinker and 15+/-5% limestone was specified in the 1987 draft of the European standard
EN 197 (Schmidt 1992). This type of Portland-limestone cement, along with other types
of blended cements, was reported to be commonly used throughout Germany by 1990. In
addition to that, the British Standard BS 7583 allowed use of up to 20% limestone in
cement in the United Kingdom in 1992 (Schmidt 1992). However, the use of it did not
become so popular until the concrete standard BS 5328 was amended in 1997 to include
its use in concrete (Kelham, 1998) Nonetheless, the standard prohibited the use Portland-
limestone cements in conditions of freezing/thawing and exposure to sulfates.
In the current version of the European standard EN 197-1 (2000), all of the 27
common types of cement recognized by the standard are allowed to contain up to 5%
minor additional constituents (MAC), the most typical of which are limestone and cement
raw meal. Moreover, four of the designated 27 types correspond to Portland-limestone
cements which allow higher amounts of limestone in two replacement level ranges,
namely CEM II/A-L and CEM II/A-LL with 6 to 20% limestone, as well as CEM II/B-L
and CEM II/B-LL with 21 to 35% limestone in addition to the 5% MAC. It should be
noted that based on different qualities of the limestone used, the Portland-limestone
cements are designated as –L and –LL. Type LL restricts the total organic carbon (TOC)
to less than 0.20% by mass while Type L restricts the TOC to 0.50% by mass. However,
for both L and LL designations, the calcium carbonate content is greater than 75% and
the clay content is less than 1.20g/100g. Adopted from a study (Sprung and Siebel, 1991),
these purity requirements are selected to minimize the risk of poor performance of non-
5
air-entrained concrete in freezing/thawing exposures. Table 1 shows the cement types
approved by EN 197-1.
Table 1.1: Types of Cement in EN 197-1 (2000)
Composition (percentage by massa) Main constituents
Clinker Blast
furnace slag
Silica fume Pozzolana Fly Ash
Burnt shale
Limestone
Minor addition
al constitu
ents Main Type
Notation of the 27 products (types of common cement)
K
S
Db
Natural
P
Natural Calcined
Q
siliceous
V
Calcare-ous
W
L
LL
0-5
CEM I Portland Cement
CEM I 95-100 - - - - - - - - - 0-5
CEM II/A-S 80-94 6-20 - - - - - - - - 0-5 Portland-slag cement CEM II/B-S 65-79 21.35 - - - - - - - - 0-5
Portland-silica fume
cement CEM II/A-D 90-94 - 6-10 - - - - - - - 0-5
CEM II/A-P 80-94 - - 6-20 - - - - - - 0-5 CEM II/B-O 65-79 - - 21-35 - - - - - - 0-5 CEM II/A-Q 80-94 - - - 6-20 - - - - - 0-5
Portland-pozzolanac
ement CEM II/B-Q 65-79 - - - 21-35 - - - - - 0-5 CEM II/A-V 80-94 - - - - 6-20 - - - - 0-5 CEM IIB-V 65-90 - - - - 21-35 - - - - 0-5
CEM II/A-W 80-94 - - - - - 6-20 - - - 0-5 Portland-fly ash cement
CEM II/B-W 65-90 - - - - - 21-35 - - - 0-5
CIM II/A-T 80-94 - - - - - - 6-20 - - 0-5 Portland-burnt shale
cement CEM II/B0T 65-79 - - - - - - 21-35 - - 0-5 CEM II./A-L 80-94 - - - - - - - 6-20 - 0-5 CEM II/B-L 65-79 - - - - - - - 21-35 - 0-5
CEM II/A-LL 80-94 - - - - - - - - 60-20 0-5
Portland-limestone cement
CEM II/B-LL 65-79 - - - - - - - - 21-35 0-5
CEM II/SA-M 80-94 �---------- 11-35 ---------� 0-5
CEM II
Portland composite cementc CEM IIB-M 65-79 �---------- 11-35 ---------� 0-5
CEM III/A 35-64 36-65 - - - - - - - - 0-5 CEM III/B 20-34 66-80 - - - - - - - - 0-5
CEM III
Blast furnace cement CEM III/C 5-19 81-95 - - - - - - - - 0-5
CEM IV/A 65-89 �-------- 11-35 -------� - - - 0-5 CEM IV
Pozzolanic cementc CEM IV/B 45-64 �--------- 11-35 -------� - - - 0-5
CEM V/A 40-64 18-30 - �----- 18-30 -----� - - - - 0-5 CEM V
Compositecementc CEM V/B 20-38 31-50 - �----- 31-50 -----� - - - - 0-5
a) The value in the table refers to the sum of the main and minor additional constituents. b) The proportion of silica fume is limited to 10%. c) In Portland-composite cements CEM II/A-M and CEM II/B-M and in composite cements CEM IV/A and CEM IV/B and in composite cements CEM V/A
and CEM V/B the main constituents other than clinker shall be declared by designation of the cement (for example see clause 8)
According to the Cement Standards of the World (CEMBUREAU 1991), more
than 25 countries allow the use of 1% to 5% limestone in their P (“Portland”) cements.
Also, many countries allow up to 35% replacement in PB (“Portland composite”)
cements. According to Hawkins et al. (2005), several countries have modified their
standards to permit limestone in some amount, including Australia, Italy, New Zealand,
6
and the United Kingdom since 1991. Taken from the latest edition of Cement Standards
of the World, Figure 1 represents how the production of Portland-limestone cement has
been increasing over time in recent years in Europe. The graph shows that the use of
CEM II limestone cements has grown from 15% in 1999 to 31.4% in 2004 and became
the single largest type of cement produced in Europe. As well, most CEM II composite
cements contain limestone.
35.4
4.8
3.7
5.4
15.0
16.8
5.6
4.03.2
6.1
34.2
5.4
2.9
6.2
18.9
16.8
5.5
5.0
3.41.7
33.7
4.22.1
5.7
24.6
14.5
6.5
51.52.1
32.1
6.8
2.7
7.0
24.0
9.6
4.8
9.5
1.91.8
31.6
5.91.2
6.9
24.5
14.3
5.1
5.4
2.92.1
27.5
4.51.4
7.4
31.4
12.5
5.6
5.7
1.82.2
0
10
20
30
40
50
60
70
80
90
100
Cem
en
t Typ
es in
Eu
rop
e (%
)
1999 2000 2001 2002 2003 2004
Others
CEM V - Composite Cement
CEM IV - Pozzolanic
CEM III - Blast furnace slag
CEM II - Portland-composite
CEM II - Portland-limestone
CEM II - Portland-fly ash
CEM II - Portland-pozzolana
CEM II - Portland-slag
CEM I - Portland
Figure 1.1: CEN Data on types of cement produced in Europe
Portland cements with some level of interground limestone have been in use in
North American countries as well. However, compared to Europe, the use of Portland-
limestone cements in North America is relatively new and limited. In the US, ASTM C
150 allowed up to 5% limestone to be used in all types Portland cements in 2004. In
Canada, the use of up to 5% ground limestone in Type 10 Portland cement (Type GU in
the new standard) has been permitted in the Canadian cement standard CSA A5 (now
7
A3001) since 1983. In 1998, the CSA standard allowed the use of up to 5% limestone in
all different types of Portland cement, and this was reaffirmed in 2003. It was not until
2006 that the concept of allowing higher levels of limestone in cements was raised at the
CSA meeting (Hooton et al, 2010). Prior to any modifications to the standard, a
comprehensive literature review was undertaken to review the use of Portland-limestone
cements (Hooton et al, 2007). Besides pointing out that limestone cements had been used
in European countries since the 1960s and that they are now the most commonly used
cements in the European Union, the authors concluded that the technical data supported
the addition of up to 15% limestone. In 2008, the CSA A3000 cements committee
approved the addition of a new class of cement, designated as Portland-Limestone
Cements (PLC), with up to 15% interground limestone. The CSA A23.1 concrete
committee also approved the use of PLC in concrete in 2009, and it was included in the
National Building Code of Canada in 2010 as well as several provincial building codes.
1.2.2. Current Situation
As mentioned before, there are four classes of composite cements with limestone
filler in the European Standard, namely CEM II/A-L or-LL (6-20% limestone) and
CEMII/B-L or -LL (21-35% limestone). It is interesting that except in Sweden and Italy,
none of these types of cement are allowed to be specified where sulfate resistance is
required (CEN, 2003). In Sweden, testing for sulfate resistance must be performed, and in
Italy, there are certain restrictions on the C3A content of the clinker and the severity of
sulfate environment. This fact reflects the concern over performance of Portland-
limestone cements in sulfate exposures.
The literature review on Portland-limestone cements (Hooton et al, 2007)
suggested that the literature was conflicting with respect to use of Portland-limestone
cements in sulfate exposures. They pointed out inconsistencies in the trends found by
different researchers as to whether limestone improved or worsened sulfate resistance of
cement, either for conventional sulfate attack or in regards with thaumasite related
deterioration. Moreover, data on the performance of Portland-limestone cements when
used in conjunction with SCMs is limited. The recommendation was that more work
8
needs to be done in the Canadian context (e.g. at CSA exposure levels and sulfate types)
on the performance of Portland-limestone cements at both 5°C and 23°C. This should
include use of plain Portland-limestone cement and its combination with levels of SCMs
currently known to provide good sulfate resistance as well as on CSA MS and HS
cements.
Due to this uncertainty about the performance of Portland-limestone cements in
sulfate environments, their use was not allowed in sulfate exposures in the 2008 revision
until further research had been conducted. By that time, the cement companies had
started trial grinds and testing the properties of Portland-limestone cement had been
initiated. However, the concern over the sulfate resistance of Portland-limestone cements,
especially in regards with thaumasite sulfate attack, became the driving force behind
several research studies. Based on the result of this research, including the present study,
the 2010 version of the CSA A3001 standard stated that Portland-limestone cements may
be used in sulfate environments provided that they are combined with the specified
minimum percentages of supplementary cementing materials and tested for sulfate
resistance at both 5°C and 23°C. According to the standard, Type MSLb (moderate-
sulfate resistant Portland-limestone) and HSLb (high-sulfate resistant Portland-limestone)
cements shall contain a minimum of 25% Type F fly ash or 40% slag or 15% metakaolin
or a combination of 5% Type SF silica fume with 25% slag or a combination of 5% Type
SF silica fume with 20% Type F fly ash However, these changes have not been adopted
by CSA A23.1 yet.
1.3. Significance of Sulfate Attack
Portland cement concrete is susceptible to attack by aqueous solutions of sulfate
salts present in some soils and groundwaters. Concrete elements such as footings,
foundation walls, retaining walls, piers, piles, culverts, pipes and surface slabs may be
exposed to attack by sulfates in certain soils and groundwaters. The occurrence and rate
of such attack depends on various factors including the concentration of sulfate ions, the
presence of water, the composition of the cement, and concrete permeability.
Progression of sulfate attack in concrete may result in extensive cracking, expansion,
9
loss of bond between the cement paste and aggregates, loss of concrete strength, and
complete deterioration of concrete. As such, sulfate attack is an important concern with
regard to concrete durability.
Sulfate attack on concrete has been reported in many parts of the world such as
western United States, the northern Great Plains area of the United States and Canada,
Spain, Great Britain, and the Middle East (Lamond and Pielert, 2006). This problem is
more common in areas where the concentration of sulfate ions is high in the soil or
groundwater. In Canada, soluble sulfates are frequently present in very high
concentrations in soils and groundwaters on the Canadian Prairies (Swenson, 1971). In
these areas, surface deposits of mainly sodium sulfate and magnesium sulfate are found,
and also the concentration of these sulfate ions in the groundwater is often high (Figure
1.2 shows the concentration of sulfate ions in groundwater in South Alberta). Although
such concentrations are not normally high enough to harm the growth of crops and
other vegetation, they are high enough to cause damage to concrete. Therefore, it is
imperative that preventive measures be taken to avoid sulfate attack on concrete,
especially in areas where damage is more likely. Examples of such preventive measures
are using sulfate-resisting cement, using supplementary cementitious materials, and
making concrete with low permeability (low water to cement ratio and appropriate
curing regimes).
10
Figure 1.2 : the concentration of sulfate ions in groundwater in Southern Alberta (Day, 2011)
1.4. Research Objectives
The objectives of this research are as follows:
• Evaluate the sulfate resistance of Portland-limestone cements, both at
23°C and 5°C.
• Develop and propose a test method for evaluation of sulfate resistance of
cement specifically addressing the thaumasite form of sulfate attack. Such
a method could be adopted by standards as a means to evaluate the
resistance of cements to thaumasite sulfate attack.
• Test the effectiveness of using supplementary cementitious materials with
Portland-limestone cements to improve their resistance to sulfate attack.
11
• Study the time evolution of sulfate attack on Portland-limestone cements
and identify the phases formed at different stages.
• Examine the strength-porosity relationship of samples made with
Portland-limestone cements and various levels of SCM replacement. This
would help to understand the effect of using SCMs with Portland-
limestone cements, and whether there is an optimum level for limestone
content of the cement if used with SCMs.
12
2. Background
The main theme of the present study is to examine the sulfate resistance of
Portland-limestone cements and their combinations with supplementary cementitious
materials (SCMs). For this purpose, it is necessary to have an understanding of the nature
of sulfate attack, its various forms, the strategies for controlling it, and the test methods
used to evaluate the sulfate resistance of cements. Next, previous research on the sulfate
resistance of Portland-limestone cements has to be studied. This will reveal the state-of-
the-art in literature, as well as the areas in which more research needs to be conducted.
Finally, since the effectiveness of using SCMs in improving the sulfate resistance of
Portland-limestone cements is being examined, it is useful to look at the properties of
such combinations in terms of hydration, strength, and porosity. This could provide some
knowledge as to how limestone interacts with SCMs, and what could be expected from
these combinations when exposed to sulfate solutions. These issues are discussed in the
following sections.
2.1 Sulfate Attack
Degradation of concrete as a result of chemical reactions between hydrated
cement and sulfate ions is referred to as sulfate attack. Based on the source from which
sulfates originate, sulfate attack is generally categorized as external sulfate attack or
internal sulfate attack. External sulfate attack occurs when sulfate solutions penetrate into
concrete from an outside source such as soil, groundwater, sea water, decaying organic
matter, and industrial effluent surrounding a concrete structure. The most common types
of solutions attacking concrete include sodium, magnesium, calcium, and potassium
sulfate. When a soluble source of sulfate is incorporated into the concrete at the time of
mixing, e.g. sulfate-rich aggregates or extra gypsum in mix design, internal sulfate attack
might happen.
Deterioration of concrete by sulfate attack is known to take two forms that are
distinctly different from each other (Mehta and Monteiro, 2006). The main form in which
sulfate attack manifests itself is expansion. The expansion is often related to the
13
formation of ettringite and gypsum, and it potentially leads to cracking of concrete. Once
concrete is cracked, its permeability increases and more aggressive materials are able to
penetrate into the concrete. However, sulfate attack can result in progressive loss of
strength and loss of mass, which happens when the cohesiveness of the cement hydration
product is depreciated. The thaumasite form of sulfate attack is an example of mass loss
and disintegration of concrete.
2.1.1. Conventional Sulfate Attack
In conventional sulfate attack, incoming sulfate ions react with calcium aluminate
phases and calcium hydroxide in hardened cement paste to form either ettringite
(3CaO·Al2O3·3CaSO4·32H2O) or gypsum (CaSO4·2H2O). The consequence is usually
expansion, cracking, and spalling of the attacked concrete or mortar. Due to the
complicated nature of the attack, several formulations have been put forward regarding
its mechanism, some of which are conflicting (Hime and Mather, 1999). The mechanism
of the attack, its progression, the symptoms of deterioration, and the time to failure seem
to be dependent on the type of attacking solution. However, it is generally accepted that
the formation of ettringite and gypsum are the main elements of this form of sulfate
attack.
2.1.1.1. Ettringite Formation
In the presence of calcium hydroxide (CH) and water (H), monosulfate hydrate
(C3A·CS·H18) and calcium aluminate hydrate (C3A·CS·H18) react with the sulfate (S) to
produce ettringite (Mehta and Monteiro, 2006):
C3A·CS·H18 + 2CH + 2S + 12H � C3A·3CS·H32
C3A·CH·H18 + 2CH + 3S + 11H � C3A·3CS·H32
It should be noted that ettringite produced by the reactions described above
occupies a smaller volume than what the reactants occupied. Therefore, the expansion
caused by sulfate attack can not be solely because of the reactions mentioned above. It is
14
generally accepted that the expansion caused by sulfate attack is the result of a particular
mechanism associated with the ettringite reaction or the result of a reaction other than the
formation of ettringite, such as gypsum formation.
Although the mechanism of expansion associated with the formation of ettringite
is still debated, two particular mechanisms have been widely published: the topochemical
reaction mechanism and the swelling mechanism (Neville, 1995). According to the
topochemical reaction theory of expansion (Mather, 1984), the reaction between C3A and
the sulfate and calcium ions in the concrete pore fluid is topochemical, i.e. in solid state.
Ettringite is then formed in the immediate vicinity of the former phase rather than evenly
within the entire volume of the paste due to the difference in solubility of C3A and
gypsum. At the same time, the gypsum present dissolves and the Ca2+ and SO42- ions and
water migrate to the place of ettringite formation. Consequently, there would be a local
increase in the volume of the solids, an overall expansion of the paste, and an increase of
its porosity (Odler and Gasser, 1988).
The swelling theory of expansion suggests that poorly crystalline ettringite which
precipitated in the solution produces expansion by adsorption of water (Mehta, 1993). In
a system containing sufficient concentrations of sulfate, hydroxyl, and calcium ions,
small, nearly colloidal ettringite is believed to form. This poorly crystalline ettringite
adsorbs water from the environment outside the concrete member which generates an
osmotic pressure. This pressure will lead to volumetric expansion of the member if the
elastic modulus of the concrete is sufficiently low (Kurtis et al, 1998).
As mentioned before, the formation of ettringite can, but does not necessarily lead
to expansion. According to Skalny et al. (2002), several conditions must be met to
produce expansion, including the following:
• There is a threshold amount for ettringite formation which if exceeded, a
pressure may be generated on the neighboring solids.
• Ettringite is formed in a topochemical process, resulting in an oriented growth
of the ettringite crystals towards the neighbouring solids. It is still not clear
15
whether and under which conditions the crystallization pressure generated
when ettringite crystallizes is sufficient to produce an expansion.
2.1.1.2. Gypsum Formation
During sulfate attack, gypsum (CaSO4·2H2O) can form as a result of reaction
between calcium hydroxide (CH) formed during cement hydration, and sulfate ions in
aqueous solution:
Ca(OH)2 + SO42- + 2H2O � CaSO4·2H2O + 2OH-
Even though it is generally accepted that the formation of gypsum has detrimental
effects, the specific mechanism is not well established (Tian and Cohen, 2000). While
some studies suggest that formation of gypsum results in expansion, others suspect
whether gypsum formation associated with sulfate attack leads to expansion. The
formation of gypsum is believed not to cause an increase in volume if the reaction
proceeds by a through-solution mechanism (Hansen, 1966). In other words, only when
gypsum crystals form in situ or on surfaces of CH particles, this gypsum formed in a
capillary cavity occupies a larger volume than that of the cavity plus the volume of the
solid CH consumed in this reaction. However, when gypsum forms from hemihydrate, a
change in the crystal structure could cause the inclusion of air voids which could result in
expansion (Sanathanam et al, 2003).
The above mentioned reaction implies that more portlandite (CH) could make
concrete more susceptible to formation of gypsum and possible damage. In other words,
an increase of CH generated during the hydration of Portland cement may result in the
production of a large amount of gypsum in concrete exposed to sulfate sources. As such,
since the amount of CH generated during hydration is related to the C3S content of
cement, it has been suggested that sulfate attack on concrete would become more serious
with increasing C3S content in modern cements (Day and Joshi, 1986) As a matter of
fact, the experiments of several researchers have indicated that samples made with C3S
cements suffered serious deterioration due to the formation of gypsum during exposure to
16
sulfate environments (Mehta, 1979; Tian and Cohen, 2000; Monteiro and Kurtis, 2003,
Irassar et al, 2005).
The concentration of the sulfate ions seems to be important with regard to the
formation of gypsum during sulfate attack. The formation of gypsum in 5% sodium
sulfate solutions, which corresponds to about 33g SO42−/L is well established in the
literature. Also, gypsum has been observed in damaged specimens after exposure to a
sodium sulfate solution of 24g SO42−/L by Gollop and Taylor (1992, 1995). Based on
calculations and experiments, Bellman et al (2005) presented the required concentration
of sulfate ions for formation of gypsum (Figure 2.1). They stated that the formation of
gypsum is either not possible or cannot lead to damage at the common moderate sulfate
concentrations of 1.5 to 3 mg/L, because supersaturation and swelling pressure are very
low. According to Biczok (1967), gypsum is rarely observed at low concentration of
sulfates (less than 1 g SO42−/L); it begins to form in the intermediate range (1–8 g
SO42−/L) and becomes the primary product deposited at high concentrations (more than 8
g SO42−/L).
Figure 2.1: Sulfate concentration that is required for a reaction of portlandite to gypsum or syngenite, calculations (line) and experimental measurements (dots) (Bellman et al, 2005).
17
2.1.2. Thaumasite Sulfate Attack
Thaumasite sulfate attack (TSA) is a form of sulfate attack attributed to the
formation of thaumasite (CaSiO3.CaCO3.CaSO4.15H2O), a calcium-silicate-sulfate-
carbonate mineral. Thaumasite forms in concrete as a result of a reaction between
calcium-silicate hydrates (C-S-H) with sulfates in the presence of carbonate ions in wet
environments (Hooton and Thomas, 2002). This reaction happens slowly, and eventually
results in a soft, white, pulpy mass which causes disintegration of the concrete. It should
be noted that low temperatures below 15°C, particularly between 0 to 5°C, are more
favorable for the formation of thaumasite (Bensted, 1999). However, a few cases have
been reported in which thaumasite sulfate attack occurred at temperatures around 20°C or
more (Collepardi, 1999; Diamond, 2003). Thus, although thaumasite formation is
accelerated at lower temperatures, it is not necessarily precluded at higher temperatures
(Sims and Huntley, 2004). According to Macphee and Barnett (2004), once formed,
thaumasite can remain stable in environments up to 30°C.
It is noteworthy that the formation of thaumasite is not always destructive. In fact,
the Thaumasite Expert Group (TEG, 1999) has termed two varieties of formation of
thaumasite as thaumasite formation (TF), and thaumasite sulfate attack (TSA). TF refers
to incidences where thaumasite can be found in preexisting voids and cracks without
necessarily causing deterioration of the host concrete or mortar. On the other hand, TSA
refers to cases where there is significant damage to the matrix of a concrete or mortar as a
result of conversion of calcium silicate hydrates (C-S-H) in the hardened Portland cement
to thaumasite.
2.1.2.1. Mechanisms of Thaumasite Formation
The mechanism of thaumasite formation in cement mixes is debated. Thaumasite
can form as a result or reaction between C-S-H, calcium sulfate, calcium carbonate and
water (Heinz and Urbonas, 2003):
C–S–H + CaCO3 + CaSO4 + xH2O � CaSiO3.CaSO4.CaCO3.15H2O (thaumasite)
18
This is known as the direct route for thaumasite formation. Irassar et al. (2005)
have proposed a sequence of stages for direct thaumasite formation as follows: (1)
diffusion of sulfate ions and CH leaching, (2) formation of ettringite, (3) gypsum
formation and depletion of CH, (4) decalcification of C–S–H, and (5) thaumasite
formation. According to this hypothesis, during the initial period, the usual cement
hydration reactions take place, and ettringite, C-S-H gel, and portlandite are formed.
Once all (or nearly all) the C3A has reacted, the external sulfate supply makes the sulfate
ions react with Ca2+ ions, more CH is decomposed in the pores causing the CH depletion,
and gypsum crystallization begins. As more gypsum is produced due to the continuous
removal of CH, the OH¯ concentration into the pores would decay, which would result in
the instability and decalcification of C–S–H. Consequently, thaumasite may form due to
the reaction of calcium carbonate, calcium sulfate, the C–S–H gel, and water. In a sodium
sulfate (Na2SO4) solution, decomposition of C–S–H is believed to be the source of silica
available in the pore solution that reacts with the carbonate and sulfate ions to form the
thaumasite. However, when magnesium ions are available (e.g. in MgSO4 solution), the
reactions seem to be different. In this case, the overall reaction of attack has been
summarized as follows (Hartshorn et al., 1999):
C3S2H3 + 3CH + 2CC + 4MS +32H � 2C3SCSH15 + 2CSH2 + 4MH
However, thaumasite can form as a result of reaction between ettringite, C–S–H
and carbonate ions in the presence of excess water, commonly known as the woodfordite
route (Bensted, 2003). In this path, besides gypsum, the aluminate C3A and ferrite C4AF
also contribute to thaumasite formation (Schmidt, 2007):
3CaO·Al2O3·3CaSO4·32H2O + 3CaO·2SiO2·3H2O + 2CaCO3 + 4H2O �
2CaSiO3·CaSO4·CaCO3·15H2O + CaSO4·2H2O + 2Al(OH)3 + 4Ca(OH)2
According to this theory, thaumasite and ettringite form as end members from a
solid solution, called woodfordite, which occurred through the reaction between
ettringite, silicate and carbonate in the presence of excess water. Although this reaction is
very slow, the rate rises significantly after an initial period when thaumasite has started to
19
form. It should be noted that the solid solution between ettringite and thaumasite is not
continuous.
Other mechanisms have been proposed for the formation of thaumasite, including
the topochemical replacement of ettringite by thaumasite (Crammond, 2003), the through
solution mechanism (Crammond, 2003), and ettringite as nucleation sites for formation of
thaumasite (Kohler, 2006). However, in summary, the formation theories commonly fall
into two categories: thaumasite is either formed through decomposition of ettringite or it
forms directly from the solution.
2.1.2.2. Thaumasite Deterioration
The formation of thaumasite can result in varying levels of deterioration in
concrete. Visual signs of deterioration due to thaumasite sulfate attack include spalling
combined with scaling of the surface of the affected concrete structure, sub-parallel
cracks filled with thaumasite, and white haloes of thaumasite occurring around aggregate
pieces (Crammond, 2002). Eventually, TSA can lead to the transformation of the cement
paste into a white, pulpy mush composed of thaumasite that has virtually no strength.
This could jeopardize the integrity of the structure. Therefore, it is critical that TSA be
avoided in the first place since stopping its progression cannot be done easily.
Many cases of damage due to thaumasite sulfate attack have been reported across
the world, especially in the past 20 years. The occurrence of thaumasite in cement-based
components exposed to sulfate attack was first reported by Erlin and Stark (1965) in
USA. They observed thaumasite sulfate attack (TSA) in two sanitary sewer pipes, at the
base of a core taken from an 11-year old concrete pavement, and a cement grout taken
from a salt mine. A few cases of thaumasite deterioration were reported in the Canadian
Arctic in early 1990s (Bickley et al, 1994). However, it was not until 1998 that TSA
began receiving great attention, when the buried elements of thirty-year old highway
bridges in the UK were found to have deteriorated as a result of TSA (Hobbs, 2003). As a
result of this discovery, the UK government commissioned the Thaumasite Expert Group
(TEG) to investigate the problem. Members of the TEG have carried out extensive
20
studies on the thaumasite form of sulfate attack since 1999. According to their recent
findings (Clark, 2007), the number of known cases of TSA in the UK had increased by
about 20 at the time of the first TEG review in March 2000 and by a further 30 by the
time of the second review in August 2002. They stated that about 70% of the structures
investigated since then have significant TSA degradation.
Oberholster (2002) found that the use of a carbonaceous, sulfide-bearing slate
aggregate in concrete floor slabs and in cement bricks resulted in serious cracking due to
thaumasite formation in several two-year-old houses in South Africa. Collepardi (1999)
reported TSA in concrete linings and historic brickwork made of limestone after almost
40 years in Italy. Deterioration due to thaumasite formation has been reported to take
place in tunnel structures in Switzerland (Romer et al, 2003). The degradation of the non-
shrinking cement mortar used in repair works in some Slovenian railway tunnels was
attributed to the formation of thaumasite (Suput et al, 2003). Instances of thaumasite
formation has been found in the concrete foundations to a housing development (Brown
and Doerr, 2000) and slabs on grade (Sahu et al, 2002) in Southern California. Ma et al
(2006) reported the first instance of the thaumasite form of sulfate attack of concrete in
China in a dam tunnel.
Figure 2.2: Soft, white mush consisting of mainly thaumasite and gypsum (Romer et al, 2003).
21
As stated earlier, in Canada, the first highly publicized deterioration occurred in
the early 1990’s, when severe deterioration caused by TSA in the reinforced concrete
foundations of two buildings in the Canadian arctic was observed (Bickley et al, 1994).
The damage was so extensive that some of the columns had to be replaced within 2 years
of casting. Thomas et al (2003) reported the observation of thaumasite in a number of
concrete samples including an 80 year-old aqueduct in Manitoba and a 33 year-old
pavement in Ontario.
2.1.3. Mitigating Sulfate Attack
Due to the extensive occurrence and destructiveness of sulfate attack, many
researchers have focused on investigating factors that influence the attack and finding
solutions to mitigate its detrimental effects. The majority of these investigations have
reached or confirmed the importance of three main elements on the sulfate resistance of
concrete (Depuy, 1994):
1. The permeability and quality of concrete plays an important role in
resistance against sulfate attack. Making high-quality, impermeable
concrete is essential in order to minimize the harmful consequences of
sulfate attack.
2. In most cases, the composition of the cement affects the deterioration due
to sulfate attack. Specifically, the amount of C3A in cement is related to
the extent of damage; as the C3A content of cement increases, the
deterioration also increases.
3. The damage also shows some relationship with the amount of calcium
hydroxide Ca(OH)2 in the hydrated cement paste. Since Ca(OH)2 is
involved in the reactions of sulfate attack, controlling the amount of
Ca(OH)2 in hydrated cement paste would improve the sulfate resistance
of cement mixes. This could be done either by limiting the amount of
Ca(OH)2 formed from hydration or by converting the Ca(OH)2 formed to
C-S-H which is more chemically resistant. The latter could be achieved
22
by the addition of SiO2, either in an active form as in slag or pozzolans,
or as pulverized quartz, and steam curing.
These mitigating steps could be put in practice through the following means: 1)
designing and making concrete with low permeability; (2) using sulfate-resistant
cements, i.e. those with low C3A or low C3A and C4AF contents; (3) using supplementary
cementitious materials (SCMs). The positive effects of using SCMs in regards with the
sulfate resistance of cement is two-fold; they not only densify the cement paste and
reduce its permeability, but also can dilute the C3A content of the system, reduce the
amount of Ca(OH)2 by transforming it to C-S-H, and reduce the alumina content of the
mixture in case of SCMs with low lime (Javed et al, 2006).
The effectiveness of implementing these factors in increasing the sulfate
resistance of cement is well established in the literature. The ACI 201 “Guide to Durable
Concrete” (2008), states that protection against sulfate attack is obtained by using
concrete that retards the ingress and movement of water through lowering the water-to-
cement ratio, minimizing shrinkage cracking, proper placement, compaction, finishing,
and curing of concrete. In ASTM C 150 Type V sulfate-resisting cement, the calculated
C3A is limited to 5% and the sum of the C4AF content plus two times the C3A content is
limited to 25%. As well, Type II moderately sulfate-resisting cement must have a
calculated C3A no greater than 8%. Moreover, the use of SCMs in making sulfate-
resisting concrete is frequently reported (Javed et al, 2006).
Although these measures for mitigating the damage of sulfate attack mainly
pertain to conventional sulfate attack, it is likely that similar actions could be of use in
regards with thaumasite sulfate attack. The effectiveness of using sulfate-resistant
Portland cements in controlling thaumasite sulfate attack is disputed since these cements
are formulated to be low in calcium aluminate, but still contain similar calcium silicate
phases as ordinary Portland cement (Sims and Huntley, 2004). Moreover, thaumasite
formation can be accelerated by Al2O3 bearing components in cements, and even the
small Al2O3 contents present in sulfate-resistant Portland cements can be enough to
accelerate the process (Nobst and Stark, 2003)
23
Nonetheless, the rate of ingress by sulfate solutions can be reduced by producing
concrete with low-permeability properties, which would reduce the rate of TSA in turn.
Results of a few studies suggest that use of pozzolans, slag, silica fume, fly ash, and
metakaloin may be helpful for this purpose, at least in the sense of lowering permeability.
The replacement of cement with ground-granulated blast-furnace slag is reported to be
effective in controlling damage due to TSA (Barnett, 2002; Higgins, 2003; Higgins and
Crammond, 2003; Brown et al, 2003; Skaropoulou et al, 2009). Mixes containing
sufficient fly ash were resistant against the formation of thaumasite (Thomas et al, 2003;
Nobst and Stark, 2003; Mulenga, 2003; Skaropoulou et al, 2009). Studies by Crammond
(2002) showed that using silica fume was effective in preventing TSA. Moreover,
Smallwood et al (2003), Tsivilis et al (2003), and Skaropoulou et al (2009) have reported
that using metakaolin with the cement is beneficial in controlling TSA.
2.1.4. Tests for Sulfate Resistance
It is evident that the nature of sulfate attack on concrete is complex. Different
forms of sulfate attack might happen based on the materials used, the exposure
conditions, the attacking agent, and other factors influencing the nature of attack.
Therefore, it is impossible to have a single test procedure covering all these situations.
Nonetheless, the sulfate resistance of cements and their combinations with various
materials can be tested either in field exposures under natural conditions or in the
laboratory under artificial and usually accelerated conditions. In laboratory tests, either
sulfate is added internally to the concrete during mixing or samples are stored in sulfate
solutions, generally sodium sulfate, magnesium sulfate, or a combination of the two. In
some cases, the samples may undergo wet-dry cycles to accelerate exposure conditions.
Assessment of the effects of sulfate attack can be estimated by several parameters
including visual inspection, dimension changes, loss of mass, loss of strength, and
changes in dynamic modulus of elasticity. As such, a variety of tests have been used,
none of which is considered to be absolute.
In the ASTM C 452 method, gypsum is added to 25 by 25 by 250-mm mortar bars
in amounts so that the total SO3 content becomes 7% by mass of cement. The samples are
24
then stored in water at 23 °C and the change in their length is measured over time. The
main application of this test is to establish whether a sulfate-resistant Portland cement
would meet the performance requirements of ASTM C 150. Since the criterion is selected
to be the amount of expansion at 14 days, the test is considered to be suitable only for
Portland cements, not blended Portland cements or blends of Portland cements with
supplementary cementitious materials since these cements develop their properties
(including sulfate resistance) more slowly. It is believed that the results of this test
commonly conform to long-term behavior of cements in concrete exposed to aggressive
sulfates and predictions based on the C3A content of the cement (Thomas and Skalny,
2006).
In the ASTM C 1012 or the CSA A3004-C8-A method, 25 by 25 by 250-mm
mortar bars are immersed in 5% Na2SO4 solution (50 g/l) at 23 °C. The length change of
the mortars is then measured periodically up to 1 year, and 18 months in some cases. This
test can be used to assess the effects of various cementitious materials on sulfate
resistance. In other words, it can be used for Portland cements, blended hydraulic
cements, and blends of Portland cement with cementitious materials such as slag and
pozzolan. It should be noted that the procedure is not representative of all sulfate solution
compositions; hence, if effects of exposure to a specific sulfate solution are to be
evaluated, that solution should be used in the test (Thomas and Skalny, 2006).
Although not stated, the implication is that the standard methods described above
concern conventional sulfate attack, usually related to the formation of ettringite. These
accelerated methods are performed at normal laboratory temperatures, while the reactions
involving thaumasite formation are very slow or are suppressed at these temperatures but
can occur relatively fast at low temperatures. Moreover, the presence of a source of
carbonate ions appears to be necessary for thaumasite formation, whereas these standard
test methods have no provisions in this regard. In addition to that, many standards have
specifications or requirements for concrete exposed to sulfate environments; however, the
lack of any specific recommendations or performance criteria regarding thaumasite
sulfate attack is apparent.
25
Prior to this study, there was no standard method for evaluating cement mixes
with regard to potential exposure to thaumasite sulfate attack However, as a result of the
present study, the CSA A3004-C8 test method was modified in 2010 to describe two
different procedures for evaluating the sulfate resistance of Portland cements. A new
procedure, denoted as Procedure B, was introduced which is used to determine resistance
to the potential for the thaumasite form of sulfate attack. It is applicable for testing
blended Portland-limestone cements as well as Portland-limestone cements used together
with supplementary cementing materials. The main difference between Procedure A and
Procedure B is that in Procedure B, the mortar bars are stored in sulfate solution at 5°C
rather than 23°C. This low temperature favors the formation of thaumasite in sulfate
attack. Also, the limestone content of the Portland-limestone cement acts as an internal
source of carbonate ions contributing to the formation of thaumasite.
2.2. Sulfate Resistance of Portland-Limestone Cements
From a theoretical point of view, it is expected that the use of limestone in
cements would dilute the C3A and other active aluminate content of cements and or
cementitious systems. Moreover, assuming that limestone would react with calcium
aluminates to form carboaluminates, the available alumina to participate in deleterious
sulfate reactions would be reduced (Hooton et al, 2007). These would potentially
contribute to the sulfate resistance of cement mixes. However, the use of excessive
amounts of limestone in cement could possibly decrease the strength of concrete and
increase its permeability, allowing faster ingress of external sulfates into the concrete.
Therefore, it is essential to investigate the sulfate resistance of Portland-limestone
cements with regard to their limestone contents. Although it is generally agreed that there
is no evidence of deleterious effects of up to 5% limestone in cements (Hooton and
Thomas, 2002), as currently allowed in both CSA A3001 and ASTM C 150 cements, the
sulfate resistance of Portland-limestone cements containing limestone contents greater
than 5% is debatable.
26
2.2.1. Conventional Sulfate Attack
The effect of inclusion of limestone on the sulfate resistance of Portland cement
mixes has been studied by several researchers. Soroka and Stern (1976) studied the effect
of calcium carbonate and calcium fluoride on the sulfate resistance of mortar bars. The
amount of these fillers was chosen to be 10, 20, 30, and 40% by mass of cement,
replacing a corresponding absolute volume of the sand. They used a water-to-cement
ratio of 0.75 to cast the bars, wet cured them for 7 days, and then air-cured them at 65%
relative humidity to 28 days. Next, they immersed the samples in 5% sodium sulfate
solution and measured the length change and time-to-cracking. Their observations
suggested that the addition of calcium carbonate to the mortars increased the sulfate
resistance as the onset of cracking was delayed with increased amount of limestone.
Moreover, the compressive strength of the samples at 28 days (prior to being exposed to
sulfate solution) increased with the addition of limestone (Table 2.1). They suggest that
the formation of calcium carboaluminates was possibly the reason for the improvement in
sulfate resistance. This was somewhat aligned with the theory previously put forward by
Chatterji and Jeffery (1963), which suggested that the formation of carboaluminates
suppressed the formation of monosulfate, and therefore reduced the potential for
ettringite formation on exposure to sulfate solutions. They also suggested that sulfate
resistance was less sensitive to variation in porosity compared to compressive strength.
Table 2.1: Sulfate Resistance of Cement with Limestone Replacements (after Soroka and Stern, 1976)
Mortar Onset of Cracking (weeks) Compressive Strength at 28
days (MPa)
Reference Mortar 6 24.8
With CaCO3 Filler (wt%) 10 20 30 40
10 12 14 16
26.5 26.8 29.1 30.3
Following the previous study, Soroka and Setter (1980) examined the expansion
and deterioration of mortar prisms made with ground limestone, dolomite, and basalt,
immersed in 5% sodium sulfate solution. They used a range of 10 to 40 percent for the
27
filler content, and their fineness (Blaine) varied from 1150 to 11000 cm2/g. They found
that the replacement of Portland cement with limestone improved its sulfate resistance in
the sense that the rate of expansion of the specimens containing limestone filler was
slower than that of the other fillers, and the onset of cracking was delayed in these
specimens. Moreover, they found that the best improvement occurred when the limestone
had the highest fineness, as shown in Table 2.2. They suspected that this improvement
was due to the increased likelihood of very fine limestone reacting to form monocalcium
carboaluminate hydrates. Nevertheless, they stated that although the limestone filler
improved the sulfate resistance of Portland cement mixtures, this improvement was not to
the extent of sulfate-resistant Portland cements.
Table 2.2: Effect of 30% filler on time to failure of mortar bars in 5% Na2SO4 (Soroka and Setter, 1980)
Onset of Cracking, Weeks
Fineness, cm2/g
Limestone Dolomite Basalt
1150-1300 3000-3700 6600-7100
9600-11200
12 10 10 18
12 (?) 6 6 6
4 4 4 2
Reference Mortar:
No Filler 6 Weeks
The effects of large quantities of limestone, namely 30% and 50%, on the sulfate
resistance of concrete were studied by Marsh and Joshi (1986). They cast 25 x 25 x 300
mm specimens and cured them in saturated limewater for 28 days at either 20°C or 50°C.
They cut the specimens to size at age 7 days and determined their resistance against
sulfate attack by measuring the length change of 125 mm cement paste prisms immersed
in 0.35 M Na2SO4 solution at 20°C. They also maintained the pH of the solution at
approximately 7 by the addition of 0.5 M H2SO4 as needed. Their findings showed that at
these dosages, the use of limestone resulted in increased expansions due to sulfate attack.
However, all of the pastes cured at 50°C were found to be sulfate resistant for exposure
periods beyond one year.
28
Hooton (1990) tested pairs of commercially produced cements with and without
limestone based on both ASTM C 452 and C 1012. The ASTM C 452 test method
determines the potential sulfate resistance of a cement by measuring the expansion of
mortar bars stored in water after their SO3 content is raised to 7.0% using gypsum. In
ASTM C 1012, mortar bars are exposed to 5% sulfate solution after the compressive
strength of companions cubes exceed 21 ± 1 MPa, and the length change is measured
periodically. The study concluded that for the limited range of C3A use (from 7.3% to
10.4%), carbonate additions did not influence the sulfate resistance of cements
significantly, and the main influencing factor was the C3A content. No clear trend was
found with regards to the effect of carbonate on sulfate resistance at room temperature
(Tables 2.3 and 2.4).
Table 2.3: ASTM C 452 Expansion of Mortars, % (Hooton, 1990)
Cement 1 1c 2 2c 3 3c
%C3A
%CaCO3 (by TGA)
10.4 0.3
10.0 4.1
9.1 0.8
9.8 4.7
8.3 0.3
7.3 2.6
Age, Days
14 28 56 91
105 119 170 261 365
0.054 0.071 0.084 0.086 0.088 0.088 0.087 0.088 0.090
0.058 0.092 0.126 0.142 0.144 0.145 0.146 0.148 0.150
0.036 0.053 0.066 0.075 0.077 0.079 0.079 0.079 0.080
0.041 0.054 0.058 0.058 0.059 0.059 0.059 0.059 0.060
0.039 0.051 0.058 0.061 0.062 0.062 0.064 0.065 0.068
0.036 0.043 0.050 0.054 0.055 0.056 0.056 0.057 0.060
Table 2.4: Time to Expansion of ASTM C 1012 Mortar Bars (Hooton, 1990)
Cement 1 1c 2 2c 3 3c
%C3A
%CaCO3 (by TGA)
10.4 0.3
10.0 4.1
9.1 0.8
9.8 4.7
8.3 0.3
7.3 2.6
Time to 0.10%
expansion, days 117 142 167 161 196 236
Matthews (1994) measured the wear rating and compressive strength of
concrete samples made with cements having different C3A contents and limestone
additions of 0%, 5%, and 25%. Concretes had water-to-cementitious ratio of about 0.6,
29
were cured for 28 days in water and then stored in sodium sulfate or magnesium sulfate
solutions at 20°C. The results indicated that when no limestone was added, concretes
made with cements having high C3A content were completely (all those with 13.1% C3A
and some of those with 10.3%) or partially (some of those with 10.3% C3A) disintegrated
at 5 years. However, the performance of concretes made with cements with C3A equal to
7.1%, 5.3% and 8.6% was satisfactory. In other words, the sulfate resistance of ordinary
Portland cement (OPC) was clearly related to the C3A content of the cement.
Nevertheless, there was little evidence of any systematic effect of 5% of limestone
addition on the sulfate resistance. Moreover, he found no consistent difference between
the performance of OPC and those with 25% limestone addition as it improved in some
mixtures while worsening in others. The overall conclusion was that the effect of 25%
limestone addition was to extend the range of performance obtained from OPC with a
wide range of C3A contents, the OPC with the lowest C3A content being improved and
the rate of deterioration of the highest C3A OPC being increased.
Sawicz and Heng (1996) investigated the influence of powdered limestone and
water to cement ratio on the sulfate resistance of concrete. In their mixes, they used
crushed limestone of 5% porosity as a coarse aggregate and limestone powder, prepared
in the laboratory by grinding coarse aggregate to Blaine fineness of 330 m2/kg as a
mineral addition. They added the limestone powder to the concrete in place of the coarse
aggregate to leave the C3A phase at a constant level of 12.1%. The water to cement ratio
of the mixes varied from 0.5 to 0.7. After curing 5 x 5 x 15 cm samples for 28 days under
standard conditions of 20 ± 2°C and relative humidity of 90%, they tested the concretes
over the range of compressive strength and absorbability. Then, they immersed the
samples in a 5% Na2SO4 solution and measured their swelling in intervals of 14 days.
Simultaneously, they investigated the phase composition of 1 x 1 x 2 cm cement pastes
with water cement ratio of 0.5, with and without addition of 20% limestone powder, by
means of X-ray diffraction (XRD). According to their findings, for each water to cement
ratio, a progressive increase of limestone powder in concrete would first improve the
sulfate resistance of concrete (decrease the amount of swelling to a minimum value) and
then impair it. Also, a higher content of limestone powder was required to achieve the
30
optimum sulfate resistance as the water cement ratio was increased. Taking advantage of
the XRD results, the authors attributed the beneficial effects of limestone on sulfate
resistance to the reversion of the initially formed ettringite (during curing) to
carboaluminate phases (monocarbonate and hemicarbonate), which are more resistant to
sulfate attack, rather than sulfoaluminates (monsulfate, hemisulfate, and solid solutions).
Figure 2.3 shows the sulfate expansion isobars for different water-to-cement ratios and
limestone levels, as reported in this study.
Figure 2.3: Sulfate expansion isobars (microstrain) for different w/c and limestone levels (Sawicz and Heng, 1996)
González and Irasser (1998) investigated the effect of limestone filler on the
sulfate resistance of low C3A Portland cement. They carried out their tests on mortars
containing one type II cement and two type V Portland cements with different C3S
contents, namely 40 and 74%. They added limestone filler containing 85% of CaCO3
with a Blaine fineness of 710 kg/m2 as 0%, 10%, and 20% of replacement by cement
weight. Their test method was based on the ASTM C 1012 procedure, where they cast a
series of 25 x 25 x 285 mm mortar specimens with sand to cement material ratio of 2.75
31
and water cement ratio of 0.485, cured them in saturated limewater solution until they
achieved 30 ± 3 MPa compressive strength, and immersed them in a 5% sodium sulfate
solution (0.35M Na2SO4). They determined the expansion, flexural and compressive
strengths, solution consumption, and X-ray diffraction analysis at different exposure
times up to 1 year. The results of their study showed a 10% replacement of limestone
filler caused no significant changes in the sulfate performance of cements while sulfate
performance of mortars containing 20% of limestone filler was worsened in all cases,
especially when high C3S or moderate C3A contents were in the Portland cement (Table
2.5). They concluded that the addition of limestone filler may increase or decrease the
sulfate performance of blended cements depending on the mineralogical composition of
the Portland clinker, the amount of filler replacement, and the equilibrium between the
increase of hydration degree before the exposure and the increase in water cement ratio
by addition of filler. Also, they stated that sulfate attack on Portland cements containing
limestone filler occurs without noticeable changes in the mineralogical nature of sulfate
attack products, gypsum and ettringite being the main ones.
Table 2.5: Sulfate Resistance of Limestone Mixtures (González and Irasser, 1998)
Cement Type V Type V Type II
C3A content, % by mass
C3S content, % by mass
0 40
1 74
6 51
Limestone Replacement 0 10 20 0 10 20 0 10 20
Time to 0.10% Expansion, days 1260 857 208 148 164 92 165 209 108
Reduction in compressive strength
(1 year in sulfate solution), % 3 4 5 29 17 50 8 25 40
Borsoi et al (2000) examined the sulfate resistance of paste and mortar specimens
made with ordinary Portland cement, C3A-free Portland cement, slag cement, and
pozzolan cement. They blended these cements with a carbonaceous or siliceous filler
(10% by cement weight), used four different water-to-cement ratios of 0.55, 0.50, 0.45,
and 0.40, and immersed the specimens in magnesium sulfate solutions with SO4-
concentrations of 350, and 3000 mg/l after 28 days of curing. They studied the
deterioration of the samples measuring the elastic modulus and compressive strength as
32
well as making visual observations and using XRD analysis to detect ettringite and/or
thaumasite formation. After 5 years of exposure, they found that the paste and mortar
specimens with slag and pozzolan cements were undamaged irrespective of the sulfate
concentration and w/c. However, paste and mortar specimens with blended limestone-
Portland and exposed to the highest concentration of MgSO4 solution (3000 mg/l)
showed surface damage due to presence of ettringite and thaumasite, yet they did not
show loss of either compressive strength or elastic modulus. The samples exposed to
lower concentrations of MgSO4 did not show any sign of deterioration after 5 years.
Moreover, the samples made with the C3A-free Portland cement showed less
deterioration, i.e. the surface damage was mitigated when the C3A content of the cement
was reduced from 8.2% to 0%.
In a study by Irrasar et al (2000), two type V cements with quite different C3S
contents and blended cements containing natural pozzolana or limestone filler were used
to evaluate their sulfate resistance. The mortars were immersed in a pH-controlled
sodium sulfate solution for 1 year and their expansion, flexural strength, and compressive
strength were measured. Their results showed that the addition of 10% limestone had no
effect on the expansion and strength of samples compared to Type V cements. In this
case, the blended limestone mixtures performed better with higher C3S content. For the
mixture with 20% limestone addition, higher expansion and lower compressive strength
was observed regardless of the C3S content.
Hekal et al (2002) immersed cement pastes blended with various amounts of silica
fume, slag, and calcium carbonate (water-to-binder ratio of 0.3 and 28 days curing) in
10% MgSO4 solution under different conditions, namely room temperature, 60°C, and
drying–immersion cycles at 60°C. The mass change and compressive strength of samples
were measured as indications of sulfate resistance. They noticed that when subjected to
cyclic wetting and drying, the 10% limestone samples exhibited similar strength loss to
that of the control paste mixture. As for the samples immersed in solution at 20 or 60°C,
the 0% and 10% limestone pastes showed similarly unaffected strengths after 180 days,
while the strength of 5% limestone samples was improved.
33
Irassar et al (2005) performed a microstructural analysis of mortars made with a
Type II (6% C3A) and two Type V cements (C3A< 2% and C3S = 40% and 74%), each
with and without 20% limestone replacement. Following the ASTM C 1012 procedure,
they immersed the specimens 5% Na2SO4 solution for two years at 20°C , and studied
the evolution of sulfate attack using XRD semi-quantitative analysis on the material
obtained by wearing in layers accompanied by SEM and EDS studies to confirm the
presence of thaumasite. They found out that all the three Portland cements used passed
ASTM C 452 using the CSA expansion limits. Also, OPC and high-C3S Type V cement
containing 20% limestone filler were more susceptible to sulfate attack than the
corresponding plain cement (Fig. 2.4). The results of XRD and SEM analysis revealed
the formation of thaumasite in all the samples, with the exception of the Type V cements
with low C3S content. They concluded that the reaction sequence in Portland limestone
cements was essentially the same as in plain Portland cements, with thaumasite forming
at later stages with decomposition of the ettringite formed during the first stage of attack.
Figure 2.4: ASTM C 1012 Expansions of Cements with and without 20% Limestone (Irassar et al, 2005)
Taylor (2001a, b) investigated the effects of limestone addition on the sulfate
resistance of low-C3A content cements. In one study (Taylor, 2001a), he interground two
different types of limestone at 2.5% or 3.5% and 3.0% or 5.0% respectively with a
34
clinker having a C3A content of 8%. The sulfate resistance was then evaluated by the
ASTM C 1012 and ASTM C 452 methods. The results showed that all the combinations
had acceptable performance as per ASTM C 150 (determining the type of cement),
ASTM C 595 (standard specification for blended cements), and ASTM 1157
(performance specifications for hydraulic cements). In fact, the mixes with limestone had
slightly lower expansions compared to the control mix.
In a similar study, Taylor (2001b) used two cements with C3A contents of 3% and
5% and interground them with limestone in amounts of 4.0% and 3.5% respectively. He
then evaluated their performance based on the ASTM C 1012 and ASTM C 452 methods,
which revealed accepted performance for all the mixes according to the aforementioned
specifications (ASTM C 150, C 595, and C 1157). Based on his studies, Taylor
concluded that low-C3A cements containing up to 5% added limestone showed
equivalent or improved sulfate resistance at room temperature to those containing no
added limestone (Fig. 2.5).
Figure 2.5: ASTM C 1012 Expansions of Cements with C3A Contents less than 5% (Taylor, 2001b, as presented in Hawkins et al.2005)
Hartshorn et al (2001) subjected mortar specimens with limestone contents of 0%,
5%, 15% and 35% replacement of cement (with C3A content of about 8%) to air curing at
5°C and 20°C, and also immersed in 1.8% MgSO4 solution at the same temperatures.
35
They used a wide range of destructive and non-destructive tests, including XRD, thermal
analysis, scanning electron microscopy, pulse velocity, dynamic modulus, shrinkage,
expansion and changes in mass, flexural, and compressive strengths to examine the
sulfate resistance. Their study showed that mortar prisms containing 35% limestone and
exposed to magnesium sulfate solution at 5°C suffered extensive damage and
deterioration within one year. In addition, they observed strong signs of impending
damage due to sulfate attack on prisms containing 15% limestone. They stated that
compression members containing 35% limestone and exposed to aggressive sulfate
environment at temperatures of about 5°C can be seriously at risk in their ability to carry
loads as the correspondent samples had lost 75% of their compressive strength, while the
loss in flexural strength was relatively small.
In a similar experiment, Collepardi et al (2003) exposed cement pastes to 10%
sodium sulfate solutions at both 5°C and 20°C after 28 days wet curing. The limestone
replacement levels were 0%, 15%, and 30%, and they used a water-to-binder ration of
0.40 for all mixes. Since they observed no evidence of damage after 3 months of
continuous immersion under the sulfate solution, they decided to expose the samples to
wet/dry weekly cycles for four months, each cycle consisting of drying at 40°C for 2 days
and immersion in the sulfate solution for 5 days at 5°C or 20°C. They used XRD
analysis, Scanning Electron Microscopy (SEM), and Energy Dispersive X-Ray Analysis
(EDX) techniques to study the sulfate attack. Their results showed that the drying caused
microcracking of the pastes; however, no sulfate-related damage was observed. The XRD
and SEM results confirmed the presence of thaumasite in cracks in all pastes stored at
5°C with slightly more in the 30% limestone cement. The authors attributed this to the
higher water to cement ratio in the 30% limestone paste (calculated to be 0.57 since the
water to binder ratio was consistent for all mixes).
A study by Barker and Hobbs (1999) consisted of examining the sulfate resistance
of mortars immersed up to 12 months at 5°C in magnesium sulfate and sodium sulfate
solutions. The mortars were prepared from OPC, a sulfate-resisting Portland cement
(SRPC), and two Portland limestone cements containing 15% by mass of an oolitic
limestone and a carboniferous limestone with water-to-cement ratios of 0.5 and 0.75.
36
Their performance was monitored by visual assessment, expansion and changes in
flexural and compressive strengths. They concluded that mortars made with a PC with a
C3A content of about 10% by mass were broadly similar in their vulnerability to sulfate
attack at 5°C as PLC mortars containing 15% limestone by mass, although the mode of
attack was different. While ettringite was observed in the OPC mortars, the PLC mortars
contained both ettringite and thaumasite. Overall, the performance of the limestone
cements was found to be similar to that of OPC in most cases, whereas the SRPC
performed better.
2.2.2. Thaumasite Sulfate Attack (TSA)
In regards to the role of limestone in thaumasite sulfate attack, Bensted (1995) has
emphasized that the limestone filler grains (as present in Portland-limestone cement)
modify the paste microstructure, and the topochemical growth of CH upon CaCO3 crystal
might occur; hence, it may facilitate the access of SO42− ions to form gypsum
expansively. Also, limestone filler increases the hydration rate of Portland cement
leading to the precipitation of CH located around the filler grains and the aggregate
surfaces; however, it does not contribute to the later CH reduction in the transition zone
as do SCMs such as granulated blast furnace slag.
The risk of destruction of mortar and concrete made with Portland-limestone
cement due to the formation of thaumasite during sulfate attack at low temperatures is
increased by the presence of fine calcite particles (Halliwell et al., 1996). It has been
postulated that the limestone content of Portland-limestone cements acts as the internal
source of carbonate ions required for thaumasite sulfate attack (Irassar et al., 2005).
Nevertheless, it is well established that the presence of carbonate, either from
limestone and dolomite in concrete aggregates or bicarbonate ions in groundwater, is
necessary for thaumasite sulfate attack to occur (Clark, 2007). As such, since presence of
soluble carbonate is essential for thaumasite sulfate attack to happen, and limestone is
basically calcium carbonate, the use of Portland-limestone cements might be of concern
with regard to thaumasite sulfate attack.
37
Heinz and Urbonas (2003) made mortars and pastes with water to cement ratio of
0.5 with CEM I Portland cement and its mixtures with 0%, 15%, and 30% limestone
filler. They stored some of these samples for 2 hours and then treated them in water at
95°C for 4 hours, while the rest of the samples were first stored for up to 24 hours at
20°C and 95% relative humidity. After this preliminary curing, the samples were stored
in water at 5°C and 20°C. They measured the length changes and resonant frequencies of
the samples during long-term water-storage, and performed XRD and SEM analysis to
study the mineralogical evolution of the samples. Their findings showed that for pastes
containing 30% limestone filler, small areas were found by SEM and XRD microanalysis
that chemically matched thaumasite or a thaumasite–ettringite solid solution.
A long-term laboratory study by Higgins and Crammond (2003) investigated the
susceptibility of concrete to the thaumasite form of sulfate attack (TSA). They used
Portland cement (C3A of 7.2%), sulfate-resisting Portland cement (0% C3A) and a
combination of 70% ground granulated blast furnace slag (GGBFS) with 30% PC as
cementitious materials and combined them with various carbonate aggregates or a non-
carbonate control. After curing in either water or air, they immersed concrete cubes in
four different sulfate solution at 5°C and 20°C, and monitored the sulfate attack through
visual examination and wear rating for up to 6 years. Based on their results, many of the
PC and SRPC concretes made with carbonate aggregate and stored in sulfate solutions at
5°C suffered deterioration consistent with TSA with SRPC providing no better resistance
to TSA than PC. However, concretes made with 70% slag showed high resistance to
TSA, and in fact, this resistance was substantially improved in the presence of carbonate.
In another project undertaken by the British Building Research Establishment
(Holton, 2003) the resistance of concretes to TSA over a 3 year period at 5°C was
evaluated. Two different types of Portland-limestone cements with medium and high C3A
content, a high-C3A Portland cement and its combination with 30% fly ash were used in
this study. Concretes with water to cement ratio of 0.40 to 0.55 and cement contents
ranging from 330 to 460 kg/m3 were cast and seal cured at 20°C for 28 days. The samples
were exposed to Class 2 exposure (1400 mg/l sulfate as calcium sulfate) and Class 3
exposure (3000 mg/l sulfate as mixed calcium and magnesium sulfate). Evaluation of the
38
samples was based on visual inspection, depth of wear ratings, compressive strength, and
XRD. After 3 years of exposure, evidence of thaumasite damage was found in both types
of sulfate solutions. The conclusion was that the performance of Portland-limestone
cement concretes (with 12% limestone and C3A content of 12%) with water-to-cement
ratios in the range 0.4 to 0.45 and minimum cement contents of 380 kg/m3 was
comparable to that of high C3A Portland cement concretes of 0.5 w/c ratio and 330 kg/m3
minimum cement content with a range of high and low carbonate content aggregates (see
Figures 2.6 and 2.7). Nonetheless, all the concretes were deteriorated due to TSA. The
sulfate resistance of the Portland-limestone cements seemed to be minimally affected by
the C3A content.
Cement Aggregate Free w/c ratio Mix number
High C3A PLC Carboniferous limestone 0.4 A00/235 Year 1 2 3
Wear rating 1 2 5
Photograph
Comment Some exposed aggregate and “peeling” on sides
Thinly exposed aggregate across most
of side faces
Aggregate thinly exposed across all of the side faces, but top
and base intact
Figure 2.6: Photos of concrete cubes: High C3A PLC (12% limestone) at 0.40 W/C (Holton 2003)
Cement Aggregate Free w/c ratio Mix number
High C3A PC Carboniferous limestone 0.5 A00/220 Year 1 2 3
Wear rating 2 3 4
Photograph
Comment Some exposed aggregate and “peeling” on sides
Thinly exposed aggregate on side
faces
Aggregate further exposed on sides with blisters developing on top, still cubic in shape
Figure 2.7: Photos of concrete cubes: High C3A PC at 0.50 W/C (Holton 2003)
39
A long-term study at Sheffield University in the UK focused on the sulfate
resistance of Portland cement combined with limestone filler (Hartshorn et al, 1999;
2001; 2002; Kakali et al, 2003; Torres et al, 2003; Torres et al, 2006). In this study, the
microstructure of Portland cement mortar specimens containing 5%, 15% and 35%
limestone filler was examined. The Portland cement was of normal typical composition
with 8.5% C3A, and the limestone was a carboniferous limestone with a calcium
carbonate content greater than 98%. After curing the specimens in water for 28 days, the
prisms were stored in 1.8% magnesium sulfate solution at 5°C and kept in these
conditions for a minimum of 4 years, after which the solution was allowed to evaporate,
and the test specimens were exposed to a drying environment at 5°C for several months
up to 5 years. The techniques used in these studies included visual examination, XRD,
Infrared Spectroscopy, and SEM. Earlier reports on this study indicated that increasing
levels of limestone increased the rate of thaumasite sulfate attack (Hartshorn et al, 1999).
After 126 days of exposure, severe deterioration was found due to thaumasite in mortar
with 35% separately added limestone filler and impending deterioration with 15% after
one year (Hartshorn et al, 2002). Later findings of this study (Torres et al, 2006)
concluded that the deterioration due to thaumasite advanced with the increased exposure
period and limestone content, especially beyond 5%. They characterized the dominant
phases within the deteriorated cement matrix as thaumasite solid solutions (TSS), which
are responsible for the cracks and delamination at the paste-aggregate interface to these
phases. They also found TSS in the control specimens with no limestone filler, attributing
it to atmospheric carbon dioxide as the most likely source for the carbonates. The authors
strongly opposed the notion that 5% replacement of Portland cement by limestone will
not adversely affect the performance of Portland cement in concrete containing either
siliceous or carbonate aggregates, because their samples containing 5% limestone had
deteriorated more extensively than the OPC control mortar after exposure to magnesium
sulfate at 5°C (see Figure 2.8).
40
Figure 2.8: The section view of prisms stored in 1.8% MgSO4 at 5°C for 5 years (L=limestone) (Torres et al, 2003)
Lipus and Puntke (2003) used Portland cement, high sulfate-resisting slag cement
and high sulfate-resisting Portland cement, and their combinations with 5% and 15%
limestone and 40% fly ash in their study. They cured 0.60 water-to-cement ratio mortar
bars for 14 days and then exposed them to 1500 and 29,800 mg/l SO3 (sodium sulfate)
solutions for 2 years and 180 days respectively at both 8°C and 20°C. They used length
change, resonant frequency, and XRD as analysis methods. According to their findings,
the SRPC (CEM I-HS) or the slag cement (CEM III/B) plus limestone mixtures did not
show any deleterious expansions, and they fulfilled the high sulfate resistance criterion of
the rapid test at 8°C and 20°C. Although XRD results showed thaumasite at the surface
layer, it was not associated with any damage. The same results were found for the mixes
containing 40% fly ash at 8°C or 20°C, except that thaumasite was not detected in these
samples. Nevertheless, the combination of 15% limestone with a high-C3A Portland
cement (CEM I), denoted as CEM II/A-LL, turned out not to be sulfate resistant.
Presence of thaumasite was observed in cracks, especially for those stored at 8°C. The
authors believed that thaumasite was only produced subsequently after primary formation
of ettringite and gypsum, as a type of tertiary product. In other words, TSA would not
have occurred if the primary ettringite cracking damage had not occurred.
41
Trägårdh and Kalinowski (2003) tested self-compacting concrete (SCC)
specimens stored at 5°C for approximately 2 years in three different magnesium sulfate
solutions with concentrations of 0.01% (100 mg/l), 0.05% (500 mg/l) and 0.14% (1400
mg/l). They used an ordinary sulfate resistant Portland cement (CEM I) with low sulfate
and C3A contents, and two blended Portland limestone cements (CEM II) with different
C3A contents in their experiments. Limestone filler was added in proportions of 0, 50,
100 and 180 kg/m3 to the mixes. The test program included measuring the compressive
strength, wear rating, weight and pH along with SEM-EDS and optical light microscopy
to study the microstructure. The results indicated that SCC with blended Portland
limestone cements (CEM II/A-LL) was vulnerable to TSA when exposed to the moderate
magnesium sulfate solution of 0.14%, even if limestone filler was not used. However, the
SCC with blended Portland cements did not show any sign of TSA at weak sulfate
concentrations (0.01 and 0.05 %) of MgSO4. After 22 months of exposure, SCC
containing cement with sulfate resistant Portland cement with a low C3A content (SRPC,
CEM I) remained intact in all solutions. The microstructural study suggested that the
formation of ettringite and ettringite solid solution compounds occurred before
thaumasite was formed. The authors concluded that the C3A content of the cement played
an important role in the extent of sulfate attack.
Mortar prisms made with OPC cement (C3A of 14% based on Bogue equations)
and 30% mass of limestone filler were stored in various sulfate solutions at different
temperatures for periods of up to 1 year (Gao et al, 2007). The water to cement ratio used
was 0.6, and samples were cured for 28 days. Subsequently, the samples were transferred
into three different sulfate solutions of 2% MgSO4, 2% Na2SO4, and 2% MgSO4 plus 2%
Na2SO4. The exposure conditions consisted of 5°C, 20°C, and cycling temperatures of
5°C and 20°C (3 days at each temperature). The methods used to evaluate the samples
included visual inspection, flexural and compressive strength, XRD, Fourier transform
infrared spectroscopy (FTIR), laser-raman spectroscopy, and scanning electron
microscopy (SEM). The results showed that MgSO4 solution was more aggressive than
Na2SO4 solution, and the magnesium ions promote the thaumasite form of sulfate attack
Thaumasite was detected in the mortars containing limestone filler after exposure to
42
sulfate solutions at both 5°C and 20°C, convincing the authors to believe that thaumasite
form of sulfate attack is not limited to low-temperature conditions. In fact, they stated
that the increase of solution temperature accelerated the sulfate attack (both magnesium
and sodium) on the samples and led to more deleterious products including gypsum,
ettringite and brucite formed on the surface of mortars.
Lee et al (2008) studied the effect of various replacement levels, namely 0, 10, 20
and 30% by mass, of limestone filler on sodium and magnesium sulfate attack at ambient
temperature. The cement had a C3A content of 9.7%, and all the mortars and pastes had a
water to cement ratio of 0.45. The specimens were cured for 7 days and then some were
immersed in sodium and magnesium sulfate solutions with 33,800 ppm of SO42-
concentration. The evolution of sulfate attack was monitored by visual inspection,
compressive strength and expansion of mortars, and microstructural analyses such as
XRD and SEM. The test results demonstrated that higher replacement levels of limestone
filler made the samples more susceptible to sulfate attack, irrespective of the type of
attacking sources. However, the deterioration modes seemed to be significantly
dependent on the type of sulfate solutions; results of XRD confirmed that the main cause
of deterioration for the 30% limestone samples exposed to sodium sulfate solution was
thaumasite formation, whereas gypsum formation as well as thaumasite formation was
primarily responsible for the deterioration of samples with high limestone replacement
levels under magnesium sulfate attack (Figure 2.9). Yet, the authors stated that the
deterioration was strongly associated with thaumasite formation in both sulfate solutions,
which suggests that TSA could occur even at normal temperatures (20 ± 1°C in this case).
43
Figure 2.9: XRD patterns of pastes exposed to Na2SO4 solution (left) and MgSO4 solution (right) for 360 days (E = ettringite, T = thaumasite, P = portlandite, C = calcite, G = gypsum, B = brucite) (Lee et al, 2008)
Schmidt et al (2009) examined the consequences of external sulfate attack on
cement pastes and mortars at 8 and 20°C. They cast samples with 0%, 5%, and 25%
limestone additions to a clinker with C3A of 7% (w/c of 0.5), and immersed in sodium
sulfate solutions of 4 g/l and 44 g/l. The methods used to evaluate the samples were
length change, compressive strength, velocity of Leaky Rayleigh waves (to measure
density), TGA, XRD, and SEM. They concluded that the addition of a few percent
limestone in Portland cement reduced the porosity and increased the resistance of
Portland cement systems to sulfate; but a higher addition of 25% increased porosity and
lowered resistance to sulfates. They found that the kinetics of degradation were
dramatically affected by the solution concentration, with gypsum forming only at the
higher concentration. They observed formation of thaumasite in limestone samples stored
at 8 °C, which led to loss of cohesion of the paste and loss of material from the surface of
the samples. They realized that thaumasite formation was always preceded by expansion
and cracking of the samples due to ettringite formation. In other words, the stages of
44
deterioration consisted of expansion at a steady stage, followed by a rapidly accelerating
stage, and finally resulting in total destruction of the samples.
Higgins (2003) examined the sulfate resistance of Portland cement with slag and
a small percentage of calcium carbonate or calcium sulfate. The slag replacement levels
chosen were 0%, 60%, and 70% by mass of cement, and 4% calcium carbonate or 2%
and 3% gypsum was added to some mixes. Concrete cubes were immersed in magnesium
and sodium sulfate solutions (1.3% SO3) and monitored for corner-loss and strength-loss,
for over six years. Also, the expansion of mortars stored in the same solutions as well as
2.4% sodium sulfate was measured for up to six years. The results suggested that both
calcium carbonate and calcium sulfate additions had a consistent beneficial effect on the
resistance of slag concrete to conventional sulfate attack, both in respect to expansion and
in respect of disintegration (see Fig. 2.10).
Figure 2.10: Expansion of prisms in Na2SO4 solutions (Higgins 2003)
2.2.3. Summary and Gaps
Based on studies presented above and other research on the sulfate resistance of
Portland-limestone cements, it is evident that the literature is conflicting with respect to
use of Portland-limestone cements in sulfate exposures, especially with respect to
45
limestone contents of greater than 5%. As Hooton et al (2007) stated, it is difficult to
evaluate these studies since performance of limestone cements is affected by the quality
of limestone used, whether the limestone was interground or blended, and changes in
particle size distribution. Moreover, the sulfate environments to which the samples are
exposed and the methods of evaluating the sulfate resistance of Portland-limestone
cements are diverse in the literature, which makes comparison complicated. In addition to
that, it is well known that the C3A content of the cement plays an important role in the
sulfate resistance of a cement; in fact, most standards including CSA A23.1 do not allow
use of high C3A cement in sulfate exposures without the use of sufficient SCMs. Since
most of the cited studies evaluated the performance of such high C3A cements with
limestone additions alone in sulfate exposures, the relevance of the information obtained
from these studies is questionable (Hooton et al, 2007).
The concern over the resistance of Portland-limestone cements against the
thaumasite form of sulfate attack seems to be valid. As stated above, several researchers
have confirmed the disintegration of samples made with Portland-limestone cements due
to formation of thaumasite, especially at low temperatures. As mentioned earlier, no
standard test method was available for evaluation of sulfate resistance of cement with
regards to the thaumasite form of sulfate attack prior to this study. As such, it was
desirable to propose such a test method that could potentially be adopted by standards
and codes. In fact, one of the main contributions of this study was that the test method
proposed for evaluation of the resistance of cements against thaumasite sulfate attack was
adopted by the CSA A3000 standards committee in 2010 and is currently in practice as
CSA A3004-C8-B.
As mentioned before, data on the performance of Portland-limestone cements
when used in conjunction with SCMs are limited. Although combining SCMs with
ordinary Portland cements is proven to be beneficial in improving their resistance against
sulfate attack, and in fact some standards mandate the usage of SCMs in moderate or
severe sulfate exposures, the applicability of this to Portland-limestone cements requires
further research. It is still unclear whether the use of PLC will increase the replacement
levels of SCMs that need to be used to improve the sulfate resistance of cement.
46
2.3. Portland-Limestone Cements and SCMs
In order to understand and predict the performance of Portland-limestone cements, and
also their combination with supplementary cementitious materials (SCMs), it is useful to
study the hydration, the compressive strength and its relationship with porosity, and the
influence of SCMs on the aforementioned properties of these cements. In general, the
principal hydration products in composite cements are similar to those found in pure
Portland cement; however, the hydration rate or the stoichiometry of hydration products
may be affected by the added constituents (Neville, 1996). Moreover, the combination of
Portland-limestone cements with SCMs would also affect the hydration process. These
changes would then be reflected in the performance of Portland-limestone cements, e.g.
the compressive strength and the porosity. Therefore, it is of interest to investigate the
performance of Portland-limestone cements with regards to their hydration.
2.3.1. Hydration of Portland-Limestone Cements
It was previously believed that the when limestone is added to cement mixes, its
main influence involves the physical characteristics of the product. The calcium
carbonate particles were believed to act as fillers, providing nucleation sites for the
products of hydration reactions (Soroka and Stern, 1976), and the inclusion of limestone
increases the rate of hydration (Soroka and Setter, 1977). According to Ellerbrock et al
(1990), the inclusion of limestone improves the particle packing of the cementitious
system. Moreover, Sprung and Siebel (1991) suggested that the calcite from limestone
participated only to a small extent, if at all, in the hydration reactions, primarily on the
surface of the limestone particles, and thus considered it to be an inert material.
However, several researchers have reported that the presence of limestone affects
the hydration of Portland cement. There exists a general agreement that limestone reacts
primarily with the C3A component of the cement to form carboaluminates (Hooton et al,
2007). According to Bonavetti et al (2001), in limestone-blended cements, calcium
monocarboaluminate could be detected as early as 3 days after casting, and would be
increased after 28 days of hydration. Voglis et al (2005) came to the conclusion that in
47
limestone cement pastes, carbonate ions become incorporated in calcium aluminate
hydrates to form carboaluminates. They observed a detectable amount of
monocarboaluminate after 1 day of hydration, and its amount continued to increase up to
28 days. Similar results have been reported by other researchers (Tsivilis et al, 1998;
Kakali et al, 2000; Lothenback et al, 2008). According to Matschei et al (2007), at low
concentrations, limestone (calcite) reacts completely to form various carboaluminate
phases. The amount of sulfate in the system controls the extent of limestone’s reactivity
(Figure 2.11). As the sulfate content increases, the likelihood of unreacted calcite
increases.
Figure 2.11: Phase equilibrium of limestone (Matschei et al, 2007)
Thermodynamic modeling and experiments done by De Weerdt et al (2011) on
the hydration of OPC and OPC-Fly ash cements containing limestone powder showed
that at 1 day, the hydrates formed are similar for all combinations, namely C–S–H,
48
portlandite, and ettringite. However, at later ages, when the reaction of the clinkers the
kind and amount of phases start to differ between the limestone containing and
limestone-free OPC and OPC-FA blends. They concluded that in the absence of
limestone powder, ettringite decomposes to monosulfate whereas in the presence of
limestone powder, the decomposition of ettringite to monosulfate is prevented as
monosulfate is rendered unstable and instead calcium mono- or hemicarboaluminate are
formed.
As for the C3S content of cement, it has been suggested that some CaCO3 can be
incorporated in C-S-H formed by C3S (Ramachandran, 1988). A Study by Pera et al
(1999) indicated that hydration of C3S in the presence of CaCO3 results in the production
of some calcium carbosilicate hydrate and good mechanical performance for amounts of
CaCO3 higher than 30%. Kakali et al. (2000) and Voglis et al (2005) reported that the
addition of CaCO3 accelerates the hydration of C3S and results in the formation of
calcium carbosilicate hydrate. Nonetheless, the formation of ettringite is under debate.
Tsivilis et al (1998) and Kakali et al (2000) suggested that calcium carbonate suppresses
the formation of ettringite, whereas Ingram et al (1990) found that ettringite formation
proceeded normally. However, other researchers (Ramachandran, 1988; Pera et al, 1999;
Voglis et al, 2005) found acceleration in the formation of ettringite.
The effect of limestone additions on the formation of calcium hydroxide, CH, is
also debated. According to Turker and Erdoğdu (2000), the production of CH is enhanced
at early ages partially due to dissolution of limestone and also due to limestone’s ability
to act as nucleation sites. Barker and Cory (1991) reported enhanced formation of
calcium hydroxide at early ages with 5% and 25% limestone. However, Tsivilis et al
(1998) suggested that the CH content is slightly lower and independent of the limestone
content in limestone cements at early ages. Yet, at ages more than 1 day, the CH content
was found to be higher in limestone cements, specifically for limestone content of 10%.
On the other hand, experiments of Voglis et al (2005) showed that addition of 15%
limestone did not affect the amount of CH formed.
49
2.3.2. Compressive Strength and Porosity
As Hooton et al (2007) have stated, the strength of concrete produced with
limestone cement is strongly dependent on the quality of the limestone used, the
manufacturing process (blending versus intergrinding) and the final particle size
distribution of the cement. It is generally accepted that limestone additions up to 5% may
increase early-age compressive. This could be attributed to improve in particle packing
(Ellerbrock et al, 1990; Sprung and Siebel, 1991), increase in the rate of cement
hydration (Soroka and Setter, 1977; Bonavetti et al, 2003), and early production of
calcium carboaluminate (Voglis et al. 2005), or possibly a combination of all. However,
replacement levels of higher than 5% seem to lead to loss in strength due to dilution
(Hooton et al, 2007). Therefore, finer grinding of limestone is proposed as a means to
compensate this strength reduction.
Several studies have examined the permeability, or resistance of concrete to the
movement or penetration of fluids, of samples made with Portland-limestone cements
(Tezuka, 1992; Matthews, 1994; Irassar et al, 2001; Tsivilis et al, 2003; Dhir et al, 2007).
Overall, it seems that concrete produced with Portland-limestone cements containing up
to 15% limestone will produce concrete with similar resistance to the penetration of
fluids (Hooton et al, 2007). It is expected that Portland cement and PLC concretes give
similar performance when they are proportioned to give the same compressive strength at
28 days.
Few studies have focused on the relationship between strength and porosity of
samples incorporating limestone additions. According to Tsivilis et al (2003), concrete
made with Portland-limestone cements that contain up to 15% limestone had the same
porosity as the ordinary Portland cement concrete. However, further increase of the
limestone content caused a relative increase of the concrete porosity. In another study,
Tsivilis et al (2002) observed that the addition of 10% limestone does not significantly
alter the compressive strength of the tested samples. In fact, it increased slightly for 5%
limestone content, and decreased to the same value as pure cement for 10% limestone.
Nevertheless, further increase of the limestone content led to the production of limestone
50
cements having compressive strength lower than the pure ones. In addition, they found
that specimens with limestone had lower porosities compared with those of the Portland
cement specimen.
Guemmadi et al (2008) reported that when limestone was used in cement pastes,
porosity of the pastes decreased with the addition of the limestone fillers up to 15%. In
other words, the optimal limestone content that allowed the obtaining of highest
resistance was found to be 15%. Moreover, the cement pastes containing limestone gave
a higher initial strength than those of the ordinary cement with the amount of limestone
up to 15%. They attributed this phenomenon to the acceleration effect of limestone
related to the formation of calcium carboaluminate hydrate, which could be explained by
the increased binding capacity of carboaluminate due to its compact structure.
De Weerdt et al (2011) concluded that when limestone is present in the mix, the
stabilization of ettringite results in an increase in the volume of hydration products, and
thus a decrease in porosity and an increase in compressive strength. Their experimental
work showed a clear increase in compressive strength when 5% limestone was added to
OPC and OPC-Fly ash mixes; however, they stated that the experimental determination
of the porosity of the cement paste did not show a clear difference between the blends
with and without limestone powder due the relatively large error associated with the
measurements.
Experimental results of studies done by Herfort (2009), as shown in Figure 2.12,
revealed that the compressive strength of cement mortars with limestone additions
increases as the limestone content increased up to about 2%. However, further increase in
the limestone content decreased the compressive strength of the mortars in a way that
addition of about 14% limestone seemed to keep the compressive strength unaffected.
The same trend was found for the calculated relative porosity of the specimens; it
decreased as the limestone content increased up to 2%, and then started decreasing with
further addition of limestone.
51
-15
-10
-5
0
5
10
15
0 2 4 6 8 10 12 14 16 18 20 22
amount of CaCO3 added [wt.-%]
rela
tive c
han
ge o
f p
oro
sit
y a
nd
co
mp
ressiv
e s
tren
gth
[%
]
decrease
increase
compressive strength measured
total porosity calculated
Figure 2.12: Relative compressive strength and porosity of PLC samples (Herfort, 2009)
2.3.3. Effect of SCMs
As several researchers have proposed, when Portland-limestone cements are used,
or more generally in the presence of limestone, the hydration reactions are changed.
Apparently limestone reacts with the alumina phases of cement and produces
monocarboaluminates, which would contribute to the strength by means of producing
more hydrates and filling the space. As Herfort (2009) has proposed, the maximum
strength and minimum porosity (as shown in Figure 2.13) corresponds to the point where
all the available alumina is used up. In other words, the additional limestone would not
react to form hydrates (ettringite and monocarboaluminate), and therefore, further
addition of limestone would not increase in strength. However, if additional aluminate is
provided, more limestone could participate in the reactions, which implies that the
optimum limestone content (for optimum strength and porosity) would increase
respectively. One possible case of this would be the addition of supplementary
cementitious materials. By adding some types of SCMs, the amount of available alumina
for reactions would increase, which could lead into more reactions products and result in
52
more strength and less porosity. In that case, the optimum level of limestone addition for
strength and porosity would also increase. Hence, addition of SCMs would allow for
more limestone content to be added in order to achieve the optimum strength. This
hypothesis is schematically portrayed in Figure 11 by the dotted line.
Figure 2.13: Hypothetical effect of availability of more alumina on strength of PLC (Herfort, 2009)
According to Damidot et al (2100), carbonates replace the sulfates in the
hydration reactions, and since this reaction may take place after several days or weeks, it
can be used to optimize late strengths. They found that the optimum strength
corresponded to about 2 to 3% limestone for ordinary Portland cements with normal C3A
contents. Based on this, they predicted that optimum late strengths could be achieved at
higher limestone contents if the aluminate content could be increased. They suggested
adding aluminosilicate supplementary cementitious materials, namely metakaolin, clay,
and blast furnace slag as sources of alumina for medium to long term hydration. Their
calculations showed that the optimum limestone contents for optimum strengths were in
-15
-10
-5
0
5
10
15
0 2 4 6 8 10 12 14 16 18 20 22
amount of CaCO3 added [wt.-%]
rela
tive c
han
ge o
f p
oro
sit
y a
nd
co
mp
ressiv
e s
tren
gth
[%
]
decrease
increase
compressive strength measured
total porosity calculated
53
the region of 15% when 30% metakaolin was used, and about 10% when a 2:1 clay
(calcined smectitte) was used (Figure 2.14).
Figure 2.14: 28 day EN 196 mortar strengths for Portland blended cement predicted from the thermodynamic model (Damidot et al 2011)
It is useful to evaluate the aforementioned hypothesis. In this regard, Portland-
limestone cements having various amounts of limestone contents, along with different
combinations of some supplementary cementitious materials, can be used to make some
samples. Once the compressive strength and porosity of the samples are measured, it
would be possible to develop a similar graph. This could provide valuable information
with regards to the range of limestone content that does not negatively affect the
compressive strength and porosity of samples, and also the optimum level of limestone
content with and without SCMs. If proven, this theory suggests that better performance is
expected from Portland-limestone cements with higher levels of limestone (between 5 to
15%) when combined with SCMs. Therefore, this is beneficial not only in terms of
durability, but also from an economical point of view.
54
3. Experiments
As stated in the previous chapters, the main objective of this study is to examine
the sulfate resistance of Portland-limestone cements and their combinations with
supplementary cementitious materials (SCMs). According to the CSA A3000-08
standard, all types of Portland cements are allowed to have up to 5% ground limestone. In
addition, a new class of Portland-limestone cements (PLC) has been approved by the
committee with a maximum addition rate of 15% interground limestone, designated as
GUL (General Use with Limestone), HEL (High Early-Strength with Limestone), MHL
(Moderate-Heat with Limestone), and LHL (Low-Heat with Limestone). However, both
CSA A3000-08 and CSA A23.1-09 do not allow the use of Portland-limestone cements in
moderate or severe sulfate exposures, mainly due to mixed information in the literature
and concerns over thaumasite form of sulfate attack. Moreover, the effectiveness of
improving the sulfate resistance of Portland-limestone cements in conjunction with SCMs
requires further research.
In this study, the sulfate resistance of one cement clinker interground to make five
types of cements with different limestone contents and their combinations with various
levels of SCMs was examined. A new test method was adopted to study the sulfate
resistance of cements in regards with the thaumasite form of sulfate attack. Moreover, the
progression of sulfate attack on samples made with Portland-limestone cements was
evaluated. Finally, the properties of combinations of Portland-limestone cements and
SCMs in terms of hydration, strength, and porosity were evaluated. Details of the
experimental program for this research are described in the following sections.
3.1. Materials
In January 2008, Holcim Canada (then St. Lawrence Cement) performed a series
of test grinds at its Mississauga plant and shipped five different types of cement along
with slag to the University of Toronto for the present study. The cements delivered were
ASTM Type I, CSA Types GU, GUL11, GUL13, and a non-standard GUL22, which had
limestone contents of 0%, 2.4%, 10.6%, 12.7%, and 21.8% respectively. This covers a
55
relatively wide range of limestone content, from 0 to about 22%, which includes four in
the allowable range of 15%. It should be noted that these cements were obtained by
intrergrinding limestone with a 12% C3A cement clinker. The chemical compositions and
physical characteristics of the cements are presented in Table 3.1.
Table 3.1: Chemical compositions and physical characteristics of the used cements
Type I GU GUL11 GUL13 GUL22
Chemical
Composition (%)
SiO2 20.61 19.67 18.88 18.46 17.44
Al2O3 5.52 5.35 5.06 4.97 4.64
Fe2O3 2.19 2.13 2.02 1.98 1.86
CaO 63.36 62.72 61.86 61.34 59.79
MgO 2.41 2.40 2.31 2.35 2.29
SO3 4.17 4.73 4.31 4.39 4.07
K2O 1.22 1.18 1.15 1.12 1.07
Na2O 0.23 0.24 0.21 0.22 0.20
TiO2 0.257 0.246 0.239 0.233 0.217
SrO 0.087 0.084 0.083 0.082 0.078
P2O5 0.130 0.125 0.124 0.118 0.115
Cl 0.029 0.020 0.030 0.023 0.020
Free CaO 0.88 1.26 1.11 1.12 0.86
LOI 0.58 1.43 4.47 5.23 8.90
Limestone 0 2.4 10.6 12.7 21.8
Physical
Characteristics
Blaine (m2/kg) 402 391 507 515 562
Retained 45µm (%)
8.56 10.95 10.91 15.05 16.66
Specific Gravity - - 3.09 3.08 3.03
As for the supplementary cementitious materials, a slag from Holcim Canada and
a metakaolin from Whitemud Resources Inc. were used in this study. The slag was used
56
in tests for sulfate resistance as well as compressive strength, porosity, and hydration.
However, the metakaolin was only used in the compressive strength, porosity, and
hydration tests. The chemical compositions of the slag and the metakaolin are presented
in Table 3.2.
Table 3.2: Chemical compositions of the slag and metakaolin
Chemical
Composition (%) Slag Metakaolin
SiO2 38.14 61
Al2O3 7.18 34
Fe2O3 0.74 1.65
CaO 39.95 0.20
MgO 10.57 0.65
SO3 2.71 -
K2O 0.46 1.25
Na2O 0.33 0.13
TiO2 0.31 0.66
P2O5 0.02 -
Cr2O3 0.04 -
MnO 0.64 -
LOI 0.27 0.53
3.2. Tests on Sulfate Resistance
The sulfate resistance of the Portland-limestone cements and their combinations
with slag were tested. Most of the current standard methods for sulfate resistance of
cement mainly concern conventional sulfate attack. Therefore, a modified method, as
prescribed by CSA A3004 for the performance of cements in regards with possibility of
thaumasite sulfate attack was implemented. The following sections discuss details of
these test methods.
57
3.2.1. Conventional Sulfate Attack
The ASTM C 1012, “Test Method for Length Change of Hydraulic Cement
Mortars to a Sulfate Solution”, or the CSA 3004-C8 (Procedure A) which is an adapted
version of it, is predominantly used for estimating the potential for sulfate attack for
cementitious mixes. In this method, the change in the length of cement mortars immersed
in a sulfate solution is used as an indication of potential for sulfate attack. One of the
main advantages of this method, compared to other methods for this purpose, is its
applicability to blended hydraulic cements and blends of Portland cement with
supplementary cementitious materials. As such, this method was used to evaluate the
resistance of limestone cement mortars against conventional sulfate attack.
For this purpose, mortar mixtures proportioned as 1 part of cement, 2.75 parts of
graded sand, and water-to-cement of 0.485 were cast. Each mix consisted of 9 mortar
cubes of 50×50×50 mm and 6 mortar bars of 25×25×285 mm in size. The mortar bars
were mixed using a mechanical mixer and cast into metal molds. Immediately after
molding, the molds were covered, placed in the curing tank, and stored in a 38°C oven
for 24 hours. After that, the molds were removed from the tank and the specimens were
demolded. After demolding, all of the bars and cubes, except two of the cubes, were
stored in a curing tank of saturated limewater at 23°C. The compressive strength of the
two cubes were determined as per ASTM C 109 (or CSA A3004-C2) standard. If the
mean strength of the two samples was 20 MPa or more, it indicated that the mortars were
set to be exposed to sulfate solution. Otherwise, the same procedure was repeated every
24 hours until a compressive strength of 20 MPa was reached.
Once the samples gained the required strength, the initial length of the mortar bars
was measured and recorded using a comparator (Figure 3.1). Then, the bars were
immersed in a 50 g/L sodium sulfate solution. In compliance with the standard, the
temperature of the solution was kept at 23°C, and the volume proportion of the solution
to mortar bars was kept within the specified range of 4 to 1. Subsequent measurements of
the length of the mortar bars were performed at 1, 2, 3, 4, 8, 13, and 15 weeks after
immersion in solution. Thereafter, the measurements were done in intervals of about 28
58
days (4 weeks). Specific measurements at 6, 12, 15, 18, and 24 months were carried out,
where applicable, to obtain values that could be compared with the specifications of
standards for sulfate-resistant cement. After each reading, the sulfate solution was
changed and the samples were stored in new solution prepared in advance. Moreover, the
mass of the mortar bars were also recorded at each reading.
Figure 3.1: Comparator used to measure the length of mortar bars
In order to evaluate effectiveness of using slag to improve the sulfate resistance of
Portland-limestone cements, the same test method was performed on various
combinations of the used cements with slag. The slag replacement levels were chosen to
be 0% (as control), 30%, and 50% by mass of the cement. Except for the mass of cement,
all the other conditions of the test were kept similar. It should be noted that in mixes
containing slag, the time required for the samples to reach 20 MPa was greater due to
secondary hydration reactions.
59
3.2.2. Thaumasite Sulfate Attack
The CSA A3004-C8 was modified in 2010 to describe two different procedures
for evaluating the sulfate resistance of Portland cements. Procedure A, as in the previous
section, is used to determine resistance of Portland cement to conventional sulfate attack
(ettringite form of sulfate attack in the standard) in moderate or severe sulfate exposures.
Procedure B is used to determine resistance to the potential for the thaumasite form of
sulfate attack. It is applicable for testing blended Portland-limestone cements or Portland-
limestone cements used together with supplementary cementing materials to evaluate the
resistance of the combinations to the thaumasite form of sulfate attack. The main
difference between the two procedures is that in Procedure B, the mortar bars are stored
in sulfate solution at 5°C rather than 23°C. This low temperature would favor the
formation of thaumasite in sulfate attack. Also, the limestone content of the Portland-
limestone cement would act as the internal source of carbonate ions contributing to the
formation of thaumasite.
The same mixes as used for Procedure A were cast for Procedure B. All the test
parameters, including mix proportions, casting, molding, and curing were identical.
Similar to the previous tests, slag replacement levels of 0% (as control), 30%, and 50%
by mass of the cement was used to prepare samples. Once the samples gained 20 MPa,
they were transferred to a refrigerator and kept there for 24 hours to reach 5°C. Next, the
initial length of the mortar bars was recorded and they were immersed in a 5°C sodium
sulfate solution. The solution was prepared in advance and stored in a refrigerator for 24
hours. The containers were then stored in a refrigerator set at 5°C (Figure 3.2).
Subsequent measurements of the length of the mortar bars were performed similar to
Procedure A. Again, the sulfate solution was changed after each reading and the samples
were stored in 5°C solutions prepared at least 24 hours prior to the measurement. The
temperature of the refrigerator was checked periodically using a thermometer to make
sure that it remains within the range of 3 to 7°C, as required by the standard. The mass of
the mortar bars were also recorded at each reading.
60
Figure 3.2: Mortar samples stored in 5°C refrigerator
61
3.3. Studying Thaumasite Sulfate Attack
As explained in the previous chapter, the mechanism of the thaumasite form of
sulfate attack is not clearly understood. Several mechanisms have been proposed by
researchers; yet, there seems to be no unanimity on this issue in the literature (refer to
section 2.1.2 of Chapter 2). Therefore, it is useful to study the evolution of thaumasite
sulfate attack, which could possibly provide a better understanding of this phenomenon.
The literature review suggested that samples made with Portland-limestone
cements and stored in sulfate solutions at cold temperatures are prone to thaumasite
sulfate attack. Moreover, the purpose of the CSA A3004-C8 Procedure B method is to
evaluate the resistance of Portland-limestone cements to the potential for the thaumasite
form of sulfate attack. Therefore, it is likely that the samples used in this study may suffer
from deterioration due to thaumasite. Hence, it seemed appropriate to examine samples
made with PLC at various stages of exposure to sulfate solution at low temperatures.
For this purpose, two sets of mortar bars, each consisting of six bars, were cast
using each of the Type I and GUL22 cements. These two cements were selected as the
two extremes for limestone content, namely 0% and 21.8%, among the cements used in
this research. The proportioning of the materials for casting the bars was exactly the same
as that of the CSA A3004-C8 Procedure B test. The only difference was that the samples
were cured at 23°C (rather than for 38°C) for 24 hours after casting. The samples were
then immersed in 5% sodium sulfate solution, and stored in a refrigerator at 5°C. The
sulfate solution was replaced with new pre-cooled solution every 4 weeks. A small piece
(about 1/8 of the length of the mortar bar) was cut and taken out of each set of the
samples every 7 days after exposure to sulfate solution. Also, a piece was taken
immediately after curing and before exposure to the solution. These pieces were tested
using X-Ray Diffraction (XRD) analysis to identify the phases present in the samples. By
this means, the compounds formed at different stages of sulfate exposure were
determined.
For the XRD tests, the small pieces of the samples were first dried in a vacuum
62
oven at 38°C for a minimum of 24 hours. Then, the sand part of the samples was
selectively removed by crushing the samples and sieving them through a 315µm and an
80µm sieve respectively. Next, the remainder was gently ground using a mortar and
pestle, and then sieved to obtain a powder with particle size of less than 45µm. This
powder was tested with the XRD equipment to identify its components. The XRD
equipment used was a Philips PW 1730 X-ray diffractometer with copper radiation,
CuKα (wavelength λ = 1.5418 Ǻ) over an angular rotation (2θ) of 5 to 65°.
3.4. Compressive Strength, Porosity, and Hydration
The same five Portland-limestone cements were used in these tests. In addition, in
order to achieve a cement with 5% limestone content, the GU and GUL11 cement were
combined with the appropriate proportion (68.3% of GU and 31.7% of GU11) to achieve
5% limestone content. Similarly, the GUL13 and the GUL22 cements were combined to
result in a cement mixture which contained 15% limestone. As such, seven values of
limestone contents were in this program, as summarized in the Table 3.3.
Table 3.3: The cements used and their correspondent limestone content
Mix Cement Constituents
Limestone
Content
(%)
1 Type I Type I – 100% 0
2 GU GU – 100% 2.4
3 GUL5 GU – 68.3% + GUL11 – 31.7% 5
4 GUL11 GUL11 – 100% 10.6
5 GUL13 GUL13 – 100% 12.7
6 GUL15 GUL13 – 74.7% + GUL22 – 25.3% 15
7 GUL22 GUL22 – 100% 21.8
63
Slag and Metakaolin were used as supplementary cementitious materials in these
tests. The replacement levels for slag were 0% (as control), 15%, 30%, and 50% by mass
of the cement. In addition to that, a 10% metakaolin replacement level was used.
3.4.2. Compressive Strength
Mortar cubes were prepared in accordance with the ASTM C 109 Standard (or
CSA A3004-C2), “Standard Test Method for Compressive Strength of Hydraulic Cement
Mortars (Using 2-in. or [50-mm] Cube Specimens)”. For this purpose, a constant water to
cement ratio of 0.485 was used throughout the investigation. Also, the proportions for the
mortars were 1 part of cement to 2.75 parts of graded standard sand by mass. The
amount of materials was selected such as to be sufficient to cast 12 mortar cubes. The
mortars were mixed in a mechanical mixer, and the specimens were then cast in 50 mm
cube moulds, covered properly, and placed in a moist room. After 24 hours, the samples
were removed from the molds and stored in saturated limewater at 23 ± 2°C. The
compressive strength of the mortar cubes was determined at 28 and 56 days. At each age,
6 of the cubes were tested according to the standard, and the average value was obtained.
3.4.3. Porosity
The porosity and pore size distribution of mortar samples was determined using
the Mercury Intrusion Porosimetry (MIP). Mercury intrusion porosimetry is based on the
principle that a non-wetting liquid will only intrude capillaries under pressure. The
relationship between the pressure and capillary diameter is described by Washburn
equation:
dP
θγ cos4−=
where P is the pressure, γ is the surface tension of the liquid, θ is the contact angle
of liquid, and d is the diameter of the capillary. In this method, the liquid non-wetting
mercury is intruded into a porous system. Since the applied pressure is inversely
proportional to the size of the pores, the pore size distribution is determined from the
64
volume intruded at each pressure increment. The total porosity is determined from the
total volume intruded.
For casting samples for MIP, the same procedure as the samples cast for
compressive strength, was followed. At each age of testing, a 50 mm cube was removed
from the limewater solution and placed in room temperature until dry. Next, the sample
was crushed using a mortar and pestle, and then sieved to obtain particles of 2mm in size
from the middle of the specimen. After that, the samples were flash frozen at −195°C by
immersion in liquid nitrogen for 5 minutes. The frozen samples were then placed in a
vacuum freezer at −20°C for a minimum of 24 hours. The total porosity and pore size
distribution of the samples were then measured using a mercury intrusion porosimeter,
Quantachrome AUTOSCAN 60. This equipment applies pressure in the range from 0 to
60,000 psi (414 MPa) and capable of measuring pore size diameter down to 2 nm.
3.4.4. Hydration
The hydration products of Portland-limestone cements and their combinations
with slag were studied. For this purpose, the Type I (0% limestone) and GUL13 cements
were used. The slag replacement levels chosen were 0%, 15%, 30%, and 50%. For each
set, two 25 by 25 by 250-mm paste bars were cast and stored at 23 ± 2°C for 24 hours.
The water to binder ratio of the mixes was 0.5. After demolding, the bars were stored in
saturated limewater at 23 ± 2°C until the time of testing. The testing ages for the samples
were 1, 7, 14, 28, and 56 days. At these ages, a small piece (about 1/8 of the length of the
bar) was cut and taken out of each set of samples. The pieces were dried in a vacuum
oven at 38°C for a minimum of 24 hours. Next, the samples were crushed, ground, and
sieved using a mortar and pestle to obtain a powder with particle size of less than 45µm.
This powder was tested using the XRD equipment to identify its components.
65
4. Results and Discussion
This chapter deals with presentation of results obtained from the experiments and
the relevant discussions. Each section of this chapter reports the results from a specific
part of the experiments, as explained in the previous chapter. Data from different
components of the experimental works are analyzed, and the findings are integrated so
that a better comprehension of the properties of Portland-limestone cements is achieved.
4.1. Sulfate Resistance Tests
The results for the tests on sulfate resistance of Portland-limestone cements and
their combinations with slag are presented. It includes results for tests regarding
conventional sulfate attack and thaumasite sulfate attack. Discussions on the performance
of PLC in sulfate exposures and the effect of limestone in this regard are provided.
Prior to discussing the results, it is essential to review the specifications of the
current standards for sulfate resistance of cements. By comparing the results obtained in
this study to the requirements of the standards, it is possible to evaluate whether Portland-
limestone cements meet the requirements of the standards.
As mentioned earlier, there was no standard method for evaluating cement mixes
with regards to potential exposure to thaumasite sulfate attack prior to this study. As a
result of the present study, a new test method was proposed for evaluating the resistance
of Portland-limestone cements against the thaumasite form of sulfate attack. This method
was a modified version of the CSA A3004-08-A (or ASTM C 1012) method where the
samples are stored at 5°C. After using the proposed test method in this study and based
on the preliminary results, the Canadian standard was modified to incorporate the
proposed test method for resistance of Portland-limestone cements against thaumasite
sulfate attack. The new test method was listed as CSA A3004-08-B in the latest version
of the standard.
66
Table 4.1 summarizes the requirements of the 2010 amendment to CSA A3000-08
for sulfate resistance of cements. According to the standard, for a Portland-limestone
cement or its blends with SCMs to be classified as moderate-sulfate resistant, the average
expansion of mortar bars shall be less than 0.10% at 6 months at 23°C (CSA A3004-C8,
Procedure A) and less than 0.10% at 18 months at 5°C (CSA A3004-C8, Procedure B).
For high-sulfate resistance, the limits are 0.05% at 6 months for Procedure A and 0.10%
at 18 months for Procedure B. It adds that if the 23°C expansion is greater than 0.05% at
6 months but less than 0.10% at 1 year, the cement shall be considered to have passed.
Moreover, for Procedure B, if the increase in expansion between 12 and 18 months
exceeds 0.03%, the sulfate expansion at 24 months shall not exceed 0.10% in order for
the cement to be deemed to have passed the sulfate resistance requirement.
Table 4.1: Current CSA A3004-08 requirements for sulfate resistance of PLC
Property MSLb HSLb Reference
Sulfate Resistance Maximum % Expansion at 6 Months
0.10 0.05* CSA A3004-C8 Procedure A
Sulfate Resistance Maximum % Expansion at 18 Months at 5°C*** 0.10** 0.10** CSA A3004-C8
Procedure B
*If the expansion is greater than 0.05% at 6 months but less than 0.10% at 1 year, the cement shall be considered to have passed. ** Sulfate resistance testing shall be run on MSLb and HSLb cement at both 23 °C and 5 °C as specified in the table. This requirement does not apply to MSb and HSb cement. In addition, MSLb and HSLb cements require special minimum proportions. *** If the increase in expansion between 12 and 18 months exceeds 0.03%, the sulfate expansion of 24 months shall not exceed 0.10% in order for the cement to be deemed to have passed the sulfate resistance requirement.
According to the ASTM C 1157-10, “Standard Performance Specification for
Hydraulic Cement”, for a cement to be considered moderate sulfate-resistant, its ASTM
C 1012 expansion shall be less than 0.10% at 6 months. For a high sulfate-resistant
cement, the expansion limits are 0.05% at 6 months and 0.10% at 1 year. However,
testing of 1 year is not required if the cement meets the 6 month limit, and the cement is
not rejected unless it fails both criteria.
67
The ACI 318-08, “Building Code Requirements for Structural Concrete and
Commentary”, has similar requirements for concrete exposed to sulfate exposures based
on the severity of the environment, as presented in Table 4.2. It specifies that
combinations of cementitious materials shall be permitted when tested for sulfate
resistance and meeting the required criteria. In other words, blends of Portland cements
with supplementary cementitious materials are allowed to be used provided that the meet
the requirements for sulfate resistance.
Table 4.2: ACI 318-08 requirements for sulfate resistance
Maximum expansion when tested using ASTM C 1012
Exposure Class
At 6 Months At 12 Months At 18 Months
S1 0.10% - -
S2 0.05% 0.10% -
S3 - - 0.10%
4.1.1. Sulfate Attack at 23°C
4.1.1.1. 100% Cement at 23°C
The results for the average change length of the mortar bars made with 100%
Portland and Portland-limestone cements are presented in Table 4.3. Figure 4.1 shows the
expansion of the samples. According to the results, all the samples exceeded the 0.10%
expansion limit before 6 months. Thus, none of the samples could be considered as
sulfate-resistant. This matches the expectation in the sense that while the C3A content of
the cements was high, no supplementary cementitious materials were used to improve the
resistance against sulfate attack. The values for 6 month expansions show that as the
limestone content of the cement increased, the expansion increased. Figure 4.2 compares
the time to failure (i.e. time to reach the 0.10% expansion limit) of the samples. Based on
the results, the time to failure of the samples decreased with increase in the limestone
68
content of the cements. In other words, the sulfate resistance of PLC decreased with
increasing limestone content within 6 months of exposure.
Table 4.3: Length change of CSA A3004-A mortar bars made with 100% cement
Average Expansion (%)
Time of
Exposure
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
3 0.037 0.057 0.051 0.066 0.207
6 0.188 0.267 0.271 0.485 0.562
12 1.690 1.554 1.854 2.976 1.479
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Av
era
ge
le
ng
th c
ha
ng
e (
%) Type I
(0% L)
GU (2.4% L)
GUL 11(10.6% L)
GUL 13(12.7% L)
GUL 22(21.8% L)
Figure 4.1: Expansion of CSA A3004-A mortar bars made with 100% cement
69
0
20
40
60
80
100
120
140
160
180
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Tim
e t
o F
ailu
re (
Days)
Figure 4.2: Time to failure of CSA A3004-A mortar bars made with 100% cement
It should be noted that although none of the samples passed the test at 6 months,
the test was carried on to monitor the degradation of the samples. After 1 year of
exposure, the expansion of all samples had exceeded 1.5%. Among the five different sets,
the GUL22 and GU samples had the lowest expansions, whereas the GUL13 and GUL11
samples showed the highest expansions respectively. Nevertheless, at this point and
beyond it, signs of severe deterioration including cracks, distortion, and falling off pieces
was observed in all samples. In fact, many of the mortar bars were bent and distorted. As
such, the results of expansion measurements were not necessarily representative for the
purpose of comparison (thus denoted as apparent expansion hereon). At some point after
16 months and before 18 months of exposure, all the mortars were severely cracked, and
it was impossible to continue the measurements. As shown in Figure 4.3, the severity of
cracking was related to the amount of apparent expansion, i.e. the GUL22 samples were
distorted and the GU samples were distorted and cracked, whereas the GUL13 and
GUL11 samples were cracked into pieces.
70
Figure 4.3: CSA A3004-A mortar bars made with 100% cement after 18 months of exposure at 23°C (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
In order to identify the phases formed, X-Ray Diffraction was carried out on the
mortar samples after 18 months of exposure. For this purpose, a small cross section was
cut off a mortar bar from every set and dried in a vacuum oven at 38°C for 24 hours.
Then, the pieces were crushed and sieved to obtain a powder with particle size of less
than 45µm. Finally, they were tested in the XRD equipment and the phases were
identified using the analysis software.
The results obtained from XRD analysis of the samples are shown in Figures 4.4
to Figure 4.8. According to the results, the main phases formed are more or less the same
in all samples. As expected, ettringite, gypsum, and calcite were observed in all samples.
Portlandite was only observed in the GU and GUL22 samples. This is in accordance with
their lower apparent expansions and the visual condition. It was explained in the previous
71
chapters that as the sulfate attack progresses, the sulfate ions attack and react with the
products of hydration and form new compounds. Specifically, the sulfate ions react with
portlandite to form gypsum. As such, less portlandite was expected to be found in
severely deteriorated samples. The results show that in the GU and GUL22 samples
which had the least apparent expansion among all, there was still some portlandite
present. However, all the portlandite in the Type I, GUL11, and GUL13 samples had
turned into gypsum as no portlandite was found in the XRD results. It should be noted the
quartz found in the XRD analysis of these samples, as well as any other mortar samples
in this study, originate from the sand particles used in making the mortars.
It is interesting that no thaumasite was found in any of the samples. This
demonstrates that the governing type of destruction is actually conventional sulfate attack
(as opposed to thaumasite sulfate attack). At 23°C, the samples were deteriorated as a
result of expansion due to the formation of ettringite and gypsum, rather than thaumasite.
This finding is in line with the literature, which suggests that formation of thaumasite is a
slow process, and it is favoured at lower temperatures.
Figure 4.4: XRD analysis of 100% Type I mortar after 18 months of exposure to sulfate solution at 23°C
72
Figure 4.5: XRD analysis of 100% GU mortar after 18 months of exposure to sulfate solution at 23°C
Figure 4.6: XRD analysis of 100% GUL11 mortar after 18 months of exposure to sulfate solution at 23°C
73
Figure 4.7: XRD analysis of 100% GUL13 mortar after 18 months of exposure to sulfate solution at 23°C
Figure 4.8: XRD analysis of 100% GUL22 mortar after 18 months of exposure to sulfate solution at 23°C
74
4.1.1.2. 70% Cement and 30% Slag at 23°C
The results for the average length change of the mortar bars made with 70%
Portland and Portland-limestone cement and 30% slag are presented in Table 4.4. Figure
4.9 shows the expansion of the samples after 2 years of exposure to sulfate solution, and
the comparison between the expansions of different sets based on the limestone content
of the cement is shown in Figure 4.10. According to the results, after 2 years (730 days)
of exposure, all the sets of samples had average expansions of less than 0.10%. In other
words, all the combinations of the cements with 30% slag are highly sulfate-resistant.
The samples were in good condition after 2 years of exposure (see Figure 4.11), and only
minor signs of deterioration in the form of narrow cracks appeared on the surface of the
mortar bars. Hence, it is concluded that this combination of the Portland and Portland-
limestone cements with slag was able to pass the sulfate resistance test; i.e. replacing the
cement with 30% slag is effective in controlling the expansion due to sulfates at 23°C,
regardless of the limestone content of the cement.
The results suggest that in general, the average expansion of the mortars bars
decreased as the limestone content of the cement increased for limestone contents less
than 13%. The GUL13 and GUL11 samples had the lowest expansions among the
samples, which shows that that the combination of Portland-limestone cements with
limestone contents of less than 15% and slag works better in terms of controlling the
expansion compared to ordinary Portland cements. Statistical analysis of the results
indicated that the difference between the expansions of GUL11 and GUL13 is not
significant at 95% confidence level, but the difference between any two other
combinations of sets were found to be significant. Based on this, it could be inferred that
as the limestone content of the cement increases, the sulfate resistance of its combination
with 30% slag increases up to about 13% limestone. Increasing the limestone content
further depresses the sulfate resistance of the combination. The limestone content of
about 11% to 13% seems to be optimum in this regard.
75
Table 4.4: Length change of CSA A3004-A mortar bars made with 70% cement and 30% slag
Average Expansion (%)
Time of
Exposure
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
6 0.040 0.036 0.031 0.029 0.030
12 0.053 0.044 0.038 0.038 0.039
18 0.065 0.050 0.045 0.043 0.049
24 0.081 0.054 0.049 0.048 0.061
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0 100 200 300 400 500 600 700 800
Time (days)
Av
era
ge
le
ng
th c
ha
ng
e (
%) Type I
(0% L)
GU (2.4% L)
GUL 11(10.6% L)
GUL 13(12.7% L)
GUL 22(21.8% L)
Figure 4.9: Expansion of CSA A3004-A mortar bars made with 70% cement and 30% slag
76
0
0.02
0.04
0.06
0.08
0.1
0.12
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Ex
pa
ns
ion
(%
)
6
Months
12
Months
18
Months
24
Months
Figure 4.10: Expansion of CSA A3004-A mortar bars made with 70% cement and 30% slag
77
Figure 4.11: CSA A3004-A mortar bars made with 70% cement and 30% slag after 24 months of exposure at 23°C (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
XRD analysis was performed on the samples to identify the phases present in the
mortar bars after 2 years of exposure, the results of which are shown in Figures 4.12 to
4.16. The main phases identified in the samples were calcite and portlandite, and quartz.
Also, some ettringite and gypsum were found in all the samples; however the
concentration of these compounds were very low compared to the other phases, as the
intensity of the peaks indicate. Also, comparison between the intensity of peaks suggests
that the concentration of ettringite and gypsum in the 30% slag mixes are extremely low
compared to those of 100% cement mixes. This verifies the ability of slag to improve the
sulfate resistance of cement mixes.
78
Figure 4.12: XRD analysis of 70% Type I and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C
Figure 4.13: XRD analysis of 70% GU and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C
79
Figure 4.14: XRD analysis of 70% GUL11 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C
Figure 4.15: XRD analysis of 70% GUL13 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C
80
Figure 4.16: XRD analysis of 70% GUL22 and 30% slag mortar after 2 years of exposure to sulfate solution at 23°C
4.1.1.3. 50% Cement and 50% Slag at 23°C
The results for the change in the length of mortar bars made with 50% Portland
and Portland-limestone cement and 50% slag are presented in Table 4.5. Figure 4.17
shows the expansion of the samples after 2 years of exposure to sulfate solution, and the
comparison between the expansions of different sets based on the limestone content of
the cement is shown in Figure 4.18. According to the results, after 2 years (730 days) of
exposure, all the sets of samples had average expansions of less than 0.10%. Therefore,
all the combinations of the Portland and Portland-limestone cements with 50% slag were
highly sulfate-resistant. The samples were in good condition after 2 years of exposure
(see Figure 4.19), and only minor signs of deterioration in the form of narrow cracks
appeared on the surface of the mortar bars. Hence, it is concluded that this combination
of the Portland and Portland-limestone cements with slag was able to pass the sulfate
81
resistance test; i.e. replacing the cement with 50% slag is effective in controlling the
expansion due to sulfates.
The results for the 50% mixes suggest that the average expansion of the mortar
bars decreased as the limestone content of the cement increased for limestone contents
less than 11%, and increased thereon. Moreover, it is understood from the results that the
combination of Portland-limestone cements with 50% slag works better in terms of
controlling the expansion compared to the 30% slag mixes or ordinary Portland cements.
Statistical analysis of the results indicated that the difference between the expansions of
GUL11, GUL13, and GUL22 is not significant at 95% level of confidence, but the
difference between these and the other two combinations of Type I and GU were found to
be significant. Based on this, it could be inferred that as the limestone content of the
cement increases, the sulfate resistance of its combination with 50% slag increases up to
about 11% limestone addition. Further increase in the limestone content of the cement
seems to be ineffective in the sulfate resistance of the cement when combined with 50%
slag. Nevertheless, the difference between the expansions of the Type I and GU samples
and the other three sets of samples were very small.
Table 4.5: Length change of CSA A3004-A mortar bars made with 50% cement and 50% slag
Average Expansion (%)
Time of
Ponding
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
6 0.025 0.026 0.016 0.018 0.018
12 0.032 0.028 0.021 0.022 0.022
18 0.037 0.031 0.024 0.024 0.025
24 0.038 0.034 0.027 0.027 0.028
82
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0 100 200 300 400 500 600 700 800
Time (days)
Avera
ge len
gth
ch
an
ge (
%)
Type I(0% L)
GU (2.4% L)
GUL11(10.6% L)
GUL13(12.7% L)
GUL22(21.8% L)
Figure 4.17: Expansion of CSA A3004-A mortar bars made with 50% cement and 50% slag
0
0.02
0.04
0.06
0.08
0.1
0.12
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Exp
an
sio
n (
%)
6
Months
12
Months
18
Months
24
Months
Figure 4.18: Expansion of CSA A3004-A mortar bars made with 50% cement and 50% slag at 23°C
83
Figure 4.19: CSA A3004-A mortar bars made with 50% cement and 50% slag after 24 months of exposure (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
It should be noted that these results matched the expectation since replacing 30%
of cement with slag was already proven to be effective in improving the sulfate resistance
of the used cements, and further increase in the slag replacement was expected to
improve the sulfate resistance even more. In fact, a comparison between the expansion of
the 30% slag mixes and the 50% slag ones, as shown in Figure 4.20 suggests that the
expansion of the 50% slag mixes are approximately half of that of the 30% slag mixes for
all the Portland and Portland-limestone cements used.
84
0
0.02
0.04
0.06
0.08
0.1
0.12
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Exp
an
sio
n (
%) 30%
Slag
50%
Slag
Figure 4.20: Comparison of expansion of mortar bars made with 30% and 50% slag at 23°C after 2 years
XRD analysis was performed on the samples to identify the phases present in the
mortar bars after 2 years of exposure, the results of which are shown in Figures 4.21 to
4.25. Similar to the 30% slag mixes, the main phases identified in the 50% slag samples
were calcite and portlandite, and quartz. Also, some ettringite and gypsum were found in
all the samples; however the concentration of these compounds were very low compared
to the other phases, as the intensity of the peaks indicate. Comparison between the
intensity of peaks suggest that the concentration of ettringite and gypsum in the 50% slag
mixes are significantly lower than those of 100% cement mixes, and relatively lower than
those of the 30% slag mixes.
85
Figure 4.21: XRD analysis of 50% Type I and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C
Figure 4.22: XRD analysis of 50% GU and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C
86
Figure 4.23: XRD analysis of 50% GUL11 and 50% slag mortar after 2 years of exposure to sulfate solution at 23°C
Figure 4.24: XRD analysis of 50% GUL13 and 50% slag mortar after 2 years of exposure to sulfate solution 23°C
87
Figure 4.25: XRD analysis of 50% GUL22 and 50% slag mortar after 2 years of exposure to sulfate solution 23°C
4.1.1.4. Overall Comparison
Figures 4.26 and 4.27 show the expansion of the CSA A3004-A mortar bars with
different slag replacement levels based on the limestone content of the cement. Figures
4.28 and 4.29 also show the effect of slag replacement on the expansion for the cements
used in this study. As the graphs show, replacing 30% or 50% of the Portland cements or
Portland-limestone cements with slag was effective in controlling the expansion due to
sulfate attack at 23°C. While the 100% cement bars had exceeded the 0.10% expansion
limit by 6 months of exposure, the 30% and 50% slag mixes had expansions less than
0.10% after 12 months of exposure. This verifies that when combined with a sufficient
amount of slag, Portland-limestone cements have good performance comparable to that
of Portland cements in sulfate exposures at 23°C. In other words, replacing cement with
slag is an effective way in increasing its resistance against sulfate attack at 23°C for both
Portland cements and Portland-limestone cements.
88
0.01
0.10
1.00
10.00
0 5 10 15 20 25
Limestone Content (%)
Exp
an
sio
n (
%)
(Lo
g)
0%
Slag
30%
Slag
50%
Slag
Figure 4.26: 6 month expansion of CSA A3004-A mortar bars with different slag replacement levels
0.01
0.10
1.00
10.00
0 5 10 15 20 25
Limestone Content (%)
Exp
an
sio
n (
%)
(Lo
g)
0%
Slag
30%
Slag
50%
Slag
Figure 4.27: 12 month expansion of CSA A3004-A mortar bars with different slag replacement levels
89
0 30 50
0.01
0.10
1.00
10.00
Slag Content (%)
Avera
ge E
xp
an
sio
n (
%)(
Lo
g)
Type I
(0% L)
GU
(2.4% L)
GUL11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
Figure 4.28: 6 month expansion of CSA A3004-A mortar bars with different limestone contents
0 30 50
0.01
0.10
1.00
10.00
Slag Content (%)
Avera
ge E
xp
an
sio
n (
%)(
Lo
g)
Type I
(0% L)
GU
(2.4% L)
GUL11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
Figure 4.29: 12 month expansion of CSA A3004-A mortar bars with different limestone contents
90
4.1.2. Thaumasite Sulfate Attack
4.1.2.1. 100% Cement at 5°C
The results for the average length change of mortar bars made with 100%
Portland and Portland-limestone cement and stored at 5°C are shown in Table 4.6 and
Figure 4.30. According to the results, all these samples reached the 0.1% expansion limit
before 1 year, and most of them reached the limit even before 6 months. In addition to
that, after 1 year of exposure, all the samples had completely disintegrated and turned
into mush. Figure 4.31 presents the time taken for each of the samples to reach the 0.1%
expansion limit (denoted as failure) and to become completely disintegrated. Once a
sample had disintegrated, it was impossible to continue the test. After complete
disintegration, each set of samples consisted of a mixture of separated mortar pieces and
a fluid mush (see Figure 4.32).
Based on the results, it is evident that the time to failure/disintegration is inversely
proportional to the limestone content of the cement. In other words, as the limestone
content of the cement increases, the samples come to failure and disintegration at earlier
ages. While the Type I cement samples (0% limestone) exceeded the 0.1% expansion
limit after 266 days and were disintegrated after 1 year of exposure, the GUL22 samples
had already passed the limit by 56 days and became disintegrated only after 90 days of
exposure.
Table 4.6: Length change of CSA A3004-B mortar bars made with 100% cement
Average Expansion (%)
Time of
Exposure
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
3 0.048 0.050 0.061 0.082 X
6 0.067 0.097 1.260 X X
9 0.140 0.605 X X X
12 X X X X X
91
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
0 50 100 150 200 250 300 350 400
Time (days)
Avera
ge
len
gth
ch
an
ge (
%)
Type I(0% L)
GU (2.4% L)
GUL 11(10.6% L)
GUL 13(12.7%L)
GUL 22(21.8% L)
Figure 4.30: Length change of CSA A3004-B mortar bars made with 100% cement
0
50
100
150
200
250
300
350
400
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Tim
e (
Days)
Disintegration
Failure
(>0.1% Expansion)
Figure 4.31: Time to failure/disintegration of CSA A3004-B mortar bars made with 100% cement
92
Figure 4.32: CSA A3004-B mortar bars made with 100% cement after disintegration at 5°C (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
Based on the observations, it was suspected that the disintegration of the samples
were due to thaumasite sulfate attack. In order to verify this theory and to identify the
phases formed, XRD was carried out on the samples upon disintegration. For each set of
samples, two specimens were selected for this purpose, one from the mortar pieces and
one from the mush. XRD analysis was performed on both the pieces and the mush
samples.
Figures 4.33 to 4.37 present the results obtained from XRD analysis of the
samples. According to the results, the main phases formed are more or less the same in all
93
samples. As for the mush, thaumasite and ettringite were the main phases observed in all
samples. As the limestone content increased, the amount of calcite found in the mush part
of the samples also increased, which is in accordance with the limestone content of the
cements. As for the mortar pieces, all the samples contained ettringite, gypsum,
portlandite, and calcite. Based on this, it is concluded that the formation of thaumasite
was the final cause of disintegration of the samples and turning them into mush.
Figure 4.33: XRD analysis of 100% Type I mortar after 365 days of exposure to sulfate solution at 5°C
94
Figure 4.34: XRD analysis of 100% GU mortar after 330 days of exposure to sulfate solution at 5°C
Figure 4.35: XRD analysis of 100% GUL11 mortar after 210 days of exposure to sulfate solution at 5°C
95
Figure 4.36: XRD analysis of 100% GUL13 mortar after 180 days of exposure to sulfate solution at 5°C
Figure 4.37: XRD analysis of 100% GUL22 mortar after 90 days of exposure to sulfate solution at 5°C
96
The results of the XRD tests confirmed that the disintegration of the samples was
due to thaumasite sulfate attack. It is interesting that all the samples, including those
without any limestone were destroyed. This shows that there must be sufficient minor
carbonate in the cement or water to allow thaumasite sulfate attack to occur. Therefore,
this test is likely more severe than in real conditions where thaumasite sulfate attack is
relatively rare.
4.1.2.2. 70% Cement and 30% Slag at 5°C
The results for the average change in length of mortar bars made from 70%
Portland and Portland-limestone cement and 30% slag are presented in Table 4.7. Figure
4.38 shows the expansion of the samples after 2 years of exposure to sulfate solution, and
the comparison between the expansions of different sets based on the limestone content
of the cement is shown in Figure 4.39. According to the results, after 18 months of
exposure, the GUL 13 and GUL22 bars exceeded the 0.10% expansion limit; therefore,
they have failed the test and cannot be considered as sulfate-resistant mixes. However,
the Type I, GU, and GUL11 mixes have expansions less than 0.10% after 18 months of
exposure. As mentioned before, the standard has stated that if the increase in expansion
between 12 and 18 months exceeds 0.03%, the sulfate expansion at 24 months shall not
exceed 0.10%. The difference in expansion of the Type I, GU, and GUL11 mixes
between 12 and 18 month are 0.027%, 0.021%, and 0.049% respectively. Therefore, the
GUL11 mix also failed the test. Nonetheless, based on the specifications of the standard,
the Type I and GU samples can be considered as sulfate-resistant at 5°C. It is concluded
that the combination of the Portland-limestone cements with 30% slag is not effective in
controlling the expansion.
97
Table 4.7: Length change of CSA A3004-B mortar bars made with 70% cement and 30% slag
Average Expansion (%)
Time of
Ponding
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
6 0.028 0.031 0.031 0.032 0.036
12 0.038 0.042 0.044 0.072 0.190
18 0.065 0.063 0.093 0.237 0.642
24 0.235 0.152 0.345 0.889 X
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
0 100 200 300 400 500 600 700 800 900
Time (days)
Av
era
ge
le
ng
th c
ha
ng
e (
%) Type I
(0% L)
GU (2.4% L)
GUL 11(10.6%L)
GUL 13(12.7%L)
GUL 22(21.8% L)
Figure 4.38: Length change of CSA A3004-B mortar bars made with 70% cement and 30% slag
98
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Ex
pa
ns
ion
(%
)
6
Months
12
Months
18
Months
24
Months
Figure 4.39: Expansion of CSA A3004-B mortar bars made with 70% cement and 30% slag at 5°C
Visual examination of the samples after 2 years of exposure shows that the
samples underwent serious deterioration, in accordance with the amount of expansion.
The GUL22 samples had completely cracked and fallen into pieces; the GUL13 samples
were cracked at several spots; and the rest of the samples showed longitudinal cracks,
worn pieces on the edges, and some softening (see Figure 4.40). Although the Type I and
GU samples have technically passed the test, the condition of the samples and the rate of
expansion suggest that they should not be used in such an exposure. According to the
CSA A3001 standard, a minimum of 40% slag is required for sulfate resistant cements to
be used in such exposures, and therefore, these mixes would not be allowed based on the
standard. However, it is interesting that although the difference in the expansion of these
two mixes between 12 and 18 month is less than 0.030%, visually these two sets are not
adequately resistant against this type of exposure to sulfate solution. This suggests that
perhaps the specification of the CSA standard could be revised, and a more conservative
approach be adopted by the standard. As mentioned, the difference between the
99
expansion of the Type I and the GU samples at 12 and 18 month of exposure was 0.027%
and 0.021% respectively. Therefore, if the 0.030% change in expansion required in the
standard was changed to 0.020%, then both of the sets would have passed this limit and
failed the test. Alternatively, the absolute expansion could be changed to less than 0.10%
at 24 months to be considered sulfate-resistant. According to Table 4.7, the expansions of
these two sets at 24 months were 0.24% and 0.15% respectively, both beyond the 0.10%
limit. As such, if the 0.10% limit at 24 months was adopted, they both would be
considered to have failed the test.
Figure 4.40: CSA A3004-B mortar bars made with 70% cement and 30% slag after 24 months of exposure at 5°C (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
100
Considering the visual examination and the significant rate of expansion of the
samples at the later stages of the test, the possibility of deterioration due to thaumasite
sulfate attack seemed to be significant. Therefore, XRD analysis was performed on a
piece taken from the cross section of each of the sets of samples. The results of the XRD
analyses are shown in Figure 4.41 to Figure 4.45. Based on the results, the main phases
present in the samples were ettringite, thaumasite, gypsum, portlandite, and calcite. The
intensity of the peaks pertaining to thaumasite increased as the limestone content of the
cement increased, indicating a higher concentration of thaumasite in the sample. This is
in accordance with the amount of expansion of the samples and the visual signs of
deterioration. Therefore, as the presence of thaumasite was confirmed, it is expected that
all these samples would disintegrate with increased time of sulfate exposure, similar to
those with no slag. Once again, this suggests that replacing 30% of the cement with slag
is not sufficient to satisfactorily increase the resistance of Portland and Portland-
limestone cements, especially those with high limestone contents (namely 11% and
higher) against the thaumasite form of sulfate attack. This finding confirms the
prescription of the CSA standard which does not allow mixes with less than 40% slag to
be used in environments where deterioration due to thaumasite sulfate attack is a concern.
Figure 4.41: XRD analysis of 70% Type I and 30% slag mortar after 2 years of exposure at 5°C
101
Figure 4.42: XRD analysis of 70% GU and 30% slag mortar after 2 years of exposure at 5°C
Figure 4.43: XRD analysis of 70% GUL11 and 30% slag mortar after 2 years of exposure at 5°C
102
Figure 4.44: XRD analysis of 70% GUL13 and 30% slag mortar after 2 years of exposure at 5°C
Figure 4.45: XRD analysis of 70% GUL22 and 30% slag mortar after 2 years of exposure at 5°C
103
4.1.2.3. 50% Cement and 50% Slag at 5°C
Table 4.8 shows the results for average length change of the mortar bars made
from mixes with 50% Portland and Portland-limestone cement and 50% slag by mass.
Figure 4.46 shows the expansion of the samples after 2 years exposure to sulfate solution,
and the comparison between the expansions of different sets based on the limestone
content of the cement is shown in Figure 4.47. According to the results, after 2 years (730
days) of exposure, all the sets of samples had average expansions of less than 0.10%.
Therefore, all the combinations of the Portland and Portland-limestone cements with 50%
slag are highly sulfate-resistant (CSA Type HSb). The samples were in good condition
after 2 years of exposure (see Figure 4.48). The results indicate that in general, the
average expansion of the mortar bars decreases as the limestone content of the cement
increases for limestone contents less than 13%. Statistical analysis of the results indicated
that the difference between the expansions of GU and GUL22 is not significant at 95%
level of confidence but the difference between any two other combinations of sets were
found to be significant. Based on this, it could be inferred that as the limestone content of
the cement increases, the sulfate resistance of its combination with 50% slag increases up
to about 13% limestone. Increasing the limestone content to 22% depresses the sulfate
resistance of the combination. The limestone content of about 13% seems to be optimum
in this regards.
Table 4.8: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag at 5°C
Average Expansion (%)
Time of
Ponding
(months)
Type I
(0% L)
GU
(2.4% L)
GUL 11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
6 0.016 0.014 0.010 0.011 0.016
12 0.025 0.025 0.017 0.018 0.023
18 0.027 0.027 0.021 0.022 0.026
24 0.037 0.032 0.027 0.026 0.031
104
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 100 200 300 400 500 600 700 800
Time (days)
Ave
rag
e l
en
gth
ch
an
ge (
%)
Type I(0% L)
GU (2.4% L)
GUL 11(10.6% L)
GUL 13(12.7% L)
GUL 22(21.8% L)
Figure 4.46: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag
0
0.02
0.04
0.06
0.08
0.1
0.12
0 2.4 10.6 12.7 21.8
Limestone Content (%)
Exp
an
sio
n (
%)
6Months
12
Months
18Months
24
Months
Figure 4.47: Length change of CSA A3004-B mortar bars made with 50% cement and 50% slag
105
Figure 4.48: CSA A3004-B mortar bars made with 50% cement and 50% slag at 5°C after 24 months of exposure (from left to right; top: Type I, GU, GUL11; bottom: GUL13, GUL22)
XRD analysis was performed on the samples to identify the phases present in the
mortar bars after 2 years of exposure, the results of which are shown in Figures 4.49 to
4.53. The main phases identified in the samples were calcite and portlandite, and quartz.
Also, some ettringite and gypsum were found in all the samples; however the
concentration of these compounds were very low compared to the other phases, as the
intensity of the peaks indicate. It should be noted that no evidence of thaumasite was
found in any of the 50% slag mortars, regardless of the limestone content of the cement.
106
Figure 4.49: XRD analysis of 50% Type I and 50% slag mortar after 2 years of exposure at 5°C
Figure 4.50: XRD analysis of 50% GU and 50% slag mortar after 2 years of exposure at 5°C
107
Figure 4.51: XRD analysis of 50% GUL11 and 50% slag mortar after 2 years of exposure at 5°C
Figure 4.52: XRD analysis of 50% GUL13 and 50% slag mortar after 2 years of exposure at 5°C
108
Figure 4.53: XRD analysis of 50% GUL22 and 50% slag mortar after 2 years of exposure at 5°C
4.2. Evolution of Thaumasite Sulfate Attack
As the results in the previous sections indicated, the mortar bars made with
Portland cements and Portland-limestone cements and stored in sulfate solution at 5°C
deteriorated due to thaumasite sulfate attack. Also, it was mentioned that the mechanism
of the thaumasite form of sulfate attack is not clearly understood. Hence, it seemed
appropriate to examine samples made with PLC exposed to sulfate solution at this
temperature. The Type I and GUL22 cements, containing 0% and 21.8% limestone
respectively were chosen for this purpose. This study consisted mainly of using XRD
analysis at various stages of exposure to the solution. The following sections discuss the
results obtained.
4.2.1. Type I Mortar Bars at 5°C
Figure 4.54 shows the result of XRD analysis on the Type I sample immediately
after demolding and prior to storage in the sulfate solution. The main phase observed, as
109
expected, was portlandite, which is a product of hydration. Calcium silicate was also
observed which relates to the unhydrated particles of cement. Calcite and ettringite were
also observed in this sample.
Figure 4.54: XRD analysis of 100% Type I mortar after demolding
Figures 4.55 to 4.61 show XRD results of the Type I sample after increasing
periods of exposure from 28 days up to 210 days. Based on the results, during this period,
the phases formed were in accordance with products of conventional sulfate attack. The
concentration of products such as ettringite and gypsum increased with time. Gypsum
was observed in the samples as early as 28 days after exposure to sulfate solution.
Moreover, the amount of portlandite present in the samples seemed to be decreasing with
time, especially after 180 days (6 months) of exposure. The samples were becoming
softer as the time of exposure increased, and it was much easier to cut a piece form the
mortar bars. In addition to that, as time passed, small pieces fell off the mortar bars and
accumulated at the bottom of the container; however, the amount was not enough to be
used for XRD analysis.
110
Figure 4.55: XRD analysis of 100% Type I mortar after 28 days of exposure to Na2SO4 at 5°C
Figure 4.56: XRD analysis of 100% Type I mortar after 56 days of exposure to Na2SO4 at 5°C
111
Figure 4.57: XRD analysis of 100% Type I mortar after 91 days of exposure to Na2SO4 at 5°C
Figure 4.58: XRD analysis of 100% Type I mortar after 120 days of exposure to Na2SO4 at 5°C
112
Figure 4.59: XRD analysis of 100% Type I mortar after 150 days of exposure to Na2SO4 at 5°C
Figure 4.60: XRD analysis of 100% Type I mortar after 180 days of exposure to Na2SO4 at 5°C
113
Figure 4.61: XRD analysis of 100% Type I mortar after 210 days of exposure to Na2SO4 at 5°C
After about 240 days of exposure (8 months), a relatively significant amount of
mush had accumulated at the bottom of the container. Therefore, it was possible to
perform the XRD analysis both on a sample of mush and a piece taken from intact mortar
bars. Figures 4.62 to 4.64 show the results of XRD analysis on the samples after 8, 9, and
10 months of exposure respectively. According to the results, thaumasite and ettringite
were the main phases in the mush whereas portlandite, calcite, gypsum, and ettringite
were the main phases in the mortar pieces. The concentration of thaumasite increased
with time in the mush part of the samples. Based on this, it can be concluded that as the
samples are stored in the solution, the onset of low-temperature sulfate attack is similar to
that of conventional sulfate attack. However, at some point, which appears to be between
7 to 8 months of exposure for Type I cement, the formation of thaumasite is initiated and
accelerates, causing disintegration of the samples. Once this happens, thaumasite
formation becomes the governing mode of deterioration, and it continues until the
samples become completely disintegrated.
114
Figure 4.62: XRD analysis of 100% Type I mortar after 240 days of exposure to Na2SO4 at 5°C
Figure 4.63: XRD analysis of 100% Type I mortar after 270 days of exposure to Na2SO4 at 5°C
115
Figure 4.64: XRD analysis of 100% Type I mortar after 300 days of exposure to Na2SO4 at 5°C
4.2.2. GUL22 Mortar Bars at 5°C
Figure 4.65 shows the result of XRD analysis on the GUL22 sample immediately
after demolding and prior to storage in the sulfate solution. The main phases observed, as
expected, were portlandite, and calcite, the former being a product of hydration and the
latter pertaining to the high limestone content of this type of cement. In fact, a
comparison between the intensity of the peak of calcite for this sample the Type I sample
clearly indicates that the limestone content of the GUL22 cement is significantly greater
than that of the Type I cement. Calcium silicate, which relates to the unhydrated particles
of cement, and ettringite were also observed in the GUL22 sample.
116
Figure 4.65: XRD analysis of 100% GUL22 mortar after demolding
Figures 4.66 and 4.67 show the XRD results of the GUL22 sample at 28 and 56
days after exposure. Based on the results, the phases formed during this period were in
accordance with products of conventional sulfate attack, similar to Type I. The
concentration of products such as ettringite and gypsum increased with time. The rate of
increase, however, was notably greater than that of the Type I samples. Gypsum was
observed in the samples as early as 28 days after exposure to sulfate solution. The
samples were becoming softer as the time of exposure increased, and it was much easier
to cut a piece form the mortar bars. In addition to that, as time passed, small pieces fell
off the mortar bars and accumulated at the bottom of the container; however, the amount
was not enough to be used for XRD analysis. These findings were very similar to the case
of Type I samples, except that they happened in a much shorter period of time.
117
Figure 4.66: XRD analysis of 100% GUL22 mortar after 28 days of exposure to Na2SO4 at 5°C
Figure 4.67: XRD analysis of 100% GUL22 mortar after 56 days of exposure to Na2SO4 at 5°C
118
After about 90 days of exposure (3 months), a relatively significant amount of
mush had accumulated at the bottom of the container. Therefore, it was possible to
perform the XRD analysis both on a sample of mush and a piece taken from the intact
mortar bars. Figures 4.68 to 4.72 show the results of XRD analysis on the samples after 3,
4, 5, 6, and 7 months of exposure respectively. According to the results, thaumasite and
ettringite were the main phases in the mush whereas portlandite, calcite, gypsum, and
ettringite were the main phases in the mortar pieces. The concentration of thaumasite
increased with time in the mush part of the samples. At 7 months, it was impossible to
completely differentiate between the mush and the pieces. Based on this, the conclusion
made from the experiments with Type I cement is reaffirmed; as the samples are stored in
the solution, the beginning phase of low-temperature sulfate attack is similar to that of
conventional sulfate attack. However, at some point in time, which appears to be between
2 to 3 months of exposure for the GUL22 cement, the formation of thaumasite is
accelerated and causes disintegration of the samples. Once this happens, thaumasite
formation becomes the governing mode of deterioration, and it continues until the
samples become disintegrated. Increasing limestone content of the cement accelerates the
damage due to thaumasite sulfate attack.
Figure 4.68: XRD analysis of 100% GUL22 mortar after 91 days of exposure to Na2SO4 at 5°C
119
Figure 4.69: XRD analysis of 100% GUL22 mortar after 120 days of exposure to Na2SO4 at 5°C
Figure 4.70: XRD analysis of 100% GUL22 mortar after 150 days of exposure to Na2SO4 at 5°C
120
Figure 4.71: XRD analysis of 100% GUL22 mortar after 180 days of exposure to Na2SO4 at 5°C
Figure 4.72: XRD analysis of 100% GUL22 mortar after 210 days of exposure to Na2SO4 at 5°C
121
4.2.3. Thaumasite Sulfate Attack and Expansion
The results for expansion of CSA A3004-B mortar bars were discussed in Section
4.1.2.1. Also, the evolution of thaumasite sulfate attack in the Type I and GUL22 mortar
bars were discussed in Sections 4.2.1 and 4.2.2. It is important to revisit the data gathered
for the expansion of the mortar bars and examine them in light of the findings of the
XRD tests on the samples. This provides useful information with regard to how
thaumasite sulfate attack and expansion are correlated.
Figure 4.73 shows the expansion of the Type I and GUL22 mortar bars with time,
as presented in Section 4.1.2.1. Considering the Type I cement samples, the trend shows
that in the first months of exposure (sections A to H of the figure), the samples expanded
at a relatively constant rate. Results from XRD (as presented in Section 4.2.1) show that
during this period, ettringite and gypsum were the main phases formed in the samples,
and their concentration increased with time (Figures 4.55 to 4.61). The expansion of
mortar bars exceeded the 0.10% limit after about 8 months of exposure. According to the
XRD results (Figure 4.62), thaumasite was detected in the samples at 240 days (8
months) after exposure, which coincided with the point where the expansion limit was
exceeded. Visual observation also showed that at this age, the bars started to fall apart; a
relatively significant amount of mush had accumulated at the bottom of the containers,
which contained more thaumasite than the remaining pieces of mortar bars. Beyond this
point, the rate of expansion of samples increased significantly (sections I to L of the
figure). XRD results revealed that the concentration of thaumasite and gypsum increased
during this period. This continued until the samples had completely disintegrated.
The results were similar for the GUL22 mortar bars. The main difference was that
the sequence of events happened much faster compared to the Type I samples. The
samples exceeded the 0.10% expansion limit after only 2 months of exposure. After this
point, the rate of expansion increased significantly. The results of the XRD tests show
that thaumasite was observed in the samples beyond this point (Figures 4.68 to 4.72).
Similar to the Type I cement samples, the concentration of thaumasite increased with
time until the samples had completely disintegrated.
122
Based on these findings, it is concluded that the formation of thaumasite occurs
at, or soon after, the samples exceed the 0.10% expansion limit. At this point, samples
showed considerable signs of deterioration; the mortar bars start falling apart and they
turned into a combination of mortar mortar pieces and soft mush. Beyond this point, the
concentration of thaumasite increased with time. This increase in the concentration of
thaumasite was accompanied by a significant increase in the rate of expansion of
samples. This continued until the samples completely disintegrated and turned into mush.
Therefore, the formation of thaumasite in these samples is an expansive phenomenon.
Moreover, the rate of deterioration due to thaumasite sulfate attack is in direct relation
with the limestone content of the cement.
Figure 4.73: Expansion of type I and GUL22 mortar bars exposed to Na2SO4 at 5°C
123
4.2.4. Progression of Thaumasite Sulfate Attack
In the course of the previous sections, it was explained that Portland cements and
Portland-limestone cements are prone to the thaumasite form of sulfate attack at low
temperatures. Also, the evolution of thaumasite sulfate attack and its various stages were
examined. It should be noted that studying the mechanism of thaumasite sulfate attack
was beyond the scope of the present research. It requires detailed investigation of the
phases formed along with visual observation of the samples in order to understand the
sequence of reactions taking place during the attack. Nevertheless, based on the limited
study on the evolution of thaumasite sulfate attack and the results of the sulfate resistance
tests, the series of events happening during the attack can be summarized as follows.
During the initial phase of the thaumasite form of sulfate attack, the sulfate ions
react with the cement paste which results in the formation of ettringite and gypsum. This
is accompanied with expansion in the samples, and they expand at a relatively constant
rate. In this period, cracking is the most visible sign of deterioration. Cracks usually
occur at the edges of the samples as well as longitudinally on the surface.
With further progression of the attack, small pieces start falling off the edges and
the surface of the samples. These falling pieces are generally soft in first place; however,
they become softer as time passes and eventually turn into a semi-fluid mush. This
procedure continues until a significant amount of mush is accumulated at the bottom of
the containers (Figure 4.74). Moreover, at this stage, the mortar pieces are generally
covered with a white soft material very similar in appearance to the mush.
In the meantime, more cracks occur on the edges and the surface of the samples.
At some point, which based on the observations of this study is fairly close to the
formation of mush, the expansion of the samples starts to increase with a considerable
rate. It seems that this stage coincides with the time when the samples exceed the 0.10%
expansion limit set by the standard. This extensive cracking and expansion causes the
mortar bars to be broken into pieces, which makes the measurement of length change
impossible (Figure 4.75)
124
Figure 4.74 : Mush accumulated at the bottom of container due to thaumasite sulfate attack
Figure 4.75 : Mortar pieces after extensive cracking due to thaumasite sulfate attack
125
Results of the XRD analysis on the samples revealed that the main phases present
in the mush were ettringite, thaumasite, and calcite, depending on the limestone content
of the cement. The mush seemed to contain none or insignificant amount of gypsum in its
composition. Moreover, although traces of portlandite might have been present in the
early stages of accumulation of mush, no portlandite was found at later ages. On the other
hand, the main phases observed in the mortar pieces were ettringite, gypsum, calcite, and
portlandite. Almost in all cases, no thaumasite was detected in the mortar pieces.
Based on these findings, it can be inferred that the formation of thaumasite mainly
takes place on the surface of the samples. It leads to a loss on cohesion of the cement
paste and to a loss of material from the sample. The absence of thaumasite in the mortar
pieces suggests that the period between the formation of thaumasite and the loss of
material (which mainly consists of thaumasite and ettringite) is short. The accumulation
of mush on the top surface of the samples also indicated that ettringite and thaumasite
coexist in the solution. This can support the theory explained in the literature that
ettringite and thaumasite are the end products from a solid solution, called woodfordite,
which occurs through the reaction between ettringite, silicate and carbonate in the
presence of excess water.
It is not possible to reach a firm conclusion, based on the findings of the present
study, whether thaumasite is formed through the direct route, the woodfordite route,
topochemical replacement of ettringite by thaumasite, the through solution mechanism, or
ettringite as nucleation sites for formation of thaumasite; however, study of the evolution
of thaumasite attack indicated that expansion related to the formation of ettringite and
gypsum preceded thaumasite formation. In fact, it seems that as Schmidt et al (2009)
have suggested, the opening up of the microstructure which is caused by extensive
cracking of the samples at the early stages of the attack is a prerequisite for the formation
of thaumasite.
126
4.3. Compressive Strength, Porosity, and Hydration
4.3.1. Compressive Strength
4.3.1.1. 100% Cement
The results for the compressive strength of the samples made with 100% cement
are presented in Table 4.9 and shown in Figure 4.76. As the results suggest, the
compressive strength of the samples increased with the limestone content up to 2.4%, and
it decreased thereon with the addition of limestone. In other words, the GU cement which
contains 2.4% limestone had the highest compressive strength among all the samples.
Compared to samples made with Type I, the GU samples had 19% and 14% more
strength at 28 days and 56 days respectively.
Statistical analysis of the results showed that the differences between the Type I,
GUL11 (10.6% limestone), GUL13 (12.7% limestone), and GUL15 samples at 28 days
were not significant at 95% level of confidence. Moreover, the differences between
results for Type I, GUL5, GUL11, GUL13, and GUL15 at 56 days were not significant at
a 95% confidence level. This suggests that limestone addition of up to about 13% does
not have a negative effect on the compressive strength of the samples. In other words, the
range of 0 to 13% limestone for optimized interground Portland-limestone cements does
not adversely affect the strength of samples.
Table 4.9: Compressive strength of mortar cubes made with 100% cement
Compressive Strength (MPa) Relative Difference in Compressive
Strength from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 39.3 43.0 0 0
2.4 46.7 48.9 19 14
5 43.2 45.1 10 5
10.6 41.8 44.3 6 3
12.7 38.2 41.8 -3 -3
15 37.7 39.2 -4 -9
21.8 33.2 34.4 -16 -20
127
0
10
20
30
40
50
60
0 2.4 5 10.6 12.7 15 21.8
Limestone Content of Cement (%)
Co
mp
ressiv
e S
tren
gth
(M
Pa)
28
Days
56
Days
Figure 4.76: Compressive strength of mortar cubes made with 100% cement
4.3.1.2. 85% Cement and 15% Slag
The results for the compressive strength of samples made with 85% cement and
15% slag are presented in Table 4.10 and shown in Figure 4.77. According to the results,
the compressive strength of the samples increased with limestone content up to about
11%, and it then decreased with further addition of limestone. The GUL11 cement which
contains 10.6% limestone had the highest compressive strength among all the samples.
Compared to samples made with Type I, the GUL11 samples had 15% and 17% higher
compressive strength at 28 days and 56 days respectively.
Statistical analysis of the results indicated that at 28 days, the differences between
the compressive strengths of the Type I, GUL15 and GUL22 samples were not significant
at a 95% confidence level. In addition to that, analysis of the results for 56 day strengths
reveals that the differences between the compressive strength of the GUL13 and GUL15
samples, and also between the Type I and GUL22 samples were not significant at a 95%
128
confidence level. As such, it can be concluded that when 15% of the cement is replaced
with slag, the range of limestone that does not affect the compressive strength negatively
is extended to about 22%. In other words, with 15% slag, the range of 0 to 22% limestone
for PLC does not adversely affect the strength of samples
Table 4.10: Compressive strength of mortar cubes made with 85% cement and 15% slag
Compressive Strength (MPa) Relative Difference in Compressive
Strength from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 40.2 42.5 0 0
2.4 42.7 46.5 6 9
5 44.3 47.4 10 12
10.6 46.1 49.0 15 15
12.7 42.6 45.0 6 6
15 40.3 44.6 0 5
21.8 39.6 42.4 -1 0
0
10
20
30
40
50
60
0 2.4 5 10.6 12.7 15 21.8
Limestone Content of Cement (%)
Co
mp
ressiv
e S
tren
gth
(M
Pa)
28
Days
56
Days
Figure 4.77: Compressive strength of mortar cubes made with 85% cement and 15% slag
129
4.3.1.3. 70% Cement and 30% Slag
The results for the compressive strength of samples made with 70% cement and
30% slag are presented in Table 4.11 and shown in Figure 4.78. Based on the results, it is
concluded that the compressive strength of the samples increased slightly as the
limestone content of the cement was increased from 0 to 2.4%, and it then decreased with
further addition of limestone. Compared to samples made with Type I, the GU samples
had only 2% and 1% more strength at 28 days and 56 days respectively.
Statistical analysis of the results indicated that at 28 days, the differences between
the compressive strength of the Type I, GUL5, and GUL11 samples were not significant
at 95% confidence level. Moreover, analysis of the results for the 56 day strength
revealed that the differences between the compressive strength of the Type I, GU, GUL5,
GUL11, and GUL13 samples were not significant. As such, it can be concluded that
when 30% of the cement is cement is replaced with slag, an increase in the limestone
content of the cement has no significant effect on the compressive strength of the samples
for limestone contents of up to about 13%.
Table 4.11: Compressive strength of mortar cubes made with 70% cement and 30% slag
Compressive Strength (MPa) Relative Difference in Compressive
Strength from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 46.7 50.9 0 0
2.4 47.6 51.2 2 1
5 46.4 50.3 -1 -1
10.6 46.3 50.2 -1 -1
12.7 45.5 50.0 -3 -2
15 44.4 48.8 -5 -4
21.8 41.7 46.3 -11 -9
130
0
10
20
30
40
50
60
0 2.4 5 10.6 12.7 15 21.8
Limestone Content of Cement (%)
Co
mp
ressiv
e S
tren
gth
(M
Pa)
28
Days
56
Days
Figure 4.78: Compressive strength of mortar cubes made with 70% cement and 30% slag
4.3.1.4. 50% Cement and 50% Slag
Table 4.12 and Figure 4.79 present the results for the compressive strength of the
samples made with 50% cement and 50%. According to the results, the general trend was
that the compressive strength of the samples decreased with the limestone content. It is
clear that the samples made with Type I cement (0% limestone) had the highest
compressive strength among all samples at both ages.
Statistical analysis of the results indicates that at both ages, the differences
between the compressive strength of the GU, GUL5, and GUL11, GUL13, and GUL15
samples were not significant at a 95% confidence level. Therefore, it can be concluded
that when 50% of the cement is replaced with slag, while the cement with no limestone
has the highest compressive strength (about 10% more than the average of the samples),
the limestone content of the cement has no significant effect on the compressive strength
of the samples (for limestone contents between 2.4% and 13%). In other words, the
131
compressive strength of the samples is independent of the limestone content of the
samples as long as it is between 2.4 and 15%. Nevertheless, it decreases when the
limestone content is increased beyond 15%.
Table 4.12: Compressive strength of mortar cubes made with 50% cement and 50% slag
Compressive Strength (MPa) Relative Difference in Compressive
Strength from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 43.4 48.5 0 0
2.4 39.5 43.9 -9 -9
5 39.3 43.9 -9 -9
10.6 39.5 43.9 -9 -9
12.7 39.5 43.4 -9 -10
15 39.0 42.9 -10 -11
21.8 33.8 37.3 -22 -23
0
10
20
30
40
50
60
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
Co
mp
ressiv
e S
tren
gth
(M
Pa)
28
Days
56
Days
Figure 4.79: Compressive strength of mortar cubes made with 50% cement and 50% slag
132
4.3.1.5. 90% Cement and 10% Metakaolin
Table 4.13 and Figure 4.80 present the results for the compressive strength of the
samples made with 90% cement and 10% Metakaolin (MK). According to the results, the
compressive strength of the samples increased with limestone contents up to about 11%,
and then decreased with further addition of limestone. The GUL11 cement which
contains 10.6% limestone had the highest compressive strength among all the samples.
Compared to samples made with Type I, the GUL11 samples had 13% and 12% higher
compressive strength at 28 days and 56 days respectively.
Statistical analysis of the results indicated that the differences between the
compressive strength of the Type I and GUl15 samples, the GU and GUL13 samples, and
the GUL5 and GUL11 samples were not significant at a 95% confidence level at both 28
and 56 days. As such, it can be concluded that when 10% of the cement is replaced with
metakaolin, the range of limestone which does not affect the compressive strength
negatively is about 15%.
Table 4.13: Compressive strength of mortar cubes made with 90% cement and 10% MK
Compressive Strength (MPa) Relative Difference in Compressive
Strength from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 43.2 42.7 0 0
2.4 44.8 44.2 4 4
5 47.9 47.7 11 12
10.6 48.8 47.9 13 12
12.7 45.1 44.6 5 4
15 43.0 42.0 -1 -2
21.8 40.7 40.7 -6 -5
133
0
10
20
30
40
50
60
0 2.4 5 10.6 12.7 15 21.8
Limestone Content of Cement (%)
Co
mp
ressiv
e S
tren
gth
(M
Pa)
28
Day
56
Day
Figure 4.80: Compressive strength of mortar cubes made with 90% cement and 10% MK
4.3.1.6. Summary
Figures 4.81 and 4.82 summarize the results for compressive strength of mortar
cubes made with Portland and Portland-limestone cements and supplementary
cementitious materials at 28 and 56 days. Also, Figures 4.81 and 4.82 show the effect of
various slag replacement levels on the compressive strength of the mortar cubes for the
cements at 28 and 56 days. According to the results shown in Figures 4.79 and 4.80, the
30% slag replacement level seems to be the most effective in improving the compressive
strength of mortar cubes. Except for the 50% slag mixes, using SCMs improved the
compressive strength of mortar cubes for all the Portland-limestone cements (limestone
contents greater than 5%). Also, by comparing the results for 100% cement mixes to the
15% slag mix, it is evident that the optimum level of limestone shifts from 2.4% in the
former to about 11% in the latter. A similar finding is also clear for the 10% metakaolin
mix.
134
The results shown in Figure 4.83 and 4.84 indicate that the effect of replacing
cement with slag on the compressive strength of mortar cubes was similar for the
Portland-limestone cements used in this study (GUL11, GUL13, and GUL22). As the
amount of slag increased from 0 to 30%, the compressive strength of the mortar cubes
made with Portland-limestone cements increased. However, further increasing the slag
replacement level from 30% to 50% resulted in a decrease in the compressive strength of
mortar cubes at 28 and 56 days.
0
10
20
30
40
50
60
0 5 10 15 20 25
Limestone Content (%)
Com
pre
ssiv
e S
trengh (M
Pa)
100%
Cement
15%
Slag
30%
Slag
50%
Slag
10% MK
Figure 4.81: 28 day compressive strength of mortar cubes made with different SCMs
135
0
10
20
30
40
50
60
0 5 10 15 20 25
Limestone Content (%)
Com
pre
ssiv
e S
trengh (M
Pa)
100%
Cement
15%
Slag
30%
Slag
50%
Slag
10% MK
Figure 4.82: 56 day compressive strength of mortar cubes made with different SCMs
0 15 30 50
0
10
20
30
40
50
60
Slag Replacement Level (%)
Co
mp
res
siv
e S
tre
ng
th (
MP
a)
Type I
(0% L)
GU
(2.4% L)
GUL11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
Figure 4.83: 28 day compressive strength of mortar cubes with various slag replacement levels
136
0 15 30 50
0
10
20
30
40
50
60
Slag Replacement Level (%)
Co
mp
res
siv
e S
tre
ng
th (
MP
a)
Type I
(0% L)
GU
(2.4% L)
GUL11
(10.6% L)
GUL13
(12.7% L)
GUL22
(21.8% L)
Figure 4.84: 56 day compressive strength of mortar cubes with various slag replacement levels
4.3.2. Porosity
4.3.2.1. 100% Cement
The results for the porosity of mortar samples made with 100% cement are
presented in Table 4.14 and shown in Figure 4.85. As the results suggest, the porosity of
the samples decreased with the limestone content up to 2.4%, and increased thereon with
further addition of limestone at both ages. In other words, the GU cement which contains
2.4% limestone had the lowest porosity among all the samples. This finding matches the
results for the compressive strength, i.e. the samples with the highest compressive
strength had the lowest porosity. At 28 days, only the GU samples had a porosity less
than that of the Type I samples; however, all the samples except the GUL22 ones had
porosities less than that of the Type I samples at 56 days.
137
Table 4.14: Total Porosity of mortar cubes made with 100% cement
Total Porosity (%) Relative Difference in Porosity
from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 10.6 9.8 0 0
2.4 10.3 8.1 -3 -17
5 11.2 9.3 6 -5
10.6 11.8 9.3 12 -5
12.7 12.1 9.63 14 -1
15 12.5 9.7 18 -1
21.8 12.6 10.4 19 6
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
To
tal
Po
rosit
y (
%)
28
Days
56
Days
Figure 4.85: Total porosity of mortar cubes made with 100% cement
138
4.3.2.2. 85% Cement and 15% Slag
The results for the porosity of samples made with 85% cement and 15% slag are
presented in Table 4.15 and shown in Figure 4.86. The porosity of the samples decreased
with limestone content up to about 11%, and then increased with further addition of
limestone. The GUL11 cement which contains 10.6% limestone had the lowest porosity
among all the samples. Once again, this finding matches the results for the compressive
strength in the sense that samples with the highest compressive strength had the lowest
porosity. The GU, GUL5, and GUL11 samples had a porosity less than that of the Type I
samples at both ages, whereas the porosity of the GUL13, GUL15, and GUL22 samples
were greater than that of Type I samples.
Table 4.15: Total Porosity of mortar cubes made with 85% cement and 15% slag
Total Porosity (%) Relative Difference in Porosity
from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 10.4 8.8 0 0
2.4 10.3 8.5 -1 -3
5 10.3 8.4 -1 -4
10.6 10.1 8.3 -3 -6
12.7 10.9 9.3 4 6
15 11.1 9.6 7 9
21.8 11.8 10.2 13 16
139
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
To
tal
Po
rosit
y (
%)
28
Days
56
Days
Figure 4.86: Total porosity of mortar cubes made with 85% cement and 15% slag
4.3.2.3. 70% Cement and 30% Slag
Table 4.16 and Figure 4.87 present the results for the porosity of samples made
with 70% cement and 30% slag. Based on the results, it is concluded that the porosity of
the samples increased slightly as the limestone content of the cement was increased from
0 to 2.4%, and then decreased with further addition of limestone. The samples made with
GU cement (2.4% limestone) had the lowest porosity among all the samples. This finding
is in accordance with the results for the compressive strength. At 28 days, the GU and
GUL5 samples had porosities less than that of the Type I samples; however, only the GU
samples had a porosity less than that of the Type I samples at 56 days.
140
Table 4.16: Total Porosity of mortar cubes made with 70% cement and 30% slag
Total Porosity (%) Relative Difference in Porosity
from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 10.1 9.2 0 0
2.4 9.8 9.1 -3 -1
5 10.0 9.3 -1 1
10.6 10.5 9.33 4 1
12.7 11.7 9.4 15 2
15 11.8 9.6 16 4
21.8 12.4 9.9 22 8
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
To
tal
Po
rosit
y (
%)
28
Days
56
Days
Figure 4.87: Total porosity of mortar cubes made with 70% cement and 30% slag
141
4.3.2.4. 50% Cement and 50% Slag
Table 4.17 and Figure 4.88 show the results for the porosity of samples made with
50% cement and 50% slag. Based on the results, the general trend was that the porosity of
the samples increased with increasing limestone content. The samples made with Type I
cement (0% limestone) had the lowest porosity among all samples at both ages, which
concurs with the results for the compressive strength where the Type I samples had the
highest values among all the samples. At both ages, all the samples had porosities greater
than that of the Type I samples. The porosities of the GU, GUL5, GUL11, GUL13, and
GUL15 samples seemed to be fairly close to each other compared to the Type I or
GUL22 samples. This suggests that when 50% of cement is replaced with slag, the
porosity of the samples remains relatively constant as long as the limestone content of the
cement is between 2.4 and 15%. It should be noted that a similar conclusion was drawn
for the compressive strength of these samples.
Table 4.17: Total Porosity of mortar cubes made with 50% cement and 50% slag
Total Porosity (%) Relative Difference in Porosity
from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 9.7 8.9 0 0
2.4 11.2 9.4 16 6
5 11.4 9.5 17 7
10.6 11.4 9.4 18 6
12.7 11.5 9.4 181 6
15 11.7 9.6 21 8
21.8 12.4 9.8 28 10
142
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
To
tal
Po
rosit
y (
%)
28
Days
56
Days
Figure 4.88: Total porosity of mortar cubes made with 50% cement and 50% slag
4.3.2.5. 90% Cement and 10% Metakaolin
The results for the porosity of samples made with 90% cement and 10%
metakaolin (MK) are presented in Table 4.18 and shown in Figure 4.89. The porosity of
the samples increased with limestone content up to about 5%, and then decreased with
further addition of limestone. The GUL5 cement had the lowest porosity among all the
samples at both ages. At 28 days, the GU and GUL5 samples had porosities less than that
of the Type I samples; however, the GU, GUL5, and GUL11 samples had porosities less
than that of the Type I samples at 56 days while the porosity of the GUL13 samples was
similar to the Type I sample at this age
143
Table 4.18: Total Porosity of mortar cubes made with 90% cement and 10% MK
Total Porosity (%) Relative Difference in Porosity
from 0% Limestone (%)
Limestone
Content (%) 28 Days 56 Days 28 Days 56 Days
0 10.3 9.8 0 0
2.4 10.1 9.4 -2 -4
5 10.0 9.3 -3 -5
10.6 10.5 9.5 2 -2
12.7 11.0 9.8 7 0
15 11.3 9.9 9 2
21.8 12.6 10.8 22 11
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 2.4 5 10.6 12.7 15 21.8
Limestone Content (%)
To
tal
Po
rosit
y (
%)
28
Days
56
Days
Figure 4.89: Total porosity of mortar cubes made with 90% cement and 10% MK
144
4.3.3. Combination of Compressive Strength and Porosity
4.3.3.1. 100% Cement
Figures 4.90 and 4.91 show the combination of results for the relative
compressive strength and porosity of mortar samples made with 100% Portland and
Portland-limestone cement at 28 and 56 days. The results clearly demonstrate a good
correlation between the compressive strength and porosity of the samples; as the
compressive strength increased, the porosity decreased and vice versa. Although the
levels at which the limestone content of the cement affected the compressive strength and
the porosity of the samples were not identical, especially at 28 days, the general trends
remained the same.
According to the results, as the limestone content of cement increased from 0 to
2.4%, the compressive strength of the samples increased while the porosity decreased.
With further addition of limestone, the compressive strength of the samples decreased
while the porosity increased. It suggests that the optimum level of limestone in terms of
the maximum compressive strength and minimum porosity is about 2.4%. One possible
explanation for this is that the limestone particles act as nucleation sites for hydration of
Portland cement phases. However, it could be explained by the previous mentioned
theory that limestone reacts with the alumina phases of cement and produces
carboaluminates, which contribute to the strength by means of producing more hydrates
and filling pore space. Once all the available alumina is consumed, the additional
limestone does not react to form hydrates, and therefore, further addition of limestone
would not contribute to increasing the strength or reducing the porosity. The results of the
tests on hydration, as will be discussed in the following sections, confirms the formation
of monocarboaluminate in Portland-limestone cement mixes (See section 4.3.4).
145
R2 = 0.98
R2 = 0.997
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
0 5 10 15 20 25
Limestone Content (%)
Rela
tiv
e C
han
ge
of C
om
pre
ss
ive
Str
en
gth
an
d
Po
rosity (%
)
28 DayStrength
28 DayPorosity(MIP)
Decrease
Increase
Figure 4.90: Relative strength and porosity of mortar cubes made with 100% cement at 28 days
R2 = 0.96
R2 = 0.98
-25
-20
-15
-10
-5
0
5
10
15
20
25
0 5 10 15 20 25
Limestone Content (%)
Rela
tive C
han
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%)
56 Day
Strength
56 Day
Porosity
(MIP)
Decrease
Increase
Figure 4.91: Relative strength and porosity of mortar cubes made with 100% cement at 56 days
146
4.3.3.2. 85% Cement and 15% Slag
Figures 4.92 and 4.93 show the combination of results for the relative
compressive strength and porosity of mortar samples made with 85% Portland and
Portland-limestone cement and 15% slag at 28 and 56 days. Once again, the inverse
correlation between the compressive strength and porosity of the samples is noticeable.
The levels at which the limestone content of the cement affected the compressive strength
and the porosity of the samples were not identical, especially at 28 days; however, the
general trends remained constant.
Based on the results, as the limestone content of cement increased from 0 to 11%,
the compressive strength of the samples increased while the porosity decreased. Further
addition of limestone resulted in decreased compressive strength and increased porosity
of the samples. The samples were divided into two sets; those with limestone contents
less than 11% and those with more than 11% limestone contents. Next, lines of best fit
were obtained for the results of each two sets. Based on these regression analyses, it is
concluded that the optimum level of limestone in terms of the maximum compressive
strength and minimum porosity is around 8%. This suggests that when slag is added to
the mix, more alumina would be available, which implies that besides acting as
nucleation sites, more limestone could participate in the hydration reactions. Hence, the
optimum limestone content (for maximum compressive strength and minimum porosity)
would shift to a higher value. In this experiment, the optimum limestone level increased
from 2.4% in the case no slag was added to about 8% when 15% slag was added to the
mix. In this scenario, once all the available alumina provided by the cement and slag is
consumed, the additional limestone does not react to form hydrates or to accelerate the
hydration of Portland cement, and therefore, further addition of limestone would not
contribute to increasing the strength or reducing the porosity.
147
R2 = 0.97
R2 = 0.70
R2 = 0.91
R2 = 0.92
-20
-15
-10
-5
0
5
10
15
20
0 5 10 15 20 25 30
Limestone Content (%)
Re
lati
ve C
ha
ng
e o
f C
om
pre
ss
ive S
tre
ng
th a
nd
Po
ros
ity
(%
)
28 DayStrength
28 DayPorosity(MIP)
Decrease
Increase
Figure 4.92: Relative strength and porosity of mortar cubes made with 85% cement and 15% slag at 28
days
R2 = 0.98
R2 = 0.78
R2 = 0.83
R2 = 0.96
-20
-15
-10
-5
0
5
10
15
20
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive
Ch
an
ge
of
Co
mp
ressiv
e S
tren
gth
an
d
Po
rosit
y (
%)
56 DayStrength
56 Day
Porosity(MIP)
Decrease
Increase
Figure 4.93: Relative strength and porosity of mortar cubes made with 85% cement and 15% slag at 56
days
148
4.3.3.3. 70% Cement and 30% Slag
The combination of results for the relative compressive strength and porosity of
mortar samples made with 70% Portland and Portland-limestone cement and 30% slag at
28 and 56 days is shown in Figures 4.94 and 4.95. According to the results, as the
limestone content of cement increased from 0 to 2.4%, the compressive strength of the
samples increased while the porosity decreased. With further addition of limestone, the
compressive strength of the samples decreased while the porosity increased. This
suggests that the optimum level of limestone in terms of the maximum compressive
strength and minimum porosity is about 2.4%. The explanation for this is that although
addition of slag provides more alumina to participate in hydration reactions with the
limestone content of the cement, since 30% of the cement is replaced with slag, the
cement content of the mix is reduced and less limestone is available to react with alumina
phases. Also, slag particles participate in secondary hydration reactions, i.e. reaction with
the products of hydration of cement clinker. Therefore, the effects of these reactions
balance out each other; the result of which is that the optimum level of limestone content
falls at 2.4%, similar to the case where no slag was used in the mixes.
The results clearly declare a good correlation between the compressive strength
and porosity of the samples; as the compressive strength increased, the porosity
decreased and vice versa. Lines of best fit were obtained for the results of samples with a
limestone content greater than 0. These lines show the trend of a decrease in compressive
strength and porosity with addition of limestone. It is interesting that for the 56 day
results, the slope of the regression line for compressive strength was 0.35 whereas the
slope of the best fit line for porosity was -0.33. This shows that the levels at which the
limestone content of the cement affected the compressive strength and the porosity of the
samples were almost identical.
149
R2 = 0.91
R2 = 0.92
-25
-20
-15
-10
-5
0
5
10
15
20
25
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive C
han
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%)
28 Day
Strength
28 DayPorosity(MIP)
Decrease
Increase
Figure 4.94: Relative strength and porosity of mortar cubes made with 70% cement and 30% slag at 28
days
R2 = 0.86
R2 = 0.89
-10
-8
-6
-4
-2
0
2
4
6
8
10
0 5 10 15 20 25 30
Limestone Content (%)
Re
lati
ve
Ch
an
ge
of
Co
mp
res
siv
e S
tre
ng
th a
nd
Po
ros
ity
(%
)
56 DayStrength
56 DayPorosity
(MIP)
Decrease
Increase
Figure 4.95: Relative strength and porosity of mortar cubes made with 70% cement and 30% slag at 56
days
150
4.3.3.4. 50% Cement and 50% Slag
Figures 4.96 and 4.97 show the combination of results for the relative
compressive strength and porosity of mortar samples made with 50% Portland and
Portland-limestone cement and 50% slag at 28 and 56 days. Once again, the inverse
correlation between the compressive strength and porosity of the samples is noticeable.
Based on the results, the general trend was that the compressive strength of the
samples decreased and the porosity increased with the limestone content. It is clear that
the samples made with Type I cement (0% limestone) had the highest compressive
strength and lowest porosity among all samples at both ages. However, the compressive
strengths of the GU, GUL5, GUL11, GUL13, and GUL15 were extremely close, and the
same was true for the porosities. This suggests that when 50% of the cement is replaced
with slag, while the cement with no limestone has the highest compressive strength and
lowest porosity, the limestone content of the cement has no significant effect on the
compressive strength or porosity of the samples (for limestone contents less than 15%).
The explanation for this is that when 50% of the cement is replaced with slag, the amount
of cement becomes significantly lower compared to the case where no slag is used. In
other words, there is less cement to be hydrated. On the other hand, since the contribution
of slag is mainly in the form of secondary hydration reactions, it mainly reacts with
products of hydration of the cement particles to form calcium silicate hydrates and fill the
voids. As such, the limestone content of the cement does not have a considerable role in
this whole process. In other words, the normal and secondary hydration reactions are the
prevailing and governing reactions as opposed to the reaction between the limestone from
the cement and alumina from the slag.
151
-30
-25
-20
-15
-10
-5
0
5
10
15
20
25
30
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive C
han
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%)
28 Day
Strength
28 DayPorosity
(MIP)
Decrease
Increase
Figure 4.96: Relative strength and porosity of mortar cubes made with 50% cement and 50% slag at 28 days
-20
-15
-10
-5
0
5
10
15
20
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive C
han
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%)
56 Day
Strength
56 Day
Porosity
(MIP)
Decrease
Increase
Figure 4.97: Relative strength and porosity of mortar cubes made with 50% cement and 50% slag at 56 days
152
4.3.3.5. 90% Cement and 10% Metakaolin
Figures 4.98 and 4.99 show the combination of results for the relative
compressive strength and porosity of mortar samples made with 90% Portland and
Portland-limestone cement and 10% metakaolin at 28 and 56 days. The results clearly
demonstrate a good correlation between the compressive strength and porosity of the
samples; as the compressive strength increased, the porosity decreased and vice versa.
Although the levels at which the limestone content of the cement affected the
compressive strength and the porosity of the samples were not identical, especially at 28
days the general trends remained constant.
Based on the results, as the limestone content of cement increased from 0 to 11%,
the compressive strength of the samples increased while the porosity decreased. Further
addition of limestone resulted in decreases in the compressive strength and increases in
the porosity of the samples. The samples were divided into two sets; those with limestone
contents less than 11% and those with more than 11% limestone contents. Next, lines of
best fit were obtained for the results of each two sets. Based on these regressions, it is
concluded that the optimum level of limestone in terms of the maximum compressive
strength and minimum porosity is around 8%. This finding is similar to the case where
15% of the cement was replaced with slag. This suggests that when metakaolin is added
to the mix, more alumina would be available, which implies that more limestone could
participate in the hydration reactions, and the optimum limestone content would shift to a
higher value. In this experiment, the optimum limestone level increased from 2.4% in the
case where no metakaolin was added to about 8% when 10% metakaolin was added to
the mix. The other conclusion that can be drawn is that 10% metakaolin is similar to 15%
slag in terms of the amount of alumina it provides for the hydration reactions with
limestone content of the cement.
153
R2 = 0.97
R2 = 0.86
R2 = 0.99
R2 = 0.96
-25
-20
-15
-10
-5
0
5
10
15
20
25
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive C
han
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%) 28 Day
Strength
28 DayPorosity(MIP)
Decrease
Increase
Figure 4.98: Relative strength and porosity of mortar cubes made with 90% cement and 10% MK at 28 days
R2 = 0.96
R2 = 0.78
R2 = 0.99
R2 = 0.94
-15
-10
-5
0
5
10
15
0 5 10 15 20 25 30
Limestone Content (%)
Rela
tive
Ch
an
ge o
f C
om
pre
ssiv
e S
tren
gth
an
d
Po
rosit
y (
%)
56 Day
Strength
56 DayPorosity(MIP)
Decrease
Increase
Figure 4.99: Relative strength and porosity of mortar cubes made with 90% cement and 10% MK at 56 days
154
4.3.4. Hydration
4.3.4.1. 100% Cement
The results of XRD analysis of 100% cement pastes made with Type I and
GUL13 cements are shown in Figure 4.100 and 4.101. The products of hydration for both
cements were fairly similar; portlandite, ettringite, and calcite were observed in both
cases. The main difference was that traces of monocarboaluminate were detected in the
GUL13 samples, especially at the ages of 28 and 56 days. This verifies the theory
explained in the previous sections that the limestone content of the cement reacts with the
alumina phases to form carboaluminate hydrates. The other difference was that the
concentration of calcite was higher in the GUL13 samples compared to the Type I ones,
due to the higher limestone content of this cement. As the age of the samples increased,
the concentration of portlandite increased in all samples since more hydration occured
with time.
5 10 15 20 25 30
E E
P
EP C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite
1 Day
7 Days
14 Days
28 Days
56 Days
Type I - 100%
Figure 4.100: XRD analysis of 100% Type I pastes at various ages
155
5 10 15 20 25 30
E E
P
EP C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; MC=Monocarboaluminate
1 Day
7 Days
14 Days
28 Days
56 Days
GUL13 - 100%
MC
Figure 4.101: XRD analysis of 100% GUL13 pastes at various ages
4.3.4.2. 85% Cement and 15% Slag
Figures 4.102 and 4.103 show the results of XRD analysis of 85% cement and
15% slag pastes made with Type I and GUL13 cements. Similar to the 100% cement
samples, the products of hydration for both cements were quite similar; portlandite,
ettringite, and calcite were observed in both cements. However, traces of
monocarboaluminate were detected in the GUL13 samples, especially at the ages of 28
and 56 days, whereas the samples made with Type I contained monosulfate instead. This
matches what is found in the literature (e.g. De Weerdt et al, 2011) that in the absence of
limestone, ettringite decomposes to monosulfate whereas in the presence of limestone,
the decomposition of ettringite to monosulfate is prevented and instead calcium mono- or
hemicarboaluminate are formed. Also, it verifies that the carboaluminate hydrates are
mostly formed at later ages.
156
The results show that the concentration of calcite was higher in the GUL13
samples compared to the Type I ones, which is due to the higher limestone content of this
cement. The concentrations of portlandite in these samples were lower than those of the
100% cement samples. This is attributed to the presence of slag in these mixes. Slag
reacts with the products of hydration of Portland cement and produces secondary C-S-H,
which results in a decrease in the amount of portlandite formed.
5 10 15 20 25 30
E EP
EP C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ; MS = Monosulfate
1 Day
7 Days
14 Days
28 Days
56 Days
Type I - S15
MS
Figure 4.102: XRD analysis of 85%Type I and 15% slag pastes at various ages
157
5 10 15 20 25 30
E E P EP C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; MC=Monocarboaluminate
1 Day
7 Days
14 Days
28 Days
56 Days
GUL13 - S15
MC
Figure 4.103: XRD analysis of 85%GUL13 and 15% slag pastes at various ages
4.3.4.3. 70% Cement and 30% Slag
The results of XRD analysis of 70% cement and 30% slag pastes made with Type
I and GUL13 cements are shown in Figure 4.104 and 4.105. Based on the results,
portlandite, ettringite, and calcite were the main phases observed in both cements.
However, The Type I samples contained monosulfate whereas the GUL13 samples
contained hemicarboaluminate and monocarboaluminate, especially at the ages of 28 and
56 days. So once again, it is concluded that ettringite decomposes to monosulfate in the
absence of limestone whereas in its presence, calcium mono- or hemicarboaluminate are
formed instead.
Due to the higher limestone content of GUL13, the concentration of calcite was
higher in these samples compared to the Type I ones. Similar to the 15% slag mixes, the
concentrations of portlandite in the 30% slag samples were lower than those of the 100%
cement samples.
158
5 10 15 20 25 30
E E P E P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ; MS = Monosulfate
1 Day
7 Days
14 Days
28 Days
56 Days
Type I - S30
MS
Figure 4.104: XRD analysis of 70%Type I and 30% slag pastes at various ages
5 10 15 20 25 30
E E P C P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ;
MC=Monocarboaluminate ; HC = Hemicarboaluminate
1 Day
7 Days
14 Days
28 Days
56 Days
GUL13 - S30MCHC
Figure 4.105: XRD analysis of 70%GUL13 and 30% slag pastes at various ages
159
4.3.4.4. 50% Cement and 50% Slag
Figures 4.106 and 4.107 show the results of XRD analysis of 70% cement and
30% slag pastes made with Type I and GUL13 cements. The results were very similar to
those of the 30% slag mixes; portlandite, ettringite, and calcite were the main phases
observed in both cements. However, the GUL13 samples contained monocarboaluminate
and hemicarboaluminate, especially at the ages of 28 and 56 days, whereas monosulfate
was detected in the GUL13 samples. The concentration of calcite was higher in the
GUL13 samples compared to the Type I ones, which is due to the higher limestone
content of this cement. The concentration of portlandite in the 50% slag samples were
lower than all the previous samples, which shows that the secondary hydration reactions
of slag consumes the portlandite and decreases its available amount.
5 10 15 20 25 30
E P P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; MS = Monosulfate
1 Day
7 Days
14 Days
28 Days
56 Days
Type I - S50
CMS E
Figure 4.106: XRD analysis of 50%Type I and 50% slag pastes at various ages
160
5 10 15 20 25 30
E E P C P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; HC = Hemicarboaluminate; MC = Monocarboaluminate
1 Day
7 Days
14 Days
28 Days
56 Days
GUL13 - S50
HC MC
Figure 4.107: XRD analysis of 50%GUL13 and 50% slag pastes at various ages
4.3.4.5. 90% Cement and 10% Metakaolin
The results of XRD analysis of 90% cement and 10% metakaolin pastes made
with Type I and GUL13 cements are shown in Figure 4.108 and 4.109. Similar to the
50% slag mixes, portlandite, ettringite, and calcite were the main phases observed in both
cements. However, the GUL13 samples contained monocarboaluminate and
hemicarboaluminate, especially at the ages of 28 and 56 days, whereas monosulfate was
detected in the Type I samples. This suggests that limestone prevents the decomposition
of ettringite to monosulfate, rather reacts with the available alumina to form
carboaluminate hydrates. A noticeable difference between these mixes and those made
with 100% cement or slag was that quartz was detected in the samples made with
metakaolin. This is attributed to the high amount of SiO2 (61%) present in the
metakaolin.
161
According to the results, the concentration of calcite was higher in the GUL13
samples compared to the Type I ones, which is due to the higher limestone content of this
cement. Once again, the concentrations of portlandite in the 10% metakaolin samples
were lower than those of the 100% cement samples, which points out to the secondary
hydration of metakaolin.
5 10 15 20 25 30
E E P E P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; Q = Quartz ; MS = Monosulfate
1 Day
7 Days
14 Days
28 Days
56 Days
Type I - MK10
QMS
Figure 4 4.108: XRD analysis of 90%Type I and 10% metakaolin pastes at various ages
162
5 10 15 20 25 30
E E P C P C
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite; Q = Quartz;
MC=Monocarboaluminate; HC=Hemicarboaluminate
1 Day
7 Days
14 Days
28 Days
56 Days
GUL13 - MK10
MCHC Q
Figure 4.109: XRD analysis of 90%GUL13 and 10% metakaolin pastes at various ages
4.3.4.6. Comparison
Figures 4.110 to 4.114 compare the XRD results obtained for the Type I and
GUL13 cements with various amounts of SCMs at 28 days. As evident in the results, the
GUL13 pastes contained carboaluminate hydrates, namely monocarboaluminate and
hemicarboaluminates whereas the Type I pastes contained monosulfate. While the 100%
cement and 15% slag mixes made with GUL13 cement contained only
monocarboaluminate, the 30% slag, 50% slag, and 10% metakaolin mixes contained both
monocarboaluminate and hemicarboaluminates. Moreover, the intensity of the ettringite
peaks in the GUL13 samples were higher than that of the Type I samples, which
indicated higher concentration of ettringite in the GUL13 samples.
Based on this, it is concluded that in the absence of limestone, ettringite
decomposes to monosulfate whereas in the presence of limestone, the decomposition of
ettringite to monosulfate is prevented and instead the limestone reacts with alumina
163
phases to form calcium mono- or hemicarboaluminate hydrates. This is in accordance
with what is found in the literature (e.g. Damidot et al, 2011; De Weerdt et al, 2011). In
addition to that, it reaffirms that if the available alumina is increased, more limestone will
react and hence, more carboaluminate hydrates will form. Since SCMs in general are
good sources of alumina, adding more slag and metakaolin would result in more available
alumina and thus more carboaluminate hydrates. This is shown in the results in the sense
the mixes with 30% slag and more or those with 10% metakaolin contained both
monocarboaluminate and hemicarboaluminates (the slag used contained 7.2% Al2O3
whereas the metakaolin contained 34% Al2O3). As such, the shift in the optimum level of
limestone used for maximum compressive strength and minimum porosity (as discussed
in the previous sections) is attributed to the availability of more alumina to react with
limestone and form carboalumiate hydrates.
5 10 15 20 25 30
P
P
E
EE
E
C
C
MCP
P
GUL13 - 100% - 28 Days
Type I - 100% - 28 Days
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ; MC = Monocarboaluminate
E
E
E C
E C
E
E
Figure 4.110 : XRD analysis of 100%Type I and GUL13 pastes at 28 days
164
5 10 15 20 25 30
P
P
E
EE
E
C
C
MC P
P
GUL13 - S15 - 28 Days
Type I - S15 - 28 Days
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ; MC = Monocarboaluminate; MS = Monosulfate
E E C
E C
E
E MS
E
Figure 4.111: XRD analysis of 85%Type I and GUL13 with 15% slag pastes at 28 days
5 10 15 20 25 30
P
P
E
E E
E
C
C
MCP
P
GUL13 - S30 - 28 Days
Type I - S30 - 28 Days
2θ, deg.
P = Portlandite ; C = Calcite ; E = Ettringite ;
MC = Monocarboaluminate ; HC = Hemicarboaluminate ; MS = Monmosulfate
E
E
E C
E C
E
HC
MS
Figure 4.112 : XRD analysis of 70%Type I and GUL13 pastes with 30% slag at 28 days
165
5 10 15 20 25 30
P
P
E
E
E
E
C
C
MC P
P
GUL13 - S50 - 28 Days
Type I - S50 - 28 Days
2θ, deg. P = Portlandite ; C = Calcite ; E = Ettringite ;
HC = Hemicarboaluminate; MC = Monocarboaluminate; MS = Monosulfate
E E C
E CE
HC
MS
Figure 4.113: XRD analysis of 50%Type I and GUL13 pastes with 50% slag at 28 days
5 10 15 20 25 30
P
P
E
EE
E
C
C
MCP
P
GUL13 - MK10 - 28 Days
Type I - MK10 - 28 Days
2θ, deg. P = Portlandite ; C = Calcite ; E = Ettringite ; Q = Quartz
HC = Hemicarboaluminate; MC = Monocarboaluminate; MS = Monosulfate
E E C
E CE
HC
MS
Q
Q
Q
Q
Figure 4.114 : XRD analysis of 90%Type I and GUL13 pastes with 10% metakaolin at 28 days
166
5. Conclusions and Recommendations
This chapter summarizes the conclusions made from the results of this study, and
provides recommendations for modification of current test methods as well as areas
where further research could be conducted. They are categorized based on the various test
methods used in this study. The conclusions cover the sulfate resistance of Portland and
Portland-limestone cements both at normal and low temperatures, the evolution of
thaumasite form of sulfate attack, and a study on the compressive strength, porosity, and
hydration of Portland and Portland-limestone cements and their combination with
supplementary cementitious materials.
5.1. Conclusions
1. At 23 °C, when no supplementary cementitious material was used, none of the
mortar bars made with Portland cements or Portland-limestone cements exposed
to 5% sodium sulfate were sulfate-resistant. This was not unexpected due to the
high tricalcium aluminate content of the cement clinker used to make the cements.
2. Replacing the Portland cements and Portland-limestone cements with 30% or
50% slag was effective in making the mixes highly sulfate-resistant when exposed
to sulfate solutions at 23 °C.
3. For the 23 °C sulfate resistance tests, the optimum limestone content of the
cement for minimum expansion with 30% slag was 11%, and for 50% slag mixes,
the Portland-limestone cements with limestone contents of 11% or greater had the
lowest expansion.
4. Overall, at 23 °C, the sulfate resistance of Portland-limestone cement combined
with slag was comparable to that of Portland cements combined with slag.
5. A new test method was proposed and developed to evaluate the resistance of
Portland-limestone cements against the low-temperature thaumasite form of
167
sulfate attack. It is essentially a modified version of the ASTM C 1012 (CSA
A3004-C8, Procedure A) in which the samples are stored in 5°C sulfate solution
rather than the standard 23°C. This test method, developed as part of this
research, was adopted in 2010 as a CSA standard and is listed as CSA A3004-C8,
Procedure B. This method was used to evaluate the resistance of the Portland
cements and Portland-limestone cements used in this study against thaumasite
sulfate attack.
6. When tested at 5 °C, and when no supplementary cementitious material was used,
all the mortar bars failed the sulfate resistance test and completely disintegrated
due to the formation of thaumasite as determined by XRD.
7. Replacing the Portland cements with 30% slag was not effective in controlling the
deterioration at 5 °C, both for Portland cements and Portland-limestone cements.
Visual examination and XRD analysis showed that the PLC-30% slag mortar bars
were damaged due to thaumasite sulfate attack.
8. All the combinations of Portland cements and Portland-limestone cements with
50% slag were found to be highly sulfate-resistant (CSA Type HSb) at 5 °C. This
is in line with the 2010 revision of the CSA A3001 standard which mandates a
minimum of 40% slag to be used in sulfate exposures where thaumasite sulfate
attack is a concern. This CSA limit was determined largely based on the results of
this study.
9. In 5 °C tests, the rate of expansion of the mortars was directly related to the
limestone content of the cement for the 100% cement and 30% slag mixes.
Nevertheless, for the 50% slag mixes, as the limestone content of the cement
increased, the sulfate resistance still increased (i.e. lower expansion) up to about
13% limestone. Even though the PLC with a limestone content of 22% when
combined with 50% slag passed the CSA expansion limits, it resulted in higher
sulfate expansion than when PLC of lower limestone content was used. This
168
suggests that going above the 15% limit on the limestone content of PLC in CSA
A3001 may not be advisable.
10. Portland cements as well as Portland-limestone cements are susceptible to the
thaumasite form of sulfate attack at low temperatures (5°C). In such
environments, the mortar bars show signs of deterioration during preliminary
stages in the form of cracking and spalling, accompanied with expansion, which is
similar to conventional sulfate attack. Eventually, after experiencing a significant
amount of expansion, the samples lose their cohesiveness and turn into a white,
pulpy mush where the C-S-H matrix has completely decomposed into thaumasite.
11. When no SCMs are used, increasing the limestone content of CSA Type GU
cement accelerates the onset of damage due to thaumasite sulfate attack.
12. Based on the above mentioned findings, it is concluded that Portland cements and
Portland-limestone cements can become resistant to sulfate attack, both the
conventional sulfate attack and the thaumasite sulfate attack, provided that they
are combined with sufficient amount of supplementary cementitious materials. As
such, the restriction on the use of Portland-limestone cements in sulfate
environments can be lifted through appropriate provisions.
13. As evidenced by XRD analysis, the limestone portion of Portland-limestone
cements reacts with the alumina phases of cement and produces
monocarboaluminates, which would contribute to the strength by means of
producing more hydrates and filling the space, and also reduces the porosity. As
such, for each clinker, there is an optimum level of limestone in terms of the
maximum strength and minimum porosity, which corresponds to the point where
all the available alumina is used up. While not investigated here, limestone
particles may also act as nucleation sites and increase the rate of reaction of
cement.
169
14. As the limestone content of the cement increased, the shift in the optimum level
of SCM in terms of the compressive strength and porosity (from about 3% to 8%
in this study) is attributed to the availability of more alumina from the SCM, that
allowed more limestone to participate in the hydration reactions, forming
additional carboaluminate hydrates.
15. In general, the findings of this study suggest that Portland-limestone cements not
only have comparable performance to that of ordinary Portland cements, but also
their combination with appropriate amounts of SCMs provides better properties,
including sulfate resistance, compressive strength, and porosity compared to OPC.
This suggests a bright future for the use of Portland-limestone cements in the
industry, as their beneficial effects in reducing the greenhouse gas emissions is
well established.
5.2. Recommendations
1. According to the thaumasite sulfate resistance test adopted by the CSA A3001,
while the initial expansion limit is 0.10% at 18 months, if the increase in
expansion between 12 and 18 months exceeds 0.03%, then the sulfate expansion
at 24 months shall not exceed 0.10%. Based on the findings of this study, it is
recommended that this 0.030% change in expansion currently required in the
standard be changed to 0.020%. Alternatively, the absolute expansion limit could
be changed to be less than 0.10% at 24 months to be considered thaumasite
sulfate-resistant.
2. The new method for evaluating the resistance of Portland-limestone cements
against thaumasite sulfate attack involves measuring the expansion of mortar bars
in the laboratory and is based on laboratory test results; however, in the field,
concrete structures are exposed to conditions where thaumasite sulfate attack is a
concern. Therefore, it is essential that the performance of concretes in these
170
conditions be studied, to confirm the laboratory test limits and to develop
additional appropriate test methods, if required.
171
6. References
ACI Committee 201 (2008). “Guide to Durable Concrete”, ACI 201.2R-08, American
Concrete Institute, Detroit, Michigan, USA.
Barker, A. and Cory, H. (1991). “The Early Hydration of Limestone-Filled Cements”, in
N. Swamy (Ed.), Blended Cements in Construction, Elsevier, London, pp. 107-124.
Barker, A. P. and Hobbs, D. W. (1999). “Performance of Portland Limestone Cements in
Mortar Prisms Immersed in Sulfate Solutions at 5°C”, Cement and Concrete Composites,
vol. 21, pp. 129-137.
Barnett, S. J., Halliwell, M. A., Crammond, N. J., Adam, C. D., Jackson, A. R. W.
(2002). “Study of Thaumasite and Ettringite Phases Formed in Sulfate/Blast Furnace Slag
Slurries using XRD Full Pattern Fitting”, Cement and Concrete Composites, vol. 24, pp.
339-346.
Bellmann, F., Möser, B., and Stark, J. (2006). “Influence of Sulfate Solution
Concentration on the Formation of Gypsum in Sulfate Resistance Test Specimen”,
Cement and Concrete Research, vol. 36, pp. 358–363.
Bensted, J. (1999). “Thaumasite—Background and Nature in Deterioration of Cements,
Mortars, and Concretes”, Cement and Concrete Composites, vol. 21, pp. 117-121.
Bickley, J. A., Hemmings, R. T., Hooton, R. D., and Balinsky, J. (1994). “Thaumasite
Related Deterioration of Concrete Structures”, In American Concrete Institution
Proceedings, Concrete Technology, Past, Present and Future, Special Publication SP 144,
pp. 159–175.
Biczok, I. (1967). Concrete Corrosion Concrete Protection, Chemical Publishing, New
York.
172
Bonavetti, V. L., Rahhal, V. F., Irassar, E. F. (2001). “Studies on the Carboaluminate
Formation in Limestone Filler-Blended Cements, Cement and Concrete Research, vol.
31, pp. 853-859.
Borsoi, A., Collepardi, S. Coppola, L., Troli, R and Collepardi, M. (2000). “Sulfate
Attack on Blended Portland Cements,” ACI SP-192, 1, pp. 417-432.
Brown, P. W. and Doerr, A. (2000). “Chemical Changes in Concrete due to the Ingress of
Aggressive Species”, Cement and Concrete Research, vol. 30, pp. 411–418.
Brown, P. W., Hooton, R. D., Clark, B. A. (2003). “The Co-Existence of Thaumasite and
Ettringite in Concrete Exposed to Magnesium Sulfate at Room Temperature and the
Influence of Blast-Furnace Slag Substitution on Sulfate Resistance”, Cement and
Concrete Composites, vol. 25, pp. 939–945.
CEN (2003). “National Specifications and Recommendations for Sulfate resisting
cement”, European Committee for Standardization, CEN TC51/WG 6 Draft Report, Dec.,
13pp.
Chatterji, S. and Jeffery, J. W. (1963). “A New Hypothesis of Sulfate Expansion,”
Magazine of Concrete Research, vol. 15, issue 44, pp. 83-86.
Collepardi, M. (1999). ”Thaumasite Formation and Deterioration in Historic Buildings”,
Cement and Concrete Composites, vol. 21, no. 2, pp. 147-154.
Collepardi, M., Monosi, S., Favoni, O., and Ogoumah-Olagot, J. J. (2003). “Resistance to
Sulfate Attack of Mixtures with Limestone-Portland blended cements,” Proceedings, 6th
International Conference on Durability of Concrete, ACI SP-212, pp. 941-949.
173
Crammond, N. J. (2002). “The Occurrence of Thaumasite in Modern Construction – A
Review”, Cement and Concrete Composites, vol. 24, pp. 393-402.
Crammond, N. J. (2003). “The Thaumasite Form of Sulfate Attack in the UK”, Cement
and Concrete Composites, vol. 25, pp. 809-819.
Damidot, D., Lothenbach, B., Herfort D., Glasser, F. P. (2011). “Thermodynamics and
Cement Sciences”, Cement and Concrete Research, vol. 41, pp. 679-695.
Day, R. L., and Joshi, R. C. (1986). “Sulfate Resistance of Mortars Containing Large
Proportions of Fly Ash”, International Seminar on Some Aspects of Admixtures and
Industrial Byproducts on Durability of Concrete, Sweden.
Day, R. L., personal communication, September 21, 2011
De Weerdt, K., Ben Haha, M., Le Saout, G., Kjellsen, K. O., Justness, H., Lothenbach, B.
(2011). “Hydration mechanisms of ternary Portland cements containing limestone
powder and fly ash”, Cement and Concrete Research, vol. 41, issue 3, pp. 279–291.
DePuy, G. W. (1994). Chemical Resistance of Concrete, in Significance of Tests and
Properties of Concrete and Concrete-Making Materials, edited by Paul Klieger and
Joseph F. Lamond , ASTM STP 169C, ASTM International, Philadelphia.
Dhir, R.K., Limbachiya, M. C., McCarthy, M. J., Chaipanich, A. (2007). “Evaluation of
Portland Limestone Cements for Use in Concrete Construction”, Materials and
Structures, vol. 40, pp. 459-473.
Diamond, S. (2003). “Thaumasite in Orange County, Southern California: An Inquiry
into the Effect of Low Temperature”, Cement and Concrete Composites, vol. 25, no. 8,
pp. 1161-1164.
174
Ellerbrock, H. G., Spung, S., Kuhlmann, K. (1990). “Particle Size Distribution and
Properties of Cement: Part III. Influence of Grinding Process”, Zem.-Kalk-Gips, vol. 43,
no. 1, pp. 13-19.
EN 197-1 (2000). “Cement Part 1: Composition, Specifications and Conformity Criteria
for Common Cements”.
Erlin, B. and Stark, D. C. (1965). “Identification and Occurrence of Thaumasite in
Concrete”, Highway Research Record, pp. 108-113.
Gao, X., Ma, B., Yang, Y., and Su, A. (2007). “Sulfate Attack of Cement-Based Material
with Limestone Filler Exposed to Different Environments”, Journal of Materials
Engineering and Performance, vol. 17, no 4, pp. 543-549.
Gollop, R. S. and Taylor, H. F. W. (1992). “Microstructural and Microanalytical Studies
of Sulfate Attack: I. Ordinary Portland Cement Paste”, Cement and Concrete Research,
vol. 22, pp. 1027-1038.
Gollop, R. S. and Taylor, H. F. W. (1995). “Microstructural and Microanalytical Studies
of Sulfate Attack: III: Sulfate-Resisting Portland Cement: Reactions with Sodium and
Magnesium Sulfate Solutions”, Cement and Concrete Research, vol. 25, no. 7, pp. 1581-
1590.
González M. A. and Irassar E. F. (1998). “Effect of Limestone Filler on the Sulfate
Resistance of Low C3A Portland Cement”, Cement and Concrete Research, vol. 28, No.
11, pp. 1655–1667.
Guemmadi, Z., Resheidat, M., Houari, H., Toumi, B. (2008). “Optimal Criteria of
Algerian Blended Cement using Limestone Fines”, Journal of Civil Engineering and
Management, vol. 14, pp. 269-275.
175
Hansen, W. C., (1966). “Attack on Portland Cement Concrete by Alkali Soils and Water-
a Critical Review”, Highway Research Record, vol. 113, pp. 1-32.
Hartshorn, S. and Sims, I. (1998). “Thaumasite–A Brief Guide for Engineers”, Concrete,
vol. 32, pp. 24-27.
Hartshorn, S. A., Sharp, J. H., and Swamy, R. N. (1999). “Thaumasite formation in
Portland limestone cement pastes”, Cement and Concrete Research, vol. 29, pp. 1331-
1340.
Hartshorn, S. A., Sharp, J. H., Swamy, R. N. (2001). “Engineering Properties and
Structural Implications of Portland Limestone Cement Mortar Exposed to Magnesium
Sulfate Attack”, Advances in Cement Research, vol. 13, no 1, pp. 31-46.
Hartshorn, S. A., Sharp, J. H., and Swamy, R. N. (2002). “The Thaumasite Form of
Sulfate Attack in Portland-Limestone Cement Mortars Stored in Magnesium Sulfate
Solution.”, Cement & Concrete Composites, vol. 24, pp.351-359.
Hawkins, P., Tennis, P., Detwiler, R. (2005). The Use of Limestone in Portland Cement,
A State-of-the-Art Review, Report EB 227.02, PCA R&D Serial Number SN2052b,
Portland Cement Association, Skokie, Illinois, USA.
Heinz, D. and Urbonas, L. (2003). “About Thaumasite Formation in Portland-Limestone
Cement Pastes and Mortars––Effect of Heat Treatment at 95 °C and Storage at 5 °C”,
Cement and Concrete Composites, vol. 25, issue 8, pp. 961-967.
Hekal, E. E., Kishar, E., and Mostafa, H. (2002). “Magnesium Sulfate Attack on
Hardened Blended Cement Pastes under Different Circumstances”, Cement and Concrete
Research, vol. 32, pp. 1421–1427.
Herfort, D., Presentation on Portland-limestone cements, 2009
176
Higgins, D. D. (2003). “Increased Sulfate Resistance of GGBFS Concrete in the Presence
of carbonate”, Cement and Concrete Composites, vol. 25, no 8, pp. 913-919.
Higgins, D. D., Crammond, N. J. (2003). “Resistance of Concrete Containing GGBS to
the Thaumasite Form of Sulfate Attack”, Cement and Concrete Composites, vol. 25, pp.
921-929.
Hime, W. G. and Mather, B. (1999). “Sulfate Attack — or Is It?”, Cement and Concrete
Research, vol. 29, pp. 789-791.
Hobbs, D. W. (2003). “Thaumasite Sulfate Attack in Field and Laboratory Concretes:
Implications for Specifications”, Cement and Concrete Composites., vol. 25, pp. 1195-
1202.
Holton, I. (2003). “Final Report on 3-Year Investigation to Establish Resistance of
Portland Limestone Cement Concretes”, BRE Report 209-164 for British Cement
Association.
Hooton, R. D., (1990). “Effects of Carbonate Additions on Heat of Hydration and Sulfate
Resistance of Portland Cement,” Carbonate Additions to Cement, ASTM STP 1064, P.
Klieger and R. D. Hooton, Eds., American Society for Testing and Materials,
Philadelphia, pp. 73-81.
Hooton, R. D. and Thomas, M. D. A. (2002). “The Use of Limestone in Portland
Cements: Effect on Thaumasite Form of Sulfate Attack”, Portland Cement Association,
PCA R&D Serial No. 2658.
Hooton, R. D., Nokken, M. R., and Thomas, M. D. A. (2007). “Portland-Limestone
Cement: State-of-the-Art Report and Gap Analysis for CSA A3000”, Portland Cement
Association, Research and Development Report SN3053, 59pp.
177
Hooton, R. D., Ramezanianpour, A. M., and Schutz, U. (2010). “Decreasing the Clinker
Component in Cementing Materials: Performance of Portland-Limestone Cements in
Concrete in combination with Supplementary Cementing Materials”, CD Proceedings,
2010 Concrete Sustainability Conference, NRMCA, Tempe, Arizona, 15pp.
Ingram, K., Poslusny, M., Daugherty, K., Rowe, K. (1990). “Carboaluminate Reactions
as Influenced by Limestone Additions”, Carbonate Additions to Cement, Ed. P. Klieger
and R. D. Hooton, ASTM Special Technical Publication, STP 1064, pp. 14-23.
Irassar, E. F., Gonzalez, M., Rahhal, V. (2000). “Sulfate Resistance of Type V Cements
with Limestone Filler and Natural Pozzolana”, Cement and Concrete Composites, vol.
22, pp. 361–368.
Irassar, E.F., Bonavetti, V.L., Menhdez, G., Donza, H. and Cabrera, O., (2001).
“Mechanical Properties and Durability of Concrete Made with Portland Limestone
Cement”, Proceedings of the 3rd CANMET/ACI International Symposium on Sustainable
Development of Concrete, ACI SP-202, American Concrete Institute, Detroit, pp. 431-
450.
Irassar, E.F., Bonavetti, V.L., Trezza, M.A., and González, M.A. (2005). “Thaumasite
Formation in Limestone Filler Cements Exposed to Sodium Sulfate Solution at 20°C,”
Cement and Concrete Composites, vol. 27, pp. 77-84.
Javed, I., Bhatty J. I., Taylor, P. C. (2006). “Sulfate Resistance of Concrete Using
Blended Cements or Supplementary Cementitious Materials”, Portland Cement
Association R&D. Serial No. 2916a.
Kakali, G., Tsivilis, S., Aggeli, E. and Bati, M. (2000). “Hydration products of C3A and
C3S and Portland cement in the presence of CaCO3”, Cement and Concrete Research, vol.
30, pp. 1073-1077.
178
Kakali, G., Tsivilis, S., Skaropoulou, A., Sharp, J. H., and Swamy, R. N. (2003).
“Parameters Affecting Thaumasite Formation in Limestone Cement Mortar”, Cement and
Concrete Composites, vol. 25, No. 8, pp. 977-981.
Kelham, S. (1998). “Portland Limestone Cements”, Concrete, vol. 32, no. 5, pp. 30-31.
Kohler, S., Heinz, D., and Urbonas, L. (2006). “Effect of Ettringite on Thaumasite
Formation”, Cement and Concrete Research, vol. 36, no. 4, pp. 697–706.
Kurtis, K. E., Monteiro, P. J. M., Brown, J. T., and Meyer-Ilse, W. (1998). “Mechanisms
of Damage Caused by Sulfate Attack Examined through Transmission X-ray
Microscopy”, ALS Compendium of User Abstracts and Technical Reports LBL-41658.
Lamond, J. L., Pielert, J. H., (2006). Significance of Tests and Properties of Concrete and
Concrete-Making Materials, ASTM STP 169D, American Society for Testing of
Materials.
Lee, S. T., Hooton, R. D., Jung, H., Park, D., and Choi, C. K. (2008). “Effect of
Limestone Filler on the Deterioration of Mortars and Pastes Exposed to Sulfate Solutions
at Ambient Temperature”, Cement and Concrete Research, vol. 38, pp. 68-76.
Lipus, K. and Puntke, S. (2003). “Sulfate Resistance of Concretes with Different
Compositions”, Beton Jg., vol. 53, 2, pp. 169-180.
Lothenbach, B., LeSaout, G., Gallucci, E., Scrivener, K. L. (2008). “Influence of
Limestone on the Hydration of Portland cements”, Cement and Concrete Research, vol.
38, no. 6, pp. 848–860.
Lounis, Z., and Diagle, L. (2010). “Environmental Benefits of Life Cycle Design of
Concrete Bridges”, NRC Publication, NRCC-49675.
179
Ma, B., Gao, X., Byars, B. A., Zhou, Q. (2006). “Thaumasite Formation in a Tunnel of
Bapanxia Dam in Western China”, Cement and Concrete Research, vol. 36, no. 4, pp.
716–722.
Macphee, D. E. and Barnett, S. J. (2004). “Solution Properties of Solids in the
Ettringite—Thaumasite Solid Solution Series”, Cement and Concrete Research, vol. 34,
no. 9, pp. 1591-1598.
Malhotra, V. M. (2000). “Role of Supplementary Cementing Materials in Reducing
Greenhouse Gas Emissions”, in Concrete Technology for a Sustainable Development in
the 21st Century, by Odd E. Gjørv and Koji Sakai, E & FN SPON.
Matthews, J.D. (1994). “Performance of Limestone Filler Cement Concrete”, in Euro-
Cements – Impact of ENV 197 on Concrete Construction, (Ed. R. K. Dhir and M. R.
Jones), E&FN Spon, London, pp. 113-147.
Mather, B., (1984). ”A Discussion of the Paper “Theories of expansion in sulfoaluminate-
type expansive cements: schools of though””, by M. D. Cohen, Cement and Concrete
Research, vol. 14, pp. 603-609.
Matschei, T., Lothenbach, B. and Glasser, F. P. (2007). “The Role of Calcium Carbonate
in Cement Hydration”, Cement and Concrete Research, vol. 37, pp. 118-130.
Mehta, P. K., Pirtz, D., and Polivka, M. (1979). “Propperties of Alite Cements”, Cement
and Concrete Research, vol. 9, pp. 439-450.
Mehta, P. K. (1993). “Sulfate Attack on Concrete - A Critical Review”, Materials
Science of Concrete III, Edited by J. Skalny, The American Ceramic Society,
Westerville, OH, pp. 105-130.
180
Mehta, P. K. and Monteiro, P. J. M. (2006). Concrete: Structure, Properties and
Materials, Prentice Hall, New Jersey, 6th edition.
Mehta, P. K. (2007). “Sustainability of the Concrete Industry- critical issues”, Concrete
Summit on Sustainable Development, Washington DC, Strategig Development Council,
American Concrete Institute, Farmington Hills, IL.
Moir, G. K. (2003). Limestone Cements, Gaining Acceptance”, International Cement
Review.
Monteiro, P. J. M. and Kurtis, K. E. (2003). “Time to Failure for Concrete Exposed to
Severe Sulfate Attack”, Cement and Concrete Research, vol. 33, no. 7, pp. 987–993.
Mulenga D.M., Stark, J., Nobst, P. (2003). “Thaumasite Formation in Concrete and
Mortars Containing Fly Ash”, Cement and Concrete Composites, vol. 25, pp. 907-912.
Neville A. M. (1995). Properties of Concrete, 4th edition, New York: John Wiley &
Sons, Inc.
Nobst P. and Stark, J. (2003). “Investigations on the Influence of Cement Type on
Thaumasite Formation”, Cement and Concrete Composites, vol. 25, pp. 899-906.
Oberholster, R. E. (2002). “Deterioration of Mortar, Plaster and Concrete: South Africa
Laboratory and Field Studies”, First International Conference on Thaumasite in
Cementitious Materials, paper 42 (Garston (UK)).
Odler, I. and Gasser, M. (1988). “Mechanism of Sulfate Expansion in Hydrated Portland
Cement”, Journal of American Ceramic Society, vol. 71, no. 11, pp. 1015-1020.
Pera, J., Husson, S., and Guilhot, B. (1999). “Influence of finely ground limestone on
cement hydration”, Cement and Concrete Composites, vol. 21, pp.99-105.
181
Ramachandran, V. S. (1988). “Thermal analysis of cement components hydrated in the
presence of Calcium Carbonate,” Thermochimica Acta, vol. 127, pp. 385-394.
Romer, M., Holzer, L., and Pfiffner, M. (2003). “Swiss Tunnel Structures: Concrete
Damage by Formation of Thaumasite”, Cement and Concrete Composites, vol. 25 no. 8,
pp. 1111-1117.
Sahu, S., Badger, S., Thaulow, N. (2002). “Evidence of Thaumasite Formation in
Southern California Concrete”, Cement and Concrete Composites, vol. 24, pp. 379-384.
Santhanam, M., Cohen, M. D., and Olek, J. (2003). “Effects of Gypsum Formation on the
Performance of Cement Mortars during External Sulfate Attack”, Cement and Concrete
Research, vol. 33, no. 3, pp. 325–332.
Sawicz Z. and Heng S. S. (1996). “Durability of Concrete with Addition of Limestone
Powder”, Magazine of Concrete Research, vol. 48, no. 175, pp. 131-137.
Schmidt, M. (1992), “Cement with Interground Additives-Capabilities and
Environmental Relief, Part 1,” Zement-Kalk-Gips, vol. 45, no. 2, pp. 64- 69 (English part
in 45, 4, 87-92.).
Schmidt, T. (2007). Sulfate Attack and the Role of Internal Carbonate on the Formation
of Thaumasite, PhD Dissertation, École Polytechnique Fédérale de Lausanne.
Schmidt, T., Lothenbach, B., Romer, M., Neuenschwander, J., and K. Scrivener (2009).
“Physical and Microstructural Aspects of Sulfate Attack on Ordinary and Limestone
Blended Portland Cements”, Cement and Concrete Research, vol. 39, pp. 1111-1121.
182
Sharp, J. H., Gartner, E. M., and Macphee, D. E. (2010). “Novel Cement Systems
(sustainability). Session 2 of the Fred Glasser Cement Science Symposium”, Advances in
Cement Research, vol. 22, issue 4, pp. 195-202.
Sims, I. and Huntley, S.A. (2004). “The Thaumasite Form of Sulfate Attack – Breaking
the Rules”, Cement and Concrete Composites, vol. 26, pp. 837-844.
Skalny, J., Marchard, J., Odler, I. (2002). Sulfate Attack on Concrete, Spon Press, London
and New York.
Skaropoulou, A., Tsivilis, S., Kakali, G., Sharp, J. H., and Swamy, R. N. (2009).
“Thaumasite Form of Sulfate Attack in Limestone Cement Mortars: A Study on Long
Term Efficiency of Mineral Admixtures, Construction and Building Materials, vol. 23,
issue 6, pp. 2338-2345.
Smallwood, I., Wild, S., and Morgan, E. (2003). “The Resistance of Metakaolin (MK)-
Portland Cement (PC) Concrete to the Thaumasite-type of Sulfate Attack (TSA)—
Program of Research and Preliminary Results”, Cement and Concrete Composites, vol.
25, pp. 931-938.
Soroka, I. and Stern, N. (1976). “Effect of Calcareous Fillers on Sulfate Resistance of
Portland Cement,” American Ceramic Society Bulletin, vol. 55, no. 6, pp. 594-595.
Soroka, I. and Setter, N. (1977). “The Effect of Fillers on Strength of Cement Mortars”
Cement and Concrete Research, vol. 7, pp. 449-456.
Soroka I. and Setter N. (1980). “Effect of Mineral Fillers on Sulfate Resistance of
Portland Cement Mortars”, Durability of Building Materials and Components, ASTM
STP 691, P.J. Sereda and G.G. Litvan, Eds., American Society for Testing and Materials,
pp. 326-335.
183
Sprung, S., and Siebel, E. (1991). “Assessment of the Sustainability of Limestone for
Producing Portland Limestone Cement (PKZ)”, Zement Kalk Gips, no. 1, pp. 1-14.
Suput, J. S., Mladenovic, A., Cernilogar, l., and Olensk, V. (2003). “Deterioration of
Mortar Caused by the Formation of Thaumasite on the Limestone Cladding of Some
Slovenian Railway Tunnels”, Cement and Concrete Composites, vol. 25, pp. 1141–1146.
Swenson, E. G. (1971). Concrete in Sulphate Environments, NRC Publication 5329381,
National Research Council of Canada.
Taylor, P. C. (2001a). Performance of Low C3A-Content Cements Containing
Interground Limestone, PCA R&D Serial No. 2182, Portland Cement Association,
Skokie, Illinois.
Taylor, P. C. (2001b). Sulfate Resistance Tests on Type V Cements Containing
Limestone, PCA R&D Serial No. 2182b, Portland Cement Association, Skokie, Illinois.
Tezuka, Y., Gomes, D., Martins, J., Djanikian, J. G. (1992). “Durability Aspect of
Cement with High Limestone Filler Content”, Proceedings of 9th International Congress
on the Chemistry of Cement., New Delhi, India, V, pp. 53–59.
Thaumasite Expert Group (1999). “The Thaumasite Form of Sulfate Attack: Risks,
Diagnosis, Remedial Works and Guidance on New Construction”, Report of the
Thaumasite Expert Working Group, DETR, London, 180p.
Thomas, M. D. A., Rogers, C. A., and Bleszynski, R. F. (2003). “Occurrences of
Thaumasite in Laboratory and Field Concrete”, Cement and Concrete Composites, vol.
25, no. 8, pp. 1045–1050.
Thomas, M. D. A. and Skalny, J. (2006). Chemical Resistance of Concrete, in
Significance of Tests and Properties of Concrete and Concrete-Making Materials, edited
184
by Joseph Lamond and James Pielert, ASTM STP 169D, ASTM International, New
York.
Tian, B., and Cohen, M. D. (2000). ”Does Gypsum Formation during Sulfate Attack on
Concrete Lead to Expansion? “, Cement and Concrete Research, vol. 30, no. 1, pp. 117-
123.
Torres, S. M., Sharp, J. H., Swamy, R. N., Lynsdale, C. J. and Huntley, S. A. (2003).
“Long- Term Durability of Portland Limestone Cement Exposed to Magnesium Sulfate
Attack.”, Cement and Concrete Composites, vol. 25, pp. 947-954.
Torres, S. M., Lynsdale, C. J., Swamy, R. N., and Sharp, J. H. (2006). “Microstructure of
5 Year Old Mortars Containing Limestone Filler Damaged by Thaumasite,” Cement and
Concrete Research, vol. 36, pp. 384-394.
Trägårdh, J. and Kalinowski, M. (2003). “Investigation of the Conditions for a
Thaumasite Form of Sulfate Attack in SCC with Limestone Filler,” Proceedings,
International RILEM Symposium on Self-Compacting Concrete, PRO 33, 844-854.
Tsivilis, S., Kakali, G., Chaniotakis, E., Souvaridou, A. (1998). “Study on the Hydration
of Portland Limestone Cement by means of TG”, Journal of Thermal Analysis and
Calorimetry, vol. 52, no. 3, pp. 863-870.
Tsivilis, S., Chaniotakis, E., Kakali, G., Batis, G. (2002). “An analysis of the properties
of Portland limestone cements and concrete”, Cement and Concrete Composites, vol. 24,
pp. 371-378.
Tsivilis, S., Tsantila, G., Kakali, G., Chaniotakis, E, Sakellariou, A. (2003). “The
permeability of Portland limestone cement concrete”, Cement and Concrete Research,
vol. 33, pp. 1465-1471.
185
Turker, P. and. Erdoğdu, K. (1999). “Effects of Limestone Addition on Microstructure
and Hydration of cements”, Proceedings of the International Conference on Cement
Microscopy Ed - Jany, L. and Nisperos, A. 22nd, pp. 265-283.
Voglis, N., Kakali, G., Chaniotakis, E., and Tsivilis, S. (2005). “Portland-limestone
cements. Their properties and hydration compared to those of other composite cements”,
Cement and Concrete Research, vol. 27, pp. 191-196.
